Merge branch 'production' into pretty
[wolnelektury.git] / apps / search / index.py
index f5489a5..cc478ee 100644 (file)
@@ -1,29 +1,47 @@
 # -*- coding: utf-8 -*-
+
 from django.conf import settings
-from lucene import SimpleFSDirectory, IndexWriter, File, Field, \
+from lucene import SimpleFSDirectory, IndexWriter, CheckIndex, \
+    File, Field, Integer, \
     NumericField, Version, Document, JavaError, IndexSearcher, \
-    QueryParser, Term, PerFieldAnalyzerWrapper, \
+    QueryParser, PerFieldAnalyzerWrapper, \
     SimpleAnalyzer, PolishAnalyzer, ArrayList, \
-    KeywordAnalyzer, NumericRangeQuery, BooleanQuery, \
-    BlockJoinQuery, BlockJoinCollector, TermsFilter, \
+    KeywordAnalyzer, NumericRangeQuery, NumericRangeFilter, BooleanQuery, \
+    BlockJoinQuery, BlockJoinCollector, Filter, TermsFilter, ChainedFilter, \
     HashSet, BooleanClause, Term, CharTermAttribute, \
-    PhraseQuery, StringReader
+    PhraseQuery, MultiPhraseQuery, StringReader, TermQuery, \
+    FuzzyQuery, FuzzyTermEnum, PrefixTermEnum, Sort, Integer, \
+    SimpleHTMLFormatter, Highlighter, QueryScorer, TokenSources, TextFragment, \
+    BooleanFilter, TermsFilter, FilterClause, QueryWrapperFilter, \
+    initVM, CLASSPATH, JArray, JavaError
     # KeywordAnalyzer
+
+# Initialize jvm
+JVM = initVM(CLASSPATH)
+
 import sys
 import os
+import re
 import errno
 from librarian import dcparser
 from librarian.parser import WLDocument
 import catalogue.models
 from multiprocessing.pool import ThreadPool
+from threading import current_thread
 import atexit
+import traceback
 
 
 class WLAnalyzer(PerFieldAnalyzerWrapper):
     def __init__(self):
         polish = PolishAnalyzer(Version.LUCENE_34)
+        #        polish_gap.setPositionIncrementGap(999)
+
         simple = SimpleAnalyzer(Version.LUCENE_34)
+        #        simple_gap.setPositionIncrementGap(999)
+
         keyword = KeywordAnalyzer(Version.LUCENE_34)
+
         # not sure if needed: there's NOT_ANALYZED meaning basically the same
 
         PerFieldAnalyzerWrapper.__init__(self, polish)
@@ -35,13 +53,31 @@ class WLAnalyzer(PerFieldAnalyzerWrapper):
         self.addAnalyzer("source_url", keyword)
         self.addAnalyzer("source_name", simple)
         self.addAnalyzer("publisher", simple)
-        self.addAnalyzer("author", simple)
+        self.addAnalyzer("authors", simple)
+        self.addAnalyzer("title", simple)
+
         self.addAnalyzer("is_book", keyword)
+        # shouldn't the title have two forms? _pl and simple?
+
+        self.addAnalyzer("themes", simple)
+        self.addAnalyzer("themes_pl", polish)
+
+        self.addAnalyzer("tag_name", simple)
+        self.addAnalyzer("tag_name_pl", polish)
+
+        self.addAnalyzer("translators", simple)
 
-        #self.addanalyzer("fragment_anchor", keyword)
+        self.addAnalyzer("KEYWORD", keyword)
+        self.addAnalyzer("SIMPLE", simple)
+        self.addAnalyzer("POLISH", polish)
 
 
 class IndexStore(object):
+    """
+    Provides access to search index.
+
+    self.store - lucene index directory
+    """
     def __init__(self):
         self.make_index_dir()
         self.store = SimpleFSDirectory(File(settings.SEARCH_INDEX))
@@ -55,9 +91,77 @@ class IndexStore(object):
             else: raise
 
 
-class Index(IndexStore):
-    def __init__(self, analyzer=None):
+class IndexChecker(IndexStore):
+    def __init__(self):
         IndexStore.__init__(self)
+
+    def check(self):
+        checker = CheckIndex(self.store)
+        status = checker.checkIndex()
+        return status
+
+
+class Snippets(object):
+    """
+    This class manages snippet files for indexed object (book)
+    the snippets are concatenated together, and their positions and
+    lengths are kept in lucene index fields.
+    """
+    SNIPPET_DIR = "snippets"
+
+    def __init__(self, book_id):
+        try:
+            os.makedirs(os.path.join(settings.SEARCH_INDEX, self.SNIPPET_DIR))
+        except OSError as exc:
+            if exc.errno == errno.EEXIST:
+                pass
+            else: raise
+        self.book_id = book_id
+        self.file = None
+
+    def open(self, mode='r'):
+        """
+        Open the snippet file. Call .close() afterwards.
+        """
+        if not 'b' in mode:
+            mode += 'b'
+        self.file = open(os.path.join(settings.SEARCH_INDEX, self.SNIPPET_DIR, str(self.book_id)), mode)
+        self.position = 0
+        return self
+
+    def add(self, snippet):
+        """
+        Append a snippet (unicode) to the snippet file.
+        Return a (position, length) tuple
+        """
+        txt = snippet.encode('utf-8')
+        l = len(txt)
+        self.file.write(txt)
+        pos = (self.position, l)
+        self.position += l
+        return pos
+
+    def get(self, pos):
+        """
+        Given a tuple of (position, length) return an unicode
+        of the snippet stored there.
+        """
+        self.file.seek(pos[0], 0)
+        txt = self.file.read(pos[1]).decode('utf-8')
+        return txt
+
+    def close(self):
+        """Close snippet file"""
+        self.file.close()
+
+
+class BaseIndex(IndexStore):
+    """
+    Base index class.
+    Provides basic operations on index: opening, closing, optimizing.
+    """
+    def __init__(self, analyzer=None):
+        super(BaseIndex, self).__init__()
         self.index = None
         if not analyzer:
             analyzer = WLAnalyzer()
@@ -70,27 +174,88 @@ class Index(IndexStore):
                                  IndexWriter.MaxFieldLength.LIMITED)
         return self.index
 
