Another upgrade.
[wolnelektury.git] / src / search / index.py
index 7120893..fc9e9d5 100644 (file)
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
-from functools import reduce, total_ordering
-from itertools import chain
-import logging
-import operator
-import os
 import re
-from django.conf import settings
-from librarian import dcparser
 from librarian.parser import WLDocument
 from lxml import etree
-import scorched
-import catalogue.models
-import picture.models
-from pdcounter.models import Author as PDCounterAuthor, BookStub as PDCounterBook
-from wolnelektury.utils import makedirs
-from . import custom
 
-log = logging.getLogger('search')
 
-
-if os.path.isfile(settings.SOLR_STOPWORDS):
-    stopwords = set(
-        line.strip()
-        for line in open(settings.SOLR_STOPWORDS) if not line.startswith('#'))
-else:
-    stopwords = set()
-
-
-class SolrIndex(object):
-    def __init__(self, mode=None):
-        self.index = custom.CustomSolrInterface(settings.SOLR, mode=mode)
-
-
-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, revision=None):
-        makedirs(os.path.join(settings.SEARCH_INDEX, self.SNIPPET_DIR))
-        self.book_id = book_id
-        self.revision = revision
-        self.file = None
-        self.position = None
-
-    @property
-    def path(self):
-        if self.revision:
-            fn = "%d.%d" % (self.book_id, self.revision)
-        else:
-            fn = "%d" % self.book_id
-
-        return os.path.join(settings.SEARCH_INDEX, self.SNIPPET_DIR, fn)
-
-    def open(self, mode='r'):
-        """
-        Open the snippet file. Call .close() afterwards.
-        """
-        if 'b' not in mode:
-            mode += 'b'
-
-        if 'w' in mode:
-            if os.path.exists(self.path):
-                self.revision = 1
-                while True:
-                    if not os.path.exists(self.path):
-                        break
-                    self.revision += 1
-
-        self.file = open(self.path, 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"""
-        if self.file:
-            self.file.close()
-
-    def remove(self):
-        self.revision = None
-        try:
-            os.unlink(self.path)
-            self.revision = 0
-            while True:
-                self.revision += 1
-                os.unlink(self.path)
-        except OSError:
-            pass
-
-
-class Index(SolrIndex):
+class Index:
     """
     Class indexing books.
     """
-    def __init__(self):
-        super(Index, self).__init__(mode='rw')
-
-    def delete_query(self, *queries):
-        """
-        index.delete(queries=...) doesn't work, so let's reimplement it
-        using deletion of list of uids.
-        """
-        uids = set()
-        for q in queries:
-            if isinstance(q, scorched.search.LuceneQuery):
-                q = self.index.query(q)
-            q.field_limiter.update(['uid'])
-            st = 0
-            rows = 100
-            while True:
-                ids = q.paginate(start=st, rows=rows).execute()
-                if not len(ids):
-                    break
-                for res in ids:
-                    uids.add(res['uid'])
-                st += rows
-        if uids:
-            self.index.delete(uids)
-            return True
-        else:
-            return False
-
-    def index_tags(self, *tags, **kw):
-        """
-        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
-        """
-        log.debug("Indexing tags")
-        remove_only = kw.get('remove_only', False)
-        # first, remove tags from index.
-        if tags:
-            tag_qs = []
-            for tag in tags:
-                q_id = self.index.Q(tag_id=tag.id)
-
-                if isinstance(tag, PDCounterAuthor):
-                    q_cat = self.index.Q(tag_category='pd_author')
-                elif isinstance(tag, PDCounterBook):
-                    q_cat = self.index.Q(tag_category='pd_book')
-                else:
-                    q_cat = self.index.Q(tag_category=tag.category)
-
-                q_id_cat = self.index.Q(q_id & q_cat)
-                tag_qs.append(q_id_cat)
-            self.delete_query(*tag_qs)
-        else:  # all
-            q = self.index.Q(tag_id__any=True)
