Switch to version 3.0.0b24

not properly closed and imported

MANIFEST.in

@@ -6,7 +6,7 @@ recursive-include doc/sphinx/source *.txt *.rst *.py
 recursive-include doc/sphinx/sphinxext *.py
 include doc/sphinx/Makefile
 include doc/sphinx/Doxyfile
-include README.txt
+include README.md
 include requirements.txt
 include scripts/obi
 

obi_completion_script.bash

@@ -1,4 +1,3 @@
-#/usr/bin/env bash
 
 _obi_comp ()
 {

python/obitools3/commands/align.pyx

@@ -266,9 +266,9 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary result view in the input DMS
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
     
-    i_dms.close()
+    i_dms.close(force=True)
     
     logger("info", "Done.")
   

python/obitools3/commands/alignpairedend.pyx

@@ -247,10 +247,10 @@ def run(config):
     #print("\n\nOutput view:\n````````````", file=sys.stderr)
     #print(repr(view), file=sys.stderr)
     
-    input[0].close()
+    input[0].close(force=True)
     if two_views:
-        rinput[0].close()
+        rinput[0].close(force=True)
-    output[0].close()
+    output[0].close(force=True)
 
     logger("info", "Done.")
     

python/obitools3/commands/annotate.pyx

@@ -13,7 +13,8 @@ from obitools3.dms.capi.obiview cimport NUC_SEQUENCE_COLUMN, \
                                         ID_COLUMN, \
                                         DEFINITION_COLUMN, \
                                         QUALITY_COLUMN, \
-                                        COUNT_COLUMN
+                                        COUNT_COLUMN, \
+                                        TAXID_COLUMN
 
 import time
 import math 
@@ -175,8 +176,8 @@ def sequenceTaggerGenerator(config, taxo=None):
             counter[0]+=1
                             
         for rank in annoteRank:
-            if 'taxid' in seq:
+            if TAXID_COLUMN in seq:
-                taxid = seq['taxid']
+                taxid = seq[TAXID_COLUMN]
                 if taxid is not None:
                     rtaxid = taxo.get_taxon_at_rank(taxid, rank)
                     if rtaxid is not None:
@@ -190,58 +191,50 @@ def sequenceTaggerGenerator(config, taxo=None):
             seq['seq_rank']=counter[0]
     
         for i,v in toSet:
-            #try:
+            try:
                 if taxo is not None:
                     environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
                 else:
                     environ = {'sequence':seq, 'counter':counter[0], 'math':math}
                 val = eval(v, environ, seq)
-            #except Exception,e:       # TODO discuss usefulness of this
+            except Exception:       # set string if not a valid expression
-            #    if options.onlyValid:
+                val = v
-            #        raise e
-            #    val = v
             seq[i]=val
 
         if length:
             seq['seq_length']=len(seq)
 
         if newId is not None:
-#            try:
+            try:
                 if taxo is not None:
                     environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
                 else:
                     environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
                 val = eval(newId, environ, seq)
-#            except Exception,e:
+            except Exception:      # set string if not a valid expression
-#                if options.onlyValid:
+                val = newId
-#                    raise e
-#                val = newId
             seq.id=val
 
         if newDef is not None:
-#            try:
+            try:
                 if taxo is not None:
                     environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
                 else:
                     environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
                 val = eval(newDef, environ, seq)
-#            except Exception,e:
+            except Exception:    # set string if not a valid expression
-#                if options.onlyValid:
+                val = newDef
-#                    raise e
-#                val = newDef
             seq.definition=val
-#             
+             
         if newSeq is not None:
-#            try:
+            try:
                 if taxo is not None:
                     environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
                 else:
                     environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
                 val = eval(newSeq, environ, seq)
-#            except Exception,e:
+            except Exception:    # set string if not a valid expression
-#                if options.onlyValid:
+                val = newSeq
-#                    raise e
-#                val = newSeq
             seq.seq=val
             if 'seq_length' in seq:
                 seq['seq_length']=len(seq)
@@ -251,15 +244,14 @@ def sequenceTaggerGenerator(config, taxo=None):
                 seq.view.delete_column(QUALITY_COLUMN)
                     
         if run is not None:
-#            try:
+            try:
                 if taxo is not None:
                     environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
                 else:
                     environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
                 eval(run, environ, seq)
-#            except Exception,e:
+            except Exception,e:
-#                if options.onlyValid:
+                raise e
-#                    raise e
 
     return sequenceTagger
 
@@ -379,7 +371,7 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary imported view used to create the final view
     if i_dms != o_dms:
         View.delete_view(o_dms, imported_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
-    i_dms.close()
+    i_dms.close(force=True)
 
     logger("info", "Done.")

python/obitools3/commands/build_ref_db.pyx

@@ -97,9 +97,9 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary result view in the input DMS
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
     
-    i_dms.close()
+    i_dms.close(force=True)
 
     logger("info", "Done.")
     

python/obitools3/commands/cat.pyx

@@ -87,6 +87,23 @@ def run(config):
         Column.new_column(o_view, REVERSE_SEQUENCE_COLUMN, OBI_SEQ)
         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
+    dict_cols = {}
+    for v in iview_list:
+        for coln in v.keys():
+            if v[coln].nb_elements_per_line > 1:
+                if coln not in dict_cols:
+                    dict_cols[coln] = {}
+                    dict_cols[coln]['eltnames'] = set(v[coln].elements_names)
+                    dict_cols[coln]['nbelts'] = v[coln].nb_elements_per_line
+                    dict_cols[coln]['obitype'] = v[coln].data_type_int
+                else:
+                    dict_cols[coln]['eltnames'] = set(v[coln].elements_names + list(dict_cols[coln]['eltnames']))
+                    dict_cols[coln]['nbelts'] = len(dict_cols[coln]['eltnames'])
+    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']))
+    
     # Initialize the progress bar
     pb = ProgressBar(total_len, config, seconde=5)
     
@@ -116,7 +133,7 @@ def run(config):
     #print(repr(view), file=sys.stderr)
 
     for d in idms_list:
-        d.close()
+        d.close(force=True)
-    o_dms.close()
+    o_dms.close(force=True)
     
     logger("info", "Done.")

python/obitools3/commands/clean.pyx

@@ -124,8 +124,8 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary result view in the input DMS
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
     
-    i_dms.close()
+    i_dms.close(force=True)
 
     logger("info", "Done.")

python/obitools3/commands/count.pyx

@@ -56,3 +56,5 @@ def run(config):
         print(count2)
     else:
         print(count1)
+    
+    input[0].close(force=True)

python/obitools3/commands/ecopcr.pyx

@@ -35,13 +35,15 @@ def addOptions(parser):
                        action="store", dest="ecopcr:primer1",
                        metavar='<PRIMER>',
                        type=str,
-                       help="Forward primer.")
+                       required=True,
+                       help="Forward primer, length must be less than or equal to 32")
 
     group.add_argument('--primer2', '-R',
                        action="store", dest="ecopcr:primer2",
                        metavar='<PRIMER>',
                        type=str,
-                       help="Reverse primer.")
+                       required=True,
+                       help="Reverse primer, length must be less than or equal to 32")
 
     group.add_argument('--error', '-e',
                        action="store", dest="ecopcr:error",
@@ -203,6 +205,7 @@ def run(config):
     #print("\n\nOutput view:\n````````````", file=sys.stderr)
     #print(repr(o_dms[o_view_name]), file=sys.stderr)
 
-    o_dms.close()
+    i_dms.close(force=True)
+    o_dms.close(force=True)
 
     logger("info", "Done.")

python/obitools3/commands/ecotag.pyx

@@ -64,9 +64,9 @@ 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(i_dms[ref_view_name].comments["ref_db_threshold"]) :
+    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'], i_dms[ref_view_name].comments["ref_db_threshold"])
+              config['ecotag']['threshold'], ref_dms[ref_view_name].comments["ref_db_threshold"])
     
     # Open the output: only the DMS
     output = open_uri(config['obi']['outputURI'],
@@ -107,8 +107,8 @@ def run(config):
     comments = View.print_config(config, "ecotag", command_line, input_dms_name=input_dms_name, input_view_name=input_view_name)
 
     if obi_ecotag(i_dms.name_with_full_path, tobytes(i_view_name), \
-                  tobytes(ref_dms_name), tobytes(ref_view_name), \
+                  ref_dms.name_with_full_path, tobytes(ref_view_name), \
-                  tobytes(taxo_dms_name), tobytes(taxonomy_name), \
+                  taxo_dms.name_with_full_path, tobytes(taxonomy_name), \
                   tobytes(o_view_name), comments, 
                   config['ecotag']['threshold']) < 0:
         raise Exception("Error running ecotag")
@@ -126,9 +126,11 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary result view in the input DMS
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
     
-    i_dms.close()
+    taxo_dms.close(force=True)
+    ref_dms.close(force=True)
+    i_dms.close(force=True)
     
     logger("info", "Done.")
 

python/obitools3/commands/export.pyx

@@ -60,11 +60,21 @@ def run(config):
     if (output[2] == Nuc_Seq) and (iview.type != b"NUC_SEQS_VIEW") :  # Nuc_Seq_Stored? TODO
         raise Exception("Error: the view to export in fasta or fastq format is not a NUC_SEQS view")
     
+    if config['obi']['only'] is not None:
+        withoutskip = min(input[4], config['obi']['only'])
+    else:
+        withoutskip = input[4]
+     
+    if config['obi']['skip'] is not None:
+        skip = min(input[4], config['obi']['skip'])
+    else:
+        skip = 0
+    
     # Initialize the progress bar
     if config['obi']['noprogressbar']:
         pb = None
     else:
-        pb = ProgressBar(len(iview), config, seconde=5)
+        pb = ProgressBar(withoutskip - skip, config, seconde=5)
 
     i=0
     for seq in iview :
@@ -86,7 +96,7 @@ def run(config):
     if not BrokenPipeError and not IOError:
         output_object.close()
     iview.close()
-    input[0].close()
+    input[0].close(force=True)
 
     logger("info", "Done.")
 

