Switch to version 3.0.1b21

`SCIENTIFIC_NAME`, and suggests using `--input-na-string` when a
sequence import fails

python/obitools3/apps/optiongroups/__init__.py

@@ -63,6 +63,12 @@ def __addImportInputOption(optionManager):
                      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,

python/obitools3/commands/alignpairedend.pyx

@@ -233,6 +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]

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
@@ -23,6 +24,7 @@ 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

@@ -109,6 +109,8 @@ def run(config):
     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
@@ -234,11 +236,15 @@ def run(config):
         def_col = view[DEFINITION_COLUMN]
         seq_col = view[NUC_SEQUENCE_COLUMN]
 
-    # Prepare taxon scientific name and taxid refs if RDP or SILVA file
+    # Prepare taxon scientific name and taxid refs if RDP/SILVA/UNITE/SINTAX formats
     silva = False
     rdp = False
     unite = False
-    if 'inputformat' in config['obi'] and (config['obi']['inputformat'] == b"silva" or config['obi']['inputformat'] == b"rdp" or config['obi']['inputformat'] == b"unite"):
+    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":
@@ -247,6 +253,8 @@ def run(config):
             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)
@@ -356,13 +364,26 @@ def run(config):
                     qual_col[i] = entry.quality
                     
                 # Parse taxon scientific name if RDP or Silva or Unite file
-                if (rdp or silva or unite):
+                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):
@@ -379,12 +400,14 @@ def run(config):
             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:]
 
@@ -397,7 +420,7 @@ def run(config):
                             pass
                     
                     if tag not in dcols :
-                                                
+                    
                         value_type = type(value)
                         nb_elts = 1
                         value_obitype = OBI_VOID
@@ -495,7 +518,7 @@ def run(config):
                                 dcols[tag][0][i] = value
         
         except Exception as e:
-            print("\nCould not import sequence:", repr(entry), "(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:

python/obitools3/commands/ngsfilter.pyx

@@ -264,6 +264,10 @@ 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()

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/dms/capi/obitaxonomy.pxd

@@ -58,7 +58,7 @@ 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)
     

python/obitools3/dms/taxo/taxo.pxd

@@ -19,8 +19,8 @@ cdef class Taxonomy(OBIWrapper) :
     cpdef Taxon get_taxon_by_idx(self, int idx)
     cpdef Taxon get_taxon_by_taxid(self, int taxid)
     cpdef Taxon get_taxon_by_name(self, object taxon_name, object restricting_taxid=*)
-    cpdef write(self, object prefix)
-    cpdef int add_taxon(self, str name, str rank_name, int parent_taxid, int min_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

@@ -174,6 +174,7 @@ cdef class Taxonomy(OBIWrapper) :
 
 
     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
@@ -282,12 +283,12 @@ cdef class Taxonomy(OBIWrapper) :
             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 :
@@ -329,6 +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/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",b"additional_info"]
+    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(maxsplit=5)
-        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[0:7]))
-    
+        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/uri/decode.pyx

@@ -506,7 +506,7 @@ def open_uri(uri,
         if format is not None:
             if seqtype==b"nuc":
                 objclass = Nuc_Seq    # Nuc_Seq_Stored? TODO
-                if format==b"fasta" or format==b"silva" or format==b"rdp" or format == b"unite":
+                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, 

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '1b17'
+serial= '1b21'
 
 version ="%d.%d.%s" % (major,minor,serial)

src/.gitignore

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

src/obidms_taxonomy.c

@@ -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);
@@ -3250,47 +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;
 }
 
@@ -3302,16 +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)
@@ -3377,9 +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));
@@ -3441,42 +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;
 }
@@ -3547,11 +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;
 }
@@ -3669,8 +3696,10 @@ 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;
 }
 

src/obidms_taxonomy.h

@@ -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);
 
 
 /**