1 /* quotearg.c - quote arguments for output
3 Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2005, 2006, 2007,
4 2008 Free Software Foundation, Inc.
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/>. */
19 /* Written by Paul Eggert <eggert@twinsun.com> */
37 #define _(msgid) gettext (msgid)
38 #define N_(msgid) msgid
41 # define SIZE_MAX ((size_t) -1)
44 #define INT_BITS (sizeof (int) * CHAR_BIT)
46 struct quoting_options
48 /* Basic quoting style. */
49 enum quoting_style style;
51 /* Additional flags. Bitwise combination of enum quoting_flags. */
54 /* Quote the characters indicated by this bit vector even if the
55 quoting style would not normally require them to be quoted. */
56 unsigned int quote_these_too[(UCHAR_MAX / INT_BITS) + 1];
59 /* Names of quoting styles. */
60 char const *const quoting_style_args[] =
73 /* Correspondences to quoting style names. */
74 enum quoting_style const quoting_style_vals[] =
76 literal_quoting_style,
78 shell_always_quoting_style,
80 c_maybe_quoting_style,
86 /* The default quoting options. */
87 static struct quoting_options default_quoting_options;
89 /* Allocate a new set of quoting options, with contents initially identical
90 to O if O is not null, or to the default if O is null.
91 It is the caller's responsibility to free the result. */
92 struct quoting_options *
93 clone_quoting_options (struct quoting_options *o)
96 struct quoting_options *p = xmemdup (o ? o : &default_quoting_options,
102 /* Get the value of O's quoting style. If O is null, use the default. */
104 get_quoting_style (struct quoting_options *o)
106 return (o ? o : &default_quoting_options)->style;
109 /* In O (or in the default if O is null),
110 set the value of the quoting style to S. */
112 set_quoting_style (struct quoting_options *o, enum quoting_style s)
114 (o ? o : &default_quoting_options)->style = s;
117 /* In O (or in the default if O is null),
118 set the value of the quoting options for character C to I.
119 Return the old value. Currently, the only values defined for I are
120 0 (the default) and 1 (which means to quote the character even if
121 it would not otherwise be quoted). */
123 set_char_quoting (struct quoting_options *o, char c, int i)
125 unsigned char uc = c;
127 (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS;
128 int shift = uc % INT_BITS;
129 int r = (*p >> shift) & 1;
130 *p ^= ((i & 1) ^ r) << shift;
134 /* In O (or in the default if O is null),
135 set the value of the quoting options flag to I, which can be a
136 bitwise combination of enum quoting_flags, or 0 for default
137 behavior. Return the old value. */
139 set_quoting_flags (struct quoting_options *o, int i)
143 o = &default_quoting_options;
149 /* Return quoting options for STYLE, with no extra quoting. */
150 static struct quoting_options
151 quoting_options_from_style (enum quoting_style style)
153 struct quoting_options o;
156 memset (o.quote_these_too, 0, sizeof o.quote_these_too);
160 /* MSGID approximates a quotation mark. Return its translation if it
161 has one; otherwise, return either it or "\"", depending on S. */
163 gettext_quote (char const *msgid, enum quoting_style s)
165 char const *translation = _(msgid);
166 if (translation == msgid && s == clocale_quoting_style)
171 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
172 argument ARG (of size ARGSIZE), using QUOTING_STYLE, FLAGS, and
173 QUOTE_THESE_TOO to control quoting.
174 Terminate the output with a null character, and return the written
175 size of the output, not counting the terminating null.
176 If BUFFERSIZE is too small to store the output string, return the
177 value that would have been returned had BUFFERSIZE been large enough.