-            self.delete_query(q)
-
-        if not remove_only:
-            # then add them [all or just one passed]
-            if not tags:
-                tags = chain(
-                    catalogue.models.Tag.objects.exclude(category='set'),
-                    PDCounterAuthor.objects.all(),
-                    PDCounterBook.objects.all())
-
-            for tag in tags:
-                if isinstance(tag, PDCounterAuthor):
-                    doc = {
-                        "tag_id": int(tag.id),
-                        "tag_name": tag.name,
-                        "tag_name_pl": tag.name,
-                        "tag_category": 'pd_author',
-                        "is_pdcounter": True,
-                        "uid": "tag%d_pd_a" % tag.id
-                        }
-                elif isinstance(tag, PDCounterBook):
-                    doc = {
-                        "tag_id": int(tag.id),
-                        "tag_name": tag.title,
-                        "tag_name_pl": tag.title,
-                        "tag_category": 'pd_book',
-                        "is_pdcounter": True,
-                        "uid": "tag%d_pd_b" % tag.id
-                        }
-                else:
-                    doc = {
-                        "tag_id": int(tag.id),
-                        "tag_name": tag.name,
-                        "tag_name_pl": tag.name,
-                        "tag_category": tag.category,
-                        "is_pdcounter": False,
-                        "uid": "tag%d" % tag.id
-                        }
-                self.index.add(doc)
-
-    def create_book_doc(self, book):
-        """
-        Create a lucene document referring book id.
-        """
-        doc = {'book_id': int(book.id)}
-        if book.parent is not None:
-            doc['parent_id'] = int(book.parent.id)
-        return doc
-
-    def remove_book(self, book_or_id, remove_snippets=True):
-        """Removes a book from search index.
-        book - Book instance."""
-        if isinstance(book_or_id, catalogue.models.Book):
-            book_id = book_or_id.id
-        else:
-            book_id = book_or_id
-
-        self.delete_query(self.index.Q(book_id=book_id))
-
-        if remove_snippets:
-            snippets = Snippets(book_id)
-            snippets.remove()
-
-    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:
-            # we don't remove snippets, since they might be still needed by
-            # threads using not reopened index
-            self.remove_book(book, remove_snippets=False)
-
-        book_doc = self.create_book_doc(book)
-        meta_fields = self.extract_metadata(book, book_info, dc_only=[
-            'source_name', 'authors', 'translators', 'title', 'epochs', 'kinds', 'genres'])
-        # let's not index it - it's only used for extracting publish date
-        if 'source_name' in meta_fields:
-            del meta_fields['source_name']
-
-        for n, f in meta_fields.items():
-            book_doc[n] = f
-
-        book_doc['uid'] = "book%s" % book_doc['book_id']
-        self.index.add(book_doc)
-        del book_doc
-        book_fields = {
-            'title': meta_fields['title'],
-            'authors': meta_fields['authors'],
-            'published_date': meta_fields['published_date']
-            }
-
-        for tag_name in ('translators', 'epochs', 'kinds', 'genres'):
-            if tag_name in meta_fields:
-                book_fields[tag_name] = meta_fields[tag_name]
-
-        self.index_content(book, book_fields=book_fields)
-
     master_tags = [
         'opowiadanie',
         'powiesc',
@@ -285,7 +23,7 @@ class Index(SolrIndex):
         'uwaga', 'extra', 'nota_red', 'abstrakt',
         'zastepnik_tekstu', 'sekcja_asterysk', 'separator_linia', 'zastepnik_wersu',
         'didaskalia',
-        'naglowek_aktu', 'naglowek_sceny', 'naglowek_czesc',
+        'naglowek_aktu', 'naglowek_sceny', 'naglowek_czesc', 'motyw'
     ]
 
     footnote_tags = ['pa', 'pt', 'pr', 'pe']
@@ -293,89 +31,41 @@ class Index(SolrIndex):
     skip_header_tags = ['autor_utworu', 'nazwa_utworu', 'dzielo_nadrzedne',
                         '{http://www.w3.org/1999/02/22-rdf-syntax-ns#}RDF']
 
