searxng/searx/utils.py

438 lines
13 KiB
Python
Raw Normal View History

# -*- coding: utf-8 -*-
2015-01-11 12:26:40 +00:00
import csv
2017-07-20 13:44:02 +00:00
import hashlib
import hmac
2015-01-11 12:26:40 +00:00
import os
import re
from babel.core import get_global
from babel.dates import format_date
2014-01-20 01:31:20 +00:00
from codecs import getincrementalencoder
from imp import load_source
from numbers import Number
from os.path import splitext, join
from io import open
2014-04-24 23:46:40 +00:00
from random import choice
import sys
import json
2014-04-24 23:46:40 +00:00
from searx import settings
2014-11-18 10:37:42 +00:00
from searx.version import VERSION_STRING
2016-09-06 14:43:48 +00:00
from searx.languages import language_codes
2014-11-18 10:37:42 +00:00
from searx import settings
2015-01-11 12:26:40 +00:00
from searx import logger
2014-11-18 10:37:42 +00:00
2016-11-30 17:43:03 +00:00
try:
from cStringIO import StringIO
except:
from io import StringIO
try:
from HTMLParser import HTMLParser
except:
from html.parser import HTMLParser
if sys.version_info[0] == 3:
unichr = chr
unicode = str
2017-11-21 15:49:28 +00:00
IS_PY2 = False
2018-06-14 09:48:31 +00:00
basestring = str
2017-11-21 15:49:28 +00:00
else:
IS_PY2 = True
2015-01-11 12:26:40 +00:00
logger = logger.getChild('utils')
2014-01-10 22:38:08 +00:00
blocked_tags = ('script',
'style')
ecma_unescape4_re = re.compile(r'%u([0-9a-fA-F]{4})', re.UNICODE)
ecma_unescape2_re = re.compile(r'%([0-9a-fA-F]{2})', re.UNICODE)
useragents = json.loads(open(os.path.dirname(os.path.realpath(__file__))
+ "/data/useragents.json", 'r', encoding='utf-8').read())
2014-01-12 19:13:14 +00:00
lang_to_lc_cache = dict()
2014-10-17 10:34:51 +00:00
def searx_useragent():
return 'searx/{searx_version} {suffix}'.format(
searx_version=VERSION_STRING,
suffix=settings['outgoing'].get('useragent_suffix', ''))
2014-10-19 10:41:04 +00:00
def gen_useragent(os=None):
return str(useragents['ua'].format(os=os or choice(useragents['os']), version=choice(useragents['versions'])))
2014-01-10 22:38:08 +00:00
def highlight_content(content, query):
if not content:
return None
# ignoring html contents
# TODO better html content detection
if content.find('<') != -1:
return content
query = query.decode('utf-8')
if content.lower().find(query.lower()) > -1:
query_regex = u'({0})'.format(re.escape(query))
2014-05-16 14:51:23 +00:00
content = re.sub(query_regex, '<span class="highlight">\\1</span>',
content, flags=re.I | re.U)
2014-01-10 22:38:08 +00:00
else:
regex_parts = []
for chunk in query.split():
if len(chunk) == 1:
2016-07-11 13:29:47 +00:00
regex_parts.append(u'\\W+{0}\\W+'.format(re.escape(chunk)))
2014-01-10 22:38:08 +00:00
else:
regex_parts.append(u'{0}'.format(re.escape(chunk)))
query_regex = u'({0})'.format('|'.join(regex_parts))
2014-05-16 14:51:23 +00:00
content = re.sub(query_regex, '<span class="highlight">\\1</span>',
content, flags=re.I | re.U)
2014-01-10 22:38:08 +00:00
return content
2013-11-08 22:44:26 +00:00
2013-11-08 22:44:26 +00:00
class HTMLTextExtractor(HTMLParser):
2013-11-08 22:44:26 +00:00
def __init__(self):
HTMLParser.__init__(self)
self.result = []
self.tags = []
def handle_starttag(self, tag, attrs):
self.tags.append(tag)
def handle_endtag(self, tag):
if not self.tags:
return
if tag != self.tags[-1]:
raise Exception("invalid html")
self.tags.pop()
def is_valid_tag(self):
return not self.tags or self.tags[-1] not in blocked_tags
2013-11-08 22:44:26 +00:00
def handle_data(self, d):
if not self.is_valid_tag():
return
2013-11-08 22:44:26 +00:00
self.result.append(d)
def handle_charref(self, number):
if not self.is_valid_tag():
return
2014-01-20 01:31:20 +00:00
if number[0] in (u'x', u'X'):
codepoint = int(number[1:], 16)
else:
codepoint = int(number)
2013-11-08 22:44:26 +00:00
self.result.append(unichr(codepoint))
def handle_entityref(self, name):
if not self.is_valid_tag():
return
2014-10-19 10:41:04 +00:00
# codepoint = htmlentitydefs.name2codepoint[name]
# self.result.append(unichr(codepoint))
2013-11-18 15:47:20 +00:00
self.result.append(name)
2013-11-08 22:44:26 +00:00
def get_text(self):
2015-01-27 19:03:33 +00:00
return u''.join(self.result).strip()
2013-11-08 22:44:26 +00:00
2013-11-08 22:44:26 +00:00
def html_to_text(html):
html = html.replace('\n', ' ')
html = ' '.join(html.split())
2013-11-08 22:44:26 +00:00
s = HTMLTextExtractor()
s.feed(html)
return s.get_text()
2013-11-15 17:55:18 +00:00
class UnicodeWriter:
"""
A CSV writer which will write rows to CSV file "f",
which is encoded in the given encoding.