python/obitools3/commands/grep.pyx

@@ -370,7 +370,7 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary imported view used to create the final view
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
-    i_dms.close()
+    i_dms.close(force=True)
 
     logger("info", "Done.")

python/obitools3/commands/head.pyx

@@ -103,7 +103,7 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary imported view used to create the final view
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
-    i_dms.close()
+    i_dms.close(force=True)
     
     logger("info", "Done.")

python/obitools3/commands/history.pyx

@@ -55,3 +55,4 @@ def run(config):
         else:
             raise Exception("ASCII history only available for views")
     
+    input[0].close(force=True)

python/obitools3/commands/import.pyx

@@ -25,7 +25,8 @@ from obitools3.dms.capi.obiview cimport VIEW_TYPE_NUC_SEQS, \
                                         DEFINITION_COLUMN, \
                                         QUALITY_COLUMN, \
                                         COUNT_COLUMN, \
-                                        TAXID_COLUMN
+                                        TAXID_COLUMN, \
+                                        MERGED_PREFIX
                                         
 from obitools3.dms.capi.obidms cimport obi_import_view
 
@@ -72,7 +73,7 @@ def addOptions(parser):
                      action="store_true", dest="import:preread",
                      default=False,
                      help="Do a first readthrough of the dataset if it contains huge dictionaries (more than 100 keys) for "
-                          "a much faster import.")
+                          "a much faster import. This option is not recommended and will slow down the import in any other case.")
 
 
 def run(config):
@@ -163,7 +164,7 @@ def run(config):
         taxo.write(taxo_name)
         taxo.close()
         o_dms.record_command_line(" ".join(sys.argv[1:]))
-        o_dms.close()
+        o_dms.close(force=True)
         logger("info", "Done.")
         return
 
@@ -217,11 +218,14 @@ def run(config):
                 logger("info", "Read %d entries", i)
     
             for tag in entry :
+                newtag = tag
+                if tag[:7] == b"merged_":
+                    newtag = MERGED_PREFIX+tag[7:]
                 if type(entry[tag]) == dict :
                     if tag in dict_dict:
-                        dict_dict[tag][0].update(entry[tag].keys())
+                        dict_dict[newtag][0].update(entry[tag].keys())
                     else:
-                        dict_dict[tag] = [set(entry[tag].keys()), get_obitype(entry[tag])]
+                        dict_dict[newtag] = [set(entry[tag].keys()), get_obitype(entry[tag])]
             i+=1
         
         if pb is not None:
@@ -232,7 +236,7 @@ def run(config):
             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])), \
-                          value_obitype)
+                          dict_dict[tag][1])
     
         
         # Reinitialize the input
@@ -256,7 +260,6 @@ def run(config):
 
         if entry is None:  # error or exception handled at lower level, not raised because Python generators can't resume after any exception is raised
             if config['obi']['skiperror']:
-                i-=1
                 continue
             else:
                 raise RollbackException("obi import error, rollbacking view", view)
@@ -266,6 +269,8 @@ def run(config):
         elif not i%50000:
             logger("info", "Imported %d entries", i)
         
+        try:
+             
             if NUC_SEQS_view: 
                 id_col[i] = entry.id
                 def_col[i] = entry.definition
@@ -288,6 +293,8 @@ def run(config):
                         tag = TAXID_COLUMN
                     if tag == b"count":
                         tag = COUNT_COLUMN
+                    if tag[:7] == b"merged_":
+                        tag = MERGED_PREFIX+tag[7:]
                         
                     if tag not in dcols :
                          
@@ -328,8 +335,8 @@ def run(config):
                         try:
                             # Check that it's not the case where the first entry contained a dict of length 1 and now there is a new key                        
                             if type(value) == dict and \
-                            dcols[tag][0].nb_elements_per_line == 1 and len(value.keys()) == 1 \
+                                dcols[tag][0].nb_elements_per_line == 1 \
-                            and dcols[tag][0].elements_names[0] != list(value.keys())[0] :
+                                and set(dcols[tag][0].elements_names) != set(value.keys()) :
                                 raise IndexError  # trigger column rewrite
                             
                             # Fill value
@@ -382,6 +389,13 @@ def run(config):
                                 # Fill value
                                 dcols[tag][0][i] = value
         
+        except Exception as e:
+            print("\nCould not import sequence id:", entry.id, "(error raised:", e, ")")
+            if 'skiperror' in config['obi'] and not config['obi']['skiperror']:
+                raise e
+            else:
+                pass
+        
         i+=1
  
     if pb is not None:
@@ -402,7 +416,7 @@ def run(config):
     except AttributeError:
         pass
     try:
-        output[0].close()
+        output[0].close(force=True)
     except AttributeError:
         pass
 

python/obitools3/commands/less.pyx

@@ -46,5 +46,5 @@ def run(config):
     process.wait()
 
     iview.close()
-    input[0].close()
+    input[0].close(force=True)
 

python/obitools3/commands/ls.pyx

@@ -36,6 +36,7 @@ def run(config):
         l = []
         for view in input[0]:
             l.append(tostr(view) + "\t(Date created: " + str(bytes2str_object(dms[view].comments["Date created"]))+")")
+            dms[view].close()
         l.sort()
         for v in l:
             print(v)
@@ -52,3 +53,4 @@ def run(config):
         print("\n### Comments:")
         print(str(input[1].comments))
     
+    input[0].close(force=True)

python/obitools3/commands/ngsfilter.pyx

@@ -42,7 +42,9 @@ def addOptions(parser):
                      metavar="<URI>",
                      type=str,
                      default=None,
-                     help="URI to the view containing the samples definition (with tags, primers, sample names,...)")
+                     required=True,
+                     help="URI to the view containing the samples definition (with tags, primers, sample names,...).\n"
+                          "\nWarning: primer lengths must be less than or equal to 32")
 
     group.add_argument('-R', '--reverse-reads',
                      action="store", dest="ngsfilter:reverse",
@@ -172,6 +174,13 @@ cdef read_info_view(info_view, max_errors=2, verbose=False, not_aligned=False):
     primer_list = []
     i=0
     for p in info_view:
+        
+        # Check primer length: should not be longer than 32, the max allowed by the apat lib
+        if len(p[b'forward_primer']) > 32:
+            raise RollbackException("Error: primers can not be longer than 32bp, rollbacking views")
+        if len(p[b'reverse_primer']) > 32:
+            raise RollbackException("Error: primers can not be longer than 32bp, rollbacking views")
+        
         forward=Primer(p[b'forward_primer'],
                        len(p[b'forward_tag']) if (b'forward_tag' in p and p[b'forward_tag']!=None) else None,
                        True,
@@ -594,7 +603,13 @@ def run(config):
     pb = ProgressBar(entries_len, config, seconde=5)
 
     # Check and store primers and tags
+    try:
         infos, primer_list = read_info_view(info_view, max_errors=config['ngsfilter']['error'], verbose=False, not_aligned=not_aligned)   # TODO obi verbose option
+    except RollbackException, e:
+        if unidentified is not None:
+            raise RollbackException("obi ngsfilter error, rollbacking views: "+str(e), o_view, unidentified)
+        else:
+            raise RollbackException("obi ngsfilter error, rollbacking view: "+str(e), o_view)
     
     aligner = Primer_search(primer_list, config['ngsfilter']['error'])
     
@@ -652,11 +667,11 @@ def run(config):
     #print("\n\nOutput view:\n````````````", file=sys.stderr)
     #print(repr(o_view), file=sys.stderr)
 
-    input[0].close()
+    input[0].close(force=True)
-    output[0].close()
+    output[0].close(force=True)
-    info_input[0].close()
+    info_input[0].close(force=True)
     if unidentified is not None:
-        unidentified_input[0].close()
+        unidentified_input[0].close(force=True)
     aligner.free()
 
     logger("info", "Done.")

python/obitools3/commands/sort.pyx

@@ -141,7 +141,7 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary imported view used to create the final view
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
-    i_dms.close()
+    i_dms.close(force=True)
 
     logger("info", "Done.")

python/obitools3/commands/stats.pyx

@@ -251,7 +251,7 @@ def run(config):
     for i in range(len(sorted_stats)):
         c = sorted_stats[i][0]
         for v in c:
-            if v is not None:
+            if type(v) == bytes:
                 print(pcat % tostr(v)+"\t", end="")
             else:
                 print(pcat % str(v)+"\t", end="")
@@ -268,6 +268,6 @@ def run(config):
         print("%7d" %catcount[c], end="")
         print("%9d" %totcount[c])
 
-    input[0].close()
+    input[0].close(force=True)
     
     logger("info", "Done.")

python/obitools3/commands/tail.pyx

@@ -106,7 +106,7 @@ def run(config):
     # If the input and the output DMS are different, delete the temporary imported view used to create the final view
     if i_dms != o_dms:
         View.delete_view(i_dms, o_view_name)
-        o_dms.close()
+        o_dms.close(force=True)
-    i_dms.close()
+    i_dms.close(force=True)
     
     logger("info", "Done.")

python/obitools3/commands/test.pyx

@@ -529,7 +529,7 @@ def run(config):
         test_taxo(config, infos)
     
     infos['view'].close()
-    infos['dms'].close()
+    infos['dms'].close(force=True)
     shutil.rmtree(config['obi']['defaultdms']+'.obidms', ignore_errors=True)
      
     print("Done.")

python/obitools3/commands/uniq.pxd

@@ -5,5 +5,5 @@ from obitools3.dms.taxo.taxo cimport Taxonomy
 from obitools3.dms.view.typed_view.view_NUC_SEQS cimport View_NUC_SEQS
 
 
-cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy)
+cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy, dict config)
-cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list mergedKeys_list=*, Taxonomy taxonomy=*, bint mergeIds=*, list categories=*, int max_elts=*)
+cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, dict config, list mergedKeys_list=*, Taxonomy taxonomy=*, bint mergeIds=*, list categories=*, int max_elts=*)

python/obitools3/commands/uniq.pyx

@@ -56,7 +56,7 @@ def addOptions(parser):
                              "(option can be used several times).")
 