-    def close(self):
+    def optimize(self):
         self.index.optimize()
+
+    def close(self):
+        try:
+            self.index.optimize()
+        except JavaError, je:
+            print "Error during optimize phase, check index: %s" % je
+
         self.index.close()
         self.index = None
 
+    def __enter__(self):
+        self.open()
+        return self
+
+    def __exit__(self, type, value, tb):
+        self.close()
+
+
+class Index(BaseIndex):
+    """
+    Class indexing books.
+    """
+    def __init__(self, analyzer=None):
+        super(Index, self).__init__(analyzer)
+
+    def index_tags(self):
+        """
+        Re-index global tag list.
+        Removes all tags from index, then index them again.
+        Indexed fields include: id, name (with and without polish stems), category
+        """
+        q = NumericRangeQuery.newIntRange("tag_id", 0, Integer.MAX_VALUE, True, True)
+        self.index.deleteDocuments(q)
+
+        for tag in catalogue.models.Tag.objects.all():
+            doc = Document()
+            doc.add(NumericField("tag_id", Field.Store.YES, True).setIntValue(tag.id))
+            doc.add(Field("tag_name", tag.name, Field.Store.NO, Field.Index.ANALYZED))
+            doc.add(Field("tag_name_pl", tag.name, Field.Store.NO, Field.Index.ANALYZED))
+            doc.add(Field("tag_category", tag.category, Field.Store.NO, Field.Index.NOT_ANALYZED))
+            self.index.addDocument(doc)
+
+    def create_book_doc(self, book):
+        """
+        Create a lucene document referring book id.
+        """
+        doc = Document()
+        doc.add(NumericField("book_id", Field.Store.YES, True).setIntValue(book.id))
+        if book.parent is not None:
+            doc.add(NumericField("parent_id", Field.Store.YES, True).setIntValue(book.parent.id))
+        return doc
+
     def remove_book(self, book):
-        q = NumericRangeQuery.newIntRange("book_id", book.id, book.id, True,True)
+        """Removes a book from search index.
+        book - Book instance."""
+        q = NumericRangeQuery.newIntRange("book_id", book.id, book.id, True, True)
         self.index.deleteDocuments(q)
 
-    def index_book(self, book, overwrite=True):
+    def index_book(self, book, book_info=None, overwrite=True):
+        """
+        Indexes the book.
+        Creates a lucene document for extracted metadata
+        and calls self.index_content() to index the contents of the book.
+        """
         if overwrite:
             self.remove_book(book)
 
-        doc = self.extract_metadata(book)
-        parts = self.extract_content(book)
-        block = ArrayList().of_(Document)
+        book_doc = self.create_book_doc(book)
+        meta_fields = self.extract_metadata(book, book_info)
+        for f in meta_fields.values():
+            if isinstance(f, list) or isinstance(f, tuple):
+                for elem in f:
+                    book_doc.add(elem)
+            else:
+                book_doc.add(f)
 
-        for p in parts:
-            block.add(p)
-        block.add(doc)
-        self.index.addDocuments(block)
+        self.index.addDocument(book_doc)
+        del book_doc
+
+        self.index_content(book, book_fields=[meta_fields['title'], meta_fields['authors']])
 
     master_tags = [
         'opowiadanie',
@@ -101,25 +266,20 @@ class Index(IndexStore):
         'wywiad'
         ]
 
-    skip_header_tags = ['autor_utworu', 'nazwa_utworu']
+    skip_header_tags = ['autor_utworu', 'nazwa_utworu', 'dzielo_nadrzedne']
 
-    def create_book_doc(self, book):
+    def extract_metadata(self, book, book_info=None):
         """
-        Create a lucene document connected to the book
+        Extract metadata from book and returns a map of fields keyed by fieldname
         """
-        doc = Document()
-        doc.add(NumericField("book_id", Field.Store.YES, True).setIntValue(book.id))
-        if book.parent is not None:
-            doc.add(NumericField("parent_id", Field.Store.YES, True).setIntValue(book.parent.id))
-        return doc
+        fields = {}
 
-    def extract_metadata(self, book):
-        book_info = dcparser.parse(book.xml_file)
+        if book_info is None:
+            book_info = dcparser.parse(open(book.xml_file.path))
 
-        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("is_book", 'true', Field.Store.NO, Field.Index.NOT_ANALYZED))
+        fields['slug'] = Field("slug", book.slug, Field.Store.NO, Field.Index.ANALYZED_NO_NORMS)
+        fields['tags'] = self.add_gaps([Field("tags", t.name, Field.Store.NO, Field.Index.ANALYZED) for t in book.tags], 'tags')
+        fields['is_book'] = Field("is_book", 'true', Field.Store.NO, Field.Index.NOT_ANALYZED)
 
