search demo
authorMarcin Koziej <marcin.koziej@nowoczesnapolska.org.pl>
Thu, 10 Nov 2011 15:57:56 +0000 (16:57 +0100)
committerMarcin Koziej <marcin.koziej@nowoczesnapolska.org.pl>
Tue, 15 Nov 2011 13:24:23 +0000 (14:24 +0100)
.gitignore
apps/search/index.py
apps/search/views.py
requirements.txt
wolnelektury/templates/newsearch/search.html

index c375441..230a876 100644 (file)
@@ -33,3 +33,5 @@ thumbs.db
 
 # Tags file
 TAGS
+
+media
index 7ec88c6..af3dd9a 100644 (file)
@@ -3,14 +3,15 @@
 from django.conf import settings
 from lucene import SimpleFSDirectory, IndexWriter, File, Field, \
     NumericField, Version, Document, JavaError, IndexSearcher, \
-    QueryParser, Term, PerFieldAnalyzerWrapper, \
+    QueryParser, PerFieldAnalyzerWrapper, \
     SimpleAnalyzer, PolishAnalyzer, ArrayList, \
     KeywordAnalyzer, NumericRangeQuery, BooleanQuery, \
     BlockJoinQuery, BlockJoinCollector, TermsFilter, \
     HashSet, BooleanClause, Term, CharTermAttribute, \
-    PhraseQuery, StringReader, TermQuery, BlockJoinQuery, \
-    Sort, Integer, \
-    initVM, CLASSPATH
+    PhraseQuery, MultiPhraseQuery, StringReader, TermQuery, BlockJoinQuery, \
+    FuzzyQuery, FuzzyTermEnum, Sort, Integer, \
+    SimpleHTMLFormatter, Highlighter, QueryScorer, TokenSources, TextFragment, \
+    initVM, CLASSPATH, JArray
     # KeywordAnalyzer
 JVM = initVM(CLASSPATH)
 import sys
@@ -24,6 +25,7 @@ from threading import current_thread
 import atexit
 import traceback
 
+
 class WLAnalyzer(PerFieldAnalyzerWrapper):
     def __init__(self):
         polish = PolishAnalyzer(Version.LUCENE_34)
@@ -43,7 +45,9 @@ class WLAnalyzer(PerFieldAnalyzerWrapper):
         self.addAnalyzer("author", simple)
         self.addAnalyzer("is_book", keyword)
 
-        #self.addanalyzer("fragment_anchor", keyword)
+        self.addAnalyzer("KEYWORD", keyword)
+        self.addAnalyzer("SIMPLE", simple)
+        self.addAnalyzer("NATURAL", polish)
 
 
 class IndexStore(object):
@@ -81,7 +85,7 @@ class Index(IndexStore):
         self.index = None
 
     def remove_book(self, book):
-        q = NumericRangeQuery.newIntRange("book_id", book.id, book.id, True,True)
+        q = NumericRangeQuery.newIntRange("book_id", book.id, book.id, True, True)
         self.index.deleteDocuments(q)
 
     def index_book(self, book, overwrite=True):
@@ -122,10 +126,10 @@ class Index(IndexStore):
         book_info = dcparser.parse(book.xml_file)
 
         print("extract metadata for book %s id=%d, thread%d" % (book.slug, book.id, current_thread().ident))
-        
+
         doc = self.create_book_doc(book)
         doc.add(Field("slug", book.slug, Field.Store.NO, Field.Index.ANALYZED_NO_NORMS))
-        #doc.add(Field("tags", ','.join([t.name for t in book.tags]), Field.Store.NO, Field.Index.ANALYZED))
+        doc.add(Field("tags", ','.join([t.name for t in book.tags]), Field.Store.NO, Field.Index.ANALYZED))
         doc.add(Field("is_book", 'true', Field.Store.NO, Field.Index.NOT_ANALYZED))
 
