Python: fasta parser: fixed a bug stopping the program when the last

line contained a single nucleotide

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:
+                if taxo is not None:
-                environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
-            else:
+                else:
-                environ = {'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'sequence':seq, 'counter':counter[0], 'math':math}
-            val = eval(v, environ, seq)
+                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:
+                if taxo is not None:
-                environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
-            else:
+                else:
-                environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
+                    environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
-            val = eval(newId, environ, seq)
+                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:
+                if taxo is not None:
-                environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
-            else:
+                else:
-                environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
+                    environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
-            val = eval(newDef, environ, seq)
+                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:
+                if taxo is not None:
-                environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
-            else:
+                else:
-                environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
+                    environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
-            val = eval(newSeq, environ, seq)
+                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:
+                if taxo is not None:
-                environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
+                    environ = {'taxonomy' : taxo, 'sequence':seq, 'counter':counter[0], 'math':math}
-            else:
+                else:
-                environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
+                    environ = {'sequence':seq, 'counter':counter[0], 'math':math}     
-            eval(run, environ, seq)
+                eval(run, environ, seq)
-#            except Exception,e:
+            except Exception,e:
-#                if options.onlyValid:
+                raise e
-#                    raise e
 
     return sequenceTagger
 

python/obitools3/commands/ecotag.pyx

@@ -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")

python/obitools3/commands/export.pyx

@@ -59,13 +59,23 @@ def run(config):
      # Check that the input view has the type NUC_SEQS if needed    # TODO discuss, maybe bool property
     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 :
         PyErr_CheckSignals()

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
 
@@ -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:
@@ -288,6 +292,8 @@ 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 +334,8 @@ 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
@@ -402,7 +408,7 @@ def run(config):
     except AttributeError:
         pass
     try:
-        output[0].close()
+        output[0].close(force=True)
     except AttributeError:
         pass
 

python/obitools3/commands/ngsfilter.pyx

@@ -42,8 +42,8 @@ def addOptions(parser):
                      metavar="<URI>",
                      type=str,
                      default=None,
-                     help="URI to the view containing the samples definition (with tags, primers, sample names,...)"
+                     help="URI to the view containing the samples definition (with tags, primers, sample names,...).\n"
-                          "Warning: primer lengths must be less than or equal to 32")
+                          "\nWarning: primer lengths must be less than or equal to 32")
 
     group.add_argument('-R', '--reverse-reads',
                      action="store", dest="ngsfilter:reverse",

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"]  
-    for seq in o_view:     
+    
-        PyErr_CheckSignals()   
+    # 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]
@@ -165,20 +178,23 @@ cdef merge_taxonomy_classification(View_NUC_SEQS o_view, Taxonomy taxonomy) :
             else:
                 fa_sn = None
                 tfa = None
-                
-            seq[b"species"] = tsp
-            seq[b"genus"] = tgn
-            seq[b"family"] = tfa
-                
-            seq[b"species_name"] = sp_sn
-            seq[b"genus_name"] = gn_sn
-            seq[b"family_name"] = fa_sn
             
-            seq[b"rank"] = taxonomy.get_rank(taxid)
+            species_column[i] = tsp
-            seq[b"scientific_name"] = taxonomy.get_scientific_name(taxid)
+            genus_column[i] = tgn
+            family_column[i] = tfa
+                
+            species_name_column[i] = sp_sn
+            genus_name_column[i] = gn_sn
+            family_name_column[i] = fa_sn
+            
+            rank_column[i] = taxonomy.get_rank(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, dict config, 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, 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
@@ -252,7 +270,12 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
     mergedKeys_m = []
     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 = []
  
@@ -320,7 +343,11 @@ 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]
-        i_col = view[key]
+        
+        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:
             str_merged_cols.append(merged_col_name)
@@ -374,12 +401,19 @@ 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("")
@@ -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]
@@ -423,12 +457,12 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
             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):
                                 
