Fri, 25 Apr 2014 22:07:19 +0200
updated CharDet to 2.2.1, updated changelog
3537
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
1 | ######################## BEGIN LICENSE BLOCK ######################## |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
2 | # The Original Code is mozilla.org code. |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
3 | # |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
4 | # The Initial Developer of the Original Code is |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
5 | # Netscape Communications Corporation. |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
6 | # Portions created by the Initial Developer are Copyright (C) 1998 |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
7 | # the Initial Developer. All Rights Reserved. |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
8 | # |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
9 | # Contributor(s): |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
10 | # Mark Pilgrim - port to Python |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
11 | # |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
12 | # This library is free software; you can redistribute it and/or |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
13 | # modify it under the terms of the GNU Lesser General Public |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
14 | # License as published by the Free Software Foundation; either |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
15 | # version 2.1 of the License, or (at your option) any later version. |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
16 | # |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
17 | # This library is distributed in the hope that it will be useful, |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
20 | # Lesser General Public License for more details. |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
21 | # |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
22 | # You should have received a copy of the GNU Lesser General Public |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
23 | # License along with this library; if not, write to the Free Software |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
24 | # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
25 | # 02110-1301 USA |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
26 | ######################### END LICENSE BLOCK ######################### |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
27 | |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
28 | from .mbcharsetprober import MultiByteCharSetProber |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
29 | from .codingstatemachine import CodingStateMachine |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
30 | from .chardistribution import EUCKRDistributionAnalysis |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
31 | from .mbcssm import EUCKRSMModel |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
32 | |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
33 | |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
34 | class EUCKRProber(MultiByteCharSetProber): |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
35 | def __init__(self): |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
36 | MultiByteCharSetProber.__init__(self) |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
37 | self._mCodingSM = CodingStateMachine(EUCKRSMModel) |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
38 | self._mDistributionAnalyzer = EUCKRDistributionAnalysis() |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
39 | self.reset() |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
40 | |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
41 | def get_charset_name(self): |
7662053c3906
updated CharDet to 2.2.1, updated changelog
T.Rzepka <Tobias.Rzepka@gmail.com>
parents:
12
diff
changeset
|
42 | return "EUC-KR" |