Merge pull request #171 from InsomniacSabbir/master

Added sanskrit corpora
cltk · Mar 5, 2016 · b73a581 · b73a581
2 parents d057ca5 + 399a653
commit b73a581
Show file tree

Hide file tree

Showing 3 changed files with 9 additions and 2 deletions.
diff --git a/cltk/corpus/sanskrit/__init__.py b/cltk/corpus/sanskrit/__init__.py
diff --git a/cltk/corpus/sanskrit/corpora.py b/cltk/corpus/sanskrit/corpora.py
@@ -0,0 +1,5 @@
+SANSKRIT_CORPORA = [
+        {'name':'sanskrit_text_jnu',
+        'location':'remote',
+        'type':'text'},
+        ]
diff --git a/cltk/corpus/utils/importer.py b/cltk/corpus/utils/importer.py
@@ -11,6 +11,7 @@
 from cltk.corpus.coptic.corpora import COPTIC_CORPORA
 from cltk.corpus.greek.corpora import GREEK_CORPORA
 from cltk.corpus.latin.corpora import LATIN_CORPORA
+from cltk.corpus.sanskrit.corpora import SANSKRIT_CORPORA
 from cltk.corpus.multilingual.corpora import MULTILINGUAL_CORPORA
 from cltk.corpus.pali.corpora import PALI_CORPORA
 from cltk.corpus.tibetan.corpora import TIBETAN_CORPORA
@@ -22,15 +23,16 @@
 from urllib.parse import urljoin
 
 
-AVAILABLE_LANGUAGES = ['chinese', 'coptic', 'greek', 'latin', 'multilingual', 'pali', 'tibetan']
+AVAILABLE_LANGUAGES = ['chinese', 'coptic', 'greek', 'latin', 'multilingual', 'pali', 'tibetan','sanskrit']
 CLTK_DATA_DIR = '~/cltk_data'
 LANGUAGE_CORPORA = {'chinese': CHINESE_CORPORA,
                     'coptic': COPTIC_CORPORA,
                     'greek': GREEK_CORPORA,
                     'latin': LATIN_CORPORA,
                     'multilingual': MULTILINGUAL_CORPORA,
                     'pali': PALI_CORPORA,
-                    'tibetan': TIBETAN_CORPORA,}
+                    'tibetan': TIBETAN_CORPORA,
+                    'sanskrit': SANSKRIT_CORPORA,}