Testing RAM instead of mmap for blob alignment

exceptions in Python>=3.7

python/obitools3/apps/optiongroups/__init__.py

@@ -1,3 +1,9 @@
+import codecs
+
+def unescaped_str(arg_str):
+    return arg_str.encode('latin-1', 'backslashreplace').decode('unicode-escape')
+
+
 def __addInputOption(optionManager):
     
     optionManager.add_argument(
@@ -39,6 +45,30 @@ def __addImportInputOption(optionManager):
                      const=b'fastq',
                      help="Input file is in fastq format")
 
+    group.add_argument('--silva-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'silva',
+                     help="Input file is in SILVA fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
+    group.add_argument('--rdp-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'rdp',
+                     help="Input file is in RDP training set fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
+    group.add_argument('--unite-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'unite',
+                     help="Input file is in UNITE fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
+    group.add_argument('--sintax-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'sintax',
+                     help="Input file is in SINTAX fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
     group.add_argument('--embl-input',
                      action="store_const", dest="obi:inputformat",
                      default=None,
@@ -119,15 +149,15 @@ def __addImportInputOption(optionManager):
 def __addTabularOption(optionManager):
     group = optionManager.add_argument_group("Input and output format options for tabular files")
 
-    group.add_argument('--header',
-                     action="store_true", dest="obi:header",
-                     default=False,
-                     help="First line of tabular file contains column names")
+    group.add_argument('--no-header',
+                     action="store_false", dest="obi:header",
+                     default=True,
+                     help="Don't print the header (first line with column names")
 
     group.add_argument('--sep',
                      action="store", dest="obi:sep",
                      default="\t",
-                     type=str,
+                     type=unescaped_str,
                      help="Column separator")
 
 
@@ -159,6 +189,16 @@ def __addTabularInputOption(optionManager):
                      help="Lines starting by this char are considered as comment")
 
 
+def __addTabularOutputOption(optionManager):
+    group = optionManager.add_argument_group("Output format options for tabular files")
+    
+    __addTabularOption(optionManager)
+
+    group.add_argument('--na-int-stay-na',
+                     action="store_false", dest="obi:na_int_to_0",
+                     help="NA (Non available) integer values should be exported as NA in tabular output (default: they are converted to 0 for tabular output).") # TODO
+
+
 def __addTaxdumpInputOption(optionManager):  # TODO maybe not the best way to do it
     group = optionManager.add_argument_group("Input format options for taxdump")
 
@@ -192,6 +232,10 @@ def addTabularInputOption(optionManager):
     __addTabularInputOption(optionManager)
 
 
+def addTabularOutputOption(optionManager):
+    __addTabularOutputOption(optionManager)
+
+
 def addTaxonomyOption(optionManager):
     __addTaxonomyOption(optionManager)
 
@@ -204,6 +248,7 @@ def addAllInputOption(optionManager):
     __addInputOption(optionManager)
     __addImportInputOption(optionManager)
     __addTabularInputOption(optionManager)
+    __addTabularOutputOption(optionManager)
     __addTaxonomyOption(optionManager)
     __addTaxdumpInputOption(optionManager)    
 
@@ -264,6 +309,35 @@ def __addExportOutputOption(optionManager):
                      const=b'tabular',
                      help="Output file is in tabular format")
 
+    group.add_argument('--metabaR-output',
+                     action="store_const", dest="obi:outputformat",
+                     default=None,
+                     const=b'metabaR',
+                     help="Export the files needed by the obifiles_to_metabarlist function of the metabaR package")
+
+    group.add_argument('--metabaR-prefix',
+                     action="store", dest="obi:metabarprefix",
+                     type=str,
+                     help="Prefix for the files when using --metabaR-output option")
+
+    group.add_argument('--metabaR-ngsfilter',
+                     action="store", dest="obi:metabarngsfilter",
+                     type=str,
+                     default=None,
+                     help="URI to the ngsfilter view when using --metabaR-output option (if not provided, it is not exported)")
+
+    group.add_argument('--metabaR-samples',
+                     action="store", dest="obi:metabarsamples",
+                     type=str,
+                     default=None,
+                     help="URI to the sample metadata view when using --metabaR-output option (if not provided, it is built as just a list of the sample names)")
+
+    group.add_argument('--only-keys',
+                       action="append", dest="obi:only_keys",
+                       type=str,
+                       default=[],
+                       help="Only export the given keys (columns).")
+
     group.add_argument('--print-na',
                      action="store_true", dest="obi:printna",
                      default=False,
@@ -296,14 +370,14 @@ def addTabularOutputOption(optionManager):
 
 def addExportOutputOption(optionManager):
     __addExportOutputOption(optionManager)
-    __addTabularOption(optionManager)
+    __addTabularOutputOption(optionManager)
 
 
 def addAllOutputOption(optionManager):
     __addOutputOption(optionManager)
     __addDMSOutputOption(optionManager)
     __addExportOutputOption(optionManager)
-    __addTabularOption(optionManager)
+    __addTabularOutputOption(optionManager)
 
 
 def addNoProgressBarOption(optionManager):

python/obitools3/commands/addtaxids.pyx

@@ -0,0 +1,231 @@
+#cython: language_level=3
+
+from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
+from obitools3.dms import DMS
+from obitools3.dms.view.view cimport View, Line_selection
+from obitools3.uri.decode import open_uri
+from obitools3.apps.optiongroups import addMinimalInputOption, addTaxonomyOption, addMinimalOutputOption, addNoProgressBarOption
+from obitools3.dms.view import RollbackException
+from obitools3.dms.column.column cimport Column
+from functools import reduce
+from obitools3.apps.config import logger
+from obitools3.utils cimport tobytes, str2bytes, tostr
+from io import BufferedWriter
+from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
+                                        ID_COLUMN, \
+                                        DEFINITION_COLUMN, \
+                                        QUALITY_COLUMN, \
+                                        COUNT_COLUMN, \
+                                        TAXID_COLUMN
+from obitools3.dms.capi.obitypes cimport OBI_INT
+from obitools3.dms.capi.obitaxonomy cimport MIN_LOCAL_TAXID
+import time
+import math 
+import sys
+
+from cpython.exc cimport PyErr_CheckSignals
+
+ 
+__title__="Annotate sequences with their corresponding NCBI taxid found from the taxon scientific name"
+
+
+ 
+def addOptions(parser):
+    
+    addMinimalInputOption(parser)
+    addTaxonomyOption(parser)
+    addMinimalOutputOption(parser)
+    addNoProgressBarOption(parser)
+
+    group=parser.add_argument_group('obi addtaxids specific options')
+
+    group.add_argument('-t', '--taxid-tag', 
+                       action="store", 
+                       dest="addtaxids:taxid_tag",
+                       metavar="<TAXID_TAG>",
+                       default=b"TAXID",
+                       help="Name of the tag to store the found taxid "
+                            "(default: 'TAXID').")
+
+    group.add_argument('-n', '--taxon-name-tag', 
+                       action="store", 
+                       dest="addtaxids:taxon_name_tag",
+                       metavar="<SCIENTIFIC_NAME_TAG>",
+                       default=b"SCIENTIFIC_NAME",
+                       help="Name of the tag giving the scientific name of the taxon "
+                            "(default: 'SCIENTIFIC_NAME').")
+
+    group.add_argument('-g', '--try-genus-match',
+                     action="store_true", dest="addtaxids:try_genus_match",
+                     default=False,
+                     help="Try matching the first word of <SCIENTIFIC_NAME_TAG> when can't find corresponding taxid for a taxon. "
+                          "If there is a match it is added in the 'parent_taxid' tag. (Can be used by 'obi taxonomy' to add the taxon under that taxid).")
+
+    group.add_argument('-a', '--restricting-ancestor', 
+                       action="store", 
+                       dest="addtaxids:restricting_ancestor",
+                       metavar="<RESTRICTING_ANCESTOR>",
+                       default=None,
+                       help="Enables to restrict the search of taxids under an ancestor specified by its taxid.")
+
+    group.add_argument('-l', '--log-file', 
+                       action="store", 
+                       dest="addtaxids:log_file",
+                       metavar="<LOG_FILE>",
+                       default='',
+                       help="Path to a log file to write informations about not found taxids.")
+
+
+def run(config):
+     
+    DMS.obi_atexit()
+    
+    logger("info", "obi addtaxids")
+
+    # Open the input
+    input = open_uri(config['obi']['inputURI'])
+    if input is None:
+        raise Exception("Could not read input view")
+    i_dms = input[0]
+    i_view = input[1]
+    i_view_name = input[1].name
+
+    # Open the output: only the DMS, as the output view is going to be created by cloning the input view
+    # (could eventually be done via an open_uri() argument)
+    output = open_uri(config['obi']['outputURI'],
+                      input=False,
+                      dms_only=True)
+    if output is None:
+        raise Exception("Could not create output view")
+    o_dms = output[0]
+    output_0 = output[0]
+    o_view_name = output[1]
+
+    # stdout output: create temporary view
+    if type(output_0)==BufferedWriter:
+        o_dms = i_dms
+        i=0
+        o_view_name = b"temp"
+        while o_view_name in i_dms: # Making sure view name is unique in output DMS
+            o_view_name = o_view_name+b"_"+str2bytes(str(i))
+            i+=1
+        imported_view_name = o_view_name
+
+    # If the input and output DMS are not the same, import the input view in the output DMS before cloning it to modify it
+    # (could be the other way around: clone and modify in the input DMS then import the new view in the output DMS)
+    if i_dms != o_dms:
+        imported_view_name = i_view_name
+        i=0
+        while imported_view_name in o_dms:  # Making sure view name is unique in output DMS
+            imported_view_name = i_view_name+b"_"+str2bytes(str(i))
+            i+=1
+        View.import_view(i_dms.full_path[:-7], o_dms.full_path[:-7], i_view_name, imported_view_name)
+        i_view = o_dms[imported_view_name]
+
+    # Clone output view from input view
+    o_view = i_view.clone(o_view_name)
+    if o_view is None:
+        raise Exception("Couldn't create output view")
+    i_view.close()
+
+    # Open taxonomy
+    taxo_uri = open_uri(config['obi']['taxoURI'])
+    if taxo_uri is None or taxo_uri[2] == bytes:
+        raise Exception("Couldn't open taxonomy")
+    taxo = taxo_uri[1]
+
+    # Initialize the progress bar
+    if config['obi']['noprogressbar'] == False:
+        pb = ProgressBar(len(o_view), config)
+    else:
+        pb = None
+        
+    try:
+        if config['addtaxids']['log_file']:
+            logfile = open(config['addtaxids']['log_file'], 'w')
+        else:
+            logfile = None
+        if config['addtaxids']['try_genus_match']:
+            try_genus = True
+        else:
+            try_genus = False
+        if 'restricting_ancestor' in config['addtaxids']:
+            res_anc = int(config['addtaxids']['restricting_ancestor'])
+        else:
+            res_anc = None
+        taxid_column_name = config['addtaxids']['taxid_tag']
+        parent_taxid_column_name = "PARENT_TAXID" # TODO macro
+        taxon_name_column_name = config['addtaxids']['taxon_name_tag']
+        taxid_column = Column.new_column(o_view, taxid_column_name, OBI_INT)
+        parent_taxid_column = Column.new_column(o_view, parent_taxid_column_name, OBI_INT)
+        taxon_name_column = o_view[taxon_name_column_name]
+        
+        found_count = 0
+        not_found_count = 0
+        parent_found_count = 0
+        
+        for i in range(len(o_view)):
+            PyErr_CheckSignals()
+            if pb is not None:
+                pb(i)
+            taxon_name = taxon_name_column[i]
+            taxon = taxo.get_taxon_by_name(taxon_name, res_anc)
+            if taxon is not None:
+                taxid_column[i] = taxon.taxid
+                found_count+=1
+            elif try_genus: # try finding genus or other parent taxon from the first word
+                #print(i, o_view[i].id)
+                taxon_name_sp = taxon_name.split(b" ")
+                taxon = taxo.get_taxon_by_name(taxon_name_sp[0], res_anc)
+                if taxon is not None:
+                    parent_taxid_column[i] = taxon.taxid
+                    parent_found_count+=1
+                    if logfile:
+                        print("Found parent taxon for", tostr(taxon_name), file=logfile)
+                else:
+                    not_found_count+=1
+                    if logfile:
+                        print("No taxid found for", tostr(taxon_name), file=logfile)
+            else:
+                not_found_count+=1
+                if logfile:
+                    print("No taxid found for", tostr(taxon_name), file=logfile)
+
+    except Exception, e:
+        raise RollbackException("obi addtaxids error, rollbacking view: "+str(e), o_view)
+
+    if pb is not None:
+        pb(i, force=True)
+        print("", file=sys.stderr)
+    
+    logger("info", "\nTaxids found: "+str(found_count)+"/"+str(len(o_view))+" ("+str(round(found_count*100.0/len(o_view), 2))+"%)")
+    if config['addtaxids']['try_genus_match']:
+        logger("info", "\nParent taxids found: "+str(parent_found_count)+"/"+str(len(o_view))+" ("+str(round(parent_found_count*100.0/len(o_view), 2))+"%)")
+    logger("info", "\nTaxids not found: "+str(not_found_count)+"/"+str(len(o_view))+" ("+str(round(not_found_count*100.0/len(o_view), 2))+"%)") 
+    
+    # Save command config in View and DMS comments
+    command_line = " ".join(sys.argv[1:])
+    input_dms_name=[input[0].name]
+    input_view_name=[i_view_name]
+    if 'taxoURI' in config['obi'] and config['obi']['taxoURI'] is not None:
+        input_dms_name.append(config['obi']['taxoURI'].split("/")[-3])
+        input_view_name.append("taxonomy/"+config['obi']['taxoURI'].split("/")[-1])
+    o_view.write_config(config, "addtaxids", command_line, input_dms_name=input_dms_name, input_view_name=input_view_name)
+    o_dms.record_command_line(command_line)
+
+    #print("\n\nOutput view:\n````````````", file=sys.stderr)
+    #print(repr(o_view), file=sys.stderr)
+
+    # stdout output: write to buffer
+    if type(output_0)==BufferedWriter:
+        logger("info", "Printing to output...")
+        o_view.print_to_output(output_0, noprogressbar=config['obi']['noprogressbar'])
+        o_view.close()
+
+    # If the input and the output DMS are different or if stdout output, delete the temporary imported view used to create the final view
+    if i_dms != o_dms or type(output_0)==BufferedWriter:
+        View.delete_view(o_dms, imported_view_name)
+        o_dms.close(force=True)
+    i_dms.close(force=True)
+
+    logger("info", "Done.")

python/obitools3/commands/align.pyx

@@ -19,7 +19,7 @@ import time
 import sys
 
 
-__title__="Aligns one sequence column with itself or two sequence columns"
+__title__="Align one sequence column with itself or two sequence columns"
 
 
 def addOptions(parser):
@@ -158,7 +158,7 @@ def run(config):
     i_view_name = i_uri.split(b"/")[0]
     i_column_name = b""
     i_element_name = b""
-    if len(i_uri.split(b"/")) == 2:
+    if len(i_uri.split(b"/")) >= 2:
         i_column_name = i_uri.split(b"/")[1]
     if len(i_uri.split(b"/")) == 3:
         i_element_name = i_uri.split(b"/")[2]
@@ -181,7 +181,7 @@ def run(config):
         i_dms_name_2 = i_dms_2.name
         i_uri_2 = input_2[1]
         original_i_view_name_2 = i_uri_2.split(b"/")[0]
-        if len(i_uri_2.split(b"/")) == 2:
+        if len(i_uri_2.split(b"/")) >= 2:
             i_column_name_2 = i_uri_2.split(b"/")[1]
         if len(i_uri_2.split(b"/")) == 3:
             i_element_name_2 = i_uri_2.split(b"/")[2]

python/obitools3/commands/alignpairedend.pyx

@@ -23,7 +23,7 @@ import os
 
 from cpython.exc cimport PyErr_CheckSignals
 
-__title__="Aligns paired-ended reads"
+__title__="Align paired-ended reads"
 
 
 
@@ -205,19 +205,25 @@ def run(config):
     if type(entries) == list:
         forward = entries[0]
         reverse = entries[1]
-        aligner = Kmer_similarity(forward, \
-                                  view2=reverse, \
-                                  kmer_size=config['alignpairedend']['kmersize'], \
-                                  reversed_column=None)
+        if len(forward) == 0 or len(reverse) == 0:
+            aligner = None
+        else:
+            aligner = Kmer_similarity(forward, \
+                                      view2=reverse, \
+                                      kmer_size=config['alignpairedend']['kmersize'], \
+                                      reversed_column=None)
     else:
-        aligner = Kmer_similarity(entries, \
-                                  column2=entries[REVERSE_SEQUENCE_COLUMN], \
-                                  qual_column2=entries[REVERSE_QUALITY_COLUMN], \
-                                  kmer_size=config['alignpairedend']['kmersize'], \
-                                  reversed_column=entries[b'reversed'])  # column created by the ngsfilter tool
+        if len(entries) == 0:
+            aligner = None
+        else:
+            aligner = Kmer_similarity(entries, \
+                                      column2=entries[REVERSE_SEQUENCE_COLUMN], \
+                                      qual_column2=entries[REVERSE_QUALITY_COLUMN], \
+                                      kmer_size=config['alignpairedend']['kmersize'], \
+                                      reversed_column=entries[b'reversed'])  # column created by the ngsfilter tool
         
     ba = alignmentIterator(entries, aligner)
-
+    
     i = 0
     for ali in ba:
         
@@ -227,6 +233,10 @@ def run(config):
         PyErr_CheckSignals()
 
         consensus = o_view[i]
+        
+        if two_views:
+            consensus[b"R1_parent"] = forward[i].id
+            consensus[b"R2_parent"] = reverse[i].id
                
         if not two_views:
             seqF = entries[i]
@@ -251,7 +261,7 @@ def run(config):
         pb(i, force=True)
         print("", file=sys.stderr)
 
-    if kmer_ali :
+    if kmer_ali and aligner is not None:
         aligner.free()
 
     # Save command config in View and DMS comments

python/obitools3/commands/annotate.pyx

@@ -16,6 +16,8 @@ from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
                                         QUALITY_COLUMN, \
                                         COUNT_COLUMN, \
                                         TAXID_COLUMN
+from obitools3.dms.capi.obitypes cimport OBI_STR
+from obitools3.dms.column.column cimport Column
 
 import time
 import math 
@@ -187,6 +189,8 @@ def sequenceTaggerGenerator(config, taxo=None):
                     else:
                         scn=None
                     seq[rank]=rtaxid
+                    if "%s_name"%rank not in seq.view:
+                        Column.new_column(seq.view, "%s_name"%rank, OBI_STR)
                     seq["%s_name"%rank]=scn
              
         if add_rank:

python/obitools3/commands/build_ref_db.pyx

@@ -16,7 +16,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Tag a set of sequences for PCR and sequencing errors identification"
+__title__="Build a reference database for ecotag"
 
 
 def addOptions(parser):
@@ -31,10 +31,9 @@ def addOptions(parser):
     group.add_argument('--threshold','-t',
                       action="store", dest="build_ref_db:threshold",
                       metavar='<THRESHOLD>',
-                      default=0.0,
+                      default=0.99,
                       type=float,
-                      help="Score threshold as a normalized identity, e.g. 0.95 for an identity of 95%%. Default: 0.00"
-                           " (no threshold).")
+                      help="Score threshold as a normalized identity, e.g. 0.95 for an identity of 95%%. Default: 0.99.")
 
 
 def run(config):

python/obitools3/commands/cat.pyx

@@ -4,7 +4,7 @@ from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
 from obitools3.dms import DMS
 from obitools3.dms.view.view cimport View
 from obitools3.uri.decode import open_uri
-from obitools3.apps.optiongroups import addMinimalOutputOption
+from obitools3.apps.optiongroups import addMinimalOutputOption, addNoProgressBarOption
 from obitools3.dms.view import RollbackException
 from obitools3.apps.config import logger
 from obitools3.utils cimport str2bytes
@@ -22,12 +22,13 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Concatenate views."
+__title__="Concatenate views"
 
  
 def addOptions(parser):
     
     addMinimalOutputOption(parser)
+    addNoProgressBarOption(parser)
 
     group=parser.add_argument_group('obi cat specific options')
 
@@ -47,9 +48,9 @@ def run(config):
     
     logger("info", "obi cat")
 
-    # Open the views to concatenate
-    iview_list = []
+    # Check the views to concatenate
     idms_list = []
+    iview_list = []
     total_len = 0
     remove_qual = False
     remove_rev_qual = False
@@ -67,8 +68,9 @@ def run(config):
         if REVERSE_QUALITY_COLUMN not in i_view: # same as above for reverse quality
             remove_rev_qual = True
         total_len += len(i_view)
-        iview_list.append(i_view)
         idms_list.append(i_dms)
+        iview_list.append(i_view.name)
+        i_view.close()
 
     # Open the output: only the DMS
     output = open_uri(config['obi']['outputURI'],
@@ -95,10 +97,12 @@ def run(config):
             Column.new_column(o_view, REVERSE_QUALITY_COLUMN, OBI_QUAL, associated_column_name=REVERSE_SEQUENCE_COLUMN, associated_column_version=o_view[REVERSE_SEQUENCE_COLUMN].version)
         
     # Initialize multiple elements columns
-    if type(output_0)==BufferedWriter: 
+    if type(output_0)!=BufferedWriter: 
         dict_cols = {}
-        for v in iview_list:
+        for v_uri in config["cat"]["views_to_cat"]:
+            v = open_uri(v_uri)[1]
             for coln in v.keys():
+                col = v[coln]
                 if v[coln].nb_elements_per_line > 1:
                     if coln not in dict_cols:
                         dict_cols[coln] = {}
@@ -108,9 +112,10 @@ def run(config):
                     else:
                         dict_cols[coln]['eltnames'] = set(v[coln].elements_names + list(dict_cols[coln]['eltnames']))
                         dict_cols[coln]['nbelts'] = len(dict_cols[coln]['eltnames'])
+            v.close()            
         for coln in dict_cols:
             Column.new_column(o_view, coln, dict_cols[coln]['obitype'], 
-                              nb_elements_per_line=dict_cols[coln]['nbelts'], elements_names=list(dict_cols[coln]['eltnames']))
+                              nb_elements_per_line=dict_cols[coln]['nbelts'], elements_names=list(dict_cols[coln]['eltnames']), dict_column=True)
     
     # Initialize the progress bar
     if not config['obi']['noprogressbar']:
@@ -119,7 +124,8 @@ def run(config):
         pb = None
     
     i = 0
-    for v in iview_list:
+    for v_uri in config["cat"]["views_to_cat"]:
+        v = open_uri(v_uri)[1]
         for entry in v:
             PyErr_CheckSignals()
             if pb is not None:
@@ -128,8 +134,13 @@ def run(config):
                 rep = repr(entry)
                 output_0.write(str2bytes(rep)+b"\n")
             else:
-                o_view[i] = entry
+                try:
+                    o_view[i] = entry
+                except:
+                    print("\nError with entry:", repr(entry))
+                    print(repr(o_view))
             i+=1
+        v.close()
 
     # Deletes quality columns if needed
     if type(output_0)!=BufferedWriter: 
@@ -144,7 +155,7 @@ def run(config):
     
     # Save command config in DMS comments
     command_line = " ".join(sys.argv[1:])
-    o_view.write_config(config, "cat", command_line, input_dms_name=[d.name for d in idms_list], input_view_name=[v.name for v in iview_list])
+    o_view.write_config(config, "cat", command_line, input_dms_name=[d.name for d in idms_list], input_view_name=[vname for vname in iview_list])
     o_dms.record_command_line(command_line)
 
     #print("\n\nOutput view:\n````````````", file=sys.stderr)

python/obitools3/commands/clean.pyx

@@ -54,11 +54,11 @@ def addOptions(parser):
                        default=False,
                        help="Only sequences labeled as heads are kept in the output. Default: False")
 
-    group.add_argument('--cluster-tags', '-C',
-                       action="store_true", 
-                       dest="clean:cluster-tags",
-                       default=False,
-                       help="Adds tags for each sequence giving its cluster's head and weight for each sample.")
+#    group.add_argument('--cluster-tags', '-C',
+#                       action="store_true", 
+#                       dest="clean:cluster-tags",
+#                       default=False,
+#                       help="Adds tags for each sequence giving its cluster's head and weight for each sample.")
 
     group.add_argument('--thread-count','-p',   # TODO should probably be in a specific option group
                        action="store", dest="clean:thread-count",
@@ -142,4 +142,5 @@ def run(config):
     
     i_dms.close(force=True)
 
-    logger("info", "Done.")
+    logger("info", "Done.")
+    

python/obitools3/commands/count.pyx

@@ -10,7 +10,7 @@ from obitools3.dms.capi.obiview cimport COUNT_COLUMN
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Counts sequence records"
+__title__="Count sequence records"
 
  
 def addOptions(parser):
@@ -29,6 +29,12 @@ def addOptions(parser):
                         default=False,
                         help="Prints only the total count of sequence records (if a sequence has no `count` attribute, its default count is 1) (default: False).")
 
+    group.add_argument('-c','--count-tag',
+                        action="store", dest="count:countcol",
+                        default='COUNT',
+                        type=str,
+                        help="Name of the tag/column associated with the count information (default: COUNT).")
+
     
 def run(config):
     
@@ -41,18 +47,20 @@ def run(config):
     if input is None:
         raise Exception("Could not read input")
     entries = input[1]
-
+    
+    countcol = config['count']['countcol'] 
+    
     count1 = len(entries)
     count2 = 0
     
-    if COUNT_COLUMN in entries and ((config['count']['sequence'] == config['count']['all']) or (config['count']['all'])) :
+    if countcol in entries and ((config['count']['sequence'] == config['count']['all']) or (config['count']['all'])) :
         for e in entries:
             PyErr_CheckSignals()
-            count2+=e[COUNT_COLUMN]
+            count2+=e[countcol]
 
-    if COUNT_COLUMN in entries and (config['count']['sequence'] == config['count']['all']):
+    if countcol in entries and (config['count']['sequence'] == config['count']['all']):
         print(count1,count2)
-    elif COUNT_COLUMN in entries and config['count']['all']:
+    elif countcol in entries and config['count']['all']:
         print(count2)
     else:
         print(count1)

python/obitools3/commands/ecopcr.pyx

@@ -175,6 +175,14 @@ def run(config):
     o_dms_name = output[0].name
     o_view_name = output[1]
 
+    # Open the taxonomy DMS
+    taxdms = open_uri(config['obi']['taxoURI'],
+                     dms_only=True)
+    if taxdms is None:
+        raise Exception("Could not open taxonomy DMS")
+    tax_dms = taxdms[0]
+    tax_dms_name = taxdms[0].name
+    
     # Read taxonomy name    
     taxonomy_name = config['obi']['taxoURI'].split("/")[-1]   # Robust in theory
 
@@ -197,7 +205,8 @@ def run(config):
 
     # TODO: primers in comments?
 
-    if obi_ecopcr(i_dms.name_with_full_path, tobytes(i_view_name), tobytes(taxonomy_name), \
+    if obi_ecopcr(i_dms.name_with_full_path, tobytes(i_view_name), 
+                  tax_dms.name_with_full_path, tobytes(taxonomy_name), \
                   o_dms.name_with_full_path, tobytes(o_view_name), comments, \
                   tobytes(config['ecopcr']['primer1']), tobytes(config['ecopcr']['primer2']), \
                   config['ecopcr']['error'], \

python/obitools3/commands/ecotag.pyx

@@ -41,6 +41,17 @@ def addOptions(parser):
                       help="Minimum identity to consider for assignment, as a normalized identity, e.g. 0.95 for an identity of 95%%. "
                            "Default: 0.00 (no threshold).")
 
+    group.add_argument('--minimum-circle','-c',
+                      action="store", dest="ecotag:bubble_threshold",
+                      metavar='<CIRCLE_THRESHOLD>',
+                      default=0.99,
+                      type=float,
+                      help="Minimum identity considered for the assignment circle "
+                           "(sequence is assigned to the LCA of all sequences within a similarity circle of the best matches; "
+                           "the threshold for this circle is the highest value between <CIRCLE_THRESHOLD> and the best assignment score found for the query sequence). "
+                           "Give value as a normalized identity, e.g. 0.95 for an identity of 95%%. "
+                           "Default: 0.99.")
+
 def run(config):
         
     DMS.obi_atexit()
@@ -66,9 +77,8 @@ def run(config):
     ref_view_name = ref[1]
 
     # Check that the threshold demanded is greater than or equal to the threshold used to build the reference database
-    if config['ecotag']['threshold'] < eval(ref_dms[ref_view_name].comments["ref_db_threshold"]) :
-        print("Error: The threshold demanded (%f) is lower than the threshold used to build the reference database (%f).", 
-              config['ecotag']['threshold'], ref_dms[ref_view_name].comments["ref_db_threshold"])
+    if config['ecotag']['bubble_threshold'] < eval(ref_dms[ref_view_name].comments["ref_db_threshold"]) :
+        raise Exception(f"Error: The threshold demanded ({config['ecotag']['bubble_threshold']}) is lower than the threshold used to build the reference database ({float(ref_dms[ref_view_name].comments['ref_db_threshold'])}).")
     
     # Open the output: only the DMS
     output = open_uri(config['obi']['outputURI'],
@@ -113,8 +123,9 @@ def run(config):
     if obi_ecotag(i_dms.name_with_full_path, tobytes(i_view_name), \
                   ref_dms.name_with_full_path, tobytes(ref_view_name), \
                   taxo_dms.name_with_full_path, tobytes(taxonomy_name), \
-                  tobytes(o_view_name), comments, 
-                  config['ecotag']['threshold']) < 0:
+                  tobytes(o_view_name), comments, \
+                  config['ecotag']['threshold'], \
+                  config['ecotag']['bubble_threshold']) < 0:
         raise Exception("Error running ecotag")
     
     # If the input and output DMS are not the same, export result view to output DMS

python/obitools3/commands/export.pyx

@@ -6,6 +6,9 @@ from obitools3.apps.config import logger
 from obitools3.dms import DMS
 from obitools3.dms.obiseq import Nuc_Seq
 from obitools3.dms.capi.obiview cimport QUALITY_COLUMN
+from obitools3.writers.tab import TabWriter
+from obitools3.format.tab import TabFormat
+from obitools3.utils cimport tobytes, tostr
 
 from obitools3.apps.optiongroups import addMinimalInputOption, \
                                         addExportOutputOption, \
@@ -76,6 +79,13 @@ def run(config):
     else:
         pb = ProgressBar(withoutskip - skip, config)
 
+    if config['obi']['outputformat'] == b'metabaR':
+        # Check prefix
+        if "metabarprefix" not in config["obi"]:
+            raise Exception("Prefix needed when exporting for metabaR (--metabaR-prefix option)")
+        else:
+            metabaRprefix = config["obi"]["metabarprefix"]
+        
     i=0
     for seq in iview :
         PyErr_CheckSignals()
@@ -89,7 +99,82 @@ def run(config):
 
     if pb is not None:
         pb(i, force=True)
-    print("", file=sys.stderr)
+        print("", file=sys.stderr)
+
+    if config['obi']['outputformat'] == b'metabaR':
+        
+        # Export ngsfilter file if view provided
+        if 'metabarngsfilter' in config['obi']:
+            ngsfilter_input = open_uri(config['obi']['metabarngsfilter'])
+            if ngsfilter_input is None:
+                raise Exception("Could not read ngsfilter view for metabaR output")
+            ngsfilter_view = ngsfilter_input[1]
+    
+            ngsfilter_output = open(config['obi']['metabarprefix']+'.ngsfilter', 'w')
+            
+            for line in ngsfilter_view:
+
+                line_to_print = b""                
+                line_to_print += line[b'experiment']
+                line_to_print += b"\t"
+                line_to_print += line[b'sample']
+                line_to_print += b"\t"
+                line_to_print += line[b'forward_tag']
+                line_to_print += b":"
+                line_to_print += line[b'reverse_tag']
+                line_to_print += b"\t"
+                line_to_print += line[b'forward_primer']
+                line_to_print += b"\t"
+                line_to_print += line[b'reverse_primer']
+                line_to_print += b"\t"
+                line_to_print += line[b'additional_info']
+                
+                print(tostr(line_to_print), file=ngsfilter_output)
+            
+            if ngsfilter_input[0] != input[0]:
+                ngsfilter_input[0].close()
+            ngsfilter_output.close()
+            
+        # Export sample metadata
+        samples_output = open(config['obi']['metabarprefix']+'_samples.csv', 'w')
+
+        # Export sample metadata file if view provided
+        if 'metabarsamples' in config['obi']:
+            samples_input = open_uri(config['obi']['metabarsamples'])
+            if samples_input is None:
+                raise Exception("Could not read sample view for metabaR output")
+            samples_view = samples_input[1]
+        
+            # Export with tab formatter
+            TabWriter(TabFormat(header=True, sep='\t',), 
+                      samples_output,
+                      header=True)
+            
+            if samples_input[0] != input[0]:
+                samples_input[0].close()
+        
+        # Else export just sample names from main view
+        else:
+            
+            sample_list = []
+            if 'MERGED_sample' in iview:
+                sample_list = iview['MERGED_sample'].keys()
+            elif 'sample' not in iview:
+                for seq in iview:
+                    sample = seq['sample']
+                    if sample not in sample_list:
+                        sample_list.append(sample)
+            else:
+                logger("warning", "Can not read sample list from main view for metabaR sample list export")            
+            
+            print("sample_id", file=samples_output)
+            for sample in sample_list:
+                line_to_print = b""                
+                line_to_print += sample
+                line_to_print += b"\t"
+                print(tostr(line_to_print), file=samples_output)
+        
+        samples_output.close()
 
     # TODO save command in input dms?
     

python/obitools3/commands/grep.pyx

@@ -91,7 +91,7 @@ def addOptions(parser):
                        metavar="<ATTRIBUTE_NAME>",
                        help="Select records with the attribute <ATTRIBUTE_NAME> "
                             "defined (not set to NA value). "
-                            "Several -a options can be used on the same "
+                            "Several -A options can be used on the same "
                             "command line.")
     
     group.add_argument("-L", "--lmax",
@@ -184,7 +184,7 @@ def Filter_generator(options, tax_filter, i_view):
     invert_selection = options["invert_selection"]
     id_set = None
     if "id_list" in options:
-        id_set = set(x.strip() for x in open(options["id_list"]))
+        id_set = set(x.strip() for x in open(options["id_list"], 'rb'))
 
     # Initialize the regular expression patterns
     seq_pattern = None
@@ -258,6 +258,13 @@ def Filter_generator(options, tax_filter, i_view):
 
 
 def Taxonomy_filter_generator(taxo, options):
+    
+    if (("required_ranks" in options and options["required_ranks"]) or \
+       ("required_taxids" in options and options["required_taxids"]) or \
+       ("ignored_taxids" in options and options["ignored_taxids"])) and \
+       (taxo is None):
+        raise RollbackException("obi grep error: can't use taxonomy options without providing a taxonomy. Rollbacking view")
+    
     if taxo is not None:
         def tax_filter(seq):
             good = True

python/obitools3/commands/head.pyx

@@ -8,6 +8,7 @@ from obitools3.apps.optiongroups import addMinimalInputOption, addMinimalOutputO
 from obitools3.dms.view import RollbackException
 from obitools3.apps.config import logger
 from obitools3.utils cimport str2bytes
+from obitools3.apps.optiongroups import addExportOutputOption
 
 import time
 import sys
@@ -16,13 +17,14 @@ from io import BufferedWriter
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Keep the N first lines of a view."
+__title__="Keep the N first lines of a view"
 
  
 def addOptions(parser):
     
     addMinimalInputOption(parser)
     addMinimalOutputOption(parser)
+    addExportOutputOption(parser)
     addNoProgressBarOption(parser)
 
     group=parser.add_argument_group('obi head specific options')

python/obitools3/commands/import.pyx

@@ -2,6 +2,7 @@
 
 import sys
 import os
+import re
 
 from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
 from obitools3.dms.view.view cimport View
@@ -26,20 +27,25 @@ from obitools3.dms.capi.obiview cimport VIEW_TYPE_NUC_SEQS, \
                                         QUALITY_COLUMN, \
                                         COUNT_COLUMN, \
                                         TAXID_COLUMN, \
-                                        MERGED_PREFIX
+                                        MERGED_PREFIX, \
+                                        SCIENTIFIC_NAME_COLUMN
                                         
 from obitools3.dms.capi.obidms cimport obi_import_view
 
 from obitools3.dms.capi.obitypes cimport obitype_t, \
                                          OBI_VOID, \
-                                         OBI_QUAL
+                                         OBI_QUAL, \
+                                         OBI_STR, \
+                                         OBI_INT
 
 from obitools3.dms.capi.obierrno cimport obi_errno
 
 from obitools3.apps.optiongroups import addImportInputOption, \
                                         addTabularInputOption, \
                                         addTaxdumpInputOption, \
-                                        addMinimalOutputOption
+                                        addMinimalOutputOption, \
+                                        addNoProgressBarOption, \
+                                        addTaxonomyOption
 
 from obitools3.uri.decode import open_uri
 
@@ -48,9 +54,10 @@ from obitools3.apps.config import logger
 from cpython.exc cimport PyErr_CheckSignals
 
 from io import BufferedWriter
+import ast
 
 
-__title__="Imports sequences from different formats into a DMS"
+__title__="Import sequences from different formats into a DMS"
  
  
 default_config = {   'destview'     : None,
@@ -67,7 +74,9 @@ def addOptions(parser):
     addImportInputOption(parser)
     addTabularInputOption(parser)
     addTaxdumpInputOption(parser)
+    addTaxonomyOption(parser)
     addMinimalOutputOption(parser)
+    addNoProgressBarOption(parser)
 
     group = parser.add_argument_group('obi import specific options')
 
@@ -77,6 +86,15 @@ def addOptions(parser):
                      help="Do a first readthrough of the dataset if it contains huge dictionaries (more than 100 keys) for "
                           "a much faster import. This option is not recommended and will slow down the import in any other case.")
 