         # validator, name
         for field in dcparser.BookInfo.FIELDS:
@@ -133,7 +293,7 @@ class Index(IndexStore):
                     if field.multiple:
                         s = ', '.join(s)
                     try:
-                        doc.add(Field(field.name, s, Field.Store.NO, Field.Index.ANALYZED))
+                        fields[field.name] = Field(field.name, s, Field.Store.NO, Field.Index.ANALYZED)
                     except JavaError as je:
                         raise Exception("failed to add field: %s = '%s', %s(%s)" % (field.name, s, je.message, je.args))
                 elif type_indicator == dcparser.as_person:
@@ -142,40 +302,43 @@ class Index(IndexStore):
                         persons = unicode(p)
                     else:
                         persons = ', '.join(map(unicode, p))
-                    doc.add(Field(field.name, persons, Field.Store.NO, Field.Index.ANALYZED))
+                    fields[field.name] = Field(field.name, persons, Field.Store.NO, Field.Index.ANALYZED)
                 elif type_indicator == dcparser.as_date:
                     dt = getattr(book_info, field.name)
-                    doc.add(Field(field.name, "%04d%02d%02d" % (dt.year, dt.month, dt.day), Field.Store.NO, Field.Index.NOT_ANALYZED))
-        return doc
+                    fields[field.name] = Field(field.name, "%04d%02d%02d" %\
+                                               (dt.year, dt.month, dt.day), Field.Store.NO, Field.Index.NOT_ANALYZED)
+
+        return fields
+
+    def add_gaps(self, fields, fieldname):
+        """
+        Interposes a list of fields with gap-fields, which are indexed spaces and returns it.
+        This allows for doing phrase queries which do not overlap the gaps (when slop is 0).
+        """
+        def gap():
+            while True:
+                yield Field(fieldname, ' ', Field.Store.NO, Field.Index.NOT_ANALYZED)
+        return reduce(lambda a, b: a + b, zip(fields, gap()))[0:-1]
 
     def get_master(self, root):
+        """
+        Returns the first master tag from an etree.
+        """
         for master in root.iter():
             if master.tag in self.master_tags:
                 return master
 
-    
-    def extract_content(self, book):
-        wld = WLDocument.from_file(book.xml_file.path)
+    def index_content(self, book, book_fields=[]):
+        """
+        Walks the book XML and extract content from it.
+        Adds parts for each header tag and for each fragment.
+        """
+        wld = WLDocument.from_file(book.xml_file.path, parse_dublincore=False)
         root = wld.edoc.getroot()
 
-        # first we build a sequence of top-level items.
-        # book_id
-        # header_index - the 0-indexed position of header element.
-        # content
         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:
-                continue
-            doc = self.create_book_doc(book)
-            doc.add(NumericField("header_index", Field.Store.YES, True).setIntValue(position))
-            doc.add(Field("header_type", header.tag, Field.Store.YES, Field.Index.NOT_ANALYZED))
-            content = u' '.join([t for t in header.itertext()])
-            doc.add(Field("content", content, Field.Store.NO, Field.Index.ANALYZED))
-            header_docs.append(doc)
 
         def walker(node):
             yield node, None
@@ -185,59 +348,119 @@ class Index(IndexStore):
             yield None, node
             return
 
-        # Then we create a document for each fragments
-        # fragment_anchor - the anchor
-        # themes - list of themes [not indexed]
-        fragment_docs = []
-        # will contain (framgent id -> { content: [], themes: [] }
-        fragments = {}
-        for start, end in walker(master):
-            if start is not None and start.tag == 'begin':
-                fid = start.attrib['id'][1:]
-                fragments[fid] = {'content': [], 'themes': []}
-                fragments[fid]['content'].append(start.tail)
-            elif start is not None and start.tag == 'motyw':
-                fid = start.attrib['id'][1:]
-                fragments[fid]['themes'].append(start.text)
-                fragments[fid]['content'].append(start.tail)
-            elif start is not None and start.tag == 'end':
-                fid = start.attrib['id'][1:]
-                if fid not in fragments:
-                    continue  # a broken <end> node, skip it
-                frag = fragments[fid]
-                del fragments[fid]
-
-                def jstr(l):
-                    return u' '.join(map(
-                        lambda x: x == None and u'(none)' or unicode(x),
-                        l))
-
-                doc = self.create_book_doc(book)
-                doc.add(Field("fragment_anchor", fid,
+        def fix_format(text):
+            return re.sub("(?m)/$", "", text)
+
+        def add_part(snippets, **fields):
+            doc = self.create_book_doc(book)
+            for f in book_fields:
+                doc.add(f)
+
+            doc.add(NumericField('header_index', Field.Store.YES, True).setIntValue(fields["header_index"]))
+            doc.add(NumericField("header_span", Field.Store.YES, True)\
+                    .setIntValue('header_span' in fields and fields['header_span'] or 1))
+            doc.add(Field('header_type', fields["header_type"], Field.Store.YES, Field.Index.NOT_ANALYZED))
+
+            doc.add(Field('content', fields["content"], Field.Store.NO, Field.Index.ANALYZED, \
+                          Field.TermVector.WITH_POSITIONS_OFFSETS))
+
+            snip_pos = snippets.add(fields["content"])
+            doc.add(NumericField("snippets_position", Field.Store.YES, True).setIntValue(snip_pos[0]))
+            doc.add(NumericField("snippets_length", Field.Store.YES, True).setIntValue(snip_pos[1]))
+
+            if 'fragment_anchor' in fields:
+                doc.add(Field("fragment_anchor", fields['fragment_anchor'],
                               Field.Store.YES, Field.Index.NOT_ANALYZED))
-                doc.add(Field("content",
-                              u' '.join(filter(lambda s: s is not None, frag['content'])),
-                              Field.Store.NO, Field.Index.ANALYZED))
-                doc.add(Field("themes",
-                              u' '.join(filter(lambda s: s is not None, frag['themes'])),
-                              Field.Store.NO, Field.Index.ANALYZED))
-
-                fragment_docs.append(doc)
-            elif start is not None:
-                for frag in fragments.values():
-                    frag['content'].append(start.text)
-            elif end is not None:
-                for frag in fragments.values():
-                    frag['content'].append(end.tail)
-
-        return header_docs + fragment_docs
 
-    def __enter__(self):
-        self.open()
-        return self
+            if 'themes' in fields:
+                themes, themes_pl = zip(*[
+                    (Field("themes", theme, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS),
+                     Field("themes_pl", theme, Field.Store.NO, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS))
+                     for theme in fields['themes']])
 
