catalogue style change

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

diff --git a/apps/search/views.py b/apps/search/views.py
index dc90841..00391f1 100644 (file)
--- a/apps/search/views.py
+++ b/apps/search/views.py
@@ -1,27 +1,40 @@
+# -*- coding: utf-8 -*-

 
 

+from django.conf import settings

 from django.shortcuts import render_to_response, get_object_or_404
 from django.template import RequestContext
 from django.contrib.auth.decorators import login_required
 from django.views.decorators import cache
 from django.shortcuts import render_to_response, get_object_or_404
 from django.template import RequestContext
 from django.contrib.auth.decorators import login_required
 from django.views.decorators import cache

+from django.http import HttpResponse, HttpResponseRedirect, Http404, HttpResponsePermanentRedirect
+from django.utils.translation import ugettext as _

 
 from catalogue.utils import get_random_hash
 
 from catalogue.utils import get_random_hash

-from catalogue.models import Book, Tag, TAG_CATEGORIES
+from catalogue.models import Book, Tag, Fragment

 from catalogue.fields import dumps
 from catalogue.views import JSONResponse
 from catalogue.fields import dumps
 from catalogue.views import JSONResponse

-from catalogue import forms
-from search import MultiSearch, JVM, SearchResult
+from search import Search, JVM, SearchResult

 from lucene import StringReader
 from lucene import StringReader

+from suggest.forms import PublishingSuggestForm

 
 import enchant
 
 dictionary = enchant.Dict('pl_PL')
 
 
 
 import enchant
 
 dictionary = enchant.Dict('pl_PL')
 
 

+def match_word_re(word):
+    if 'sqlite' in settings.DATABASES['default']['ENGINE']:
+        return r"\b%s\b" % word
+    elif 'mysql' in settings.DATABASES['default']['ENGINE']:
+        return "[[:<:]]%s[[:>:]]" % word
+
+

 def did_you_mean(query, tokens):
     change = {}
 def did_you_mean(query, tokens):
     change = {}

-    

     for t in tokens:
     for t in tokens:

-        print("%s ok? %s, sug: %s" %(t, dictionary.check(t), dictionary.suggest(t)))
+        authors = Tag.objects.filter(category='author', name__iregex=match_word_re(t))
+        if len(authors) > 0:
+            continue
+        

         if not dictionary.check(t):
             try:
                 change[t] = dictionary.suggest(t)[0]
         if not dictionary.check(t):
             try:
                 change[t] = dictionary.suggest(t)[0]


@@ -37,30 +50,38 @@ def did_you_mean(query, tokens):
     return query
 
 
     return query
 
 

-def category_name(category):
-    try:
-        return filter(lambda c: c[0] == category, TAG_CATEGORIES)[0][1].encode('utf-8')
-    except IndexError:
-        raise KeyError("No category %s" % category)
-
-

 def hint(request):
     prefix = request.GET.get('term', '')
     if len(prefix) < 2:
 def hint(request):
     prefix = request.GET.get('term', '')
     if len(prefix) < 2:

-        return JSONResponse(dumps({}))
+        return JSONResponse([])

     JVM.attachCurrentThread()
     JVM.attachCurrentThread()

-    s = MultiSearch()
+    s = Search()
+
+    hint = s.hint()
+    try:
+        tags = request.GET.get('tags', '')
+        hint.tags(Tag.get_tag_list(tags))
+    except:
+        pass
+
+    # tagi beda ograniczac tutaj
+    # ale tagi moga byc na ksiazce i na fragmentach
+    # jezeli tagi dot tylko ksiazki, to wazne zeby te nowe byly w tej samej ksiazce
+    # jesli zas dotycza themes, to wazne, zeby byly w tym samym fragmencie.
+

     tags = s.hint_tags(prefix)
     books = s.hint_books(prefix)
 
     tags = s.hint_tags(prefix)
     books = s.hint_books(prefix)
 

+    # TODO DODAC TU HINTY
+

     return JSONResponse(
         [{'label': t.name,
     return JSONResponse(
         [{'label': t.name,

-          'category': category_name(t.category),
+          'category': _(t.category),

           'id': t.id,
           'url': t.get_absolute_url()}
           for t in tags] + \
           [{'label': b.title,
           'id': t.id,
           'url': t.get_absolute_url()}
           for t in tags] + \
           [{'label': b.title,

-            'category': category_name('book'),
+            'category': _('book'),

             'id': b.id,
             'url': b.get_absolute_url()}
             for b in books])
             'id': b.id,
             'url': b.get_absolute_url()}
             for b in books])


@@ -69,7 +90,7 @@ def hint(request):
 def main(request):
     results = {}
     JVM.attachCurrentThread()  # where to put this?
 def main(request):
     results = {}
     JVM.attachCurrentThread()  # where to put this?