+    group.add_argument('--space-priority',
+                     action="store_true", dest="import:space_priority",
+                     default=False,
+                     help="If importing a view into another DMS, do it by importing each line, saving disk space if the original view "
+                          "has a line selection associated.")
+
+#    group.add_argument('--only-id',
+#                     action="store", dest="import:onlyid",
+#                     help="only id")
 
 def run(config):
     
@@ -89,6 +107,10 @@ def run(config):
     cdef   obitype_t   new_type
     cdef   bint        get_quality
     cdef   bint        NUC_SEQS_view
+    cdef   bint        silva
+    cdef   bint        rdp
+    cdef   bint        unite
+    cdef   bint        sintax
     cdef   int         nb_elts
     cdef   object      d
     cdef   View        view
@@ -99,6 +121,8 @@ def run(config):
     cdef   Column      seq_col
     cdef   Column      qual_col
     cdef   Column      old_column
+    cdef   Column      sci_name_col
+    cdef   bytes       sci_name
     cdef   bint        rewrite
     cdef   dict        dcols
     cdef   int         skipping
@@ -142,7 +166,7 @@ def run(config):
     else:
         v = None
     
-    if config['obi']['taxdump'] or isinstance(input[1], View): 
+    if config['obi']['taxdump'] or (isinstance(input[1], View) and not config['import']['space_priority']): 
         dms_only=True
     else:
         dms_only=False
@@ -170,17 +194,33 @@ def run(config):
         logger("info", "Done.")
         return
 
-    # If importing a view between two DMS, use C API
-    if isinstance(input[1], View):
+    # Open taxonomy if there is one
+    if 'taxoURI' in config['obi'] and config['obi']['taxoURI'] is not None:
+        taxo_uri = open_uri(config['obi']['taxoURI'])
+        if taxo_uri is None or taxo_uri[2] == bytes:
+            raise Exception("Couldn't open taxonomy")
+        taxo = taxo_uri[1]
+
+    else :
+        taxo = None
+
+    # If importing a view between two DMS and not wanting to save space if line selection in original view, use C API
+    if isinstance(input[1], View) and not config['import']['space_priority']:
         if obi_import_view(input[0].name_with_full_path, o_dms.name_with_full_path, input[1].name, tobytes((config['obi']['outputURI'].split('/'))[-1])) < 0 :
+            input[0].close(force=True)
+            output[0].close(force=True)
             raise Exception("Error importing a view in a DMS")
         o_dms.record_command_line(" ".join(sys.argv[1:]))
-        o_dms.close()
+        input[0].close(force=True)
+        output[0].close(force=True)
         logger("info", "Done.")
         return
 
-    if entry_count >= 0:
+    # Reinitialize the progress bar
+    if entry_count >= 0 and config['obi']['noprogressbar'] == False:
         pb = ProgressBar(entry_count, config)
+    else:
+        pb = None
         
     NUC_SEQS_view = False
     if isinstance(output[1], View) :
@@ -189,15 +229,38 @@ def run(config):
             NUC_SEQS_view = True
     else: 
         raise NotImplementedError()
-        
+
     # Save basic columns in variables for optimization
     if NUC_SEQS_view :
         id_col = view[ID_COLUMN]
         def_col = view[DEFINITION_COLUMN]
         seq_col = view[NUC_SEQUENCE_COLUMN]
-        
+
+    # Prepare taxon scientific name and taxid refs if RDP/SILVA/UNITE/SINTAX formats
+    silva = False
+    rdp = False
+    unite = False
+    sintax=False
+    if 'inputformat' in config['obi'] and (config['obi']['inputformat'] == b"silva" or \
+                                           config['obi']['inputformat'] == b"rdp" or \
+                                           config['obi']['inputformat'] == b"unite" or \
+                                           config['obi']['inputformat'] == b"sintax"):
+        #if taxo is None:
+        #    raise Exception("A taxonomy (as built by 'obi import --taxdump') must be provided for SILVA and RDP files")
+        if config['obi']['inputformat'] == b"silva":
+            silva = True
+        elif config['obi']['inputformat'] == b"rdp":
+            rdp = True
+        elif config['obi']['inputformat'] == b"unite":
+            unite = True
+        elif config['obi']['inputformat'] == b"sintax":
+            sintax = True
+        sci_name_col = Column.new_column(view, SCIENTIFIC_NAME_COLUMN, OBI_STR)
+        if taxo is not None:
+            taxid_col = Column.new_column(view, TAXID_COLUMN, OBI_INT)
+
     dcols = {}
-        
+
     # First read through the entries to prepare columns with dictionaries as they are very time-expensive to rewrite
     if config['import']['preread']:
         logger("info", "First readthrough...")
@@ -237,15 +300,21 @@ def run(config):
         for tag in dict_dict:
             dcols[tag] = (Column.new_column(view, tag, dict_dict[tag][1], \
                               nb_elements_per_line=len(dict_dict[tag][0]), \
-                              elements_names=list(dict_dict[tag][0])), \
+                              elements_names=list(dict_dict[tag][0]), \
+                              dict_column=True), \
                           dict_dict[tag][1])
     
         
         # Reinitialize the input
         if isinstance(input[0], CompressedFile):
             input_is_file = True
-        if entry_count >= 0:
+        
+        # Reinitialize the progress bar
+        if entry_count >= 0 and config['obi']['noprogressbar'] == False:
             pb = ProgressBar(entry_count, config)
+        else:
+            pb = None
+
         try:
             input[0].close()
         except AttributeError:
@@ -254,6 +323,11 @@ def run(config):
         if input is None:
             raise Exception("Could not open input URI")
     
+#    if 'onlyid' in config['import']:
+#        onlyid = tobytes(config['import']['onlyid'])
+#    else:
+#        onlyid = None
+    
     entries = input[1]
     i = 0
     for entry in entries :
@@ -271,10 +345,13 @@ def run(config):
         elif not i%50000:
             logger("info", "Imported %d entries", i)
         
+#        if onlyid is not None and entry.id != onlyid:
+#            continue
+        
         try:
              
             if NUC_SEQS_view: 
-                id_col[i] = entry.id
+                id_col[i] = entry.id           
                 def_col[i] = entry.definition
                 seq_col[i] = entry.seq
                 # Check if there is a sequencing quality associated by checking the first entry    # TODO haven't found a more robust solution yet
@@ -285,40 +362,89 @@ def run(config):
                         qual_col = view[QUALITY_COLUMN]
                 if get_quality:
                     qual_col[i] = entry.quality
-             
+                    
+                # Parse taxon scientific name if RDP or Silva or Unite file
+                if (rdp or silva or unite or sintax):
+                    if rdp or silva:
+                        sci_names = entry.definition.split(b";")
+                        sci_name_col[i] = sci_names[-1]
+                    elif unite:
+                        sci_names = entry.id.split(b'|')[-1].split(b';')
+                        sci_name_col[i] = re.sub(b'[a-zA-Z]__', b'', sci_names[-1])
+                    elif sintax:
+                        reconstructed_line = entry.id+b' '+entry.definition[:-1]
+                        splitted_reconstructed_line = reconstructed_line.split(b';')
+                        taxa = splitted_reconstructed_line[1].split(b'=')[1]
+                        taxa = splitted_reconstructed_line[1].split(b',')
+                        sci_names = []
+                        for t in taxa:
+                            tf = t.split(b':')[1]
+                            sci_names.append(tf)
+                        sci_name_col[i] = sci_names[-1]
+                        id_col[i] = reconstructed_line.split(b';')[0]    
+                        def_col[i] = reconstructed_line
+                        
+                    # Fond taxid if taxonomy provided
+                    if taxo is not None :
+                        for sci_name in reversed(sci_names):
+                            if unite:
+                                sci_name = re.sub(b'[a-zA-Z]__', b'', sci_name)
+                            if sci_name.split()[0] != b'unidentified' and sci_name.split()[0] != b'uncultured' and sci_name.split()[0] != b'metagenome':
+                                taxon = taxo.get_taxon_by_name(sci_name)
+                                if taxon is not None:
+                                    sci_name_col[i] = taxon.name
+                                    taxid_col[i] = taxon.taxid
+                                    #print(taxid_col[i], sci_name_col[i])
+                                    break
+
             for tag in entry :
                 
                 if tag != ID_COLUMN and tag != DEFINITION_COLUMN and tag != NUC_SEQUENCE_COLUMN and tag != QUALITY_COLUMN :  # TODO dirty 
-                                    
+                
                     value = entry[tag]
                     if tag == b"taxid":
                         tag = TAXID_COLUMN
                     if tag == b"count":
                         tag = COUNT_COLUMN
+                    if tag == b"scientific_name":
+                        tag = SCIENTIFIC_NAME_COLUMN
                     if tag[:7] == b"merged_":
                         tag = MERGED_PREFIX+tag[7:]
-                        
+
+                    if type(value) == bytes and value[:1]==b"[" :
+                        try:
+                            if type(eval(value)) == list:
+                                value = eval(value)
+                                #print(value)
+                        except:
+                            pass
+                    
                     if tag not in dcols :
-                         
+                    
                         value_type = type(value)
                         nb_elts = 1
                         value_obitype = OBI_VOID
-                         
-                        if value_type == dict or value_type == list :
+                        dict_col = False
+                        
+                        if value_type == dict :
                             nb_elts = len(value)
                             elt_names = list(value)
+                            dict_col = True
                         else :
                             nb_elts = 1
                             elt_names = None
-                         
+                        
+                        if value_type == list :
+                            tuples = True
+                        else:
+                            tuples = False
+                        
                         value_obitype = get_obitype(value)
-                         
+                                                
                         if value_obitype != OBI_VOID :
-                            dcols[tag] = (Column.new_column(view, tag, value_obitype, nb_elements_per_line=nb_elts, elements_names=elt_names), value_obitype)
+                            dcols[tag] = (Column.new_column(view, tag, value_obitype, nb_elements_per_line=nb_elts, elements_names=elt_names, dict_column=dict_col, tuples=tuples), value_obitype)
                                                      
                             # Fill value
-                            if value_type == dict and nb_elts == 1:  # special case that makes the OBI3 create a 1 elt/line column which won't read a dict value
-                                value = value[list(value.keys())[0]]       # The solution is to transform the value in a simple atomic one acceptable by the column
                             dcols[tag][0][i] = value
                          
                         # TODO else log error?
@@ -344,8 +470,8 @@ def run(config):
                             # Fill value
                             dcols[tag][0][i] = value
                          
-                        except IndexError :
-                                                     
+                        except (IndexError, OverflowError):
+                                                                                 
                             value_type = type(value)
                             old_column = dcols[tag][0]
                             old_nb_elements_per_line = old_column.nb_elements_per_line
@@ -392,18 +518,19 @@ def run(config):
                                 dcols[tag][0][i] = value
         
         except Exception as e:
-            print("\nCould not import sequence id:", entry.id, "(error raised:", e, ")")
+            print("\nCould not import sequence:\n", repr(entry), "\nError raised:", e, "\n/!\ Check if '--input-na-string' option needs to be set")
             if 'skiperror' in config['obi'] and not config['obi']['skiperror']:
                 raise e
             else:
                 pass
+                i-=1  # overwrite problematic entry
         
         i+=1
  
     if pb is not None:
         pb(i, force=True)
         print("", file=sys.stderr)
-    logger("info", "Imported %d entries", i)
+    logger("info", "Imported %d entries", len(view))
     
     # Save command config in View and DMS comments
     command_line = " ".join(sys.argv[1:])

python/obitools3/commands/ls.pyx

@@ -31,27 +31,11 @@ def run(config):
     input = open_uri(config['obi']['inputURI'])
     if input is None:
         raise Exception("Could not read input")
-    if input[2] == DMS and not config['ls']['longformat']:
-        dms = input[0]
-        l = []
-        for viewname in input[0]:
-            view = dms[viewname]
-            l.append(tostr(viewname) + "\t(Date created: " + str(bytes2str_object(view.comments["Date created"]))+")")
-            view.close()
-        l.sort()
-        for v in l:
-            print(v)
+    
+    # Print representation
+    if config['ls']['longformat']:
+        print(input[1].repr_longformat())
     else:
         print(repr(input[1]))
-    if input[2] == DMS:
-        taxolist = ["\n### Taxonomies:"]
-        for t in Taxonomy.list_taxos(input[0]):
-            taxolist.append("\t"+tostr(t))
-        if len(taxolist) > 1:
-            for t in taxolist:
-                print(t)
-    if config['ls']['longformat'] and len(input[1].comments) > 0:
-        print("\n### Comments:")
-        print(str(input[1].comments))
     
     input[0].close(force=True)

python/obitools3/commands/ngsfilter.pyx

@@ -24,11 +24,10 @@ from cpython.exc cimport PyErr_CheckSignals
 from io import BufferedWriter
 
 
-#REVERSE_SEQ_COLUMN_NAME = b"REVERSE_SEQUENCE"      # used by alignpairedend tool
-#REVERSE_QUALITY_COLUMN_NAME = b"REVERSE_QUALITY"   # used by alignpairedend tool
+MAX_PAT_LEN = 31
 
 
-__title__="Assigns sequence records to the corresponding experiment/sample based on DNA tags and primers"
+__title__="Assign sequence records to the corresponding experiment/sample based on DNA tags and primers"
 
 
 def addOptions(parser):
@@ -88,6 +87,8 @@ class Primer:
         @type direct:
         '''
         
+        assert len(sequence) <= MAX_PAT_LEN, "Primer %s is too long, 31 bp max" % sequence
+        
         assert sequence not in Primer.collection        \
             or Primer.collection[sequence]==taglength,  \
             "Primer %s must always be used with tags of the same length" % sequence
@@ -268,14 +269,18 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
     
     not_aligned = len(sequences) > 1
     sequences[0] = sequences[0].clone()
+    
+    if not_aligned:
+        sequences[0][b"R1_parent"] = sequences[0].id
+        sequences[0][b"R2_parent"] = sequences[1].id
 
     if not_aligned:
         sequences[1] = sequences[1].clone()
-        sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq             # used by alignpairedend tool
+        sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq        # used by alignpairedend tool
         sequences[0][REVERSE_QUALITY_COLUMN] = sequences[1].quality     # used by alignpairedend tool
 
     for seq in sequences:
-        if hasattr(seq, "quality_array"): 
+        if hasattr(seq, "quality_array") and seq.quality_array is not None: 
             q = -reduce(lambda x,y:x+y,(math.log10(z) for z in seq.quality_array),0)/len(seq.quality_array)*10
             seq[b'avg_quality']=q
             q = -reduce(lambda x,y:x+y,(math.log10(z) for z in seq.quality_array[0:10]),0)
@@ -299,7 +304,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
             directmatch.append((p, p(seq, same_sequence=not new_seq, pattern=pattern), seq, p))
             new_seq = False
             pattern+=1
-    
+        
     # Choose match closer to the start of (one of the) sequence(s)
     directmatch = sorted(directmatch, key=sortMatch)
     all_direct_matches = directmatch
@@ -326,7 +331,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
         sequences[0] = sequences[0][directmatch[1][2]:]
     else:
         sequences[1] = sequences[1][directmatch[1][2]:]
-        sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq           # used by alignpairedend tool
+        sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq      # used by alignpairedend tool
         sequences[0][REVERSE_QUALITY_COLUMN] = sequences[1].quality   # used by alignpairedend tool
     
     if directmatch[0].forward:
@@ -373,7 +378,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
                 sequences[0] = sequences[0][:r[1]]
             else:
                 sequences[1] = sequences[1][:r[1]]
-                sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq           # used by alignpairedend tool
+                sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq      # used by alignpairedend tool
                 sequences[0][REVERSE_QUALITY_COLUMN] = sequences[1].quality   # used by alignpairedend tool
         # do the same on the other seq
         if first_match_first_seq: 
@@ -398,7 +403,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
             seq_to_match = sequences[0]
         reversematch = []
         # Compute begin
-        begin=directmatch[1][2]+1  # end of match + 1 on the same sequence
+        #begin=directmatch[1][2]+1  # end of match + 1 on the same sequence -- No, already cut out forward primer        
         # Try reverse matching on the other sequence:
         new_seq = True
         pattern = 0
@@ -412,7 +417,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
                 primer=p
             # Saving original primer as 4th member of the tuple to serve as correct key in infos dict even if it might have been reversed complemented
             # (3rd member already used by directmatch)
-            reversematch.append((primer, primer(seq_to_match, same_sequence=not new_seq, pattern=pattern, begin=begin), None, p))
+            reversematch.append((primer, primer(seq_to_match, same_sequence=not new_seq, pattern=pattern, begin=0), None, p))
             new_seq = False
             pattern+=1
         # Choose match closer to the end of the sequence
@@ -454,7 +459,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
             sequences[1] = sequences[1][reversematch[1][2]:]
             if not directmatch[0].forward:
                 sequences[1] = sequences[1].reverse_complement
-            sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq           # used by alignpairedend tool
+            sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq      # used by alignpairedend tool
             sequences[0][REVERSE_QUALITY_COLUMN] = sequences[1].quality   # used by alignpairedend tool
         else:
             sequences[0] = sequences[0][reversematch[1][2]:]
@@ -649,6 +654,7 @@ def run(config):
     
     g = 0
     u = 0
+    i = 0
     no_tags = config['ngsfilter']['notags']
     try:
         for i in range(entries_len):

python/obitools3/commands/rm.pyx

@@ -0,0 +1,87 @@
+#cython: language_level=3
+
+from obitools3.uri.decode import open_uri
+from obitools3.apps.config import logger
+from obitools3.dms import DMS
+from obitools3.apps.optiongroups import addMinimalInputOption
+from obitools3.dms.view.view cimport View
+from obitools3.utils cimport tostr
+import os
+import shutil
+ 
+__title__="Delete a view"
+
+
+def addOptions(parser):    
+    addMinimalInputOption(parser)
+
+def run(config):
+
+    DMS.obi_atexit()
+    
+    logger("info", "obi rm")
+
+    # Open the input
+    input = open_uri(config['obi']['inputURI'])
+    if input is None:
+        raise Exception("Could not read input")
+    
+    # Check that it's a view
+    if isinstance(input[1], View) :
+        view = input[1]
+    else: 
+        raise NotImplementedError()
+    
+    dms = input[0]
+    
+    # Get the path to the view file to remove
+    path = dms.full_path  # dms path
+    view_path=path+b"/VIEWS/"
+    view_path+=view.name
+    view_path+=b".obiview"
+    
+    to_remove = {}
+    # For each column:
+    for col_alias in view.keys():
+        col = view[col_alias]
+        col_name = col.original_name
+        col_version = col.version
+        col_type = col.data_type
+        col_ref = (col_name, col_version)
+        # build file name and AVL file names
+        col_file_name = f"{tostr(path)}/{tostr(col.original_name)}.obicol/{tostr(col.original_name)}@{col.version}.odc"
+        if col_type in [b'OBI_CHAR', b'OBI_QUAL', b'OBI_STR', b'OBI_SEQ']:
+            avl_file_name = f"{tostr(path)}/OBIBLOB_INDEXERS/{tostr(col.original_name)}_{col.version}_indexer"
+        else:
+            avl_file_name = None
+        to_remove[col_ref] = [col_file_name, avl_file_name]
+    
+    # For each view:
+    do_not_remove = []
+    for vn in dms:
+        v = dms[vn]
+        # ignore the one being deleted
+        if v.name != view.name:
+            # check that none of the column is referenced, if referenced, remove from list to remove
+            cols = [(v[c].original_name, v[c].version) for c in v.keys()]
+            for col_ref in to_remove:
+                if col_ref in cols:
+                    do_not_remove.append(col_ref)
+    
+    for nr in do_not_remove:
+        to_remove.pop(nr)
+    
+    # Close the view and the DMS
+    view.close()
+    input[0].close(force=True)
+    
+    #print(to_remove)
+    
+    # rm AFTER view and DMS close
+    os.remove(view_path)
+    for col in to_remove:
+        os.remove(to_remove[col][0])
+        if to_remove[col][1] is not None:
+            shutil.rmtree(to_remove[col][1])
+        
+        

python/obitools3/commands/sort.pyx

@@ -36,7 +36,7 @@ NULL_VALUE = {OBI_BOOL: OBIBool_NA,
               OBI_STR: b""}
  
 
-__title__="Sort view lines according to the value of a given attribute."
+__title__="Sort view lines according to the value of a given attribute"
 
  
 def addOptions(parser):

python/obitools3/commands/split.pyx

@@ -0,0 +1,105 @@
+#cython: language_level=3
+
+from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
+from obitools3.dms import DMS
+from obitools3.dms.view.view cimport View, Line_selection
+from obitools3.uri.decode import open_uri
+from obitools3.apps.optiongroups import addMinimalInputOption, addTaxonomyOption, addMinimalOutputOption, addNoProgressBarOption
+from obitools3.dms.view import RollbackException
+from obitools3.apps.config import logger
+from obitools3.utils cimport tobytes
+
+import sys
+from cpython.exc cimport PyErr_CheckSignals
+
+ 
+__title__="Split"
+
+ 
+def addOptions(parser):
+    
+    addMinimalInputOption(parser)
+    addNoProgressBarOption(parser)
+
+    group=parser.add_argument_group("obi split specific options")
+ 
+    group.add_argument('-p','--prefix',
+                       action="store", dest="split:prefix",
+                       metavar="<PREFIX>",
+                       help="Prefix added to each subview name (included undefined)")
+
+    group.add_argument('-t','--tag-name',
+                       action="store", dest="split:tagname",
+                       metavar="<TAG_NAME>",
+                       help="Attribute/tag used to split the input")
+      
+    group.add_argument('-u','--undefined',
+                       action="store", dest="split:undefined",
+                       default=b'UNDEFINED',
+                       metavar="<VIEW_NAME>",
+                       help="Name of the view where undefined sequenced are stored (will be PREFIX_VIEW_NAME)")
+
+
+def run(config):
+     
+    DMS.obi_atexit()
+    
+    logger("info", "obi split")
+
+    # Open the input
+    input = open_uri(config["obi"]["inputURI"])
+    if input is None:
+        raise Exception("Could not read input view")
+    i_dms = input[0]
+    i_view = input[1]
+
+    # Initialize the progress bar
+    if config['obi']['noprogressbar'] == False:
+        pb = ProgressBar(len(i_view), config)
+    else:
+        pb = None
+    
+    tag_to_split = config["split"]["tagname"]
+    undefined = tobytes(config["split"]["undefined"])
+    selections = {}
+    
+    # Go through input view and split
+    for i in range(len(i_view)):
+        PyErr_CheckSignals()
+        if pb is not None:
+            pb(i)
+        line = i_view[i]
+        if tag_to_split not in line or line[tag_to_split] is None or len(line[tag_to_split])==0:
+            value = undefined
+        else:
+            value = line[tag_to_split]
+        if value not in selections:
+            selections[value] = Line_selection(i_view)
+        selections[value].append(i)
+        
+    if pb is not None:
+        pb(len(i_view), force=True)
+        print("", file=sys.stderr)
+
+    # Create output views with the line selection
+    try:
+        for cat in selections:
+            o_view_name = config["split"]["prefix"].encode()+cat
+            o_view = selections[cat].materialize(o_view_name)
+            # Save command config in View and DMS comments
+            command_line = " ".join(sys.argv[1:])
+            input_dms_name=[input[0].name]
+            input_view_name=[input[1].name]
+            if 'taxoURI' in config['obi'] and config['obi']['taxoURI'] is not None:
+                input_dms_name.append(config['obi']['taxoURI'].split("/")[-3])
+                input_view_name.append("taxonomy/"+config['obi']['taxoURI'].split("/")[-1])
+            o_view.write_config(config, "split", command_line, input_dms_name=input_dms_name, input_view_name=input_view_name)
+            o_view.close()
+    except Exception, e:
+        raise RollbackException("obi split error, rollbacking view: "+str(e), o_view)
+
+    i_dms.record_command_line(command_line)
+    i_dms.close(force=True)
+
+    logger("info", "Done.")
+

python/obitools3/commands/stats.pyx

@@ -16,7 +16,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Compute basic statistics for attribute values."
+__title__="Compute basic statistics for attribute values"
 
 '''
 `obi stats` computes basic statistics for attribute values of sequence records.
@@ -119,9 +119,12 @@ def mean(values, options):
 
 def variance(v):
     if len(v)==1: 
-        return 0 
+        return 0
     s = reduce(lambda x,y:(x[0]+y,x[1]+y**2),v,(0.,0.))
-    return s[1]/(len(v)-1) - s[0]**2/len(v)/(len(v)-1)
+    var = round(s[1]/(len(v)-1) - s[0]**2/len(v)/(len(v)-1), 5) # round to go around shady python rounding stuff when var is actually 0
+    if var == -0.0: # then fix -0 to +0 if was rounded to -0
+        var = 0.0
+    return var
 
 
 def varpop(values, options):
@@ -154,7 +157,7 @@ def run(config):
     else :
         taxo = None
 
-    statistics = set(config['stats']['minimum']) | set(config['stats']['maximum']) | set(config['stats']['mean'])
+    statistics = set(config['stats']['minimum']) | set(config['stats']['maximum']) | set(config['stats']['mean']) | set(config['stats']['var']) | set(config['stats']['sd'])
     total = 0
     catcount={}
     totcount={}
@@ -195,7 +198,7 @@ def run(config):
         except KeyError:
             totcount[category]=totcount.get(category,0)+1
         for var in statistics:
-            if var in line:
+            if var in line and line[var] is not None:
                 v = line[var]
                 if var not in values:
                     values[var]={}
@@ -238,14 +241,34 @@ def run(config):
     else:
         sdvar= "%s"
         
-    hcat = "\t".join([pcat % x for x in config['stats']['categories']]) + "\t" +\
-           "\t".join([minvar % x for x in config['stats']['minimum']])  + "\t" +\
-           "\t".join([maxvar % x for x in config['stats']['maximum']])  + "\t" +\
-           "\t".join([meanvar % x for x in config['stats']['mean']])  + "\t" +\
-           "\t".join([varvar % x for x in config['stats']['var']])  + "\t" +\
-           "\t".join([sdvar % x for x in config['stats']['sd']]) + \
-           "\t   count" + \
-           "\t   total" 
+    hcat = ""
+    
+    for x in config['stats']['categories']:
+        hcat += pcat % x
+        hcat += "\t"
+    
+    for x in config['stats']['minimum']:
+        hcat += minvar % x
+        hcat += "\t"
+
+    for x in config['stats']['maximum']:
+        hcat += maxvar % x
+        hcat += "\t"
+
+    for x in config['stats']['mean']:
+        hcat += meanvar % x
+        hcat += "\t"
+    
+    for x in config['stats']['var']:
+        hcat += varvar % x
+        hcat += "\t"
+
+    for x in config['stats']['sd']:
+        hcat += sdvar % x
+        hcat += "\t"
+
+    hcat += "count\ttotal" 
+    
     print(hcat)
     sorted_stats = sorted(catcount.items(), key = lambda kv:(totcount[kv[0]]), reverse=True)
     for i in range(len(sorted_stats)):
@@ -265,8 +288,8 @@ def run(config):
             print((("%%%df" % lvarp[m]) % varp[m][c])+"\t", end="")
         for m in config['stats']['sd']:
             print((("%%%df" % lsigma[m]) % sigma[m][c])+"\t", end="")
-        print("%7d" %catcount[c], end="")
-        print("%9d" %totcount[c])
+        print("%d" %catcount[c]+"\t", end="")
+        print("%d" %totcount[c]+"\t")
 
     input[0].close(force=True)
     

python/obitools3/commands/tail.pyx

@@ -15,7 +15,7 @@ from cpython.exc cimport PyErr_CheckSignals
 from io import BufferedWriter
 