-    published_date_re = re.compile("([0-9]+)[\]. ]*$")
-
-    def extract_metadata(self, book, book_info=None, dc_only=None):
-        """
-        Extract metadata from book and returns a map of fields keyed by fieldname
-        """
-        fields = {}
-
-        if book_info is None:
-            book_info = dcparser.parse(open(book.xml_file.path))
-
-        fields['slug'] = book.slug
-        fields['is_book'] = True
-
-        # validator, name
-        for field in dcparser.BookInfo.FIELDS:
-            if dc_only and field.name not in dc_only:
-                continue
-            if hasattr(book_info, field.name):
-                if not getattr(book_info, field.name):
-                    continue
-                # since no type information is available, we use validator
-                type_indicator = field.validator
-                if type_indicator == dcparser.as_unicode:
-                    s = getattr(book_info, field.name)
-                    if field.multiple:
-                        s = ', '.join(s)
-                    fields[field.name] = s
-                elif type_indicator == dcparser.as_person:
-                    p = getattr(book_info, field.name)
-                    if isinstance(p, dcparser.Person):
-                        persons = str(p)
-                    else:
-                        persons = ', '.join(map(str, p))
-                    fields[field.name] = persons
-                elif type_indicator == dcparser.as_date:
-                    dt = getattr(book_info, field.name)
-                    fields[field.name] = dt
-
-        # get published date
-        pd = None
-        if hasattr(book_info, 'source_name') and book_info.source_name:
-            match = self.published_date_re.search(book_info.source_name)
-            if match is not None:
-                pd = str(match.groups()[0])
-        if not pd:
-            pd = ""
-        fields["published_date"] = pd
-
-        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):
+    @classmethod
+    def get_master(cls, root):
         """
         Returns the first master tag from an etree.
         """
         for master in root.iter():
-            if master.tag in self.master_tags:
+            if master.tag in cls.master_tags:
                 return master
 
-    def index_content(self, book, book_fields):
+    @staticmethod
+    def add_snippet(book, text, position):
+        book.snippet_set.create(
+            sec=position + 1,
+            text=text
+        )
+
+    @classmethod
+    def index_book(cls, book):
         """
         Walks the book XML and extract content from it.
         Adds parts for each header tag and for each fragment.
         """
+        if not book.xml_file: return
+
+        book.snippet_set.all().delete()
+
         wld = WLDocument.from_file(book.xml_file.path, parse_dublincore=False)
         root = wld.edoc.getroot()
 
-        master = self.get_master(root)
+        master = cls.get_master(root)
         if master is None:
             return []
 
         def walker(node):
-            if node.tag not in self.ignore_content_tags:
+            if node.tag not in cls.ignore_content_tags:
                 yield node, None, None
                 if node.text is not None:
                     yield None, node.text, None
@@ -389,616 +79,43 @@ class Index(SolrIndex):
             return
 
         def fix_format(text):
-            # separator = [u" ", u"\t", u".", u";", u","]
             if isinstance(text, list):
-                # need to join it first
                 text = filter(lambda s: s is not None, content)
-                text = u' '.join(text)
-                # for i in range(len(text)):
-                #     if i > 0:
-                #         if text[i][0] not in separator\
-                #             and text[i - 1][-1] not in separator:
-                #          text.insert(i, u" ")
+                text = ' '.join(text)
 
             return re.sub("(?m)/$", "", text)
 