         # validator, name
@@ -160,7 +164,6 @@ class Index(IndexStore):
             if master.tag in self.master_tags:
                 return master
 
-    
     def extract_content(self, book):
         wld = WLDocument.from_file(book.xml_file.path)
         root = wld.edoc.getroot()
@@ -172,7 +175,7 @@ class Index(IndexStore):
         master = self.get_master(root)
         if master is None:
             return []
-        
+
         header_docs = []
         for header, position in zip(list(master), range(len(master))):
             if header.tag in self.skip_header_tags:
@@ -224,7 +227,7 @@ class Index(IndexStore):
                               Field.Store.YES, Field.Index.NOT_ANALYZED))
                 doc.add(Field("content",
                               u' '.join(filter(lambda s: s is not None, frag['content'])),
-                              Field.Store.YES, Field.Index.ANALYZED))
+                              Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS))
                 doc.add(Field("themes",
                               u' '.join(filter(lambda s: s is not None, frag['themes'])),
                               Field.Store.NO, Field.Index.ANALYZED))
@@ -249,12 +252,12 @@ class Index(IndexStore):
 
 def log_exception_wrapper(f):
     def _wrap(*a):
-       try:
-           f(*a)
-       except Exception, e:
-           print("Error in indexing thread: %s" % e)
-           traceback.print_exc()
-           raise e
+        try:
+            f(*a)
+        except Exception, e:
+            print("Error in indexing thread: %s" % e)
+            traceback.print_exc()
+            raise e
     return _wrap
 
 
@@ -307,7 +310,7 @@ class ReusableIndex(Index):
 class Search(IndexStore):
     def __init__(self, default_field="content"):
         IndexStore.__init__(self)
-        self.analyzer = PolishAnalyzer(Version.LUCENE_34)
+        self.analyzer = WLAnalyzer() #PolishAnalyzer(Version.LUCENE_34)
         ## self.analyzer = WLAnalyzer()
         self.searcher = IndexSearcher(self.store, True)
         self.parser = QueryParser(Version.LUCENE_34, default_field,
@@ -387,7 +390,7 @@ class Search(IndexStore):
 
 
 class SearchResult(object):
-    def __init__(self, searcher, scoreDocs, score=None):
+    def __init__(self, searcher, scoreDocs, score=None, highlight_query=None):
         if score:
             self.score = score
         else:
@@ -411,6 +414,12 @@ class SearchResult(object):
             self.sections.append(sec)
             self.scores[sec] = scoreDocs.score
 
+        self.snippets = []
+
+    def add_snippets(self, snippets):
+        self.snippets += snippets
+        return self
+
     def get_book(self):
         return catalogue.models.Book.objects.get(id=self.book_id)
 
@@ -432,6 +441,7 @@ class SearchResult(object):
             raise ValueError("this search result is or book %d; tried to merge with %d" % (self.book_id, other.book_id))
         self.fragments += other.fragments
         self.sections += other.sections
+        self.snippets += other.snippets
         self.scores.update(other.scores)
         if other.score > self.score:
             self.score = other.score
@@ -458,38 +468,75 @@ class SearchResult(object):
 
 class MultiSearch(Search):
     """Class capable of IMDb-like searching"""
-    def get_tokens(self, queryreader):
-        if isinstance(queryreader, str) or isinstance(queryreader, unicode):
-            queryreader = StringReader(queryreader)
-        queryreader.reset()
-        tokens = self.analyzer.reusableTokenStream('content', queryreader)
+    def get_tokens(self, searched, field='content'):
+        """returns tokens analyzed by a proper (for a field) analyzer
+        argument can be: StringReader, string/unicode, or tokens. In the last case
+        they will just be returned (so we can reuse tokens, if we don't change the analyzer)
+        """
+        if isinstance(searched, str) or isinstance(searched, unicode):
+            searched = StringReader(searched)
+        elif isinstance(searched, list):
+            return searched
+
+        searched.reset()
+        tokens = self.analyzer.reusableTokenStream(field, searched)
         toks = []
         while tokens.incrementToken():
             cta = tokens.getAttribute(CharTermAttribute.class_)
             toks.append(cta.toString())
         return toks
 
