+2002-08-18 Bruno Haible <bruno@clisp.org>
+
+ * gettext.texi (php-format): New subsection.
+ (PHP): Update.
+ * xgettext.texi: Mention language PHP.
+
+2002-08-06 Bruno Haible <bruno@clisp.org>
+
+ * gettext-0.11.5 released.
+
2002-07-16 Bruno Haible <bruno@clisp.org>
* matrix.texi: Update.
* object-pascal-format:: Object Pascal Format Strings
* ycp-format:: YCP Format Strings
* tcl-format:: Tcl Format Strings
+* php-format:: PHP Format Strings
Individual Programming Languages
* object-pascal-format:: Object Pascal Format Strings
* ycp-format:: YCP Format Strings
* tcl-format:: Tcl Format Strings
+* php-format:: PHP Format Strings
@end menu
@node c-format, python-format, Translators for other Languages, Translators for other Languages
In summary, a directive starts with @samp{%} and is followed by @samp{%}
or a nonzero digit (@samp{1} to @samp{9}).
-@node tcl-format, , ycp-format, Translators for other Languages
+@node tcl-format, php-format, ycp-format, Translators for other Languages
@subsection Tcl Format Strings
Tcl format strings are described in the @file{format.n} manual page,
@uref{http://www.scriptics.com/man/tcl8.3/TclCmd/format.htm}.
+@node php-format, , tcl-format, Translators for other Languages
+@subsection PHP Format Strings
+
+PHP format strings are described in the documentation of the PHP function
+@code{sprintf}, in @file{phpdoc/manual/function.sprintf.html}.
+
@node Maintainers for other Languages, List of Programming Languages, Translators for other Languages, Programming Languages
@section The Maintainer's View
@c Perl 1911
@c C++ 1379 *
@c Java 1200 *
-@c PHP 1051
+@c PHP 1051 *
@c Python 613 *
@c Unix Shell 357
@c Tcl 266 *
@table @asis
@item RPMs
-mod_php4, phplib, phpdoc
+mod_php4, mod_php4-core, phplib, phpdoc
@item File extension
@code{php}, @code{php3}, @code{php4}
@item String syntax
-@code{"abc"}
+@code{"abc"}, @code{'abc'}
@item gettext shorthand
@code{_("abc")}
@code{bindtextdomain} function
@item setlocale
-@code{setlocale} function
+Programmer must call @code{setlocale (LC_ALL, "")}
@item Prerequisite
---
use
@item Extractor
----
+@code{xgettext}
@item Formatting with positions
----
+@code{printf "%2\$d %1\$d"}
@item Portability
On platforms without gettext, the functions are not available.
Specifies the language of the input files. The supported languages
are @code{C}, @code{C++}, @code{ObjectiveC}, @code{PO}, @code{Python},
@code{Lisp}, @code{EmacsLisp}, @code{librep}, @code{Java}, @code{awk},
-@code{YCP}, @code{Tcl}, @code{RST}, @code{Glade}.
+@code{YCP}, @code{Tcl}, @code{PHP}, @code{RST}, @code{Glade}.
@item -C
@itemx --c++
+2002-08-18 Bruno Haible <bruno@clisp.org>
+
+ * message.h (enum format_type): New enum value 'format_php'.
+ (NFORMATS): Increment.
+ * message.c (format_language, format_language_pretty): Add entry
+ for php.
+ * format.h (formatstring_php): New declaration.
+ * format-php.c: New file.
+ * format.c (formatstring_parsers): Add entry for php.
+ * x-php.h: New file.
+ * x-php.c: New file.
+ * xgettext.c: Include x-php.c.
+ (main): Call x_php_extract_all, x_php_keyword.
+ (language_to_scanner): Add PHP rule.
+ (extension_to_language): Add PHP rule.
+ * Makefile.am (noinst_HEADERS): Add x-php.h.
+ (FORMAT_SOURCE): Add format-php.c.
+ (xgettext_SOURCES): Add x-php.c.
+
2002-08-17 Bruno Haible <bruno@clisp.org>
* urlget.c (fetch): Also try invoking the 'curl' program.
msgl-ascii.h msgl-cat.h msgl-english.h msgfmt.h msgunfmt.h read-mo.h \
write-mo.h read-java.h write-java.h read-tcl.h write-tcl.h po-time.h \
plural-table.h format.h xgettext.h x-c.h x-po.h x-python.h x-lisp.h \
-x-elisp.h x-librep.h x-java.h x-awk.h x-ycp.h x-tcl.h x-rst.h x-glade.h
+x-elisp.h x-librep.h x-java.h x-awk.h x-ycp.h x-tcl.h x-php.h x-rst.h x-glade.h
EXTRA_DIST = FILES project-id msgunfmt.tcl \
gnu/gettext/DumpResource.java gnu/gettext/GetURL.java
# xgettext and msgfmt deal with format strings.
FORMAT_SOURCE = format.c \
format-c.c format-python.c format-lisp.c format-elisp.c format-librep.c \
-format-java.c format-awk.c format-pascal.c format-ycp.c format-tcl.c
+format-java.c format-awk.c format-pascal.c format-ycp.c format-tcl.c \
+format-php.c
# libgettextsrc contains all code that is needed by at least two programs.
libgettextsrc_la_SOURCES = \
msgunfmt_SOURCES = msgunfmt.c read-mo.c read-java.c read-tcl.c
xgettext_SOURCES = xgettext.c \
x-c.c x-po.c x-python.c x-lisp.c x-elisp.c x-librep.c x-java.l x-awk.c \
- x-ycp.c x-tcl.c x-rst.c x-glade.c
+ x-ycp.c x-tcl.c x-php.c x-rst.c x-glade.c
msgattrib_SOURCES = msgattrib.c
msgcat_SOURCES = msgcat.c
msgcomm_SOURCES = msgcomm.c
--- /dev/null
+/* PHP format strings.
+ Copyright (C) 2001-2002 Free Software Foundation, Inc.
+ Written by Bruno Haible <bruno@clisp.org>, 2002.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#include <stdbool.h>
+#include <stdlib.h>
+
+#include "format.h"
+#include "xmalloc.h"
+#include "error.h"
+#include "progname.h"
+#include "gettext.h"
+
+#define _(str) gettext (str)
+
+/* PHP format strings are described in phpdoc-4.0.6, file
+ phpdoc/manual/function.sprintf.html, and are implemented in
+ php-4.1.0/ext/standard/formatted_print.c.
+ A directive
+ - starts with '%' or '%m$' where m is a positive integer,
+ - is optionally followed by any of the characters '0', '-', ' ', or
+ "'<anychar>", each of which acts as a flag,
+ - is optionally followed by a width specification: a nonempty digit
+ sequence,
+ - is optionally followed by '.' and a precision specification: a nonempty
+ digit sequence,
+ - is optionally followed by a size specifier 'l', which is ignored,
+ - is finished by a specifier
+ - 's', that needs a string argument,
+ - 'b', 'd', 'u', 'o', 'x', 'X', that need an integer argument,
+ - 'e', 'f', that need a floating-point argument,
+ - 'c', that needs a character argument.
+ Additionally there is the directive '%%', which takes no argument.
+ Numbered and unnumbered argument specifications can be used in the same
+ string. Numbered argument specifications have no influence on the
+ "current argument index", that is incremented each time an argument is read.
+ */
+
+enum format_arg_type
+{
+ FAT_INTEGER,
+ FAT_FLOAT,
+ FAT_CHARACTER,
+ FAT_STRING
+};
+
+struct numbered_arg
+{
+ unsigned int number;
+ enum format_arg_type type;
+};
+
+struct spec
+{
+ unsigned int directives;
+ unsigned int numbered_arg_count;
+ unsigned int allocated;
+ struct numbered_arg *numbered;
+};
+
+/* Locale independent test for a decimal digit.
