1 /* Shell format strings.
2 Copyright (C) 2003-2004, 2006-2007, 2009, 2015 Free Software
4 Written by Bruno Haible <bruno@clisp.org>, 2003.
6 This program is free software: you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 3 of the License, or
9 (at your option) any later version.
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with this program. If not, see <http://www.gnu.org/licenses/>. */
30 #include "format-invalid.h"
33 #define _(str) gettext (str)
35 /* Shell format strings are simply strings subjects to variable substitution.
36 A variable substitution starts with '$' and is finished by either
37 - a nonempty sequence of alphanumeric ASCII characters, the first being
39 - an opening brace '{', a nonempty sequence of alphanumeric ASCII
40 characters, the first being not a digit, and a closing brace '}'.
41 We don't support variable references like $1, $$ or $? since they make
42 no sense when 'envsubst' is invoked.
43 We don't support non-ASCII variable names, to avoid dependencies w.r.t. the
44 current encoding: While "${\xe0}" looks like a variable access in ISO-8859-1
45 encoding, it doesn't look like one in the BIG5, BIG5-HKSCS, GBK, GB18030,
46 SHIFT_JIS, JOHAB encodings, because \xe0\x7d is a single character in these
48 We don't support the POSIX syntax for default or alternate values:
49 ${variable-default} ${variable:-default}
50 ${variable=default} ${variable:=default}
51 ${variable+replacement} ${variable:+replacement}
52 ${variable?ignored} ${variable:?ignored}
53 because the translator might be tempted to change the default value; if
54 we allow it we have a security problem; if we don't allow it the translator
65 unsigned int directives;
66 unsigned int named_arg_count;
67 unsigned int allocated;
68 struct named_arg *named;
73 named_arg_compare (const void *p1, const void *p2)
75 return strcmp (((const struct named_arg *) p1)->name,
76 ((const struct named_arg *) p2)->name);
79 #define INVALID_NON_ASCII_VARIABLE() \
80 xstrdup (_("The string refers to a shell variable with a non-ASCII name."))
81 #define INVALID_SHELL_SYNTAX() \
82 xstrdup (_("The string refers to a shell variable with complex shell brace syntax. This syntax is unsupported here due to security reasons."))
83 #define INVALID_CONTEXT_DEPENDENT_VARIABLE() \
84 xstrdup (_("The string refers to a shell variable whose value may be different inside shell functions."))
85 #define INVALID_EMPTY_VARIABLE() \
86 xstrdup (_("The string refers to a shell variable with an empty name."))
89 format_parse (const char *format, bool translated, char *fdi,
90 char **invalid_reason)
92 const char *const format_start = format;
97 spec.named_arg_count = 0;
101 for (; *format != '\0';)
102 if (*format++ == '$')
104 /* A variable substitution. */
107 FDI_SET (format - 1, FMTDIR_START);
112 const char *name_start;
113 const char *name_end;
116 name_start = ++format;
117 for (; *format != '\0'; format++)
121 if (!c_isascii (*format))
123 *invalid_reason = INVALID_NON_ASCII_VARIABLE ();
124 FDI_SET (format, FMTDIR_ERROR);
127 if (format > name_start
128 && (*format == '-' || *format == '=' || *format == '+'
129 || *format == '?' || *format == ':'))
131 *invalid_reason = INVALID_SHELL_SYNTAX ();
132 FDI_SET (format, FMTDIR_ERROR);
135 if (!(c_isalnum (*format) || *format == '_')
136 || (format == name_start && c_isdigit (*format)))
138 *invalid_reason = INVALID_CONTEXT_DEPENDENT_VARIABLE ();
139 FDI_SET (format, FMTDIR_ERROR);
145 *invalid_reason = INVALID_UNTERMINATED_DIRECTIVE ();
146 FDI_SET (format - 1, FMTDIR_ERROR);
151 n = name_end - name_start;
154 *invalid_reason = INVALID_EMPTY_VARIABLE ();
155 FDI_SET (format - 1, FMTDIR_ERROR);
158 name = XNMALLOC (n + 1, char);
159 memcpy (name, name_start, n);
162 else if (c_isalpha (*format) || *format == '_')
