Files
odoo_source/addons/website/tools.py
T
Benoit Socias c6ba756e4b [IMP] core, website: use fuzzy matching in web-based search
Before this commit searched terms were the ones entered by the end-user.

After this commit if the searched term is not found then the search is
performed on a resembling term.
The similarity is obtained from a Levenshtein distance combined with
ratios of common and different letters.
By default the dictionary against which the search term is matched is
based on the content of the search fields containing a word that starts
with the first letter of the search term.
If the `pg_trgm` Postgresql extension is installed the dictionary is
built from matches using the `<%%` operator (similar to the
`word_similarity` function).
Several approaches were benchmarked during development, those results
are available through the task record.

task-2379555
https://github.com/odoo/odoo/pull/65871

Part-of: odoo/odoo#65871
2021-09-03 06:59:33 +00:00

202 lines
6.7 KiB
Python

# -*- encoding: utf-8 -*-
# Part of Odoo. See LICENSE file for full copyright and licensing details.
import contextlib
import re
from lxml import etree
from psycopg2 import sql
from unittest.mock import Mock, MagicMock, patch
import werkzeug
import odoo
from odoo.tools.misc import DotDict
def get_video_embed_code(video_url):
''' Computes the valid iframe from given URL that can be embedded
(or False in case of invalid URL).
'''
if not video_url:
return False
# To detect if we have a valid URL or not
validURLRegex = r'^(http:\/\/|https:\/\/|\/\/)[a-z0-9]+([\-\.]{1}[a-z0-9]+)*\.[a-z]{2,5}(:[0-9]{1,5})?(\/.*)?$'
# Regex for few of the widely used video hosting services
ytRegex = r'^(?:(?:https?:)?\/\/)?(?:www\.)?(?:youtu\.be\/|youtube(-nocookie)?\.com\/(?:embed\/|v\/|watch\?v=|watch\?.+&v=))((?:\w|-){11})(?:\S+)?$'
vimeoRegex = r'\/\/(player.)?vimeo.com\/([a-z]*\/)*([0-9]{6,11})[?]?.*'
dmRegex = r'.+dailymotion.com\/(video|hub|embed)\/([^_?]+)[^#]*(#video=([^_&]+))?'
igRegex = r'(.*)instagram.com\/p\/(.[a-zA-Z0-9]*)'
ykuRegex = r'(.*).youku\.com\/(v_show\/id_|embed\/)(.+)'
if not re.search(validURLRegex, video_url):
return False
else:
embedUrl = False
ytMatch = re.search(ytRegex, video_url)
vimeoMatch = re.search(vimeoRegex, video_url)
dmMatch = re.search(dmRegex, video_url)
igMatch = re.search(igRegex, video_url)
ykuMatch = re.search(ykuRegex, video_url)
if ytMatch and len(ytMatch.groups()[1]) == 11:
embedUrl = '//www.youtube%s.com/embed/%s?rel=0' % (ytMatch.groups()[0] or '', ytMatch.groups()[1])
elif vimeoMatch:
embedUrl = '//player.vimeo.com/video/%s' % (vimeoMatch.groups()[2])
elif dmMatch:
embedUrl = '//www.dailymotion.com/embed/video/%s' % (dmMatch.groups()[1])
elif igMatch:
embedUrl = '//www.instagram.com/p/%s/embed/' % (igMatch.groups()[1])
elif ykuMatch:
ykuLink = ykuMatch.groups()[2]
if '.html?' in ykuLink:
ykuLink = ykuLink.split('.html?')[0]
embedUrl = '//player.youku.com/embed/%s' % (ykuLink)
else:
# We directly use the provided URL as it is
embedUrl = video_url
return '<iframe class="embed-responsive-item" src="%s" allowFullScreen="true" frameborder="0"></iframe>' % embedUrl
def werkzeugRaiseNotFound(*args, **kwargs):
raise werkzeug.exceptions.NotFound()
@contextlib.contextmanager
def MockRequest(
env, *, routing=True, multilang=True,
context=None,
cookies=None, country_code=None, website=None, sale_order_id=None,
website_sale_current_pl=None,