-    def __exit__(self, type, value, tb):
-        self.close()
+                themes = self.add_gaps(themes, 'themes')
+                themes_pl = self.add_gaps(themes_pl, 'themes_pl')
+
+                for t in themes:
+                    doc.add(t)
+                for t in themes_pl:
+                    doc.add(t)
+
+            return doc
+
+        def give_me_utf8(s):
+            if isinstance(s, unicode):
+                return s.encode('utf-8')
+            else:
+                return s
+
+        fragments = {}
+        snippets = Snippets(book.id).open('w')
+        try:
+            for header, position in zip(list(master), range(len(master))):
+
+                if header.tag in self.skip_header_tags:
+                    continue
+
+                content = u' '.join([t for t in header.itertext()])
+                content = fix_format(content)
+
+                doc = add_part(snippets, header_index=position, header_type=header.tag, content=content)
+
+                self.index.addDocument(doc)
+
+                for start, end in walker(header):
+                    if start is not None and start.tag == 'begin':
+                        fid = start.attrib['id'][1:]
+                        fragments[fid] = {'content': [], 'themes': [], 'start_section': position, 'start_header': header.tag}
+                        fragments[fid]['content'].append(start.tail)
+                    elif start is not None and start.tag == 'motyw':
+                        fid = start.attrib['id'][1:]
+                        if start.text is not None:
+                            fragments[fid]['themes'] += map(str.strip, map(give_me_utf8, start.text.split(',')))
+                        fragments[fid]['content'].append(start.tail)
+                    elif start is not None and start.tag == 'end':
+                        fid = start.attrib['id'][1:]
+                        if fid not in fragments:
+                            continue  # a broken <end> node, skip it
+                        frag = fragments[fid]
+                        if frag['themes'] == []:
+                            continue  # empty themes list.
+                        del fragments[fid]
+
+                        def jstr(l):
+                            return u' '.join(map(
+                                lambda x: x == None and u'(none)' or unicode(x),
+                                l))
+
+                        doc = add_part(snippets,
+                                       header_type=frag['start_header'],
+                                       header_index=frag['start_section'],
+                                       header_span=position - frag['start_section'] + 1,
+                                       fragment_anchor=fid,
+                                       content=u' '.join(filter(lambda s: s is not None, frag['content'])),
+                                       themes=frag['themes'])
+
+                        self.index.addDocument(doc)
+                    elif start is not None:
+                        for frag in fragments.values():
+                            frag['content'].append(start.text)
+                    elif end is not None:
+                        for frag in fragments.values():
+                            frag['content'].append(end.tail)
+        finally:
+            snippets.close()
+
+
+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
+    return _wrap
 
 
 class ReusableIndex(Index):
@@ -249,35 +472,23 @@ class ReusableIndex(Index):
     if you cannot rely on atexit, use ReusableIndex.close_reusable() yourself.
     """
     index = None
-    pool = None
-    pool_jobs = None
 
     def open(self, analyzer=None, threads=4):
         if ReusableIndex.index is not None:
             self.index = ReusableIndex.index
         else:
-            ReusableIndex.pool = ThreadPool(threads)
-            ReusableIndex.pool_jobs = []
+            print("opening index")
             Index.open(self, analyzer)
             ReusableIndex.index = self.index
             atexit.register(ReusableIndex.close_reusable)
 
-    def index_book(self, *args, **kw):
-        job = ReusableIndex.pool.apply_async(Index.index_book, args, kw)
-        ReusableIndex.pool_jobs.append(job)
+    def index_book(self, *args, **kw):
+    #     job = ReusableIndex.pool.apply_async(log_exception_wrapper(Index.index_book), (self,) + args, kw)
+        ReusableIndex.pool_jobs.append(job)
 
     @staticmethod
     def close_reusable():
         if ReusableIndex.index is not None:
-            all_jobs = len(ReusableIndex.pool_jobs)
-            waited=1
-            for job in ReusableIndex.pool_jobs:
-                sys.stdout.write("\rWaiting for search index job: %d/%d..." % 
-                job.wait()
-                waited+=1
-            print("Indexing done.")
-            ReusableIndex.pool.close()
-
             ReusableIndex.index.optimize()
             ReusableIndex.index.close()
             ReusableIndex.index = None
@@ -286,20 +497,13 @@ class ReusableIndex(Index):
         pass
 
 
-class Search(IndexStore):
-    def __init__(self, default_field="content"):
-        IndexStore.__init__(self)
-        self.analyzer = PolishAnalyzer(Version.LUCENE_34)
-        ## self.analyzer = WLAnalyzer()
-        self.searcher = IndexSearcher(self.store, True)
-        self.parser = QueryParser(Version.LUCENE_34, default_field,
-                                  self.analyzer)
-
-        self.parent_filter = TermsFilter()
-        self.parent_filter.addTerm(Term("is_book", "true"))
-
-    def query(self, query):
-        return self.parser.parse(query)
+class JoinSearch(object):
+    """
+    This mixin could be used to handle block join queries.
+    (currently unused)
+    """
+    def __init__(self, *args, **kw):
+        super(JoinSearch, self).__init__(*args, **kw)
 
     def wrapjoins(self, query, fields=[]):
         """
@@ -324,125 +528,647 @@ class Search(IndexStore):
             return BlockJoinQuery(query, self.parent_filter,
                                   BlockJoinQuery.ScoreMode.Total)
 
-    def simple_search(self, query, max_results=50):
-        """Returns (books, total_hits)
-        """
+    def bsearch(self, query, max_results=50):
+        q = self.query(query)
+        bjq = BlockJoinQuery(q, self.parent_filter, BlockJoinQuery.ScoreMode.Avg)
 
-        tops = self.searcher.search(self.query(query), max_results)
+        tops = self.searcher.search(bjq, max_results)
         bks = []
         for found in tops.scoreDocs:
             doc = self.searcher.doc(found.doc)
             bks.append(catalogue.models.Book.objects.get(id=doc.get("book_id")))
         return (bks, tops.totalHits)
 