"""
def __init__(self, f, dialect=csv.excel, encoding="utf-8", **kwds):
# Redirect output to a queue
2016-11-30 17:43:03 +00:00
self.queue = StringIO()
2013-11-15 17:55:18 +00:00
self.writer = csv.writer(self.queue, dialect=dialect, **kwds)
self.stream = f
2014-01-20 01:31:20 +00:00
self.encoder = getincrementalencoder(encoding)()
2013-11-15 17:55:18 +00:00
def writerow(self, row):
2017-11-21 15:49:28 +00:00
if IS_PY2:
row = [s.encode("utf-8") if hasattr(s, 'encode') else s for s in row]
self.writer.writerow(row)
2013-11-15 17:55:18 +00:00
# Fetch UTF-8 output from the queue ...
2017-11-21 15:49:28 +00:00
data = self.queue.getvalue()
if IS_PY2:
data = data.decode("utf-8")
else:
data = data.strip('\x00')
2013-11-15 17:55:18 +00:00
# ... and reencode it into the target encoding
data = self.encoder.encode(data)
# write to the target stream
2017-11-21 15:49:28 +00:00
if IS_PY2:
self.stream.write(data)
else:
self.stream.write(data.decode("utf-8"))
2013-11-15 17:55:18 +00:00
# empty queue
self.queue.truncate(0)
def writerows(self, rows):
for row in rows:
self.writerow(row)
2014-04-24 23:46:40 +00:00
def get_resources_directory(searx_directory, subdirectory, resources_directory):
if not resources_directory:
resources_directory = os.path.join(searx_directory, subdirectory)
if not os.path.isdir(resources_directory):
raise Exception(resources_directory + " is not a directory")
return resources_directory
2014-04-24 23:46:40 +00:00
def get_themes(templates_path):
"""Returns available themes list."""
themes = os.listdir(templates_path)
if '__common__' in themes:
themes.remove('__common__')
return themes
2015-01-01 16:48:12 +00:00
def get_static_files(static_path):
2015-01-01 16:48:12 +00:00
static_files = set()
static_path_length = len(static_path) + 1
for directory, _, files in os.walk(static_path):
2015-01-01 16:48:12 +00:00
for filename in files:
f = os.path.join(directory[static_path_length:], filename)
2015-01-01 16:48:12 +00:00
static_files.add(f)
return static_files
2015-01-01 17:59:53 +00:00
def get_result_templates(templates_path):
2015-01-01 17:59:53 +00:00
result_templates = set()
templates_path_length = len(templates_path) + 1
for directory, _, files in os.walk(templates_path):
2015-01-01 17:59:53 +00:00
if directory.endswith('result_templates'):
for filename in files:
f = os.path.join(directory[templates_path_length:], filename)
2015-01-01 17:59:53 +00:00
result_templates.add(f)
return result_templates
2015-01-11 12:26:40 +00:00
def format_date_by_locale(date, locale_string):
2015-01-11 12:26:40 +00:00
# strftime works only on dates after 1900
if date.year <= 1900:
return date.isoformat().split('T')[0]
if locale_string == 'all':
locale_string = settings['ui']['default_locale'] or 'en_US'
# to avoid crashing if locale is not supported by babel
try:
formatted_date = format_date(date, locale=locale_string)
except:
formatted_date = format_date(date, "YYYY-MM-dd")
return formatted_date
def dict_subset(d, properties):
result = {}
for k in properties:
if k in d:
result[k] = d[k]
return result
2015-01-29 18:44:52 +00:00
2015-09-07 17:22:01 +00:00
def prettify_url(url, max_length=74):
if len(url) > max_length:
2016-11-30 17:43:03 +00:00
chunk_len = int(max_length / 2 + 1)
2015-09-07 17:22:01 +00:00
return u'{0}[...]{1}'.format(url[:chunk_len], url[-chunk_len:])
2015-01-29 18:44:52 +00:00
else:
return url
# get element in list or default value
def list_get(a_list, index, default=None):
if len(a_list) > index:
return a_list[index]
else:
return default
def get_torrent_size(filesize, filesize_multiplier):
try:
filesize = float(filesize)
if filesize_multiplier == 'TB':
filesize = int(filesize * 1024 * 1024 * 1024 * 1024)
elif filesize_multiplier == 'GB':
filesize = int(filesize * 1024 * 1024 * 1024)
elif filesize_multiplier == 'MB':
filesize = int(filesize * 1024 * 1024)
elif filesize_multiplier == 'KB':
filesize = int(filesize * 1024)
2016-10-11 17:31:42 +00:00
elif filesize_multiplier == 'TiB':
filesize = int(filesize * 1000 * 1000 * 1000 * 1000)
elif filesize_multiplier == 'GiB':
filesize = int(filesize * 1000 * 1000 * 1000)
elif filesize_multiplier == 'MiB':
filesize = int(filesize * 1000 * 1000)
elif filesize_multiplier == 'KiB':
