QueryParser refactoring step 1: add the lexicon to the constructor args.

lib/python/Products/ZCTextIndex/QueryParser.py

@@ -92,8 +92,8 @@ _tokenizer_regex = re.compile(r"""
 
 class QueryParser:
 
-    def __init__(self):
-        pass # This parser has no persistent state
+    def __init__(self, lexicon):
+        self._lexicon = lexicon
 
     def parseQuery(self, query):
         # Lexical analysis.

lib/python/Products/ZCTextIndex/ZCTextIndex.py

@@ -39,14 +39,18 @@ class ZCTextIndex(Persistent, Acquisition.Implicit, SimpleItem):
 
     __implements__ = PluggableIndexInterface
 
+    ## Magic class attributes ##
+
     meta_type = 'ZCTextIndex'
 
-    manage_options= (
+    manage_options = (
         {'label': 'Settings', 'action': 'manage_main'},
     )
 
     query_options = ['query']
 
+    ## Constructor ##
+
     def __init__(self, id, extra, caller, index_factory=OkapiIndex):
         self.id = id
         self._fieldname = extra.doc_attr
@@ -64,13 +68,15 @@ class ZCTextIndex(Persistent, Acquisition.Implicit, SimpleItem):
         self._index_factory = index_factory
         self.clear()
 
+    ## External methods not in the Pluggable Index API ##
+
     def query(self, query, nbest=10):
         """Return pair (mapping from docids to scores, num results).
 
         The num results is the total number of results before trimming
         to the nbest results.
         """
-        tree = QueryParser().parseQuery(query)
+        tree = QueryParser(self.lexicon).parseQuery(query)
         results = tree.executeQuery(self.index)
         if results is None:
             return [], 0
@@ -107,7 +113,7 @@ class ZCTextIndex(Persistent, Acquisition.Implicit, SimpleItem):
         if record.keys is None:
             return None
         query_str = ' '.join(record.keys)
-        tree = QueryParser().parseQuery(query_str)
+        tree = QueryParser(self.lexicon).parseQuery(query_str)
         results = tree.executeQuery(self.index)
         return  results, (self._fieldname,)
 
@@ -120,10 +126,14 @@ class ZCTextIndex(Persistent, Acquisition.Implicit, SimpleItem):
         get_word = self.lexicon.get_word
         return [get_word(wid) for wid in word_ids]
 
+    ## XXX To which API does this conform? ##
+
     def clear(self):
         """reinitialize the index"""
         self.index = self._index_factory(self.lexicon)
 
+    ## Helper ##
+
     def _get_object_text(self, obj):
         x = getattr(obj, self._fieldname)
         if callable(x):

lib/python/Products/ZCTextIndex/tests/mhindex.py

@@ -394,7 +394,7 @@ class TextIndex(Persistent):
 
     def query(self, query, nbest=10):
         # returns a total hit count and a mapping from docids to scores
-        parser = QueryParser()
+        parser = QueryParser(self.lexicon)
         tree = parser.parseQuery(query)
         results = tree.executeQuery(self.index)
         if results is None:
@@ -404,7 +404,7 @@ class TextIndex(Persistent):
         return chooser.getbest(), len(results)
 
     def query_weight(self, query):
-        parser = QueryParser()
+        parser = QueryParser(self.lexicon)
         tree = parser.parseQuery(query)
         terms = tree.terms()
         return self.index.query_weight(terms)

lib/python/Products/ZCTextIndex/tests/testQueryEngine.py

@@ -18,6 +18,7 @@ from BTrees.IIBTree import IIBucket
 
 from Products.ZCTextIndex.QueryParser import QueryParser
 from Products.ZCTextIndex.ParseTree import ParseError, QueryError
+from Products.ZCTextIndex.Lexicon import Lexicon, Splitter
 
 class FauxIndex:
 
@@ -34,7 +35,8 @@ class FauxIndex:
 class TestQueryEngine(TestCase):
 
     def setUp(self):
-        self.parser = QueryParser()
+        self.lexicon = Lexicon(Splitter())
+        self.parser = QueryParser(self.lexicon)
         self.index = FauxIndex()
 
     def compareSet(self, set, dict):

lib/python/Products/ZCTextIndex/tests/testQueryParser.py

@@ -19,6 +19,7 @@ from Products.ZCTextIndex.QueryParser import QueryParser
 from Products.ZCTextIndex.ParseTree import ParseError, ParseTreeNode
 from Products.ZCTextIndex.ParseTree import OrNode, AndNode, NotNode
 from Products.ZCTextIndex.ParseTree import AtomNode, PhraseNode, GlobNode
+from Products.ZCTextIndex.Lexicon import Lexicon, Splitter
 
 class TestQueryParser(TestCase):
 
@@ -54,7 +55,8 @@ class TestQueryParser(TestCase):
         self.assertRaises(ParseError, self.p.parseQuery, input)
 
     def setUp(self):
-        self.p = QueryParser()
+        self.lexicon = Lexicon(Splitter())
+        self.p = QueryParser(self.lexicon)
 
     def testParseQuery(self):
         self.expect("foo", AtomNode("foo"))

lib/python/Products/ZCTextIndex/tests/testZCTextIndex.py

@@ -265,7 +265,7 @@ class CosineIndexTests(ZCIndexTestsBase, testIndex.CosineIndexTest):
                    [(1, 0.19), (2, 0.18), (3, 0.63), (5, 0.22), (6, 0.39)]]
         for i in range(len(queries)):
             raw = queries[i]
-            q = QueryParser().parseQuery(raw)
+            q = QueryParser(self.lexicon).parseQuery(raw)
             wq = self.index.query_weight(q.terms())
             eq(wq, scaled_int(wqs[i]))
             r, n = self.zc_index.query(raw)
@@ -428,10 +428,11 @@ class QueryTestsBase(testQueryEngine.TestQueryEngine,
         extra = Extra()
         extra.doc_attr = 'text'
         extra.lexicon_id = 'lexicon'
-        caller = LexiconHolder(Lexicon(Splitter(), CaseNormalizer(),
-                               StopWordRemover()))
+        self.lexicon = Lexicon(Splitter(), CaseNormalizer(),
+                               StopWordRemover())
+        caller = LexiconHolder(self.lexicon)
         self.zc_index = ZCTextIndex('name', extra, caller, self.IndexFactory)
-        self.p = self.parser = QueryParser()
+        self.p = self.parser = QueryParser(self.lexicon)
         self.index = self.zc_index.index
         self.add_docs()