+ Argument can be 'char' or 'unsigned char'. (Whereas the argument of
+ <ctype.h> isdigit must be an 'unsigned char'.) */
+#undef isdigit
+#define isdigit(c) ((unsigned int) ((c) - '0') < 10)
+
+
+/* Prototypes for local functions. Needed to ensure compiler checking of
+ function argument counts despite of K&R C function definition syntax. */
+static int numbered_arg_compare PARAMS ((const void *p1, const void *p2));
+static void *format_parse PARAMS ((const char *format));
+static void format_free PARAMS ((void *descr));
+static int format_get_number_of_directives PARAMS ((void *descr));
+static bool format_check PARAMS ((const lex_pos_ty *pos,
+ void *msgid_descr, void *msgstr_descr,
+ bool equality,
+ bool noisy, const char *pretty_msgstr));
+
+
+static int
+numbered_arg_compare (p1, p2)
+ const void *p1;
+ const void *p2;
+{
+ unsigned int n1 = ((const struct numbered_arg *) p1)->number;
+ unsigned int n2 = ((const struct numbered_arg *) p2)->number;
+
+ return (n1 > n2 ? 1 : n1 < n2 ? -1 : 0);
+}
+
+static void *
+format_parse (format)
+ const char *format;
+{
+ unsigned int directives;
+ unsigned int numbered_arg_count;
+ unsigned int allocated;
+ struct numbered_arg *numbered;
+ unsigned int unnumbered_arg_count;
+ struct spec *result;
+
+ directives = 0;
+ numbered_arg_count = 0;
+ allocated = 0;
+ numbered = NULL;
+ unnumbered_arg_count = 0;
+
+ for (; *format != '\0';)
+ if (*format++ == '%')
+ {
+ /* A directive. */
+ directives++;
+
+ if (*format != '%')
+ {
+ /* A complex directive. */
+ unsigned int number;
+ enum format_arg_type type;
+
+ number = ++unnumbered_arg_count;
+ if (isdigit (*format))
+ {
+ const char *f = format;
+ unsigned int m = 0;
+
+ do
+ {
+ m = 10 * m + (*f - '0');
+ f++;
+ }
+ while (isdigit (*f));
+
+ if (*f == '$')
+ {
+ if (m == 0)
+ goto bad_format;
+ number = m;
+ format = ++f;
+ --unnumbered_arg_count;
+ }
+ }
+
+ /* Parse flags. */
+ for (;;)
+ {
+ if (*format == '0' || *format == '-' || *format == ' ')
+ format++;
+ else if (*format == '\'')
+ {
+ format++;
+ if (*format == '\0')
+ goto bad_format;
+ format++;
+ }
+ else
+ break;
+ }
+
+ /* Parse width. */
+ if (isdigit (*format))
+ {
+ do
+ format++;
+ while (isdigit (*format));
+ }
+
+ /* Parse precision. */
+ if (*format == '.')
+ {
+ format++;
+
+ if (isdigit (*format))
+ {
+ do
+ format++;
+ while (isdigit (*format));
+ }
+ else
+ --format; /* will jump to bad_format */
+ }
+
+ /* Parse size. */
+ if (*format == 'l')
+ format++;
+
+ switch (*format)
+ {
+ case 'b': case 'd': case 'u': case 'o': case 'x': case 'X':
+ type = FAT_INTEGER;
+ break;
+ case 'e': case 'f':
+ type = FAT_FLOAT;
+ break;
+ case 'c':
+ type = FAT_CHARACTER;
+ break;
+ case 's':
+ type = FAT_STRING;
+ break;
+ default:
+ goto bad_format;
+ }
+
+ if (allocated == numbered_arg_count)
+ {
+ allocated = 2 * allocated + 1;
+ numbered = (struct numbered_arg *) xrealloc (numbered, allocated * sizeof (struct numbered_arg));
+ }
+ numbered[numbered_arg_count].number = number;
+ numbered[numbered_arg_count].type = type;
+ numbered_arg_count++;
+ }
+
+ format++;
+ }
+
+ /* Sort the numbered argument array, and eliminate duplicates. */
+ if (numbered_arg_count > 1)
+ {
+ unsigned int i, j;
+ bool err;
+
+ qsort (numbered, numbered_arg_count,
+ sizeof (struct numbered_arg), numbered_arg_compare);
+
+ /* Remove duplicates: Copy from i to j, keeping 0 <= j <= i. */
+ err = false;
+ for (i = j = 0; i < numbered_arg_count; i++)
+ if (j > 0 && numbered[i].number == numbered[j-1].number)
+ {
+ enum format_arg_type type1 = numbered[i].type;
+ enum format_arg_type type2 = numbered[j-1].type;
+ enum format_arg_type type_both;
+
+ if (type1 == type2)
+ type_both = type1;
+ else
+ /* Incompatible types. */
+ type_both = type1, err = true;
+
+ numbered[j-1].type = type_both;
+ }
+ else
+ {
+ if (j < i)
+ {
+ numbered[j].number = numbered[i].number;
+ numbered[j].type = numbered[i].type;
+ }
+ j++;
+ }
+ numbered_arg_count = j;
+ if (err)
+ goto bad_format;
+ }
+
+ result = (struct spec *) xmalloc (sizeof (struct spec));
+ result->directives = directives;
+ result->numbered_arg_count = numbered_arg_count;
+ result->allocated = allocated;
+ result->numbered = numbered;
+ return result;
+
+ bad_format:
+ if (numbered != NULL)
+ free (numbered);
+ return NULL;
+}
+
+static void
+format_free (descr)
+ void *descr;
+{
+ struct spec *spec = (struct spec *) descr;
+
+ if (spec->numbered != NULL)
+ free (spec->numbered);
+ free (spec);
+}
+
+static int
+format_get_number_of_directives (descr)
+ void *descr;
+{
+ struct spec *spec = (struct spec *) descr;
+
+ return spec->directives;
+}
+
+static bool
+format_check (pos, msgid_descr, msgstr_descr, equality, noisy, pretty_msgstr)
+ const lex_pos_ty *pos;
+ void *msgid_descr;
+ void *msgstr_descr;
+ bool equality;
+ bool noisy;
+ const char *pretty_msgstr;
+{
+ struct spec *spec1 = (struct spec *) msgid_descr;
+ struct spec *spec2 = (struct spec *) msgstr_descr;
+ bool err = false;
+
+ if (spec1->numbered_arg_count + spec2->numbered_arg_count > 0)
+ {
+ unsigned int i, j;
+ unsigned int n1 = spec1->numbered_arg_count;
+ unsigned int n2 = spec2->numbered_arg_count;
+
+ /* Check the argument names are the same.
