Python Search.search_everywhere方法代码示例

# 需要导入模块: from search import Search [as 别名]
# 或者: from search.Search import search_everywhere [as 别名]
def main(request):
    results = {}

    results = None
    query = None

    query = request.GET.get('q', '')

    if len(query) < 2:
        return render_to_response('catalogue/search_too_short.html',
                                  {'prefix': query},
            context_instance=RequestContext(request))

    query = remove_query_syntax_chars(query)
    
    search = Search()

    theme_terms = search.index.analyze(text=query, field="themes_pl") \
        + search.index.analyze(text=query, field="themes")

            # change hints
    tags = search.hint_tags(query, pdcounter=True, prefix=False)
    tags = split_tags(tags)

    author_results = search.search_phrase(query, 'authors', book=True)
    translator_results = search.search_phrase(query, 'translators', book=True)

    title_results = search.search_phrase(query, 'title', book=True)

    # 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 = search.search_some(query, ['authors', 'translators', 'title', 'tags'], query_terms=theme_terms)
    author_title_rest = []

    for b in author_title_mixed:
        also_in_mixed = filter(lambda ba: ba.book_id == b.book_id, author_results + translator_results + title_results)
        for b2 in also_in_mixed:
            b2.boost *= 1.1
        if also_in_mixed is []:
            author_title_rest.append(b)

    # Do a phrase search but a term search as well - this can give us better snippets then search_everywhere,
    # Because the query is using only one field.
    text_phrase = SearchResult.aggregate(
        search.search_phrase(query, 'text', snippets=True, book=False),
        search.search_some(query, ['text'], snippets=True, book=False, query_terms=theme_terms))

    everywhere = search.search_everywhere(query, query_terms=theme_terms)

    def already_found(results):
        def f(e):
            for r in results:
                if e.book_id == r.book_id:
                    e.boost = 0.9
                    results.append(e)
                    return True
            return False
        return f
    f = already_found(author_results + translator_results + title_results + text_phrase)
    everywhere = filter(lambda x: not f(x), everywhere)

    author_results = SearchResult.aggregate(author_results)
    translator_results = SearchResult.aggregate(translator_results)
    title_results = SearchResult.aggregate(title_results)

    everywhere = SearchResult.aggregate(everywhere, author_title_rest)

    for field, res in [('authors', author_results),
                       ('translators', translator_results),
                       ('title', title_results),
                       ('text', text_phrase),
                       ('text', everywhere)]:
        res.sort(reverse=True)
        for r in res:
            search.get_snippets(r, query, field, 3)

    suggestion = u''

    def ensure_exists(r):
        try:
            return r.book
        except Book.DoesNotExist:
            return False

    author_results = filter(ensure_exists, author_results)
    translator_results = filter(ensure_exists, translator_results)
    title_results = filter(ensure_exists, title_results)
    text_phrase = filter(ensure_exists, text_phrase)
    everywhere = filter(ensure_exists, everywhere)

    results = author_results + translator_results + title_results + text_phrase + everywhere
    # ensure books do exists & sort them
    for res in (author_results, translator_results, title_results, text_phrase, everywhere):
        res.sort(reverse=True)

    # We don't want to redirect to book text, but rather display result page even with one result.
    # 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']
#.........这里部分代码省略.........

# 需要导入模块: from search import Search [as 别名]
# 或者: from search.Search import search_everywhere [as 别名]
class BookSearchTests(WLTestCase):
    def setUp(self):
        JVM.attachCurrentThread()
        WLTestCase.setUp(self)
        settings.NO_SEARCH_INDEX = False
        settings.SEARCH_INDEX = path.join(settings.MEDIA_ROOT, 'search')

        txt = path.join(path.dirname(__file__), 'files/fraszka-do-anusie.xml')
        self.book = models.Book.from_xml_file(txt)

        index = Index()
        index.open()
        try:
            index.index_book(self.book)
        except:
            index.close()

        self.search = Search()

    def test_search_perfect_book_author(self):
        books = self.search.search_perfect_book("sęp szarzyński")
        assert len(books) == 1
        assert books[0].book_id == self.book.id

    def test_search_perfect_book_title(self):
        books = self.search.search_perfect_book("fraszka anusie")
        assert len(books) == 1
        assert books[0].book_id == self.book.id

    def test_search_perfect_parts(self):
        books = self.search.search_perfect_parts("Jakoż hamować")
        assert len(books) == 2
        for b in books:
            b.book_id == self.book.id
        a = SearchResult.aggregate(books)
        # just one fragment hit.
        assert len(filter(lambda x: x[1], a[0].hits)) == 1
        print a[0].process_hits()

    def test_search_perfect_author_title(self):
        books = self.search.search_perfect_book("szarzyński anusie")
        assert books == []

        books = self.search.search_book("szarzyński anusie")
        assert len(books) == 1

        books = self.search.search_book("szarzyński fraszka")
        assert len(books) == 1

    def test_search_everywhere(self):
        books = self.search.search_everywhere("szarzyński kochanek")
        print 'szarzyński kochanek %s' % [b.hits for b in books]

        books = self.search.search_everywhere("szarzyński narcyz")
        print 'szarzyński narcyz %s' % [b.hits for b in books]

        books = self.search.search_everywhere("anusie narcyz")
        print 'anusie narcyz %s' % [b.hits for b in books]

        # theme content cross
        books = self.search.search_everywhere("wzrok  boginie")
        print 'wzrok boginie %s' % [b.hits for b in books]

        books = self.search.search_everywhere("anusie płynęły zdroje")
        print 'anusie płynęły zdroje %s' % [b.hits for b in books]