Fix section counting/targetting from search snippet

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

diff --git a/apps/search/index.py b/apps/search/index.py
index 33836ad..6d97047 100644 (file)
--- a/apps/search/index.py
+++ b/apps/search/index.py
@@ -276,7 +276,7 @@ class Index(BaseIndex):
 
     footnote_tags = ['pa', 'pt', 'pr', 'pe']
 
 
     footnote_tags = ['pa', 'pt', 'pr', 'pe']
 

-    skip_header_tags = ['autor_utworu', 'nazwa_utworu', 'dzielo_nadrzedne']
+    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]+)[\]. ]*$")
 
 
     published_date_re = re.compile("([0-9]+)[\]. ]*$")
 


@@ -323,7 +323,6 @@ class Index(BaseIndex):
         # get published date
         source = book_info.source_name
         match = self.published_date_re.search(source)
         # get published date
         source = book_info.source_name
         match = self.published_date_re.search(source)

-        print("published date is %s %s" % (match, match is not None and match.groups()))

         if match is not None:
             fields["published_date"] = Field("published_date", str(match.groups()[0]), Field.Store.YES, Field.Index.NOT_ANALYZED)
 
         if match is not None:
             fields["published_date"] = Field("published_date", str(match.groups()[0]), Field.Store.YES, Field.Index.NOT_ANALYZED)
 


@@ -426,9 +425,8 @@ class Index(BaseIndex):
 
         fragments = {}
         snippets = Snippets(book.id).open('w')
 
         fragments = {}
         snippets = Snippets(book.id).open('w')

-        position = 0

         try:
         try:

-            for header in list(master):
+            for header, position in zip(list(master), range(len(master))):

 
                 if header.tag in self.skip_header_tags:
                     continue
 
                 if header.tag in self.skip_header_tags:
                     continue


@@ -441,15 +439,15 @@ class Index(BaseIndex):
 
                 for start, end in walker(header, ignore_tags=self.ignore_content_tags):
                     # handle footnotes
 
                 for start, end in walker(header, ignore_tags=self.ignore_content_tags):
                     # handle footnotes

-                    if start is not None and start.tag in self.footnote_tags:
-                        footnote = ' '.join(start.itertext())
-                    elif end is not None and footnote is not None and end.tag in self.footnote_tags:
-                        doc = add_part(snippets, header_index=position, header_type=header.tag,
-                                       content=footnote)
+                    # if start is not None and start.tag in self.footnote_tags:
+                    #     footnote = ' '.join(start.itertext())
+                    # elif end is not None and footnote is not None and end.tag in self.footnote_tags:
+                    #     doc = add_part(snippets, header_index=position, header_type=header.tag,
+                    #                    content=footnote)

 
 

-                        self.index.addDocument(doc)
+                    #     self.index.addDocument(doc)

 
 

-                        footnote = None
+                    #     footnote = None

 
                     # handle fragments and themes.
                     if start is not None and start.tag == 'begin':
 
                     # handle fragments and themes.
                     if start is not None and start.tag == 'begin':


@@ -496,7 +494,6 @@ class Index(BaseIndex):
                                content=fix_format(content))
 
                 self.index.addDocument(doc)
                                content=fix_format(content))
 
                 self.index.addDocument(doc)

-                position += 1

 
         finally:
             snippets.close()
 
         finally:
             snippets.close()


@@ -619,7 +616,6 @@ class SearchResult(object):
 
         pd = stored.get("published_date")
         if pd is None:
 
         pd = stored.get("published_date")
         if pd is None:

-            print "published_date is none for book %d" % self.book_id

             pd = 0
         self.published_date = int(pd)
 
             pd = 0
         self.published_date = int(pd)
 


@@ -642,7 +638,7 @@ class SearchResult(object):
             raise ValueError("this search result is or book %d; tried to merge with %d" % (self.book_id, other.book_id))
         self._hits += other._hits
         if other.score > self.score:
             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
+            self._score = other._score

         return self
 
     def get_book(self):
         return self
 
     def get_book(self):


@@ -714,7 +710,6 @@ class SearchResult(object):
                 tokens = self.search.get_tokens(self.searched, 'POLISH', cached=self.tokens_cache)
                 for theme in themes:
                     name_tokens = self.search.get_tokens(theme.name, 'POLISH')
                 tokens = self.search.get_tokens(self.searched, 'POLISH', cached=self.tokens_cache)
                 for theme in themes:
                     name_tokens = self.search.get_tokens(theme.name, 'POLISH')

-                    print "THEME HIT: %s in %s" % (tokens, name_tokens)

                     for t in tokens:
                         if t in name_tokens:
                             if not theme in themes_hit:
                     for t in tokens:
                         if t in name_tokens:
                             if not theme in themes_hit:


@@ -967,13 +962,13 @@ class Search(IndexStore):
         return q
 
     def search_phrase(self, searched, field, book=True, max_results=20, fuzzy=False,
         return q
 
     def search_phrase(self, searched, field, book=True, max_results=20, fuzzy=False,

-                      filters=None, tokens_cache=None, boost=None, snippets=False):
+                      filters=None, tokens_cache=None, boost=None, snippets=False, slop=2):

         if filters is None: filters = []
         if tokens_cache is None: tokens_cache = {}
 
         tokens = self.get_tokens(searched, field, cached=tokens_cache)
 
         if filters is None: filters = []
         if tokens_cache is None: tokens_cache = {}
 
         tokens = self.get_tokens(searched, field, cached=tokens_cache)
 

-        query = self.make_phrase(tokens, field=field, fuzzy=fuzzy)
+        query = self.make_phrase(tokens, field=field, fuzzy=fuzzy, slop=slop)

         if book:
             filters.append(self.term_filter(Term('is_book', 'true')))
         top = self.searcher.search(query, self.chain_filters(filters), max_results)
         if book:
             filters.append(self.term_filter(Term('is_book', 'true')))
         top = self.searcher.search(query, self.chain_filters(filters), max_results)


@@ -981,7 +976,7 @@ class Search(IndexStore):
         return [SearchResult(self, found, snippets=(snippets and self.get_snippets(found, query) or None), searched=searched) for found in top.scoreDocs]
 
     def search_some(self, searched, fields, book=True, max_results=20, fuzzy=False,
         return [SearchResult(self, found, snippets=(snippets and self.get_snippets(found, query) or None), searched=searched) for found in top.scoreDocs]
 
     def search_some(self, searched, fields, book=True, max_results=20, fuzzy=False,

-                    filters=None, tokens_cache=None, boost=None):
+                    filters=None, tokens_cache=None, boost=None, snippets=True):

         if filters is None: filters = []
         if tokens_cache is None: tokens_cache = {}
 
         if filters is None: filters = []
         if tokens_cache is None: tokens_cache = {}
 


@@ -999,7 +994,7 @@ class Search(IndexStore):
         top = self.searcher.search(query, self.chain_filters(filters), max_results)
 
         return [SearchResult(self, found, searched=searched, tokens_cache=tokens_cache,
         top = self.searcher.search(query, self.chain_filters(filters), max_results)
 
         return [SearchResult(self, found, searched=searched, tokens_cache=tokens_cache,

-                             snippets=self.get_snippets(found, query)) for found in top.scoreDocs]
+                             snippets=(snippets and self.get_snippets(found, query) or None)) for found in top.scoreDocs]

 
     def search_perfect_book(self, searched, max_results=20, fuzzy=False, hint=None):
         """
 
     def search_perfect_book(self, searched, max_results=20, fuzzy=False, hint=None):
         """