 
-__title__="Keep the N last lines of a view."
+__title__="Keep the N last lines of a view"
 
  
 def addOptions(parser):

python/obitools3/commands/taxonomy.pyx

@@ -0,0 +1,230 @@
+#cython: language_level=3
+
+from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
+from obitools3.dms import DMS
+from obitools3.dms.view.view cimport View, Line_selection
+from obitools3.uri.decode import open_uri
+from obitools3.apps.optiongroups import addMinimalInputOption, addTaxonomyOption, addMinimalOutputOption, addNoProgressBarOption
+from obitools3.dms.view import RollbackException
+from obitools3.dms.column.column cimport Column
+from functools import reduce
+from obitools3.apps.config import logger
+from obitools3.utils cimport tobytes, str2bytes, tostr
+from io import BufferedWriter
+from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
+                                        ID_COLUMN, \
+                                        DEFINITION_COLUMN, \
+                                        QUALITY_COLUMN, \
+                                        COUNT_COLUMN, \
+                                        TAXID_COLUMN
+from obitools3.dms.capi.obitypes cimport OBI_INT
+from obitools3.dms.capi.obitaxonomy cimport MIN_LOCAL_TAXID
+import time
+import math
+import sys
+
+from cpython.exc cimport PyErr_CheckSignals
+
+
+__title__="Add taxa with a new generated taxid to an NCBI taxonomy database"
+
+
+
+def addOptions(parser):
+
+    addMinimalInputOption(parser)
+    addTaxonomyOption(parser)
+    addMinimalOutputOption(parser)
+    addNoProgressBarOption(parser)
+
+    group=parser.add_argument_group('obi taxonomy specific options')
+
+    group.add_argument('-n', '--taxon-name-tag',
+                       action="store",
+                       dest="taxonomy:taxon_name_tag",
+                       metavar="<SCIENTIFIC_NAME_TAG>",
+                       default=b"SCIENTIFIC_NAME",
+                       help="Name of the tag giving the scientific name of the taxon "
+                            "(default: 'SCIENTIFIC_NAME').")
+
+#    group.add_argument('-g', '--try-genus-match',
+#                     action="store_true", dest="taxonomy:try_genus_match",
+#                     default=False,
+#                     help="Try matching the first word of <SCIENTIFIC_NAME_TAG> when can't find corresponding taxid for a taxon. "
+#                          "If there is a match it is added in the 'parent_taxid' tag. (Can be used by 'obi taxonomy' to add the taxon under that taxid).")
+
+    group.add_argument('-a', '--restricting-ancestor',
+                       action="store",
+                       dest="taxonomy:restricting_ancestor",
+                       metavar="<RESTRICTING_ANCESTOR>",
+                       default=None,
+                       help="Enables to restrict the addition of taxids under an ancestor specified by its taxid.")
+
+    group.add_argument('-t', '--taxid-tag',
+                       action="store",
+                       dest="taxonomy:taxid_tag",
+                       metavar="<TAXID_TAG>",
+                       default=b"TAXID",
+                       help="Name of the tag to store the new taxid "
+                            "(default: 'TAXID').")
+
+    group.add_argument('-l', '--log-file',
+                       action="store",
+                       dest="taxonomy:log_file",
+                       metavar="<LOG_FILE>",
+                       default='',
+                       help="Path to a log file to write informations about added taxids.")
+
+
+def run(config):
+
+    DMS.obi_atexit()
+
+    logger("info", "obi taxonomy")
+
+    # Open the input
+    input = open_uri(config['obi']['inputURI'])
+    if input is None:
+        raise Exception("Could not read input view")
+    i_dms = input[0]
+    i_view = input[1]
+    i_view_name = input[1].name
+
+    # Open the output: only the DMS, as the output view is going to be created by cloning the input view
+    # (could eventually be done via an open_uri() argument)
+    output = open_uri(config['obi']['outputURI'],
+                      input=False,
+                      dms_only=True)
+    if output is None:
+        raise Exception("Could not create output view")
+    o_dms = output[0]
+    output_0 = output[0]
+    o_view_name = output[1]
+
+    # stdout output: create temporary view
+    if type(output_0)==BufferedWriter:
+        o_dms = i_dms
+        i=0
+        o_view_name = b"temp"
+        while o_view_name in i_dms: # Making sure view name is unique in output DMS
+            o_view_name = o_view_name+b"_"+str2bytes(str(i))
+            i+=1
+        imported_view_name = o_view_name
+
+    # If the input and output DMS are not the same, import the input view in the output DMS before cloning it to modify it
+    # (could be the other way around: clone and modify in the input DMS then import the new view in the output DMS)
+    if i_dms != o_dms:
+        imported_view_name = i_view_name
+        i=0
+        while imported_view_name in o_dms:  # Making sure view name is unique in output DMS
+            imported_view_name = i_view_name+b"_"+str2bytes(str(i))
+            i+=1
+        View.import_view(i_dms.full_path[:-7], o_dms.full_path[:-7], i_view_name, imported_view_name)
+        i_view = o_dms[imported_view_name]
+
+    # Clone output view from input view
+    o_view = i_view.clone(o_view_name)
+    if o_view is None:
+        raise Exception("Couldn't create output view")
+    i_view.close()
+
+    # Open taxonomy
+    taxo_uri = open_uri(config['obi']['taxoURI'])
+    if taxo_uri is None or taxo_uri[2] == bytes:
+        raise Exception("Couldn't open taxonomy")
+    taxo = taxo_uri[1]
+
+    # Initialize the progress bar
+    if config['obi']['noprogressbar'] == False:
+        pb = ProgressBar(len(o_view), config)
+    else:
+        pb = None
+
+    try:
+        if config['taxonomy']['log_file']:
+            logfile = open(config['taxonomy']['log_file'], 'w')
+        else:
+            logfile = sys.stdout
+        if 'restricting_ancestor' in config['taxonomy']:
+            res_anc = int(config['taxonomy']['restricting_ancestor'])
+        else:
+            res_anc = None
+        taxid_column_name = config['taxonomy']['taxid_tag']
+        parent_taxid_column_name = "PARENT_TAXID" # TODO macro
+        taxon_name_column_name = config['taxonomy']['taxon_name_tag']
+        taxid_column = Column.new_column(o_view, taxid_column_name, OBI_INT)
+        if parent_taxid_column_name in o_view:
+            parent_taxid_column = o_view[parent_taxid_column_name]
+        else:
+            parent_taxid_column = None
+        #parent_taxid_column = Column.new_column(o_view, parent_taxid_column_name, OBI_INT)
+        taxon_name_column = o_view[taxon_name_column_name]
+
+        for i in range(len(o_view)):
+            PyErr_CheckSignals()
+            #if pb is not None:
+            #    pb(i)
+            taxon_name = taxon_name_column[i]
+            taxon = taxo.get_taxon_by_name(taxon_name, res_anc)
+            if taxon is not None:
+                taxid_column[i] = taxon.taxid
+                if logfile:
+                    print(f"Found taxon '{tostr(taxon_name)}' already existing with taxid {taxid_column[i]}", file=logfile)
+            else: # try finding genus or other parent taxon from the first word
+                #print(i, o_view[i].id)
+                if parent_taxid_column is not None and parent_taxid_column[i] is not None:
+                    taxid_column[i] = taxo.add_taxon(taxon_name, 'species', parent_taxid_column[i])
+                    if logfile:
+                        print(f"Adding taxon '{tostr(taxon_name)}' under provided parent {parent_taxid_column[i]} with taxid {taxid_column[i]}", file=logfile)
+                else:
+                    taxon_name_sp = taxon_name.split(b" ")
+                    taxon = taxo.get_taxon_by_name(taxon_name_sp[0], res_anc)
+                    if taxon is not None:
+                        parent_taxid_column[i] = taxon.taxid
+                        taxid_column[i] = taxo.add_taxon(taxon_name, 'species', taxon.taxid)
+                        if logfile:
+                            print(f"Adding taxon '{tostr(taxon_name)}' under '{tostr(taxon.name)}' ({taxon.taxid}) with taxid {taxid_column[i]}", file=logfile)
+                    else:
+                        taxid_column[i] = taxo.add_taxon(taxon_name, 'species', res_anc)
+                        if logfile:
+                            print(f"Adding taxon '{tostr(taxon_name)}' under provided restricting ancestor {res_anc} with taxid {taxid_column[i]}", file=logfile)
+
+        taxo.write(taxo.name, update=True)
+        
+    except Exception, e:
+        raise RollbackException("obi taxonomy error, rollbacking view: "+str(e), o_view)
+
+    #if pb is not None:
+    #    pb(i, force=True)
+    #    print("", file=sys.stderr)
+
+    #logger("info", "\nTaxa already in the taxonomy: "+str(found_count)+"/"+str(len(o_view))+" ("+str(round(found_count*100.0/len(o_view), 2))+"%)")
+    #logger("info", "\nParent taxids found: "+str(parent_found_count)+"/"+str(len(o_view))+" ("+str(round(parent_found_count*100.0/len(o_view), 2))+"%)")
+    #logger("info", "\nTaxids not found: "+str(not_found_count)+"/"+str(len(o_view))+" ("+str(round(not_found_count*100.0/len(o_view), 2))+"%)")
+
+    # Save command config in View and DMS comments
+    command_line = " ".join(sys.argv[1:])
+    input_dms_name=[input[0].name]
+    input_view_name=[i_view_name]
+    if 'taxoURI' in config['obi'] and config['obi']['taxoURI'] is not None:
+        input_dms_name.append(config['obi']['taxoURI'].split("/")[-3])
+        input_view_name.append("taxonomy/"+config['obi']['taxoURI'].split("/")[-1])
+    o_view.write_config(config, "taxonomy", command_line, input_dms_name=input_dms_name, input_view_name=input_view_name)
+    o_dms.record_command_line(command_line)
+
+    #print("\n\nOutput view:\n````````````", file=sys.stderr)
+    #print(repr(o_view), file=sys.stderr)
+
+    # stdout output: write to buffer
+    if type(output_0)==BufferedWriter:
+        logger("info", "Printing to output...")
+        o_view.print_to_output(output_0, noprogressbar=config['obi']['noprogressbar'])
+        o_view.close()
+
+    # If the input and the output DMS are different or if stdout output, delete the temporary imported view used to create the final view
+    if i_dms != o_dms or type(output_0)==BufferedWriter:
+        View.delete_view(o_dms, imported_view_name)
+        o_dms.close(force=True)
+    i_dms.close(force=True)
+
+    logger("info", "Done.")

python/obitools3/commands/test.pyx

@@ -23,6 +23,7 @@ from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
 import shutil
 import string
 import random
+import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
@@ -38,7 +39,7 @@ COL_COMMENTS_MAX_LEN = 2048
 MAX_INT = 2147483647    # used to generate random float values
  
  
-__title__="Tests if the obitools are working properly"
+__title__="Test if the obitools are working properly"
 
 
 default_config = {
@@ -300,8 +301,11 @@ def fill_column(config, infos, col) :
 def create_random_column(config, infos) :
     alias = random.choice([b'', random_unique_name(infos)])
     tuples = random.choice([True, False])
+    dict_column = False
     if not tuples :
         nb_elements_per_line=random.randint(1, config['test']['maxelts'])
+        if nb_elements_per_line > 1:
+            dict_column = True
         elements_names = []
         for i in range(nb_elements_per_line) :
             elements_names.append(random_unique_element_name(config, infos))
@@ -317,6 +321,7 @@ def create_random_column(config, infos) :
                                data_type, 
                                nb_elements_per_line=nb_elements_per_line,
                                elements_names=elements_names,
+                               dict_column=dict_column,
                                tuples=tuples,
                                comments=random_comments(config),
                                alias=alias
@@ -366,7 +371,7 @@ def random_new_view(config, infos, first=False):
             infos['view'] = View_NUC_SEQS.new(infos['dms'], random_unique_name(infos), comments=random_comments(config))   # TODO quality column
         else :
             infos['view'] = View.new(infos['dms'], random_unique_name(infos), comments=random_comments(config))   # TODO quality column
-        
+        infos['view'].write_config(config, "test", infos["command_line"], input_dms_name=[infos['dms'].name], input_view_name=["random"])
     print_test(config, repr(infos['view']))
     if v_to_clone is not None :
         if line_selection is None:
@@ -441,7 +446,7 @@ def addOptions(parser):
                        default=20,
                        type=int,
                        help="Maximum length of tuples. "
-                            "Default: 200")
+                            "Default: 20")
   
     group.add_argument('--max_ini_col_count','-o',
                        action="store", dest="test:maxinicolcount",
@@ -454,10 +459,10 @@ def addOptions(parser):
     group.add_argument('--max_line_nb','-l',
                        action="store", dest="test:maxlinenb",
                        metavar='<MAX_LINE_NB>',
-                       default=10000,
+                       default=1000,
                        type=int,
                        help="Maximum number of lines in a column. "
-                            "Default: 10000")
+                            "Default: 1000")
  
     group.add_argument('--max_elts_per_line','-e',
                        action="store", dest="test:maxelts",
@@ -497,7 +502,8 @@ def run(config):
                                     (b"OBI_SEQ", False): random_seq, (b"OBI_SEQ", True): random_seq_tuples,
                                     (b"OBI_STR", False): random_bytes, (b"OBI_STR", True): random_bytes_tuples
                                   },
-             'tests': [test_set_and_get, test_add_col, test_delete_col, test_col_alias, test_new_view]
+             'tests': [test_set_and_get, test_add_col, test_delete_col, test_col_alias, test_new_view],
+             'command_line': " ".join(sys.argv[1:])
             }
  
     # TODO ???

python/obitools3/commands/uniq.pyx

@@ -354,6 +354,9 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, di
         key = mergedKeys[k]
         merged_col_name = mergedKeys_m[k]
         
+#        if merged_infos[merged_col_name]['nb_elts'] == 1:
+#            raise Exception("Can't merge information from a tag with only one element (e.g. one sample ; don't use -m option)")
+        
         if merged_col_name in view:
             i_col = view[merged_col_name]
         else:
@@ -375,6 +378,7 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, di
                               OBI_INT,
                               nb_elements_per_line=merged_infos[merged_col_name]['nb_elts'],
                               elements_names=list(merged_infos[merged_col_name]['elt_names']),
+                              dict_column=True,
                               comments=i_col.comments,
                               alias=merged_col_name
                              )
@@ -397,6 +401,7 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, di
                               OBI_INT,
                               nb_elements_per_line=len(view),
                               elements_names=[id for id in i_id_col],
+                              dict_column=True,
                               alias=TAXID_DIST_COLUMN
                              )
 

python/obitools3/dms/capi/obidms.pxd

@@ -34,6 +34,7 @@ cdef extern from "obidms.h" nogil:
     int obi_close_dms(OBIDMS_p dms, bint force)
     char* obi_dms_get_dms_path(OBIDMS_p dms)
     char* obi_dms_get_full_path(OBIDMS_p dms, const_char_p path_name)
+    char* obi_dms_formatted_infos(OBIDMS_p dms, bint detailed)
     void obi_close_atexit()
     
     obiversion_t obi_import_column(const char* dms_path_1, const char* dms_path_2, const char* column_name, obiversion_t version_number)

python/obitools3/dms/capi/obidmscolumn.pxd

@@ -31,6 +31,7 @@ cdef extern from "obidmscolumn.h" nogil:
         const_char_p        elements_names
         OBIType_t           returned_data_type
         OBIType_t           stored_data_type
+        bint                dict_column
         bint                tuples
         bint                to_eval
         time_t              creation_date
@@ -63,10 +64,11 @@ cdef extern from "obidmscolumn.h" nogil:
     
     char* obi_get_elements_names(OBIDMS_column_p column)
     
-    char* obi_column_formatted_infos(OBIDMS_column_p column)
-
     index_t obi_column_get_element_index_from_name(OBIDMS_column_p column, const char* element_name)
     
     int obi_column_write_comments(OBIDMS_column_p column, const char* comments)
 
     int obi_column_add_comment(OBIDMS_column_p column, const char* key, const char* value)
+
+    char* obi_column_formatted_infos(OBIDMS_column_p column, bint detailed)
+    

python/obitools3/dms/capi/obiecopcr.pxd

@@ -8,6 +8,7 @@ cdef extern from "obi_ecopcr.h" nogil:
 
     int obi_ecopcr(const char* input_dms_name,
                    const char* i_view_name,
+                   const char* tax_dms_name,
                    const char* taxonomy_name,
                    const char* output_dms_name,
                    const char* o_view_name,

python/obitools3/dms/capi/obiecotag.pxd

@@ -11,4 +11,5 @@ cdef extern from "obi_ecotag.h" nogil:
                    const char* taxonomy_name,
                    const char* output_view_name,
                    const char* output_view_comments,
-                   double ecotag_threshold)
+                   double ecotag_threshold,
+                   double bubble_threshold)

python/obitools3/dms/capi/obitaxonomy.pxd

@@ -7,6 +7,8 @@ from libc.stdint   cimport int32_t
 
 cdef extern from "obidms_taxonomy.h" nogil:
 
+    extern int MIN_LOCAL_TAXID
+    
     struct ecotxnode :
         int32_t           taxid
         int32_t           rank
@@ -18,6 +20,13 @@ cdef extern from "obidms_taxonomy.h" nogil:
     ctypedef ecotxnode ecotx_t
 
 
+    struct econame_t : # can't get this struct to be accepted by Cython ('unknown size')
+        char*      name
+        char*      class_name
+        int32_t    is_scientific_name
+        ecotxnode* taxon
+
+
     struct ecotxidx_t :   
         int32_t  count
         int32_t  max_taxid
@@ -30,9 +39,14 @@ cdef extern from "obidms_taxonomy.h" nogil:
         char**   label
 
 
+    struct econameidx_t :   
+        int32_t    count
+        econame_t* names
+
+
     struct OBIDMS_taxonomy_t :
         ecorankidx_t* ranks
-#        econameidx_t* names
+        econameidx_t* names
         ecotxidx_t*   taxa
 
     ctypedef OBIDMS_taxonomy_t* OBIDMS_taxonomy_p
@@ -44,14 +58,18 @@ cdef extern from "obidms_taxonomy.h" nogil:
 
     OBIDMS_taxonomy_p obi_read_taxdump(const_char_p taxdump)
     
-    int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const_char_p tax_name)
+    int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const_char_p tax_name, bint update)
 
     int obi_close_taxonomy(OBIDMS_taxonomy_p taxonomy)
     
     ecotx_t* obi_taxo_get_parent_at_rank(ecotx_t* taxon, int32_t rankidx)
     
     ecotx_t* obi_taxo_get_taxon_with_taxid(OBIDMS_taxonomy_p taxonomy, int32_t taxid)
+
+    char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
     
+    ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
+
     bint obi_taxo_is_taxon_under_taxid(ecotx_t* taxon, int32_t other_taxid)
     
     ecotx_t* obi_taxo_get_species(ecotx_t* taxon, OBIDMS_taxonomy_p taxonomy)
@@ -71,4 +89,4 @@ cdef extern from "obidms_taxonomy.h" nogil:
     int obi_taxo_add_preferred_name_with_taxon(OBIDMS_taxonomy_p tax, ecotx_t* taxon, const char* preferred_name)
 
     const char* obi_taxo_rank_index_to_label(int32_t rank_idx, ecorankidx_t* ranks)
-    
+

python/obitools3/dms/capi/obitypes.pxd

@@ -53,6 +53,8 @@ cdef extern from "obitypes.h" nogil:
     extern const_char_p  OBIQual_char_NA
     extern uint8_t*      OBIQual_int_NA
     extern void*         OBITuple_NA
+    
+    extern obiint_t      OBI_INT_MAX
 
     const_char_p name_data_type(int data_type)
 

python/obitools3/dms/capi/obiview.pxd

@@ -27,6 +27,7 @@ cdef extern from "obiview.h" nogil:
     extern const_char_p REVERSE_QUALITY_COLUMN
     extern const_char_p REVERSE_SEQUENCE_COLUMN
     extern const_char_p COUNT_COLUMN
+    extern const_char_p SCIENTIFIC_NAME_COLUMN
     extern const_char_p TAXID_COLUMN
     extern const_char_p MERGED_TAXID_COLUMN
     extern const_char_p MERGED_PREFIX
@@ -94,6 +95,7 @@ cdef extern from "obiview.h" nogil:
                             index_t nb_elements_per_line,
                             char* elements_names,
                             bint  elt_names_formatted,
+                            bint dict_column,
                             bint tuples,
                             bint to_eval,
                             const_char_p indexer_name,
@@ -103,13 +105,17 @@ cdef extern from "obiview.h" nogil:
                             bint create)
 
     int obi_view_delete_column(Obiview_p view, const_char_p column_name, bint delete_file)
-            
+
     OBIDMS_column_p obi_view_get_column(Obiview_p view, const_char_p column_name)
 
     OBIDMS_column_p* obi_view_get_pointer_on_column_in_view(Obiview_p view, const_char_p column_name)
 
     int obi_view_create_column_alias(Obiview_p view, const_char_p current_name, const_char_p alias)    
-    
+
+    char* obi_view_formatted_infos(Obiview_p view, bint detailed)
+
+    char* obi_view_formatted_infos_one_line(Obiview_p view)
+
     int obi_view_write_comments(Obiview_p view, const_char_p comments)
 
     int obi_view_add_comment(Obiview_p view, const_char_p key, const_char_p value)

python/obitools3/dms/column/column.pyx

@@ -7,7 +7,8 @@ __OBIDMS_COLUMN_CLASS__ = {}
 from ..capi.obitypes cimport name_data_type, \
                              obitype_t, \
                              obiversion_t, \
-                             OBI_QUAL
+                             OBI_QUAL, \
+                             OBI_STR
 
 from ..capi.obidms cimport obi_import_column
 
@@ -40,7 +41,8 @@ from obitools3.utils cimport tobytes, \
 from obitools3.dms.column import typed_column
  
 from libc.stdlib cimport free
- 
+from libc.string cimport strcpy
+  
 import importlib
 import inspect
 import pkgutil
@@ -89,6 +91,7 @@ cdef class Column(OBIWrapper) :
                    obitype_t data_type,
                    index_t   nb_elements_per_line=1,
                    list      elements_names=None,
+                   bint      dict_column=False,
                    bint      tuples=False,
                    bint      to_eval=False,
                    object    associated_column_name=b"",
@@ -97,6 +100,7 @@ cdef class Column(OBIWrapper) :
                    object    alias=b""):
         # TODO indexer_name?
         
+        cdef Column column
         cdef bytes  column_name_b = tobytes(column_name)
         cdef bytes  alias_b = tobytes(alias)
         cdef bytes  comments_b = str2bytes(json.dumps(bytes2str_object(comments)))
@@ -125,6 +129,10 @@ cdef class Column(OBIWrapper) :
         else:
             elements_names_p = NULL
         
+        if column_name_b == b"SAMPLE" or column_name_b == b"sample":
+        # force str type
+            data_type = OBI_STR
+        
         if data_type == OBI_QUAL:
             if associated_column_name_b == b"":
                 if column_name == QUALITY_COLUMN:
@@ -132,13 +140,14 @@ cdef class Column(OBIWrapper) :
                          raise RuntimeError("Cannot create column %s in view %s: trying to create quality column but no NUC_SEQ column to associate it with in the view" % (bytes2str(column_name_b),
                                                                                bytes2str(view.name)))
                     associated_column_name_b = NUC_SEQUENCE_COLUMN
-                    associated_column_version = view[NUC_SEQUENCE_COLUMN].version
+                    associated_column_version = view[NUC_SEQUENCE_COLUMN].version                    
                 elif column_name == REVERSE_QUALITY_COLUMN:
                     if REVERSE_SEQUENCE_COLUMN not in view:
                          raise RuntimeError("Cannot create column %s in view %s: trying to create reverse quality column but no REVERSE_SEQUENCE column to associate it with in the view" % (bytes2str(column_name_b),
                                                                                bytes2str(view.name)))
                     associated_column_name_b = REVERSE_SEQUENCE_COLUMN
                     associated_column_version = view[REVERSE_SEQUENCE_COLUMN].version
+                    
         
         if (obi_view_add_column(view                      = view.pointer(),
                                 column_name               = column_name_b,
@@ -149,6 +158,7 @@ cdef class Column(OBIWrapper) :
                                 nb_elements_per_line      = nb_elements_per_line,
                                 elements_names            = elements_names_p,
                                 elt_names_formatted       = False,
+                                dict_column               = dict_column,
                                 tuples                    = tuples,
                                 to_eval                   = to_eval,
                                 indexer_name              = NULL,
@@ -158,8 +168,19 @@ cdef class Column(OBIWrapper) :
                                 create                    = True)<0):
             raise RuntimeError("Cannot create column %s in view %s" % (bytes2str(column_name_b),
                                                                        bytes2str(view.name)))
-
-        return Column.open(view, alias_b)
+        
+        column = Column.open(view, alias_b)
+        
+        # Automatically associate nuc sequence column to quality column if necessary
+        if data_type == OBI_QUAL:
+            if column_name == QUALITY_COLUMN:
+                view[NUC_SEQUENCE_COLUMN].associated_column_name = column.name
+                view[NUC_SEQUENCE_COLUMN].associated_column_version = column.version
+            elif column_name == REVERSE_QUALITY_COLUMN:
+                view[REVERSE_SEQUENCE_COLUMN].associated_column_name = column.name
+                view[REVERSE_SEQUENCE_COLUMN].associated_column_version = column.version
+        
+        return column
  
  
     @staticmethod
@@ -186,7 +207,7 @@ cdef class Column(OBIWrapper) :
         
         column_p = column_pp[0]
         column_type = column_p.header.returned_data_type
-        column_class = Column.get_column_class(column_type, (column_p.header.nb_elements_per_line > 1), column_p.header.tuples)
+        column_class = Column.get_column_class(column_type, (column_p.header.nb_elements_per_line > 1 or column_p.header.dict_column == True), column_p.header.tuples)
         column = OBIWrapper.new_wrapper(column_class, column_pp)
         
         column._view = view
@@ -222,6 +243,7 @@ cdef class Column(OBIWrapper) :
                                 nb_elements_per_line      = -1,
                                 elements_names            = NULL,
                                 elt_names_formatted       = False,
+                                dict_column               = False,
                                 tuples                    = False,
                                 to_eval                   = False,
                                 indexer_name              = NULL,
@@ -288,15 +310,24 @@ cdef class Column(OBIWrapper) :
     
     @OBIWrapper.checkIsActive
     def __repr__(self) :
-        cdef bytes s
-        #cdef char* s_b
-        #cdef str s_str
-        #s_b = obi_column_formatted_infos(self.pointer())
-        #s_str = bytes2str(s_b)
-        #free(s_b)
-        s = self._alias + b", data type: " + self.data_type
-        #return s_str
-        return bytes2str(s)
+        cdef str s
+        cdef char* sc
+        cdef OBIDMS_column_p pointer = self.pointer()
+        sc = obi_column_formatted_infos(pointer, False)
+        s = bytes2str(sc)
+        free(sc)
+        return s
+
+
+    @OBIWrapper.checkIsActive
+    def repr_longformat(self) :
+        cdef str s
+        cdef char* sc
+        cdef OBIDMS_column_p pointer = self.pointer()
+        sc = obi_column_formatted_infos(pointer, True)
+        s = bytes2str(sc)
+        free(sc)
+        return s
     
          
     def close(self):  # TODO discuss, can't be called bc then bug when closing view that tries to close it in C
@@ -351,6 +382,13 @@ cdef class Column(OBIWrapper) :
             raise OBIDeactivatedInstanceError()
         return self.pointer().header.nb_elements_per_line
  
+    # dict_column property getter
+    @property
+    def dict_column(self):
+        if not self.active() :
+            raise OBIDeactivatedInstanceError()
+        return self.pointer().header.dict_column
+ 
     # data_type property getter
     @property
     def data_type(self):
@@ -407,6 +445,31 @@ cdef class Column(OBIWrapper) :
             raise OBIDeactivatedInstanceError()
         return obi_format_date(self.pointer().header.creation_date)
 
+
+    # associated_column name property getter and setter
+    @property
+    def associated_column_name(self):
+        if not self.active() :
+            raise OBIDeactivatedInstanceError()
+        return self.pointer().header.associated_column.column_name
+
+    @associated_column_name.setter
+    def associated_column_name(self, object new_name):
+        strcpy(self.pointer().header.associated_column.column_name, tobytes(new_name))
+
+
+    # associated_column version property getter and setter
+    @property
+    def associated_column_version(self):
+        if not self.active() :
+            raise OBIDeactivatedInstanceError()
+        return self.pointer().header.associated_column.version
+
+    @associated_column_version.setter
+    def associated_column_version(self, int new_version):
+        self.pointer().header.associated_column.version = new_version
+
+
     # comments property getter
     @property
     def comments(self):

python/obitools3/dms/column/typed_column/bool.pyx

@@ -38,11 +38,13 @@ cdef class Column_bool(Column):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         return Column.new_column(view, column_name, OBI_BOOL,
                           nb_elements_per_line=nb_elements_per_line, 
-                          elements_names=elements_names, 
+                          elements_names=elements_names,
+                          dict_column=dict_column, 
                           tuples=tuples,
                           comments=comments)
                 

python/obitools3/dms/column/typed_column/char.pyx

@@ -36,12 +36,14 @@ cdef class Column_char(Column):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         
         return Column.new_column(view, column_name, OBI_CHAR,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=tuples,
                                  comments=comments)
 

python/obitools3/dms/column/typed_column/float.pyx

@@ -36,12 +36,14 @@ cdef class Column_float(Column):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         
         return Column.new_column(view, column_name, OBI_FLOAT,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=tuples,
                                  comments=comments)
 

python/obitools3/dms/column/typed_column/int.pyx

@@ -38,12 +38,14 @@ cdef class Column_int(Column):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         
         return Column.new_column(view, column_name, OBI_INT,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=tuples,
                                  comments=comments)
 

python/obitools3/dms/column/typed_column/qual.pyx

@@ -38,6 +38,7 @@ cdef class Column_qual(Column_idx):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             object  associated_column_name=b"",
             int     associated_column_version=-1,
             object  comments={}):
@@ -45,6 +46,7 @@ cdef class Column_qual(Column_idx):
         return Column.new_column(view, column_name, OBI_QUAL,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=False,
                                  associated_column_name=associated_column_name,
                                  associated_column_version=associated_column_name,

python/obitools3/dms/column/typed_column/seq.pyx

@@ -39,12 +39,14 @@ cdef class Column_seq(Column_idx):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         
         return Column.new_column(view, column_name, OBI_SEQ,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=tuples,
                                  comments=comments)
 

python/obitools3/dms/column/typed_column/str.pyx

@@ -38,12 +38,14 @@ cdef class Column_str(Column_idx):
             object  column_name,
             index_t nb_elements_per_line=1,
             object  elements_names=None,
+            bint    dict_column=False,
             bint    tuples=False,
             object  comments={}):
         
         return Column.new_column(view, column_name, OBI_STR,
                                  nb_elements_per_line=nb_elements_per_line, 
                                  elements_names=elements_names, 
+                                 dict_column=dict_column, 
                                  tuples=tuples,
                                  comments=comments)
 
@@ -72,6 +74,9 @@ cdef class Column_str(Column_idx):
         if value is None :
             value_b = <char*>OBIStr_NA
         else :
+            if self.name == b'sample' or self.name == b'SAMPLE':
+                if type(value) == int:
+                    value = str(value) # force sample ids to be str
             value_bytes = tobytes(value)
             value_b = <char*>value_bytes
 
@@ -135,6 +140,9 @@ cdef class Column_multi_elts_str(Column_multi_elts_idx):
         if value is None :
             value_b = <char*>OBIStr_NA
         else :
+            if self.name == b'sample' or self.name == b'SAMPLE':
+                if type(value) == int:
+                    value = str(value) # force sample ids to be str
             value_bytes = tobytes(value)
             value_b = <char*>value_bytes
                 
@@ -204,6 +212,9 @@ cdef class Column_tuples_str(Column_idx):
             i = 0
             for elt in value :
                 if elt is not None and elt != '':
+                    if self.name == b'sample' or self.name == b'SAMPLE':
+                        if type(elt) == int:
+                            elt = str(elt) # force sample ids to be str
                     elt_b = tobytes(elt)
                     strcpy(array+i, <char*>elt_b)
                     i = i + len(elt_b) + 1

python/obitools3/dms/dms.pyx

@@ -10,7 +10,8 @@ from .capi.obidms cimport obi_open_dms, \
                           obi_dms_exists, \
                           obi_dms_get_full_path, \
                           obi_close_atexit, \
-                          obi_dms_write_comments
+                          obi_dms_write_comments, \
+                          obi_dms_formatted_infos
                                                       
 from .capi.obitypes cimport const_char_p
                             
@@ -32,6 +33,8 @@ from .object import OBIWrapper
 import json
 import time
 
+from libc.stdlib cimport free
+
 
 cdef class DMS(OBIWrapper):    
     
@@ -223,13 +226,24 @@ cdef class DMS(OBIWrapper):
 
 
     @OBIWrapper.checkIsActive
-    def __repr__(self):
-        cdef str s 
-        s=""
-        for view_name in self.keys():
-            view = self.get_view(view_name)
-            s = s + repr(view) + "\n"
-            view.close()
+    def __repr__(self) :
+        cdef str s
+        cdef char* sc
+        cdef OBIDMS_p pointer = self.pointer()
+        sc = obi_dms_formatted_infos(pointer, False)
+        s = bytes2str(sc)
+        free(sc)
+        return s
+
+
+    @OBIWrapper.checkIsActive
+    def repr_longformat(self) :
+        cdef str s
+        cdef char* sc
+        cdef OBIDMS_p pointer = self.pointer()
+        sc = obi_dms_formatted_infos(pointer, True)
+        s = bytes2str(sc)
+        free(sc)
         return s
     
 

python/obitools3/dms/taxo/taxo.pxd

@@ -11,13 +11,16 @@ cdef class Taxonomy(OBIWrapper) :
     cdef bytes  _name
     cdef DMS    _dms
     cdef list   _ranks
+    cdef dict   _name_dict
     
     cdef inline OBIDMS_taxonomy_p pointer(self)
+    cdef fill_name_dict(self)
 
     cpdef Taxon get_taxon_by_idx(self, int idx)
     cpdef Taxon get_taxon_by_taxid(self, int taxid)
-    cpdef write(self, object prefix)
-    cpdef int add_taxon(self, str name, str rank_name, int parent_taxid, int min_taxid=*)
+    cpdef Taxon get_taxon_by_name(self, object taxon_name, object restricting_taxid=*)
+    cpdef write(self, object prefix, bint update=*)
+    cpdef int add_taxon(self, object name, object rank_name, int parent_taxid, int min_taxid=*)
     cpdef object get_species(self, int taxid)
     cpdef object get_genus(self, int taxid)
     cpdef object get_family(self, int taxid)

python/obitools3/dms/taxo/taxo.pyx

@@ -1,5 +1,7 @@
 #cython: language_level=3
 
+import sys
+
 from obitools3.utils cimport str2bytes, bytes2str, tobytes, tostr
 from ..capi.obidms cimport OBIDMS_p, obi_dms_get_full_path
                           
