obi import: fixed a bug when skipping an entry

triggering error

python/obitools3/commands/import.pyx

@@ -260,7 +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)
@@ -270,6 +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
@@ -388,6 +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:

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
 
 

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '0b20'
+serial= '0b22'
 
 version ="%d.%d.%s" % (major,minor,serial)