178 If ARGSIZE is SIZE_MAX, use the string length of the argument for ARGSIZE.
180 This function acts like quotearg_buffer (BUFFER, BUFFERSIZE, ARG,
181 ARGSIZE, O), except it breaks O into its component pieces and is
182 not careful about errno. */
185 quotearg_buffer_restyled (char *buffer, size_t buffersize,
186 char const *arg, size_t argsize,
187 enum quoting_style quoting_style, int flags,
188 unsigned int const *quote_these_too)
192 char const *quote_string = 0;
193 size_t quote_string_len = 0;
194 bool backslash_escapes = false;
195 bool unibyte_locale = MB_CUR_MAX == 1;
196 bool elide_outer_quotes = (flags & QA_ELIDE_OUTER_QUOTES) != 0;
201 if (len < buffersize) \
207 switch (quoting_style)
209 case c_maybe_quoting_style:
210 quoting_style = c_quoting_style;
211 elide_outer_quotes = true;
213 case c_quoting_style:
214 if (!elide_outer_quotes)
216 backslash_escapes = true;
218 quote_string_len = 1;
221 case escape_quoting_style:
222 backslash_escapes = true;
223 elide_outer_quotes = false;
226 case locale_quoting_style:
227 case clocale_quoting_style:
230 Get translations for open and closing quotation marks.
232 The message catalog should translate "`" to a left
233 quotation mark suitable for the locale, and similarly for
234 "'". If the catalog has no translation,
235 locale_quoting_style quotes `like this', and
236 clocale_quoting_style quotes "like this".
238 For example, an American English Unicode locale should
239 translate "`" to U+201C (LEFT DOUBLE QUOTATION MARK), and
240 should translate "'" to U+201D (RIGHT DOUBLE QUOTATION
241 MARK). A British English Unicode locale should instead
242 translate these to U+2018 (LEFT SINGLE QUOTATION MARK) and
243 U+2019 (RIGHT SINGLE QUOTATION MARK), respectively.
245 If you don't know what to put here, please see
246 <http://en.wikipedia.org/wiki/Quotation_mark#Glyphs>
247 and use glyphs suitable for your language. */
249 char const *left = gettext_quote (N_("`"), quoting_style);
250 char const *right = gettext_quote (N_("'"), quoting_style);
251 if (!elide_outer_quotes)
252 for (quote_string = left; *quote_string; quote_string++)
253 STORE (*quote_string);
254 backslash_escapes = true;
255 quote_string = right;
256 quote_string_len = strlen (quote_string);
260 case shell_quoting_style:
261 quoting_style = shell_always_quoting_style;
262 elide_outer_quotes = true;
264 case shell_always_quoting_style:
265 if (!elide_outer_quotes)
268 quote_string_len = 1;
271 case literal_quoting_style:
272 elide_outer_quotes = false;
279 for (i = 0; ! (argsize == SIZE_MAX ? arg[i] == '\0' : i == argsize); i++)
284 if (backslash_escapes
286 && i + quote_string_len <= argsize
287 && memcmp (arg + i, quote_string, quote_string_len) == 0)
289 if (elide_outer_quotes)
290 goto force_outer_quoting_style;
298 if (backslash_escapes)
300 if (elide_outer_quotes)
301 goto force_outer_quoting_style;
303 if (i + 1 < argsize && '0' <= arg[i + 1] && arg[i + 1] <= '9')
310 else if (flags & QA_ELIDE_NULL_BYTES)
315 switch (quoting_style)
317 case shell_always_quoting_style:
318 if (elide_outer_quotes)
319 goto force_outer_quoting_style;
322 case c_quoting_style:
323 if ((flags & QA_SPLIT_TRIGRAPHS)
324 && i + 2 < argsize && arg[i + 1] == '?')