 
-cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
+cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy, dict config) :
     
     cdef int             taxid
     cdef Nuc_Seq_Stored  seq
@@ -69,7 +69,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
     cdef object          gn_sn
     cdef object          fa_sn
 
-    # Create columns
+    # Create columns and save them for efficiency
     if b"species" in o_view and o_view[b"species"].data_type_int != OBI_INT :
         o_view.delete_column(b"species")
     if b"species" not in o_view:
@@ -77,6 +77,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"species", 
                           OBI_INT
                          )
+    species_column = o_view[b"species"]
 
     if b"genus" in o_view and o_view[b"genus"].data_type_int != OBI_INT :
         o_view.delete_column(b"genus")
@@ -85,6 +86,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"genus", 
                           OBI_INT
                          )
+    genus_column = o_view[b"genus"]
 
     if b"family" in o_view and o_view[b"family"].data_type_int != OBI_INT :
         o_view.delete_column(b"family")
@@ -93,6 +95,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"family", 
                           OBI_INT
                          )
+    family_column = o_view[b"family"]
 
     if b"species_name" in o_view and o_view[b"species_name"].data_type_int != OBI_STR :
         o_view.delete_column(b"species_name")
@@ -101,6 +104,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"species_name", 
                           OBI_STR
                          )
+    species_name_column = o_view[b"species_name"]
 
     if b"genus_name" in o_view and o_view[b"genus_name"].data_type_int != OBI_STR :
         o_view.delete_column(b"genus_name")
@@ -109,6 +113,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"genus_name", 
                           OBI_STR
                          )
+    genus_name_column = o_view[b"genus_name"]
 
     if b"family_name" in o_view and o_view[b"family_name"].data_type_int != OBI_STR :
         o_view.delete_column(b"family_name")
@@ -117,6 +122,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"family_name", 
                           OBI_STR
                          )
+    family_name_column = o_view[b"family_name"]
 
     if b"rank" in o_view and o_view[b"rank"].data_type_int != OBI_STR :
         o_view.delete_column(b"rank")
@@ -125,6 +131,7 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"rank", 
                           OBI_STR
                          )
+    rank_column = o_view[b"rank"]
 
     if b"scientific_name" in o_view and o_view[b"scientific_name"].data_type_int != OBI_STR :
         o_view.delete_column(b"scientific_name")
@@ -133,9 +140,15 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                           b"scientific_name", 
                           OBI_STR
                          )
+    scientific_name_column = o_view[b"scientific_name"]  
     
+    # Initialize the progress bar
+    pb = ProgressBar(len(o_view), config, seconde=5)
+    
+    i=0
     for seq in o_view:
         PyErr_CheckSignals()
+        pb(i)
         if MERGED_TAXID_COLUMN in seq :
             m_taxids = []            
             m_taxids_dict = seq[MERGED_TAXID_COLUMN]
@@ -166,19 +179,22 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
                 fa_sn = None
                 tfa = None
             
-            seq[b"species"] = tsp
+            species_column[i] = tsp
-            seq[b"genus"] = tgn
+            genus_column[i] = tgn
-            seq[b"family"] = tfa
+            family_column[i] = tfa
                 
-            seq[b"species_name"] = sp_sn
+            species_name_column[i] = sp_sn
-            seq[b"genus_name"] = gn_sn
+            genus_name_column[i] = gn_sn
-            seq[b"family_name"] = fa_sn
+            family_name_column[i] = fa_sn
             
-            seq[b"rank"] = taxonomy.get_rank(taxid)
+            rank_column[i] = taxonomy.get_rank(taxid)
-            seq[b"scientific_name"] = taxonomy.get_scientific_name(taxid)
+            scientific_name_column[i] = taxonomy.get_scientific_name(taxid)
+        i+=1
+    
+    pb(len(o_view), force=True)
     
 
-cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list mergedKeys_list=None, Taxonomy taxonomy=None, bint mergeIds=False, list categories=None, int max_elts=1000000) :
+cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, dict config, list mergedKeys_list=None, Taxonomy taxonomy=None, bint mergeIds=False, list categories=None, int max_elts=1000000) :
      
     cdef int            i
     cdef int            k
@@ -187,6 +203,7 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     cdef int            u_idx
     cdef int            i_idx
     cdef int            i_count
+    cdef int            o_count
     cdef str            key_str
     cdef bytes          key
     cdef bytes          mkey
@@ -209,7 +226,6 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     cdef Nuc_Seq_Stored i_seq
     cdef Nuc_Seq_Stored o_seq
     cdef Nuc_Seq_Stored u_seq
-    cdef Column         i_col
     cdef Column         i_seq_col
     cdef Column         i_id_col
     cdef Column         i_taxid_col
@@ -217,6 +233,8 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     cdef Column         o_id_col
     cdef Column         o_taxid_dist_col
     cdef Column         o_merged_col
+    cdef Column         o_count_col
+    cdef Column         i_count_col
     cdef Column_line    i_mcol  
     cdef object         taxid_dist_dict
     cdef object         iter_view
@@ -253,6 +271,11 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     for k in range(k_count):
         mergedKeys_m.append(MERGED_PREFIX + mergedKeys[k])
     
+    # Check that not trying to remerge without total count information
+    for key in mergedKeys_m:
+        if key in view and COUNT_COLUMN not in view:
+            raise Exception("\n>>>>\nError: trying to re-merge tags without total count tag. Run obi annotate to add the count tag from the relevant merged tag, i.e.: \nobi annotate --set-tag COUNT:'sum([value for key,value in sequence['MERGED_sample'].items()])' dms/input dms/output\n")
+    
     if categories is None:
         categories = []
  
@@ -320,6 +343,10 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     for k in range(k_count):
         key = mergedKeys[k]
         merged_col_name = mergedKeys_m[k]
+        
+        if merged_col_name in view:
+            i_col = view[merged_col_name]
+        else:
             i_col = view[key]
         
         if merged_infos[merged_col_name]['nb_elts'] > max_elts:
@@ -374,23 +401,30 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
                           alias=MERGED_COLUMN
                          )
 
-    # Keep columns that are going to be used a lot in variables 
+    # Keep columns in variables for efficiency
     o_id_col = o_view[ID_COLUMN]
     if TAXID_DIST_COLUMN in o_view:
         o_taxid_dist_col = o_view[TAXID_DIST_COLUMN]
     if MERGED_COLUMN in o_view:
         o_merged_col = o_view[MERGED_COLUMN]
+    if COUNT_COLUMN not in o_view:
+        Column.new_column(o_view,
+                          COUNT_COLUMN,
+                          OBI_INT)
+    o_count_col = o_view[COUNT_COLUMN]
+    if COUNT_COLUMN in view:
+        i_count_col = view[COUNT_COLUMN]
         
     pb(len(view), force=True)
     print("")
     logger("info", "Second browsing through the input")
     # Initialize the progress bar
-    pb = ProgressBar(len(uniques), seconde=5)
+    pb = ProgressBar(len(view), seconde=5)
     o_idx = 0
+    total_treated = 0
     
     for unique_id in uniques :
         PyErr_CheckSignals()
-        pb(o_idx)
         
         merged_sequences = uniques[unique_id]
         
@@ -407,7 +441,7 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
             merged_list = list(set(merged_list)) # deduplicate the list
             o_merged_col[o_idx] = merged_list
 
-        o_seq[COUNT_COLUMN] = 0
+        o_count = 0
 
         if TAXID_DIST_COLUMN in u_seq and i_taxid_dist_col[u_idx] is not None:
             taxid_dist_dict = i_taxid_dist_col[u_idx]
@@ -419,16 +453,17 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
             merged_dict[mkey] = {}
 
         for i_idx in merged_sequences:
+            pb(total_treated)
  
             i_id = i_id_col[i_idx]
             i_seq = view[i_idx]
 
-            if COUNT_COLUMN not in i_seq or i_seq[COUNT_COLUMN] is None:
+            if COUNT_COLUMN not in i_seq or i_count_col[i_idx] is None:
                 i_count = 1
             else:
-                i_count = i_seq[COUNT_COLUMN]
+                i_count = i_count_col[i_idx]
  
-            o_seq[COUNT_COLUMN] += i_count
+            o_count += i_count
         
             for k in range(k_count):
                                 
@@ -464,12 +499,14 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
                             else:
                                 mcol[key2] = mcol[key2] + i_mcol[key2]
                                         
-            # Write taxid_dist
+            for key in i_seq.keys():
-            if mergeIds and TAXID_COLUMN in mergedKeys:
+                # Delete informations that differ between the merged sequences
-                if TAXID_DIST_COLUMN in str_merged_cols:
+                # TODO make special columns list? // could be more efficient
-                    o_taxid_dist_col[o_idx] = str(taxid_dist_dict)
+                if key != COUNT_COLUMN and key != ID_COLUMN and key != NUC_SEQUENCE_COLUMN and key in o_seq and o_seq[key] != i_seq[key] \
-                else:
+                    and key not in merged_dict :
-                    o_taxid_dist_col[o_idx] = taxid_dist_dict
+                    o_seq[key] = None
+            
+            total_treated += 1
             
         # Write merged dicts
         for mkey in merged_dict: 
@@ -482,25 +519,33 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
                 #    if mkey_cols[mkey][o_idx][key] is None:
                 #        mkey_cols[mkey][o_idx][key] = 0
 
-            for key in i_seq.keys():
+        # Write taxid_dist
-                # Delete informations that differ between the merged sequences
+        if mergeIds and TAXID_COLUMN in mergedKeys:
-                # TODO make special columns list?
+            if TAXID_DIST_COLUMN in str_merged_cols:
-                if key != COUNT_COLUMN and key != ID_COLUMN and key != NUC_SEQUENCE_COLUMN and key in o_seq and o_seq[key] != i_seq[key] \
+                o_taxid_dist_col[o_idx] = str(taxid_dist_dict)
-                    and key not in merged_dict :
+            else:
-                    o_seq[key] = None
+                o_taxid_dist_col[o_idx] = taxid_dist_dict
                 
+        o_count_col[o_idx] = o_count
         o_idx += 1
     
+    pb(len(view), force=True)
+    
     # Deletes quality columns if there is one because the matching between sequence and quality will be broken (quality set to NA when sequence not)
     if QUALITY_COLUMN in view:
         o_view.delete_column(QUALITY_COLUMN)
     if REVERSE_QUALITY_COLUMN in view:
         o_view.delete_column(REVERSE_QUALITY_COLUMN)
     
