Add grammar to tags. Also: respect custom tag slugs on publishing.

[wolnelektury.git] / src / search / index.py

diff --git a/src/search/index.py b/src/search/index.py
index a712b07..2be60fd 100644 (file)
--- a/src/search/index.py
+++ b/src/search/index.py
@@ -1,7 +1,7 @@
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
 # 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 total_ordering
+from functools import reduce, total_ordering

 from itertools import chain
 import logging
 import operator
 from itertools import chain
 import logging
 import operator


@@ -95,7 +95,10 @@ class Snippets(object):
         of the snippet stored there.
         """
         self.file.seek(pos[0], 0)
         of the snippet stored there.
         """
         self.file.seek(pos[0], 0)

-        txt = self.file.read(pos[1]).decode('utf-8')
+        try:
+            txt = self.file.read(pos[1]).decode('utf-8')
+        except:
+            return ''

         return txt
 
     def close(self):
         return txt
 
     def close(self):


@@ -142,7 +145,8 @@ class Index(SolrIndex):
                     uids.add(res['uid'])
                 st += rows
         if uids:
                     uids.add(res['uid'])
                 st += rows
         if uids:

-            self.index.delete(uids)
+            # FIXME: With Solr API change, this doesn't work.
+            #self.index.delete(uids)

             return True
         else:
             return False
             return True
         else:
             return False


@@ -389,16 +393,16 @@ class Index(SolrIndex):
             return
 
         def fix_format(text):
             return
 
         def fix_format(text):

-            # separator = [u" ", u"\t", u".", u";", u","]
+            # separator = [" ", "\t", ".", ";", ","]

             if isinstance(text, list):
                 # need to join it first
                 text = filter(lambda s: s is not None, content)
             if isinstance(text, list):
                 # need to join it first
                 text = filter(lambda s: s is not None, content)

-                text = u' '.join(text)
+                text = ' '.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:
                 # 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.insert(i, " ")

 
             return re.sub("(?m)/$", "", text)
 
 
             return re.sub("(?m)/$", "", text)
 


@@ -462,7 +466,7 @@ class Index(SolrIndex):
                     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,
                     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),
+                                       text=''.join(footnote),

                                        is_footnote=True)
                         self.index.add(doc)
                         footnote = []
                                        is_footnote=True)
                         self.index.add(doc)
                         footnote = []


@@ -612,7 +616,7 @@ class SearchResult(object):
         return result
 
     def __str__(self):
         return result
 
     def __str__(self):

-        return u"<SR id=%d %d(%d) hits score=%f %d snippets>" % \
+        return "<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))
             (self.book_id, len(self._hits),
              len(self._processed_hits) if self._processed_hits else -1,
              self._score, len(self.snippets))


@@ -634,7 +638,10 @@ class SearchResult(object):
     def get_book(self):
         if self._book is not None:
             return self._book
     def get_book(self):
         if self._book is not None:
             return self._book

-        self._book = catalogue.models.Book.objects.get(id=self.book_id)
+        try:
+            self._book = catalogue.models.Book.objects.get(id=self.book_id, findable=True)
+        except catalogue.models.Book.DoesNotExist:
+            self._book = None

         return self._book
 
     book = property(get_book)
         return self._book
 
     book = property(get_book)


@@ -661,20 +668,18 @@ class SearchResult(object):
             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)
 
             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):
+        def remove_duplicates(lst, keyfn, larger):

             els = {}
             for e in lst:
                 eif = keyfn(e)
                 if eif in els:
             els = {}
             for e in lst:
                 eif = keyfn(e)
                 if eif in els:

-                    if compare(els[eif], e) >= 1:
+                    if larger(els[eif], e):

                         continue
                 els[eif] = e
             return els.values()
 
         # remove fragments with duplicated fid's and duplicated snippets
                         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]))
+        frags = remove_duplicates(frags, lambda f: f[self.FRAGMENT], lambda a, b: a[self.SCORE] > b[self.SCORE])

 
         # remove duplicate sections
         sections = {}
 
         # remove duplicate sections
         sections = {}


@@ -713,12 +718,12 @@ class SearchResult(object):
                             break
 
             def theme_by_name(n):
                             break
 
             def theme_by_name(n):

-                th = filter(lambda t: t.name == n, themes)
+                th = list(filter(lambda t: t.name == n, themes))

                 if th:
                     return th[0]
                 else:
                     return None
                 if th:
                     return th[0]
                 else:
                     return None

-            themes_hit = filter(lambda a: a is not None, map(theme_by_name, themes_hit))
+            themes_hit = list(filter(lambda a: a is not None, map(theme_by_name, themes_hit)))

 
             m = {'score': f[self.SCORE],
                  'fragment': frag,
 
             m = {'score': f[self.SCORE],
                  'fragment': frag,


@@ -746,13 +751,17 @@ class SearchResult(object):
                     books[r.book_id] = r
         return books.values()
 
                     books[r.book_id] = r
         return books.values()
 

+    def get_sort_key(self):
+        return (-self.score,
+                self.published_date,
+                self.book.sort_key_author if self.book else '',
+                self.book.sort_key if self.book else '')
+

     def __lt__(self, other):
     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)
+        return self.get_sort_key() > other.get_sort_key()

 
     def __eq__(self, other):
 
     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)
+        return self.get_sort_key() == other.get_sort_key()

 
     def __len__(self):
         return len(self.hits)
 
     def __len__(self):
         return len(self.hits)


@@ -794,7 +803,7 @@ class PictureResult(object):
             self._hits.append(hit)
 
     def __str__(self):
             self._hits.append(hit)
 
     def __str__(self):

-        return u"<PR id=%d score=%f >" % (self.picture_id, self._score)
+        return "<PR id=%d score=%f >" % (self.picture_id, self._score)

 
     def __repr__(self):
         return str(self)
 
     def __repr__(self):
         return str(self)


@@ -897,7 +906,7 @@ class Search(SolrIndex):
 
     def search_by_author(self, words):
         from catalogue.models import Book
 
     def search_by_author(self, words):
         from catalogue.models import Book

-        books = Book.objects.filter(parent=None).order_by('-popularity__count')
+        books = Book.objects.filter(parent=None, findable=True).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]]
         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]]


@@ -971,7 +980,7 @@ class Search(SolrIndex):
                 idx += 1
 
         except IOError as e:
                 idx += 1
 
         except IOError as e:

-            book = catalogue.models.Book.objects.filter(id=book_id)
+            book = catalogue.models.Book.objects.filter(id=book_id, findable=True)

             if not book:
                 log.error("Book does not exist for book id = %d" % book_id)
             elif not book.get().children.exists():
             if not book:
                 log.error("Book does not exist for book id = %d" % book_id)
             elif not book.get().children.exists():