-        def add_part(snippets, **fields):
-            doc = self.create_book_doc(book)
-            for n, v in book_fields.items():
-                doc[n] = v
-
-            doc['header_index'] = fields["header_index"]
-            doc['header_span'] = 'header_span' in fields and fields['header_span'] or 1
-            doc['header_type'] = fields['header_type']
-
-            doc['text'] = fields['text']
-
-            # snippets
-            snip_pos = snippets.add(fields["text"])
-
-            doc['snippets_position'] = snip_pos[0]
-            doc['snippets_length'] = snip_pos[1]
-            if snippets.revision:
-                doc["snippets_revision"] = snippets.revision
-
-            if 'fragment_anchor' in fields:
-                doc["fragment_anchor"] = fields['fragment_anchor']
-
-            if 'themes' in fields:
-                doc['themes'] = fields['themes']
-            doc['uid'] = "part%s-%s-%s-%s" % (
-                book.id, doc['header_index'], doc['header_span'], doc.get('fragment_anchor', ''))
-            return doc
-
-        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
-                if header.tag is etree.Comment:
-                    continue
-
-                # section content
-                content = []
-                footnote = []
-
-                def all_content(text):
-                    for frag in fragments.values():
-                        frag['text'].append(text)
-                    content.append(text)
-                handle_text = [all_content]
-
-                for start, text, end in walker(header):
-                    # handle footnotes
-                    if start is not None and start.tag in self.footnote_tags:
-                        footnote = []
-
-                        def collect_footnote(t):
-                            footnote.append(t)
-
-                        handle_text.append(collect_footnote)
-                    elif end is not None and footnote is not [] and end.tag in self.footnote_tags:
-                        handle_text.pop()
-                        doc = add_part(snippets, header_index=position, header_type=header.tag,
-                                       text=u''.join(footnote),
-                                       is_footnote=True)
-                        self.index.add(doc)
-                        footnote = []
-
-                    # handle fragments and themes.
-                    if start is not None and start.tag == 'begin':
-                        fid = start.attrib['id'][1:]
-                        fragments[fid] = {
-                            'text': [], 'themes': [], 'start_section': position, 'start_header': header.tag}
-
-                    # themes for this fragment
-                    elif start is not None and start.tag == 'motyw':
-                        fid = start.attrib['id'][1:]
-                        handle_text.append(lambda text: None)
-                        if start.text is not None:
-                            fragments[fid]['themes'] += map(str.strip, map(str, (start.text.split(','))))
-                    elif end is not None and end.tag == 'motyw':
-                        handle_text.pop()
-
-                    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 not frag['themes']:
-                            continue  # empty themes list.
-                        del fragments[fid]
-
-                        doc = add_part(snippets,
-                                       header_type=frag['start_header'],
-                                       header_index=frag['start_section'],
-                                       header_span=position - frag['start_section'] + 1,
-                                       fragment_anchor=fid,
-                                       text=fix_format(frag['text']),
-                                       themes=frag['themes'])
-                        self.index.add(doc)
-
-                        # Collect content.
-
-                    if text is not None and handle_text is not []:
-                        hdl = handle_text[-1]
-                        hdl(text)
-
-                        # in the end, add a section text.
-                doc = add_part(snippets, header_index=position,
-                               header_type=header.tag, text=fix_format(content))
-
-                self.index.add(doc)
-
-        finally:
-            snippets.close()
-
-    def remove_picture(self, picture_or_id):
-        """Removes a picture from search index."""
-        if isinstance(picture_or_id, picture.models.Picture):
-            picture_id = picture_or_id.id
-        else:
-            picture_id = picture_or_id
-        self.delete_query(self.index.Q(picture_id=picture_id))
-
-    def index_picture(self, picture, picture_info=None, overwrite=True):
-        """
-        Indexes the picture.
-        Creates a lucene document for extracted metadata
-        and calls self.index_area() to index the contents of the picture.
-        """
-        if overwrite:
-            # we don't remove snippets, since they might be still needed by
-            # threads using not reopened index
-            self.remove_picture(picture)
-
-        picture_doc = {'picture_id': int(picture.id)}
-        meta_fields = self.extract_metadata(picture, picture_info, dc_only=[
-            'authors', 'title', 'epochs', 'kinds', 'genres'])
-
-        picture_doc.update(meta_fields)
-
-        picture_doc['uid'] = "picture%s" % picture_doc['picture_id']
-        self.index.add(picture_doc)