@@ -15,7 +17,11 @@ from ..capi.obitaxonomy cimport obi_taxonomy_exists, \
                                 obi_taxo_get_species, \
                                 obi_taxo_get_genus, \
                                 obi_taxo_get_family, \
-                                ecotx_t                
+                                ecotx_t, \
+                                econame_t, \
+                                obi_taxo_get_name_from_name_idx, \
+                                obi_taxo_get_taxon_from_name_idx
+                                
                                            
 from cpython.pycapsule cimport PyCapsule_New, PyCapsule_GetPointer
 import tarfile
@@ -24,11 +30,29 @@ from libc.stdlib  cimport free
 
 
 cdef class Taxonomy(OBIWrapper) :
-    # TODO function to import taxonomy?
-    
+    # TODO function to import taxonomy?    
+     
     cdef inline OBIDMS_taxonomy_p pointer(self) :
         return <OBIDMS_taxonomy_p>(self._pointer)        
 
+    cdef fill_name_dict(self):
+        print("Indexing taxon names...", file=sys.stderr)
+        
+        cdef OBIDMS_taxonomy_p pointer = self.pointer()
+        cdef ecotx_t*     taxon_p
+        cdef object       taxon_capsule
+        cdef bytes        name
+        cdef int          count
+        cdef int          n
+        
+        count = (<OBIDMS_taxonomy_p>pointer).names.count
+        
+        for n in range(count) :
+            name = obi_taxo_get_name_from_name_idx(pointer, n)
+            taxon_p = obi_taxo_get_taxon_from_name_idx(pointer, n)
+            taxon_capsule = PyCapsule_New(taxon_p, NULL, NULL)
+            self._name_dict[name] = Taxon(taxon_capsule, self)
+
 
     @staticmethod
     def exists(DMS dms, object name) :
@@ -69,13 +93,16 @@ cdef class Taxonomy(OBIWrapper) :
             raise RuntimeError("Error : Cannot read taxonomy %s" 
                                % tostr(name))
 
+        print("Taxonomy read", file=sys.stderr)
+
         taxo = OBIWrapper.new_wrapper(Taxonomy, pointer)
 
         dms.register(taxo)
 
         taxo._dms = dms
         taxo._name = tobytes(name)
-        
+        taxo._name_dict = {}
+        taxo.fill_name_dict()
         taxo._ranks = []
         for r in range((<OBIDMS_taxonomy_p>pointer).ranks.count) :
             taxo._ranks.append(obi_taxo_rank_index_to_label(r, (<OBIDMS_taxonomy_p>pointer).ranks))
@@ -118,19 +145,22 @@ cdef class Taxonomy(OBIWrapper) :
 
         taxo._dms = dms
         taxo._name = folder_path
-        
+        taxo._name_dict = {}
+        taxo.fill_name_dict()
         taxo._ranks = []
         for r in range((<OBIDMS_taxonomy_p>pointer).ranks.count) :
             taxo._ranks.append(obi_taxo_rank_index_to_label(r, (<OBIDMS_taxonomy_p>pointer).ranks))
-                
+        
+        print('Read %d taxa' % len(taxo), file=sys.stderr)
+        
         return taxo
 
     
     def __getitem__(self, object ref):     
         if type(ref) == int :
             return self.get_taxon_by_taxid(ref)
-        else :
-            raise NotImplementedError()
+        elif type(ref) == str or type(ref) == bytes :
+            return self.get_taxon_by_name(ref)
 
 
     cpdef Taxon get_taxon_by_taxid(self, int taxid):
@@ -143,6 +173,20 @@ cdef class Taxonomy(OBIWrapper) :
         return Taxon(taxon_capsule, self)
 
 
+    cpdef Taxon get_taxon_by_name(self, object taxon_name, object restricting_taxid=None):            
+        #print(taxon_name)
+        taxon = self._name_dict.get(tobytes(taxon_name), None)
+        if not taxon:
+            return None
+        elif restricting_taxid:
+            if self.is_ancestor(restricting_taxid, taxon.taxid):
+                return taxon
+            else:
+                return None
+        else:
+            return taxon
+
+
     cpdef Taxon get_taxon_by_idx(self, int idx):
         cdef ecotx_t* taxa
         cdef ecotx_t* taxon_p
@@ -232,19 +276,19 @@ cdef class Taxonomy(OBIWrapper) :
  
         taxa = self.pointer().taxa.taxon
  
-        # Yield each taxid
+        # Yield each taxon
         for t in range(self.pointer().taxa.count):
             taxon_p = <ecotx_t*> (taxa+t)
             taxon_capsule = PyCapsule_New(taxon_p, NULL, NULL)
             yield Taxon(taxon_capsule, self)
 
 
-    cpdef write(self, object prefix) :
-        if obi_write_taxonomy(self._dms.pointer(), self.pointer(), tobytes(prefix)) < 0 :
+    cpdef write(self, object prefix, bint update=False) :
+        if obi_write_taxonomy(self._dms.pointer(), self.pointer(), tobytes(prefix), update) < 0 :
             raise Exception("Error writing the taxonomy to binary files")
     
     
-    cpdef int add_taxon(self, str name, str rank_name, int parent_taxid, int min_taxid=10000000) :
+    cpdef int add_taxon(self, object name, object rank_name, int parent_taxid, int min_taxid=10000000) :
         cdef int taxid
         taxid = obi_taxo_add_local_taxon(self.pointer(), tobytes(name), tobytes(rank_name), parent_taxid, min_taxid)
         if taxid < 0 :
@@ -267,6 +311,11 @@ cdef class Taxonomy(OBIWrapper) :
     def name(self):
         return self._name
     
+    # ranks property getter
+    @property
+    def ranks(self):
+        return self._ranks
+
     
     def parental_tree_iterator(self, int taxid):
         """
@@ -281,6 +330,7 @@ cdef class Taxonomy(OBIWrapper) :
         if taxon is not None:
             while taxon.taxid != 1:
                 yield taxon
+                #print(taxon.taxid)
                 taxon = taxon.parent
             yield taxon
         else:

python/obitools3/dms/view/view.pyx

@@ -7,6 +7,7 @@ cdef dict __VIEW_CLASS__= {}
 from libc.stdlib  cimport malloc
 
 from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
+from obitools3.version import version
 
 from ..capi.obiview cimport Alias_column_pair_p, \
                             obi_new_view, \
@@ -18,7 +19,9 @@ from ..capi.obiview cimport Alias_column_pair_p, \
                             obi_view_delete_column, \
                             obi_view_create_column_alias, \
                             obi_view_write_comments, \
-                            obi_delete_view
+                            obi_delete_view, \
+                            obi_view_formatted_infos, \
+                            obi_view_formatted_infos_one_line
                            
 from ..capi.obidmscolumn cimport OBIDMS_column_p
 from ..capi.obidms cimport OBIDMS_p
@@ -58,6 +61,8 @@ import pkgutil
 import json
 import sys
 
+from libc.stdlib cimport free
+
 
 cdef class View(OBIWrapper) :
 
@@ -72,7 +77,7 @@ cdef class View(OBIWrapper) :
  
  
     @staticmethod
-    def import_view(object dms_1, object dms_2, object view_name_1, object view_name_2):
+    def import_view(object dms_1, object dms_2, object view_name_1, object view_name_2): # TODO argument to import line by line to avoid huge AVL copy
         if obi_import_view(tobytes(dms_1), tobytes(dms_2), tobytes(view_name_1), tobytes(view_name_2)) < 0 :
             raise Exception("Error importing a view")
 
@@ -183,11 +188,24 @@ cdef class View(OBIWrapper) :
     
     
     @OBIWrapper.checkIsActive
-    def __repr__(self) :         
-        cdef str s = "#View name:\n{name:s}\n#Line count:\n{line_count:d}\n#Columns:\n".format(name       = bytes2str(self.name),
-                                                                                                                         line_count = self.line_count)
-        for column_name in self.keys() :
-            s = s + repr(self[column_name]) + '\n'
+    def __repr__(self) :
+        cdef str s
+        cdef char* sc
+        cdef Obiview_p pointer = self.pointer()
+        sc = obi_view_formatted_infos(pointer, False)
+        s = bytes2str(sc)
+        free(sc)
+        return s
+
+
+    @OBIWrapper.checkIsActive
+    def repr_longformat(self) :
+        cdef str s
+        cdef char* sc
+        cdef Obiview_p pointer = self.pointer()
+        sc = obi_view_formatted_infos(pointer, True)
+        s = bytes2str(sc)
+        free(sc)
         return s
 
 
@@ -325,9 +343,9 @@ cdef class View(OBIWrapper) :
 
         new_column = Column.new_column(self, old_column.pointer().header.name, new_data_type, 
                                        nb_elements_per_line=new_nb_elements_per_line, elements_names=new_elements_names, 
-                                       comments=old_column.comments, alias=column_name_b+tobytes('___new___'))
+                                       dict_column=(new_nb_elements_per_line>1), comments=old_column.comments, alias=column_name_b+tobytes('___new___'))
         
-        switch_to_dict = old_column.nb_elements_per_line == 1 and new_nb_elements_per_line > 1
+        switch_to_dict = not old_column.dict_column and new_nb_elements_per_line > 1
         ori_key = old_column._elements_names[0]
         
         for i in range(length) :
@@ -386,6 +404,7 @@ cdef class View(OBIWrapper) :
                                   col.data_type_int, 
                                   nb_elements_per_line = col.nb_elements_per_line,
                                   elements_names = col._elements_names,
+                                  dict_column = col.dict_column,
                                   tuples = col.tuples,
                                   to_eval = col.to_eval,
                                   comments = col.comments,
@@ -434,6 +453,7 @@ cdef class View(OBIWrapper) :
             for i in range(len(input_view_name)):
                 input_str.append(tostr(input_dms_name[i])+"/"+tostr(input_view_name[i]))
         comments["input_str"] = input_str
+        comments["version"] = version
         return bytes2str_object(comments)
 
     
@@ -580,7 +600,8 @@ cdef class View(OBIWrapper) :
                 if element is not None:
                     if element.comments[b"input_dms_name"] is not None :
                         for i in range(len(element.comments[b"input_dms_name"])) :
-                            if element.comments[b"input_dms_name"][i] == element.dms.name and b"/" not in element.comments[b"input_view_name"][i]:  # Same DMS and not a special element like a taxonomy
+                            if b"/" not in element.comments[b"input_view_name"][i] and element.comments[b"input_view_name"][i] in element.dms \
+                            and element.comments[b"input_dms_name"][i] == element.dms.name :  # Same DMS and not a special element like a taxonomy and view was not deleted
                                 top_level.append(element.dms[element.comments[b"input_view_name"][i]])
                             else:
                                 top_level.append(None)
@@ -778,7 +799,8 @@ cdef class Line :
  
  
     def keys(self):
-        return self._view.keys()
+        cdef bytes key
+        return [key for key in self._view.keys()]
  
  
     def __contains__(self, object column_name):
@@ -786,7 +808,7 @@ cdef class Line :
 
 
     def __repr__(self):
-        return bytes2str(self).repr_bytes()
+        return bytes2str(self.repr_bytes())
 
 
     cpdef repr_bytes(self):

python/obitools3/format/fasta.pyx

@@ -7,11 +7,12 @@ from obitools3.utils cimport bytes2str
 
 cdef class FastaFormat:
 	
-	def __init__(self, list tags=[], bint printNAKeys=False, bytes NAString=b"NA"):
+	def __init__(self, list tags=[], bint printNAKeys=False, bytes NAString=b"NA", bint NAIntTo0=False):
 		self.headerFormatter = HeaderFormat("fasta",
 										    tags=tags,
 										    printNAKeys=printNAKeys,
-										    NAString=NAString)
+										    NAString=NAString,
+										    NAIntTo0=NAIntTo0)
 		
 	@cython.boundscheck(False)	
 	def __call__(self, object data):

python/obitools3/format/fastq.pyx

@@ -8,11 +8,12 @@ from obitools3.utils cimport bytes2str, str2bytes, tobytes
 # TODO quality offset option?
 cdef class FastqFormat:
 	
-	def __init__(self, list tags=[], bint printNAKeys=False, bytes NAString=b"NA"):
+	def __init__(self, list tags=[], bint printNAKeys=False, bytes NAString=b"NA", bint NAIntTo0=False):
 		self.headerFormatter = HeaderFormat("fastq",
 										    tags=tags,
 										    printNAKeys=printNAKeys,
-										    NAString=NAString)
+										    NAString=NAString,
+                                            NAIntTo0=NAIntTo0)
 		
 	@cython.boundscheck(False)	
 	def __call__(self, object data):

python/obitools3/format/header.pxd

@@ -4,5 +4,6 @@ cdef class HeaderFormat:
     cdef set    tags
     cdef bint   printNAKeys
     cdef bytes  NAString
+    cdef bint   NAIntTo0
     cdef size_t headerBufferLength
     

python/obitools3/format/header.pyx

@@ -8,13 +8,14 @@ from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
 
 from obitools3.utils cimport str2bytes, bytes2str_object
 from obitools3.dms.column.column cimport Column_line
+from obitools3.dms.column.typed_column.int cimport Column_int, Column_multi_elts_int
 
 
 cdef class HeaderFormat:
     
     SPECIAL_KEYS = [NUC_SEQUENCE_COLUMN, ID_COLUMN, DEFINITION_COLUMN, QUALITY_COLUMN]
     
-    def __init__(self, str format="fasta", list tags=[], bint printNAKeys=False, bytes NAString=b"NA"):
+    def __init__(self, str format="fasta", list tags=[], bint printNAKeys=False, bytes NAString=b"NA", bint NAIntTo0=False):
         '''
             @param format: 
             @type  format: `str`
@@ -32,6 +33,7 @@ cdef class HeaderFormat:
         self.tags           = set(tags)
         self.printNAKeys    = printNAKeys
         self.NAString       = NAString
+        self.NAIntTo0       = NAIntTo0
         
         if format=="fasta":
             self.start=b">"
@@ -57,17 +59,25 @@ cdef class HeaderFormat:
             if k in tags:
                 value = data[k]
                 if value is None or (isinstance(value, Column_line) and value.is_NA()):
-                    if self.printNAKeys:
+                    if isinstance(data.view[k], Column_int) and self.NAIntTo0: # people want missing int values to be 0
+                        value = b'0'
+                    elif self.printNAKeys:
                         value = self.NAString
                     else:
                         value = None
                 else:
                     if type(value) == Column_line:
-                        value = value.bytes()
+                        if isinstance(data.view[k], Column_multi_elts_int) and self.NAIntTo0:
+                            value = dict(value)
+                            for key in data.view[k].keys():
+                                if key not in value or value[key]:
+                                    value[key] = 0
+                        else:
+                            value = value.bytes()
                     else:
                         if type(value) == tuple:
                             value=list(value)
-                        value = str2bytes(str(bytes2str_object(value))) # genius programming
+                    value = str2bytes(str(bytes2str_object(value))) # genius programming
                 if value is not None:
                     lines.append(k + b"=" + value + b";")   
                 

python/obitools3/format/tab.pxd

@@ -4,5 +4,8 @@ cdef class TabFormat:
     cdef bint header
     cdef bint first_line
     cdef bytes NAString
-    cdef list tags
-    cdef bytes sep
+    cdef set   tags
+    cdef bytes sep
+    cdef bint NAIntTo0
+    cdef bint metabaR
+    cdef bint ngsfilter

python/obitools3/format/tab.pyx

@@ -4,51 +4,87 @@ cimport cython
 from obitools3.dms.view.view cimport Line
 from obitools3.utils cimport bytes2str_object, str2bytes, tobytes
 from obitools3.dms.column.column cimport Column_line, Column_multi_elts
+from obitools3.dms.column.typed_column.int cimport Column_int, Column_multi_elts_int
 
+import sys
 
 cdef class TabFormat:
     
-    def __init__(self, header=True, bytes NAString=b"NA", bytes sep=b"\t"):
+    def __init__(self, list tags=[], header=True, bytes NAString=b"NA", bytes sep=b"\t", bint NAIntTo0=True, metabaR=False, ngsfilter=False):
+        self.tags = set(tags)
         self.header = header
         self.first_line = True
         self.NAString = NAString
         self.sep = sep
+        self.NAIntTo0 = NAIntTo0
+        self.metabaR = metabaR
+        self.ngsfilter = ngsfilter
         
     @cython.boundscheck(False)    
     def __call__(self, object data):
         
+        cdef set ktags
+        cdef list tags = [key for key in data]
+        
         line = []
         
-        if self.first_line:
-            self.tags = [k for k in data.keys()]
-        
-        for k in self.tags:
-            
-            if self.header and self.first_line:
-                if isinstance(data.view[k], Column_multi_elts):
-                    for k2 in data.view[k].keys():
-                        line.append(tobytes(k)+b':'+tobytes(k2))
-                else:
-                    line.append(tobytes(k))
-            else:
+        if self.tags is not None and self.tags:
+            ktags = self.tags
+        else:
+            ktags = set(tags)            
+                
+        if self.header and self.first_line:
+            for k in ktags:
+                if k in tags:
+                    if self.metabaR:
+                        if k == b'NUC_SEQ':
+                            ktoprint = b'sequence'
+                        else:
+                            ktoprint = k.lower()
+                        ktoprint = ktoprint.replace(b'merged_', b'')
+                    else:
+                        ktoprint = k
+                    if isinstance(data.view[k], Column_multi_elts):
+                        keys = data.view[k].keys()
+                        keys.sort()
+                        for k2 in keys:
+                            line.append(tobytes(ktoprint)+b':'+tobytes(k2))
+                    else:
+                        line.append(tobytes(ktoprint))
+            r = self.sep.join(value for value in line)
+            r += b'\n'
+            line = []
+                    
+        for k in ktags:
+            if k in tags:
                 value = data[k]
                 if isinstance(data.view[k], Column_multi_elts):
+                    keys = data.view[k].keys()
+                    keys.sort()
                     if value is None:  # all keys at None
-                        for k2 in data.view[k].keys(): # TODO could be much more efficient
+                        for k2 in keys: # TODO could be much more efficient
                             line.append(self.NAString)
                     else:
-                        for k2 in data.view[k].keys(): # TODO could be much more efficient
+                        for k2 in keys: # TODO could be much more efficient
                             if value[k2] is not None:
                                 line.append(str2bytes(str(bytes2str_object(value[k2]))))  # genius programming
                             else:
-                                line.append(self.NAString)
+                                if self.NAIntTo0 and isinstance(data.view[k], Column_multi_elts_int):
+                                    line.append(b"0")
+                                else:
+                                    line.append(self.NAString)
                 else:
-                    if value is not None:
+                    if value is not None or (self.NAIntTo0 and isinstance(data.view[k], Column_int)):
                         line.append(str2bytes(str(bytes2str_object(value))))
                     else:
                         line.append(self.NAString)
-                  	
+                  	      	
+        if self.header and self.first_line:
+            r += self.sep.join(value for value in line)
+        else:
+            r = self.sep.join(value for value in line)
+
         if self.first_line:
             self.first_line = False
-      		
-        return self.sep.join(value for value in line)
+
+        return r

python/obitools3/libalign/_solexapairend.pyx

@@ -259,7 +259,7 @@ def buildJoinedSequence(ali, reverse, seq, forward=None):
     seq[b"pairedend_limit"]=len(forward)
     seq[b"seq_length"] = ali.consensus_len
     seq[b"overlap_length"] = ali.overlap_len
-    if ali.consensus_len > 0:
+    if ali.overlap_len > 0:
         seq[b'score_norm']=round(float(ali.score)/ali.overlap_len, 3)
     else:
         seq[b"score_norm"]=0.0

python/obitools3/parsers/fastq.pyx

@@ -103,7 +103,11 @@ def fastqWithQualityIterator(lineiterator,
         yield seq
 
         read+=1
-        hline = next(i)
+        try:
+            hline = next(i)
+        except StopIteration:
+            return
+
     
 
 def fastqWithoutQualityIterator(lineiterator, 
@@ -174,5 +178,7 @@ def fastqWithoutQualityIterator(lineiterator,
         yield seq
     
         read+=1
-        hline = next(i)
-        
+        try:
+            hline = next(i)
+        except StopIteration:
+            return        

python/obitools3/parsers/genbank.pyx

@@ -22,11 +22,11 @@ from libc.stdlib  cimport free, malloc, realloc
 from libc.string cimport strcpy, strlen
 
 
-_featureMatcher = re.compile(b'^FEATURES.+\n(?=ORIGIN)',re.DOTALL + re.M)
+_featureMatcher = re.compile(b'^FEATURES.+\n(?=ORIGIN(\s*))',re.DOTALL + re.M)
 
 _headerMatcher = re.compile(b'^LOCUS.+(?=\nFEATURES)', re.DOTALL + re.M)
-_seqMatcher    = re.compile(b'ORIGIN.+(?=//\n)', re.DOTALL + re.M)
-_cleanSeq1     = re.compile(b'ORIGIN.+\n')
+_seqMatcher    = re.compile(b'^ORIGIN.+(?=//\n)', re.DOTALL + re.M)
+_cleanSeq1     = re.compile(b'ORIGIN(\s*)\n')
 _cleanSeq2     = re.compile(b'[ \n0-9]+')
 _acMatcher     = re.compile(b'(?<=^ACCESSION   ).+',re.M)
 _deMatcher     = re.compile(b'(?<=^DEFINITION  ).+\n( .+\n)*',re.M)
@@ -155,10 +155,10 @@ def genbankIterator_file(lineiterator,
         yield seq
         read+=1
     
-    # Last sequence
-    seq = genbankParser(entry)
-    
-    yield seq
+    # Last sequence if not empty lines
+    if entry.strip():
+        seq = genbankParser(entry)
+        yield seq
     
     free(entry)
 

python/obitools3/parsers/ngsfilter.pyx

@@ -48,24 +48,25 @@ def ngsfilterIterator(lineiterator,
         all_lines.insert(0, firstline)
         
     # Insert header for column names
-    column_names = [b"experiment", b"sample", b"forward_tag", b"reverse_tag", b"forward_primer", b"reverse_primer"]
+    column_names = [b"experiment", b"sample", b"forward_tag", b"reverse_tag", b"forward_primer", b"reverse_primer"] #,b"additional_info"]
     header = out_sep.join(column_names)
     
     new_lines.append(header)
         
     for line in all_lines:
-        split_line = line.split()
-        tags = split_line.pop(2)
-        tags = tags.split(b":")
-        for t_idx in range(len(tags)):
-            if tags[t_idx]==b"-" or tags[t_idx]==b"None" or tags[t_idx]==b"":
-                tags[t_idx] = nastring
-        if len(tags) == 1:          # Forward and reverse tags are the same
-            tags.append(tags[0])
-        split_line.insert(2, tags[0])
-        split_line.insert(3, tags[1])
-        new_lines.append(out_sep.join(split_line[0:6]))
-    
+        split_line = line.split(maxsplit=5)
+        if split_line:
+            tags = split_line.pop(2)
+            tags = tags.split(b":")
+            for t_idx in range(len(tags)):
+                if tags[t_idx]==b"-" or tags[t_idx]==b"None" or tags[t_idx]==b"":
+                    tags[t_idx] = nastring
+            if len(tags) == 1:          # Forward and reverse tags are the same
+                tags.append(tags[0])
+            split_line.insert(2, tags[0])
+            split_line.insert(3, tags[1])
+            new_lines.append(out_sep.join(split_line[0:6]))
+        
     return tabIterator(iter(new_lines),
                        header = True,
                        sep = out_sep,

python/obitools3/parsers/tab.pyx

@@ -8,7 +8,7 @@ Created on feb 20th 2018
 
 import types
 from obitools3.utils cimport __etag__
-
+from obitools3.utils cimport str2bytes
 
 def tabIterator(lineiterator, 
                 bint header = False,
@@ -75,7 +75,7 @@ def tabIterator(lineiterator,
                 continue
             else:
                 # TODO ??? default column names? like R?
-                keys = [i for i in range(len(line.split(sep)))]
+                keys = [str2bytes(str(i)) for i in range(len(line.split(sep)))]
                 
         while skipped < skip :
             line = next(iterator)
@@ -99,7 +99,10 @@ def tabIterator(lineiterator,
         
         read+=1
         
-        line = next(iterator)
+        try:
+            line = next(iterator)
+        except StopIteration:
+            return 
     
         
         

python/obitools3/parsers/universal.pyx

@@ -53,7 +53,11 @@ def entryIteratorFactory(lineiterator,
 
     i = iterator
         
-    first=next(i)    
+    try:
+        first=next(i)
+    except StopIteration:
+        first=""
+        pass
 
     format=b"tabular"
     

python/obitools3/uri/decode.pyx

@@ -173,7 +173,10 @@ def open_uri(uri,
              type newviewtype=View,
              dms_only=False,
              force_file=False):
-    
+
+    if type(uri) == str and not uri.isascii():
+        raise Exception("Paths must be ASCII characters only")
+            
     cdef bytes urib = tobytes(uri)
     cdef bytes scheme
     cdef tuple dms
@@ -192,7 +195,7 @@ def open_uri(uri,
     
     config = getConfiguration()
     urip = urlparse(urib)
-            
+        
     if 'obi' not in config:
         config['obi']={}
     
@@ -209,13 +212,14 @@ def open_uri(uri,
     scheme = urip.scheme
     
     error = None
-    
-    if urib != b"-" and \
+        
+    if b'/taxonomy/' in urib or \
+        (urib != b"-" and \
         (scheme==b"dms" or \
          (scheme==b"" and \
           (((not input) and "outputformat" not in config["obi"]) or \
-          (input and "inputformat" not in config["obi"])))):   # TODO maybe not best way
-        
+          (input and "inputformat" not in config["obi"]))))):   # TODO maybe not best way
+                
         if default_dms is not None and b"/" not in urip.path:   # assuming view to open in default DMS (TODO discuss)
             dms=(default_dms, urip.path)
         else:
@@ -223,7 +227,7 @@ def open_uri(uri,
         
         if dms is None and default_dms is not None:
             dms=(default_dms, urip.path)
-
+                
         if dms is not None:
             if dms_only:
                 return (dms[0],
@@ -248,7 +252,7 @@ def open_uri(uri,
                 
                 if default_dms is None:
                     config["obi"]["defaultdms"]=resource[0]
-                                
+                
                 return (resource[0],
                         resource[1],
                         type(resource[1]),
@@ -276,11 +280,16 @@ def open_uri(uri,
             iseq = urib
             objclass = bytes
     else:  # TODO update uopen to be able to write? 
-        if urip.path == b'-':
+        if config['obi']['outputformat'] == b'metabaR':
+            if 'metabarprefix' not in config['obi']:
+                raise Exception("Prefix needed when exporting for metabaR (--metabaR-prefix option)")
+            else:
+                file = open(config['obi']['metabarprefix']+'.tab', 'wb')
+        elif not urip.path or urip.path == b'-':
             file = sys.stdout.buffer
-        elif urip.path :
+        else:
             file = open(urip.path, 'wb')
-        
+
     if file is not None:
         qualifiers=parse_qs(urip.query)
         
@@ -298,11 +307,11 @@ def open_uri(uri,
                     format=config["obi"][formatkey]
                 except KeyError:
                     format=None
-                
+                       
         if b'seqtype' in qualifiers:
             seqtype=qualifiers[b'seqtype'][0]
         else:
-            if format == b"ngsfilter" or format == b"tabular":  # TODO discuss
+            if format == b"ngsfilter" or format == b"tabular" or format == b"metabaR":  # TODO discuss
                 seqtype=None
             else:
                 try:
@@ -386,10 +395,10 @@ def open_uri(uri,
             raise MalformedURIException('Malformed header argument in URI')
        
         if b"sep" in qualifiers:
-            sep=tobytes(qualifiers[b"sep"][0][0])
+            sep = tobytes(qualifiers[b"sep"][0][0])
         else:
             try:
-                sep=tobytes(config["obi"]["sep"])
+                sep = tobytes(config["obi"]["sep"])
             except KeyError:
                 sep=None
         
@@ -426,7 +435,21 @@ def open_uri(uri,
                 nastring=tobytes(config["obi"][nakey])
             except KeyError:
                 nastring=b'NA'
-                
+
+        if b"na_int_to_0" in qualifiers:
+            try:
+                na_int_to_0=eval(qualifiers[b"na_int_to_0"][0])
+            except Exception as e:
+                raise MalformedURIException("Malformed 'NA_int_to_0' argument in URI")
+        else:
+            try:
+                na_int_to_0=config["obi"]["na_int_to_0"]
+            except KeyError:
+                if format==b"tabular" or format==b"metabaR":
+                    na_int_to_0=True
+                else:
+                    na_int_to_0=False
+                       
         if b"stripwhite" in qualifiers:
             try:
                 stripwhite=eval(qualifiers[b"stripwhite"][0])
@@ -461,17 +484,36 @@ def open_uri(uri,
             except KeyError:
                 commentchar=b'#'
 
+        if b"only_keys" in qualifiers:
+            only_keys=qualifiers[b"only_keys"][0] # not sure that works but no one ever uses qualifiers
+        else:
+            try:
+                only_keys_str=config["obi"]["only_keys"]
+                only_keys=[]
+                for key in only_keys_str:
+                    only_keys.append(tobytes(key))
+            except KeyError:
+                only_keys=[]
+
+        if b"metabaR_prefix" in qualifiers:
+            metabaR_prefix = tobytes(qualifiers[b"metabaR_prefix"][0][0])
+        else:
+            try:
+                metabaR_prefix = tobytes(config["obi"]["metabarprefix"])
+            except KeyError:
+                metabaR_prefix=None
+
         if format is not None:
             if seqtype==b"nuc":
                 objclass = Nuc_Seq    # Nuc_Seq_Stored? TODO
-                if format==b"fasta":
+                if format==b"fasta" or format==b"silva" or format==b"rdp" or format == b"unite" or format == b"sintax":
                     if input:
                         iseq = fastaNucIterator(file, 
                                                 skip=skip, 
                                                 only=only,
                                                 nastring=nastring)
                     else:
-                        iseq = FastaNucWriter(FastaFormat(printNAKeys=printna, NAString=nastring), 
+                        iseq = FastaNucWriter(FastaFormat(tags=only_keys, printNAKeys=printna, NAString=nastring), 
                                               file,
                                               skip=skip,
                                               only=only)
@@ -484,7 +526,7 @@ def open_uri(uri,
                                              noquality=noquality,
                                              nastring=nastring)
                     else:
-                        iseq = FastqWriter(FastqFormat(printNAKeys=printna, NAString=nastring), 
+                        iseq = FastqWriter(FastqFormat(tags=only_keys, printNAKeys=printna, NAString=nastring), 
                                            file,
                                            skip=skip,
                                            only=only)
@@ -520,7 +562,17 @@ def open_uri(uri,
                                        skip = skip,
                                        only = only)
                 else:
-                    iseq = TabWriter(TabFormat(header=header, NAString=nastring, sep=sep), 
+                    iseq = TabWriter(TabFormat(tags=only_keys, header=header, NAString=nastring, sep=sep, NAIntTo0=na_int_to_0), 
+                                               file,
+                                               skip=skip,
+                                               only=only,
+                                               header=header)
+            elif format==b"metabaR":
+                objclass = dict
+                if input:
+                    raise NotImplementedError('Input data file format not implemented')
+                else:
+                    iseq = TabWriter(TabFormat(tags=only_keys, header=header, NAString=nastring, sep=sep, NAIntTo0=na_int_to_0, metabaR=True), 
                                                file,
                                                skip=skip,
                                                only=only,
@@ -538,7 +590,7 @@ def open_uri(uri,
                                              skip = skip,
                                              only = only)
                 else:
-                    raise NotImplementedError('Output sequence file format not implemented')
+                    raise NotImplementedError('Output data file format not implemented')
         else:
             if input:
                 iseq, objclass, format = entryIteratorFactory(file,
@@ -556,7 +608,7 @@ def open_uri(uri,
                                                               commentchar)
             else:    # default export is in fasta? or tab? TODO
                 objclass = Nuc_Seq   # Nuc_Seq_Stored? TODO
-                iseq = FastaNucWriter(FastaFormat(printNAKeys=printna, NAString=nastring), 
+                iseq = FastaNucWriter(FastaFormat(tags=only_keys, printNAKeys=printna, NAString=nastring), 
                                       file,
                                       skip=skip,
                                       only=only)
@@ -565,6 +617,6 @@ def open_uri(uri,
         
         entry_count = -1
         if input:
-            entry_count = count_entries(file, format)
+            entry_count = count_entries(file, format, header)
         
         return (file, iseq, objclass, urib, entry_count)

python/obitools3/utils.pxd

@@ -2,8 +2,8 @@
 
 from obitools3.dms.capi.obitypes cimport obitype_t, index_t
 
-cpdef bytes format_separator(bytes format)
-cpdef int count_entries(file, bytes format)
+cpdef bytes format_uniq_pattern(bytes format)
+cpdef int count_entries(file, bytes format, bint header)
 
 cdef obi_errno_to_exception(index_t line_nb=*, object elt_id=*, str error_message=*)
 
@@ -18,7 +18,7 @@ cdef object clean_empty_values_from_object(object value, exclude=*)
 
 cdef obitype_t get_obitype_single_value(object value)
 cdef obitype_t update_obitype(obitype_t obitype, object new_value)
-cdef obitype_t get_obitype_iterable_value(object value)
+cdef obitype_t get_obitype_iterable_value(object value, type t)
 cdef obitype_t get_obitype(object value)
 
 cdef object __etag__(bytes x, bytes nastring=*)

python/obitools3/utils.pyx

@@ -9,7 +9,8 @@ from obitools3.dms.capi.obitypes cimport is_a_DNA_seq, \
                                          OBI_QUAL, \
                                          OBI_SEQ, \
                                          OBI_STR, \
-                                         index_t
+                                         index_t, \
+                                         OBI_INT_MAX
 
 from obitools3.dms.capi.obierrno cimport OBI_LINE_IDX_ERROR, \
                                          OBI_ELT_IDX_ERROR, \
@@ -24,11 +25,11 @@ import glob
 import gzip
 