328 case '(': case ')': case '-': case '/':
329 case '<': case '=': case '>':
330 /* Escape the second '?' in what would otherwise be
332 if (elide_outer_quotes)
333 goto force_outer_quoting_style;
352 case '\a': esc = 'a'; goto c_escape;
353 case '\b': esc = 'b'; goto c_escape;
354 case '\f': esc = 'f'; goto c_escape;
355 case '\n': esc = 'n'; goto c_and_shell_escape;
356 case '\r': esc = 'r'; goto c_and_shell_escape;
357 case '\t': esc = 't'; goto c_and_shell_escape;
358 case '\v': esc = 'v'; goto c_escape;
360 /* No need to escape the escape if we are trying to elide
361 outer quotes and nothing else is problematic. */
362 if (backslash_escapes && elide_outer_quotes && quote_string_len)
366 if (quoting_style == shell_always_quoting_style
367 && elide_outer_quotes)
368 goto force_outer_quoting_style;
371 if (backslash_escapes)
378 case '{': case '}': /* sometimes special if isolated */
379 if (! (argsize == SIZE_MAX ? arg[1] == '\0' : argsize == 1))
387 case '!': /* special in bash */
388 case '"': case '$': case '&':
389 case '(': case ')': case '*': case ';':
391 case '=': /* sometimes special in 0th or (with "set -k") later args */
393 case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */
395 /* A shell special character. In theory, '$' and '`' could
396 be the first bytes of multibyte characters, which means
397 we should check them with mbrtowc, but in practice this
398 doesn't happen so it's not worth worrying about. */
399 if (quoting_style == shell_always_quoting_style
400 && elide_outer_quotes)
401 goto force_outer_quoting_style;
405 if (quoting_style == shell_always_quoting_style)
407 if (elide_outer_quotes)
408 goto force_outer_quoting_style;
415 case '%': case '+': case ',': case '-': case '.': case '/':
416 case '0': case '1': case '2': case '3': case '4': case '5':
417 case '6': case '7': case '8': case '9': case ':':
418 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
419 case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
420 case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
421 case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
422 case 'Y': case 'Z': case ']': case '_': case 'a': case 'b':
423 case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
424 case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
425 case 'o': case 'p': case 'q': case 'r': case 's': case 't':
426 case 'u': case 'v': case 'w': case 'x': case 'y': case 'z':
427 /* These characters don't cause problems, no matter what the
428 quoting style is. They cannot start multibyte sequences. */
432 /* If we have a multibyte sequence, copy it until we reach
433 its end, find an error, or come back to the initial shift
434 state. For C-like styles, if the sequence has
435 unprintable characters, escape the whole sequence, since
436 we can't easily escape single characters within it. */
438 /* Length of multibyte sequence found so far. */
446 printable = isprint (c) != 0;
451 memset (&mbstate, 0, sizeof mbstate);
455 if (argsize == SIZE_MAX)
456 argsize = strlen (arg);
461 size_t bytes = mbrtowc (&w, &arg[i + m],
462 argsize - (i + m), &mbstate);
465 else if (bytes == (size_t) -1)
470 else if (bytes == (size_t) -2)
473 while (i + m < argsize && arg[i + m])
479 /* Work around a bug with older shells that "see" a '\'
480 that is really the 2nd byte of a multibyte character.
481 In practice the problem is limited to ASCII
482 chars >= '@' that are shell special chars. */
483 if ('[' == 0x5b && elide_outer_quotes
484 && quoting_style == shell_always_quoting_style)
487 for (j = 1; j < bytes; j++)
488 switch (arg[i + m + j])
490 case '[': case '\\': case '^':
492 goto force_outer_quoting_style;
504 while (! mbsinit (&mbstate));
507 if (1 < m || (backslash_escapes && ! printable))
509 /* Output a multibyte sequence, or an escaped
510 unprintable unibyte character. */
515 if (backslash_escapes && ! printable)
517 if (elide_outer_quotes)
518 goto force_outer_quoting_style;
520 STORE ('0' + (c >> 6));
521 STORE ('0' + ((c >> 3) & 7));
535 if (! ((backslash_escapes || elide_outer_quotes)
537 && quote_these_too[c / INT_BITS] & (1 << (c % INT_BITS))))
541 if (elide_outer_quotes)
542 goto force_outer_quoting_style;
549 if (len == 0 && quoting_style == shell_always_quoting_style
550 && elide_outer_quotes)
551 goto force_outer_quoting_style;
553 if (quote_string && !elide_outer_quotes)
554 for (; *quote_string; quote_string++)
555 STORE (*quote_string);
557 if (len < buffersize)
561 force_outer_quoting_style:
562 /* Don't reuse quote_these_too, since the addition of outer quotes
563 sufficiently quotes the specified characters. */
564 return quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
566 flags & ~QA_ELIDE_OUTER_QUOTES, NULL);
569 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
570 argument ARG (of size ARGSIZE), using O to control quoting.