@@ -463,44 +497,52 @@ cdef uniq_sequences(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, li
                                 mcol[key2] = i_mcol[key2]
                             else:
                                 mcol[key2] = mcol[key2] + i_mcol[key2]
-            
+                                        
-            # Write taxid_dist
-            if mergeIds and TAXID_COLUMN in mergedKeys:
-                if TAXID_DIST_COLUMN in str_merged_cols:
-                    o_taxid_dist_col[o_idx] = str(taxid_dist_dict)
-                else:
-                    o_taxid_dist_col[o_idx] = taxid_dist_dict
-            
-            # Write merged dicts
-            for mkey in merged_dict: 
-                if mkey in str_merged_cols:
-                    mkey_cols[mkey][o_idx] = str(merged_dict[mkey])
-                else:
-                    mkey_cols[mkey][o_idx] = merged_dict[mkey]
-                    # Sets NA values to 0  # TODO discuss, for now keep as None and test for None instead of testing for 0 in tools
-                    #for key in mkey_cols[mkey][o_idx]:
-                    #    if mkey_cols[mkey][o_idx][key] is None:
-                    #        mkey_cols[mkey][o_idx][key] = 0
-                            
             for key in i_seq.keys():
                 # Delete informations that differ between the merged sequences
-                # TODO make special columns list?
+                # TODO make special columns list? // could be more efficient
                 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] \
                     and key not in merged_dict :
                     o_seq[key] = None
 
+        # Write merged dicts
+        for mkey in merged_dict: 
+            if mkey in str_merged_cols:
+                mkey_cols[mkey][o_idx] = str(merged_dict[mkey])
+            else:
+                mkey_cols[mkey][o_idx] = merged_dict[mkey]
+                # Sets NA values to 0  # TODO discuss, for now keep as None and test for None instead of testing for 0 in tools
+                #for key in mkey_cols[mkey][o_idx]:
+                #    if mkey_cols[mkey][o_idx][key] is None:
+                #        mkey_cols[mkey][o_idx][key] = 0
+
+        # Write taxid_dist
+        if mergeIds and TAXID_COLUMN in mergedKeys:
+            if TAXID_DIST_COLUMN in str_merged_cols:
+                o_taxid_dist_col[o_idx] = str(taxid_dist_dict)
+            else:
+                o_taxid_dist_col[o_idx] = taxid_dist_dict
+                
+        o_count_col[o_idx] = o_count
         o_idx += 1
     
+    pb(len(uniques), 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 +589,10 @@ 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 +608,8 @@ 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/view/view.pyx

@@ -531,8 +531,8 @@ cdef class View(OBIWrapper) :
         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/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/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '0-beta14'
+serial= '0-beta17'
 
 version ="%d.%02d.%s" % (major,minor,serial)

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/obi_ecopcr.c

@@ -1061,7 +1061,7 @@ 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)) &&
@@ -1151,7 +1151,7 @@ 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)) &&
@@ -1232,7 +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_ecotag.c

@@ -455,7 +455,7 @@ int obi_ecotag(const char* dms_name,
 
 	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;
@@ -562,7 +562,7 @@ 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 +570,12 @@ 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)
 					{

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/obidmscolumn.c

@@ -1974,6 +1974,10 @@ int obi_enlarge_column(OBIDMS_column_p column)
 	// Calculate the new file size
 	old_line_count = (column->header)->line_count;
 	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)
 	{

src/obidmscolumn.h

@@ -34,7 +34,7 @@
 #define NB_ELTS_MAX_IF_DEFAULT_NAME (1000000) 	/**< The maximum number of elements per line if the default element names
 										   	   	 *   are used ("0\01\02\0...\0n"), considering ELEMENTS_NAMES_MAX.  // TODO not up to date
 										   	   	 */
-#define COLUMN_GROWTH_FACTOR (1.3)	 	  		/**< The growth factor when a column is enlarged.
+#define COLUMN_GROWTH_FACTOR (2)	 	  		/**< The growth factor when a column is enlarged.
                                 	   	   	   	 */
 #define MAXIMUM_LINE_COUNT (1000000000)   		/**< The maximum line count for the data of a column (1E9). //TODO
                                 	       	   	 */

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)
 		{

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));
 }