 
-cpdef bytes format_separator(bytes format):
+cpdef bytes format_uniq_pattern(bytes format):
     if format == b"fasta":
         return b"\n>"
     elif format == b"fastq":
-        return b"\n@"
+        return b"\n\+\n"
     elif format == b"ngsfilter" or format == b"tabular":
         return b"\n"
     elif format == b"genbank" or format == b"embl":
@@ -39,10 +40,10 @@ cpdef bytes format_separator(bytes format):
         return None
 
 
-cpdef int count_entries(file, bytes format):
+cpdef int count_entries(file, bytes format, bint header):
     
     try:
-        sep = format_separator(format)
+        sep = format_uniq_pattern(format)
         if sep is None:
             return -1
         sep = re.compile(sep)
@@ -72,8 +73,10 @@ cpdef int count_entries(file, bytes format):
                 return -1
             mmapped_file = mmap.mmap(f.fileno(), 0, access=mmap.ACCESS_READ)
             total_count += len(re.findall(sep, mmapped_file))
-            if format != b"ngsfilter" and format != b"tabular" and format != b"embl" and format != b"genbank":
+            if format != b"ngsfilter" and format != b"tabular" and format != b"embl" and format != b"genbank" and format != b"fastq":
                 total_count += 1 # adding +1 for 1st entry because separators include \n (ngsfilter and tabular already count one more because of last \n)
+            if format == b"tabular" and header: # not counting header as an entry
+                total_count -= 1
             
     except:
         if len(files) > 1:
@@ -257,38 +260,51 @@ cdef obitype_t update_obitype(obitype_t obitype, object new_value) :
     
     new_type = type(new_value)
     
-    if obitype == OBI_INT :
-        if new_type == float :
-            return OBI_FLOAT
-    # TODO BOOL vers INT/FLOAT
-    elif new_type == str or new_type == bytes :
+    #if new_type == NoneType:  # doesn't work because Cython sucks
+    if new_value == None or new_type==list or new_type==dict or new_type==tuple:
+        return obitype
+    
+    # TODO BOOL to INT/FLOAT
+    if new_type == str or new_type == bytes :
         if obitype == OBI_SEQ and is_a_DNA_seq(tobytes(new_value)) :
             pass
         else :
             return OBI_STR
-    
+    elif obitype == OBI_INT :
+        if new_type == float or new_value > OBI_INT_MAX :
+            return OBI_FLOAT
+        
     return obitype
 
 
-cdef obitype_t get_obitype_iterable_value(object value) :
+cdef obitype_t get_obitype_iterable_value(object value, type t) :
     
     cdef obitype_t value_obitype
     
     value_obitype = OBI_VOID
     
-    for k in value :
-        if value_obitype == OBI_VOID :
-            value_obitype = get_obitype_single_value(value[k])
-        else :
-            value_obitype = update_obitype(value_obitype, value[k])
+    if t == dict:
+        for k in value :
+            if value_obitype == OBI_VOID :
+                value_obitype = get_obitype_single_value(value[k])
+            else :
+                value_obitype = update_obitype(value_obitype, value[k])
+    
+    elif t == list or t == tuple:
+        for v in value :
+            if value_obitype == OBI_VOID :
+                value_obitype = get_obitype_single_value(v)
+            else :
+                value_obitype = update_obitype(value_obitype, v)
     
     return value_obitype
 
 
 cdef obitype_t get_obitype(object value) :
     
-    if type(value) == dict or type(value) == list or type(value) == tuple :
-        return get_obitype_iterable_value(value)
+    t = type(value)
+    if t == dict or t == list or t == tuple :
+        return get_obitype_iterable_value(value, t)
     
     else :
         return get_obitype_single_value(value)

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '0b28'
+serial= '1b25'
 
 version ="%d.%d.%s" % (major,minor,serial)

python/obitools3/writers/tab.pyx

@@ -20,8 +20,6 @@ cdef class TabWriter:
             self.only = -1
         else:
             self.only = int(only)
-            if header:
-                self.only += 1
 
         self.formatter = formatter
         self.output = output_object

src/.gitignore

@@ -3,3 +3,4 @@
 /cmake_install.cmake
 /libcobitools3.dylib
 /Makefile
+/build/

src/build_reference_db.c

@@ -77,6 +77,7 @@ static inline ecotx_t* get_lca_from_merged_taxids(Obiview_p view, OBIDMS_column_
 {
 	ecotx_t* taxon = NULL;
 	ecotx_t* lca = NULL;
+	ecotx_t* lca1 = NULL;
 	int32_t taxid;
 	index_t taxid_idx;
 	int64_t taxid_str_idx;
@@ -108,10 +109,11 @@ static inline ecotx_t* get_lca_from_merged_taxids(Obiview_p view, OBIDMS_column_
 			else
 			{
 				// Compute LCA
+				lca1 = lca;
 				lca = obi_taxo_get_lca(taxon, lca);
 				if (lca == NULL)
 				{
-					obidebug(1, "\nError getting the last common ancestor of two taxa when building a reference database");
+					obidebug(1, "\nError getting the last common ancestor of two taxa when building a reference database, %d %d", taxid, lca1->taxid);
 					return NULL;
 				}
 			}
@@ -185,7 +187,7 @@ int build_reference_db(const char* dms_name,
 	matrix_view_name = strcpy(matrix_view_name, o_view_name);
 	strcat(matrix_view_name, "_matrix");
 
-	fprintf(stderr, "Aligning queries with reference database...\n");
+	fprintf(stderr, "Aligning sequences...\n");
 	if (obi_lcs_align_one_column(dms_name,
 								 refs_view_name,
 								 "",
@@ -243,6 +245,7 @@ int build_reference_db(const char* dms_name,
 							false,
 							false,
 							false,
+							false,
 							"",
 							"",
 							-1,
@@ -392,6 +395,7 @@ int build_reference_db(const char* dms_name,
 							1,
 							"",
 							false,
+							false,
 							true,
 							false,
 							"",
@@ -415,6 +419,7 @@ int build_reference_db(const char* dms_name,
 							1,
 							"",
 							false,
+							false,
 							true,
 							false,
 							"",
@@ -860,7 +865,8 @@ int build_reference_db(const char* dms_name,
 	fprintf(stderr,"\rDone : 100 %%           \n");
 
 	// Add information about the threshold used to build the DB
-	snprintf(threshold_str, 5, "%f", threshold);
+#define snprintf_nowarn(...) (snprintf(__VA_ARGS__) < 0 ? abort() : (void)0)
+	snprintf_nowarn(threshold_str, 5, "%f", threshold);
 
 	new_comments = obi_add_comment((o_view->infos)->comments, DB_THRESHOLD_KEY_IN_COMMENTS, threshold_str);
 	if (new_comments == NULL)

src/encode.c

@@ -36,10 +36,12 @@ bool only_ATGC(const char* seq)
 	{
 		if (!((*c == 'A') || \
 			  (*c == 'T') || \
+			  (*c == 'U') || \
 			  (*c == 'G') || \
 			  (*c == 'C') || \
 			  (*c == 'a') || \
 			  (*c == 't') || \
+			  (*c == 'u') || \
 			  (*c == 'g') || \
 			  (*c == 'c')))
 		{
@@ -182,6 +184,8 @@ byte_t* encode_seq_on_2_bits(const char* seq, int32_t length)
 			break;
 		case 't':
 		case 'T':
+		case 'u':
+		case 'U':
 			seq_b[i/4] |= NUC_T_2b;
 			break;
 		default:
@@ -288,6 +292,8 @@ byte_t* encode_seq_on_4_bits(const char* seq, int32_t length)
 			break;
 		case 't':
 		case 'T':
+		case 'u': // discussable
+		case 'U':
 			seq_b[i/2] |= NUC_T_4b;
 			break;
 		case 'r':

src/encode.h

@@ -64,7 +64,7 @@ enum
 
 
 /**
- * @brief Checks if there are only 'atgcATGC' characters in a
+ * @brief Checks if there are only 'atgcuATGCU' characters in a
  *        character string.
  *
  * @param seq The sequence to check.
@@ -129,12 +129,13 @@ byte_t get_nucleotide_from_encoded_seq(byte_t* seq, int32_t idx, uint8_t encodin
 /**
  * @brief Encodes a DNA sequence with each nucleotide coded on 2 bits.
  *
- *    A or a : 00
- *    C or c : 01
- *    T or t : 10
- *    G or g : 11
+ *    A or a :           00
+ *    C or c :           01
+ *    T or t or U or u : 10
+ *    G or g :           11
  *
- * @warning The DNA sequence must contain only 'atgcATGC' characters.
+ * @warning The DNA sequence must contain only 'atgcuATGCU' characters.
+ * @warning Uracil ('U') bases are encoded as Thymine ('T') bases.
  *
  * @param seq The sequence to encode.
  * @param length The length of the sequence to encode.
@@ -169,23 +170,24 @@ char* decode_seq_on_2_bits(byte_t* seq_b, int32_t length_seq);
 /**
  * @brief Encodes a DNA sequence with each nucleotide coded on 4 bits.
  *
- *		A or a : 0001
- *      C or c : 0010
- *      G or g : 0011
- *      T or t : 0100
- *      R or r : 0101
- *      Y or y : 0110
- *      S or s : 0111
- *      W or w : 1000
- *      K or k : 1001
- *      M or m : 1010
- *      B or b : 1011
- *      D or d : 1100
- *      H or h : 1101
- *      V or v : 1110
- *      N or n : 1111
+ *		A or a :           0001
+ *      C or c :           0010
+ *      G or g :           0011
+ *      T or t or U or u : 0100
+ *      R or r :           0101
+ *      Y or y :           0110
+ *      S or s :           0111
+ *      W or w :           1000
+ *      K or k :           1001
+ *      M or m :           1010
+ *      B or b :           1011
+ *      D or d :           1100
+ *      H or h :           1101
+ *      V or v :           1110
+ *      N or n :           1111
  *
  * @warning The DNA sequence must contain only IUPAC characters.
+ * @warning Uracil ('U') bases are encoded as Thymine ('T') bases.
  *
  * @param seq The sequence to encode.
  * @param length The length of the sequence to encode.

src/kmer_similarity.c

@@ -77,7 +77,6 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	int32_t*        shift_count_array;
 	Obi_ali_p       ali = NULL;
 	int 			i, j;
-	bool 			switched_seqs;
 	bool			reversed;
 	int 			score = 0;
 	Obi_blob_p      blob1 = NULL;
@@ -124,6 +123,8 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	bool			keep_seq2_start;
 	bool			keep_seq1_end;
 	bool			keep_seq2_end;
+	bool			left_ali;
+	bool 			rev_quals = false;
 
 	// Check kmer size
 	if ((kmer_size < 1) || (kmer_size > 4))
@@ -148,19 +149,8 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	}
 
 	// Choose the shortest sequence to save kmer positions in array
-	switched_seqs = false;
 	len1 = blob1->length_decoded_value;
 	len2 = blob2->length_decoded_value;
-	if (len2 < len1)
-	{
-		switched_seqs = true;
-		temp_len = len1;
-		len1 = len2;
-		len2 = temp_len;
-		temp_blob = blob1;
-		blob1 = blob2;
-		blob2 = temp_blob;
-	}
 
 	// Force encoding on 2 bits by replacing ambiguous nucleotides by 'a's
 	if (blob1->element_size == 4)
@@ -196,7 +186,47 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	else
 		reversed = false;
 	if (reversed)
-		switched_seqs = !switched_seqs;
+		// unreverse to make cases simpler. Costly but rare (direct match is reverse primer match)
+	{
+		if (seq2 == NULL)
+			seq2 = obi_blob_to_seq(blob2);
+		seq2 = reverse_complement_sequence(seq2);
+		blob2 = obi_seq_to_blob(seq2);
+
+		if (seq1 == NULL)
+			seq1 = obi_blob_to_seq(blob1);
+		seq1 = reverse_complement_sequence(seq1);
+		blob1 = obi_seq_to_blob(seq1);
+		free_blob1 = true;
+
+		// Get the quality arrays
+		qual1 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view1, qual_col1, idx1, 0, &qual1_len);
+		if (qual1 == NULL)
+		{
+			obidebug(1, "\nError getting the quality of the 1st sequence when computing the kmer similarity between two sequences");
+			return NULL;
+		}
+		qual2 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view2, qual_col2, idx2, 0, &qual2_len);
+		if (qual2 == NULL)
+		{
+			obidebug(1, "\nError getting the quality of the 2nd sequence when computing the kmer similarity between two sequences");
+			return NULL;
+		}
+
+		uint8_t* newqual1 = malloc(qual1_len*sizeof(uint8_t));
+		uint8_t* newqual2 = malloc(qual2_len*sizeof(uint8_t));
+
+		for (i=0;i<qual1_len;i++)
+			newqual1[i] = qual1[qual1_len-1-i];
+
+		for (i=0;i<qual2_len;i++)
+			newqual2[i] = qual2[qual2_len-1-i];
+
+		qual1 = newqual1;
+		qual2 = newqual2;
+
+		rev_quals = true;
+	}
 
 	// Save total length for the shift counts array
 	total_len = len1 + len2 + 1;   // +1 for shift 0
@@ -237,7 +267,7 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 		 	return NULL;
 		}
 	}
-	else if (len1 >= shift_array_height)
+	else if (total_len >= shift_array_height)
 	{
 		shift_array_height = total_len;
 		*shift_array_p = (int32_t*) realloc(*shift_array_p, ARRAY_LENGTH * shift_array_height * sizeof(int32_t));
@@ -291,7 +321,7 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	*shift_array_height_p = shift_array_height;
 	*shift_count_array_length_p = shift_count_array_length;
 
-	// Fill array with positions of kmers in the shortest sequence
+	// Fill array with positions of kmers in the first sequence
 	encoding = blob1->element_size;
 	kmer_count = len1 - kmer_size + 1;
 	for (kmer_idx=0; kmer_idx < kmer_count; kmer_idx++)
@@ -310,7 +340,7 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 		kmer_pos_array[(kmer*kmer_pos_array_height)+i] = kmer_idx;
 	}
 
-	// Compare positions of kmers between both sequences and store shifts
+	// Compare positions of kmers between both sequences and store shifts (a shift corresponds to pos2 - pos1)
 	kmer_count = blob2->length_decoded_value - kmer_size + 1;
 	for (kmer_idx=0; kmer_idx < kmer_count; kmer_idx++)
 	{
@@ -374,35 +404,42 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	// The 873863 cases of hell
 	if (best_shift > 0)
 	{
+		left_ali = false;
 		overlap_len = len2 - best_shift;
 		if (len1 <= overlap_len)
 		{
 			overlap_len = len1;
-			if (! switched_seqs)
-				keep_seq2_end = true;
-			else
-				keep_seq2_start = true;
-		}
-		else if (switched_seqs)
-		{
-			keep_seq2_start = true;
-			keep_seq1_end = true;
+			keep_seq2_end = true;
 		}
 	}
 	else if (best_shift < 0)
 	{
+		left_ali = true;
 	    overlap_len = len1 + best_shift;
-	    if (!switched_seqs)
-	    {
-	    	keep_seq1_start = true;
-	    	keep_seq2_end = true;
-	    }
-	}
-	else
-	{
-		overlap_len = len1;
-		if ((!switched_seqs) && (len2 > len1))
+		if (len2 <= overlap_len)
+		{
+			overlap_len = len2;
+			keep_seq1_start = true;
+		}
+		else
+		{
+			keep_seq1_start = true;
 			keep_seq2_end = true;
+		}
+	}
+	else // if (best_shift == 0)
+	{
+		if (len2 >= len1)
+		{
+			overlap_len = len1;
+			keep_seq2_end = true;
+			left_ali = true;
+		}
+		else
+		{
+			overlap_len = len2;
+			left_ali = false; // discussable
+		}
 	}
 
 	ali = (Obi_ali_p) malloc(sizeof(Obi_ali_t));
@@ -433,7 +470,7 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 		ali->direction[0] = '\0';
 	else
 	{
-		if (((best_shift <= 0) && (!switched_seqs)) || ((best_shift > 0) && switched_seqs))
+		if (left_ali)
 			strcpy(ali->direction, "left");
 		else
 			strcpy(ali->direction, "right");
@@ -442,28 +479,28 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	// Build the consensus sequence if asked
 	if (build_consensus)
 	{
-		// Get the quality arrays
-		qual1 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view1, qual_col1, idx1, 0, &qual1_len);
-		if (qual1 == NULL)
+		if (! rev_quals)
 		{
-			obidebug(1, "\nError getting the quality of the 1st sequence when computing the kmer similarity between two sequences");
-			return NULL;
-		}
-		qual2 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view2, qual_col2, idx2, 0, &qual2_len);
-		if (qual2 == NULL)
-		{
-			obidebug(1, "\nError getting the quality of the 2nd sequence when computing the kmer similarity between two sequences");
-			return NULL;
+			// Get the quality arrays
+			qual1 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view1, qual_col1, idx1, 0, &qual1_len);
+			if (qual1 == NULL)
+			{
+				obidebug(1, "\nError getting the quality of the 1st sequence when computing the kmer similarity between two sequences");
+				return NULL;
+			}
+			qual2 = obi_get_qual_int_with_elt_idx_and_col_p_in_view(view2, qual_col2, idx2, 0, &qual2_len);
+			if (qual2 == NULL)
+			{
+				obidebug(1, "\nError getting the quality of the 2nd sequence when computing the kmer similarity between two sequences");
+				return NULL;
+			}
 		}
 
 		// Decode the first sequence if not already done
 		if (seq1 == NULL)
 			seq1 = obi_blob_to_seq(blob1);
 
-		if (! switched_seqs)
-		    consensus_len = len2 - best_shift;
-		else
-		    consensus_len = len1 + best_shift;
+		consensus_len = len2 - best_shift;
 
 		// Allocate memory for consensus sequence
 		consensus_seq = (char*) malloc(consensus_len + 1 * sizeof(char)); // TODO keep malloced too maybe
@@ -557,6 +594,12 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	free(seq2);
 	free(blob2);
 
+	if (rev_quals)
+	{
+		free(qual1);
+		free(qual2);
+	}
+
 	return ali;
 }
 

src/obi_clean.c

@@ -88,42 +88,42 @@ static int create_output_columns(Obiview_p o_view,
 								 int sample_count)
 {
 	// Status column
-	if (obi_view_add_column(o_view, CLEAN_STATUS_COLUMN_NAME, -1, NULL, OBI_CHAR, 0, sample_count, (sample_column->header)->elements_names, true, false, false, NULL, NULL, -1, CLEAN_STATUS_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_STATUS_COLUMN_NAME, -1, NULL, OBI_CHAR, 0, sample_count, (sample_column->header)->elements_names, true, true, false, false, NULL, NULL, -1, CLEAN_STATUS_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_STATUS_COLUMN_NAME);
 		return -1;
 	}
 
 	// Head column
-	if (obi_view_add_column(o_view, CLEAN_HEAD_COLUMN_NAME, -1, NULL, OBI_BOOL, 0, 1, NULL, false, false, false, NULL, NULL, -1, CLEAN_HEAD_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_HEAD_COLUMN_NAME, -1, NULL, OBI_BOOL, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, CLEAN_HEAD_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_HEAD_COLUMN_NAME);
 		return -1;
 	}
 
 	// Sample count column
-	if (obi_view_add_column(o_view, CLEAN_SAMPLECOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, CLEAN_SAMPLECOUNT_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_SAMPLECOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, CLEAN_SAMPLECOUNT_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_SAMPLECOUNT_COLUMN_NAME);
 		return -1;
 	}
 
 	// Head count column
-	if (obi_view_add_column(o_view, CLEAN_HEADCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, CLEAN_HEADCOUNT_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_HEADCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, CLEAN_HEADCOUNT_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_HEADCOUNT_COLUMN_NAME);
 		return -1;
 	}
 
 	// Internal count column
-	if (obi_view_add_column(o_view, CLEAN_INTERNALCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, CLEAN_INTERNALCOUNT_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_INTERNALCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, CLEAN_INTERNALCOUNT_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_INTERNALCOUNT_COLUMN_NAME);
 		return -1;
 	}
 
 	// Singleton count column
-	if (obi_view_add_column(o_view, CLEAN_SINGLETONCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, CLEAN_SINGLETONCOUNT_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, CLEAN_SINGLETONCOUNT_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, CLEAN_SINGLETONCOUNT_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", CLEAN_SINGLETONCOUNT_COLUMN_NAME);
 		return -1;
@@ -229,6 +229,8 @@ int obi_clean(const char* dms_name,
 		return -1;
 	}
 
+	seq_count = (i_view->infos)->line_count;
+
 	// Open the sequence column
 	if (strcmp((i_view->infos)->view_type, VIEW_TYPE_NUC_SEQS) == 0)
 		iseq_column = obi_view_get_column(i_view, NUC_SEQUENCE_COLUMN);
@@ -245,7 +247,7 @@ int obi_clean(const char* dms_name,
 	}
 
 	// Open the sample column if there is one
-	if ((strcmp(sample_column_name, "") == 0) || (sample_column_name == NULL))
+	if ((strcmp(sample_column_name, "") == 0) || (sample_column_name == NULL) || (seq_count == 0))
 	{
 		fprintf(stderr, "Info: No sample information provided, assuming one sample.\n");
 		sample_column = obi_view_get_column(i_view, COUNT_COLUMN);
@@ -340,66 +342,67 @@ int obi_clean(const char* dms_name,
 		return -1;
 	}
 
-	// Build kmer tables
-	ktable = hash_seq_column(i_view, iseq_column, 0);
-	if (ktable == NULL)
+	if (seq_count > 0)
 	{
-		obi_set_errno(OBI_CLEAN_ERROR);
-		obidebug(1, "\nError building kmer tables before aligning");
-		return -1;
-	}
+		// Build kmer tables
+		ktable = hash_seq_column(i_view, iseq_column, 0);
+		if (ktable == NULL)
+		{
+			obi_set_errno(OBI_CLEAN_ERROR);
+			obidebug(1, "\nError building kmer tables before aligning");
+			return -1;
+		}
 
-	seq_count = (i_view->infos)->line_count;
-
-	// Allocate arrays for sample counts otherwise reading in mapped files takes longer
-	complete_sample_count_array = (int*) malloc(seq_count * sample_count * sizeof(int));
-	if (complete_sample_count_array == NULL)
-	{
-		obi_set_errno(OBI_MALLOC_ERROR);
-		obidebug(1, "\nError allocating memory for the array of sample counts, size: %lld", seq_count * sample_count * sizeof(int));
-		return -1;
-	}
-	for (samp=0; samp < sample_count; samp++)
-	{
-		for (k=0; k<seq_count; k++)
-			complete_sample_count_array[k+(samp*seq_count)] = obi_get_int_with_elt_idx_and_col_p_in_view(i_view, sample_column, k, samp);
-	}
-
-	// Allocate arrays for blobs otherwise reading in mapped files takes longer
-	blob_array = (Obi_blob_p*) malloc(seq_count * sizeof(Obi_blob_p));
-	if (blob_array == NULL)
-	{
-		obi_set_errno(OBI_MALLOC_ERROR);
-		obidebug(1, "\nError allocating memory for the array of blobs");
-		return -1;
-	}
-	for (k=0; k<seq_count; k++)
-	{
-		blob_array[k] = obi_get_blob_with_elt_idx_and_col_p_in_view(i_view, iseq_column, k, 0);
-	}
-
-	// Allocate alignment result array (byte at 0 if not aligned yet,
-	//											1 if sequence at index has a similarity above the threshold with the current sequence,
-	//											2 if sequence at index has a similarity below the threshold with the current sequence)
-	alignment_result_array = (byte_t*) calloc(seq_count, sizeof(byte_t));
-	if (alignment_result_array == NULL)
-	{
-		obi_set_errno(OBI_MALLOC_ERROR);
-		obidebug(1, "\nError allocating memory for alignment result array");
-		return -1;
-	}
-
-	// Initialize all sequences to singletons or NA if no sequences in that sample
-	for (k=0; k<seq_count; k++)
-	{
+		// Allocate arrays for sample counts otherwise reading in mapped files takes longer
+		complete_sample_count_array = (int*) malloc(seq_count * sample_count * sizeof(int));
+		if (complete_sample_count_array == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for the array of sample counts, size: %lld", seq_count * sample_count * sizeof(int));
+			return -1;
+		}
 		for (samp=0; samp < sample_count; samp++)
 		{
-			if (obi_get_int_with_elt_idx_and_col_p_in_view(i_view, sample_column, k, samp) != OBIInt_NA)  // Only initialize samples where there are some sequences
+			for (k=0; k<seq_count; k++)
+				complete_sample_count_array[k+(samp*seq_count)] = obi_get_int_with_elt_idx_and_col_p_in_view(i_view, sample_column, k, samp);
+		}
+
+		// Allocate arrays for blobs otherwise reading in mapped files takes longer
+		blob_array = (Obi_blob_p*) malloc(seq_count * sizeof(Obi_blob_p));
+		if (blob_array == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for the array of blobs");
+			return -1;
+		}
+		for (k=0; k<seq_count; k++)
+		{
+			blob_array[k] = obi_get_blob_with_elt_idx_and_col_p_in_view(i_view, iseq_column, k, 0);
+		}
+
+		// Allocate alignment result array (byte at 0 if not aligned yet,
+		//											1 if sequence at index has a similarity above the threshold with the current sequence,
+		//											2 if sequence at index has a similarity below the threshold with the current sequence)
+		alignment_result_array = (byte_t*) calloc(seq_count, sizeof(byte_t));
+		if (alignment_result_array == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for alignment result array");
+			return -1;
+		}
+
+		// Initialize all sequences to singletons or NA if no sequences in that sample
+		for (k=0; k<seq_count; k++)
+		{
+			for (samp=0; samp < sample_count; samp++)
 			{
-				if (obi_set_char_with_elt_idx_and_col_p_in_view(o_view, status_column, k, samp, 's') < 0)
+				if (obi_get_int_with_elt_idx_and_col_p_in_view(i_view, sample_column, k, samp) != OBIInt_NA)  // Only initialize samples where there are some sequences
 				{
-					obidebug(1, "\nError initializing all sequences to singletons");
-					return -1;
+					if (obi_set_char_with_elt_idx_and_col_p_in_view(o_view, status_column, k, samp, 's') < 0)
+					{
+						obidebug(1, "\nError initializing all sequences to singletons");
+						return -1;
+					}
 				}
 			}
 		}
@@ -551,17 +554,20 @@ int obi_clean(const char* dms_name,
 		}
 	}
 
-	free_kmer_tables(ktable, seq_count);
-	free(complete_sample_count_array);
-	free(blob_array);
-	free(alignment_result_array);
+	if (seq_count > 0)
+	{
+		free_kmer_tables(ktable, seq_count);
+		free(complete_sample_count_array);
+		free(blob_array);
+		free(alignment_result_array);
+	}
 
 	fprintf(stderr, "\n");
 
 	if (stop)
 		return -1;
 
-	if (heads_only)
+	if (heads_only && (seq_count > 0))
 	{
 		line_selection = malloc((((o_view->infos)->line_count) + 1) * sizeof(index_t));
 		if (line_selection == NULL)
@@ -635,7 +641,7 @@ int obi_clean(const char* dms_name,
 	}
 
 	// Flag the end of the line selection
-	if (heads_only)
+	if (heads_only && (seq_count > 0))
 		line_selection[l] = -1;
 
 	// Create new view with line selection if heads only

src/obi_ecopcr.c

@@ -150,49 +150,49 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 {
 	// Original length column
-	if (obi_view_add_column(o_view, ECOPCR_SEQLEN_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SEQLEN_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_SEQLEN_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SEQLEN_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_SEQLEN_COLUMN_NAME);
 		return -1;
 	}
 
 	// Amplicon length column
-	if (obi_view_add_column(o_view, ECOPCR_AMPLICONLEN_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_AMPLICONLEN_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_AMPLICONLEN_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_AMPLICONLEN_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_AMPLICONLEN_COLUMN_NAME);
 		return -1;
 	}
 
 	// Taxid column
-	if (obi_view_add_column(o_view, TAXID_COLUMN, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, TAXID_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, TAXID_COLUMN, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, TAXID_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", TAXID_COLUMN);
 		return -1;
 	}
 
 	// Taxonomic rank column
-	if (obi_view_add_column(o_view, ECOPCR_RANK_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_RANK_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_RANK_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_RANK_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_RANK_COLUMN_NAME);
 		return -1;
 	}
 
 	// Species taxid column
-	if (obi_view_add_column(o_view, ECOPCR_SPECIES_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SPECIES_TAXID_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_SPECIES_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SPECIES_TAXID_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_SPECIES_TAXID_COLUMN_NAME);
 		return -1;
 	}
 
 	// Genus taxid column
-	if (obi_view_add_column(o_view, ECOPCR_GENUS_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_GENUS_TAXID_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_GENUS_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_GENUS_TAXID_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_GENUS_TAXID_COLUMN_NAME);
 		return -1;
 	}
 
 	// Family taxid column
-	if (obi_view_add_column(o_view, ECOPCR_FAMILY_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_FAMILY_TAXID_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_FAMILY_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_FAMILY_TAXID_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_FAMILY_TAXID_COLUMN_NAME);
 		return -1;
@@ -201,7 +201,7 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	if (kingdom_mode)
 	{
 		// Kingdom taxid column
-		if (obi_view_add_column(o_view, ECOPCR_KINGDOM_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_KINGDOM_TAXID_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(o_view, ECOPCR_KINGDOM_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_KINGDOM_TAXID_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the %s column", ECOPCR_KINGDOM_TAXID_COLUMN_NAME);
 			return -1;
@@ -210,7 +210,7 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	else
 	{
 		// Superkingdom taxid column
-		if (obi_view_add_column(o_view, ECOPCR_SUPERKINGDOM_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SUPERKINGDOM_TAXID_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(o_view, ECOPCR_SUPERKINGDOM_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SUPERKINGDOM_TAXID_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the %s column", ECOPCR_SUPERKINGDOM_TAXID_COLUMN_NAME);
 			return -1;
@@ -218,28 +218,28 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	}
 
 	// Scientific name column
-	if (obi_view_add_column(o_view, ECOPCR_SCIENTIFIC_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SCIENTIFIC_NAME_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_SCIENTIFIC_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SCIENTIFIC_NAME_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_SCIENTIFIC_NAME_COLUMN_NAME);
 		return -1;
 	}
 
 	// Species name column
-	if (obi_view_add_column(o_view, ECOPCR_SPECIES_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SPECIES_NAME_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_SPECIES_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SPECIES_NAME_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_SPECIES_NAME_COLUMN_NAME);
 		return -1;
 	}
 
 	// Genus name column
-	if (obi_view_add_column(o_view, ECOPCR_GENUS_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_GENUS_NAME_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_GENUS_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_GENUS_NAME_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_GENUS_NAME_COLUMN_NAME);
 		return -1;
 	}
 
 	// Family name column
-	if (obi_view_add_column(o_view, ECOPCR_FAMILY_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_FAMILY_NAME_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_FAMILY_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_FAMILY_NAME_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_FAMILY_NAME_COLUMN_NAME);
 		return -1;
@@ -248,7 +248,7 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	if (kingdom_mode)
 	{
 		// Kingdom name column
-		if (obi_view_add_column(o_view, ECOPCR_KINGDOM_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_KINGDOM_NAME_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(o_view, ECOPCR_KINGDOM_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_KINGDOM_NAME_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the %s column", ECOPCR_KINGDOM_NAME_COLUMN_NAME);
 			return -1;
@@ -257,7 +257,7 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	else
 	{
 		// Superkingdom name column
-		if (obi_view_add_column(o_view, ECOPCR_SUPERKINGDOM_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_SUPERKINGDOM_NAME_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(o_view, ECOPCR_SUPERKINGDOM_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_SUPERKINGDOM_NAME_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the %s column", ECOPCR_SUPERKINGDOM_NAME_COLUMN_NAME);
 			return -1;
@@ -265,49 +265,49 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode)
 	}
 
 	// Strand column
-	if (obi_view_add_column(o_view, ECOPCR_STRAND_COLUMN_NAME, -1, NULL, OBI_CHAR, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_STRAND_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_STRAND_COLUMN_NAME, -1, NULL, OBI_CHAR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_STRAND_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_STRAND_COLUMN_NAME);
 		return -1;
 	}
 
 	// Primer 1 column
-	if (obi_view_add_column(o_view, ECOPCR_PRIMER1_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_PRIMER1_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_PRIMER1_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_PRIMER1_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_PRIMER1_COLUMN_NAME);
 		return -1;
 	}
 
 	// Primer 2 column
-	if (obi_view_add_column(o_view, ECOPCR_PRIMER2_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_PRIMER2_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_PRIMER2_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_PRIMER2_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_PRIMER2_COLUMN_NAME);
 		return -1;
 	}
 
 	// Error 1 column
-	if (obi_view_add_column(o_view, ECOPCR_ERROR1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_ERROR1_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_ERROR1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_ERROR1_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_ERROR1_COLUMN_NAME);
 		return -1;
 	}
 
 	// Error 2 column
-	if (obi_view_add_column(o_view, ECOPCR_ERROR2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_ERROR2_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_ERROR2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_ERROR2_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_ERROR2_COLUMN_NAME);
 		return -1;
 	}
 
 	// Temperature 1 column
-	if (obi_view_add_column(o_view, ECOPCR_TEMP1_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_TEMP1_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_TEMP1_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_TEMP1_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_TEMP1_COLUMN_NAME);
 		return -1;
 	}
 