-    def make_phrase(self, tokens, field='content', slop=2):
-        phrase = PhraseQuery()
-        phrase.setSlop(slop)
-        for t in tokens:
-            term = Term(field, t)
-            phrase.add(term)
+    def fuzziness(self, fuzzy):
+        if not fuzzy:
+            return None
+        if isinstance(fuzzy, float) and fuzzy > 0.0 and fuzzy <= 1.0:
+            return fuzzy
+        else:
+            return 0.5
+
+    def make_phrase(self, tokens, field='content', slop=2, fuzzy=False):
+        if fuzzy:
+            phrase = MultiPhraseQuery()
+            for t in tokens:
+                term = Term(field, t)
+                fuzzterm = FuzzyTermEnum(self.searcher.getIndexReader(), term, self.fuzziness(fuzzy))
+                fuzzterms = []
+
+                while True:
+                    #                    print("fuzz %s" % unicode(fuzzterm.term()).encode('utf-8'))
+                    ft = fuzzterm.term()
+                    if ft:
+                        fuzzterms.append(ft)
+                    if not fuzzterm.next(): break
+                if fuzzterms:
+                    phrase.add(JArray('object')(fuzzterms, Term))
+                else:
+                    phrase.add(term)
+        else:
+            phrase = PhraseQuery()
+            phrase.setSlop(slop)
+            for t in tokens:
+                term = Term(field, t)
+                phrase.add(term)
         return phrase
 
-    def make_term_query(self, tokens, field='content', modal=BooleanClause.Occur.SHOULD):
+    def make_term_query(self, tokens, field='content', modal=BooleanClause.Occur.SHOULD, fuzzy=False):
         q = BooleanQuery()
         for t in tokens:
             term = Term(field, t)
-            q.add(BooleanClause(TermQuery(term), modal))
+            if fuzzy:
+                term = FuzzyQuery(term, self.fuzziness(fuzzy))
+            else:
+                term = TermQuery(term)
+            q.add(BooleanClause(term, modal))
         return q
 
     def content_query(self, query):
         return BlockJoinQuery(query, self.parent_filter,
                               BlockJoinQuery.ScoreMode.Total)
 
-    def search_perfect_book(self, tokens, max_results=20):
-        qrys = [self.make_phrase(tokens, field=fld) for fld in ['author', 'title']]
+    def search_perfect_book(self, searched, max_results=20, fuzzy=False):
+        qrys = [self.make_phrase(self.get_tokens(searched, field=fld), field=fld, fuzzy=fuzzy) for fld in ['author', 'title']]
 
         books = []
         for q in qrys:
@@ -498,38 +545,52 @@ class MultiSearch(Search):
                 books.append(SearchResult(self.searcher, found))
         return books
 
-    def search_perfect_parts(self, tokens, max_results=20):
-        qrys = [self.make_phrase(tokens, field=fld) for fld in ['content']]
+    def search_perfect_parts(self, searched, max_results=20, fuzzy=False):
+        qrys = [self.make_phrase(self.get_tokens(searched), field=fld, fuzzy=fuzzy) for fld in ['content']]
 
         books = []
         for q in qrys:
             top = self.searcher.search(q, max_results)
             for found in top.scoreDocs:
-                books.append(SearchResult(self.searcher, found))
+                books.append(SearchResult(self.searcher, found).add_snippets(self.get_snippets(found, q)))
 