-        del picture_doc['is_book']
-        for area in picture.areas.all():
-            self.index_area(area, picture_fields=picture_doc)
-
-    def index_area(self, area, picture_fields):
-        """
-        Indexes themes and objects on the area.
-        """
-        doc = dict(picture_fields)
-        doc['area_id'] = area.id
-        doc['themes'] = list(area.tags.filter(category__in=('theme', 'thing')).values_list('name', flat=True))
-        doc['uid'] = 'area%s' % area.id
-        self.index.add(doc)
-
-
-@total_ordering
-class SearchResult(object):
-    def __init__(self, doc, how_found=None, query_terms=None):
-        self.boost = 1.0
-        self._hits = []
-        self._processed_hits = None  # processed hits
-        self.snippets = []
-        self.query_terms = query_terms
-        self._book = None
-
-        if 'score' in doc:
-            self._score = doc['score']
-        else:
-            self._score = 0
-
-        self.book_id = int(doc["book_id"])
-
-        try:
-            self.published_date = int(doc.get("published_date"))
-        except ValueError:
-            self.published_date = 0
-
-        # content hits
-        header_type = doc.get("header_type", None)
-        # we have a content hit in some header of fragment
-        if header_type is not None:
-            sec = (header_type, int(doc["header_index"]))
-            header_span = doc['header_span']
-            header_span = header_span is not None and int(header_span) or 1
-            fragment = doc.get("fragment_anchor", None)
-            snippets_pos = (doc['snippets_position'], doc['snippets_length'])
-            snippets_rev = doc.get('snippets_revision', None)
-
-            hit = (sec + (header_span,), fragment, self._score, {
-                'how_found': how_found,
-                'snippets_pos': snippets_pos,
-                'snippets_revision': snippets_rev,
-                'themes': doc.get('themes', []),
-                'themes_pl': doc.get('themes_pl', [])
-                })
-
-            self._hits.append(hit)
-
-    @classmethod
-    def from_book(cls, book, how_found=None, query_terms=None):
-        doc = {
-            'score': book.popularity.count,
-            'book_id': book.id,
-            'published_date': 0,
-        }
-        result = cls(doc, how_found=how_found, query_terms=query_terms)
-        result._book = book
-        return result
-
-    def __str__(self):
-        return u"<SR id=%d %d(%d) hits score=%f %d snippets>" % \
-            (self.book_id, len(self._hits),
-             len(self._processed_hits) if self._processed_hits else -1,
-             self._score, len(self.snippets))
-
-    def __bytes__(self):
-        return str(self).encode('utf-8')
-
-    @property
-    def score(self):
-        return self._score * self.boost
-
-    def merge(self, other):
-        if self.book_id != other.book_id:
-            raise ValueError("this search result is for book %d; tried to merge with %d" % (self.book_id, other.book_id))
-        self._hits += other._hits
-        self._score += max(other._score, 0)
-        return self
-
-    def get_book(self):
-        if self._book is not None:
-            return self._book
-        self._book = catalogue.models.Book.objects.get(id=self.book_id)
-        return self._book
-
-    book = property(get_book)
-
-    POSITION = 0
-    FRAGMENT = 1
-    POSITION_INDEX = 1
-    POSITION_SPAN = 2
-    SCORE = 2
-    OTHER = 3
-
-    @property
-    def hits(self):
-        if self._processed_hits is not None:
-            return self._processed_hits
-
-        # to sections and fragments
-        frags = filter(lambda r: r[self.FRAGMENT] is not None, self._hits)
-
-        sect = [hit for hit in self._hits if hit[self.FRAGMENT] is None]
-
-        # sections not covered by fragments
-        sect = filter(lambda s: 0 == len(list(filter(
-            lambda f: f[self.POSITION][self.POSITION_INDEX] <= s[self.POSITION][self.POSITION_INDEX] <
-                      f[self.POSITION][self.POSITION_INDEX] + f[self.POSITION][self.POSITION_SPAN], frags))), sect)
-
-        def remove_duplicates(lst, keyfn, compare):
-            els = {}
-            for e in lst:
-                eif = keyfn(e)
-                if eif in els:
-                    if compare(els[eif], e) >= 1:
-                        continue
-                els[eif] = e
-            return els.values()
-
-        # remove fragments with duplicated fid's and duplicated snippets
-        frags = remove_duplicates(frags, lambda f: f[self.FRAGMENT], lambda a, b: cmp(a[self.SCORE], b[self.SCORE]))
-        # frags = remove_duplicates(frags, lambda f: f[OTHER]['snippet_pos'] and f[OTHER]['snippet_pos'] or f[FRAGMENT],
-        #                           lambda a, b: cmp(a[SCORE], b[SCORE]))
-
-        # remove duplicate sections
-        sections = {}
-
-        for s in sect:
-            si = s[self.POSITION][self.POSITION_INDEX]
-            # skip existing
-            if si in sections:
-                if sections[si]['score'] >= s[self.SCORE]:
-                    continue
-
-            m = {'score': s[self.SCORE],
-                 'section_number': s[self.POSITION][self.POSITION_INDEX] + 1,
-                 }
-            m.update(s[self.OTHER])
-            sections[si] = m
-
-        hits = list(sections.values())
-
-        for f in frags:
-            try:
-                frag = catalogue.models.Fragment.objects.get(anchor=f[self.FRAGMENT], book__id=self.book_id)