+    # Delete old columns that are now merged
+    for k in range(k_count):
+        if mergedKeys[k] in o_view:
+            o_view.delete_column(mergedKeys[k])
+    
     if taxonomy is not None:
         print("")  # TODO because in the middle of progress bar. Better solution?
         logger("info", "Merging taxonomy classification")
-        merge_taxonomy_classification(o_view, taxonomy)
+        merge_taxonomy_classification(o_view, taxonomy, config)
 
 
 
@@ -547,11 +592,10 @@ def run(config):
     pb = ProgressBar(len(entries), config, seconde=5)
     
     try:
-        uniq_sequences(entries, o_view, pb, mergedKeys_list=config['uniq']['merge'], taxonomy=taxo, mergeIds=config['uniq']['mergeids'], categories=config['uniq']['categories'], max_elts=config['obi']['maxelts'])       
+        uniq_sequences(entries, o_view, pb, config, mergedKeys_list=config['uniq']['merge'], taxonomy=taxo, mergeIds=config['uniq']['mergeids'], categories=config['uniq']['categories'], max_elts=config['obi']['maxelts'])       
     except Exception, e:
         raise RollbackException("obi uniq error, rollbacking view: "+str(e), o_view)
     
-    pb(len(entries), force=True)
     print("", file=sys.stderr)
 
     # Save command config in View and DMS comments
@@ -567,8 +611,8 @@ def run(config):
     #print("\n\nOutput view:\n````````````", file=sys.stderr)
     #print(repr(o_view), file=sys.stderr)
     
-    input[0].close()
+    input[0].close(force=True)
-    output[0].close()
+    output[0].close(force=True)
 
     logger("info", "Done.")
 

python/obitools3/dms/capi/obidmscolumn.pxd

@@ -63,6 +63,8 @@ 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)

python/obitools3/dms/column/column.pyx

@@ -14,6 +14,7 @@ from ..capi.obidms cimport obi_import_column
 from ..capi.obidmscolumn cimport OBIDMS_column_header_p, \
                                  obi_close_column, \
                                  obi_get_elements_names, \
+                                 obi_column_formatted_infos, \
                                  obi_column_write_comments
                                      
 from ..capi.obiutils cimport obi_format_date
@@ -288,7 +289,13 @@ 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)
     
          

python/obitools3/dms/dms.pyx

@@ -94,16 +94,16 @@ cdef class DMS(OBIWrapper):
         return dms
     
     
-    def close(self) :
+    def close(self, force=False) :
         '''
-        Closes the DMS instance and free the associated memory
+        Closes the DMS instance and free the associated memory (no counter, closing is final)
         
         The `close` method is automatically called by the object destructor.
         '''
         cdef OBIDMS_p pointer = self.pointer()
         if self.active() :
             OBIWrapper.close(self)
-            if (obi_close_dms(pointer, False)) < 0 :
+            if (obi_close_dms(pointer, force=force)) < 0 :
                 raise Exception("Problem closing an OBIDMS")
 
 
@@ -254,7 +254,8 @@ cdef class DMS(OBIWrapper):
     # bash command history property getter
     @property
     def bash_history(self):
-        s = b"#!/bin/bash\n\n"
+        #s = b"#!${bash}/bin/bash\n\n"
+        s = b""
         first = True
         for command in self.command_line_history:
             s+=b"#"

python/obitools3/dms/view/view.pyx

@@ -526,11 +526,12 @@ cdef class View(OBIWrapper) :
     # bash command history property getter
     @property
     def bash_history(self):
-        s = b"#!/bin/bash\n\n"
+        s = b""
         first = True
         for level in self.view_history:
             command_list = [level[input][b"command_line"] for input in level.keys()]
             for command in command_list:
+                s+=b"obi "
                 s+=command
                 s+=b"\n"
         return s

python/obitools3/libalign/_solexapairend.pyx

@@ -188,7 +188,7 @@ def buildConsensus(ali, seq, ref_tags=None):
         seq[b'shift']=ali.shift
     else:
         if len(ali[0])>999:   # TODO why?
-            raise AssertionError,"Too long alignemnt"
+            raise AssertionError,"Too long alignment"
     
         ic=IterOnConsensus(ali)
     
@@ -250,11 +250,21 @@ def buildJoinedSequence(ali, reverse, seq, forward=None):
     quality.extend(reverse.quality)
     seq.set(forward.id +b"_PairedEnd", s, definition=forward.definition, quality=quality)
     seq[b"score"]=ali.score
+    if len(ali.direction) > 0:
         seq[b"ali_direction"]=ali.direction
+    else:
+        seq[b"ali_direction"]=None
     seq[b"mode"]=b"joined"
     seq[b"pairedend_limit"]=len(forward)
+    seq[b"ali_length"] = ali.consensus_len
+    if ali.consensus_len > 0:
+        seq[b"score_norm"]=float(ali.score)/ali.consensus_len
+    else:
+        seq[b"score_norm"]=0.0
+  
     for tag in forward:
-        if tag != REVERSE_SEQUENCE_COLUMN and tag != REVERSE_QUALITY_COLUMN:
+        if tag != REVERSE_SEQUENCE_COLUMN and tag != REVERSE_QUALITY_COLUMN and \
+            tag != NUC_SEQUENCE_COLUMN and tag != QUALITY_COLUMN:
             seq[tag] = forward[tag]
     return seq
 

python/obitools3/parsers/embl.pyx

@@ -156,6 +156,9 @@ def emblIterator_file(lineiterator,
         yield seq
         read+=1
     
+    # Last sequence
+    seq = emblParser(entry)
+        
     yield seq
     
     free(entry)
@@ -174,7 +177,7 @@ def emblIterator_dir(dir_path,
     for filename in files:
         if read==only:
             return
-        print("Parsing file %s (%d/%d)" % (tostr(filename), read_files, len(files)))
+        print("Parsing file %s (%d/%d)" % (tostr(filename), read_files+1, len(files)))
         f = uopen(filename)
         if only is not None:
             only_f = only-read

python/obitools3/parsers/fasta.pyx

@@ -104,6 +104,7 @@ def fastaNucIterator(lineiterator,
     cdef bytes      sequence
     cdef int        skipped, ionly, read
     cdef Nuc_Seq    seq
+    cdef bint       stop
     
     if only is None:
         ionly = -1
@@ -130,7 +131,8 @@ def fastaNucIterator(lineiterator,
     else:
         line = firstline       
     
-    while True:
+    stop=False
+    while not stop:
                 
         if ionly >= 0 and read >= ionly:
             break
@@ -153,7 +155,7 @@ def fastaNucIterator(lineiterator,
                 s.append(line[0:-1])
                 line = next(iterator)
         except StopIteration:
-            pass
+            stop=True
         
         sequence  = b"".join(s)        
         

python/obitools3/parsers/genbank.pyx

@@ -25,8 +25,9 @@ from libc.string cimport strcpy, strlen
 _featureMatcher = re.compile(b'^FEATURES.+\n(?=ORIGIN)',re.DOTALL + re.M)
 
 _headerMatcher = re.compile(b'^LOCUS.+(?=\nFEATURES)', re.DOTALL + re.M)
-_seqMatcher    = re.compile(b'(?<=ORIGIN).+(?=//\n)', re.DOTALL + re.M)
+_seqMatcher    = re.compile(b'ORIGIN.+(?=//\n)', re.DOTALL + re.M)
-_cleanSeq      = re.compile(b'[ \n0-9]+')
+_cleanSeq1     = re.compile(b'ORIGIN.+\n')
+_cleanSeq2     = re.compile(b'[ \n0-9]+')
 _acMatcher     = re.compile(b'(?<=^ACCESSION   ).+',re.M)
 _deMatcher     = re.compile(b'(?<=^DEFINITION  ).+\n( .+\n)*',re.M)
 _cleanDe       = re.compile(b'\n *')
@@ -42,7 +43,8 @@ def genbankParser(bytes text):
         ft     = _featureMatcher.search(text).group()
         
         s      = _seqMatcher.search(text).group()
-        s      = _cleanSeq.sub(b'', s).upper()
+        s      = _cleanSeq1.sub(b'', s)
+        s      = _cleanSeq2.sub(b'', s)
         
         acs    = _acMatcher.search(text).group()
         acs    = acs.split()
@@ -52,12 +54,6 @@ def genbankParser(bytes text):
         de     = _deMatcher.search(header).group()
         de     = _cleanDe.sub(b' ',de).strip().strip(b'.')
     
-    except Exception as e:
-        print("\nCould not import sequence id:", text.split()[1], "(error raised:", e, ")")
-        # Do not raise any Exception if you need the possibility to resume the generator
-        # (Python generators can't resume after any exception is raised)
-        return None
-    
         tags = {}
         extractTaxon(ft, tags)
             
@@ -68,6 +64,12 @@ def genbankParser(bytes text):
                       offset=-1,
                       tags=tags)
     
+    except Exception as e:
+        print("\nCould not import sequence id:", text.split()[1], "(error raised:", e, ")")
+        # Do not raise any Exception if you need the possibility to resume the generator
+        # (Python generators can't resume after any exception is raised)
+        return None
+
     return seq
 
 
@@ -153,6 +155,9 @@ def genbankIterator_file(lineiterator,
         yield seq
         read+=1
     
+    # Last sequence
+    seq = genbankParser(entry)
+    
     yield seq
     
     free(entry)
@@ -168,10 +173,12 @@ def genbankIterator_dir(dir_path,
     read = 0
     read_files = 0
     files = [filename for filename in glob.glob(os.path.join(path, b'*.gbff*'))]
+    files.extend([filename for filename in glob.glob(os.path.join(path, b'*.seq*'))])  # new genbank extension
+    files = list(set(files))
     for filename in files:
         if read==only:
             return
-        print("Parsing file %s (%d/%d)" % (tostr(filename), read_files, len(files)))
+        print("Parsing file %s (%d/%d)" % (tostr(filename), read_files+1, len(files)))
         f = uopen(filename)
         if only is not None:
             only_f = only-read

python/obitools3/utils.pyx

@@ -72,7 +72,7 @@ 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":
+            if format != b"ngsfilter" and format != b"tabular" and format != b"embl" and format != b"genbank":
                 total_count += 1 # adding +1 for 1st entry because separators include \n (ngsfilter and tabular already count one more because of last \n)
             
     except:

