Mercurial > emacs
annotate admin/charsets/cp932.awk @ 105270:3c67a7fe36f8
*** empty log message ***
author | Juanma Barranquero <lekktu@gmail.com> |
---|---|
date | Tue, 29 Sep 2009 01:01:43 +0000 |
parents | ce88a631c161 |
children | 1d1d5d9bd884 |
rev | line source |
---|---|
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
1 # cp932.awk -- Add sort keys and append user defined area to CP932-2BYTE.map. |
100971 | 2 # Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009 |
89745 | 3 # National Institute of Advanced Industrial Science and Technology (AIST) |
4 # Registration Number H13PRO009 | |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
5 |
89745 | 6 # This file is part of GNU Emacs. |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
7 |
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
8 # GNU Emacs is free software: you can redistribute it and/or modify |
89745 | 9 # it under the terms of the GNU General Public License as published by |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
10 # the Free Software Foundation, either version 3 of the License, or |
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
11 # (at your option) any later version. |
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
12 |
89745 | 13 # GNU Emacs is distributed in the hope that it will be useful, |
14 # but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 # GNU General Public License for more details. | |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
17 |
89745 | 18 # You should have received a copy of the GNU General Public License |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
19 # along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. |
89745 | 20 |
94832
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
21 # Commentary: |
eb2d9dfc8486
Switch to recommended form of GPLv3 permissions notice.
Glenn Morris <rgm@gnu.org>
parents:
91414
diff
changeset
|
22 |
89745 | 23 # Add a sort key 0, 1, 2, or 3 at the tail of each line as a comment |
24 # to realize the round trip mapping to Unicode works as described in | |
25 # this page: | |
26 # http://support.microsoft.com/default.aspx?scid=kb;EN-US;170559 | |
27 # Each sort key means as below: | |
28 # 0: JISX0208 characters. | |
29 # 1: NEC special characters. | |
30 # 2: IBM extension characters. | |
31 # 3: NEC selection of IBM extension characters. | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
32 # 4: user defined area |
89745 | 33 |
34 BEGIN { | |
35 tohex["A"] = 10; | |
36 tohex["B"] = 11; | |
37 tohex["C"] = 12; | |
38 tohex["D"] = 13; | |
39 tohex["E"] = 14; | |
40 tohex["F"] = 15; | |
41 } | |
42 | |
43 function decode_hex(str) { | |
44 n = 0; | |
45 len = length(str); | |
46 for (i = 1; i <= len; i++) | |
47 { | |
48 c = substr(str, i, 1); | |
49 if (c >= "0" && c <= "9") | |
50 n = n * 16 + (c - "0"); | |
51 else | |
52 n = n * 16 + tohex[c]; | |
53 } | |
54 return n; | |
55 } | |
56 | |
57 function sjis_to_jis_ku(code) | |
58 { | |
59 s1 = int(code / 256); | |
60 s2 = code % 256; | |
61 if (s2 >= 159) # s2 >= 0x9F | |
62 { | |
63 if (s1 >= 224) # s1 >= 0xE0 | |
64 j1 = s1 * 2 - 352; # j1 = s1 * 2 - 0x160 | |
65 else | |
66 j1 = s1 * 2 - 224; # j1 = s1 * 2 - 0xE0 | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
67 j2 = s2 - 126 # j2 = s2 - #x7E |
89745 | 68 } |
69 else | |
70 { | |
71 if (s1 >= 224) | |
72 j1 = s1 * 2 - 353; # j1 = s1 * 2 - 0x161 | |
73 else | |
74 j1 = s1 * 2 - 225; # j1 = s1 * 2 - 0xE1 | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
75 if (s2 >= 127) # s2 >= #x7F |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
76 j2 = s2 - 32; |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
77 else |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
78 j2 = s2 - 31; |
89745 | 79 } |
80 return j1 - 32; | |
81 } | |
82 | |
83 /^0x[89E]/ { | |
84 sjis=decode_hex(substr($1, 3, 4)) | |
85 ku=sjis_to_jis_ku(sjis); | |
86 if (ku == 13) | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
87 printf "%s # 1 %02X%02X\n", $0, j1, j2; |
89745 | 88 else if (ku >= 89 && ku <= 92) |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
89 printf "%s # 3 %02X%02X\n", $0, j1, j2; |
89745 | 90 else |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
91 printf "%s # 0 %02X%02X\n", $0, j1, j2; |
89745 | 92 next; |
93 } | |
94 | |
95 /^0xF/ { | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
96 printf "%s # 2\n", $0; |
89745 | 97 next; |
98 } | |
99 | |
100 { | |
101 print; | |
102 } | |
89786
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
103 |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
104 END { |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
105 code = 57344; # 0xE000 |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
106 for (i = 240; i < 250; i++) |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
107 { |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
108 for (j = 64; j <= 126; j++) |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
109 printf "0x%02X%02X 0x%04X # 4\n", i, j, code++; |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
110 for (j = 128; j <= 158; j++) |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
111 printf "0x%02X%02X 0x%04X # 4\n", i, j, code++; |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
112 for (; j <= 252; j++) |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
113 printf "0x%02X%02X 0x%04X # 4\n", i, j, code++; |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
114 } |
cce22763e5b1
Append user defined area of CP932.
Kenichi Handa <handa@m17n.org>
parents:
89745
diff
changeset
|
115 } |
89916
e0e4e6a0599f
Changes from arch/CVS synchronization
Miles Bader <miles@gnu.org>
parents:
89786
diff
changeset
|
116 |
e0e4e6a0599f
Changes from arch/CVS synchronization
Miles Bader <miles@gnu.org>
parents:
89786
diff
changeset
|
117 # arch-tag: 998dc444-759d-43ef-87e3-2ab205011394 |