-            except catalogue.models.Fragment.DoesNotExist:
-                # stale index
+        for position, header in enumerate(master):
+            if header.tag in cls.skip_header_tags:
                 continue
-            # Figure out if we were searching for a token matching some word in theme name.
-            themes = frag.tags.filter(category='theme')
-            themes_hit = set()
-            if self.query_terms is not None:
-                for i in range(0, len(f[self.OTHER]['themes'])):
-                    tms = f[self.OTHER]['themes'][i].split(r' +') + f[self.OTHER]['themes_pl'][i].split(' ')
-                    tms = map(str.lower, tms)
-                    for qt in self.query_terms:
-                        if qt in tms:
-                            themes_hit.add(f[self.OTHER]['themes'][i])
-                            break
-
-            def theme_by_name(n):
-                th = list(filter(lambda t: t.name == n, themes))
-                if th:
-                    return th[0]
-                else:
-                    return None
-            themes_hit = list(filter(lambda a: a is not None, map(theme_by_name, themes_hit)))
-
-            m = {'score': f[self.SCORE],
-                 'fragment': frag,
-                 'section_number': f[self.POSITION][self.POSITION_INDEX] + 1,
-                 'themes': themes,
-                 'themes_hit': themes_hit
-                 }
-            m.update(f[self.OTHER])
-            hits.append(m)
-
-        hits.sort(key=lambda h: h['score'], reverse=True)
-
-        self._processed_hits = hits
-
-        return hits
-
-    @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)
-                else:
-                    books[r.book_id] = r
-        return books.values()
-
-    def __lt__(self, other):
-        return (-self.score, self.published_date, self.book.sort_key_author, self.book.sort_key) > \
-               (-other.score, other.published_date, other.book.sort_key_author, other.book.sort_key)
-
-    def __eq__(self, other):
-        return (self.score, self.published_date, self.book.sort_key_author, self.book.sort_key) == \
-               (other.score, other.published_date, other.book.sort_key_author, other.book.sort_key)
-
-    def __len__(self):
-        return len(self.hits)
-
-    def snippet_pos(self, idx=0):
-        return self.hits[idx]['snippets_pos']
-
-    def snippet_revision(self, idx=0):
-        try:
-            return self.hits[idx]['snippets_revision']
-        except (IndexError, KeyError):
-            return None
-
-
-@total_ordering
-class PictureResult(object):
-    def __init__(self, doc, how_found=None, query_terms=None):
-        self.boost = 1.0
-        self.query_terms = query_terms
-        self._picture = None
-        self._hits = []
-        self._processed_hits = None
-
-        if 'score' in doc:
-            self._score = doc['score']
-        else:
-            self._score = 0
-
-        self.picture_id = int(doc["picture_id"])
-
-        if doc.get('area_id'):
-            hit = (self._score, {
-                'how_found': how_found,
-                'area_id': doc['area_id'],
-                'themes': doc.get('themes', []),
-                'themes_pl': doc.get('themes_pl', []),
-            })
-
-            self._hits.append(hit)
-
-    def __str__(self):
-        return u"<PR id=%d score=%f >" % (self.picture_id, self._score)
-
-    def __repr__(self):
-        return str(self)
-
-    @property
-    def score(self):
-        return self._score * self.boost
-
-    def merge(self, other):
-        if self.picture_id != other.picture_id:
-            raise ValueError(
-                "this search result is for picture %d; tried to merge with %d" % (self.picture_id, other.picture_id))
-        self._hits += other._hits
-        self._score += max(other._score, 0)
-        return self
-
-    SCORE = 0
-    OTHER = 1
-
-    @property
-    def hits(self):
-        if self._processed_hits is not None:
-            return self._processed_hits
-
-        hits = []
-        for hit in self._hits:
-            try:
-                area = picture.models.PictureArea.objects.get(id=hit[self.OTHER]['area_id'])
-            except picture.models.PictureArea.DoesNotExist:
-                # stale index
+            if header.tag is etree.Comment:
                 continue
-            # Figure out if we were searching for a token matching some word in theme name.
-            themes_hit = set()
-            if self.query_terms is not None:
-                for i in range(0, len(hit[self.OTHER]['themes'])):
-                    tms = hit[self.OTHER]['themes'][i].split(r' +') + hit[self.OTHER]['themes_pl'][i].split(' ')
-                    tms = map(str.lower, tms)
-                    for qt in self.query_terms:
-                        if qt in tms:
-                            themes_hit.add(hit[self.OTHER]['themes'][i])
-                            break
-
-            m = {
-                'score': hit[self.SCORE],
-                'area': area,
-                'themes_hit': themes_hit,
-            }
-            m.update(hit[self.OTHER])
-            hits.append(m)
-
-        hits.sort(key=lambda h: h['score'], reverse=True)
-        hits = hits[:1]
-        self._processed_hits = hits
-        return hits
-
-    def get_picture(self):
-        if self._picture is None:
-            self._picture = picture.models.Picture.objects.get(id=self.picture_id)
-        return self._picture
-
-    picture = property(get_picture)
 
