From e8090a44c9d90cbfcdb3d82245628267e4fa2458 Mon Sep 17 00:00:00 2001
From: Celine Mercier <celine.mercier@metabarcoding.org>
Date: Fri, 15 Mar 2019 16:06:06 +0100
Subject: [PATCH] Fixed the ultimate bug with embl (and genbank) parsers:
 raising any exception in a python generator makes it unable to resume. So
 now, exceptions are not raised but printed, then functions return None and
 that's handled at higher level.

---
 python/obitools3/commands/import.pyx |  8 ++++++++
 python/obitools3/parsers/embl.pyx    | 12 ++++++------
 2 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/python/obitools3/commands/import.pyx b/python/obitools3/commands/import.pyx
index ecf8479..53a63b1 100755
--- a/python/obitools3/commands/import.pyx
+++ b/python/obitools3/commands/import.pyx
@@ -5,6 +5,7 @@ import os
 
 from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
 from obitools3.dms.view.view cimport View
+from obitools3.dms.view import RollbackException
 from obitools3.dms.view.typed_view.view_NUC_SEQS cimport View_NUC_SEQS
 from obitools3.dms.column.column cimport Column
 from obitools3.dms.obiseq cimport Nuc_Seq
@@ -153,6 +154,13 @@ def run(config):
     i = 0
     for entry in entries :
         
+        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)
+        
         pb(i)
                         
         if NUC_SEQS_view: 
diff --git a/python/obitools3/parsers/embl.pyx b/python/obitools3/parsers/embl.pyx
index 070c911..84342da 100755
--- a/python/obitools3/parsers/embl.pyx
+++ b/python/obitools3/parsers/embl.pyx
@@ -45,7 +45,7 @@ def emblParser(bytes text):
         
         s      = _seqMatcher.search(text).group()
         s      = _cleanSeq.sub(b'', s).upper()
-        
+
         acs    = _acMatcher.search(text).group()
         acs    = acs.replace(b';', b' ')
         acs    = acs.split()
@@ -55,11 +55,11 @@ def emblParser(bytes text):
         de     = _deMatcher.search(header).group()
         de     = _cleanDe.sub(b' ', de).strip().strip(b'.')
 
-    except AttributeError,e:
-        print('=======================================================')
-        print(text)
-        print('=======================================================')
-        raise e
+    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)