+ Both arrays are sorted. We search for the first difference. */
+ for (i = 0, j = 0; i < n1 || j < n2; )
+ {
+ int cmp = (i >= n1 ? 1 :
+ j >= n2 ? -1 :
+ spec1->numbered[i].number > spec2->numbered[j].number ? 1 :
+ spec1->numbered[i].number < spec2->numbered[j].number ? -1 :
+ 0);
+
+ if (cmp > 0)
+ {
+ if (noisy)
+ {
+ error_with_progname = false;
+ error_at_line (0, 0, pos->file_name, pos->line_number,
+ _("a format specification for argument %u, as in '%s', doesn't exist in 'msgid'"),
+ spec2->numbered[j].number, pretty_msgstr);
+ error_with_progname = true;
+ }
+ err = true;
+ break;
+ }
+ else if (cmp < 0)
+ {
+ if (equality)
+ {
+ if (noisy)
+ {
+ error_with_progname = false;
+ error_at_line (0, 0, pos->file_name, pos->line_number,
+ _("a format specification for argument %u doesn't exist in '%s'"),
+ spec1->numbered[i].number, pretty_msgstr);
+ error_with_progname = true;
+ }
+ err = true;
+ break;
+ }
+ else
+ i++;
+ }
+ else
+ j++, i++;
+ }
+ /* Check the argument types are the same. */
+ if (!err)
+ for (i = 0, j = 0; j < n2; )
+ {
+ if (spec1->numbered[i].number == spec2->numbered[j].number)
+ {
+ if (spec1->numbered[i].type != spec2->numbered[j].type)
+ {
+ if (noisy)
+ {
+ error_with_progname = false;
+ error_at_line (0, 0, pos->file_name, pos->line_number,
+ _("format specifications in 'msgid' and '%s' for argument %u are not the same"),
+ pretty_msgstr,
+ spec2->numbered[j].number);
+ error_with_progname = true;
+ }
+ err = true;
+ break;
+ }
+ j++, i++;
+ }
+ else
+ i++;
+ }
+ }
+
+ return err;
+}
+
+
+struct formatstring_parser formatstring_php =
+{
+ format_parse,
+ format_free,
+ format_get_number_of_directives,
+ format_check
+};
+
+
+#ifdef TEST
+
+/* Test program: Print the argument list specification returned by
+ format_parse for strings read from standard input. */
+
+#include <stdio.h>
+#include "getline.h"
+
+static void
+format_print (descr)
+ void *descr;
+{
+ struct spec *spec = (struct spec *) descr;
+ unsigned int last;
+ unsigned int i;
+
+ if (spec == NULL)
+ {
+ printf ("INVALID");
+ return;
+ }
+
+ printf ("(");
+ last = 1;
+ for (i = 0; i < spec->numbered_arg_count; i++)
+ {
+ unsigned int number = spec->numbered[i].number;
+
+ if (i > 0)
+ printf (" ");
+ if (number < last)
+ abort ();
+ for (; last < number; last++)
+ printf ("_ ");
+ switch (spec->numbered[i].type)
+ {
+ case FAT_INTEGER:
+ printf ("i");
+ break;
+ case FAT_FLOAT:
+ printf ("f");
+ break;
+ case FAT_CHARACTER:
+ printf ("c");
+ break;
+ case FAT_STRING:
+ printf ("s");
+ break;
+ default:
+ abort ();
+ }
+ last = number + 1;
+ }
+ printf (")");
+}
+
+int
+main ()
+{
+ for (;;)
+ {
+ char *line = NULL;
+ size_t line_len = 0;
+ void *descr;
+
+ if (getline (&line, &line_len, stdin) < 0)
+ break;
+
+ descr = format_parse (line);
+
+ format_print (descr);
+ printf ("\n");
+
+ free (line);
+ }
+
+ return 0;
+}
+
+/*
+ * For Emacs M-x compile
+ * Local Variables:
+ * compile-command: "/bin/sh ../libtool --mode=link gcc -o a.out -static -O -g -Wall -I.. -I../lib -I../intl -DHAVE_CONFIG_H -DTEST format-php.c ../lib/libgettextlib.la"
+ * End:
+ */
+
+#endif /* TEST */
/* format_awk */ &formatstring_awk,
/* format_pascal */ &formatstring_pascal,
/* format_ycp */ &formatstring_ycp,
- /* format_tcl */ &formatstring_tcl
+ /* format_tcl */ &formatstring_tcl,
+ /* format_php */ &formatstring_php
};
extern struct formatstring_parser formatstring_pascal;
extern struct formatstring_parser formatstring_ycp;
extern struct formatstring_parser formatstring_tcl;
+extern struct formatstring_parser formatstring_php;
/* Table of all format string parsers. */
extern struct formatstring_parser *formatstring_parsers[NFORMATS];
/* format_awk */ "awk",
/* format_pascal */ "object-pascal",
/* format_ycp */ "ycp",
- /* format_tcl */ "tcl"
+ /* format_tcl */ "tcl",
+ /* format_php */ "php"
};
const char *const format_language_pretty[NFORMATS] =
/* format_awk */ "awk",
/* format_pascal */ "Object Pascal",
/* format_ycp */ "YCP",
- /* format_tcl */ "Tcl"
+ /* format_tcl */ "Tcl",
+ /* format_php */ "PHP"
};
format_awk,
format_pascal,
format_ycp,
- format_tcl
+ format_tcl,
+ format_php
};
-#define NFORMATS 11 /* Number of format_type enum values. */
+#define NFORMATS 12 /* Number of format_type enum values. */
extern const char *const format_language[NFORMATS];
extern const char *const format_language_pretty[NFORMATS];
--- /dev/null
+/* xgettext PHP backend.
+ Copyright (C) 2001-2002 Free Software Foundation, Inc.
+
+ This file was written by Bruno Haible <bruno@clisp.org>, 2002.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
+
+#ifdef HAVE_CONFIG_H
+# include "config.h"
+#endif
+
+#include <errno.h>
+#include <stdbool.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "message.h"
+#include "x-php.h"
+#include "xgettext.h"
+#include "error.h"
+#include "xmalloc.h"
+#include "exit.h"
+#include "gettext.h"
+
+#define _(s) gettext(s)
+
+
+/* The PHP syntax is defined in phpdoc/manual/langref.html.
+ See also php-4.1.0/Zend/zend_language_scanner.l. */
+
+enum token_type_ty
+{
+ token_type_eof,
+ token_type_lparen, /* ( */
+ token_type_rparen, /* ) */
+ token_type_comma, /* , */
+ token_type_string_literal, /* "abc" */
+ token_type_symbol, /* symbol, number */
+ token_type_other /* misc. operator */
+};
+typedef enum token_type_ty token_type_ty;
+
+typedef struct token_ty token_ty;
+struct token_ty
+{
+ token_type_ty type;
+ char *string; /* for token_type_string_literal, token_type_symbol */
+ int line_number;
+};
+
+
+/* Prototypes for local functions. Needed to ensure compiler checking of
+ function argument counts despite of K&R C function definition syntax. */
+static void init_keywords PARAMS ((void));
+static int phase1_getc PARAMS ((void));
+static void phase1_ungetc PARAMS ((int c));
+static void skip_html PARAMS ((void));
+#if 0
+static int phase2_getc PARAMS ((void));
+static void phase2_ungetc PARAMS ((int c));
+#endif
+static inline void comment_start PARAMS ((void));
+static inline void comment_add PARAMS ((int c));
+static inline void comment_line_end PARAMS ((size_t chars_to_remove));
+static int phase3_getc PARAMS ((void));
+static void phase3_ungetc PARAMS ((int c));
+static inline void free_token PARAMS ((token_ty *tp));
+static void x_php_lex PARAMS ((token_ty *tp));
+static bool extract_parenthesized PARAMS ((message_list_ty *mlp,
+ int commas_to_skip,
+ int plural_commas));
+
+
+/* ====================== Keyword set customization. ====================== */
+
+/* If true extract all strings. */
+static bool extract_all = false;
+
+static hash_table keywords;
+static bool default_keywords = true;
+
+
+void
+x_php_extract_all ()
+{
+ extract_all = true;
+}
+
+
+void
+x_php_keyword (name)
+ const char *name;
+{
+ if (name == NULL)
+ default_keywords = false;
+ else
+ {
+ const char *end;
+ int argnum1;
+ int argnum2;
+ const char *colon;
+
+ if (keywords.table == NULL)
+ init_hash (&keywords, 100);
+
+ split_keywordspec (name, &end, &argnum1, &argnum2);
+
+ /* The characters between name and end should form a valid C identifier.
