Mercurial > libervia-backend

annotate src/plugins/plugin_exp_lang_detect.py @ 1965:4c5d8cd35690

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
plugin exp_lang_detect: language detection plugin, first draft This plugin try to autodetect outgoing and ingoing messages language when it is not specified by sender
author Goffi <goffi@goffi.org>
date Mon, 20 Jun 2016 18:44:27 +0200
parents
children d95a6d553bec
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
1965
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
1 #!/usr/bin/env python2
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
2 # -*- coding: utf-8 -*-
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
3
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
4 # SAT plugin to detect language (experimental)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
5 # Copyright (C) 2009-2016 Jérôme Poisson (goffi@goffi.org)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
6
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
7 # This program is free software: you can redistribute it and/or modify
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
8 # it under the terms of the GNU Affero General Public License as published by
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
9 # the Free Software Foundation, either version 3 of the License, or
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
10 # (at your option) any later version.
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
11
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
12 # This program is distributed in the hope that it will be useful,
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
15 # GNU Affero General Public License for more details.
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
16
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
17 # You should have received a copy of the GNU Affero General Public License
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
18 # along with this program. If not, see <http://www.gnu.org/licenses/>.
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
19
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
20 from sat.core.i18n import _
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
21 from sat.core.log import getLogger
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
22 log = getLogger(__name__)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
23 from sat.core import exceptions
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
24
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
25 try:
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
26 from langid.langid import LanguageIdentifier, model
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
27 except ImportError:
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
28 raise exceptions.MissingModule(u'Missing module langid, please download/install it with "pip install langid")')
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
29
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
30 identifier = LanguageIdentifier.from_modelstring(model, norm_probs=False)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
31
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
32
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
33 PLUGIN_INFO = {
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
34 "name": "Language detection plugin",
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
35 "import_name": "EXP-LANG-DETECT",
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
36 "type": "EXP",
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
37 "protocols": [],
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
38 "dependencies": [],
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
39 "main": "LangDetect",
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
40 "handler": "no",
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
41 "description": _("""Detect and set message language when unknown""")
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
42 }
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
43
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
44
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
45 class LangDetect(object):
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
46
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
47 def __init__(self, host):
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
48 log.info(_(u"Language detection plugin initialization"))
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
49 self.host = host
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
50 host.trigger.add("MessageReceived", self.MessageReceivedTrigger)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
51 host.trigger.add("messageSend", self.MessageSendTrigger)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
52
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
53 def addLanguage(self, mess_data):
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
54 message = mess_data['message']
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
55 if len(message) == 1 and message.keys()[0] == '':
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
56 msg = message.values()[0]
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
57 lang = identifier.classify(msg)[0]
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
58 mess_data["message"] = {lang: msg}
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
59 return mess_data
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
60
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
61 def MessageReceivedTrigger(self, client, message_elt, post_treat):
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
62 """ Check if source is linked and repeat message, else do nothing """
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
63 post_treat.addCallback(self.addLanguage)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
64 return True
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
65
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
66 def MessageSendTrigger(self, client, data, pre_xml_treatments, post_xml_treatments):
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
67 self.addLanguage(data)
4c5d8cd35690 plugin exp_lang_detect: language detection plugin, first draft
Goffi <goffi@goffi.org>
parents:
diff changeset
68 return True