         return books
 
-    def search_everywhere(self, tokens, max_results=20):
+    def search_everywhere(self, searched, max_results=20, fuzzy=False):
+        books = []
+
+        # content only query : themes x content
+        q = BooleanQuery()
+
+        tokens = self.get_tokens(searched)
+        q.add(BooleanClause(self.make_term_query(tokens, field='themes', fuzzy=fuzzy), BooleanClause.Occur.MUST))
+        q.add(BooleanClause(self.make_term_query(tokens, field='content', fuzzy=fuzzy), BooleanClause.Occur.SHOULD))
+
+        topDocs = self.searcher.search(q, max_results)
+        for found in topDocs.scoreDocs:
+            books.append(SearchResult(self.searcher, found))
+
+        # joined query themes/content x author/title/epochs/genres/kinds
         q = BooleanQuery()
         in_meta = BooleanQuery()
         in_content = BooleanQuery()
 
         for fld in ['themes', 'content']:
-            in_content.add(BooleanClause(self.make_term_query(tokens, field=fld), BooleanClause.Occur.SHOULD))
+            in_content.add(BooleanClause(self.make_term_query(tokens, field=fld, fuzzy=False), BooleanClause.Occur.SHOULD))
+
+        in_meta.add(BooleanClause(self.make_term_query(self.get_tokens(searched, field='author'), field='author', fuzzy=fuzzy), BooleanClause.Occur.SHOULD))
 
-        for fld in ['author', 'title', 'epochs', 'genres', 'kinds']:
-            in_meta.add(BooleanClause(self.make_term_query(tokens, field=fld), BooleanClause.Occur.SHOULD))
+        for fld in ['title', 'epochs', 'genres', 'kinds']:
+            in_meta.add(BooleanClause(self.make_term_query(tokens, field=fld, fuzzy=fuzzy), BooleanClause.Occur.SHOULD))
 
         q.add(BooleanClause(in_meta, BooleanClause.Occur.MUST))
         in_content_join = self.content_query(in_content)
         q.add(BooleanClause(in_content_join, BooleanClause.Occur.MUST))
-
+        #        import pdb; pdb.set_trace()
         collector = BlockJoinCollector(Sort.RELEVANCE, 100, True, True)
 
         self.searcher.search(q, collector)
 
-        books = []
-
         top_groups = collector.getTopGroups(in_content_join, Sort.RELEVANCE, 0, max_results, 0, True)
         if top_groups:
             for grp in top_groups.groups:
@@ -579,7 +640,7 @@ class MultiSearch(Search):
 
         return None
 
-    
+
     def do_search(self, query, max_results=50, collector=None):
         tops = self.searcher.search(query, max_results)
         #tops = self.searcher.search(p_content, max_results)
@@ -591,3 +652,16 @@ class MultiSearch(Search):
             bks.append(b)
             print "%s (%d) -> %f" % (b, b.id, found.score)
         return (bks, tops.totalHits)
+
+    def get_snippets(self, scoreDoc, query, field='content'):
+        htmlFormatter = SimpleHTMLFormatter()
+        highlighter = Highlighter(htmlFormatter, QueryScorer(query))
+
+        stored = self.searcher.doc(scoreDoc.doc)
+        text = stored.get(field)
+        tokenStream = TokenSources.getAnyTokenStream(self.searcher.getIndexReader(), scoreDoc.doc, field, self.analyzer)
+        #  highlighter.getBestTextFragments(tokenStream, text, False, 10)
+        snip = highlighter.getBestFragments(tokenStream, text, 3, "...")
+        print('snips: %s' % snip)
+
+        return [snip]
index fc14b06..d9b2f26 100644 (file)
@@ -1,5 +1,4 @@
 