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '0-beta10'
+serial= '0b24'
 
-version ="%d.%02d.%s" % (major,minor,serial)
+version ="%d.%d.%s" % (major,minor,serial)

requirements.txt

@@ -0,0 +1,5 @@
+--extra-index-url https://pypi.python.org/simple/
+Cython>=0.24
+Sphinx>=1.2.0
+ipython>=3.0.0
+breathe>=4.0.0

setup.py

@@ -5,7 +5,8 @@ import re
 import subprocess
 
 from distutils import log
-from distutils.core import setup
+#from distutils.core import setup
+from setuptools import setup    # to work with pip
 
 from distutils.core import Extension
 from distutils.sysconfig import get_python_lib
@@ -26,10 +27,11 @@ class Distribution(ori_Distribution):
         
         ori_Distribution.__init__(self, attrs)
         
-        self.global_options.insert(0,('cobitools3', None, "intall location of the C library"
+        self.global_options.insert(0,('cobitools3', None, "install location of the C library"
                                      ))
 
 from distutils.command.build import build as build_ori
+from setuptools.command.bdist_egg import bdist_egg as bdist_egg_ori
 from distutils.core import Command
 
 
@@ -70,6 +72,12 @@ class build(build_ori):
         build_ori.run(self)
 
 
+class bdist_egg(bdist_egg_ori):
+    def run(self):
+        self.run_command('build_clib')
+        bdist_egg_ori.run(self)
+
+
 sys.path.append(os.path.abspath("python"))
 
 
@@ -88,9 +96,10 @@ PACKAGE     = "OBITools3"
 VERSION     = version
 AUTHOR      = 'Celine Mercier'
 EMAIL       = 'celine.mercier@metabarcoding.org'
-URL         = "http://metabarcoding.org/obitools3"
+URL         = "https://metabarcoding.org/obitools3"
+PLATFORMS   = "posix"
 LICENSE     = "CeCILL-V2"
-DESCRIPTION = "Tools and library for DNA metabarcoding",
+DESCRIPTION = "A package for the management of analyses and data in DNA metabarcoding."
 PYTHONMIN   = '3.5'
 
 SRC       = 'python'
@@ -147,17 +156,24 @@ classifiers=['Development Status :: 4 - Beta',
              'Topic :: Utilities',
              ]
 
+with open("README.md", "r") as fh:
+    long_description = fh.read()
+
 setup(name=PACKAGE,
       description=DESCRIPTION,
+      long_description=long_description,
+      long_description_content_type="text/markdown",
       classifiers=classifiers,
       version=VERSION,
       author=AUTHOR,
       author_email=EMAIL,
+      platforms=PLATFORMS,
       license=LICENSE,
       url=URL,
       ext_modules=xx,
       distclass=Distribution,
       cmdclass={'build': build,
+                'bdist_egg': bdist_egg,
                 'build_clib': build_clib},
       cobitools3=get_python_lib(),
       packages = findPackage('python'),

src/build_reference_db.c

@@ -157,7 +157,7 @@ int build_reference_db(const char* dms_name,
 	ecotx_t* lca_2 = NULL;
 	ecotx_t* lca = NULL;
 	index_t idx1, idx2;
-	index_t i, j, k;
+	index_t i, j, k, count;
 	int32_t taxid_array_length;
 	int32_t score_array_length;
 	int32_t taxid_array_writable_length;
@@ -185,6 +185,7 @@ 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");
 	if (obi_lcs_align_one_column(dms_name,
 								 refs_view_name,
 								 "",
@@ -320,13 +321,19 @@ int build_reference_db(const char* dms_name,
 		return -1;
 	}
 
+	count = (matrix_with_lca_view->infos)->line_count;
+	fprintf(stderr, "Computing LCAs...\n");
+
 	// Compute all the LCAs
 		// For each pair
-	for (i=0; i<(matrix_with_lca_view->infos)->line_count; i++)
+	for (i=0; i<count; i++)
 	{
 		if (! keep_running)
 			return -1;
 
+		if (i%1000 == 0)
+			fprintf(stderr,"\rDone : %f %%       ", (i / (float) count)*100);
+
 		// Read all taxids associated with the first sequence and compute their LCA
 		// Read line index
 		idx1 = obi_get_int_with_elt_idx_and_col_p_in_view(matrix_with_lca_view, matrix_idx1_column, i, 0);
@@ -363,6 +370,7 @@ int build_reference_db(const char* dms_name,
 			return -1;
 		}
 	}
+	fprintf(stderr,"\rDone : 100 %%           \n");
 
 	// Clone refs view, add 2 arrays columns for lca and score, compute and write them
 
@@ -442,13 +450,18 @@ int build_reference_db(const char* dms_name,
 		return -1;
 	}
 
+	fprintf(stderr, "Building LCA arrays...\n");
+
 	// For each sequence, look for all its alignments in the matrix, and for each different LCA taxid/score, order them and write them
 	// Going through matrix once, filling refs arrays on the go for efficiency
-	for (i=0; i<(matrix_with_lca_view->infos)->line_count; i++)
+	for (i=0; i<count; i++)
 	{
 		if (! keep_running)
 			return -1;
 
+		if (i%1000 == 0)
+			fprintf(stderr,"\rDone : %f %%       ", (i / (float) count)*100);
+
 		// Read ref line indexes
 		idx1 = obi_get_int_with_elt_idx_and_col_p_in_view(matrix_with_lca_view, matrix_idx1_column, i, 0);
 		idx2 = obi_get_int_with_elt_idx_and_col_p_in_view(matrix_with_lca_view, matrix_idx2_column, i, 0);
@@ -464,6 +477,8 @@ int build_reference_db(const char* dms_name,
 		// Read alignment score
 		score = obi_get_float_with_elt_idx_and_col_p_in_view(matrix_with_lca_view, matrix_score_column, i, 0);
 
+		//fprintf(stderr, "\n\ntaxid_lca=%d, score=%f, idx1=%d, idx2=%d", taxid_lca, score, idx1, idx2);
+
 		///////////////// Compute for first sequence \\\\\\\\\\\\\\\\\\\\\\\     (TODO function)
 
 		// Read arrays
@@ -480,9 +495,11 @@ int build_reference_db(const char* dms_name,
 //			return -1;
 //		}
 
+		//fprintf(stderr, "\n1st sequence");
 		// If empty, add values
 		if (taxid_array_length == 0)
 		{
+			//fprintf(stderr, "\nEmpty, add value");
 			if (obi_set_array_with_col_p_in_view(o_view, final_lca_taxid_a_column, idx1, &taxid_lca, (uint8_t) (obi_sizeof(OBI_INT) * 8), 1) < 0)
 			{
 				obidebug(1, "\nError setting a LCA taxid array in a column when building a reference database");
@@ -496,6 +513,8 @@ int build_reference_db(const char* dms_name,
 		}
 		else
 		{
+			//fprintf(stderr, "\nNot empty");
+
 			j = 0;
 			modified = false;
 			while (j < taxid_array_length)
@@ -509,6 +528,9 @@ int build_reference_db(const char* dms_name,
 						memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 						modified = true;
 
+						//fprintf(stderr, "\nSame LCA, replace %d and %f with %d and %f", lca_taxid_array_writable[j],
+						//		score_array_writable[j], taxid_lca, score);
+
 						// Better score for the same LCA, replace this LCA/score pair
 						lca_taxid_array_writable[j] = taxid_lca;
 						score_array_writable[j] = score;
@@ -535,6 +557,8 @@ int build_reference_db(const char* dms_name,
 				{
 					if (score > score_array[j])
 					{
+						//fprintf(stderr, "\nInsert new");
+
 						memcpy(lca_taxid_array_writable, lca_taxid_array, taxid_array_length*sizeof(obiint_t));
 						memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 						modified = true;
@@ -579,10 +603,15 @@ int build_reference_db(const char* dms_name,
 				memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 				modified = true;
 
+				//fprintf(stderr, "\nAppend at the end");
+
 				// Append LCA
 				lca_taxid_array_writable[taxid_array_writable_length] = taxid_lca;
 				score_array_writable[score_array_writable_length] = score;
 
+				taxid_array_writable_length++;
+				score_array_writable_length++;
+
 				// Remove the previous (children) LCAs from the array if their score is equal or lower
 				while ((j>0) && (score_array_writable[j-1] <= score))
 				{
@@ -603,6 +632,13 @@ int build_reference_db(const char* dms_name,
 			// Write new arrays
 			if (modified)
 			{
+//				fprintf(stderr, "\n\nnew array:");
+//				for (k=0;k<taxid_array_writable_length;k++)
+//				{
+//					lca = obi_taxo_get_taxon_with_taxid(tax, lca_taxid_array_writable[k]);
+//					fprintf(stderr, "\nLCA=%d, %s, score=%f", lca_taxid_array_writable[k], lca->name, score_array_writable[k]);
+//				}
+
 				if (obi_set_array_with_col_p_in_view(o_view, final_lca_taxid_a_column, idx1, lca_taxid_array_writable, (uint8_t) (obi_sizeof(OBI_INT) * 8), taxid_array_writable_length) < 0)
 				{
 					obidebug(1, "\nError setting a LCA taxid array in a column when building a reference database");
@@ -632,9 +668,13 @@ int build_reference_db(const char* dms_name,
 //			return -1;
 //		}
 
+		//fprintf(stderr, "\n2nd sequence");
+
 		// If empty, add values
 		if (taxid_array_length == 0)
 		{
+			//fprintf(stderr, "\nEmpty, add value");
+
 			if (obi_set_array_with_col_p_in_view(o_view, final_lca_taxid_a_column, idx2, &taxid_lca, (uint8_t) (obi_sizeof(OBI_INT) * 8), 1) < 0)
 			{
 				obidebug(1, "\nError setting a LCA taxid array in a column when building a reference database");
@@ -648,6 +688,8 @@ int build_reference_db(const char* dms_name,
 		}
 		else
 		{
+			//fprintf(stderr, "\nNot empty");
+
 			j = 0;
 			modified = false;
 			while (j < taxid_array_length)
@@ -661,6 +703,9 @@ int build_reference_db(const char* dms_name,
 						memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 						modified = true;
 