-    srch = MultiSearch()
+    srch = Search()

 
     results = None
     query = None
 
     results = None
     query = None


@@ -77,30 +98,80 @@ def main(request):
 
     if 'q' in request.GET:
         tags = request.GET.get('tags', '')
 
     if 'q' in request.GET:
         tags = request.GET.get('tags', '')

+        query = request.GET['q']
+        book_id = request.GET.get('book', None)
+        book = None
+        if book_id is not None:
+            book = get_object_or_404(Book, id=book_id)
+
+        hint = srch.hint()

         try:
             tag_list = Tag.get_tag_list(tags)
         except:
             tag_list = []
 
         try:
             tag_list = Tag.get_tag_list(tags)
         except:
             tag_list = []
 

-            #        tag_filter = srch.
+        if len(query) < 2:
+            return render_to_response('catalogue/search_too_short.html', {'tags': tag_list, 'prefix': query},
+                                      context_instance=RequestContext(request))
+
+        hint.tags(tag_list)
+        if book:
+            hint.books(book)

 
 

-        query = request.GET['q']

         toks = StringReader(query)
         toks = StringReader(query)

+        tokens_cache = {}

         fuzzy = 'fuzzy' in request.GET
         if fuzzy:
             fuzzy = 0.7
 
         fuzzy = 'fuzzy' in request.GET
         if fuzzy:
             fuzzy = 0.7
 

-
-        results = SearchResult.aggregate(srch.search_perfect_book(toks, fuzzy=fuzzy),
-                                         srch.search_perfect_parts(toks, fuzzy=fuzzy),
-                                         srch.search_everywhere(toks, fuzzy=fuzzy))
+        author_results = srch.search_phrase(toks, 'authors', fuzzy=fuzzy, tokens_cache=tokens_cache)
+        title_results = srch.search_phrase(toks, 'title', fuzzy=fuzzy, tokens_cache=tokens_cache)
+
+        # Boost main author/title results with mixed search, and save some of its results for end of list.
+        # boost author, title results
+        author_title_mixed = srch.search_some(toks, ['authors', 'title', 'tags'], fuzzy=fuzzy, tokens_cache=tokens_cache)
+        author_title_rest = []
+        for b in author_title_mixed:
+            bks = filter(lambda ba: ba.book_id == b.book_id, author_results + title_results)
+            for b2 in bks:
+                b2.boost *= 1.1
+            if bks is []:
+                author_title_rest.append(b)
+        
+        text_phrase = SearchResult.aggregate(srch.search_phrase(toks, 'content', fuzzy=fuzzy, tokens_cache=tokens_cache, snippets=True, book=False))
+        
+        everywhere = SearchResult.aggregate(srch.search_everywhere(toks, fuzzy=fuzzy, tokens_cache=tokens_cache), author_title_rest)
+
+        for res in [author_results, title_results, text_phrase, everywhere]:
+            res.sort(reverse=True)
+
+        suggestion = did_you_mean(query, srch.get_tokens(toks, field="SIMPLE"))
+
+        results = author_results + title_results + text_phrase + everywhere

         results.sort(reverse=True)
         results.sort(reverse=True)

-
-        for r in results:
-            print r.parts
-
-    return render_to_response('newsearch/search.html', {'results': results,
-                                                        'did_you_mean': (query is not None) and 
-                                                        did_you_mean(query, srch.get_tokens(query, field='SIMPLE')),
-                                                        'fuzzy': fuzzy},
-                              context_instance=RequestContext(request))
+        
+        if len(results) == 1:
+            fragment_hits = filter(lambda h: 'fragment' in h, results[0].hits)
+            if len(fragment_hits) == 1:
+                anchor = fragment_hits[0]['fragment']
+                frag = Fragment.objects.get(anchor=anchor)
+                return HttpResponseRedirect(frag.get_absolute_url())
+            return HttpResponseRedirect(results[0].book.get_absolute_url())
+        elif len(results) == 0:
+            form = PublishingSuggestForm(initial={"books": query + ", "})
+            return render_to_response('catalogue/search_no_hits.html',
+                                      {'tags': tag_list,
+                                       'prefix': query,
+                                       "form": form,
+                                       'did_you_mean': suggestion},
+                context_instance=RequestContext(request))
+
+        return render_to_response('catalogue/search_multiple_hits.html',
+                                  {'tags': tag_list,
+                                   'prefix': query,
+                                   'results': { 'author': author_results,
+                                                'title': title_results,
+                                                'content': text_phrase,
+                                                'other': everywhere},
+                                   'did_you_mean': suggestion},
+            context_instance=RequestContext(request))