| 250 |
kaklik |
1 |
#!/bin/sh |
|
|
2 |
# $Id: sync_lang.sh,v 2.22 2005/12/07 23:03:46 nijel Exp $ |
|
|
3 |
## |
|
|
4 |
# Shell script that synchronises all translations in phpMyAdmin |
|
|
5 |
## |
|
|
6 |
# Any parameters (except --iconv/--recode) will be passed to grep to filter |
|
|
7 |
# processed translation, for example: './sync_lang.sh czech' will process only |
|
|
8 |
# czech translation, './sync_lang.sh -e czech -e english' will process czech |
|
|
9 |
# and english translations. |
|
|
10 |
## |
|
|
11 |
# Written by Michal Cihar <nijel at users.sourceforge.net> |
|
|
12 |
## |
|
|
13 |
# Changes: |
|
|
14 |
# 2005-12-08 |
|
|
15 |
# * less verbose output to allow quick overview |
|
|
16 |
# 2005-11-29 |
|
|
17 |
# * hack for multibyte chars, so that \'; at the end will not fool PHP |
|
|
18 |
# 2004-09-22 |
|
|
19 |
# * default to iconv, as it doesn't break things as recode does |
|
|
20 |
# 2004-09-03 |
|
|
21 |
# * hack for hebrew |
|
|
22 |
# 2003-11-18 |
|
|
23 |
# * switch php3 -> php |
|
|
24 |
# 2003-04-14 |
|
|
25 |
# * convert only files that are needed to convert (checks mtime), --force to |
|
|
26 |
# avoid this checking |
|
|
27 |
# * get charset from filename when reading from file failed |
|
|
28 |
# * report failed translations at the end |
|
|
29 |
# 2002-09-18 |
|
|
30 |
# * now accepts parameters --iconv/--recode for specifying which convertor |
|
|
31 |
# to use |
|
|
32 |
# 2002-08-13 |
|
|
33 |
# * support for synchronisation only for selected language(s) |
|
|
34 |
# 2002-07-18 |
|
|
35 |
# * can exclude some languages from conversion |
|
|
36 |
# 2002-07-17 |
|
|
37 |
# * support for multiple convertors (recode added) |
|
|
38 |
## |
|
|
39 |
|
|
|
40 |
## |
|
|
41 |
# convertor setup |
|
|
42 |
## |
|
|
43 |
# CONVERTOR_PARAMS is used for printf and it also receives two params: source |
|
|
44 |
# and target charset |
|
|
45 |
# |
|
|
46 |
|
|
|
47 |
case "$1" in |
|
|
48 |
--iconv) |
|
|
49 |
echo Using iconv on user request |
|
|
50 |
CONVERTOR=iconv |
|
|
51 |
# the space on following is REQUIRED |
|
|
52 |
CONVERTOR_PARAMS=" -f %s -t %s" |
|
|
53 |
shift |
|
|
54 |
;; |
|
|
55 |
--recode) |
|
|
56 |
echo Using recode on user request |
|
|
57 |
echo '(please use iconv for arabic)' |
|
|
58 |
CONVERTOR=recode |
|
|
59 |
CONVERTOR_PARAMS=" -f %s..%s" |
|
|
60 |
shift |
|
|
61 |
;; |
|
|
62 |
*) |
|
|
63 |
echo Using iconv as default, force with --iconv/--recode |
|
|
64 |
CONVERTOR=iconv |
|
|
65 |
# the space on following is REQUIRED |
|
|
66 |
CONVERTOR_PARAMS=" -f %s -t %s" |
|
|
67 |
;; |
|
|
68 |
esac |
|
|
69 |
|
|
|
70 |
if [ "$1" = "--force" ] ; then |
|
|
71 |
FORCE=1 |
|
|
72 |
shift |
|
|
73 |
else |
|
|
74 |
FORCE=0 |
|
|
75 |
fi |
|
|
76 |
|
|
|
77 |
|
|
|
78 |
## |
|
|
79 |
# names of translations to process |
|
|
80 |
## |
|
|
81 |
# Here should be listed all translations for which conversion should be done. |
|
|
82 |
# The name is filename without inc.php. |
|
|
83 |
# |
|
|
84 |
BASE_TRANSLATIONS="afrikaans-iso-8859-1 |
|
|
85 |
albanian-iso-8859-1 |
|
|
86 |
arabic-windows-1256 |
|
|
87 |
azerbaijani-iso-8859-9 |
|
|
88 |
basque-iso-8859-1 |
|
|
89 |
belarusian_cyrillic-windows-1251 |
|
|
90 |
belarusian_latin-utf-8 |
|
|
91 |
bosnian-windows-1250 |
|
|
92 |
brazilian_portuguese-iso-8859-1 |
|
|
93 |
bulgarian-utf-8 |
|
|
94 |
catalan-iso-8859-1 |
|
|
95 |
chinese_traditional-utf-8 |