+						//fprintf(stderr, "\nSame LCA, replace %d and %f with %d and %f", lca_taxid_array_writable[j],
+						//		score_array_writable[j], taxid_lca, score);
+
 						// Better score for the same LCA, replace this LCA/score pair
 						lca_taxid_array_writable[j] = taxid_lca;
 						score_array_writable[j] = score;
@@ -687,6 +732,8 @@ int build_reference_db(const char* dms_name,
 				{
 					if (score > score_array[j])
 					{
+						//fprintf(stderr, "\nInsert new");
+
 						memcpy(lca_taxid_array_writable, lca_taxid_array, taxid_array_length*sizeof(obiint_t));
 						memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 						modified = true;
@@ -727,6 +774,8 @@ int build_reference_db(const char* dms_name,
 
 			if (j == taxid_array_length) // same or parent LCA not found, need to be appended at the end
 			{
+				//fprintf(stderr, "\nAppend at the end");
+
 				memcpy(lca_taxid_array_writable, lca_taxid_array, taxid_array_length*sizeof(obiint_t));
 				memcpy(score_array_writable, score_array, score_array_length*sizeof(obifloat_t));
 				modified = true;
@@ -735,6 +784,9 @@ int build_reference_db(const char* dms_name,
 				lca_taxid_array_writable[taxid_array_writable_length] = taxid_lca;
 				score_array_writable[score_array_writable_length] = score;
 
+				taxid_array_writable_length++;
+				score_array_writable_length++;
+
 				// Remove the previous (children) LCAs from the array if their score is equal or lower
 				while ((j>0) && (score_array_writable[j-1] <= score))
 				{
@@ -769,11 +821,17 @@ int build_reference_db(const char* dms_name,
 			}
 		}
 	}
+	fprintf(stderr,"\rDone : 100 %%           \n");
 
+	fprintf(stderr, "Writing results...\n");
+	count = (o_view->infos)->line_count;
 	// Fill empty LCA informations (because filling from potentially sparse alignment matrix) with the sequence taxid
 	score=1.0;  // technically getting LCA of identical sequences
-	for (i=0; i<(o_view->infos)->line_count; i++)
+	for (i=0; i<count; i++)
 	{
+		if (i%1000 == 0)
+			fprintf(stderr,"\rDone : %f %%       ", (i / (float) count)*100);
+
 		obi_get_array_with_col_p_in_view(o_view, final_lca_taxid_a_column, i, &taxid_array_length);
 		if (taxid_array_length == 0)  // no LCA set
 		{
@@ -799,6 +857,7 @@ 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);
@@ -858,7 +917,6 @@ int build_reference_db(const char* dms_name,
 	free(matrix_view_name);
 	free(matrix_with_lca_view_name);
 
-	fprintf(stderr,"\rDone : 100 %%           \n");
 	return 0;
 }
 

src/kmer_similarity.c

@@ -413,7 +413,10 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	 	return NULL;
 	}
 
+	if (max_common_kmers > 0)
 		score = max_common_kmers + kmer_size - 1;  // aka the number of nucleotides in the longest stretch of kmers perfectly matching
+	else
+		score = 0;
 	abs_shift = abs(best_shift);
 
 	// Save result in Obi_ali structure
@@ -423,10 +426,15 @@ Obi_ali_p kmer_similarity(Obiview_p view1, OBIDMS_column_p column1, index_t idx1
 	ali->shift = abs_shift;
 	ali->consensus_seq = NULL;
 	ali->consensus_qual = NULL;
+	if (score == 0)
+		ali->direction[0] = '\0';
+	else
+	{
 		if (((best_shift <= 0) && (!switched_seqs)) || ((best_shift > 0) && switched_seqs))
 			strcpy(ali->direction, "left");
 		else
 			strcpy(ali->direction, "right");
+	}
 
 	// Build the consensus sequence if asked
 	if (build_consensus)

src/obi_ecopcr.c

@@ -105,7 +105,8 @@ static int create_output_columns(Obiview_p o_view, bool kingdom_mode);
  * @param o_temp1_column A pointer on the output column for the temperature for the first primer.
  * @param o_temp2_column A pointer on the output column for the temperature for the second primer.
  *
- * @retval 0 if the operation was successfully completed.
+ * @retval 0 if the sequence was skipped (taxid not found, warning printed).
+ * @retval 1 if the sequence was successfully printed to the output.
  * @retval -1 if an error occurred.
  *
  * @since July 2018
@@ -366,6 +367,17 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 
 	// 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);
+
+	// Get the taxon structure
+	main_taxon = obi_taxo_get_taxon_with_taxid(taxonomy, taxid);
+	if (main_taxon == NULL)
+	{
+		obidebug(1, "\nWarning: error reading the taxonomic information of a sequence. Seq id: %s, taxid: %d. Probably deprecated taxid. Skipping this sequence.", seq_id, taxid);
+		return 0;
+	}
+
 	ldelta = (pos1 <= keep_nucleotides)?pos1:keep_nucleotides;
 	rdelta = ((pos2+keep_nucleotides)>=seq_len)?seq_len-pos2:keep_nucleotides;
 
@@ -431,16 +443,7 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 	if (isnan(tm2))
 		tm2 = OBIFloat_NA;
 
-	// Get the taxon structure
-	main_taxon = obi_taxo_get_taxon_with_taxid(taxonomy, taxid);
-	if (main_taxon == NULL)
-	{
-		obidebug(1, "\nError reading the taxonomic information of a sequence");
-		return -1;
-	}
-
 	// Write sequence id
-	seq_id = obi_get_str_with_elt_idx_and_col_p_in_view(i_view, i_id_column, i_idx, 0);
 	if (obi_set_str_with_elt_idx_and_col_p_in_view(o_view, o_id_column, o_idx, 0, seq_id) < 0)
 	{
 		obidebug(1, "\nError writing the sequence id");
@@ -629,7 +632,7 @@ static int print_seq(Obiview_p i_view, Obiview_p o_view,
 		return -1;
 	}
 
-	return 0;
+	return 1;
 }
 
 
@@ -698,6 +701,7 @@ int obi_ecopcr(const char* i_dms_name,
 
 	obiint_t      taxid;
 	char*         sequence;
+	int			  printed;
 
 	SeqPtr        apatseq=NULL;
 	int32_t       o1Hits;
@@ -1057,14 +1061,14 @@ int obi_ecopcr(const char* i_dms_name,
 											length = 0;
 											if (posj > posi)
 												length = posj - posi - o1->patlen - o2->patlen;
-											if (posj < posi)
+											else if (circular > 0)
 												length = posj + apatseq->seqlen - posi - o1->patlen - o2->patlen;
 											if ((length>0) &&	// For when primers touch or overlap
 												(!min_len || (length >= min_len)) &&
 												(!max_len || (length <= max_len)))
 											{
 												// Print the found amplicon
-												if (print_seq(i_view, o_view,
+												printed = print_seq(i_view, o_view,
 														  	  i_idx, o_idx,
 															  taxonomy,
 															  sequence,
@@ -1090,11 +1094,13 @@ int obi_ecopcr(const char* i_dms_name,
 															  o_strand_column,
 															  o_primer1_column, o_primer2_column,
 															  o_error1_column, o_error2_column,
-															  o_temp1_column,  o_temp2_column) < 0)
+															  o_temp1_column,  o_temp2_column);
+												if (printed < 0)
 												{
 													obidebug(1, "\nError writing the ecopcr result");
 													return -1;
 												}
+												else if (printed > 0)
 													o_idx++;
 											}
 										}
@@ -1145,14 +1151,14 @@ int obi_ecopcr(const char* i_dms_name,
 											length = 0;
 											if (posj > posi)
 												length = posj - posi + 1  - o2->patlen - o1->patlen; /* - o1->patlen : deleted by <EC> (prior to the OBITools3) */
-											if (posj < posi)
+											else if (circular > 0)
 												length = posj + apatseq->seqlen - posi - o1->patlen - o2->patlen;
 											if ((length>0) &&	// For when primers touch or overlap
 												(!min_len || (length >= min_len)) &&
 												(!max_len || (length <= max_len)))
 											{
 												// Print the found amplicon
-												if (print_seq(i_view, o_view,
+												printed = print_seq(i_view, o_view,
 														  	  i_idx, o_idx,
 															  taxonomy,
 															  sequence,
@@ -1178,11 +1184,13 @@ int obi_ecopcr(const char* i_dms_name,
 															  o_strand_column,
 															  o_primer1_column, o_primer2_column,
 															  o_error1_column, o_error2_column,
-															  o_temp1_column,  o_temp2_column) < 0)
+															  o_temp1_column,  o_temp2_column);
+												if (printed < 0)
 												{
 													obidebug(1, "\nError writing the ecopcr result");
 													return -1;
 												}
+												else if (printed > 0)
 													o_idx++;
 											}
 										}
@@ -1224,7 +1232,7 @@ int obi_ecopcr(const char* i_dms_name,
 		return -1;
 	}
 
-	fprintf(stderr,"\rDone : 100 %%           ");
+	fprintf(stderr,"\rDone : 100 %%           \n");
 	return 0;
 
 	return 0;

src/obi_ecopcr.h

@@ -81,8 +81,8 @@
  * @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.
- * @param primer1 The first primer.
+ * @param primer1 The first primer, length must be less than or equal to 32 (because of apat lib limitation).
- * @param primer2 The second primer.
+ * @param primer2 The second primer, length must be less than or equal to 32 (because of apat lib limitation).
  * @param error_max The maximum number of errors allowed per primer for amplification.
  * @param min_len The minimum length of an amplicon.
  * @param max_len The maximum length of an amplicon.