filesize = int(filesize * 1000)
except:
filesize = None
return filesize
2016-09-06 14:43:48 +00:00
2016-10-11 17:31:42 +00:00
def convert_str_to_int(number_str):
if number_str.isdigit():
return int(number_str)
else:
return 0
# convert a variable to integer or return 0 if it's not a number
def int_or_zero(num):
if isinstance(num, list):
if len(num) < 1:
return 0
num = num[0]
return convert_str_to_int(num)
2016-09-06 14:43:48 +00:00
def is_valid_lang(lang):
is_abbr = (len(lang) == 2)
lang = lang.lower().decode('utf-8')
2016-09-06 14:43:48 +00:00
if is_abbr:
for l in language_codes:
if l[0][:2] == lang:
return (True, l[0][:2], l[3].lower())
2016-09-06 14:43:48 +00:00
return False
else:
for l in language_codes:
if l[1].lower() == lang or l[3].lower() == lang:
return (True, l[0][:2], l[3].lower())
2016-09-06 14:43:48 +00:00
return False
def _get_lang_to_lc_dict(lang_list):
key = str(lang_list)
value = lang_to_lc_cache.get(key, None)
if value is None:
value = dict()
for lc in lang_list:
value.setdefault(lc.split('-')[0], lc)
lang_to_lc_cache[key] = value
return value
# auxiliary function to match lang_code in lang_list
def _match_language(lang_code, lang_list=[], custom_aliases={}):
# replace language code with a custom alias if necessary
if lang_code in custom_aliases:
lang_code = custom_aliases[lang_code]
if lang_code in lang_list:
return lang_code
# try to get the most likely country for this language
subtags = get_global('likely_subtags').get(lang_code)
if subtags:
subtag_parts = subtags.split('_')
new_code = subtag_parts[0] + '-' + subtag_parts[-1]
if new_code in custom_aliases:
new_code = custom_aliases[new_code]
if new_code in lang_list:
return new_code
# try to get the any supported country for this language
return _get_lang_to_lc_dict(lang_list).get(lang_code, None)
# get the language code from lang_list that best matches locale_code
def match_language(locale_code, lang_list=[], custom_aliases={}, fallback='en-US'):
# try to get language from given locale_code
language = _match_language(locale_code, lang_list, custom_aliases)
if language:
return language
locale_parts = locale_code.split('-')
lang_code = locale_parts[0]
# try to get language using an equivalent country code
if len(locale_parts) > 1:
country_alias = get_global('territory_aliases').get(locale_parts[-1])
if country_alias:
language = _match_language(lang_code + '-' + country_alias[0], lang_list, custom_aliases)
if language:
return language
# try to get language using an equivalent language code
alias = get_global('language_aliases').get(lang_code)
if alias:
language = _match_language(alias, lang_list, custom_aliases)
if language:
return language
if lang_code != locale_code:
# try to get language from given language without giving the country
language = _match_language(lang_code, lang_list, custom_aliases)
return language or fallback
def load_module(filename, module_dir):
modname = splitext(filename)[0]
if modname in sys.modules:
del sys.modules[modname]
filepath = join(module_dir, filename)
module = load_source(modname, filepath)
module.name = modname
return module
2017-07-20 13:44:02 +00:00
def new_hmac(secret_key, url):
try:
secret_key_bytes = bytes(secret_key, 'utf-8')
except TypeError as err:
if isinstance(secret_key, bytes):
secret_key_bytes = secret_key
else:
raise err
2017-07-20 13:44:02 +00:00
if sys.version_info[0] == 2:
return hmac.new(bytes(secret_key), url, hashlib.sha256).hexdigest()
else:
return hmac.new(secret_key_bytes, url, hashlib.sha256).hexdigest()
def to_string(obj):
if isinstance(obj, basestring):
return obj
if isinstance(obj, Number):
return unicode(obj)
if hasattr(obj, '__str__'):
return obj.__str__()
if hasattr(obj, '__repr__'):
return obj.__repr__()
def ecma_unescape(s):
"""
python implementation of the unescape javascript function
https://www.ecma-international.org/ecma-262/6.0/#sec-unescape-string
https://developer.mozilla.org/fr/docs/Web/JavaScript/Reference/Objets_globaux/unescape
"""
# s = unicode(s)
# "%u5409" becomes "吉"
s = ecma_unescape4_re.sub(lambda e: unichr(int(e.group(1), 16)), s)
# "%20" becomes " ", "%F3" becomes "ó"
s = ecma_unescape2_re.sub(lambda e: unichr(int(e.group(1), 16)), s)
return s