Merge pull request #51 from open-craft/kshitij/fix-empty-search-quotes

Fix missing search excerpts when query has quotes
openedx · Jul 5, 2018 · 4062361 · 4062361
2 parents efe2243 + 629c70b
commit 4062361
Show file tree

Hide file tree

Showing 3 changed files with 54 additions and 62 deletions.
diff --git a/search/result_processor.py b/search/result_processor.py
@@ -4,6 +4,7 @@
 import json
 import logging
 import re
+import shlex
 import textwrap
 
 from django.conf import settings
@@ -128,19 +129,24 @@ def excerpt(self):
         if "content" not in self._results_fields:
             return None
 
-        match_words = [self._match_phrase]
-        separate_words = self._match_phrase.split(' ')
-        if len(separate_words) > 1:
-            match_words.extend(self._match_phrase.split(' '))
+        match_phrases = [self._match_phrase]
+        separate_phrases = [
+            phrase.decode('utf-8')
+            for phrase in shlex.split(self._match_phrase.encode('utf-8'))
+        ]
+        if len(separate_phrases) > 1:
+            match_phrases.extend(separate_phrases)
+        else:
+            match_phrases = separate_phrases
 
         matches = SearchResultProcessor.find_matches(
             SearchResultProcessor.strings_in_dictionary(self._results_fields["content"]),
-            match_words,
+            match_phrases,
             DESIRED_EXCERPT_LENGTH
         )
         excerpt_text = ELLIPSIS.join(matches)
 
-        for match_word in match_words:
+        for match_word in match_phrases:
             excerpt_text = SearchResultProcessor.decorate_matches(excerpt_text, match_word)
 
         return excerpt_text
diff --git a/search/tests/test_search_result_processor.py b/search/tests/test_search_result_processor.py
@@ -1,12 +1,15 @@
 # -*- coding: utf-8 -*-
 """ Tests for result processors """
+import ddt
+
 from django.test import TestCase
 from django.test.utils import override_settings
 from search.result_processor import SearchResultProcessor
 
 
 # Any class that inherits from TestCase will cause too-many-public-methods pylint error
 # pylint: disable=too-many-public-methods
+@ddt.ddt
 class SearchResultProcessorTests(TestCase):
     """ Tests to check SearchResultProcessor is working as desired """
 
@@ -182,33 +185,7 @@ def test_excerpt_front(self):
                     "Dog - match upon first word "
                     "The long and winding road "
                     "That leads to your door "
-                    "Will never disappear "
-                    "I've seen that road before "
-                    "It always leads me here "
-                    "Lead me to you door "
-                    "The wild and windy night "
-                    "That the rain washed away "
-                    "Has left a pool of tears "
-                    "Crying for the day "
-                    "Why leave me standing here "
-                    "Let me know the way "
-                    "Many times I've been alone "
-                    "And many times I've cried "
-                    "Any way you'll never know "
-                    "The many ways I've tried "
-                    "But still they lead me back "
-                    "To the long winding road "
-                    "You left me standing here "
-                    "A long long time ago "
-                    "Don't leave me waiting here "
-                    "Lead me to your door "
-                    "But still they lead me back "
-                    "To the long winding road "
-                    "You left me standing here "
-                    "A long long time ago "
-                    "Don't leave me waiting here "
-                    "Lead me to your door "
-                    "Yeah, yeah, yeah, yeah "
+                    "Will never disappear ..."
                 ),
             }
         }
@@ -229,41 +206,50 @@ def test_excerpt_back(self):
             "content": {
                 "notes": (
                     "The long and winding road "
-                    "That leads to your door "
-                    "Will never disappear "
-                    "I've seen that road before "
-                    "It always leads me here "
-                    "Lead me to you door "
-                    "The wild and windy night "
-                    "That the rain washed away "
-                    "Has left a pool of tears "
-                    "Crying for the day "
-                    "Why leave me standing here "
-                    "Let me know the way "
-                    "Many times I've been alone "
-                    "And many times I've cried "
-                    "Any way you'll never know "
-                    "The many ways I've tried "
-                    "But still they lead me back "
-                    "To the long winding road "
-                    "You left me standing here "
-                    "A long long time ago "
-                    "Don't leave me waiting here "
-                    "Lead me to your door "
-                    "But still they lead me back "
-                    "To the long winding road "
-                    "You left me standing here "
-                    "A long long time ago "
-                    "Don't leave me waiting here "
-                    "Lead me to your door "
-                    "Yeah, yeah, yeah, yeah "
+                    "That leads to your door ..."
+                    "... Yeah, yeah, yeah, yeah "
                     "Match upon last word - Dog"
                 ),
             }
         }
         srp = SearchResultProcessor(test_result, "dog")
         self.assertEqual(srp.excerpt[-33:], "Match upon last word - <b>Dog</b>")
 
+    @ddt.data(
+        (u'"never disappear"', u"leads to your door Will <b>never disappear</b>"),
+        (u'"I\'ve seen"', u"<b>I've seen</b> that road before It always"),
+        (
+            u'"long and winding" leads', 
+            u'The <b>long and winding</b> road That <b>leads</b> to your door'
+        ),
+        (u'"search"', u"इसको <b>search</b> करें| Lead"),
+        (u'"हिंदी में"', u"It always leads me here यह एक <b>हिंदी में</b>"),
+        (u'"इसको search"', u"वाक्य है| <b>इसको search</b> करें| Lead me"),
+        # Match at the beginning
+        (u'"The long"', u'<b>The long</b> and winding road That'),
+        # Match at the end
+        (u'"rain washed away"', u'windy night That the <b>rain washed away</b>'),
+    )
+    @ddt.unpack
+    def test_excerpt_quoted(self, search_phrase, expected_excerpt):
+        test_result = {
+            "content": {
+                "notes": (
+                    u"The long and winding road "
+                    u"That leads to your door "
+                    u"Will never disappear "
+                    u"I've seen that road before "
+                    u"It always leads me here "
+                    u"यह एक हिंदी में लिखा हुआ वाक्य है| इसको search करें| "
+                    u"Lead me to you door "
+                    u"The wild and windy night "
+                    u"That the rain washed away "                    
+                ),
+            }
+        }
+        srp = SearchResultProcessor(test_result, search_phrase)
+        self.assertIn(expected_excerpt, srp.excerpt)
+
 
 class TestSearchResultProcessor(SearchResultProcessor):
     """

diff --git a/setup.py b/setup.py
@@ -5,7 +5,7 @@
 
 setup(
     name='edx-search',
-    version='1.2.0',
+    version='1.2.1',
     description='Search and index routines for index access',
     author='edX',
     author_email='[email protected]',