src/obi_ecotag.c

@@ -71,9 +71,12 @@ static int create_output_columns(Obiview_p o_view);
  * @param name The assigned scientific name.
  * @param assigned_status_column A pointer on the column where the assigned status should be written.
  * @param assigned The assigned status (whether the sequence was assigned to a taxon or not).
- * @param best_match_column A pointer on the column where the list of ids of the best matches should be written.
+ * @param best_match_ids_column A pointer on the column where the list of ids of the best matches should be written.
  * @param best_match_ids The list of ids of the best matches as an array of the concatenated ids separated by '\0'.
  * @param best_match_ids_length The total length of the array of ids of best matches.
+ * @param best_match_taxids_column A pointer on the column where the list of taxids of the best matches should be written.
+ * @param best_match_taxids The list of taxids of the best matches as an array of the taxids.
+ * @param best_match_taxids_length The length of the array of taxids of best matches.
  * @param score_column A pointer on the column where the score should be written.
  * @param score The similarity score of the sequence with its best match(es).
  *
@@ -87,7 +90,8 @@ int print_assignment_result(Obiview_p output_view, index_t line,
 							OBIDMS_column_p assigned_taxid_column, int32_t taxid,
 							OBIDMS_column_p assigned_name_column, const char* name,
 							OBIDMS_column_p assigned_status_column, bool assigned,
-							OBIDMS_column_p best_match_column, const char* best_match_ids, int best_match_ids_length,
+							OBIDMS_column_p best_match_ids_column, const char* best_match_ids, int best_match_ids_length,
+							OBIDMS_column_p best_match_taxids_column, const int32_t* best_match_taxids, int best_match_taxids_length,
 							OBIDMS_column_p score_column, double score);
 
 
@@ -100,37 +104,44 @@ 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, ECOTAG_SCORE_COLUMN_NAME, true) < 0)
+	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)
 	{
 		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, ECOTAG_TAXID_COLUMN_NAME, true) < 0)
+	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)
 	{
 		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, ECOTAG_NAME_COLUMN_NAME, true) < 0)
+	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)
 	{
 		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, ECOTAG_STATUS_COLUMN_NAME, true) < 0)
+	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)
 	{
 		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, ECOTAG_BEST_MATCH_IDS_COLUMN_NAME, true) < 0)
+	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)
 	{
-		obidebug(1, "\nError creating the column for the array of ids of the best match in ecotag");
+		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)
+	{
+		obidebug(1, "\nError creating the column for the array of taxids of best matches in ecotag");
 		return -1;
 	}
 
@@ -142,7 +153,8 @@ int print_assignment_result(Obiview_p output_view, index_t line,
 							OBIDMS_column_p assigned_taxid_column, int32_t taxid,
 							OBIDMS_column_p assigned_name_column, const char* name,
 							OBIDMS_column_p assigned_status_column, bool assigned,
-							OBIDMS_column_p best_match_column, const char* best_match_ids, int best_match_ids_length,
+							OBIDMS_column_p best_match_ids_column, const char* best_match_ids, int best_match_ids_length,
+							OBIDMS_column_p best_match_taxids_column, const int32_t* best_match_taxids, int best_match_taxids_length,
 							OBIDMS_column_p score_column, double score)
 {
 	// Write the assigned taxid
@@ -167,9 +179,16 @@ int print_assignment_result(Obiview_p output_view, index_t line,
 	}
 
 	// Write the best match ids
-	if (obi_set_array_with_col_p_in_view(output_view, best_match_column, line, best_match_ids, (uint8_t)(sizeof(char)*8), best_match_ids_length) < 0)
+	if (obi_set_array_with_col_p_in_view(output_view, best_match_ids_column, line, best_match_ids, (uint8_t)(sizeof(char)*8), best_match_ids_length) < 0)
 	{
-		obidebug(1, "\nError writing a assignment status in a column when writing ecotag results");
+		obidebug(1, "\nError writing the array of best match ids in a column when writing ecotag results");
+		return -1;
+	}
+
+	// Write the best match taxids
+	if (obi_set_array_with_col_p_in_view(output_view, best_match_taxids_column, line, best_match_taxids, (uint8_t)(sizeof(OBI_INT)*8), best_match_taxids_length) < 0)
+	{
+		obidebug(1, "\nError writing the array of best match taxids in a column when writing ecotag results");
 		return -1;
 	}
 
@@ -235,6 +254,8 @@ int obi_ecotag(const char* dms_name,
 	char*     		best_match_ids;
 	char*			best_match_ids_to_store;
 	int32_t         best_match_ids_length;
+	int32_t*   		best_match_taxids;
+	int32_t*		best_match_taxids_to_store;
 	int				best_match_count;
 	int             buffer_size;
 	int 			best_match_ids_buffer_size;
@@ -263,7 +284,8 @@ int obi_ecotag(const char* dms_name,
 	OBIDMS_column_p assigned_taxid_column = NULL;
 	OBIDMS_column_p assigned_name_column = NULL;
 	OBIDMS_column_p assigned_status_column = NULL;
-	OBIDMS_column_p best_match_column = NULL;
+	OBIDMS_column_p best_match_ids_column = NULL;
+	OBIDMS_column_p best_match_taxids_column = NULL;
 	OBIDMS_column_p lca_taxid_a_column = NULL;
 	OBIDMS_column_p score_a_column = NULL;
 	OBIDMS_column_p ref_taxid_column = NULL;
@@ -396,7 +418,8 @@ int obi_ecotag(const char* dms_name,
 	assigned_taxid_column = obi_view_get_column(output_view, ECOTAG_TAXID_COLUMN_NAME);
 	assigned_name_column = obi_view_get_column(output_view, ECOTAG_NAME_COLUMN_NAME);
 	assigned_status_column = obi_view_get_column(output_view, ECOTAG_STATUS_COLUMN_NAME);
-	best_match_column = obi_view_get_column(output_view, ECOTAG_BEST_MATCH_IDS_COLUMN_NAME);
+	best_match_ids_column = obi_view_get_column(output_view, ECOTAG_BEST_MATCH_IDS_COLUMN_NAME);
+	best_match_taxids_column = obi_view_get_column(output_view, ECOTAG_BEST_MATCH_TAXIDS_COLUMN_NAME);
 	score_column = obi_view_get_column(output_view, ECOTAG_SCORE_COLUMN_NAME);
 
 	// Open the used reference columns
@@ -453,9 +476,17 @@ int obi_ecotag(const char* dms_name,
 		return -1;
 	}
 
+	best_match_taxids = (int32_t*) malloc(buffer_size* sizeof(int32_t));
+	if (best_match_taxids == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError allocating memory for the best match taxid array in ecotag");
+		return -1;
+	}
+
 	for (i=0; i < query_count; i++)
 	{
-		if (i%100 == 0)
+		if (i%1000 == 0)
 			fprintf(stderr,"\rDone : %f %%       ", (i / (float) query_count)*100);
 
 		best_match_count = 0;
@@ -514,7 +545,7 @@ int obi_ecotag(const char* dms_name,
 
 				// Store in best match array
 
-				// Grow match array if needed
+				// Grow match and taxid array if needed
 				if (best_match_count == buffer_size)
 				{
 					buffer_size = buffer_size*2;
@@ -525,6 +556,13 @@ int obi_ecotag(const char* dms_name,
 						obidebug(1, "\nError reallocating match array when assigning");
 						return -1;
 					}
+					best_match_taxids = (int32_t*) realloc(best_match_taxids, buffer_size*sizeof(int32_t));
+					if (best_match_taxids == NULL)
+					{
+						obi_set_errno(OBI_MALLOC_ERROR);
+						obidebug(1, "\nError reallocating match taxids array when assigning");
+						return -1;
+					}
 				}
 
 				id = obi_get_str_with_elt_idx_and_col_p_in_view(ref_view, ref_id_column, j, 0);
@@ -545,6 +583,7 @@ 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++;
 				strcpy(best_match_ids+best_match_ids_length, id);
 				best_match_ids_length = best_match_ids_length + id_len + 1;
@@ -562,7 +601,7 @@ int obi_ecotag(const char* dms_name,
 			score_array = obi_get_array_with_col_p_in_view(ref_view, score_a_column, best_match_idx, &lca_array_length);
 
 			k = 0;
-			while ((k < lca_array_length) && (score_array[k] >= ecotag_threshold))
+			while ((k < lca_array_length) && (score_array[k] >= best_score))
 				k++;
 
 			if (k>0)
@@ -570,12 +609,12 @@ int obi_ecotag(const char* dms_name,
 				lca_array = obi_get_array_with_col_p_in_view(ref_view, lca_taxid_a_column, best_match_idx, &lca_array_length);
 				if (j>0)
 				{
-					lca = obi_taxo_get_taxon_with_taxid(taxonomy, lca_taxid);
+//					lca = obi_taxo_get_taxon_with_taxid(taxonomy, lca_taxid);
-					if (lca == NULL)
+//					if (lca == NULL)
-					{
+//					{
-						obidebug(1, "\nError getting a taxon from a taxid when doing taxonomic assignment");
+//						obidebug(1, "\nError getting a taxon from a taxid when doing taxonomic assignment");
-						return -1;
+//						return -1;
-					}
+//					}
 					lca_in_array = obi_taxo_get_taxon_with_taxid(taxonomy, lca_array[k-1]);
 					if (lca_in_array == NULL)
 					{
@@ -629,6 +668,7 @@ int obi_ecotag(const char* dms_name,
 			else
 				lca_name = lca->name;
 			best_match_ids_to_store = best_match_ids;
+			best_match_taxids_to_store = best_match_taxids;
 		}
 		else
 		{
@@ -636,6 +676,7 @@ int obi_ecotag(const char* dms_name,
 			lca_name = OBIStr_NA;
 			lca_taxid = OBIInt_NA;
 			best_match_ids_to_store = OBITuple_NA;
+			best_match_taxids_to_store = OBITuple_NA;
 			score = OBIFloat_NA;
 		}
 
