From 242c80ede3ba60152da7eee21703b48657cb260e Mon Sep 17 00:00:00 2001
From: jeni Shah <jenishah@users.noreply.github.com>
Date: Sat, 20 Oct 2018 18:52:10 +0530
Subject: [PATCH] fix phraser memory

---
 gensim/models/phrases.py                  |   4 ++--
 gensim/test/test_data/phraser_model_3dot6 | Bin 0 -> 543 bytes
 gensim/test/test_phrases.py               |  10 ++++++++++
 3 files changed, 12 insertions(+), 2 deletions(-)
 create mode 100644 gensim/test/test_data/phraser_model_3dot6

diff --git a/gensim/models/phrases.py b/gensim/models/phrases.py
index 9d8a5f5da6..247f5532a9 100644
--- a/gensim/models/phrases.py
+++ b/gensim/models/phrases.py
@@ -805,7 +805,7 @@ def __init__(self, phrases_model):
         for bigram, score in phrases_model.export_phrases(corpus, self.delimiter, as_tuples=True):
             if bigram in self.phrasegrams:
                 logger.info('Phraser repeat %s', bigram)
-            self.phrasegrams[bigram] = (phrases_model.vocab[self.delimiter.join(bigram)], score)
+            self.phrasegrams[bigram] = (None, score)
             count += 1
             if not count % 50000:
                 logger.info('Phraser added %i phrasegrams', count)
@@ -848,7 +848,7 @@ def score_item(self, worda, wordb, components, scorer):
 
         """
         try:
-            return self.phrasegrams[tuple(components)][1]
+            return self.phrasegrams[tuple(components)][-1]
         except KeyError:
             return -1
 
diff --git a/gensim/test/test_data/phraser_model_3dot6 b/gensim/test/test_data/phraser_model_3dot6
new file mode 100644
index 0000000000000000000000000000000000000000..ea6e2643744ebc22c39c2d744eac07203da18b2c
GIT binary patch
literal 543
zcmZ{g%We}f6o!*Dmu6ZZ6aoc$0b1G=XxhRHuw#)8nw2*rYvQ;wmd18|d!|w=RicZ|
z!|)=!7RS?Uy1=qz>pTB{ZofMslg2AO)pj8ZKV9Z7_0mVbn%6~Wo!p^)i_U}%Tdh`^
zJLz*<EU-3fH$*GLg{@49?ut+zrBx{%dV)d=DSRYN_(bTlHf7-$us*#-S}Jpa^%}z%
z5gQQl!a8L#Z2YG$ie}2B1rI_={cW;MvI(oTHJplSk0*ljd8G<U&Utj{><x`5%Lq>s
zJiEsz#&fnoriWQGmy&1e7B41wpLL0sN=v-VsjYF0S1VX~8t8R^+cDnMcsuL-{IUI~
zIW^vqsu!$IN{^i&+l{feqEup;V?WUMF+SAz7|bL5-Jx6K;F0Vw$Uen5A{v%N3%1VV
zIN(W)lNz5NmHtw(b&_#zD!sh+_?qB!MZ8dfOcR{3sL`m5lF5a~d4dZzZkUsza$a3Y
R?$aw-*p$NG5`3@b(|^@5sb&BG

literal 0
HcmV?d00001

diff --git a/gensim/test/test_phrases.py b/gensim/test/test_phrases.py
index e83bf5b2b9..717a8dda25 100644
--- a/gensim/test/test_phrases.py
+++ b/gensim/test/test_phrases.py
@@ -646,6 +646,16 @@ def testEncoding(self):
         self.assertTrue(isinstance(transformed, six.text_type))
 
 
+class TestPhraserModelCompatibilty(unittest.TestCase):
+
+    def testCompatibilty(self):
+        bigram_loaded = Phraser.load(datapath("phraser_model_3dot6"))
+        test_sentences = [u'trees', u'graph', u'minors']
+        prev_ver = bigram_loaded[test_sentences]
+        expected_res = ['trees_graph', 'minors']
+        self.assertEqual(prev_ver, expected_res)
+
+
 if __name__ == '__main__':
     logging.basicConfig(format='%(asctime)s : %(levelname)s : %(message)s', level=logging.DEBUG)
     unittest.main()