+ A colon means an invalid parse in split_keywordspec(). */
+ colon = strchr (name, ':');
+ if (colon == NULL || colon >= end)
+ {
+ if (argnum1 == 0)
+ argnum1 = 1;
+ insert_entry (&keywords, name, end - name,
+ (void *) (long) (argnum1 + (argnum2 << 10)));
+ }
+ }
+}
+
+/* Finish initializing the keywords hash table.
+ Called after argument processing, before each file is processed. */
+static void
+init_keywords ()
+{
+ if (default_keywords)
+ {
+ x_php_keyword ("_");
+ x_php_keyword ("gettext");
+ x_php_keyword ("dgettext:2");
+ x_php_keyword ("dcgettext:2");
+ default_keywords = false;
+ }
+}
+
+
+/* ======================== Reading of characters. ======================== */
+
+
+/* Real filename, used in error messages about the input file. */
+static const char *real_file_name;
+
+/* Logical filename and line number, used to label the extracted messages. */
+static char *logical_file_name;
+static int line_number;
+
+/* The input file stream. */
+static FILE *fp;
+
+
+/* 1. line_number handling. */
+
+/* Maximum used guaranteed to be < 4. */
+static unsigned char phase1_pushback[4];
+static int phase1_pushback_length;
+
+static int
+phase1_getc ()
+{
+ int c;
+
+ if (phase1_pushback_length)
+ c = phase1_pushback[--phase1_pushback_length];
+ else
+ {
+ c = getc (fp);
+
+ if (c == EOF)
+ {
+ if (ferror (fp))
+ error (EXIT_FAILURE, errno, _("error while reading \"%s\""),
+ real_file_name);
+ return EOF;
+ }
+ }
+
+ if (c == '\n')
+ line_number++;
+
+ return c;
+}
+
+static void
+phase1_ungetc (c)
+ int c;
+{
+ if (c != EOF)
+ {
+ if (c == '\n')
+ --line_number;
+
+ phase1_pushback[phase1_pushback_length++] = c;
+ }
+}
+
+
+/* 2. Ignore HTML sections. They are equivalent to PHP echo commands and
+ therefore don't contain translatable strings. */
+
+static void
+skip_html ()
+{
+ for (;;)
+ {
+ int c = phase1_getc ();
+
+ if (c == EOF)
+ return;
+
+ if (c == '<')
+ {
+ int c2 = phase1_getc ();
+
+ if (c2 == EOF)
+ break;
+
+ if (c2 == '?')
+ {
+ /* <?php is the normal way to enter PHP mode. <? and <?= are
+ recognized by PHP depending on a configuration setting. */
+ int c3 = phase1_getc ();
+
+ if (c3 != '=')
+ phase1_ungetc (c3);
+
+ return;
+ }
+
+ if (c2 == '%')
+ {
+ /* <% and <%= are recognized by PHP depending on a configuration
+ setting. */
+ int c3 = phase1_getc ();
+
+ if (c3 != '=')
+ phase1_ungetc (c3);
+
+ return;
+ }
+
+ if (c2 == '<')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+
+ /* < script language = php >
+ < script language = "php" >
+ < script language = 'php' >
+ are always recognized. */
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 != 's' && c2 != 'S')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'c' && c2 != 'C')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'r' && c2 != 'R')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'i' && c2 != 'I')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'p' && c2 != 'P')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 't' && c2 != 'T')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (!(c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r'))
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ do
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r');
+ if (c2 != 'l' && c2 != 'L')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'a' && c2 != 'A')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'n' && c2 != 'N')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'g' && c2 != 'G')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'u' && c2 != 'U')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'a' && c2 != 'A')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'g' && c2 != 'G')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'e' && c2 != 'E')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 != '=')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 == '"')
+ {
+ c2 = phase1_getc ();
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'h')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != '"')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ }
+ else if (c2 == '\'')
+ {
+ c2 = phase1_getc ();
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'h')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != '\'')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ }
+ else
+ {
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'h')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ c2 = phase1_getc ();
+ if (c2 != 'p')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ }
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 != '>')
+ {
+ phase1_ungetc (c2);
+ continue;
+ }
+ return;
+ }
+ }
+}
+
+#if 0
+
+static unsigned char phase2_pushback[1];
+static int phase2_pushback_length;
+
+static int
+phase2_getc ()
+{
+ int c;
+
+ if (phase2_pushback_length)
+ return phase2_pushback[--phase2_pushback_length];
+
+ c = phase1_getc ();
+ switch (c)
+ {
+ case '?':
+ case '%':
+ {
+ int c2 = phase1_getc ();
+ if (c2 == '>')
+ {
+ /* ?> and %> terminate PHP mode and switch back to HTML mode. */
+ skip_html ();
+ return ' ';
+ }
+ phase1_ungetc (c2);
+ }
+ break;
+
+ case '<':
+ {
+ int c2 = phase1_getc ();
+
+ /* < / script > terminates PHP mode and switches back to HTML mode. */
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 == '/')
+ {
+ do
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r');
+ if (c2 == 's' || c2 == 'S')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'c' || c2 == 'C')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'r' || c2 == 'R')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'i' || c2 == 'I')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'p' || c2 == 'P')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 't' || c2 == 'T')
+ {
+ do
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t'
+ || c2 == '\n' || c2 == '\r');
+ if (c2 == '>')
+ {
+ skip_html ();
+ return ' ';
+ }
+ }
+ }
+ }
+ }
+ }
+ }
+ }
+ phase1_ungetc (c2);
+ }
+ break;
+ }
+
+ return c;
+}
+
+static void
+phase2_ungetc (c)
+ int c;
+{
+ if (c != EOF)
+ phase2_pushback[phase2_pushback_length++] = c;
+}
+
+#endif
+
+
+/* Accumulating comments. */
+
+static char *buffer;
+static size_t bufmax;
+static size_t buflen;
+
+static inline void
+comment_start ()
+{
+ buflen = 0;
+}
+
+static inline void
+comment_add (c)
+ int c;
+{
+ if (buflen >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[buflen++] = c;
+}
+
+static inline void
+comment_line_end (chars_to_remove)
+ size_t chars_to_remove;
+{
+ buflen -= chars_to_remove;
+ while (buflen >= 1
+ && (buffer[buflen - 1] == ' ' || buffer[buflen - 1] == '\t'))
+ --buflen;
+ if (chars_to_remove == 0 && buflen >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[buflen] = '\0';
+ xgettext_comment_add (buffer);
+}
+
+
+/* 3. Replace each comment that is not inside a string literal with a
+ space character. We need to remember the comment for later, because
+ it may be attached to a keyword string. */
+
+/* These are for tracking whether comments count as immediately before
+ keyword. */
+static int last_comment_line;
+static int last_non_comment_line;
+
+static unsigned char phase3_pushback[1];
+static int phase3_pushback_length;
+