):
router = MagicMock()
match = router.return_value.bind.return_value.match
if routing:
match.return_value[0].routing = {
'type': 'http',
'website': True,
'multilang': multilang
}
else:
match.side_effect = werkzeugRaiseNotFound
if context is None:
context = {}
lang_code = context.get('lang', env.context.get('lang', 'en_US'))
context.setdefault('lang', lang_code)
request = Mock(
context=context,
db=None,
endpoint=match.return_value[0] if routing else None,
env=env,
httprequest=Mock(
host='localhost',
path='/hello/',
app=odoo.http.root,
environ={'REMOTE_ADDR': '127.0.0.1'},
cookies=cookies or {},
referrer='',
),
lang=env['res.lang']._lang_get(lang_code),
redirect=env['ir.http']._redirect,
session=DotDict(
geoip={'country_code': country_code},
debug=False,
sale_order_id=sale_order_id,
website_sale_current_pl=website_sale_current_pl,
),
website=website
)
with contextlib.ExitStack() as s:
odoo.http._request_stack.push(request)
s.callback(odoo.http._request_stack.pop)
s.enter_context(patch('odoo.http.root.get_db_router', router))
yield request
# Fuzzy matching tools
def distance(s1="", s2="", limit=4):
"""
Limited Levenshtein-ish distance (inspired from Apache text common)
Note: this does not return quick results for simple cases (empty string, equal strings)
those checks should be done outside loops that use this function.
:param s1: first string
:param s2: second string
:param limit: maximum distance to take into account, return -1 if exceeded
:return: number of character changes needed to transform s1 into s2 or -1 if this exceeds the limit
"""
BIG = 100000 # never reached integer
if len(s1) > len(s2):
s1, s2 = s2, s1
l1 = len(s1)
l2 = len(s2)
if l2 - l1 > limit:
return -1
boundary = min(l1, limit) + 1
p = [i if i < boundary else BIG for i in range(0, l1 + 1)]
d = [BIG for _ in range(0, l1 + 1)]
for j in range(1, l2 + 1):
j2 = s2[j -1]
d[0] = j
range_min = max(1, j - limit)
range_max = min(l1, j + limit)
if range_min > 1:
d[range_min -1] = BIG
for i in range(range_min, range_max + 1):
if s1[i - 1] == j2:
d[i] = p[i - 1]
else:
d[i] = 1 + min(d[i - 1], p[i], p[i - 1])
p, d = d, p
return p[l1] if p[l1] <= limit else -1
def similarity_score(s1, s2):
"""
Computes a score that describes how much two strings are matching.
:param s1: first string
:param s2: second string
:return: float score, the higher the more similar
pairs returning non-positive scores should be considered non similar
"""
dist = distance(s1, s2)
if dist == -1:
return -1
set1 = set(s1)
score = len(set1.intersection(s2)) / len(set1)
score -= dist / len(s1)
score -= len(set1.symmetric_difference(s2)) / (len(s1) + len(s2))
return score
def text_from_html(html_fragment):
"""
Returns the plain non-tag text from an html
:param html_fragment: document from which text must be extracted
:return: text extracted from the html
"""
# lxml requires one single root element
tree = etree.fromstring('<p>%s</p>' % html_fragment, etree.XMLParser(recover=True))
return ' '.join(tree.itertext())
def get_unaccent_sql_wrapper(cr):
"""
Returns a function that wraps SQL within unaccent if available
TODO remove when this tool becomes globally available
:param cr: cursor on which the wrapping is done
:return: function that wraps SQL with unaccent if available
"""
if odoo.registry(cr.dbname).has_unaccent:
return lambda x: sql.SQL("unaccent({wrapped_sql})").format(wrapped_sql=x)
return lambda x: x