switch to version 3.0.1b7

should not have

python/obitools3/apps/optiongroups/__init__.py

@@ -1,3 +1,9 @@
+import codecs
+
+def unescaped_str(arg_str):
+    return arg_str.encode('latin-1', 'backslashreplace').decode('unicode-escape')
+
+
 def __addInputOption(optionManager):
     
     optionManager.add_argument(
@@ -133,7 +139,7 @@ def __addTabularOption(optionManager):
     group.add_argument('--sep',
                      action="store", dest="obi:sep",
                      default="\t",
-                     type=str,
+                     type=unescaped_str,
                      help="Column separator")
 
 

python/obitools3/commands/addtaxids.pyx

@@ -26,7 +26,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
  
-__title__="Annotate sequences with their corresponding NCBI taxid found from the taxon scientific name."
+__title__="Annotate sequences with their corresponding NCBI taxid found from the taxon scientific name"
 
 
  

python/obitools3/commands/align.pyx

@@ -19,7 +19,7 @@ import time
 import sys
 
 
-__title__="Aligns one sequence column with itself or two sequence columns"
+__title__="Align one sequence column with itself or two sequence columns"
 
 
 def addOptions(parser):

python/obitools3/commands/alignpairedend.pyx

@@ -23,7 +23,7 @@ import os
 
 from cpython.exc cimport PyErr_CheckSignals
 
-__title__="Aligns paired-ended reads"
+__title__="Align paired-ended reads"
 
 
 

python/obitools3/commands/build_ref_db.pyx

@@ -16,7 +16,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Tag a set of sequences for PCR and sequencing errors identification"
+__title__="Build a reference database for ecotag"
 
 
 def addOptions(parser):

python/obitools3/commands/cat.pyx

@@ -22,7 +22,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Concatenate views."
+__title__="Concatenate views"
 
  
 def addOptions(parser):

python/obitools3/commands/count.pyx

@@ -10,7 +10,7 @@ from obitools3.dms.capi.obiview cimport COUNT_COLUMN
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Counts sequence records"
+__title__="Count sequence records"
 
  
 def addOptions(parser):

python/obitools3/commands/head.pyx

@@ -16,7 +16,7 @@ from io import BufferedWriter
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Keep the N first lines of a view."
+__title__="Keep the N first lines of a view"
 
  
 def addOptions(parser):

python/obitools3/commands/import.pyx

@@ -52,7 +52,7 @@ from cpython.exc cimport PyErr_CheckSignals
 from io import BufferedWriter
 
 
-__title__="Imports sequences from different formats into a DMS"
+__title__="Import sequences from different formats into a DMS"
  
  
 default_config = {   'destview'     : None,
@@ -430,7 +430,7 @@ def run(config):
     if pb is not None:
         pb(i, force=True)
         print("", file=sys.stderr)
-    logger("info", "Imported %d entries", i)
+    logger("info", "Imported %d entries", len(view))
     
     # Save command config in View and DMS comments
     command_line = " ".join(sys.argv[1:])

python/obitools3/commands/ngsfilter.pyx

@@ -24,7 +24,7 @@ from cpython.exc cimport PyErr_CheckSignals
 from io import BufferedWriter
 
 
-__title__="Assigns sequence records to the corresponding experiment/sample based on DNA tags and primers"
+__title__="Assign sequence records to the corresponding experiment/sample based on DNA tags and primers"
 
 
 def addOptions(parser):
@@ -450,7 +450,7 @@ cdef tuple annotate(sequences, infos, no_tags, verbose=False):
             sequences[1] = sequences[1][reversematch[1][2]:]
             if not directmatch[0].forward:
                 sequences[1] = sequences[1].reverse_complement
-            sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq           # used by alignpairedend tool
+            sequences[0][REVERSE_SEQUENCE_COLUMN] = sequences[1].seq      # used by alignpairedend tool
             sequences[0][REVERSE_QUALITY_COLUMN] = sequences[1].quality   # used by alignpairedend tool
         else:
             sequences[0] = sequences[0][reversematch[1][2]:]

python/obitools3/commands/sort.pyx

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

python/obitools3/commands/stats.pyx

@@ -16,7 +16,7 @@ import sys
 from cpython.exc cimport PyErr_CheckSignals
 
 
-__title__="Compute basic statistics for attribute values."
+__title__="Compute basic statistics for attribute values"
 
 '''
 `obi stats` computes basic statistics for attribute values of sequence records.

python/obitools3/commands/tail.pyx

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

python/obitools3/commands/test.pyx

@@ -39,7 +39,7 @@ COL_COMMENTS_MAX_LEN = 2048
 MAX_INT = 2147483647    # used to generate random float values
  
  
-__title__="Tests if the obitools are working properly"
+__title__="Test if the obitools are working properly"
 
 
 default_config = {

python/obitools3/uri/decode.pyx

@@ -389,10 +389,7 @@ def open_uri(uri,
             sep = tobytes(qualifiers[b"sep"][0][0])
         else:
             try:
-                sep = config["obi"]["sep"]
-                if sep == '\\t':   # dirty workaround for flake8(?) issue that reads '\t' as '\'+'t' when parsing the option value
-                    sep = '\t'
-                sep = tobytes(sep)
+                sep = tobytes(config["obi"]["sep"])
             except KeyError:
                 sep=None
         
@@ -568,6 +565,6 @@ def open_uri(uri,
         
         entry_count = -1
         if input:
-            entry_count = count_entries(file, format)
+            entry_count = count_entries(file, format, header)
         
         return (file, iseq, objclass, urib, entry_count)

python/obitools3/utils.pxd

@@ -3,7 +3,7 @@
 from obitools3.dms.capi.obitypes cimport obitype_t, index_t
 
 cpdef bytes format_uniq_pattern(bytes format)
-cpdef int count_entries(file, bytes format)
+cpdef int count_entries(file, bytes format, bint header)
 
 cdef obi_errno_to_exception(index_t line_nb=*, object elt_id=*, str error_message=*)
 

python/obitools3/utils.pyx

@@ -40,7 +40,7 @@ cpdef bytes format_uniq_pattern(bytes format):
         return None
 
 
-cpdef int count_entries(file, bytes format):
+cpdef int count_entries(file, bytes format, bint header):
     
     try:
         sep = format_uniq_pattern(format)
@@ -75,6 +75,8 @@ cpdef int count_entries(file, bytes format):
             total_count += len(re.findall(sep, mmapped_file))
             if format != b"ngsfilter" and format != b"tabular" and format != b"embl" and format != b"genbank" and format != b"fastq":
                 total_count += 1 # adding +1 for 1st entry because separators include \n (ngsfilter and tabular already count one more because of last \n)
+            if format == b"tabular" and header: # not counting header as an entry
+                total_count -= 1
             
     except:
         if len(files) > 1:

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '1b6'
+serial= '1b7'
 
 version ="%d.%d.%s" % (major,minor,serial)

src/kmer_similarity.c

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