+static int
+phase3_getc ()
+{
+ int lineno;
+ int c;
+
+ if (phase3_pushback_length)
+ return phase3_pushback[--phase3_pushback_length];
+
+ c = phase1_getc ();
+
+ if (c == '#')
+ {
+ /* sh comment. */
+ bool last_was_qmark;
+
+ comment_start ();
+ lineno = line_number;
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == '\n' || c == EOF)
+ {
+ comment_line_end (0);
+ break;
+ }
+ if (last_was_qmark && c == '>')
+ {
+ comment_line_end (1);
+ skip_html ();
+ break;
+ }
+ /* We skip all leading white space, but not EOLs. */
+ if (!(buflen == 0 && (c == ' ' || c == '\t')))
+ comment_add (c);
+ last_was_qmark = (c == '?' || c == '%');
+ }
+ last_comment_line = lineno;
+ return '\n';
+ }
+ else if (c == '/')
+ {
+ c = phase1_getc ();
+
+ switch (c)
+ {
+ default:
+ phase1_ungetc (c);
+ return '/';
+
+ case '*':
+ {
+ /* C comment. */
+ bool last_was_star;
+
+ comment_start ();
+ lineno = line_number;
+ last_was_star = false;
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == EOF)
+ break;
+ /* We skip all leading white space, but not EOLs. */
+ if (buflen == 0 && (c == ' ' || c == '\t'))
+ continue;
+ comment_add (c);
+ switch (c)
+ {
+ case '\n':
+ comment_line_end (1);
+ comment_start ();
+ lineno = line_number;
+ last_was_star = false;
+ continue;
+
+ case '*':
+ last_was_star = true;
+ continue;
+
+ case '/':
+ if (last_was_star)
+ {
+ comment_line_end (2);
+ break;
+ }
+ /* FALLTHROUGH */
+
+ default:
+ last_was_star = false;
+ continue;
+ }
+ break;
+ }
+ last_comment_line = lineno;
+ return ' ';
+ }
+
+ case '/':
+ {
+ /* C++ comment. */
+ bool last_was_qmark;
+
+ comment_start ();
+ lineno = line_number;
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == '\n' || c == EOF)
+ {
+ comment_line_end (0);
+ break;
+ }
+ if (last_was_qmark && c == '>')
+ {
+ comment_line_end (1);
+ skip_html ();
+ break;
+ }
+ /* We skip all leading white space, but not EOLs. */
+ if (!(buflen == 0 && (c == ' ' || c == '\t')))
+ comment_add (c);
+ last_was_qmark = (c == '?' || c == '%');
+ }
+ last_comment_line = lineno;
+ return '\n';
+ }
+ }
+ }
+ else
+ return c;
+}
+
+static void
+phase3_ungetc (c)
+ int c;
+{
+ if (c != EOF)
+ phase3_pushback[phase3_pushback_length++] = c;
+}
+
+
+/* Free the memory pointed to by a 'struct token_ty'. */
+static inline void
+free_token (tp)
+ token_ty *tp;
+{
+ if (tp->type == token_type_string_literal || tp->type == token_type_symbol)
+ free (tp->string);
+}
+
+
+/* 4. Combine characters into tokens. Discard whitespace. */
+
+static void
+x_php_lex (tp)
+ token_ty *tp;
+{
+ static char *buffer;
+ static int bufmax;
+ int bufpos;
+ int c;
+
+ tp->string = NULL;
+
+ for (;;)
+ {
+ tp->line_number = line_number;
+ c = phase3_getc ();
+ switch (c)
+ {
+ case EOF:
+ tp->type = token_type_eof;
+ return;
+
+ case '\n':
+ if (last_non_comment_line > last_comment_line)
+ xgettext_comment_reset ();
+ /* FALLTHROUGH */
+ case ' ':
+ case '\t':
+ case '\r':
+ /* Ignore whitespace. */
+ continue;
+ }
+
+ last_non_comment_line = tp->line_number;
+
+ switch (c)
+ {
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
+ case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
+ case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
+ case 'V': case 'W': case 'X': case 'Y': case 'Z':
+ case '_':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
+ case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
+ case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
+ case 'v': case 'w': case 'x': case 'y': case 'z':
+ bufpos = 0;
+ for (;;)
+ {
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos++] = c;
+ c = phase1_getc ();
+ switch (c)
+ {
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case '_':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ case '0': case '1': case '2': case '3': case '4':
+ case '5': case '6': case '7': case '8': case '9':
+ continue;
+
+ default:
+ phase1_ungetc (c);
+ break;
+ }
+ break;
+ }
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos] = 0;
+ tp->string = xstrdup (buffer);
+ tp->type = token_type_symbol;
+ return;
+
+ case '\'':
+ /* Single-quoted string literal. */
+ bufpos = 0;
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == EOF || c == '\'')
+ break;
+ if (c == '\\')
+ {
+ c = phase1_getc ();
+ if (c != '\\' && c != '\'')
+ {
+ phase1_ungetc (c);
+ c = '\\';
+ }
+ }
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos++] = c;
+ }
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos] = 0;
+ tp->type = token_type_string_literal;
+ tp->string = xstrdup (buffer);
+ return;
+
+ case '"':
+ /* Double-quoted string literal. */
+ tp->type = token_type_string_literal;
+ bufpos = 0;
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == EOF || c == '"')
+ break;
+ if (c == '$')
+ {
+ c = phase1_getc ();
+ if ((c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z')
+ || c == '_' || c == '{' || c >= 0x7f)
+ {
+ /* String with variables. */
+ tp->type = token_type_other;
+ continue;
+ }
+ phase1_ungetc (c);
+ c = '$';
+ }
+ if (c == '{')
+ {
+ c = phase1_getc ();
+ if (c == '$')
+ {
+ /* String with expressions. */
+ tp->type = token_type_other;
+ continue;
+ }
+ phase1_ungetc (c);
+ c = '{';
+ }
+ if (c == '\\')
+ {
+ int n, j;
+
+ c = phase1_getc ();
+ switch (c)
+ {
+ case '"':
+ case '\\':
+ case '$':
+ break;
+
+ case '0': case '1': case '2': case '3':
+ case '4': case '5': case '6': case '7':
+ n = 0;
+ for (j = 0; j < 3; ++j)
+ {
+ n = n * 8 + c - '0';
+ c = phase1_getc ();
+ switch (c)
+ {
+ default:
+ break;
+
+ case '0': case '1': case '2': case '3':
+ case '4': case '5': case '6': case '7':
+ continue;
+ }
+ break;
+ }
+ phase1_ungetc (c);
+ c = n;
+ break;
+
+ case 'x':
+ n = 0;
+ for (j = 0; j < 2; ++j)
+ {
+ c = phase1_getc ();
+ switch (c)
+ {
+ case '0': case '1': case '2': case '3': case '4':
+ case '5': case '6': case '7': case '8': case '9':
+ n = n * 16 + c - '0';
+ break;
+ case 'A': case 'B': case 'C': case 'D': case 'E':
+ case 'F':
+ n = n * 16 + 10 + c - 'A';
+ break;
+ case 'a': case 'b': case 'c': case 'd': case 'e':
+ case 'f':
+ n = n * 16 + 10 + c - 'a';
+ break;
+ default:
+ phase1_ungetc (c);
+ c = 0;
+ break;
+ }
+ if (c == 0)
+ break;
+ }
+ if (j == 0)
+ {
+ phase1_ungetc ('x');
+ c = '\\';
+ }
+ else
+ c = n;
+ break;
+
+ case 'n':
+ c = '\n';
+ break;
+ case 't':
+ c = '\t';
+ break;
+ case 'r':
+ c = '\r';
+ break;
+
+ default:
+ phase1_ungetc (c);
+ c = '\\';
+ break;
+ }
+ }
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos++] = c;
+ }
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos] = 0;
+ if (tp->type == token_type_string_literal)
+ tp->string = xstrdup (buffer);
+ return;
+
+ case '?':
+ case '%':
+ {
+ int c2 = phase1_getc ();
+ if (c2 == '>')
+ {
+ /* ?> and %> terminate PHP mode and switch back to HTML
+ mode. */
+ skip_html ();
+ }
+ else
+ phase1_ungetc (c2);
+ tp->type = token_type_other;
+ return;
+ }
+
+ case '(':
+ tp->type = token_type_lparen;
+ return;
+
+ case ')':
+ tp->type = token_type_rparen;
+ return;
+
+ case ',':
+ tp->type = token_type_comma;
+ return;
+
+ case '<':
+ {
+ int c2 = phase1_getc ();
+ if (c2 == '<')
+ {
+ int c3 = phase1_getc ();
+ if (c3 == '<')
+ {
+ /* Start of here document.