 	// Temperature 2 column
-	if (obi_view_add_column(o_view, ECOPCR_TEMP2_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ECOPCR_TEMP2_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(o_view, ECOPCR_TEMP2_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ECOPCR_TEMP2_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the %s column", ECOPCR_TEMP2_COLUMN_NAME);
 		return -1;
@@ -365,8 +365,6 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 
 	int32_t i;
 
-	// TODO add check for primer longer than MAX_PAT_LEN (32)
-
 	// Get sequence id
 	seq_id = obi_get_str_with_elt_idx_and_col_p_in_view(i_view, i_id_column, i_idx, 0);
 
@@ -645,7 +643,8 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 
 int obi_ecopcr(const char* i_dms_name,
 		       const char* i_view_name,
-			   const char* taxonomy_name,     // TODO discuss that input dms assumed
+			   const char* tax_dms_name,
+			   const char* taxonomy_name,
 			   const char* o_dms_name,
 			   const char* o_view_name,
 			   const char* o_view_comments,
@@ -678,6 +677,7 @@ int obi_ecopcr(const char* i_dms_name,
 
 	OBIDMS_p          i_dms = NULL;
 	OBIDMS_p          o_dms = NULL;
+	OBIDMS_p          tax_dms = NULL;
 	OBIDMS_taxonomy_p taxonomy = NULL;
 	Obiview_p         i_view = NULL;
 	Obiview_p         o_view = NULL;
@@ -749,6 +749,20 @@ int obi_ecopcr(const char* i_dms_name,
 	o1c = complementPattern(o1);
 	o2c = complementPattern(o2);
 
+	// check for primers equal or longer than MAX_PAT_LEN (32)
+	if (strlen(primer1) >= MAX_PAT_LEN)
+	{
+		obi_set_errno(OBI_ECOPCR_ERROR);
+		obidebug(1, "\nError: first primer is too long, needs to be < 32bp (%s)", primer1);
+		return -1;
+	}
+	if (strlen(primer2) >= MAX_PAT_LEN)
+	{
+		obi_set_errno(OBI_ECOPCR_ERROR);
+		obidebug(1, "\nError: second primer is too long, needs to be < 32bp (%s)", primer2);
+		return -1;
+	}
+
 	// Open input DMS
 	i_dms = obi_open_dms(i_dms_name, false);
 	if (i_dms == NULL)
@@ -965,8 +979,16 @@ int obi_ecopcr(const char* i_dms_name,
 		return -1;
 	}
 
+	// Open taxonomy DMS
+	tax_dms = obi_open_dms(tax_dms_name, false);
+	if (tax_dms == NULL)
+	{
+		obidebug(1, "\nError opening the taxonomy DMS");
+		return -1;
+	}
+
 	// Open the taxonomy
-	taxonomy = obi_read_taxonomy(i_dms, taxonomy_name, false);
+	taxonomy = obi_read_taxonomy(tax_dms, taxonomy_name, false);
 	if (taxonomy == NULL)
 	{
 		obidebug(1, "\nError opening the taxonomy");

src/obi_ecopcr.h

@@ -77,7 +77,8 @@
  *
  * @param i_dms_name The path to the input DMS.
  * @param i_view_name The name of the input view.
- * @param taxonomy_name The name of the taxonomy in the input DMS.
+ * @param tax_dms_name The path to the DMS containing the taxonomy.
+ * @param taxonomy_name The name of the taxonomy.
  * @param o_dms_name The path to the output DMS.
  * @param o_view_name The name of the output view.
  * @param o_view_comments The comments to associate with the output view.
@@ -106,6 +107,7 @@
  */
 int obi_ecopcr(const char* i_dms_name,
 		       const char* i_view_name,
+			   const char* tax_dms_name,
 			   const char* taxonomy_name,
 			   const char* o_dms_name,
 			   const char* o_view_name,

src/obi_ecotag.c

@@ -104,42 +104,42 @@ int print_assignment_result(Obiview_p output_view, index_t line,
 static int create_output_columns(Obiview_p o_view)
 {
 	// Score column
-	if (obi_view_add_column(o_view, ECOTAG_SCORE_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_SCORE_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the score in ecotag");
 		return -1;
 	}
 
 	// Assigned taxid column
-	if (obi_view_add_column(o_view, ECOTAG_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_TAXID_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the assigned taxid in ecotag");
 		return -1;
 	}
 
 	// Assigned scientific name column
-	if (obi_view_add_column(o_view, ECOTAG_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_NAME_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the assigned scientific name in ecotag");
 		return -1;
 	}
 
 	// Assignement status column
-	if (obi_view_add_column(o_view, ECOTAG_STATUS_COLUMN_NAME, -1, NULL, OBI_BOOL, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_STATUS_COLUMN_NAME, -1, NULL, OBI_BOOL, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the assignment status in ecotag");
 		return -1;
 	}
 
 	// Column for array of best match ids
-	if (obi_view_add_column(o_view, ECOTAG_BEST_MATCH_IDS_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, true, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_BEST_MATCH_IDS_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, true, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the array of ids of best matches in ecotag");
 		return -1;
 	}
 
 	// Column for array of best match taxids
-	if (obi_view_add_column(o_view, ECOTAG_BEST_MATCH_TAXIDS_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, true, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(o_view, ECOTAG_BEST_MATCH_TAXIDS_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, true, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the array of taxids of best matches in ecotag");
 		return -1;
@@ -218,7 +218,8 @@ int obi_ecotag(const char* dms_name,
 			   const char* taxonomy_name,
 			   const char* output_view_name,
 			   const char* output_view_comments,
-			   double ecotag_threshold)    // TODO different threshold for the similarity sphere around ref seqs
+			   double ecotag_threshold,
+			   double bubble_threshold)
 {
 
 	// For each sequence
@@ -232,13 +233,14 @@ int obi_ecotag(const char* dms_name,
 		// Write result (max score, threshold, LCA assigned, list of the ids of the best matches)
 
 
-	index_t         i, j, k;
+	index_t         i, j, k, t;
 	ecotx_t*        lca;
 	ecotx_t*        lca_in_array;
 	ecotx_t*        best_match;
 	index_t         query_seq_idx, ref_seq_idx;
 	double          score, best_score;
 	double			threshold;
+	double			lca_threshold;
 	int             lcs_length;
 	int             ali_length;
 	Kmer_table_p    ktable;
@@ -257,16 +259,20 @@ int obi_ecotag(const char* dms_name,
 	int32_t*   		best_match_taxids;
 	int32_t*		best_match_taxids_to_store;
 	int				best_match_count;
+	int				best_match_taxid_count;
 	int             buffer_size;
 	int 			best_match_ids_buffer_size;
 	index_t         best_match_idx;
 	int32_t         lca_array_length;
 	int32_t         lca_taxid;
 	int32_t         taxid_best_match;
+	int32_t         taxid;
+	int32_t         taxid_to_store;
 	bool   			assigned;
 	const char*		lca_name;
 	const char*		id;
 	int				id_len;
+	bool			already_in;
 
 	OBIDMS_p          dms = NULL;
 	OBIDMS_p		  ref_dms = NULL;
@@ -389,10 +395,10 @@ int obi_ecotag(const char* dms_name,
 		return -1;
 	}
 	free(db_threshold_str);
-	if (ecotag_threshold < db_threshold)
+	if (bubble_threshold < db_threshold)
 	{
 		fprintf(stderr, "\nError: The threshold demanded (%f) is lower than the threshold used to build the reference database (%f).\n\n",
-				ecotag_threshold, db_threshold);
+				bubble_threshold, db_threshold);
 		return -1;
 	}
 
@@ -486,10 +492,11 @@ int obi_ecotag(const char* dms_name,
 
 	for (i=0; i < query_count; i++)
 	{
-		if (i%1000 == 0)
+		if (i%10 == 0)
 			fprintf(stderr,"\rDone : %f %%       ", (i / (float) query_count)*100);
 
 		best_match_count = 0;
+		best_match_taxid_count = 0;
 		best_match_ids_length = 0;
 		threshold = ecotag_threshold;
 		best_score = 0.0;
@@ -541,6 +548,7 @@ int obi_ecotag(const char* dms_name,
 					// Reset the array with that match
 					best_match_ids_length = 0;
 					best_match_count = 0;
+					best_match_taxid_count = 0;
 				}
 
 				// Store in best match array
@@ -583,8 +591,27 @@ int obi_ecotag(const char* dms_name,
 
 				// Save match
 				best_match_array[best_match_count] = j;
-				best_match_taxids[best_match_count] = obi_get_int_with_elt_idx_and_col_p_in_view(ref_view, ref_taxid_column, j, 0);
 				best_match_count++;
+
+				// Save best match taxid only if not already in array
+				taxid_to_store = obi_get_int_with_elt_idx_and_col_p_in_view(ref_view, ref_taxid_column, j, 0);
+				already_in = false;
+				for (t=0; t<best_match_taxid_count; t++)
+				{
+					taxid = best_match_taxids[t];
+					//fprintf(stderr, "\ntaxid %d, taxid_to_store %d\n", taxid, taxid_to_store);
+					if (taxid == taxid_to_store)
+					{
+						already_in = true;
+						break;
+					}
+				}
+				if (! already_in)
+				{
+					best_match_taxids[best_match_taxid_count] = taxid_to_store;
+					best_match_taxid_count++;
+				}
+
 				strcpy(best_match_ids+best_match_ids_length, id);
 				best_match_ids_length = best_match_ids_length + id_len + 1;
 			}
@@ -597,11 +624,16 @@ int obi_ecotag(const char* dms_name,
 		{
 			best_match_idx = best_match_array[j];
 
-			// Find the LCA for the chosen threshold
+			// Find the LCA for the highest threshold between best_score and the chosen bubble threshold
 			score_array = obi_get_array_with_col_p_in_view(ref_view, score_a_column, best_match_idx, &lca_array_length);
 
+			if (bubble_threshold < best_score)
+				lca_threshold = best_score;
+			else
+				lca_threshold = bubble_threshold;
+
 			k = 0;
-			while ((k < lca_array_length) && (score_array[k] >= best_score))
+			while ((k < lca_array_length) && (score_array[k] >= lca_threshold))
 				k++;
 
 			if (k>0)
@@ -686,7 +718,7 @@ int obi_ecotag(const char* dms_name,
 									assigned_name_column, lca_name,
 									assigned_status_column, assigned,
 									best_match_ids_column, best_match_ids_to_store, best_match_ids_length,
-									best_match_taxids_column, best_match_taxids_to_store, best_match_count,
+									best_match_taxids_column, best_match_taxids_to_store, best_match_taxid_count,
 									score_column, best_score
 									) < 0)
 							return -1;

src/obi_ecotag.h

@@ -42,12 +42,14 @@
  * @param output_view_name The name to give to the output view.
  * @param output_view_comments The comments to associate to the output view.
  * @param ecotag_threshold The threshold at which to assign.
+ * @param bubble_threshold The threshold at which to look for an LCA (i.e. minimum identity considered for the assignment circle);
+ *                         the threshold actually used will be the highest between this value and the best assignment score found.
  *
  * 	The algorithm works like this:
  * 		For each query sequence:
  *			Align with reference database
  *			Keep the indices of all the best matches
- *			For each kept index, get the LCA at that threshold as stored in the reference database, then the LCA of those LCAs
+ *			For each kept index, get the LCA at the highest threshold between bubble_threshold and the best assignment score found (as stored in the reference database), then the LCA of those LCAs
  *			Write result (max score, threshold, taxid and scientific name of the LCA assigned, list of the ids of the best matches)
  *
  * @returns A value indicating the success of the operation.
@@ -65,7 +67,8 @@ int obi_ecotag(const char* dms_name,
 			   const char* taxonomy_name,
 			   const char* output_view_name,
 			   const char* output_view_comments,
-			   double ecotag_threshold);
+			   double ecotag_threshold,
+			   double bubble_threshold);
 
 
 #endif /* OBI_ECOTAG_H_ */

src/obi_lcs.c

@@ -10,8 +10,9 @@
  */
 
 //#define OMP_SUPPORT // TODO
-#ifdef OMP_SUPPORT
+#ifdef _OPENMP
 #include <omp.h>
+#include <pthread.h>
 #endif
 
 #include <stdlib.h>
@@ -155,35 +156,35 @@ static int create_alignment_output_columns(Obiview_p output_view,
 										   bool normalize, int reference, bool similarity_mode)
 {
 	// Create the column for the ids of the 1st sequence aligned
-	if (obi_view_add_column(output_view, ID1_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, id1_indexer_name, NULL, -1, ID1_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(output_view, ID1_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, id1_indexer_name, NULL, -1, ID1_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the first column for the sequence ids when aligning");
 		return -1;
 	}
 
 	// Create the column for the ids of the 2nd sequence aligned
-	if (obi_view_add_column(output_view, ID2_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, id2_indexer_name, NULL, -1, ID2_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(output_view, ID2_COLUMN_NAME, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, id2_indexer_name, NULL, -1, ID2_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the second column for the sequence ids when aligning");
 		return -1;
 	}
 
 	// Create the column for the index (in the input view) of the first sequences aligned
-	if (obi_view_add_column(output_view, IDX1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, IDX1_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(output_view, IDX1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, IDX1_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the first column for the sequence indices when aligning");
 		return -1;
 	}
 
 	// Create the column for the index (in the input view) of the second sequences aligned
-	if (obi_view_add_column(output_view, IDX2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, IDX2_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(output_view, IDX2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, IDX2_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the second column for the sequence indices when aligning");
 		return -1;
 	}
 
 	// Create the column for the LCS length
-	if (obi_view_add_column(output_view, LCS_LENGTH_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, LCS_LENGTH_COLUMN_COMMENTS, true) < 0)
+	if (obi_view_add_column(output_view, LCS_LENGTH_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, LCS_LENGTH_COLUMN_COMMENTS, true) < 0)
 	{
 		obidebug(1, "\nError creating the column for the LCS length when aligning");
 		return -1;
@@ -192,7 +193,7 @@ static int create_alignment_output_columns(Obiview_p output_view,
 	// Create the column for the alignment length if it is computed
 	if ((reference == ALILEN) && (normalize || !similarity_mode))
 	{
-		if (obi_view_add_column(output_view, ALI_LENGTH_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, ALI_LENGTH_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(output_view, ALI_LENGTH_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, ALI_LENGTH_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the column for the alignment length when aligning");
 			return -1;
@@ -201,7 +202,7 @@ static int create_alignment_output_columns(Obiview_p output_view,
 	// Create the column for the alignment score
 	if (normalize)
 	{
-		if (obi_view_add_column(output_view, SCORE_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, NULL, NULL, -1, SCORE_COLUMN_NAME, true) < 0)
+		if (obi_view_add_column(output_view, SCORE_COLUMN_NAME, -1, NULL, OBI_FLOAT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, SCORE_COLUMN_NAME, true) < 0)
 		{
 			obidebug(1, "\nError creating the column for the score when aligning");
 			return -1;
@@ -209,7 +210,7 @@ static int create_alignment_output_columns(Obiview_p output_view,
 	}
 	else
 	{
-		if (obi_view_add_column(output_view, SCORE_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, SCORE_COLUMN_NAME, true) < 0)
+		if (obi_view_add_column(output_view, SCORE_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, SCORE_COLUMN_NAME, true) < 0)
 		{
 			obidebug(1, "\nError creating the column for the score when aligning");
 			return -1;
@@ -219,14 +220,14 @@ static int create_alignment_output_columns(Obiview_p output_view,
 	if (print_seq)
 	{
 		// Create the column for the first sequences aligned
-		if (obi_view_add_column(output_view, SEQ1_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, seq1_indexer_name, NULL, -1, SEQ1_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(output_view, SEQ1_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, false, seq1_indexer_name, NULL, -1, SEQ1_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the first column for the sequences when aligning");
 			return -1;
 		}
 
 		// Create the column for the second sequences aligned
-		if (obi_view_add_column(output_view, SEQ2_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, seq2_indexer_name, NULL, -1, SEQ2_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(output_view, SEQ2_COLUMN_NAME, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, false, seq2_indexer_name, NULL, -1, SEQ2_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the second column for the sequences when aligning");
 			return -1;
@@ -235,14 +236,14 @@ static int create_alignment_output_columns(Obiview_p output_view,
 	if (print_count)
 	{
 		// Create the column for the count of the first sequences aligned
-		if (obi_view_add_column(output_view, COUNT1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, COUNT1_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(output_view, COUNT1_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, COUNT1_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the first column for the sequence counts when aligning");
 			return -1;
 		}
 
 		// Create the column for the count of the second sequences aligned
-		if (obi_view_add_column(output_view, COUNT2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, COUNT2_COLUMN_COMMENTS, true) < 0)
+		if (obi_view_add_column(output_view, COUNT2_COLUMN_NAME, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, COUNT2_COLUMN_COMMENTS, true) < 0)
 		{
 			obidebug(1, "\nError creating the second column for the sequence counts when aligning");
 			return -1;
@@ -407,10 +408,15 @@ int obi_lcs_align_one_column(const char* dms_name,
 	int             lcs_length;
 	int             ali_length;
 	Kmer_table_p    ktable;
+	Obi_blob_p      blob1_mapped;
+	Obi_blob_p   	blob2_mapped;
 	Obi_blob_p      blob1;
-	Obi_blob_p   	blob2;
+	Obi_blob_p      blob2;
+	int				blob1_size;
+	int             blob2_size;
 	int				lcs_min;
 	index_t         seq_elt_idx;
+	bool			stop = false;
 
 	OBIDMS_p        dms = NULL;
 	Obiview_p       seq_view = NULL;
@@ -568,10 +574,16 @@ int obi_lcs_align_one_column(const char* dms_name,
 
 	seq_count = (seq_view->infos)->line_count;
 
-	#ifdef OMP_SUPPORT
-	omp_set_num_threads(thread_count);
-	#pragma omp parallel for
-	#endif
+//	#ifdef _OPENMP
+//	int max_threads = omp_get_max_threads();
+//	if ((thread_count == -1) || (thread_count > max_threads))
+//		thread_count = max_threads;
+//	omp_set_num_threads(thread_count);
+//	fprintf(stderr, "Running on %d thread(s)\n", thread_count);
+//	pthread_mutex_t mutex;
+//	if (pthread_mutex_init(&mutex, NULL) < 0)
+//		return -1;
+//	#endif
 
 	for (i=0; i < (seq_count - 1); i++)
 	{
@@ -585,23 +597,45 @@ int obi_lcs_align_one_column(const char* dms_name,
 		id1_idx = obi_get_index_with_elt_idx_and_col_p_in_view(seq_view, id_column, i, 0);	// TODO Could there be multiple IDs per line?
 		// Get first sequence and its index
 		seq1_idx = obi_get_index_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, i, seq_elt_idx);
-		blob1 = obi_get_blob_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, i, seq_elt_idx);
+		blob1_mapped = obi_get_blob_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, i, seq_elt_idx);
+		if (blob1_mapped == NULL)
+		{
+			obidebug(1, "\nError retrieving sequences to align");
+			return -1;
+		}
+		blob1_size = obi_blob_sizeof(blob1_mapped);
+		blob1 = (Obi_blob_p) malloc(blob1_size);
 		if (blob1 == NULL)
 		{
 			obidebug(1, "\nError retrieving sequences to align");
 			return -1;
 		}
+		blob1 = memcpy(blob1, blob1_mapped, blob1_size);
 
+//#pragma omp parallel shared(blob1_mapped, blob1, blob1_size, id1_idx, seq1_idx, stop, seq_count, output_view, k, ktable, \
+//		   idx1_column, idx2_column, i, id1_column, id2_column, print_seq, seq1_column, seq2_column, \
+//		   print_count, count1_column, count2_column, ali_length_column, lcs_length_column, score_column, reference, normalize, similarity_mode) \
+//		private(blob2_mapped, blob2, blob2_size, j, id2_idx, seq2_idx, count2, count1, score, ali_length, lcs_length)
+//{
+//		#pragma omp for schedule(dynamic, seq_count/thread_count + (seq_count % thread_count != 0))  // Avoid 0 which blocks the program
 		for (j=i+1; j < seq_count; j++)
 		{
 			// Get second sequence and its index
 			seq2_idx = obi_get_index_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, j, seq_elt_idx);
-			blob2 = obi_get_blob_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, j, seq_elt_idx);
+			blob2_mapped = obi_get_blob_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, j, seq_elt_idx);
+			if (blob2_mapped == NULL)
+			{
+				obidebug(1, "\nError retrieving sequences to align");
+				stop = true;
+			}
+			blob2_size = obi_blob_sizeof(blob2_mapped);
+			blob2 = (Obi_blob_p) malloc(blob2_size);
 			if (blob2 == NULL)
 			{
 				obidebug(1, "\nError retrieving sequences to align");
-				return -1;
+				stop = true;
 			}
+			blob2 = memcpy(blob2, blob2_mapped, blob2_size);
 
 			// Check if the sequences are identical in a quick way (same index in the same indexer)
 			if (obi_get_index_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, i, seq_elt_idx) == obi_get_index_with_elt_idx_and_col_p_in_view(seq_view, iseq_column, j, seq_elt_idx))
@@ -624,6 +658,8 @@ int obi_lcs_align_one_column(const char* dms_name,
 					score = obiblob_sse_banded_lcs_align(blob1, blob2, threshold, normalize, reference, similarity_mode, &lcs_length, &ali_length);
 			}
 
+			free(blob2);
+
 			if ((score >= 0) && (((normalize || similarity_mode) && (score >= threshold)) || ((!similarity_mode && !normalize) && (score <= threshold))))
 			{	// Print result
 
@@ -637,6 +673,10 @@ int obi_lcs_align_one_column(const char* dms_name,
 					count2 = obi_get_int_with_elt_idx_and_col_p_in_view(seq_view, i_count_column, j, 0);
 				}
 
+//				#ifdef _OPENMP
+//				if (pthread_mutex_lock(&mutex) < 0)
+//					stop = true;
+//				#endif
 				if (print_alignment_result(output_view, k,
 										   idx1_column, idx2_column, i, j,
 										   id1_column, id2_column, id1_idx, id2_idx,
@@ -646,15 +686,24 @@ int obi_lcs_align_one_column(const char* dms_name,
 										   lcs_length_column, lcs_length,
 										   score_column, score,
 										   reference, normalize, similarity_mode) < 0)
-					return -1;
+					stop = true;
+//				#ifdef _OPENMP
+//				if (pthread_mutex_unlock(&mutex) < 0)
+//					stop = true;
+//				#endif
 
 				k++;
 			}
 		}
+//}
+		free(blob1);
 	}
 
 	fprintf(stderr,"\rDone : 100 %%         \n");
 
+	//fprintf(stderr,"\nstop=%d\n", stop);
+
+
 	// Close views
 	if (obi_save_and_close_view(seq_view) < 0)
 	{
@@ -675,6 +724,11 @@ int obi_lcs_align_one_column(const char* dms_name,
 
 	free_kmer_tables(ktable, seq_count);
 
+//	#ifdef _OPENMP
+//	if (pthread_mutex_destroy(&mutex) < 0)
+//		return -1;
+//	#endif
+
 	return 0;
 }
 
@@ -1087,4 +1141,3 @@ int obi_lcs_align_two_columns(const char* dms_name,
 
 	return 0;
 }
-

src/obiavl.c

@@ -582,6 +582,7 @@ int truncate_avl_to_size_used(OBIDMS_avl_p avl)	// TODO is it necessary to unmap
 {
 	size_t  			file_size;
 	size_t  			new_data_size;
+	size_t  			header_size;
 	double  			multiple;
 	int 				file_descriptor;
 
@@ -589,6 +590,8 @@ int truncate_avl_to_size_used(OBIDMS_avl_p avl)	// TODO is it necessary to unmap
 	multiple = ceil((double) (ONE_IF_ZERO((avl->header)->nb_items * sizeof(AVL_node_t))) / (double) getpagesize());
 	new_data_size = ((size_t) multiple) * getpagesize();
 
+	header_size = (avl->header)->header_size;
+
 	// Check that it is actually greater than the current size of the file, otherwise no need to truncate
 	if ((avl->header)->avl_size == new_data_size)
 		return 0;
@@ -596,16 +599,22 @@ int truncate_avl_to_size_used(OBIDMS_avl_p avl)	// TODO is it necessary to unmap
 	// Get the file descriptor
 	file_descriptor = avl->avl_fd;
 
-	// Unmap the tree before truncating the file
+	// Unmap the entire file before truncating it (WSL requirement)
 	if (munmap(avl->tree, (avl->header)->avl_size) < 0)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
 		obidebug(1, "\nError munmapping the tree of an AVL before truncating");
 		return -1;
 	}
+	if (munmap(avl->header, header_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the tree of an AVL before truncating");
+		return -1;
+	}
 
 	// Truncate the file
-	file_size = (avl->header)->header_size + new_data_size;
+	file_size = header_size + new_data_size;
 	if (ftruncate(file_descriptor, file_size) < 0)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
@@ -613,7 +622,22 @@ int truncate_avl_to_size_used(OBIDMS_avl_p avl)	// TODO is it necessary to unmap
 		return -1;
 	}
 
-	// Remap the data
+	// Remap the header and the data
+
+	avl->header    = mmap(NULL,
+			         header_size,
+					 PROT_READ | PROT_WRITE,
+					 MAP_SHARED,
+					 file_descriptor,
+					 0
+					);
+	if (avl->header == MAP_FAILED)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError re-mmapping the header of an AVL after truncating");
+		return -1;
+	}
+
 	avl->tree = mmap(NULL,
 			         new_data_size,
 					 PROT_READ | PROT_WRITE,
@@ -640,6 +664,7 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 {
 	size_t  			file_size;
 	index_t  			new_data_size;
+	size_t  			header_size;
 	double  			multiple;
 	int 				file_descriptor;
 
@@ -647,6 +672,8 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 	multiple = ceil((double) (ONE_IF_ZERO((avl_data->header)->data_size_used)) / (double) getpagesize());
 	new_data_size = ((index_t) multiple) * getpagesize();
 
+	header_size = (avl_data->header)->header_size;
+
 	// Check that it is actually greater than the current size of the file, otherwise no need to truncate
 	if ((avl_data->header)->data_size_max >= new_data_size)
 		return 0;
@@ -654,7 +681,8 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 	// Get the file descriptor
 	file_descriptor = avl_data->data_fd;
 
-	// Unmap the data before truncating the file
+	// Unmap the entire file before truncating it (WSL requirement)
+
 	if (munmap(avl_data->data, (avl_data->header)->data_size_max) < 0)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
@@ -662,8 +690,15 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 		return -1;
 	}
 
+	if (munmap(avl_data->header, header_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the header of an AVL before truncating");
+		return -1;
+	}
+
 	// Truncate the file
-	file_size = (avl_data->header)->header_size + new_data_size;
+	file_size = header_size + new_data_size;
 	if (ftruncate(file_descriptor, file_size) < 0)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
@@ -672,6 +707,22 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 	}
 
 	// Remap the data
+
+	avl_data->header = mmap(NULL,
+					        header_size,
+						    PROT_READ | PROT_WRITE,
+						    MAP_SHARED,
+						    file_descriptor,
+						    0
+					      );
+
+	if (avl_data->header == MAP_FAILED)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError re-mmapping the header of an AVL after truncating");
+		return -1;
+	}
+
 	avl_data->data = mmap(NULL,
 					      new_data_size,
 						  PROT_READ | PROT_WRITE,
@@ -710,6 +761,20 @@ int grow_avl(OBIDMS_avl_p avl)		// TODO Lock when needed
 	header_size = (avl->header)->header_size;
 	file_size = header_size + new_data_size;
 
+	// Unmap the entire file before truncating it (WSL requirement)
+	if (munmap(avl->tree, old_data_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the tree of an AVL tree file before enlarging");
+		return -1;
+	}
+	if (munmap(avl->header, header_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the header of an AVL tree file before enlarging");
+		return -1;
+	}
+
 	// Enlarge the file
 	if (ftruncate(avl_file_descriptor, file_size) < 0)
 	{
@@ -718,12 +783,20 @@ int grow_avl(OBIDMS_avl_p avl)		// TODO Lock when needed
 		return -1;
 	}
 
-	// Unmap and re-map the data
+	// Re-map
 
-	if (munmap(avl->tree, old_data_size) < 0)
+	avl->header = mmap(NULL,
+					   header_size,
+					   PROT_READ | PROT_WRITE,
+					   MAP_SHARED,
+					   avl_file_descriptor,
+					   0
+					  );
+
+	if (avl->header == MAP_FAILED)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
-		obidebug(1, "\nError munmapping the tree of an AVL tree file before enlarging");
+		obidebug(1, "\nError re-mmapping the header of an AVL tree file after enlarging the file");
 		return -1;
 	}
 
@@ -768,6 +841,20 @@ int grow_avl_data(OBIDMS_avl_data_p avl_data)		// TODO Lock when needed
 	header_size = (avl_data->header)->header_size;
 	file_size = header_size + new_data_size;
 
+	// Unmap the entire file before truncating it (WSL requirement)
+	if (munmap(avl_data->data, old_data_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the data of an AVL tree data file before enlarging");
+		return -1;
+	}
+	if (munmap(avl_data->header, header_size) < 0)
+	{
+		obi_set_errno(OBI_AVL_ERROR);
+		obidebug(1, "\nError munmapping the header of an AVL tree data file before enlarging");
+		return -1;
+	}
+
 	// Enlarge the file
 	if (ftruncate(avl_data_file_descriptor, file_size) < 0)
 	{
@@ -776,12 +863,19 @@ int grow_avl_data(OBIDMS_avl_data_p avl_data)		// TODO Lock when needed
 		return -1;
 	}
 
-	// Unmap and re-map the data
+	// Re-map
 
-	if (munmap(avl_data->data, old_data_size) < 0)
+	avl_data->header = mmap(NULL,
+						    header_size,
+						    PROT_READ | PROT_WRITE,
+						    MAP_SHARED,
+						    avl_data_file_descriptor,
+						    0
+						   );
+	if (avl_data->header == MAP_FAILED)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
-		obidebug(1, "\nError munmapping the data of an AVL tree data file before enlarging");
+		obidebug(1, "\nError re-mmapping the header of an AVL tree data file after enlarging the file");
 		return -1;
 	}
 
@@ -792,7 +886,6 @@ int grow_avl_data(OBIDMS_avl_data_p avl_data)		// TODO Lock when needed
 						  avl_data_file_descriptor,
 						  header_size
 						 );
-
 	if (avl_data->data == MAP_FAILED)
 	{
 		obi_set_errno(OBI_AVL_ERROR);

src/obidms.c

@@ -316,6 +316,15 @@ static int enlarge_infos_file(OBIDMS_p dms, size_t new_size)
 	multiple = 	ceil((double) new_size / (double) getpagesize());
 	rounded_new_size = multiple * getpagesize();
 
+	// Unmap the entire file before truncating it (WSL requirement)
+	if (munmap(dms->infos, (dms->infos)->file_size) < 0)
+	{
+		obi_set_errno(OBIDMS_UNKNOWN_ERROR);
+		obidebug(1, "\nError munmapping a DMS information file when enlarging");
+		close(infos_file_descriptor);
+		return -1;
+	}
+
 	// Enlarge the file
 	if (ftruncate(infos_file_descriptor, rounded_new_size) < 0)
 	{
@@ -325,15 +334,7 @@ static int enlarge_infos_file(OBIDMS_p dms, size_t new_size)
 		return -1;
 	}
 
-	// Unmap and remap the file
-	if (munmap(dms->infos, (dms->infos)->file_size) < 0)
-	{
-		obi_set_errno(OBIDMS_UNKNOWN_ERROR);
-		obidebug(1, "\nError munmapping a DMS information file when enlarging");
-		close(infos_file_descriptor);
-		return -1;
-	}
-
+	// Remap the file
 	dms->infos = mmap(NULL,
 					  rounded_new_size,
 					  PROT_READ | PROT_WRITE,
@@ -1409,6 +1410,111 @@ DIR* opendir_in_dms(OBIDMS_p dms, const char* path_name)
 }
 
