Use a csv.Sniffer to detect the delimiter of LCSV files

Thus making the "delimiter" parameter of LCSV2RDF and lcsv_extentities()
optional.

lcsv.py

@@ -14,6 +14,7 @@
 # You should have received a copy of the GNU Lesser General Public License along
 # with this program. If not, see <http://www.gnu.org/licenses/>.
 
+import csv
 from itertools import count
 
 from six import text_type
@@ -38,9 +39,13 @@ class LCSV2RDF(object):
     reference : W3C unofficial draft, http://jenit.github.io/linked-csv/
     """
 
-    def __init__(self, stream, delimiter, encoding='utf-8',
+    def __init__(self, stream, delimiter=None, encoding='utf-8',
                  uri_generator=None, uri_cls=text_type, default_lang=None):
         """ check stream validity and init attribute"""
+        if delimiter is None:
+            sample = stream.read(1024)
+            stream.seek(0)
+            delimiter = csv.Sniffer().sniff(sample).delimiter
         self.source_file = ucsvreader(stream, encoding=encoding, delimiter=delimiter)
         # raise an error value if the line doesn't contains a #
         if uri_generator is None:

sobjects.py

@@ -156,7 +156,7 @@ class SKOSParser(datafeed.DataFeedParser):
                                        source=self.source, raise_on_error=raise_on_error, **kwargs)
 
 
-def lcsv_extentities(stream, scheme_uri, delimiter, encoding='utf-8', language_code=None):
+def lcsv_extentities(stream, scheme_uri, delimiter=None, encoding='utf-8', language_code=None):
     """Return external entities generator from SKOS LCSV stream or URL (by transforming it to RDF
     first).
 
@@ -171,8 +171,9 @@ def lcsv_extentities(stream, scheme_uri, delimiter, encoding='utf-8', language_c
     """
     graph = rdfio.RDFLibRDFGraph()
     # add LCSV statements to the RDF graph
-    lcsv2rdf = lcsv.LCSV2RDF(stream, delimiter, encoding=encoding, default_lang=language_code,
-                             uri_cls=graph.uri, uri_generator=lambda x: str(uuid4()) + x)
+    lcsv2rdf = lcsv.LCSV2RDF(stream, delimiter=delimiter, encoding=encoding,
+                             default_lang=language_code, uri_cls=graph.uri,
+                             uri_generator=lambda x: str(uuid4()) + x)
     for (subj, pred, obj) in lcsv2rdf.triples():
         graph.add(subj, pred, obj)
 

test/unittest_lcsv.py

@@ -81,6 +81,12 @@ class LCSV2RDFTC(TestCase):
                               '#4')
                          ]))
 
+    def test_lcsv_parsing_sniff(self):
+        fpath = self.datapath('lcsv_example_shortened.csv')
+        lcsv2rdf = lcsv.LCSV2RDF(open(fpath),
+                                 uri_generator=lambda x: x, default_lang='es')
+        self.assertEqual(len(list(lcsv2rdf.triples())), 18)
+
 
 if __name__ == "__main__":
     from unittest import main