-    def search(self, query, max_results=50):
-        query = self.query(query)
-        query = self.wrapjoins(query, ["content", "themes"])
 
-        tops = self.searcher.search(query, max_results)
-        bks = []
-        for found in tops.scoreDocs:
-            doc = self.searcher.doc(found.doc)
-            bks.append(catalogue.models.Book.objects.get(id=doc.get("book_id")))
-        return (bks, tops.totalHits)
+class SearchResult(object):
+    def __init__(self, searcher, scoreDocs, score=None, how_found=None, snippets=None):
+        if score:
+            self.score = score
+        else:
+            self.score = scoreDocs.score
 
-    def bsearch(self, query, max_results=50):
-        q = self.query(query)
-        bjq = BlockJoinQuery(q, self.parent_filter, BlockJoinQuery.ScoreMode.Avg)
+        self._hits = []
+        self.hits = None  # processed hits
 
-        tops = self.searcher.search(bjq, max_results)
+        stored = searcher.doc(scoreDocs.doc)
+        self.book_id = int(stored.get("book_id"))
+
+        header_type = stored.get("header_type")
+        if not header_type:
+            return
+
+        sec = (header_type, int(stored.get("header_index")))
+        header_span = stored.get('header_span')
+        header_span = header_span is not None and int(header_span) or 1
+
+        fragment = stored.get("fragment_anchor")
+
+        hit = (sec + (header_span,), fragment, scoreDocs.score, {'how_found': how_found, 'snippets': snippets and [snippets] or []})
+
+        self._hits.append(hit)
+
+    def merge(self, other):
+        if self.book_id != other.book_id:
+            raise ValueError("this search result is or book %d; tried to merge with %d" % (self.book_id, other.book_id))
+        self._hits += other._hits
+        if other.score > self.score:
+            self.score = other.score
+        return self
+
+    def get_book(self):
+        return catalogue.models.Book.objects.get(id=self.book_id)
+
+    book = property(get_book)
+
+    def process_hits(self):
+        POSITION = 0
+        FRAGMENT = 1
+        POSITION_INDEX = 1
+        POSITION_SPAN = 2
+        SCORE = 2
+        OTHER = 3
+
+        # to sections and fragments
+        frags = filter(lambda r: r[FRAGMENT] is not None, self._hits)
+        sect = filter(lambda r: r[FRAGMENT] is None, self._hits)
+        sect = filter(lambda s: 0 == len(filter(
+            lambda f: s[POSITION][POSITION_INDEX] >= f[POSITION][POSITION_INDEX]
+            and s[POSITION][POSITION_INDEX] < f[POSITION][POSITION_INDEX] + f[POSITION][POSITION_SPAN],
+            frags)), sect)
+
+        hits = []
+
+        # remove duplicate fragments
+        fragments = {}
+        for f in frags:
+            fid = f[FRAGMENT]
+            if fid in fragments:
+                if fragments[fid][SCORE] >= f[SCORE]:
+                    continue
+            fragments[fid] = f
+        frags = fragments.values()
+
+        # remove duplicate sections
+        sections = {}
+
+        for s in sect:
+            si = s[POSITION][POSITION_INDEX]
+            # skip existing
+            if si in sections:
+                if sections[si]['score'] >= s[SCORE]:
+                    continue
+
+            m = {'score': s[SCORE],
+                 'section_number': s[POSITION][POSITION_INDEX] + 1,
+                 }
+            m.update(s[OTHER])
+            sections[si] = m
+
+        hits = sections.values()
+
+        for f in frags:
+            frag = catalogue.models.Fragment.objects.get(anchor=f[FRAGMENT])
+            m = {'score': f[SCORE],
+                 'fragment': frag,
+                 'themes': frag.tags.filter(category='theme')
+                 }
+            m.update(f[OTHER])
+            hits.append(m)
+
+        hits.sort(lambda a, b: cmp(a['score'], b['score']), reverse=True)
+
+        self.hits = hits
+
+        return self
+
+    def __unicode__(self):
+        return u'SearchResult(book_id=%d, score=%d)' % (self.book_id, self.score)
+
+    @staticmethod
+    def aggregate(*result_lists):
+        books = {}
+        for rl in result_lists:
+            for r in rl:
+                if r.book_id in books:
+                    books[r.book_id].merge(r)
+                    #print(u"already have one with score %f, and this one has score %f" % (books[book.id][0], found.score))
+                else:
+                    books[r.book_id] = r
+        return books.values()
+
+    def __cmp__(self, other):
+        return cmp(self.score, other.score)
+
+
+class Hint(object):
+    """
+    Given some hint information (information we already know about)
+    our search target - like author, title (specific book), epoch, genre, kind
+    we can narrow down search using filters.
+    """
+    def __init__(self, search):
+        """
+        Accepts a Searcher instance.
+        """
+        self.search = search
+        self.book_tags = {}
+        self.part_tags = []
+        self._books = []
+
+    def books(self, *books):
+        """
+        Give a hint that we search these books.
+        """
+        self._books = books
+
+    def tags(self, tags):
+        """
+        Give a hint that these Tag objects (a list of)
+        is necessary.
+        """
+        for t in tags:
+            if t.category in ['author', 'title', 'epoch', 'genre', 'kind']:
+                lst = self.book_tags.get(t.category, [])
+                lst.append(t)
+                self.book_tags[t.category] = lst
+            if t.category in ['theme', 'theme_pl']:
+                self.part_tags.append(t)
+
+    def tag_filter(self, tags, field='tags'):
+        """
+        Given a lsit of tags and an optional field (but they are normally in tags field)
+        returns a filter accepting only books with specific tags.
+        """
+        q = BooleanQuery()
+
+        for tag in tags:
+            toks = self.search.get_tokens(tag.name, field=field)
+            tag_phrase = PhraseQuery()
+            for tok in toks:
+                tag_phrase.add(Term(field, tok))
+            q.add(BooleanClause(tag_phrase, BooleanClause.Occur.MUST))
+
+        return QueryWrapperFilter(q)
+
+    def book_filter(self):
+        """
+        Filters using book tags (all tag kinds except a theme)
+        """
+        tags = reduce(lambda a, b: a + b, self.book_tags.values(), [])
+        if tags:
+            return self.tag_filter(tags)
+        else:
+            return None
+
+    def part_filter(self):
+        """
+        This filter can be used to look for book parts.
+        It filters on book id and/or themes.
+        """
+        fs = []
+        if self.part_tags:
+            fs.append(self.tag_filter(self.part_tags, field='themes'))
+
+        if self._books != []:
+            bf = BooleanFilter()
+            for b in self._books:
+                id_filter = NumericRangeFilter.newIntRange('book_id', b.id, b.id, True, True)
+                bf.add(FilterClause(id_filter, BooleanClause.Occur.SHOULD))
+            fs.append(bf)
+
+        return Search.chain_filters(fs)
+
+    def should_search_for_book(self):
+        return self._books == []
+
+    def just_search_in(self, all):
+        """Holds logic to figure out which indexes should be search, when we have some hinst already"""
+        some = []
+        for field in all:
+            if field == 'authors' and 'author' in self.book_tags:
+                continue
+            if field == 'title' and self._books != []:
+                continue
+            if (field == 'themes' or field == 'themes_pl') and self.part_tags:
+                continue
+            some.append(field)
+        return some
+
+
+class Search(IndexStore):
+    """
+    Search facilities.
+    """
+    def __init__(self, default_field="content"):
+        IndexStore.__init__(self)
+        self.analyzer = WLAnalyzer()  # PolishAnalyzer(Version.LUCENE_34)
+        # self.analyzer = WLAnalyzer()
+        self.searcher = IndexSearcher(self.store, True)
+        self.parser = QueryParser(Version.LUCENE_34, default_field,
+                                  self.analyzer)
+
+        self.parent_filter = TermsFilter()
+        self.parent_filter.addTerm(Term("is_book", "true"))
+
+    def query(self, query):
+        """Parse query in default Lucene Syntax. (for humans)
+        """
+        return self.parser.parse(query)
+
+    def simple_search(self, query, max_results=50):
+        """Runs a query for books using lucene syntax. (for humans)
+        Returns (books, total_hits)
+        """
+
+        tops = self.searcher.search(self.query(query), max_results)
         bks = []
         for found in tops.scoreDocs:
             doc = self.searcher.doc(found.doc)
             bks.append(catalogue.models.Book.objects.get(id=doc.get("book_id")))
         return (bks, tops.totalHits)
 