+ Parse whitespace, then label, then newline. */
+ do
+ c = phase3_getc ();
+ while (c == ' ' || c == '\t' || c == '\n' || c == '\r');
+
+ bufpos = 0;
+ do
+ {
+ if (bufpos >= bufmax)
+ {
+ bufmax += 100;
+ buffer = xrealloc (buffer, bufmax);
+ }
+ buffer[bufpos++] = c;
+ c = phase3_getc ();
+ }
+ while (c != EOF && c != '\n' && c != '\r');
+ /* buffer[0..bufpos-1] now contains the label. */
+
+ /* Now skip the here document. */
+ for (;;)
+ {
+ c = phase1_getc ();
+ if (c == EOF)
+ break;
+ if (c == '\n' || c == '\r')
+ {
+ int bufidx = 0;
+
+ while (bufidx < bufpos)
+ {
+ c = phase1_getc ();
+ if (c == EOF)
+ break;
+ if (c != buffer[bufidx])
+ {
+ phase1_ungetc (c);
+ break;
+ }
+ }
+ c = phase1_getc ();
+ if (c != ';')
+ phase1_ungetc (c);
+ c = phase1_getc ();
+ if (c == '\n' || c == '\r')
+ break;
+ }
+ }
+
+ /* FIXME: Ideally we should turn the here document into a
+ string literal if it didn't contain $ substitution. And
+ we should also respect backslash escape sequences like
+ in double-quoted strings. */
+ tp->type = token_type_other;
+ return;
+ }
+ phase1_ungetc (c3);
+ }
+
+ /* < / script > terminates PHP mode and switches back to HTML
+ mode. */
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r')
+ c2 = phase1_getc ();
+ if (c2 == '/')
+ {
+ do
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t' || c2 == '\n' || c2 == '\r');
+ if (c2 == 's' || c2 == 'S')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'c' || c2 == 'C')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'r' || c2 == 'R')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'i' || c2 == 'I')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 'p' || c2 == 'P')
+ {
+ c2 = phase1_getc ();
+ if (c2 == 't' || c2 == 'T')
+ {
+ do
+ c2 = phase1_getc ();
+ while (c2 == ' ' || c2 == '\t'
+ || c2 == '\n' || c2 == '\r');
+ if (c2 == '>')
+ {
+ skip_html ();
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+ }
+ else
+ phase1_ungetc (c2);
+
+ tp->type = token_type_other;
+ return;
+ }
+
+ case '`':
+ /* Execution operator. */
+ default:
+ /* We could carefully recognize each of the 2 and 3 character
+ operators, but it is not necessary, as we only need to recognize
+ gettext invocations. Don't bother. */
+ tp->type = token_type_other;
+ return;
+ }
+ }
+}
+
+/* ========================= Extracting strings. ========================== */
+
+/* The file is broken into tokens. Scan the token stream, looking for
+ a keyword, followed by a left paren, followed by a string. When we
+ see this sequence, we have something to remember. We assume we are
+ looking at a valid C or C++ program, and leave the complaints about
+ the grammar to the compiler.
+
+ Normal handling: Look for
+ keyword ( ... msgid ... )
+ Plural handling: Look for
+ keyword ( ... msgid ... msgid_plural ... )
+
+ We use recursion because the arguments before msgid or between msgid
+ and msgid_plural can contain subexpressions of the same form. */
+
+
+/* Extract messages until the next balanced closing parenthesis.
+ Extracted messages are added to MLP.
+ When a specific argument shall be extracted, COMMAS_TO_SKIP >= 0 and,
+ if also a plural argument shall be extracted, PLURAL_COMMAS > 0,
+ otherwise PLURAL_COMMAS = 0.
+ When no specific argument shall be extracted, COMMAS_TO_SKIP < 0.
+ Return true upon eof, false upon closing parenthesis. */
+static bool
+extract_parenthesized (mlp, commas_to_skip, plural_commas)
+ message_list_ty *mlp;
+ int commas_to_skip;
+ int plural_commas;
+{
+ /* Remember the message containing the msgid, for msgid_plural. */
+ message_ty *plural_mp = NULL;
+
+ /* 0 when no keyword has been seen. 1 right after a keyword is seen. */
+ int state;
+ /* Parameters of the keyword just seen. Defined only in state 1. */
+ int next_commas_to_skip = -1;
+ int next_plural_commas = 0;
+
+ /* Start state is 0. */
+ state = 0;
+
+ for (;;)
+ {
+ token_ty token;
+
+ x_php_lex (&token);
+ switch (token.type)
+ {
+ case token_type_symbol:
+ /* No need to bother if we extract all strings anyway. */
+ if (!extract_all)
+ {
+ void *keyword_value;
+
+ if (find_entry (&keywords, token.string, strlen (token.string),
+ &keyword_value)
+ == 0)
+ {
+ int argnum1 = (int) (long) keyword_value & ((1 << 10) - 1);
+ int argnum2 = (int) (long) keyword_value >> 10;
+
+ next_commas_to_skip = argnum1 - 1;
+ next_plural_commas = (argnum2 > argnum1 ? argnum2 - argnum1 : 0);
+ state = 1;
+ }
+ else
+ state = 0;
+ }
+ free (token.string);
+ continue;
+
+ case token_type_lparen:
+ /* No need to recurse if we extract all strings anyway. */
+ if (extract_all)
+ continue;
+ if (state
+ ? extract_parenthesized (mlp, next_commas_to_skip,
+ next_plural_commas)
+ : extract_parenthesized (mlp, -1, 0))
+ return true;
+ state = 0;
+ continue;
+
+ case token_type_rparen:
+ /* No need to return if we extract all strings anyway. */
+ if (extract_all)
+ continue;
+ return false;
+
+ case token_type_comma:
+ /* No need to bother if we extract all strings anyway. */
+ if (extract_all)
+ continue;
+ if (commas_to_skip >= 0)
+ {
+ if (commas_to_skip > 0)
+ commas_to_skip--;
+ else
+ if (plural_mp != NULL && plural_commas > 0)
+ {
+ commas_to_skip = plural_commas - 1;
+ plural_commas = 0;
+ }
+ else
+ commas_to_skip = -1;
+ }
+ state = 0;
+ continue;
+
+ case token_type_string_literal:
+ {
+ lex_pos_ty pos;
+ pos.file_name = logical_file_name;
+ pos.line_number = token.line_number;
+
+ if (extract_all)
+ remember_a_message (mlp, token.string, &pos);
+ else
+ {
+ if (commas_to_skip == 0)
+ {
+ if (plural_mp == NULL)
+ {
+ /* Seen an msgid. */
+ message_ty *mp = remember_a_message (mlp, token.string,
+ &pos);
+ if (plural_commas > 0)
+ plural_mp = mp;
+ }
+ else
+ {
+ /* Seen an msgid_plural. */
+ remember_a_message_plural (plural_mp, token.string,
+ &pos);
+ plural_mp = NULL;
+ }
+ }
+ else
+ free (token.string);
+ state = 0;
+ }
+ continue;
+ }
+
+ case token_type_other:
+ state = 0;
+ continue;
+
+ case token_type_eof:
+ return true;
+
+ default:
+ abort ();
+ }
+ }
+}
+
+
+void
+extract_php (f, real_filename, logical_filename, mdlp)
+ FILE *f;
+ const char *real_filename;
+ const char *logical_filename;
+ msgdomain_list_ty *mdlp;
+{
+ message_list_ty *mlp = mdlp->item[0]->messages;
+
+ fp = f;
+ real_file_name = real_filename;
+ logical_file_name = xstrdup (logical_filename);
+ line_number = 1;
+
+ last_comment_line = -1;
+ last_non_comment_line = -1;
+
+ init_keywords ();
+
+ /* Initial mode is HTML mode, not PHP mode. */
+ skip_html ();
+
+ /* Eat tokens until eof is seen. When extract_parenthesized returns
+ due to an unbalanced closing parenthesis, just restart it. */
+ while (!extract_parenthesized (mlp, -1, 0))
+ ;
+
+ /* Close scanner. */
+ fp = NULL;
+ real_file_name = NULL;
+ logical_file_name = NULL;
+ line_number = 0;
+}
--- /dev/null
+/* xgettext PHP backend.
