summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorBruno Haible <bruno@clisp.org>2008-08-16 10:08:53 +0000
committerBruno Haible <bruno@clisp.org>2009-06-23 12:15:47 +0200
commite3af853612c1398b62190e14d6e711559e0b0ef5 (patch)
tree9b4de73913121ff8d775f906624bb9f136552756
parent25360c47bf5554419146fe65855582e393730bcd (diff)
downloadexternal_gettext-e3af853612c1398b62190e14d6e711559e0b0ef5.zip
external_gettext-e3af853612c1398b62190e14d6e711559e0b0ef5.tar.gz
external_gettext-e3af853612c1398b62190e14d6e711559e0b0ef5.tar.bz2
Improve interpretation of the [ ] operators.
-rw-r--r--gettext-tools/src/ChangeLog11
-rw-r--r--gettext-tools/src/x-python.c75
2 files changed, 68 insertions, 18 deletions
diff --git a/gettext-tools/src/ChangeLog b/gettext-tools/src/ChangeLog
index 5bc89d2..85427d4 100644
--- a/gettext-tools/src/ChangeLog
+++ b/gettext-tools/src/ChangeLog
@@ -1,5 +1,16 @@
2008-08-16 Bruno Haible <bruno@clisp.org>
+ * x-python.c (enum token_type_ty): New values token_type_lbracket,
+ token_type_rbracket.
+ (phase5_get): Recognize also token_type_lbracket, token_type_rbracket.
+ (extract_balanced): Renamed from extract_parenthesized. Add 'delim'
+ argument. Handle token_type_lbracket and token_type_rbracket.
+ (extract_python): Update.
+ Reported by Claude Paroz <claude@2xlibre.net>
+ via <http://savannah.gnu.org/bugs/?23824>.
+
+2008-08-16 Bruno Haible <bruno@clisp.org>
+
* x-php.c (extract_balanced): Fix small bug in 2007-03-17 commit.
2008-08-15 Bruno Haible <bruno@clisp.org>
diff --git a/gettext-tools/src/x-python.c b/gettext-tools/src/x-python.c
index e05aca4..e12b0e9 100644
--- a/gettext-tools/src/x-python.c
+++ b/gettext-tools/src/x-python.c
@@ -988,6 +988,8 @@ enum token_type_ty
token_type_lparen, /* ( */
token_type_rparen, /* ) */
token_type_comma, /* , */
+ token_type_lbracket, /* [ */
+ token_type_rbracket, /* ] */
token_type_string, /* "abc", 'abc', """abc""", '''abc''' */
token_type_symbol, /* symbol, number */
token_type_other /* misc. operator */
@@ -1581,13 +1583,13 @@ phase5_get (token_ty *tp)
case '[': case '{':
open_pbb++;
- tp->type = token_type_other;
+ tp->type = (c == '[' ? token_type_lbracket : token_type_other);
return;
case ']': case '}':
if (open_pbb > 0)
open_pbb--;
- tp->type = token_type_other;
+ tp->type = (c == ']' ? token_type_rbracket : token_type_other);
return;
default:
@@ -1664,14 +1666,17 @@ static flag_context_list_table_ty *flag_context_list_table;
and msgid_plural can contain subexpressions of the same form. */
-/* Extract messages until the next balanced closing parenthesis.
+/* Extract messages until the next balanced closing parenthesis or bracket.
Extracted messages are added to MLP.
- Return true upon eof, false upon closing parenthesis. */
+ DELIM can be either token_type_rparen or token_type_rbracket, or
+ token_type_eof to accept both.
+ Return true upon eof, false upon closing parenthesis or bracket. */
static bool
-extract_parenthesized (message_list_ty *mlp,
- flag_context_ty outer_context,
- flag_context_list_iterator_ty context_iter,
- struct arglist_parser *argparser)
+extract_balanced (message_list_ty *mlp,
+ token_type_ty delim,
+ flag_context_ty outer_context,
+ flag_context_list_iterator_ty context_iter,
+ struct arglist_parser *argparser)
{
/* Current argument number. */
int arg = 1;
@@ -1720,9 +1725,10 @@ extract_parenthesized (message_list_ty *mlp,
continue;
case token_type_lparen:
- if (extract_parenthesized (mlp, inner_context, next_context_iter,
- arglist_parser_alloc (mlp,
- state ? next_shapes : NULL)))
+ if (extract_balanced (mlp, token_type_rparen,
+ inner_context, next_context_iter,
+ arglist_parser_alloc (mlp,
+ state ? next_shapes : NULL)))
{
xgettext_current_source_encoding = po_charset_utf8;
arglist_parser_done (argparser, arg);
@@ -1734,10 +1740,16 @@ extract_parenthesized (message_list_ty *mlp,
continue;
case token_type_rparen:
- xgettext_current_source_encoding = po_charset_utf8;
- arglist_parser_done (argparser, arg);
- xgettext_current_source_encoding = xgettext_current_file_source_encoding;
- return false;
+ if (delim == token_type_rparen || delim == token_type_eof)
+ {
+ xgettext_current_source_encoding = po_charset_utf8;
+ arglist_parser_done (argparser, arg);
+ xgettext_current_source_encoding = xgettext_current_file_source_encoding;
+ return false;
+ }
+ next_context_iter = null_context_list_iterator;
+ state = 0;
+ continue;
case token_type_comma:
arg++;
@@ -1749,6 +1761,32 @@ extract_parenthesized (message_list_ty *mlp,
state = 0;
continue;
+ case token_type_lbracket:
+ if (extract_balanced (mlp, token_type_rbracket,
+ null_context, null_context_list_iterator,
+ arglist_parser_alloc (mlp, NULL)))
+ {
+ xgettext_current_source_encoding = po_charset_utf8;
+ arglist_parser_done (argparser, arg);
+ xgettext_current_source_encoding = xgettext_current_file_source_encoding;
+ return true;
+ }
+ next_context_iter = null_context_list_iterator;
+ state = 0;
+ continue;
+
+ case token_type_rbracket:
+ if (delim == token_type_rbracket || delim == token_type_eof)
+ {
+ xgettext_current_source_encoding = po_charset_utf8;
+ arglist_parser_done (argparser, arg);
+ xgettext_current_source_encoding = xgettext_current_file_source_encoding;
+ return false;
+ }
+ next_context_iter = null_context_list_iterator;
+ state = 0;
+ continue;
+
case token_type_string:
{
lex_pos_ty pos;
@@ -1825,10 +1863,11 @@ extract_python (FILE *f,
init_keywords ();
- /* Eat tokens until eof is seen. When extract_parenthesized returns
+ /* Eat tokens until eof is seen. When extract_balanced returns
due to an unbalanced closing parenthesis, just restart it. */
- while (!extract_parenthesized (mlp, null_context, null_context_list_iterator,
- arglist_parser_alloc (mlp, NULL)))
+ while (!extract_balanced (mlp, token_type_eof,
+ null_context, null_context_list_iterator,
+ arglist_parser_alloc (mlp, NULL)))
;
fp = NULL;