 
+char* obi_dms_formatted_infos(OBIDMS_p dms, bool detailed)
+{
+	char*          dms_infos = NULL;
+	char*   	   view_infos = NULL;
+	char*          view_name = NULL;
+	char*          tax_name = NULL;
+	char*          all_tax_dir_path = NULL;
+	int            i, last_dot_pos;
+	struct dirent* dp;
+	Obiview_p      view;
+
+	// DMS name
+	dms_infos = (char*) malloc((strlen("# DMS name: ")+strlen(dms->dms_name)+strlen("\n# Views:\n")+1) * sizeof(char));
+	if (dms_infos == NULL)
+	{
+		obidebug(1, "\nError allocating memory for DMS formatted infos");
+		return NULL;
+	}
+	strcpy(dms_infos, "# DMS name: ");
+	strcat(dms_infos, dms->dms_name);
+	strcat(dms_infos, "\n# Views:\n");
+
+	// Go through views and get their infos
+	rewinddir(dms->view_directory);
+	while ((dp = readdir(dms->view_directory)) != NULL)
+	{
+		if ((dp->d_name)[0] == '.')
+			continue;
+		i=0;
+		while (i < strlen(dp->d_name))
+		{
+			if ((dp->d_name)[i] == '.')
+				last_dot_pos = i;
+			i++;
+		}
+		view_name = (char*) malloc((last_dot_pos+1) * sizeof(char));
+		if (view_name == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for a view name when getting formatted DMS infos: file %s", dp->d_name);
+			return NULL;
+		}
+		strncpy(view_name, dp->d_name, last_dot_pos);
+		view_name[last_dot_pos] = '\0';
+		view = obi_open_view(dms, view_name);
+		if (view == NULL)
+		{
+			obidebug(1, "\nError opening a view to get DMS formatted infos");
+			return NULL;
+		}
+		if (detailed)
+			view_infos = obi_view_formatted_infos(view, detailed);
+		else
+			view_infos = obi_view_formatted_infos_one_line(view);
+		if (view_infos == NULL)
+		{
+			obidebug(1, "\nError getting a view infos to get DMS formatted infos");
+			return NULL;
+		}
+		dms_infos = realloc(dms_infos, (strlen(dms_infos)+strlen(view_infos)+1) * sizeof(char));
+		if (dms_infos == NULL)
+		{
+			obidebug(1, "\nError reallocating memory for DMS formatted infos");
+			return NULL;
+		}
+		strcat(dms_infos, view_infos);
+		if (obi_save_and_close_view(view) < 0)
+		{
+			obidebug(1, "\nError closing view while getting DMS formatted infos");
+			return NULL;
+		}
+		if (detailed)
+		{
+			dms_infos = realloc(dms_infos, (strlen(dms_infos)+2) * sizeof(char));
+			strcat(dms_infos, "\n");
+		}
+	}
+
+	// Add taxonomies
+	dms_infos = realloc(dms_infos, (strlen(dms_infos)+strlen("\n# Taxonomies:\n")+1) * sizeof(char));
+	if (dms_infos == NULL)
+	{
+		obidebug(1, "\nError reallocating memory for DMS formatted infos");
+		return NULL;
+	}
+	strcat(dms_infos, "# Taxonomies:\n");
+	rewinddir(dms->tax_directory);
+	while ((dp = readdir(dms->tax_directory)) != NULL)
+	{
+		if ((dp->d_name)[0] == '.')
+			continue;
+		tax_name = dp->d_name;
+		dms_infos = realloc(dms_infos, (strlen(dms_infos)+strlen("  # ")+strlen(view_infos)+1) * sizeof(char));
+		if (dms_infos == NULL)
+		{
+			obidebug(1, "\nError reallocating memory for DMS formatted infos");
+			return NULL;
+		}
+		strcat(dms_infos, "  # ");
+		strcat(dms_infos, tax_name);
+	}
+	return dms_infos;
+}
+
+
 // TODO move somewhere else maybe
 // TODO discuss arguments
 obiversion_t obi_import_column(const char* dms_path_1, const char* dms_path_2, const char* column_name, obiversion_t version_number)
@@ -1474,8 +1580,8 @@ obiversion_t obi_import_column(const char* dms_path_1, const char* dms_path_2, c
 
 	// Create new column
 	column_2 = obi_create_column(dms_2, column_name, header_1->returned_data_type, header_1->line_count,
-								 header_1->nb_elements_per_line, header_1->elements_names, true, header_1->tuples,
-								 header_1->to_eval, new_avl_name, (header_1->associated_column).column_name,
+								 header_1->nb_elements_per_line, header_1->elements_names, true, header_1->dict_column,
+								 header_1->tuples, header_1->to_eval, new_avl_name, (header_1->associated_column).column_name,
 								 (header_1->associated_column).version, header_1->comments);
 
 	if (column_2 == NULL)
@@ -1659,6 +1765,12 @@ int obi_import_view(const char* dms_path_1, const char* dms_path_2, const char*
 	else	// Non-typed view
 		view_2 = obi_new_view(dms_2, view_name_2, NULL, NULL, (view_1->infos)->comments);
 
+	if (view_2 == NULL)
+	{
+		obidebug(1, "\nError creating the new view to import a view in a DMS");
+		return -1;
+	}
+
 	// Import line count
 	view_2->infos->line_count = view_1->infos->line_count;
 
@@ -1706,6 +1818,7 @@ int obi_import_view(const char* dms_path_1, const char* dms_path_2, const char*
 								false,
 								false,
 								false,
+								false,
 								NULL,
 								NULL,
 								-1,

src/obidms.h

@@ -40,7 +40,7 @@
                                 	 	 	 	 */
 #define MAX_NB_OPENED_INDEXERS (1000) 			/**< The maximum number of indexers open at the same time.
                                 	 	 	 	 */
-#define MAX_PATH_LEN (1024)						/**< Maximum length for the character string defining a
+#define MAX_PATH_LEN (2048)						/**< Maximum length for the character string defining a
  	 	 	 	 	 	 	 	 	 	 	 	 *	 file or directory path.
  	 	 	 	 	 	 	 	 	 	 	 	 */
 
@@ -459,6 +459,23 @@ char* obi_dms_get_full_path(OBIDMS_p dms, const char* path_name);
 DIR* opendir_in_dms(OBIDMS_p dms, const char* path_name);
 
 
+/**
+ * @brief Returns the informations of a DMS with a human readable format (dms name, taxonomies and view infos).
+ *
+ * @warning The returned pointer has to be freed by the caller.
+ *
+ * @param column A pointer on a DMS.
+ * @param detailed Whether the informations should contain detailed view infos.
+ *
+ * @returns A pointer on a character array where the formatted DMS informations are stored.
+ * @retval NULL if an error occurred.
+ *
+ * @since September 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_dms_formatted_infos(OBIDMS_p dms, bool detailed);
+
+
 /**
  * @brief Imports a column, copying it from a DMS to another DMS, and returns the version of the column in the destination DMS.
  *

src/obidms_taxonomy.c

@@ -873,7 +873,7 @@ static ecotxidx_t* read_taxonomy_idx(const char* taxa_file_name, const char* loc
 	taxa_index->buffer_size = taxa_index->count;
 
 	taxa_index->max_taxid = 0;
-	printf("Reading %d taxa...\n", count_taxa);
+	fprintf(stderr, "Reading %d taxa...\n", count_taxa);
 	for (i=0; i<count_taxa; i++)
 	{
 		readnext_ecotaxon(f_taxa, &(taxa_index->taxon[i]));
@@ -886,9 +886,9 @@ static ecotxidx_t* read_taxonomy_idx(const char* taxa_file_name, const char* loc
 	}
 
 	if (count_local_taxa > 0)
-		printf("Reading %d local taxa...\n", count_local_taxa);
+		fprintf(stderr, "Reading %d local taxa...\n", count_local_taxa);
 	else
-		printf("No local taxa\n");
+		fprintf(stderr, "No local taxa\n");
 
 	count_taxa = taxa_index->count;
 
@@ -1092,7 +1092,7 @@ static int write_ranks_idx(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* taxo
 	free(taxonomy_path);
 
 	// Create file
-	file_descriptor = open(file_name, O_RDWR | O_CREAT | O_EXCL, 0777);
+	file_descriptor = open(file_name, O_RDWR | O_CREAT, 0777);
 	if (file_descriptor < 0)
 	{
 		obi_set_errno(OBI_TAXONOMY_ERROR);
@@ -1196,7 +1196,7 @@ static int write_taxonomy_idx(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* t
 	free(taxonomy_path);
 
 	// Create file
-	file_descriptor = open(file_name, O_RDWR | O_CREAT | O_EXCL, 0777);
+	file_descriptor = open(file_name, O_RDWR | O_CREAT, 0777);
 	if (file_descriptor < 0)
 	{
 		obi_set_errno(OBI_TAXONOMY_ERROR);
@@ -1472,7 +1472,7 @@ static int write_names_idx(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* taxo
 	free(taxonomy_path);
 
 	// Create file
-	file_descriptor = open(file_name, O_RDWR | O_CREAT | O_EXCL, 0777);
+	file_descriptor = open(file_name, O_RDWR | O_CREAT, 0777);
 	if (file_descriptor < 0)
 	{
 		obi_set_errno(OBI_TAXONOMY_ERROR);
@@ -1760,7 +1760,7 @@ static int write_merged_idx(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* tax
 	free(taxonomy_path);
 
 	// Create file
-	file_descriptor = open(file_name, O_RDWR | O_CREAT | O_EXCL, 0777);
+	file_descriptor = open(file_name, O_RDWR | O_CREAT, 0777);
 	if (file_descriptor < 0)
 	{
 		obi_set_errno(OBI_TAXONOMY_ERROR);
@@ -2463,6 +2463,32 @@ int read_merged_dmp(const char* taxdump, OBIDMS_taxonomy_p tax, int32_t* delnode
 				return -1;
 			}
 
+			// Write the rest of the taxa from the current taxa list
+			while (nT < (tax->taxa)->count)
+			{
+			// Add element from taxa list
+				// Enlarge structure if needed
+				if (n == buffer_size)
+				{
+					buffer_size = buffer_size * 2;
+					tax->merged_idx = (ecomergedidx_t*) realloc(tax->merged_idx, sizeof(ecomergedidx_t) + sizeof(ecomerged_t) * buffer_size);
+					if (tax->merged_idx == NULL)
+					{
+						obi_set_errno(OBI_MALLOC_ERROR);
+						obidebug(1, "\nError reallocating memory for a taxonomy structure");
+						closedir(tax_dir);
+						fclose(file);
+						return -1;
+					}
+				}
+
+				(tax->merged_idx)->merged[n].taxid = (tax->taxa)->taxon[nT].taxid;
+				(tax->merged_idx)->merged[n].idx = nT;
+
+				nT++;
+				n++;
+			}
+
 			// Store count
 			(tax->merged_idx)->count = n;
 
@@ -3224,47 +3250,48 @@ OBIDMS_taxonomy_p obi_read_taxonomy(OBIDMS_p dms, const char* taxonomy_name, boo
 }
 
 
-int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* tax_name)
+int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* tax_name, bool update)
 {
 	char* taxonomy_path;
 
-	// Build the taxonomy directory path
-	taxonomy_path = get_taxonomy_path(dms, tax_name);
-	if (taxonomy_path == NULL)
-		return -1;
-
-	// Try to create the directory
-	if (mkdir(taxonomy_path, 00777) < 0)
-	{
-		if (errno == EEXIST)
-			obidebug(1, "\nA taxonomy already exists with this name.");
-		obidebug(1, "\nProblem creating a new taxonomy directory");
+	if (!update) {
+		// Build the taxonomy directory path
+		taxonomy_path = get_taxonomy_path(dms, tax_name);
+		if (taxonomy_path == NULL)
+			return -1;
+		// Try to create the directory
+		if (mkdir(taxonomy_path, 00777) < 0)
+		{
+			if (errno == EEXIST)
+				obidebug(1, "\nA taxonomy already exists with this name.");
+			obidebug(1, "\nProblem creating a new taxonomy directory");
+			free(taxonomy_path);
+			return -1;
+		}
 		free(taxonomy_path);
-		return -1;
 	}
 
-	free(taxonomy_path);
-
-    if (write_ranks_idx(dms, tax, tax_name) < 0)
-        return -1;
-    if (write_taxonomy_idx(dms, tax, tax_name) < 0)
-        return -1;
-    if (write_names_idx(dms, tax, tax_name) < 0)
-    	return -1;
-    if (write_merged_idx(dms, tax, tax_name) < 0)
-    	return -1;
-    // Check if there are local taxa (if so last taxon is local)
+	if (write_ranks_idx(dms, tax, tax_name) < 0)
+		return -1;
+	if (write_taxonomy_idx(dms, tax, tax_name) < 0)
+		return -1;
+	if (write_names_idx(dms, tax, tax_name) < 0)
+		return -1;
+	if (write_merged_idx(dms, tax, tax_name) < 0)
+		return -1;
+	// Write preferred names if there are some
+	if (tax->preferred_names != NULL)
+	{
+		if (write_preferred_names_idx(dms, tax, tax_name) < 0)
+			return -1;
+	}
+    // Write local taxa if there are some
     if ((tax->taxa)->local_count > 0)
     {
     	if (write_local_taxonomy_idx(dms, tax, tax_name) < 0)
     		return -1;
     }
-    // Write preferred names if there are some
-    if (tax->preferred_names != NULL)
-    {
-    	if (write_preferred_names_idx(dms, tax, tax_name) < 0)
-    		return -1;
-    }
+
 	return 0;
 }
 
@@ -3276,16 +3303,17 @@ int obi_close_taxonomy(OBIDMS_taxonomy_p taxonomy)
 	if (taxonomy)
 	{
 		// Update local informations (local taxa and preferred names) if there are any
-		if ((taxonomy->taxa)->local_count > 0)
-		{
-			if (taxonomy->dms == NULL)
-			{
-				obi_set_errno(OBI_TAXONOMY_ERROR);
-				obidebug(1, "\nError closing a taxonomy with local files but no DMS associated (probably read directly from taxdump)");		// TODO discuss
-			}
-			if (write_local_taxonomy_idx(taxonomy->dms, taxonomy, taxonomy->tax_name) < 0)
-				return -1;
-		}
+		// Done with write_taxo, edits all needed files. Only ldx file was edited in OBI1 but it led to issues. Discussable
+//		if ((taxonomy->taxa)->local_count > 0)
+//		{
+//			if (taxonomy->dms == NULL)
+//			{
+//				obi_set_errno(OBI_TAXONOMY_ERROR);
+//				obidebug(1, "\nError closing a taxonomy with local files but no DMS associated (probably read directly from taxdump)");		// TODO discuss
+//			}
+//			if (write_local_taxonomy_idx(taxonomy->dms, taxonomy, taxonomy->tax_name) < 0)
+//				return -1;
+//		}
 
 		// Write preferred names if there are some
 		if (taxonomy->preferred_names)
@@ -3351,9 +3379,10 @@ int obi_close_taxonomy(OBIDMS_taxonomy_p taxonomy)
 int obi_taxo_add_local_taxon(OBIDMS_taxonomy_p tax, const char* name, const char* rank_name, int32_t parent_taxid, int32_t min_taxid)
 {
 	int32_t    taxid;
+	int32_t    count;
 	ecotx_t*   taxon;
 	int        i;
-//	econame_t* name_struct;
+	econame_t* name_struct;
 
 	// Enlarge the structure memory for a new taxon
 	tax->taxa = (ecotxidx_t*) realloc(tax->taxa, sizeof(ecotxidx_t) + sizeof(ecotx_t) * (((tax->taxa)->count) + 1));
@@ -3415,42 +3444,65 @@ int obi_taxo_add_local_taxon(OBIDMS_taxonomy_p tax, const char* name, const char
 	((tax->taxa)->local_count)++;
 	(tax->taxa)->buffer_size = (tax->taxa)->count;
 
-//	// Add new name in names structure     // Commented because the new name was not added in the .ndx file in the OBITools1
-//	// Allocate memory for new name
-//	tax->names = (econameidx_t*) realloc(tax->names, sizeof(econameidx_t) + sizeof(econame_t) * ((tax->names)->count + 1));
-//	if (tax->names == NULL)
-//	{
-//		obi_set_errno(OBI_MALLOC_ERROR);
-//		obidebug(1, "\nError reallocating memory for a taxonomy structure to add a new taxon");
-//		return -1;
-//	}
-//
-//	// Add new name
-//	name_struct = (tax->names)->names + ((tax->names)->count);
-//	name_struct->name = (char*) malloc((strlen(name) + 1) * sizeof(char));
-//	if (name_struct->name == NULL)
-//	{
-//		obi_set_errno(OBI_MALLOC_ERROR);
-//		obidebug(1, "\nError allocating memory for a taxon name to add a new taxon");
-//		return -1;
-//	}
-//	strcpy(name_struct->name, name);
-//	name_struct->class_name = (char*) malloc((strlen("scientific name") + 1) * sizeof(char));
-//	if (name_struct->class_name == NULL)
-//	{
-//		obi_set_errno(OBI_MALLOC_ERROR);
-//		obidebug(1, "\nError allocating memory for a taxon class name to add a new taxon");
-//		return -1;
-//	}
-//	strcpy(name_struct->class_name, "scientific name");
-//	name_struct->is_scientific_name = true;
-//	name_struct->taxon = ((tax->taxa)->taxon) + ((tax->taxa)->count) - 1;
-//
-//	// Sort names in alphabetical order
-//	qsort((tax->names)->names, (tax->names)->count, sizeof(econame_t), cmp_names);
-//
-//	// Update name count
-//	((tax->names)->count)++;
+	// Add new name in names structure // On the OBI1, the new name was not added in the .ndx file but it could create issues
+	// Allocate memory for new name
+	tax->names = (econameidx_t*) realloc(tax->names, sizeof(econameidx_t) + sizeof(econame_t) * ((tax->names)->count + 1));
+	if (tax->names == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError reallocating memory for a taxonomy structure to add a new taxon");
+		return -1;
+	}
+
+	// Add new name
+	name_struct = (tax->names)->names + ((tax->names)->count);
+	name_struct->name = (char*) malloc((strlen(name) + 1) * sizeof(char));
+	if (name_struct->name == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError allocating memory for a taxon name to add a new taxon");
+		return -1;
+	}
+	strcpy(name_struct->name, name);
+	name_struct->class_name = (char*) malloc((strlen("scientific name") + 1) * sizeof(char));
+	if (name_struct->class_name == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError allocating memory for a taxon class name to add a new taxon");
+		return -1;
+	}
+	strcpy(name_struct->class_name, "scientific name");
+	name_struct->is_scientific_name = true;
+	name_struct->taxon = ((tax->taxa)->taxon) + ((tax->taxa)->count) - 1;
+
+	// Update name count
+	((tax->names)->count)++;
+
+	// Sort names in alphabetical order
+	qsort((tax->names)->names, (tax->names)->count, sizeof(econame_t), cmp_names);
+
+	// Add to merged index
+	tax->merged_idx = (ecomergedidx_t*) realloc(tax->merged_idx, sizeof(ecomergedidx_t) + sizeof(ecomerged_t) * ((tax->merged_idx)->count + 1));
+	if (tax->merged_idx == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError reallocating memory for a taxonomy structure");
+		return -1;
+	}
+
+	count = (tax->merged_idx)->count;
+	(tax->merged_idx)->count = count + 1;
+	(tax->merged_idx)->merged[count].taxid = taxid;
+	(tax->merged_idx)->merged[count].idx = taxon->idx;
+
+	//fprintf(stderr, "\nEntered in merged taxon.idx=%d", (tax->merged_idx)->merged[(tax->merged_idx)->count -1].idx);
+	//fprintf(stderr, "\nEntered in merged taxon.taxid=%d", (tax->merged_idx)->merged[(tax->merged_idx)->count -1].taxid);
+	//fprintf(stderr, "\nEntered in merged at %d", (tax->merged_idx)->count -1);
+	//taxon = obi_taxo_get_taxon_with_taxid(tax, taxid);
+	//fprintf(stderr, "\ntaxon=%x", taxon);
+	//fprintf(stderr, "\ntaxon.taxid=%d", taxon->taxid);
+	//fprintf(stderr, "\ntaxon.name=%s", taxon->name);
+	//fprintf(stderr, "\ntaxon.idx=%d\n\n", ((tax->merged_idx)->count));
 
 	return taxid;
 }
@@ -3521,11 +3573,12 @@ int obi_taxo_add_preferred_name_with_taxon(OBIDMS_taxonomy_p tax, ecotx_t* taxon
 	name_struct->is_scientific_name = false;
 	name_struct->taxon = taxon;
 
+	// Update preferred name count
+	((tax->preferred_names)->count)++;
+
 	// Sort preferred names in alphabetical order
 	qsort((tax->preferred_names)->names, (tax->preferred_names)->count, sizeof(econame_t), cmp_names);
 
-	// Update preferred name count
-	((tax->preferred_names)->count)++;
 
 	return 0;
 }
@@ -3643,12 +3696,26 @@ ecotx_t* obi_taxo_get_taxon_with_taxid(OBIDMS_taxonomy_p taxonomy, int32_t taxid
 	else if (indexed_taxon->idx == -1)
 		current_taxon = NULL;	// TODO discuss what to do when old deleted taxon
 	else
+	{
 		current_taxon = (taxonomy->taxa->taxon)+(indexed_taxon->idx);
-
+		//fprintf(stderr, "\n>>>idx %d, taxid %d<<<\n", indexed_taxon->idx, indexed_taxon->taxid);
+	}
 	return current_taxon;
 }
 
 
+char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
+{
+	return (((taxonomy->names)->names)[idx]).name;
+}
+
+
+ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
+{
+	return (((taxonomy->names)->names)[idx]).taxon;
+}
+
+
 int obi_taxo_is_taxon_under_taxid(ecotx_t* taxon, int32_t other_taxid)		// TODO discuss that this doesn't work with deprecated taxids
 {
 	ecotx_t* next_parent;

src/obidms_taxonomy.h

@@ -75,7 +75,7 @@ typedef struct {
  	 	 	 	 	 	 	 */
 	int32_t max_taxid;		/**< Maximum taxid existing in the taxon index.
 	 	 	 	 	 	 	 */
-	int32_t buffer_size;	/**< Number of taxa.	// TODO kept this but not sure of its use
+	int32_t buffer_size;	/**< .	// TODO kept this but not sure of its use
  	 	 	 	 	 	 	 */
 	ecotx_t taxon[];		/**< Taxon array.
  	 	 	 	 	 	 	 */
@@ -239,6 +239,7 @@ OBIDMS_taxonomy_p obi_read_taxonomy(OBIDMS_p dms, const char* taxonomy_name, boo
  * @param dms A pointer on the DMS to which the taxonomy belongs.
  * @param tax A pointer on the taxonomy structure.
  * @param tax_name The name (prefix) of the taxonomy.
+ * @param update Whether files should be rewritten or if it's a new taxonomy (set to true e.g. after adding local taxa).
  *
  * @returns An integer value indicating the success of the operation.
  * @retval 0 on success.
@@ -247,7 +248,7 @@ OBIDMS_taxonomy_p obi_read_taxonomy(OBIDMS_p dms, const char* taxonomy_name, boo
  * @since 2016
  * @author Celine Mercier (celine.mercier@metabarcoding.org)
  */
-int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* tax_name);
+int obi_write_taxonomy(OBIDMS_p dms, OBIDMS_taxonomy_p tax, const char* tax_name, bool update);
 
 
 /**
@@ -447,8 +448,51 @@ ecotx_t* obi_taxo_get_superkingdom(ecotx_t* taxon, OBIDMS_taxonomy_p taxonomy);
 const char* obi_taxo_rank_index_to_label(int32_t rank_idx, ecorankidx_t* ranks);
 
 
-// TODO
+/**
+ * @brief Function checking whether a taxid is included in a subset of the taxonomy.
+ *
+ * @param taxonomy A pointer on the taxonomy structure.
+ * @param restrict_to_taxids An array of taxids. The researched taxid must be under at least one of those array taxids.
+ * @param count Number of taxids in restrict_to_taxids.
+ * @param taxid The taxid to check.
+ *
+ * @returns A value indicating whether the taxid is included in the chosen subset of the taxonomy.
+ * @retval 0 if the taxid is not included in the subset of the taxonomy.
+ * @retval 1 if the taxid is included in the subset of the taxonomy.
+ *
+ * @since October 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
 int obi_taxo_is_taxid_included(OBIDMS_taxonomy_p taxonomy,
 			     			   int32_t* restrict_to_taxids,
 				    		   int32_t count,
 					    	   int32_t taxid);
+
+
+/**
+ * @brief Function returning the name of a taxon from its index in the taxonomy name index (econameidx_t).
+ *
+ * @param taxonomy A pointer on the taxonomy structure.
+ * @param idx The index at which the name is in the taxonomy name index (econameidx_t).
+ *
+ * @returns The taxon name.
+ *
+ * @since October 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx);
+
+
+/**
+ * @brief Function returning a taxon structure from its index in the taxonomy name index (econameidx_t).
+ *
+ * @param taxonomy A pointer on the taxonomy structure.
+ * @param idx The index at which the taxon is in the taxonomy name index (econameidx_t).
+ *
+ * @returns The taxon structure.
+ *
+ * @since October 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx);
+

src/obidmscolumn.c

@@ -1024,6 +1024,7 @@ OBIDMS_column_p obi_create_column(OBIDMS_p     dms,
 								  index_t      nb_elements_per_line,
 								  char*        elements_names,
 								  bool		   elt_names_formatted,
+								  bool         dict_column,
 								  bool		   tuples,
 								  bool         to_eval,
 								  const char*  indexer_name,
@@ -1282,6 +1283,7 @@ OBIDMS_column_p obi_create_column(OBIDMS_p     dms,
 	header->nb_elements_per_line          = nb_elements_per_line;
 	header->stored_data_type     		  = stored_data_type;
 	header->returned_data_type   		  = returned_data_type;
+	header->dict_column     			  = dict_column;
 	header->tuples     					  = tuples;
 	header->to_eval                       = to_eval;
 	header->creation_date 		 		  = time(NULL);
@@ -1312,19 +1314,10 @@ OBIDMS_column_p obi_create_column(OBIDMS_p     dms,
 		return NULL;
 	}
 
-	// Store the associated column reference if needed // TODO discuss cases
-	if (data_type == OBI_QUAL)
+	// Store the associated column reference if needed
+	if ((associated_column_name != NULL) && (*associated_column_name != '\0'))
 	{
-		if ((associated_column_name == NULL) || (*associated_column_name == '\0'))
-		{
-			obidebug(1, "\nError: The name of the associated column when creating a new column is NULL");
-			munmap(new_column->header, header_size);
-			close(column_file_descriptor);
-			free(new_column);
-			return NULL;
-		}
 		strcpy((header->associated_column).column_name, associated_column_name);
-
 		if (associated_column_version == -1)
 		{
 			obidebug(1, "\nError: The version of the associated column when creating a new column is not defined");
@@ -1336,6 +1329,7 @@ OBIDMS_column_p obi_create_column(OBIDMS_p     dms,
 		(header->associated_column).version = associated_column_version;
 	}
 
+
 	// If the data type is OBI_STR, OBI_SEQ or OBI_QUAL, the associated obi_indexer is opened or created
 	if ((returned_data_type == OBI_STR) || (returned_data_type == OBI_SEQ) || (returned_data_type == OBI_QUAL) || tuples)
 	{
@@ -1619,6 +1613,7 @@ OBIDMS_column_p obi_clone_column(OBIDMS_p         dms,
 								   nb_elements_per_line,
 								   (column_to_clone->header)->elements_names,
 								   true,
+								   (column_to_clone->header)->dict_column,
 								   (column_to_clone->header)->tuples,
 								   (column_to_clone->header)->to_eval,
 								   (column_to_clone->header)->indexer_name,
@@ -1733,16 +1728,32 @@ int obi_close_column(OBIDMS_column_p column)
 int obi_clone_column_indexer(OBIDMS_column_p column)
 {
 	char* new_indexer_name;
+	int i;
 
-	new_indexer_name = obi_build_indexer_name((column->header)->name, (column->header)->version);
-	if (new_indexer_name == NULL)
-		return -1;
-
-	column->indexer = obi_clone_indexer(column->indexer, new_indexer_name);	// TODO Need to lock this somehow?
-	if (column->indexer == NULL)
+	i=0;
+	while (true) // find avl name not already used
 	{
-		obidebug(1, "\nError cloning a column's indexer to make it writable");
-		return -1;
+		new_indexer_name = obi_build_indexer_name((column->header)->name, ((column->header)->version)+i);
+		if (new_indexer_name == NULL)
+			return -1;
+
+		column->indexer = obi_clone_indexer(column->indexer, new_indexer_name);	// TODO Need to lock this somehow?
+		if (column->indexer == NULL)
+		{
+			if (errno == EEXIST)
+			{
+				free(new_indexer_name);
+				i++;
+			}
+			else
+			{
+				free(new_indexer_name);
+				obidebug(1, "\nError cloning a column's indexer to make it writable");
+				return -1;
+			}
+		}
+		else
+			break;
 	}
 
 	strcpy((column->header)->indexer_name, new_indexer_name);
@@ -1758,6 +1769,7 @@ int obi_truncate_column(OBIDMS_column_p column)	// TODO is it necessary to unmap
 {
 	size_t  file_size;
 	size_t  data_size;
+	size_t  header_size;
 	index_t new_line_count;
 	double  multiple;
 	int 	column_file_descriptor;
@@ -1780,6 +1792,8 @@ int obi_truncate_column(OBIDMS_column_p column)	// TODO is it necessary to unmap
 
 	data_size = obi_array_sizeof((column->header)->stored_data_type, new_line_count, (column->header)->nb_elements_per_line);
 
+	header_size = (column->header)->header_size;
+
 	// Check that it is actually greater than the current data size, otherwise no need to truncate
 	if ((column->header)->data_size == data_size)
 		return 0;
@@ -1844,7 +1858,7 @@ int obi_truncate_column(OBIDMS_column_p column)	// TODO is it necessary to unmap
 		return -1;
 	}
 
-	// Unmap the data before truncating the file
+	// Unmap the entire file before truncating it (WSL requirement)
 	if (munmap(column->data, (column->header)->data_size) < 0)
 	{
 		obi_set_errno(OBICOL_UNKNOWN_ERROR);
@@ -1852,9 +1866,16 @@ int obi_truncate_column(OBIDMS_column_p column)	// TODO is it necessary to unmap
 		close(column_file_descriptor);
 		return -1;
 	}
+	if (munmap(column->header, header_size) < 0)
+	{
+		obi_set_errno(OBICOL_UNKNOWN_ERROR);
+		obidebug(1, "\nError munmapping the header of a column before truncating");
+		close(column_file_descriptor);
+		return -1;
+	}
 
 	// Truncate the column file
-	file_size = (column->header)->header_size + data_size;
+	file_size = header_size + data_size;
 	if (ftruncate(column_file_descriptor, file_size) < 0)
 	{
 		obi_set_errno(OBICOL_UNKNOWN_ERROR);
@@ -1863,13 +1884,30 @@ int obi_truncate_column(OBIDMS_column_p column)	// TODO is it necessary to unmap
 		return -1;
 	}
 
-	// Remap the data
+	// Remap the header and the data
+
+	column->header = mmap(NULL,
+						  header_size,
+						  PROT_READ | PROT_WRITE,
+						  MAP_SHARED,
+						  column_file_descriptor,
+						  0
+					     );
+
+	if (column->header == MAP_FAILED)
+	{
+		obi_set_errno(OBICOL_UNKNOWN_ERROR);
+		obidebug(1, "\nError re-mmapping the header of a column after truncating");
+		close(column_file_descriptor);
+		return -1;
+	}
+
 	column->data = mmap(NULL,
 						data_size,
 						PROT_READ | PROT_WRITE,
 						MAP_SHARED,
 						column_file_descriptor,
-						(column->header)->header_size
+						header_size
 					   );
 
 	if (column->data == MAP_FAILED)
@@ -2423,17 +2461,81 @@ char* obi_get_formatted_elements_names(OBIDMS_column_p column)
 }
 
 
-char* obi_column_formatted_infos(OBIDMS_column_p column)
+char* obi_column_formatted_infos(OBIDMS_column_p column, bool detailed)
 {
-	char* column_infos;
-	char* elt_names;
-
-	column_infos = malloc(1024 * sizeof(char));
+	char* column_infos = NULL;
+	char* elt_names = NULL;
+	char* data_type_str = NULL;
+	char* comments = NULL;
 
+	// Get element names informations
 	elt_names = obi_get_formatted_elements_names(column);
+	if (elt_names == NULL)
+	{
+		obidebug(1, "\nError getting formatted elements names for formatted columns infos");
+		return NULL;
+	}
 
+	// Get data type informations
+	data_type_str = name_data_type((column->header)->returned_data_type);
+	if (data_type_str == NULL)
+	{
+		obidebug(1, "\nError getting formatted data type for formatted columns infos");
+		return NULL;
+	}
+
+	// Get commments if detailed informations required
+	if (detailed)
+		comments = (column->header)->comments;
+
+	// Build the string of formatted infos, allocating memory as needed
+
+	// Data type
+	column_infos = (char*) malloc((strlen("data type: ")+strlen(data_type_str)+1) * sizeof(char));
+	if (column_infos == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError allocating memory for formatted column infos");
+		return NULL;
+	}
+
+	strcpy(column_infos, "data type: ");
+	strcat(column_infos, data_type_str);
+
+	// Element names if more than 1
+	if ((column->header)->nb_elements_per_line > 1)
+	{
+		column_infos = realloc(column_infos, (strlen(column_infos)+strlen(", elements: ")+strlen(elt_names)+1) * sizeof(char));
+		if (column_infos == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for formatted column infos");
+			return NULL;
+		}
+
+		strcat(column_infos, ", elements: ");
+		strcat(column_infos, elt_names);
+	}
+
+	if (detailed && (strlen(comments)>2)) // Add all comments if required and not empty
+	{
+		column_infos = realloc(column_infos, (strlen(column_infos)+strlen("\nComments:\n")+strlen(comments)+1) * sizeof(char));
+		if (column_infos == NULL)
+		{
+			obi_set_errno(OBI_MALLOC_ERROR);
+			obidebug(1, "\nError allocating memory for formatted column infos");
+			return NULL;
+		}
+
+		strcat(column_infos, "\nComments:\n");
+		strcat(column_infos, comments);
+	}
+
+//  "data type: OBI_TYPE, element names: [formatted element names](, all comments)"
 
 	free(elt_names);
+	free(data_type_str);
+
 	return column_infos;
 }
 
@@ -2480,7 +2582,6 @@ int obi_column_prepare_to_set_value(OBIDMS_column_p column, index_t line_nb, ind
 }
 