-# TokenStream tokenStream = analyzer.tokenStream(fieldName, reader);
-# OffsetAttribute offsetAttribute = tokenStream.getAttribute(OffsetAttribute.class);
-# CharTermAttribute charTermAttribute = tokenStream.getAttribute(CharTermAttribute.class);
-
-# while (tokenStream.incrementToken()) {
-#     int startOffset = offsetAttribute.startOffset();
-#     int endOffset = offsetAttribute.endOffset();
-#     String term = charTermAttribute.toString();
-# }
-
+    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
 
-class MultiSearch(Search):
-    """Class capable of IMDb-like searching"""
-    def get_tokens(self, queryreader):
-        if isinstance(queryreader, str):
-            queryreader = StringReader(queryreader)
-        queryreader.reset()
-        tokens = self.analyzer.reusableTokenStream('content', queryreader)
+        searched.reset()
+        tokens = self.analyzer.reusableTokenStream(field, searched)
         toks = []
         while tokens.incrementToken():
             cta = tokens.getAttribute(CharTermAttribute.class_)
-            toks.append(cta)
+            toks.append(cta.toString())
         return toks
 
-    def make_phrase(self, tokens, field='content', joined=False):
-        phrase = PhraseQuery()
-        for t in tokens:
-            term = Term(field, t)
-            phrase.add(term)
-        if joined:
-            phrase = self.content_query(phrase)
+    def fuzziness(self, fuzzy):
+        """Helper method to sanitize fuzziness"""
+        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):
+        """
+        Return a PhraseQuery with a series of tokens.
+        """
+        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, joined=False):
+    def make_term_query(self, tokens, field='content', modal=BooleanClause.Occur.SHOULD, fuzzy=False):
+        """
+        Returns term queries joined by boolean query.
+        modal - applies to boolean query
+        fuzzy - should the query by fuzzy.
+        """
         q = BooleanQuery()
         for t in tokens:
             term = Term(field, t)
+            if fuzzy:
+                term = FuzzyQuery(term, self.fuzziness(fuzzy))
+            else:
+                term = TermQuery(term)
             q.add(BooleanClause(term, modal))
-        if joined:
-            self.content_query(q)
         return q
 
-    def content_query(self, query):
-        return BlockJoinQuery(query, self.parent_filter,
-                              BlockJoinQuery.ScoreMode.Total)
+    def content_query(self, query):
+        return BlockJoinQuery(query, self.parent_filter,
+                              BlockJoinQuery.ScoreMode.Total)
 