-    @staticmethod
-    def aggregate(*result_lists):
-        books = {}
-        for rl in result_lists:
-            for r in rl:
-                if r.picture_id in books:
-                    books[r.picture_id].merge(r)
-                else:
-                    books[r.picture_id] = r
-        return books.values()
-
-    def __lt__(self, other):
-        return self.score < other.score
-
-    def __eq__(self, other):
-        return self.score == other.score
-
-
-class Search(SolrIndex):
-    """
-    Search facilities.
-    """
-    def __init__(self, default_field="text"):
-        super(Search, self).__init__(mode='r')
-
-    def make_term_query(self, query, field='text', modal=operator.or_):
-        """
-        Returns term queries joined by boolean query.
-        modal - applies to boolean query
-        fuzzy - should the query by fuzzy.
-        """
-        if query is None:
-            query = ''
-        q = self.index.Q()
-        q = reduce(modal, map(lambda s: self.index.Q(**{field: s}), query.split(r" ")), q)
-
-        return q
+            # section content
+            content = []
+            footnote = []
 
-    def search_by_author(self, words):
-        from catalogue.models import Book
-        books = Book.objects.filter(parent=None).order_by('-popularity__count')
-        for word in words:
-            books = books.filter(cached_author__iregex='\m%s\M' % word).select_related('popularity__count')
-        return [SearchResult.from_book(book, how_found='search_by_author', query_terms=words) for book in books[:30]]
+            def all_content(text):
+                content.append(text)
+            handle_text = [all_content]
 