-
 from django.shortcuts import render_to_response, get_object_or_404
 from django.template import RequestContext
 from django.contrib.auth.decorators import login_required
@@ -9,6 +8,31 @@ from catalogue.utils import get_random_hash
 from catalogue.models import Book, Tag
 from catalogue import forms
 from search import MultiSearch, JVM, SearchResult
+from lucene import StringReader
+
+import enchant
+
+dictionary = enchant.Dict('pl_PL')
+
+
+def did_you_mean(query, tokens):
+    change = {}
+    
+    for t in tokens:
+        print("%s ok? %s, sug: %s" %(t, dictionary.check(t), dictionary.suggest(t)))
+        if not dictionary.check(t):
+            try:
+                change[t] = dictionary.suggest(t)[0]
+            except IndexError:
+                pass
+
+    if change == {}:
+        return None
+
+    for frm, to in change.items():
+        query = query.replace(frm, to)
+        
+    return query
 
 
 def main(request):
@@ -17,15 +41,26 @@ def main(request):
     srch = MultiSearch()
 
     results = None
+    query = None
+    fuzzy = False
     if 'q' in request.GET:
-        toks = srch.get_tokens(request.GET['q'])
-        results = SearchResult.aggregate(srch.search_perfect_book(toks),
-                                         srch.search_perfect_parts(toks),
-                                         srch.search_everywhere(toks))
+        query = request.GET['q']
+        toks = StringReader(query)
+        fuzzy = 'fuzzy' in request.GET
+        if fuzzy:
+            fuzzy = 0.7
+
+
+        results = SearchResult.aggregate(srch.search_perfect_book(toks, fuzzy=fuzzy),
+                                         srch.search_perfect_parts(toks, fuzzy=fuzzy),
+                                         srch.search_everywhere(toks, fuzzy=fuzzy))
         results.sort(reverse=True)
 
         for r in results:
             print r.parts
 
-    return render_to_response('newsearch/search.html', {"results": results},
+    return render_to_response('newsearch/search.html', {'results': results,
+                                                        'did_you_mean': (query is not None) and 
+                                                        did_you_mean(query, srch.get_tokens(query, field='SIMPLE')),
+                                                        'fuzzy': fuzzy},
                               context_instance=RequestContext(request))
index 4fd263e..dfb01f1 100644 (file)
@@ -28,3 +28,6 @@ lxml>=2.2.2
 # celery tasks
 django-celery
 django-kombu
+
+# spell checking
+pyenchant
index 4dfc858..af36b59 100644 (file)
     <h1>Search</h1>
     <form action="{% url newsearch %}" method="get" accept-charset="utf-8" id="search-form-x">
         <p>
-         <input type="text" name="q" style="width:250px; font-size: 1.2em;">
+         <input type="text" name="q" value="{{request.GET.q}}" style="width:250px; font-size: 1.2em;">
          <input type="submit" value="{% trans "Search" %}" /> 
+         <br />
+         <input type="checkbox" value="true" name="fuzzy" {% if fuzzy %}checked{% endif %}/> fuzzy.
        </p>
     </form>
+    {% if did_you_mean %}
+    Czy miałeś na mysli <a href="?q={{did_you_mean|urlencode}}">{{did_you_mean}}</a>?
+    {% endif %}
+
 
     <div id="results">
       <ol>
       {% for result in results %}
       <li>
-       <p>{{result.book.pretty_title}} (id: {{result.book.id}}, score: {{result.score}})</p>
+       <p><a href="{{result.book.get_absolute_url}}">{{result.book.pretty_title}}</a> (id: {{result.book.id}}, score: {{result.score}})</p>
        <ul>
+         {% for snippet in result.snippets %}
+         <li>{{snippet|safe}}</li>
+         {% endfor %}
+
          {% for part in result.parts %}
          {% if part.header %}
          <li>W {{part.header}} nr {{part.position}}</li>
          {% else %} 
          {% if part.fragment %}
          <li>
-           <div style="font-weight: bold;">{% for tag in part.fragment.tags %}{{tag.name}} {% endfor %}</div>
-           {{part.fragment.text|safe}} </li>
+           <div style="">Tagi/Motywy: {% for tag in part.fragment.tags %}{{tag.name}} {% endfor %}</div>
+           {{part.fragment.short_html|safe}}
+         </li>
          {% endif %}
          {% endif %}
          {% endfor %}