xgettext: Support message syntax checks

With this change, xgettext could report common syntactic problems
in extracted strings.  The current built-in checks are
ellipsis-unicode, space-ellipsis, and quote-unicode.  Those checks
can be enabled with --check option of xgettext and disabled with
special "xgettext:" comment in source files.
Feature suggested by Philip Withnall in:
https://savannah.gnu.org/bugs/?44098
* gettext-tools/src/message.h (enum syntax_check_type): New enum.
(NSYNTAXCHECKS): New constant.
(enum is_syntax_check): New enum.
(struct message_ty): New field 'do_syntax_check'.
(syntax_check_name): New variable declaration.
* gettext-tools/src/message.c (syntax_check_name): New variable.
* gettext-tools/src/msgl-cat.c (catenate_msgdomain_list): Propagate
mp->do_syntax_check.
* gettext-tools/src/msgmerge.c (message_merge): Propagate
ref->do_syntax_check.
* gettext-tools/src/msgl-check.h (syntax_check_message_list): New
declaration.
* gettext-tools/src/msgl-check.c (sentence_end): New function.
(syntax_check_ellipsis_unicode): New function.
(syntax_check_space_ellipsis): New function.
(syntax_check_quote_unicode): New function.
(syntax_check_message): New function.
(syntax_check_message_list): New function.
* gettext-tools/src/read-catalog-abstract.h (po_parse_comment_special):
Adjust function declaration.
* gettext-tools/src/read-catalog-abstract.c (po_parse_comment_special):
Add new argument SCP for syntax checking; all callers changed.
* gettext-tools/src/read-catalog.h (DEFAULT_CATALOG_READER_TY): New
field 'do_syntax_check'.
* gettext-tools/src/read-catalog.c (default_constructor): Initialize
this->do_syntax_check.
(default_copy_comment_state): Propagate this->do_syntax_check.
* gettext-tools/src/xgettext.c (long_options): Add --check option.
(main): Handle --check option.
(usage): Document --check option.
(remember_a_message): Propagate do_syntax_check value.

* gettext-tools/tests/xgettext-14: New file.
* gettext-tools/tests/sentence.c: New file.
* gettext-tools/tests/Makefile.am (TESTS): Add new test.
(noinst_PROGRAMS): Add 'sentence'.
(sentence_SOURCES): New variable.
(sentence_CPPFLAGS): New variable.
(sentence_LDADD): New variable.

* gettext-tools/doc/xgettext.texi: Document --check option.

1 files changed, 54 insertions, 0 deletions

diff --git a/gettext-tools/doc/xgettext.texi b/gettext-tools/doc/xgettext.texi
index 451e25f..f01c9d2 100644
--- a/gettext-tools/doc/xgettext.texi
+++ b/gettext-tools/doc/xgettext.texi
@@ -144,6 +144,60 @@ gettext (
 The second comment line will not be extracted, because there is one
 blank line between the comment line and the keyword.
 
+@item --check[=@var{CHECK}]
+@opindex --check@r{, @code{xgettext} option}
+@cindex supported syntax checks, @code{xgettext}
+Perform a syntax check on msgid and msgid_plural.  The supported checks
+are:
+
+@table @samp
+@item ellipsis-unicode
+Prefer Unicode ellipsis character over ASCII @code{...}
+
+@item space-ellipsis
+Prohibit whitespace before an ellipsis character
+
+@item quote-unicode
+Prefer Unicode quotation marks over ASCII @code{"'`}
+
+@end table
+
+The option has an effect on all input files.  To enable or disable
+checks for a certain string, you can mark it with an @code{xgettext:}
+special comment in the source file.  For example, if you specify the
+@code{--check=space-ellipsis} option, but want to suppress the check on
+a particular string, add the following comment:
+
+@example
+/* xgettext: no-space-ellipsis-check */
+gettext ("We really want a space before ellipsis here ...");
+@end example
+
+The @code{xgettext:} comment can be followed by flags separated with a
+comma.  The possible flags are of the form @samp{[no-]@var{name}-check},
+where @var{name} is the name of a valid syntax check.  If a flag is
+prefixed by @code{no-}, the meaning is negated.
+
+Some tests apply the checks to each sentence within the msgid, rather
+than the whole string.  xgettext detects the end of sentence by
+performing a pattern match, which usually looks for a period followed by
+a certain number of spaces.  The number is specified with the
+@code{--sentence-end} option.
+
+@item --sentence-end[=@var{TYPE}]
+@opindex --sentence-end@r{, @code{xgettext} option}
+@cindex sentence end markers, @code{xgettext}
+The supported values are:
+
+@table @samp
+@item single-space
+Expect at least one whitespace after a period
+
+@item double-space
+Expect at least two whitespaces after a period
+
+@end table
+
 @end table
 
 @subsection Language specific options