-    def search_words(self, words, fields, required=None, book=True, picture=False):
-        if book and not picture and fields == ['authors']:
-            return self.search_by_author(words)
-        filters = []
-        for word in words:
-            if book or picture or (word not in stopwords):
-                word_filter = None
-                for field in fields:
-                    q = self.index.Q(**{field: word})
-                    if word_filter is None:
-                        word_filter = q
-                    else:
-                        word_filter |= q
-                filters.append(word_filter)
-        if required:
-            required_filter = None
-            for field in required:
-                for word in words:
-                    if book or picture or (word not in stopwords):
-                        q = self.index.Q(**{field: word})
-                        if required_filter is None:
-                            required_filter = q
-                        else:
-                            required_filter |= q
-            filters.append(required_filter)
-        if not filters:
-            return []
-        params = {}
-        if book:
-            params['is_book'] = True
-        if picture:
-            params['picture_id__gt'] = 0
-        else:
-            params['book_id__gt'] = 0
-        query = self.index.query(**params)
-        query = self.apply_filters(query, filters).field_limit(score=True, all_fields=True)
-        result_class = PictureResult if picture else SearchResult
-        return [result_class(found, how_found='search_words', query_terms=words) for found in query.execute()]
-
-    def get_snippets(self, searchresult, query, field='text', num=1):
-        """
-        Returns a snippet for found scoreDoc.
-        """
-        maxnum = len(searchresult)
-        if num is None or num < 0 or num > maxnum:
-            num = maxnum
-        book_id = searchresult.book_id
-        revision = searchresult.snippet_revision()
-        snippets = Snippets(book_id, revision=revision)
-        snips = [None] * maxnum
-        try:
-            snippets.open()
-            idx = 0
-            while idx < maxnum and num > 0:
-                position, length = searchresult.snippet_pos(idx)
-                if position is None or length is None:
-                    continue
-                text = snippets.get((int(position),
-                                     int(length)))
-                snip = self.index.highlight(text=text, field=field, q=query)
-                if not snip and field == 'text':
-                    snip = self.index.highlight(text=text, field='text_nonstem', q=query)
-                if snip not in snips:
-                    snips[idx] = snip
-                    if snip:
-                        num -= 1
-                idx += 1
-
-        except IOError as e:
-            book = catalogue.models.Book.objects.filter(id=book_id)
-            if not book:
-                log.error("Book does not exist for book id = %d" % book_id)
-            elif not book.get().children.exists():
-                log.error("Cannot open snippet file for book id = %d [rev=%s], %s" % (book_id, revision, e))
-            return []
-        finally:
-            snippets.close()
+            for start, text, end in walker(header):
+                # handle footnotes
+                if start is not None and start.tag in cls.footnote_tags:
+                    footnote = []
 
-        # remove verse end markers..
-        snips = [s.replace("/\n", "\n") if s else s for s in snips]
+                    def collect_footnote(t):
+                        footnote.append(t)
 
-        searchresult.snippets = snips
-
-        return snips
-
-    @staticmethod
-    def apply_filters(query, filters):
-        """
-        Apply filters to a query
-        """
-        if filters is None:
-            filters = []
-        filters = filter(lambda x: x is not None, filters)
-        for f in filters:
-            query = query.query(f)
-        return query
+                    handle_text.append(collect_footnote)
+                elif end is not None and footnote is not [] and end.tag in cls.footnote_tags:
+                    handle_text.pop()
+                    cls.add_snippet(book, ''.join(footnote), position)
+                    footnote = []
 
+                if text is not None and handle_text is not []:
+                    hdl = handle_text[-1]
+                    hdl(text)
 
-if getattr(settings, 'SEARCH_MOCK', False):
-    from .mock_search import Search
+            # in the end, add a section text.
+            cls.add_snippet(book, fix_format(content), position)