164 const char *name_start;
165 const char *name_end;
171 while (*format != '\0' && (c_isalnum (*format) || *format == '_'));
174 n = name_end - name_start;
175 name = XNMALLOC (n + 1, char);
176 memcpy (name, name_start, n);
179 else if (*format != '\0')
181 if (!c_isascii (*format))
183 *invalid_reason = INVALID_NON_ASCII_VARIABLE ();
184 FDI_SET (format, FMTDIR_ERROR);
189 *invalid_reason = INVALID_CONTEXT_DEPENDENT_VARIABLE ();
190 FDI_SET (format, FMTDIR_ERROR);
196 *invalid_reason = INVALID_UNTERMINATED_DIRECTIVE ();
197 FDI_SET (format - 1, FMTDIR_ERROR);
201 /* Named argument. */
202 if (spec.allocated == spec.named_arg_count)
204 spec.allocated = 2 * spec.allocated + 1;
205 spec.named = (struct named_arg *) xrealloc (spec.named, spec.allocated * sizeof (struct named_arg));
207 spec.named[spec.named_arg_count].name = name;
208 spec.named_arg_count++;
210 FDI_SET (format - 1, FMTDIR_END);
213 /* Sort the named argument array, and eliminate duplicates. */
214 if (spec.named_arg_count > 1)
218 qsort (spec.named, spec.named_arg_count, sizeof (struct named_arg),
221 /* Remove duplicates: Copy from i to j, keeping 0 <= j <= i. */
222 for (i = j = 0; i < spec.named_arg_count; i++)
223 if (j > 0 && strcmp (spec.named[i].name, spec.named[j-1].name) == 0)
224 free (spec.named[i].name);
228 spec.named[j].name = spec.named[i].name;
231 spec.named_arg_count = j;
234 result = XMALLOC (struct spec);
239 if (spec.named != NULL)
242 for (i = 0; i < spec.named_arg_count; i++)
243 free (spec.named[i].name);
250 format_free (void *descr)
252 struct spec *spec = (struct spec *) descr;
254 if (spec->named != NULL)
257 for (i = 0; i < spec->named_arg_count; i++)
258 free (spec->named[i].name);
265 format_get_number_of_directives (void *descr)
267 struct spec *spec = (struct spec *) descr;
269 return spec->directives;
273 format_check (void *msgid_descr, void *msgstr_descr, bool equality,
274 formatstring_error_logger_t error_logger,
275 const char *pretty_msgid, const char *pretty_msgstr)
277 struct spec *spec1 = (struct spec *) msgid_descr;
278 struct spec *spec2 = (struct spec *) msgstr_descr;
281 if (spec1->named_arg_count + spec2->named_arg_count > 0)
284 unsigned int n1 = spec1->named_arg_count;
285 unsigned int n2 = spec2->named_arg_count;
287 /* Check the argument names are the same.
288 Both arrays are sorted. We search for the first difference. */
289 for (i = 0, j = 0; i < n1 || j < n2; )
291 int cmp = (i >= n1 ? 1 :
293 strcmp (spec1->named[i].name, spec2->named[j].name));
298 error_logger (_("a format specification for argument '%s', as in '%s', doesn't exist in '%s'"),
299 spec2->named[j].name, pretty_msgstr,
309 error_logger (_("a format specification for argument '%s' doesn't exist in '%s'"),
310 spec1->named[i].name, pretty_msgstr);
326 struct formatstring_parser formatstring_sh =
330 format_get_number_of_directives,
338 /* Test program: Print the argument list specification returned by
339 format_parse for strings read from standard input. */
344 format_print (void *descr)
346 struct spec *spec = (struct spec *) descr;
356 for (i = 0; i < spec->named_arg_count; i++)
360 printf ("'%s'", spec->named[i].name);
371 size_t line_size = 0;
373 char *invalid_reason;
376 line_len = getline (&line, &line_size, stdin);
379 if (line_len > 0 && line[line_len - 1] == '\n')
380 line[--line_len] = '\0';
382 invalid_reason = NULL;
383 descr = format_parse (line, false, NULL, &invalid_reason);
385 format_print (descr);
388 printf ("%s\n", invalid_reason);
390 free (invalid_reason);
398 * For Emacs M-x compile
400 * compile-command: "/bin/sh ../libtool --tag=CC --mode=link gcc -o a.out -static -O -g -Wall -I.. -I../gnulib-lib -I../intl -DHAVE_CONFIG_H -DTEST format-sh.c ../gnulib-lib/libgettextlib.la"