def __init__(self, books_folder, pages_index_folder, csv_path):
self.books_folder = books_folder
self.pages_index_folder = pages_index_folder
self.csv_path = csv_path
self.pages_search_engine = TSearchEngine(index_location=self.pages_index_folder)
self.cfields_search_engine = TCustomFieldsSearchEngine(csv_path)
class TSearchServer():
def __init__(self, books_folder, pages_index_folder, csv_path):
self.books_folder = books_folder
self.pages_index_folder = pages_index_folder
self.csv_path = csv_path
self.pages_search_engine = TSearchEngine(index_location=self.pages_index_folder)
self.cfields_search_engine = TCustomFieldsSearchEngine(csv_path)
def get_pages_segment_data(self, segment_id):
obj_id, field_id, start, length = self.pages_search_engine.segment_index.get_segment(segment_id)
import os
location = os.path.join(self.books_folder, obj_id, field_id)
f = open(location, "rb")
f.seek(start)
snippet = f.read(length)
return obj_id, field_id, snippet
def select_words_in_snippet(self, words2select, snippet):
matches = self.pages_search_engine.parsers.parse_buffer(snippet, "windows-1251")
to_select = []
for token, position in words2select:
to_select += [(matches[position].start, matches[position].start + matches[position].length)]
to_select.sort()
for sel_index in xrange(len(to_select) - 1, -1, -1):
sel_start, sel_end = to_select[sel_index]
snippet = snippet[:sel_start] + "<b>" + snippet[sel_start:sel_end] + "</b>" + snippet[sel_end:]
snippet = snippet.decode("windows-1251").replace(chr(13), " ").replace(chr(10), " ").replace('"', "'")
import re
snippet = re.subn("\s+", " ", snippet)[0]
return snippet
""" returns object of class TBook"""
def get_book_data(self, object_id):
return self.cfields_search_engine.objects[int(object_id)]
def search(self, params):
EMPTY_RESPONSE = [[], 0]
filtered_object_ids = self.cfields_search_engine.process_query(title=params["title"],
author=params["author"],
udc=params["udc"],
year=params["year"],
year_max=params["year_max"],
year_min=params["year_min"],
pages_count=params["pages_count"] ,
pages_count_max=params["pages_count_max"],
pages_count_min=params["pages_count_min"],
lib_section=params["filter_lib_section"])
#no books satisfying filters
if filtered_object_ids == -1:
return EMPTY_RESPONSE
if params["filter_object_id"]:
filter_object_id = int(params["filter_object_id"])
if filtered_object_ids == 0: # no restrictions introduced
filtered_object_ids = [filter_object_id]
elif filter_object_id in filtered_object_ids:
filtered_object_ids = [filter_object_id]
else:
return EMPTY_RESPONSE
if filtered_object_ids == 0: #all books are accepted
filtered_object_ids = None
if not filtered_object_ids and not params["pages_query"]: #all accepted
return EMPTY_RESPONSE
if not params["pages_query"]:# no query to the pages index
return EMPTY_RESPONSE
objects_matching_custom_field = self.cfields_search_engine.find_mentions_of_author_and_title(params["pages_query"])
if filtered_object_ids != None:
if len(filtered_object_ids) == 1:#searching inside one book:
objects_matching_custom_field = []
else:
objects_matching_custom_field = [obj_id for obj_id in objects_matching_custom_field \
if obj_id in filtered_object_ids]
first_object2return = params["start"]
objects2return = params["len"]
first_from_custom_matchings = min(len(objects_matching_custom_field), first_object2return)
take_custom_matchings_count = min(objects2return, max(0, len(objects_matching_custom_field) - first_object2return))
objects2return = max(0, objects2return - take_custom_matchings_count)
first_object2return = max(0, first_object2return - len(objects_matching_custom_field))
total_results_count = take_custom_matchings_count
joined_results = []
for obj_id in objects_matching_custom_field[first_from_custom_matchings: first_from_custom_matchings + take_custom_matchings_count]:
search_result = TSearchEngineResult(obj_id, 0, 0)
joined_results += [search_result]
pages_results, pages_results_count = self.pages_search_engine.search(query=params["pages_query"],
filter_objects=filtered_object_ids,
first_object2return=first_object2return,
objects2return=objects2return)
joined_results += pages_results
total_results_count += pages_results_count
return joined_results, total_results_count