-    def multiseach(self, query, max_results=50):
+    def search_perfect_book(self, searched, max_results=20, fuzzy=False, hint=None):
         """
-        Search strategy:
-        - (phrase) OR -> content
-                      -> title
-                      -> author
-        - (keywords)  -> author
-                      -> motyw
-                      -> tags
-                      -> content
+        Search for perfect book matches. Just see if the query matches with some author or title,
+        taking hints into account.
         """
-        queryreader = StringReader(query)
-        tokens = self.get_tokens(queryreader)
+        fields_to_search = ['authors', 'title']
+        only_in = None
+        if hint:
+            if not hint.should_search_for_book():
+                return []
+            fields_to_search = hint.just_search_in(fields_to_search)
+            only_in = hint.book_filter()
+
+        qrys = [self.make_phrase(self.get_tokens(searched, field=fld), field=fld, fuzzy=fuzzy) for fld in fields_to_search]
+
+        books = []
+        for q in qrys:
+            top = self.searcher.search(q,
+                self.chain_filters([only_in, self.term_filter(Term('is_book', 'true'))]),
+                max_results)
+            for found in top.scoreDocs:
+                books.append(SearchResult(self.searcher, found))
+        return books
+
+    def search_book(self, searched, max_results=20, fuzzy=False, hint=None):
+        fields_to_search = ['tags', 'authors', 'title']
+
+        only_in = None
+        if hint:
+            if not hint.should_search_for_book():
+                return []
+            fields_to_search = hint.just_search_in(fields_to_search)
+            only_in = hint.book_filter()
+
+        tokens = self.get_tokens(searched, field='SIMPLE')
 
-        top_level = BooleanQuery()
-        Should = BooleanClause.Occur.SHOULD
+        q = BooleanQuery()
+
+        for fld in fields_to_search:
+            q.add(BooleanClause(self.make_term_query(tokens, field=fld,
+                                fuzzy=fuzzy), BooleanClause.Occur.SHOULD))
+
+        books = []
+        top = self.searcher.search(q,
+                                   self.chain_filters([only_in, self.term_filter(Term('is_book', 'true'))]),
+            max_results)
+        for found in top.scoreDocs:
+            books.append(SearchResult(self.searcher, found))
+
+        return books
+
+    def search_perfect_parts(self, searched, max_results=20, fuzzy=False, hint=None):
+        """
+        Search for book parts which containt a phrase perfectly matching (with a slop of 2, default for make_phrase())
+        some part/fragment of the book.
+        """
+        qrys = [self.make_phrase(self.get_tokens(searched), field=fld, fuzzy=fuzzy) for fld in ['content']]
+
+        flt = None
+        if hint:
+            flt = hint.part_filter()
+
+        books = []
+        for q in qrys:
+            top = self.searcher.search(q,
+                                       self.chain_filters([self.term_filter(Term('is_book', 'true'), inverse=True),
+                                                           flt]),
+                                       max_results)
+            for found in top.scoreDocs:
+                books.append(SearchResult(self.searcher, found, snippets=self.get_snippets(found, q)))
+
+        return books
+
+    def search_everywhere(self, searched, max_results=20, fuzzy=False, hint=None):
+        """
+        Tries to use search terms to match different fields of book (or its parts).
+        E.g. one word can be an author survey, another be a part of the title, and the rest
+        are some words from third chapter.
+        """
+        books = []
+        only_in = None
+
+        if hint:
+            only_in = hint.part_filter()
+
+        # content only query : themes x content
+        q = BooleanQuery()
+
+        tokens_pl = self.get_tokens(searched, field='content')
+        tokens = self.get_tokens(searched, field='SIMPLE')
+
+        # only search in themes when we do not already filter by themes
+        if hint is None or hint.just_search_in(['themes']) != []:
+            q.add(BooleanClause(self.make_term_query(tokens_pl, field='themes_pl',
+                                                     fuzzy=fuzzy), BooleanClause.Occur.MUST))
+
+        q.add(BooleanClause(self.make_term_query(tokens_pl, field='content',
+                                                 fuzzy=fuzzy), BooleanClause.Occur.SHOULD))
+
+        topDocs = self.searcher.search(q, only_in, max_results)
+        for found in topDocs.scoreDocs:
+            books.append(SearchResult(self.searcher, found))
+            print "* %s theme x content: %s" % (searched, books[-1]._hits)
+
+        # query themes/content x author/title/tags
+        q = BooleanQuery()
+        in_content = BooleanQuery()
+        in_meta = BooleanQuery()
+
+        for fld in ['themes_pl', 'content']:
+            in_content.add(BooleanClause(self.make_term_query(tokens_pl, field=fld, fuzzy=False), BooleanClause.Occur.SHOULD))
+
+        for fld in ['tags', 'authors', 'title']:
+            in_meta.add(BooleanClause(self.make_term_query(tokens, field=fld, fuzzy=False), BooleanClause.Occur.SHOULD))
+
+        q.add(BooleanClause(in_content, BooleanClause.Occur.MUST))
+        q.add(BooleanClause(in_meta, BooleanClause.Occur.SHOULD))
+
+        topDocs = self.searcher.search(q, only_in, max_results)
+        for found in topDocs.scoreDocs:
+            books.append(SearchResult(self.searcher, found))
+            print "* %s scatter search: %s" % (searched, books[-1]._hits)
+
+        return books
+
+    # def multisearch(self, query, max_results=50):
+    #     """
+    #     Search strategy:
+    #     - (phrase) OR -> content
+    #                   -> title
+    #                   -> authors
+    #     - (keywords)  -> authors
+    #                   -> motyw
+    #                   -> tags
+    #                   -> content
+    #     """
+        # queryreader = StringReader(query)
+        # tokens = self.get_tokens(queryreader)
+
+        # top_level = BooleanQuery()
+        # Should = BooleanClause.Occur.SHOULD
+
+        # phrase_level = BooleanQuery()
+        # phrase_level.setBoost(1.3)
+
+        # p_content = self.make_phrase(tokens, joined=True)
+        # p_title = self.make_phrase(tokens, 'title')
+        # p_author = self.make_phrase(tokens, 'author')
+
+        # phrase_level.add(BooleanClause(p_content, Should))
+        # phrase_level.add(BooleanClause(p_title, Should))
+        # phrase_level.add(BooleanClause(p_author, Should))
+
+        # kw_level = BooleanQuery()
+
+        # kw_level.add(self.make_term_query(tokens, 'author'), Should)
+        # j_themes = self.make_term_query(tokens, 'themes', joined=True)
+        # kw_level.add(j_themes, Should)
+        # kw_level.add(self.make_term_query(tokens, 'tags'), Should)
+        # j_con = self.make_term_query(tokens, joined=True)
+        # kw_level.add(j_con, Should)
+
+        # top_level.add(BooleanClause(phrase_level, Should))
+        # top_level.add(BooleanClause(kw_level, Should))
+
+        # return None
+
+    def get_snippets(self, scoreDoc, query, field='content'):
+        """
+        Returns a snippet for found scoreDoc.
+        """
+        htmlFormatter = SimpleHTMLFormatter()
+        highlighter = Highlighter(htmlFormatter, QueryScorer(query))
 