+ Copyright (C) 2002 Free Software Foundation, Inc.
+ Written by Bruno Haible <bruno@clisp.org>, 2002.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
+
+
+#define EXTENSIONS_PHP \
+ { "php", "PHP" }, \
+ { "php3", "PHP" }, \
+ { "php4", "PHP" }, \
+
+#define SCANNERS_PHP \
+ { "PHP", extract_php, &formatstring_php }, \
+
+/* Scan a PHP file and add its translatable strings to mdlp. */
+extern void extract_php PARAMS ((FILE *fp, const char *real_filename,
+ const char *logical_filename,
+ msgdomain_list_ty *mdlp));
+
+extern void x_php_keyword PARAMS ((const char *keyword));
+extern void x_php_extract_all PARAMS ((void));
#include "x-awk.h"
#include "x-ycp.h"
#include "x-tcl.h"
+#include "x-php.h"
#include "x-rst.h"
#include "x-glade.h"
x_java_extract_all ();
x_awk_extract_all ();
x_tcl_extract_all ();
+ x_php_extract_all ();
x_glade_extract_all ();
break;
case 'c':
x_java_keyword (optarg);
x_awk_keyword (optarg);
x_tcl_keyword (optarg);
+ x_php_keyword (optarg);
x_glade_keyword (optarg);
}
break;
-L, --language=NAME recognise the specified language\n\
(C, C++, ObjectiveC, PO, Python, Lisp,\n\
EmacsLisp, librep, Java, awk, YCP, Tcl,\n\
- RST, Glade)\n\
+ PHP, RST, Glade)\n\
-C, --c++ shorthand for --language=C++\n\
By default the language is guessed depending on the input file name extension.\n\
"));
SCANNERS_AWK
SCANNERS_YCP
SCANNERS_TCL
+ SCANNERS_PHP
SCANNERS_RST
SCANNERS_GLADE
/* Here will follow more languages and their scanners: perl, etc...
EXTENSIONS_AWK
EXTENSIONS_YCP
EXTENSIONS_TCL
+ EXTENSIONS_PHP
EXTENSIONS_RST
EXTENSIONS_GLADE
/* Here will follow more file extensions: sh, pl ... */
+2002-08-18 Bruno Haible <bruno@clisp.org>
+
+ * format-php-1: New file.
+ * format-php-2: New file.
+ * lang-php: New file.
+ * Makefile.am (TESTS): Add them.
+
+2002-08-06 Bruno Haible <bruno@clisp.org>
+
+ * gettext-0.11.5 released.
+
2002-08-06 Bruno Haible <bruno@clisp.org>
* msgunfmt-2: Fix typo.
format-java-1 format-java-2 \
format-librep-1 format-librep-2 \
format-lisp-1 format-lisp-2 \
+ format-php-1 format-php-2 \
format-python-1 format-python-2 \
format-pascal-1 format-pascal-2 \
format-tcl-1 format-tcl-2 \
format-ycp-1 format-ycp-2 \
plural-1 plural-2 \
- lang-c lang-c++ lang-objc lang-python lang-clisp lang-elisp lang-librep lang-java lang-gawk lang-pascal lang-ycp lang-tcl lang-po lang-rst \
+ lang-c lang-c++ lang-objc lang-python lang-clisp lang-elisp lang-librep lang-java lang-gawk lang-pascal lang-ycp lang-tcl lang-php lang-po lang-rst \
rpath-1a rpath-1b \
rpath-2aaa rpath-2aab rpath-2aac rpath-2aad \
rpath-2aba rpath-2abb rpath-2abc rpath-2abd \
--- /dev/null
+#! /bin/sh
+
+# Test recognition of PHP format strings.
+
+tmpfiles=""
+trap 'rm -fr $tmpfiles' 1 2 3 15
+
+tmpfiles="$tmpfiles f-ph-1.data"
+cat <<\EOF > f-ph-1.data
+# Valid: no argument
+"abc%%"
+# Valid: one string argument
+"abc%s"
+# Valid: one integer argument
+"abc%b"
+# Valid: one integer argument
+"abc%d"
+# Valid: one integer argument
+"abc%u"
+# Valid: one integer argument
+"abc%o"
+# Valid: one integer argument
+"abc%x"
+# Valid: one integer argument
+"abc%X"
+# Valid: one floating-point argument
+"abc%e"
+# Valid: one floating-point argument
+"abc%f"
+# Valid: one character argument
+"abc%c"
+# Valid: one argument with flags
+"abc%-f"
+# Valid: one argument with padding flags
+"abc%'=f"
+# Valid: one argument with width
+"abc%2f"
+# Valid: one argument with precision
+"abc%.4f"
+# Valid: one argument with width and precision
+"abc%14.4f"
+# Invalid: unterminated
+"abc%"
+# Invalid: unknown format specifier
+"abc%y"
+# Invalid: unknown format specifier
+"abc%F"
+# Invalid: flags after width
+"abc%5-f"
+# Invalid: twice precision
+"abc%.4.2f"
+# Valid: three arguments
+"abc%d%x%x"
+# Valid: a numbered argument
+"abc%1$d"
+# Invalid: zero
+"abc%0$d"
+# Valid: two-digit numbered arguments
+"abc%11$def%10$dgh%9$dij%8$dkl%7$dmn%6$dop%5$dqr%4$dst%3$duv%2$dwx%1$dyz"
+# Invalid: unterminated number
+"abc%1"
+# Invalid: flags before number
+"abc%-1$d"
+# Valid: three arguments, two with same number
+"abc%1$4x,%2$c,%1$u"
+# Invalid: argument with conflicting types
+"abc%1$4x,%2$c,%1$s"
+# Valid: no conflict
+"abc%1$4x,%2$c,%1$u"
+# Valid: mixing of numbered and unnumbered arguments
+"abc%d%2$x"
+# Valid: numbered argument with constant precision
+"abc%1$.9x"
+# Valid: missing non-final argument
+"abc%2$x%3$s"
+# Valid: permutation
+"abc%2$ddef%1$d"
+# Valid: multiple uses of same argument
+"abc%2$xdef%1$sghi%2$x"
+EOF
+
+: ${XGETTEXT=xgettext}
+n=0
+while read comment; do
+ read string
+ n=`expr $n + 1`
+ tmpfiles="$tmpfiles f-ph-1-$n.in f-ph-1-$n.po"
+ echo "<?= gettext(${string}) ?>" | sed -e 's/\$/\\\$/g' > f-ph-1-$n.in
+ ${XGETTEXT} -L PHP -o f-ph-1-$n.po f-ph-1-$n.in || exit 1
+ test -f f-ph-1-$n.po || exit 1
+ fail=
+ if echo "$comment" | grep 'Valid:' > /dev/null; then
+ if grep php-format f-ph-1-$n.po > /dev/null; then
+ :
+ else
+ fail=yes
+ fi
+ else
+ if grep php-format f-ph-1-$n.po > /dev/null; then
+ fail=yes
+ else
+ :
+ fi
+ fi
+ if test -n "$fail"; then
+ echo "Format string recognition error:" 1>&2
+ cat f-ph-1-$n.in 1>&2
+ echo "Got:" 1>&2
+ cat f-ph-1-$n.po 1>&2
+ exit 1
+ fi
+done < f-ph-1.data
+
+rm -fr $tmpfiles
+
+exit 0
--- /dev/null
+#! /bin/sh
+
+# Test checking of PHP format strings.