 
-
 int obi_column_prepare_to_get_value(OBIDMS_column_p column, index_t line_nb)
 {
 	if ((line_nb+1) > ((column->header)->line_count))

src/obidmscolumn.h

@@ -36,7 +36,7 @@
 										   	   	 */
 #define COLUMN_GROWTH_FACTOR (2)	 	  		/**< The growth factor when a column is enlarged.
                                 	   	   	   	 */
-#define MAXIMUM_LINE_COUNT (1000000000)   		/**< The maximum line count for the data of a column (1E9). //TODO
+#define MAXIMUM_LINE_COUNT (1000000000000)      /**< The maximum line count for the data of a column (1E12). //TODO
                                 	       	   	 */
 #define COMMENTS_MAX_LENGTH (4096)        		/**< The maximum length for comments.
  	 	 	 	 	 	 	 	 	       	   	 */
@@ -77,6 +77,8 @@ typedef struct OBIDMS_column_header {
 	OBIType_t			stored_data_type;		    						/**< Type of the data that is actually stored in the data
 															 	 	 	 	 *   part of the column.
 															 	 	 	 	 */
+	bool                dict_column;										/**< Whether the column contains dictionary-like values.
+	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 */
 	bool				tuples;												/**< A boolean indicating whether the column contains indices referring to indexed tuples.
 																			 */
 	bool				to_eval;											/**< A boolean indicating whether the column contains expressions that should be evaluated
@@ -249,6 +251,7 @@ size_t obi_calculate_header_size(index_t nb_elements_per_line, int64_t elts_name
  * @param elements_names The names of the elements with ';' as separator (no terminal ';'),
  *                       NULL or "" if the default names are to be used ("0\01\02\0...\0n").
  * @param elt_names_formatted Whether the separator for the elements names is ';' (false), or '\0' (true, as formatted by format_elements_names()).
+ * @param dict_column A boolean indicating whether the column should contain dictionary-like values.
  * @param tuples A boolean indicating whether the column should contain indices referring to indexed tuples.
  * @param to_eval A boolean indicating whether the column contains expressions that should be evaluated
  *                (typically OBI_STR columns containing character strings to be evaluated by Python).
@@ -271,6 +274,7 @@ OBIDMS_column_p obi_create_column(OBIDMS_p     dms,
 								  index_t      nb_elements_per_line,
 								  char*        elements_names,
 								  bool		   elt_names_formatted,
+								  bool         dict_column,
 								  bool         tuples,
 								  bool         to_eval,
 								  const char*  indexer_name,
@@ -505,12 +509,37 @@ index_t obi_column_get_element_index_from_name(OBIDMS_column_p column, const cha
 char* obi_get_elements_names(OBIDMS_column_p column);
 
 
-// TODO
-//char* obi_get_formatted_elements_names(OBIDMS_column_p column);
+/**
+ * @brief Recovers the elements names of the lines of a column with a human readable format ("0; 1; 2; ...; n\0").
+ *
+ * @warning The returned pointer has to be freed by the caller.
+ *
+ * @param column A pointer on an OBIDMS column.
+ *
+ * @returns A pointer on a character array where the elements names are stored.
+ * @retval NULL if an error occurred.
+ *
+ * @since September 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_get_formatted_elements_names(OBIDMS_column_p column);
 
 
-// TODO
-//char* obi_column_formatted_infos(OBIDMS_column_p column);
+/**
+ * @brief Returns the informations of a column with a human readable format (data type, element names, comments).
+ *
+ * @warning The returned pointer has to be freed by the caller.
+ *
+ * @param column A pointer on an OBIDMS column.
+ * @param detailed Whether the informations should contain column comments or just data type and element names.
+ *
+ * @returns A pointer on a character array where the formatted column informations are stored.
+ * @retval NULL if an error occurred.
+ *
+ * @since September 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_column_formatted_infos(OBIDMS_column_p column, bool detailed);
 
 
 /**

src/obisig.h

@@ -25,7 +25,7 @@
  * @author Celine Mercier (celine.mercier@metabarcoding.org)
  *
  */
-bool volatile keep_running;
+extern bool volatile keep_running;
 void sig_handler(int signum);
 
 

src/obitypes.h

@@ -29,6 +29,8 @@
 #define OBIQual_int_NA (NULL)		/**< NA value for the type OBI_QUAL if the quality is in integer format */
 #define OBITuple_NA (NULL)			/**< NA value for tuples of any type */
 
+#define OBI_INT_MAX (INT32_MAX)     /**< Maximum value for the type OBI_INT */
+
 
 /**
  * @brief enum for the boolean OBIType.

src/obiview.c

@@ -17,6 +17,7 @@
 #include <sys/mman.h>
 #include <inttypes.h>
 #include <math.h>
+#include <time.h>
 //#include <ctype.h>
 
 #include "obiview.h"
@@ -254,11 +255,15 @@ static int update_lines(Obiview_p view, index_t line_count);
 /**
  * @brief Internal function to clone a column in the context of a view.
  *
+ * Used to edit a closed column.
+ *
  * Clones with the right line selection and replaces the cloned columns with the new ones in the view.
  * If there is a line selection, all columns have to be cloned, otherwise only the column of interest is cloned.
  *
  * @param view A pointer on the view.
  * @param column_name The name of the column in the view that should be cloned.
+ * @param clone_associated Whether to clone the associated column
+ *  	  (should always be true except when calling from the function itself to avoid infinite recursion).
  *
  * @returns A pointer on the new column.
  * @retval NULL if an error occurred.
@@ -266,7 +271,7 @@ static int update_lines(Obiview_p view, index_t line_count);
  * @since February 2016
  * @author Celine Mercier (celine.mercier@metabarcoding.org)
  */
-static OBIDMS_column_p clone_column_in_view(Obiview_p view, const char* column_name);
+static OBIDMS_column_p clone_column_in_view(Obiview_p view, const char* column_name, bool clone_associated);
 
 
 /**
@@ -633,6 +638,15 @@ static int enlarge_view_file(Obiview_p view, size_t new_size)
 	multiple = 	ceil((double) new_size / (double) getpagesize());
 	rounded_new_size = multiple * getpagesize();
 
+	// Unmap the entire file before truncating it (WSL requirement)
+	if (munmap(view->infos, (view->infos)->file_size) < 0)
+	{
+		obi_set_errno(OBIVIEW_ERROR);
+		obidebug(1, "\nError munmapping a view file when enlarging");
+		close(obiview_file_descriptor);
+		return -1;
+	}
+
 	// Enlarge the file
 	if (ftruncate(obiview_file_descriptor, rounded_new_size) < 0)
 	{
@@ -642,15 +656,7 @@ static int enlarge_view_file(Obiview_p view, size_t new_size)
 		return -1;
 	}
 
-	// Unmap and remap the file
-	if (munmap(view->infos, (view->infos)->file_size) < 0)
-	{
-		obi_set_errno(OBIVIEW_ERROR);
-		obidebug(1, "\nError munmapping a view file when enlarging");
-		close(obiview_file_descriptor);
-		return -1;
-	}
-
+	// Remap the file
 	view->infos = mmap(NULL,
 					   rounded_new_size,
 					   PROT_READ | PROT_WRITE,
@@ -845,7 +851,7 @@ static int update_lines(Obiview_p view, index_t line_count)
 		// Clone the column first if needed
 		if (!(column->writable))
 		{
-			column = clone_column_in_view(view, (((view->infos)->column_references)[i]).alias);
+			column = clone_column_in_view(view, (((view->infos)->column_references)[i]).alias, true);
 			if (column == NULL)
 			{
 				obidebug(1, "\nError cloning a column in a view when updating its line count");
@@ -870,12 +876,14 @@ static int update_lines(Obiview_p view, index_t line_count)
 }
 
 
-static OBIDMS_column_p clone_column_in_view(Obiview_p view, const char* column_name)
+static OBIDMS_column_p clone_column_in_view(Obiview_p view, const char* column_name, bool clone_associated)
 {
-	int i;
+	int i, j;
 	OBIDMS_column_p column = NULL;
 	OBIDMS_column_p new_column = NULL;
 	OBIDMS_column_p column_buffer;
+	OBIDMS_column_p associated_cloned_column = NULL;
+	char* associated_column_alias = NULL;
 
 	// Check that the view is not read-only
 	if (view->read_only)
@@ -916,11 +924,62 @@ static OBIDMS_column_p clone_column_in_view(Obiview_p view, const char* column_n
 				return NULL;
 			}
 
+			// Look for associated column to clone and reassociate
+			if ((column_buffer->header->associated_column).column_name[0] != '\0')
+			{
+				// Get the associated column alias
+				j=0;
+				while (((strcmp((((view->infos)->column_references)[j]).column_refs.column_name, (column_buffer->header->associated_column).column_name)) ||
+						((((view->infos)->column_references)[j]).column_refs.version != (column_buffer->header->associated_column).version)) &&
+						j<(view->infos)->column_count)  // TODO function for that
+					j++;
+
+				if (j == (view->infos)->column_count) // not found
+				{
+					obi_set_errno(OBIVIEW_ERROR);
+					obidebug(1, "\nCould not find associated column when cloning a column for editing");
+					return NULL;
+				}
+
+				// No line selection: only this column is cloned, clone and reassociate the associated column
+				if ((view->line_selection == NULL) && clone_associated)
+				{
+					associated_column_alias = (((view->infos)->column_references)[j]).alias;
+					// Clone the associated column
+					associated_cloned_column = clone_column_in_view(view, associated_column_alias, false);
+					// Reassociate both ways
+					strcpy((associated_cloned_column->header->associated_column).column_name, column->header->name);
+					(associated_cloned_column->header->associated_column).version = column->header->version;
+					strcpy((column->header->associated_column).column_name, associated_cloned_column->header->name);
+					(column->header->associated_column).version = associated_cloned_column->header->version;
+				}
+				else
+				{
+					// Line selection: all columns are cloned, check if associated column has been cloned previously (it precedes this one in the list) to reassociate
+					if (j < i)
+					{
+						// Get pointer to associated column
+						associated_cloned_column = *((OBIDMS_column_p*)ll_get(view->columns, j));
+						if (associated_cloned_column == NULL)
+						{
+							obi_set_errno(OBIVIEW_ERROR);
+							obidebug(1, "\nError getting a column to clone from the linked list of column pointers of a view");
+							return NULL;
+						}
+						// Reassociate both ways
+						strcpy((associated_cloned_column->header->associated_column).column_name, column->header->name);
+						(associated_cloned_column->header->associated_column).version = column->header->version;
+						strcpy((column->header->associated_column).column_name, associated_cloned_column->header->name);
+						(column->header->associated_column).version = associated_cloned_column->header->version;
+					}
+				}
+			}
+
 			// Close old cloned column
 			obi_close_column(column_buffer);
 
 			if (!strcmp((((view->infos)->column_references)[i]).alias, column_name))
-				// Found the column to return
+				// Get the column to return
 				new_column = column;
 		}
 	}
@@ -1128,6 +1187,7 @@ static int close_view(Obiview_p view)
 			obidebug(1, "\nError getting a column to close from the linked list of column pointers of a view");
 			return -1;
 		}
+
 		if (obi_close_column(column) < 0)
 		{
 			obidebug(1, "\nError closing a column while closing a view");
@@ -1193,7 +1253,7 @@ static int prepare_to_set_value_in_column(Obiview_p view, OBIDMS_column_p* colum
 			return -1;
 		}
 
-		(*column_pp) = clone_column_in_view(view, column_name);
+		(*column_pp) = clone_column_in_view(view, column_name, true);
 		if ((*column_pp) == NULL)
 		{
 			obidebug(1, "\nError trying to clone a column to modify it");
@@ -1653,7 +1713,7 @@ Obiview_p obi_new_view(OBIDMS_p dms, const char* view_name, Obiview_p view_to_cl
 		// If there is a new line selection, build it by combining it with the one from the view to clone if there is one
 		else if (line_selection != NULL)
 		{
-			view->line_selection = obi_create_column(view->dms, LINES_COLUMN_NAME, OBI_IDX, 0, 1, NULL, false, false, false, NULL, NULL, -1, NULL);
+			view->line_selection = obi_create_column(view->dms, LINES_COLUMN_NAME, OBI_IDX, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, NULL);
 			if ((view->line_selection) == NULL)
 			{
 				obidebug(1, "\nError creating a column corresponding to a line selection");
@@ -1803,6 +1863,7 @@ Obiview_p obi_new_view(OBIDMS_p dms, const char* view_name, Obiview_p view_to_cl
 									false,
 									false,
 									false,
+									false,
 									NULL,
 									NULL,
 									-1,
@@ -1844,6 +1905,7 @@ Obiview_p obi_new_view_nuc_seqs(OBIDMS_p dms, const char* view_name, Obiview_p v
 {
 	Obiview_p 		view;
 	OBIDMS_column_p associated_nuc_column;
+	OBIDMS_column_p associated_qual_column;
 	int				nb_predicates;
 
 	if (view_to_clone != NULL)
@@ -1870,19 +1932,19 @@ Obiview_p obi_new_view_nuc_seqs(OBIDMS_p dms, const char* view_name, Obiview_p v
 	if ((view_to_clone == NULL) && create_default_columns)
 	{
 		// Adding sequence column
-		if (obi_view_add_column(view, NUC_SEQUENCE_COLUMN, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)	// discuss using same indexer "NUC_SEQ_INDEXER"
+		if (obi_view_add_column(view, NUC_SEQUENCE_COLUMN, -1, NULL, OBI_SEQ, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)	// discuss using same indexer "NUC_SEQ_INDEXER"
 		{
 			obidebug(1, "Error adding an obligatory column in a nucleotide sequences view");
 			return NULL;
 		}
 		// Adding id column
-		if (obi_view_add_column(view, ID_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+		if (obi_view_add_column(view, ID_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 		{
 			obidebug(1, "Error adding an obligatory column in a nucleotide sequences view");
 			return NULL;
 		}
 		// Adding definition column
-		if (obi_view_add_column(view, DEFINITION_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+		if (obi_view_add_column(view, DEFINITION_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 		{
 			obidebug(1, "Error adding an obligatory column in a nucleotide sequences view");
 			return NULL;
@@ -1891,11 +1953,15 @@ Obiview_p obi_new_view_nuc_seqs(OBIDMS_p dms, const char* view_name, Obiview_p v
 		if (quality_column)
 		{
 			associated_nuc_column = obi_view_get_column(view, NUC_SEQUENCE_COLUMN);
-			if (obi_view_add_column(view, QUALITY_COLUMN, -1, NULL, OBI_QUAL, 0, 1, NULL, false, false, false, NULL, (associated_nuc_column->header)->name, (associated_nuc_column->header)->version, "{}", true) < 0)		// TODO discuss automatic association
+			if (obi_view_add_column(view, QUALITY_COLUMN, -1, NULL, OBI_QUAL, 0, 1, NULL, false, false, false, false, NULL, (associated_nuc_column->header)->name, (associated_nuc_column->header)->version, "{}", true) < 0)		// TODO discuss automatic association
 			{
 				obidebug(1, "Error adding an obligatory column in a nucleotide sequences view");
 				return NULL;
 			}
+			// Associating both ways: associating nuc sequences column to quality column
+			associated_qual_column = obi_view_get_column(view, QUALITY_COLUMN);
+			strcpy((associated_nuc_column->header->associated_column).column_name, associated_qual_column->header->name);
+			(associated_nuc_column->header->associated_column).version = associated_qual_column->header->version;
 		}
 	}
 
@@ -1922,7 +1988,7 @@ Obiview_p obi_new_view_nuc_seqs(OBIDMS_p dms, const char* view_name, Obiview_p v
 	(view->predicate_functions)[(view->nb_predicates)] = view_has_nuc_sequence_column;
 	(view->predicate_functions)[(view->nb_predicates) + 1] = view_has_id_column;
 	(view->predicate_functions)[(view->nb_predicates) + 2] = view_has_definition_column;
-//	if (quality_column)   # TODO discuss. Commented bc for example with obi annotate, clone view so clone predicate, then modify seq, so quality is deleted, and predicate boom
+//	if (quality_column)   # TODO fix by triggering predicate deleting if quality deleting. Commented bc for example with obi annotate, clone view so clone predicate, then modify seq, so quality is deleted, and predicate boom
 //		(view->predicate_functions)[(view->nb_predicates) + 3] = view_has_quality_column;
 
 	view->nb_predicates = nb_predicates;
@@ -2212,7 +2278,7 @@ Obiview_p obi_open_view(OBIDMS_p dms, const char* view_name)
 
 // TODO return a pointer on the column?
 int obi_view_add_column(Obiview_p    view,
-						char*  column_name,
+						char*        column_name,
 						obiversion_t version_number,
 						const char*  alias,
 						OBIType_t    data_type,
@@ -2220,6 +2286,7 @@ int obi_view_add_column(Obiview_p    view,
 						index_t      nb_elements_per_line,
 						char*        elements_names,
 						bool         elt_names_formatted,
+						bool         dict_column,
 						bool		 tuples,
 						bool         to_eval,
 						const char*  indexer_name,
@@ -2302,7 +2369,7 @@ int obi_view_add_column(Obiview_p    view,
 	// Open or create the column
 	if (create)
 	{	// Create column
-		column = obi_create_column(view->dms, column_name, data_type, nb_lines, nb_elements_per_line, elements_names, elt_names_formatted, tuples, to_eval, indexer_name, associated_column_name, associated_column_version, comments);
+		column = obi_create_column(view->dms, column_name, data_type, nb_lines, nb_elements_per_line, elements_names, elt_names_formatted, dict_column, tuples, to_eval, indexer_name, associated_column_name, associated_column_version, comments);
 		if (column == NULL)
 		{
 			obidebug(1, "\nError creating a column to add to a view");
@@ -2541,6 +2608,144 @@ int obi_view_create_column_alias(Obiview_p view, const char* current_name, const
 }
 
 
+char* obi_view_formatted_infos(Obiview_p view, bool detailed)
+{
+	int             i;
+	char*   		view_infos = NULL;
+	char*   		view_name = NULL;
+	time_t  		creation_date;
+	char*			creation_date_str = NULL;
+	index_t 		line_count;
+	char    		line_count_str[256];
+	OBIDMS_column_p column;
+	char*			column_alias = NULL;
+	char*   		column_infos = NULL;
+	char*			comments = NULL;
+
+	// View name
+	view_name = (view->infos)->name;
+	view_infos = (char*) malloc((strlen("# View name:\n")+strlen(view_name)+1) * sizeof(char));
+	strcpy(view_infos, "# View name:\n");
+	strcat(view_infos, view_name);
+
+	// Date created
+	if (view->read_only) // Date not saved until view is finished writing
+	{
+		creation_date = (view->infos)->creation_date;
+		creation_date_str = ctime(&creation_date);
+		view_infos = realloc(view_infos, (strlen(view_infos)+strlen("\n# Date created:\n")+strlen(creation_date_str)+1) * sizeof(char));
+		strcat(view_infos, "\n# Date created:\n");
+		strcat(view_infos, creation_date_str);
+	}
+
+	// Line count
+	line_count = (view->infos)->line_count;
+	snprintf(line_count_str, sizeof line_count_str, "%lld", line_count);
+	view_infos = realloc(view_infos, (strlen(view_infos)+strlen("\n# Line count:\n")+strlen(line_count_str)+1) * sizeof(char));
+	strcat(view_infos, "# Line count:\n");
+	strcat(view_infos, line_count_str);
+
+	// Columns: go through each, print their alias then their infos
+	view_infos = realloc(view_infos, (strlen(view_infos)+strlen("\n# Columns:")+1) * sizeof(char));
+	strcat(view_infos, "\n# Columns:");
+	for (i=0; i<((view->infos)->column_count); i++)
+	{
+		column = *((OBIDMS_column_p*)ll_get(view->columns, i));
+		if (column == NULL)
+		{
+			obidebug(1, "\nError getting a column from the linked list of column pointers of a view to format view infos");
+			return NULL;
+		}
+
+		// Column alias
+		column_alias = (((view->infos)->column_references)[i]).alias;
+		view_infos = realloc(view_infos, (strlen(view_infos)+strlen("\n")+strlen(column_alias)+strlen(", ")+1) * sizeof(char));
+		strcat(view_infos, "\n");
+		strcat(view_infos, column_alias);
+		strcat(view_infos, ", ");
+
+		// Column infos
+		column_infos = obi_column_formatted_infos(column, detailed);
+		if (column_infos == NULL)
+		{
+			obidebug(1, "\nError getting column infos to format view infos");
+			return NULL;
+		}
+
+		view_infos = realloc(view_infos, (strlen(view_infos)+strlen(column_infos)+1) * sizeof(char));
+		strcat(view_infos, column_infos);
+		free(column_infos);
+	}
+
+	// Get commments if detailed informations required
+	if (detailed)
+	{
+		comments = (view->infos)->comments;
+		if (strlen(comments)>2) // Add all comments if not empty
+		{
+			view_infos = realloc(view_infos, (strlen(view_infos)+strlen("\n# Comments:\n")+strlen(comments)+1) * sizeof(char));
+			if (view_infos == NULL)
+			{
+				obi_set_errno(OBI_MALLOC_ERROR);
+				obidebug(1, "\nError allocating memory for formatted view infos");
+				return NULL;
+			}
+
+			strcat(view_infos, "\n# Comments:\n");
+			strcat(view_infos, comments);
+		}
+	}
+
+	view_infos = realloc(view_infos, (strlen(view_infos)+2) * sizeof(char));
+	strcat(view_infos, "\n");
+
+	return view_infos;
+}
+
+
+char* obi_view_formatted_infos_one_line(Obiview_p view)
+{
+	int             i;
+	char*   		view_infos = NULL;
+	char*   		view_name = NULL;
+	time_t  		creation_date;
+	char*			creation_date_str = NULL;
+	index_t 		line_count;
+	char    		line_count_str[256];
+
+	// View name
+	view_name = (view->infos)->name;
+	view_infos = (char*) malloc((strlen("  # ")+strlen(view_name)+2) * sizeof(char));
+	strcpy(view_infos, "  # ");
+	strcat(view_infos, view_name);
+	strcat(view_infos, ":");
+
+	// Date created
+	if (view->read_only) // Date not saved until view is finished writing
+	{
+		creation_date = (view->infos)->creation_date;
+		creation_date_str = ctime(&creation_date);
+		// Delete \n added by ctime
+		creation_date_str[strlen(creation_date_str)-1] = '\0';
+		view_infos = realloc(view_infos, (strlen(view_infos)+strlen(" Date created: ")+strlen(creation_date_str)+1) * sizeof(char));
+		strcat(view_infos, " Date created: ");
+		strcat(view_infos, creation_date_str);
+	}
+
+	// Line count
+	line_count = (view->infos)->line_count;
+	snprintf(line_count_str, sizeof line_count_str, "%lld", line_count);
+	view_infos = realloc(view_infos, (strlen(view_infos)+strlen(" ; Line count: ")+strlen(line_count_str)+1) * sizeof(char));
+	strcat(view_infos, " ; Line count: ");
+	strcat(view_infos, line_count_str);
+
+	view_infos = realloc(view_infos, (strlen(view_infos)+2) * sizeof(char));
+	strcat(view_infos, "\n");
+
+	return view_infos;
+}
+
+
 int obi_view_write_comments(Obiview_p view, const char* comments)
 {
 	size_t new_size;
@@ -2705,7 +2910,7 @@ int obi_clean_unfinished_views(OBIDMS_p dms)
 		if ((dp->d_name)[0] == '.')
 			continue;
 		i=0;
-		while ((dp->d_name)[i] != '.')
+		while (strncmp((dp->d_name)+i, ".obiview", 8))
 			i++;
 		relative_path = (char*) malloc(strlen(VIEW_DIR_NAME) + strlen(dp->d_name) + 2);
 		strcpy(relative_path, VIEW_DIR_NAME);
@@ -3028,7 +3233,7 @@ int obi_create_auto_count_column(Obiview_p view)
 		return -1;
 	}
 
-	if (obi_view_add_column(view, COUNT_COLUMN, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(view, COUNT_COLUMN, -1, NULL, OBI_INT, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "Error adding an automatic count column in a view");
 		return -1;
@@ -3080,7 +3285,7 @@ int obi_create_auto_id_column(Obiview_p view, const char* prefix)
 	}
 
 	// Create the new ID column
-	if (obi_view_add_column(view, ID_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, NULL, NULL, -1, "{}", true) < 0)
+	if (obi_view_add_column(view, ID_COLUMN, -1, NULL, OBI_STR, 0, 1, NULL, false, false, false, false, NULL, NULL, -1, "{}", true) < 0)
 	{
 		obidebug(1, "Error adding an automatic ID column in a view");
 		return -1;

src/obiview.h

@@ -30,54 +30,56 @@
 #include "obiblob.h"
 
 
-#define OBIVIEW_NAME_MAX_LENGTH (249)   		/**< The maximum length of an OBIDMS view name, without the extension.
-                                	 	  	  	 */
-#define VIEW_TYPE_MAX_LENGTH (1024)   			/**< The maximum length of the type name of a view.
-                                	 	  	  	 */
-#define LINES_COLUMN_NAME "LINES"				/**< The name of the column containing the line selections
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in all views.
-                                	 	  	  	 */
-#define VIEW_TYPE_NUC_SEQS "NUC_SEQS_VIEW"   	/**< The type name of views based on nucleotide sequences
-												 *   and their metadata.
-                                	 	  	  	 */
-#define NUC_SEQUENCE_COLUMN "NUC_SEQ"			/**< The name of the column containing the nucleotide sequences
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	  	 */
-#define ID_COLUMN "ID"							/**< The name of the column containing the sequence identifiers
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	  	 */
-#define DEFINITION_COLUMN "DEFINITION"			/**< The name of the column containing the sequence definitions
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	  	 */
-#define QUALITY_COLUMN "QUALITY"				/**< The name of the column containing the sequence qualities
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	   	 */
-#define REVERSE_QUALITY_COLUMN "REVERSE_QUALITY" /**< The name of the column containing the sequence qualities
- 	 	 	 	 	 	 	 	 	 	 	 	 *    of the reverse read (generated by ngsfilter, used by alignpairedend).
-                                	 	  	   	 */
+#define OBIVIEW_NAME_MAX_LENGTH (249)   		   /**< The maximum length of an OBIDMS view name, without the extension.
+                                	 	  	  	    */
+#define VIEW_TYPE_MAX_LENGTH (1024)   			   /**< The maximum length of the type name of a view.
+                                	 	  	  	    */
+#define LINES_COLUMN_NAME "LINES"				   /**< The name of the column containing the line selections
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in all views.
+                                	 	  	  	    */
+#define VIEW_TYPE_NUC_SEQS "NUC_SEQS_VIEW"   	   /**< The type name of views based on nucleotide sequences
+												    *   and their metadata.
+                                	 	  	  	    */
+#define NUC_SEQUENCE_COLUMN "NUC_SEQ"			   /**< The name of the column containing the nucleotide sequences
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	  	    */
+#define ID_COLUMN "ID"							   /**< The name of the column containing the sequence identifiers
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	  	    */
+#define DEFINITION_COLUMN "DEFINITION"			   /**< The name of the column containing the sequence definitions
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	  	    */
+#define QUALITY_COLUMN "QUALITY"				   /**< The name of the column containing the sequence qualities
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	   	    */
+#define REVERSE_QUALITY_COLUMN "REVERSE_QUALITY"   /**< The name of the column containing the sequence qualities
+ 	 	 	 	 	 	 	 	 	 	 	 	    *    of the reverse read (generated by ngsfilter, used by alignpairedend).
+                                	 	  	   	    */
 #define REVERSE_SEQUENCE_COLUMN "REVERSE_SEQUENCE" /**< The name of the column containing the sequence
- 	 	 	 	 	 	 	 	 	 	 	 	 *    of the reverse read (generated by ngsfilter, used by alignpairedend).
-                                	 	  	   	 */
-#define QUALITY_COLUMN "QUALITY"				/**< The name of the column containing the sequence qualities
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	   	 */
-#define COUNT_COLUMN "COUNT"				    /**< The name of the column containing the sequence counts
- 	 	 	 	 	 	 	 	 	 	 	 	 *   in NUC_SEQS_VIEW views.
-                                	 	  	  	 */
-#define TAXID_COLUMN "TAXID"				    /**< The name of the column containing the taxids.       TODO subtype of INT column?
-                                	             */
-#define MERGED_TAXID_COLUMN "MERGED_TAXID"		/**< The name of the column containing the merged taxids information.
-                                	             */
-#define MERGED_PREFIX "MERGED_"		            /**< The prefix to prepend to column names when merging informations during obi uniq.
-                                	             */
-#define TAXID_DIST_COLUMN "TAXID_DIST"			/**< The name of the column containing a dictionary of taxid:[list of ids] when merging informations during obi uniq.
-                                	             */
-#define MERGED_COLUMN "MERGED"					/**< The name of the column containing a list of ids when merging informations during obi uniq.
-                                	             */
-#define ID_PREFIX "seq"						    /**< The default prefix of sequence identifiers in automatic ID columns.
-                                	 	  	  	 */
-#define PREDICATE_KEY "predicates"		        /**< The key used in the json-formatted view comments to store predicates.
-                                	 	  	  	 */
+ 	 	 	 	 	 	 	 	 	 	 	 	    *    of the reverse read (generated by ngsfilter, used by alignpairedend).
+                                	 	  	   	    */
+#define QUALITY_COLUMN "QUALITY"				   /**< The name of the column containing the sequence qualities
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	   	    */
+#define COUNT_COLUMN "COUNT"				       /**< The name of the column containing the sequence counts
+ 	 	 	 	 	 	 	 	 	 	 	 	    *   in NUC_SEQS_VIEW views.
+                                	 	  	  	    */
+#define SCIENTIFIC_NAME_COLUMN "SCIENTIFIC_NAME"   /**< The name of the column containing the taxon scientific name.
+                                	                */
+#define TAXID_COLUMN "TAXID"				       /**< The name of the column containing the taxids.       TODO subtype of INT column?
+                                	                */
+#define MERGED_TAXID_COLUMN "MERGED_TAXID"		   /**< The name of the column containing the merged taxids information.
+                                	                */
+#define MERGED_PREFIX "MERGED_"		               /**< The prefix to prepend to column names when merging informations during obi uniq.
+                                	                */
+#define TAXID_DIST_COLUMN "TAXID_DIST"			   /**< The name of the column containing a dictionary of taxid:[list of ids] when merging informations during obi uniq.
+                                	                */
+#define MERGED_COLUMN "MERGED"					   /**< The name of the column containing a list of ids when merging informations during obi uniq.
+                                	                */
+#define ID_PREFIX "seq"						       /**< The default prefix of sequence identifiers in automatic ID columns.
+                                	 	  	  	    */
+#define PREDICATE_KEY "predicates"		           /**< The key used in the json-formatted view comments to store predicates.
+                                	 	  	  	    */
 
 
 /**
@@ -398,6 +400,7 @@ Obiview_p obi_open_view(OBIDMS_p dms, const char* view_name);
  * @param elements_names The names of the elements with ';' as separator (no terminal ';'),
  * 						 if the column is created; NULL or "" if the default names are to be used ("0\01\02\0...\0n").
  * @param elt_names_formatted Whether the separator for the elements names is ';' (false), or '\0' (true, as formatted by format_elements_names()).
+ * @param dict_column Whether the column contains dictionary-like values.
  * @param tuples A boolean indicating whether the column should contain indices referring to indexed tuples.
  * @param to_eval A boolean indicating whether the column contains expressions that should be evaluated
  *                (typically OBI_STR columns containing character strings to be evaluated by Python).
@@ -406,7 +409,7 @@ Obiview_p obi_open_view(OBIDMS_p dms, const char* view_name);
  * @param associated_column_name The name of the associated column if there is one (otherwise NULL or ""), if the column is created.
  * @param associated_column_version The version of the associated column if there is one (otherwise -1), if the column is created.
  * @param comments Optional comments associated with the column if it is created (NULL or "" if no comments associated).
- * @param create Whether the column should be created (create == true) or opened (create == false).
+ * @param create Whether the column should be created (create == true) or already exists (create == false).
  *
  * @returns A value indicating the success of the operation.
  * @retval 0 if the operation was successfully completed.
@@ -416,7 +419,7 @@ Obiview_p obi_open_view(OBIDMS_p dms, const char* view_name);
  * @author Celine Mercier (celine.mercier@metabarcoding.org)
  */
 int obi_view_add_column(Obiview_p    view,
-						char*  column_name,
+						char*        column_name,
 						obiversion_t version_number,
 						const char*  alias,
 						OBIType_t    data_type,
@@ -424,6 +427,7 @@ int obi_view_add_column(Obiview_p    view,
 						index_t      nb_elements_per_line,
 						char*        elements_names,
 						bool         elt_names_formatted,
+						bool         dict_column,
 						bool         tuples,
 						bool         to_eval,
 						const char*  indexer_name,
@@ -519,6 +523,39 @@ OBIDMS_column_p* obi_view_get_pointer_on_column_in_view(Obiview_p view, const ch
 int obi_view_create_column_alias(Obiview_p view, const char* current_name, const char* alias);
 
 
+/**
+ * @brief Returns the informations of a view with a human readable format (view name, date created, line count, column informations, comments).
+ *
+ * @warning The returned pointer has to be freed by the caller.
+ *
+ * @param column A pointer on a view.
+ * @param detailed Whether the informations should contain view comments.
+ *
+ * @returns A pointer on a character array where the formatted view informations are stored.
+ * @retval NULL if an error occurred.
+ *
+ * @since September 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_view_formatted_infos(Obiview_p view, bool detailed);
+
+
+/**
+ * @brief Returns the informations of a view with a human readable format on one line (view name, date created, line count).
+ *
+ * @warning The returned pointer has to be freed by the caller.
+ *
+ * @param column A pointer on a view.
+ *
+ * @returns A pointer on a character array where the formatted view informations are stored.
+ * @retval NULL if an error occurred.
+ *
+ * @since September 2020
+ * @author Celine Mercier (celine.mercier@metabarcoding.org)
+ */
+char* obi_view_formatted_infos_one_line(Obiview_p view);
+
+
 /**
  * @brief Internal function writing new comments in a view file.
  *