-        phrase_level = BooleanQuery()
+        stored = self.searcher.doc(scoreDoc.doc)
 
-        p_content = self.make_phrase(tokens, joined=True)
-        p_title = self.make_phrase(tokens, 'title')
-        p_author = self.make_phrase(tokens, 'author')
+        # locate content.
+        snippets = Snippets(stored.get('book_id')).open()
+        try:
+            text = snippets.get((int(stored.get('snippets_position')),
+                                 int(stored.get('snippets_length'))))
+        finally:
+            snippets.close()
+
+        tokenStream = TokenSources.getAnyTokenStream(self.searcher.getIndexReader(), scoreDoc.doc, field, self.analyzer)
+        #  highlighter.getBestTextFragments(tokenStream, text, False, 10)
+        snip = highlighter.getBestFragments(tokenStream, text, 3, "...")
+
+        return snip
+
+    @staticmethod
+    def enum_to_array(enum):
+        """
+        Converts a lucene TermEnum to array of Terms, suitable for
+        addition to queries
+        """
+        terms = []
+
+        while True:
+            t = enum.term()
+            if t:
+                terms.append(t)
+            if not enum.next(): break
 
-        phrase_level.add(BooleanClause(p_content, Should))
-        phrase_level.add(BooleanClause(p_title, Should))
-        phrase_level.add(BooleanClause(p_author, Should))
+        if terms:
+            return JArray('object')(terms, Term)
 
-        kw_level = BooleanQuery()
+    def search_tags(self, query, filter=None, max_results=40):
+        """
+        Search for Tag objects using query.
+        """
+        tops = self.searcher.search(query, filter, max_results)
 
-        kw_level.add(self.make_term_query(tokens, 'author'), Should)
-        kw_level.add(self.make_term_query(tokens, 'themes', joined=True), Should)
-        kw_level.add(self.make_term_query(tokens, 'tags'), Should)
-        kw_level.add(self.make_term_query(tokens, joined=True), Should)
+        tags = []
+        for found in tops.scoreDocs:
+            doc = self.searcher.doc(found.doc)
+            tag = catalogue.models.Tag.objects.get(id=doc.get("tag_id"))
+            tags.append(tag)
+            print "%s (%d) -> %f" % (tag, tag.id, found.score)
 
-        top_level.add(BooleanClause(phrase_level, Should))
-        top_level.add(BooleanClause(kw_level, Should))
+        return tags
 
-        tops = self.searcher.search(top_level, max_results)
+    def search_books(self, query, filter=None, max_results=10):
+        """
+        Searches for Book objects using query
+        """
         bks = []
+        tops = self.searcher.search(query, filter, max_results)
         for found in tops.scoreDocs:
             doc = self.searcher.doc(found.doc)
             bks.append(catalogue.models.Book.objects.get(id=doc.get("book_id")))
-        return (bks, tops.totalHits)
+        return bks
+
+    def create_prefix_phrase(self, toks, field):
+        q = MultiPhraseQuery()
+        for i in range(len(toks)):
+            t = Term(field, toks[i])
+            if i == len(toks) - 1:
+                pterms = Search.enum_to_array(PrefixTermEnum(self.searcher.getIndexReader(), t))
+                if pterms:
+                    q.add(pterms)
+                else:
+                    q.add(t)
+            else:
+                q.add(t)
+        return q
+
+    @staticmethod
+    def term_filter(term, inverse=False):
+        only_term = TermsFilter()
+        only_term.addTerm(term)
+
+        if inverse:
+            neg = BooleanFilter()
+            neg.add(FilterClause(only_term, BooleanClause.Occur.MUST_NOT))
+            only_term = neg
+
+        return only_term
+
+    def hint_tags(self, string, max_results=50):
+        """
+        Return auto-complete hints for tags
+        using prefix search.
+        """
+        toks = self.get_tokens(string, field='SIMPLE')
+        top = BooleanQuery()
+
+        for field in ['tag_name', 'tag_name_pl']:
+            q = self.create_prefix_phrase(toks, field)
+            top.add(BooleanClause(q, BooleanClause.Occur.SHOULD))
+
+        no_book_cat = self.term_filter(Term("tag_category", "book"), inverse=True)
+
+        return self.search_tags(top, no_book_cat, max_results=max_results)
+
+    def hint_books(self, string, max_results=50):
+        """
+        Returns auto-complete hints for book titles
+        Because we do not index 'pseudo' title-tags.
+        Prefix search.
+        """
+        toks = self.get_tokens(string, field='SIMPLE')
+
+        q = self.create_prefix_phrase(toks, 'title')
+
+        return self.search_books(q, self.term_filter(Term("is_book", "true")), max_results=max_results)
+
+    @staticmethod
+    def chain_filters(filters, op=ChainedFilter.AND):
+        """
+        Chains a filter list together
+        """
+        filters = filter(lambda x: x is not None, filters)
+        if not filters:
+            return None
+        chf = ChainedFilter(JArray('object')(filters, Filter), op)
+        return chf
+
+    def filtered_categories(self, tags):
+        """
+        Return a list of tag categories, present in tags list.
+        """
+        cats = {}
+        for t in tags:
+            cats[t.category] = True
+        return cats.keys()
+
+    def hint(self):
+        return Hint(self)