@@ -644,7 +685,8 @@ int obi_ecotag(const char* dms_name,
 								    assigned_taxid_column, lca_taxid,
 									assigned_name_column, lca_name,
 									assigned_status_column, assigned,
-									best_match_column, best_match_ids_to_store, best_match_ids_length,
+									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,
 									score_column, best_score
 									) < 0)
 							return -1;
@@ -652,6 +694,7 @@ int obi_ecotag(const char* dms_name,
 
 	free(best_match_array);
 	free(best_match_ids);
+	free(best_match_taxids);
 
 	obi_close_taxonomy(taxonomy);
 	obi_save_and_close_view(query_view);

src/obi_ecotag.h

@@ -23,7 +23,8 @@
 #define ECOTAG_TAXID_COLUMN_NAME "TAXID"
 #define ECOTAG_NAME_COLUMN_NAME "SCIENTIFIC_NAME"
 #define ECOTAG_STATUS_COLUMN_NAME "ID_STATUS"
-#define ECOTAG_BEST_MATCH_IDS_COLUMN_NAME "BEST_MATCH"
+#define ECOTAG_BEST_MATCH_IDS_COLUMN_NAME "BEST_MATCH_IDS"
+#define ECOTAG_BEST_MATCH_TAXIDS_COLUMN_NAME "BEST_MATCH_TAXIDS"
 #define ECOTAG_SCORE_COLUMN_NAME "BEST_IDENTITY"
 
 

src/obiavl.c

@@ -648,7 +648,7 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 	new_data_size = ((index_t) multiple) * getpagesize();
 
 	// 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)
+	if ((avl_data->header)->data_size_max >= new_data_size)
 		return 0;
 
 	// Get the file descriptor
@@ -667,7 +667,7 @@ int truncate_avl_data_to_size_used(OBIDMS_avl_data_p avl_data)	// TODO is it nec
 	if (ftruncate(file_descriptor, file_size) < 0)
 	{
 		obi_set_errno(OBI_AVL_ERROR);
-		obidebug(1, "\nError truncating an AVL data file");
+		obidebug(1, "\nError truncating an AVL data file, old data size = %lld, new data size = %lld", (avl_data->header)->data_size_max, new_data_size);
 		return -1;
 	}
 

src/obidms.c

@@ -1496,7 +1496,7 @@ obiversion_t obi_import_column(const char* dms_path_1, const char* dms_path_2, c
 	memcpy(column_2->data, column_1->data, header_1->data_size);
 
 	// Copy the AVL files if there are some (overwriting the automatically created files)
-	if ((header_1->returned_data_type == OBI_STR) || (header_1->returned_data_type == OBI_SEQ) || (header_1->returned_data_type == OBI_QUAL))
+	if ((header_1->tuples) || ((header_1->returned_data_type == OBI_STR) || (header_1->returned_data_type == OBI_SEQ) || (header_1->returned_data_type == OBI_QUAL)))
 	{
 		avl_name_1 = (char*) malloc((strlen(header_1->indexer_name) + 1)  * sizeof(char));
 		if (avl_name_1 == NULL)

src/obidms_taxonomy.c

@@ -2376,9 +2376,10 @@ int read_merged_dmp(const char* taxdump, OBIDMS_taxonomy_p tax, int32_t* delnode
 				// and the deleted taxids with no current reference. An element of the list is composed of the taxid, and the index
 				// of the taxon in the taxa structure, or -1 for deleted taxids.
 				// Creating the merged list requires to merge the 3 ordered lists into one.
-				while (((nT < (tax->taxa)->count) && ((tax->taxa)->taxon[nT].taxid < old_taxid)) || ((nD >= 0) && (delnodes[nD] < old_taxid)))
+				while (((nT < (tax->taxa)->count) && ((tax->taxa)->taxon[nT].taxid < old_taxid)) ||
+						((nD >= 0) && (delnodes[nD] < old_taxid)))
 				{
-					if ((tax->taxa)->taxon[nT].taxid < delnodes[nD])
+					if ((nT < (tax->taxa)->count) && (tax->taxa)->taxon[nT].taxid < delnodes[nD])
 					{ // Add element from taxa list
 						// Enlarge structure if needed
 						if (n == buffer_size)
@@ -2401,7 +2402,7 @@ int read_merged_dmp(const char* taxdump, OBIDMS_taxonomy_p tax, int32_t* delnode
 						nT++;
 						n++;
 					}
-					else if (delnodes[nD] < (tax->taxa)->taxon[nT].taxid)
+					else
 					{ // Add element from deleted taxids list
 						// Enlarge structure if needed
 						if (n == buffer_size)
@@ -3036,12 +3037,12 @@ OBIDMS_taxonomy_p obi_read_taxonomy(OBIDMS_p dms, const char* taxonomy_name, boo
 
 	strcpy(tax->tax_name, taxonomy_name);
 
-	buffer_size = 2048;
-
 	taxonomy_path = get_taxonomy_path(dms, taxonomy_name);
 	if (taxonomy_path == NULL)
 		return NULL;
 
+	buffer_size = strlen(taxonomy_path) + strlen(taxonomy_name) + 6;
+
 	// Read ranks
 	ranks_file_name = (char*) malloc(buffer_size*sizeof(char));
 	if (ranks_file_name == NULL)

src/obidmscolumn.c

@@ -1693,8 +1693,8 @@ int obi_close_column(OBIDMS_column_p column)
 		if (obi_dms_unlist_column(column->dms, column) < 0)
 			ret_val = -1;
 
-		// If the data type is OBI_STR, OBI_SEQ or OBI_QUAL, the associated indexer is closed
+		// If it's a tuple column or the data type is OBI_STR, OBI_SEQ or OBI_QUAL, the associated indexer is closed
-		if (((column->header)->returned_data_type == OBI_STR) || ((column->header)->returned_data_type == OBI_SEQ) || ((column->header)->returned_data_type == OBI_QUAL))
+		if (((column->header)->tuples) || (((column->header)->returned_data_type == OBI_STR) || ((column->header)->returned_data_type == OBI_SEQ) || ((column->header)->returned_data_type == OBI_QUAL)))
 			if (obi_close_indexer(column->indexer) < 0)
 				ret_val = -1;
 
@@ -1973,7 +1973,11 @@ int obi_enlarge_column(OBIDMS_column_p column)
 
 	// Calculate the new file size
 	old_line_count = (column->header)->line_count;
-	new_line_count = old_line_count * COLUMN_GROWTH_FACTOR;
+	new_line_count = ceil((double) old_line_count * (double) COLUMN_GROWTH_FACTOR);
+	if (new_line_count > old_line_count+100000)
+		new_line_count = old_line_count+100000;
+	else if (new_line_count < old_line_count+1000)
+		new_line_count = old_line_count+1000;
 
 	if (new_line_count > MAXIMUM_LINE_COUNT)
 	{
@@ -2381,6 +2385,54 @@ char* obi_get_elements_names(OBIDMS_column_p column)
 }
 
 
+char* obi_get_formatted_elements_names(OBIDMS_column_p column)
+{
+	char* elements_names;
+	int   i, j;
+	int   elt_idx;
+	int   len;
+
+	elements_names = (char*) malloc(((column->header)->elements_names_length + (column->header)->nb_elements_per_line) * sizeof(char));
+	if (elements_names == NULL)
+	{
+		obi_set_errno(OBI_MALLOC_ERROR);
+		obidebug(1, "\nError allocating memory for elements names");
+		return NULL;
+	}
+
+	j = 0;
+	for (i=0; i < (column->header)->nb_elements_per_line; i++)
+	{
+		elt_idx = ((column->header)->elements_names_idx)[i];
+		len = strlen(((column->header)->elements_names)+elt_idx);
+		memcpy(elements_names+j, ((column->header)->elements_names)+elt_idx, len*sizeof(char));
+		j = j + len;
+		elements_names[j] = ';';
+		j++;
+		elements_names[j] = ' ';
+		j++;
+	}
+
+	elements_names[j - 1] = '\0';
+
+	return elements_names;
+}
+
+
+char* obi_column_formatted_infos(OBIDMS_column_p column)
+{
+	char* column_infos;
+	char* elt_names;
+
+	column_infos = malloc(1024 * sizeof(char));
+
+	elt_names = obi_get_formatted_elements_names(column);
+
+
+	free(elt_names);
+	return column_infos;
+}
+
 
 int obi_column_prepare_to_set_value(OBIDMS_column_p column, index_t line_nb, index_t elt_idx)
 {

src/obidmscolumn.h

@@ -505,6 +505,14 @@ 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);
+
+
+// TODO
+//char* obi_column_formatted_infos(OBIDMS_column_p column);
+
+
 /**
  * @brief Prepares a column to set a value.
  *

src/obiview.c

@@ -1037,8 +1037,9 @@ static int finish_view(Obiview_p view)
 		return -1;
 	}
 
-	// Add count column if it's a NUC_SEQ_VIEW with no count column // TODO discuss
+	// Add count column if it's a NUC_SEQ_VIEW with no count column (and there's no MERGED_sample column) // TODO discuss
-	if ((!strcmp((view->infos)->view_type, VIEW_TYPE_NUC_SEQS)) && (!obi_view_column_exists(view, COUNT_COLUMN)))
+	if ((!strcmp((view->infos)->view_type, VIEW_TYPE_NUC_SEQS)) && (!obi_view_column_exists(view, COUNT_COLUMN))
+			&& (!obi_view_column_exists(view, "MERGED_sample")))  // TODO should eventually compute from merged samples?
 	{
 		if (obi_create_auto_count_column(view) < 0)
 		{
@@ -1407,7 +1408,7 @@ static char* view_check_qual_match_seqs(Obiview_p view)
 						// Test that the lengths of the quality and the sequence are equal
 						if ((size_t)qual_len != strlen(seq))
 						{
-							obidebug(1, "\nError checking the predicate for view %s: The sequences and sequence quality arrays match.", (view->infos)->name);
+							obidebug(1, "\nError checking the predicate for view %s: The sequences and sequence quality arrays match (index %lld, seq=%s, quality length = %d).", (view->infos)->name, j, seq, qual_len);
 							return NULL;
 						}
 					}

src/sse_banded_LCS_alignment.c

@@ -686,6 +686,9 @@ int calculateSizeToAllocate(int maxLen, int LCSmin)
 	size *=  3;
 	size +=  16;
 
+	size += 10;  // band-aid for memory bug I don't understand (triggered on specific db on ubuntu)
+				 // bug might have to do with the way different systems behave when aligning the address in obi_get_memory_aligned_on_16
+
 	return(size*sizeof(int16_t));
 }