|
|
96 |
chinese_simplified-gb2312 |
|
|
97 |
croatian-iso-8859-2 |
|
|
98 |
czech-utf-8 |
|
|
99 |
danish-iso-8859-1 |
|
|
100 |
dutch-iso-8859-1 |
|
|
101 |
english-iso-8859-1 |
|
|
102 |
estonian-iso-8859-1 |
|
|
103 |
finnish-iso-8859-1 |
|
|
104 |
french-iso-8859-1 |
|
|
105 |
galician-iso-8859-1 |
|
|
106 |
german-utf-8 |
|
|
107 |
greek-iso-8859-7 |
|
|
108 |
hebrew-iso-8859-8-i |
|
|
109 |
hungarian-iso-8859-2 |
|
|
110 |
indonesian-iso-8859-1 |
|
|
111 |
italian-iso-8859-1 |
|
|
112 |
japanese-utf-8 |
|
|
113 |
korean-euc-kr |
|
|
114 |
latvian-windows-1257 |
|
|
115 |
lithuanian-windows-1257 |
|
|
116 |
malay-iso-8859-1 |
|
|
117 |
norwegian-iso-8859-1 |
|
|
118 |
persian-windows-1256 |
|
|
119 |
polish-iso-8859-2 |
|
|
120 |
portuguese-iso-8859-1 |
|
|
121 |
romanian-iso-8859-1 |
|
|
122 |
russian-windows-1251 |
|
|
123 |
serbian_cyrillic-windows-1251 |
|
|
124 |
serbian_latin-windows-1250 |
|
|
125 |
slovenian-iso-8859-2 |
|
|
126 |
slovak-iso-8859-2 |
|
|
127 |
spanish-iso-8859-1 |
|
|
128 |
swedish-iso-8859-1 |
|
|
129 |
tatarish-iso-8859-9 |
|
|
130 |
thai-tis-620 |
|
|
131 |
turkish-utf-8 |
|
|
132 |
ukrainian-windows-1251" |
|
|
133 |
|
|
|
134 |
## |
|
|
135 |
# which translations should not be translated to utf-8 |
|
|
136 |
## |
|
|
137 |
# List here any translation that should not be converted to utf-8. The name is |
|
|
138 |
# same as above. |
|
|
139 |
# |
|
|
140 |
IGNORE_UTF="" |
|
|
141 |
|
|
|
142 |
## |
|
|
143 |
# which translations should not be automatically generated |
|
|
144 |
## |
|
|
145 |
# List here any translation should not be automatically generated from base |
|
|
146 |
# translation for that language (usually for those which are not correctly |
|
|
147 |
# supported by convertor). |
|
|
148 |
# |
|
|
149 |
IGNORE_TRANSLATIONS=" |
|
|
150 |
russian-cp-866" |
|
|
151 |
|
|
|
152 |
## |
|
|
153 |
# end of configuration, you hopefully won't need to edit anything bellow |
|
|
154 |
## |
|
|
155 |
|
|
|
156 |
TEMPFILE=`mktemp /tmp/pma-sync-lang.XXXXXX` |
|
|
157 |
|
|
|
158 |
cleanup() { |
|
|
159 |
rm -f $TEMPFILE |
|
|
160 |
} |
|
|
161 |
|
|
|
162 |
trap cleanup INT ABRT TERM |
|
|
163 |
|
|
|
164 |
FAILED="" |
|
|
165 |
|
|
|
166 |
echo "-------------------------------------------------------------------" |
|
|
167 |
# go through all file we should process |
|
|
168 |
for base in $BASE_TRANSLATIONS ; do |
|
|
169 |
if [ "$#" -gt 0 ] ; then |
|
|
170 |
if ( echo $base | grep -q "$@" ) ; then |
|
|
171 |
true |
|
|
172 |
else |
|
|
173 |
continue |
|
|
174 |
fi |
|
|
175 |
fi |
|
|
176 |
# grep language from basename |
|
|
177 |
lang=$(echo $base|sed 's%-.*%%') |
|
|
178 |
# which files will we create from current? |
|
|
179 |
create_files=$(ls --color=none -1 $lang*.inc.php|grep -v $base.inc.php) |
|
|
180 |
|
|
|
181 |
for ignore in $IGNORE_TRANSLATIONS ; do |
|
|
182 |
create_files=$(echo "$create_files" | grep -v $ignore) |
|
|
183 |
done |
|
|
184 |
|
|
|
185 |
# charset of source file |
|
|
186 |
src_charset=$(grep '\$charset' $base.inc.php | sed "s%^[^'\"]*['\"]\\([^'\"]*\\)['\"][^'\"]*$%\\1%") |
|
|
187 |
replace_charset=$src_charset |
|
|
188 |
# special case for hebrew |
|
|
189 |
if [ $src_charset = 'iso-8859-8-i' ] ; then |
|
|
190 |
src_charset=iso-8859-8 |
|
|
191 |
fi |
|
|
192 |
echo -n "$base [charset $src_charset]" |
|
|
193 |
|
|
|
194 |
# do we already have utf-8 translation? |
|