+
+tmpfiles=""
+trap 'rm -fr $tmpfiles' 1 2 3 15
+
+tmpfiles="$tmpfiles f-ph-2.data"
+cat <<\EOF > f-ph-2.data
+# Valid: %% doesn't count
+msgid "abc%%def"
+msgstr "xyz"
+# Invalid: invalid msgstr
+msgid "abc%%def"
+msgstr "xyz%"
+# Valid: same arguments
+msgid "abc%s%gdef"
+msgstr "xyz%s%g"
+# Valid: same arguments, with different widths
+msgid "abc%2sdef"
+msgstr "xyz%3s"
+# Valid: same arguments but in numbered syntax
+msgid "abc%s%gdef"
+msgstr "xyz%1$s%2$g"
+# Valid: permutation
+msgid "abc%s%g%cdef"
+msgstr "xyz%3$c%2$g%1$s"
+# Invalid: too few arguments
+msgid "abc%2$udef%1$s"
+msgstr "xyz%1$s"
+# Invalid: too few arguments
+msgid "abc%sdef%u"
+msgstr "xyz%s"
+# Invalid: too many arguments
+msgid "abc%udef"
+msgstr "xyz%uvw%c"
+# Valid: same numbered arguments, with different widths
+msgid "abc%2$5s%1$4s"
+msgstr "xyz%2$4s%1$5s"
+# Invalid: missing argument
+msgid "abc%2$sdef%1$u"
+msgstr "xyz%1$u"
+# Invalid: missing argument
+msgid "abc%1$sdef%2$u"
+msgstr "xyz%2$u"
+# Invalid: added argument
+msgid "abc%1$udef"
+msgstr "xyz%1$uvw%2$c"
+# Valid: type compatibility
+msgid "abc%b"
+msgstr "xyz%d"
+# Valid: type compatibility
+msgid "abc%u"
+msgstr "xyz%d"
+# Valid: type compatibility
+msgid "abc%o"
+msgstr "xyz%d"
+# Valid: type compatibility
+msgid "abc%x"
+msgstr "xyz%d"
+# Valid: type compatibility
+msgid "abc%X"
+msgstr "xyz%d"
+# Valid: type compatibility
+msgid "abc%e"
+msgstr "xyz%f"
+# Invalid: type incompatibility
+msgid "abc%s"
+msgstr "xyz%d"
+# Invalid: type incompatibility
+msgid "abc%s"
+msgstr "xyz%e"
+# Invalid: type incompatibility
+msgid "abc%s"
+msgstr "xyz%c"
+# Invalid: type incompatibility
+msgid "abc%d"
+msgstr "xyz%e"
+# Invalid: type incompatibility
+msgid "abc%d"
+msgstr "xyz%c"
+# Invalid: type incompatibility
+msgid "abc%e"
+msgstr "xyz%c"
+EOF
+
+: ${MSGFMT=msgfmt}
+n=0
+while read comment; do
+ read msgid_line
+ read msgstr_line
+ n=`expr $n + 1`
+ tmpfiles="$tmpfiles f-ph-2-$n.po f-ph-2-$n.mo"
+ cat <<EOF > f-ph-2-$n.po
+#, php-format
+${msgid_line}
+${msgstr_line}
+EOF
+ fail=
+ if echo "$comment" | grep 'Valid:' > /dev/null; then
+ if ${MSGFMT} --check-format -o f-ph-2-$n.mo f-ph-2-$n.po; then
+ :
+ else
+ fail=yes
+ fi
+ else
+ ${MSGFMT} --check-format -o f-ph-2-$n.mo f-ph-2-$n.po 2> /dev/null
+ if test $? = 1; then
+ :
+ else
+ fail=yes
+ fi
+ fi
+ if test -n "$fail"; then
+ echo "Format string checking error:" 1>&2
+ cat f-ph-2-$n.po 1>&2
+ exit 1
+ fi
+done < f-ph-2.data
+
+rm -fr $tmpfiles
+
+exit 0
--- /dev/null
+#! /bin/sh
+
+# Test of gettext facilities in the PHP language.
+# Assumes an fr_FR locale is installed.
+# Assumes the following packages are installed: mod_php4-core.
+
+tmpfiles=""
+trap 'rm -fr $tmpfiles' 1 2 3 15
+
+tmpfiles="$tmpfiles prog.php"
+cat <<\EOF > prog.php
+<?
+ setlocale (LC_ALL, "");
+ textdomain ("prog");
+ bindtextdomain ("prog", ".");
+ echo _("'Your command, please?', asked the waiter.");
+ echo "\n";
+ echo printf(_("%s is replaced by %s."), "FF", "EUR");
+ echo "\n";
+?>
+EOF
+
+tmpfiles="$tmpfiles prog.pot"
+: ${XGETTEXT=xgettext}
+${XGETTEXT} -o prog.pot --omit-header --no-location prog.php
+
+tmpfiles="$tmpfiles prog.ok"
+cat <<EOF > prog.ok
+msgid "'Your command, please?', asked the waiter."
+msgstr ""
+
+#, php-format
+msgid "%s is replaced by %s."
+msgstr ""
+EOF
+
+: ${DIFF=diff}
+${DIFF} prog.ok prog.pot || exit 1
+
+tmpfiles="$tmpfiles fr.po"
+cat <<\EOF > fr.po
+msgid ""
+msgstr "Content-Type: text/plain; charset=ISO-8859-1\n"
+
+msgid "'Your command, please?', asked the waiter."
+msgstr "«Votre commande, s'il vous plait», dit le garçon."
+
+# Reverse the arguments.
+#, php-format
+msgid "%s is replaced by %s."
+msgstr "%2$s remplace %1$s."
+EOF
+
+tmpfiles="$tmpfiles fr.po.new"
+: ${MSGMERGE=msgmerge}
+${MSGMERGE} -q -o fr.po.new fr.po prog.pot
+
+: ${DIFF=diff}
+${DIFF} fr.po fr.po.new || exit 1
+
+tmpfiles="$tmpfiles fr"
+test -d fr || mkdir fr
+test -d fr/LC_MESSAGES || mkdir fr/LC_MESSAGES
+
+: ${MSGFMT=msgfmt}
+${MSGFMT} -o fr/LC_MESSAGES/prog.mo fr.po
+
+# Test for presence of gawk version 4.0 or newer.
+case `(php -v) 2>/dev/null` in
+ [4-9].*) ;;
+ *) echo "SKIP: lang-php"; rm -fr $tmpfiles; exit 77;;
+esac
+
+tmpfiles="$tmpfiles prog.ok prog.out"
+: ${DIFF=diff}
+cat <<\EOF > prog.ok
+«Votre commande, s'il vous plait», dit le garçon.
+EUR remplace FF.
+EOF
+
+LANGUAGE= LC_ALL=fr_FR php -q prog.php > prog.out || exit 1
+${DIFF} prog.ok prog.out || exit 1
+
+rm -fr $tmpfiles
+
+exit 0
projects / thirdparty / gettext.git / commitdiff