|
195 |
if [ $src_charset = 'utf-8' ] ; then |
|
|
196 |
is_utf=yes |
|
|
197 |
else |
|
|
198 |
is_utf=no |
|
|
199 |
fi |
|
|
200 |
|
|
|
201 |
# at first update existing translations |
|
|
202 |
for file in $create_files ; do |
|
|
203 |
# charset of destination file |
|
|
204 |
|
|
|
205 |
# grepping from file causes problems when it is empty... |
|
|
206 |
charset=$(grep '\$charset' $file | sed "s%^[^'\"]*['\"]\\([^'\"]*\\)['\"][^'\"]*$%\\1%") |
|
|
207 |
if [ -z "$charset" ] ; then |
|
|
208 |
charset=$(echo $file | sed -e 's/^[^-]*-//' -e 's/\.inc\.php\?$//') |
|
|
209 |
fi |
|
|
210 |
|
|
|
211 |
if [ $charset = 'utf-8' ] ; then |
|
|
212 |
is_utf=yes |
|
|
213 |
fi |
|
|
214 |
|
|
|
215 |
# check whether we need to update translation |
|
|
216 |
if [ ! "$base.inc.php" -nt "$file" -a "$FORCE" -eq 0 -a -s "$file" ] ; then |
|
|
217 |
echo -n " ($file:ok)" |
|
|
218 |
continue |
|
|
219 |
fi |
|
|
220 |
|
|
|
221 |
echo -n " ($file:to $charset:" |
|
|
222 |
if [ $charset = 'utf-8' ] ; then |
|
|
223 |
# if we convert to utf-8, we should add allow_recoding |
|
|
224 |
is_utf=yes |
|
|
225 |
$CONVERTOR $(printf "$CONVERTOR_PARAMS" $src_charset $charset) < $base.inc.php| sed -e "s/$replace_charset/$charset/" -e '/\$charset/a\ |
|
|
226 |
$allow_recoding = TRUE;' > $TEMPFILE |
|
|
227 |
elif [ $src_charset = 'utf-8' ] ; then |
|
|
228 |
is_utf=yes |
|
|
229 |
# if we convert from utf-8, we should remove allow_recoding |
|
|
230 |
$CONVERTOR $(printf "$CONVERTOR_PARAMS" $src_charset $charset) < $base.inc.php| grep -v allow_recoding | sed "s/$replace_charset/$charset/" > $TEMPFILE |
|
|
231 |
else |
|
|
232 |
# just convert |
|
|
233 |
$CONVERTOR $(printf "$CONVERTOR_PARAMS" $src_charset $charset) < $base.inc.php| sed "s/$replace_charset/$charset/" > $TEMPFILE |
|
|
234 |
fi |
|
|
235 |
if [ -s $TEMPFILE ] ; then |
|
|
236 |
sed "s/\\\\';[[:space:]]*$/\\\\\\\\';/" $TEMPFILE > $file |
|
|
237 |
echo -n 'done)' |
|
|
238 |
else |
|
|
239 |
FAILED="$FAILED $file" |
|
|
240 |
echo -n 'FAILED)' |
|
|
241 |
fi |
|
|
242 |
done |
|
|
243 |
|
|
|
244 |
# now check whether we found utf-8 translation |
|
|
245 |
if [ $is_utf = no ] ; then |
|
|
246 |
if ( echo $IGNORE_UTF | grep -q $base ) ; then |
|
|
247 |
# utf-8 should not be created |
|
|
248 |
true |
|
|
249 |
else |
|
|
250 |
# we should create utf-8 translation |
|
|
251 |
charset=utf-8 |
|
|
252 |
file=$lang-$charset.inc.php |
|
|
253 |
echo -n " [$file:$charset:" |
|
|
254 |
$CONVERTOR $(printf "$CONVERTOR_PARAMS" $src_charset $charset) < $base.inc.php| sed -e "s/$replace_charset/$charset/" -e '/\$charset/a\ |
|
|
255 |
$allow_recoding = TRUE;' > $TEMPFILE |
|
|
256 |
if [ -s $TEMPFILE ] ; then |
|
|
257 |
cat $TEMPFILE > $file |
|
|
258 |
echo -n 'done)' |
|
|
259 |
else |
|
|
260 |
FAILED="$FAILED $file" |
|
|
261 |
echo -n 'FAILED)' |
|
|
262 |
fi |
|
|
263 |
fi |
|
|
264 |
fi |
|
|
265 |
echo |
|
|
266 |
done |
|
|
267 |
|
|
|
268 |
echo "-------------------------------------------------------------------" |
|
|
269 |
|
|
|
270 |
if [ -z "$FAILED" ] ; then |
|
|
271 |
echo "Everything seems to went okay" |
|
|
272 |
else |
|
|
273 |
echo "!!!SOME CONVERSION FAILED!!!" |
|
|
274 |
echo "Following file were NOT updated:" |
|
|
275 |
echo |
|
|
276 |
echo "$FAILED" |
|
|
277 |
echo |
|
|
278 |
echo "!!!SOME CONVERSION FAILED!!!" |
|
|
279 |
fi |
|
|
280 |
|
|
|
281 |
cleanup |