571 If O is null, use the default.
572 Terminate the output with a null character, and return the written
573 size of the output, not counting the terminating null.
574 If BUFFERSIZE is too small to store the output string, return the
575 value that would have been returned had BUFFERSIZE been large enough.
576 If ARGSIZE is SIZE_MAX, use the string length of the argument for
579 quotearg_buffer (char *buffer, size_t buffersize,
580 char const *arg, size_t argsize,
581 struct quoting_options const *o)
583 struct quoting_options const *p = o ? o : &default_quoting_options;
585 size_t r = quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
586 p->style, p->flags, p->quote_these_too);
591 /* Equivalent to quotearg_alloc (ARG, ARGSIZE, NULL, O). */
593 quotearg_alloc (char const *arg, size_t argsize,
594 struct quoting_options const *o)
596 return quotearg_alloc_mem (arg, argsize, NULL, o);
599 /* Like quotearg_buffer (..., ARG, ARGSIZE, O), except return newly
600 allocated storage containing the quoted string, and store the
601 resulting size into *SIZE, if non-NULL. The result can contain
602 embedded null bytes only if ARGSIZE is not SIZE_MAX, SIZE is not
603 NULL, and set_quoting_flags has not set the null byte elision
606 quotearg_alloc_mem (char const *arg, size_t argsize, size_t *size,
607 struct quoting_options const *o)
609 struct quoting_options const *p = o ? o : &default_quoting_options;
611 /* Elide embedded null bytes if we can't return a size. */
612 int flags = p->flags | (size ? 0 : QA_ELIDE_NULL_BYTES);
613 size_t bufsize = quotearg_buffer_restyled (0, 0, arg, argsize, p->style,
614 flags, p->quote_these_too) + 1;
615 char *buf = xcharalloc (bufsize);
616 quotearg_buffer_restyled (buf, bufsize, arg, argsize, p->style, flags,
624 /* A storage slot with size and pointer to a value. */
631 /* Preallocate a slot 0 buffer, so that the caller can always quote
632 one small component of a "memory exhausted" message in slot 0. */
633 static char slot0[256];
634 static unsigned int nslots = 1;
635 static struct slotvec slotvec0 = {sizeof slot0, slot0};
636 static struct slotvec *slotvec = &slotvec0;
641 struct slotvec *sv = slotvec;
643 for (i = 1; i < nslots; i++)
645 if (sv[0].val != slot0)
648 slotvec0.size = sizeof slot0;
649 slotvec0.val = slot0;
659 /* Use storage slot N to return a quoted version of argument ARG.
660 ARG is of size ARGSIZE, but if that is SIZE_MAX, ARG is a
661 null-terminated string.
662 OPTIONS specifies the quoting options.
663 The returned value points to static storage that can be
664 reused by the next call to this function with the same value of N.
665 N must be nonnegative. N is deliberately declared with type "int"
666 to allow for future extensions (using negative values). */
668 quotearg_n_options (int n, char const *arg, size_t argsize,
669 struct quoting_options const *options)
674 struct slotvec *sv = slotvec;
681 /* FIXME: technically, the type of n1 should be `unsigned int',
682 but that evokes an unsuppressible warning from gcc-4.0.1 and
683 older. If gcc ever provides an option to suppress that warning,
684 revert to the original type, so that the test in xalloc_oversized
685 is once again performed only at compile time. */
687 bool preallocated = (sv == &slotvec0);
689 if (xalloc_oversized (n1, sizeof *sv))
692 slotvec = sv = xrealloc (preallocated ? NULL : sv, n1 * sizeof *sv);
695 memset (sv + nslots, 0, (n1 - nslots) * sizeof *sv);
700 size_t size = sv[n].size;
701 char *val = sv[n].val;
702 /* Elide embedded null bytes since we don't return a size. */
703 int flags = options->flags | QA_ELIDE_NULL_BYTES;
704 size_t qsize = quotearg_buffer_restyled (val, size, arg, argsize,
705 options->style, flags,
706 options->quote_these_too);
710 sv[n].size = size = qsize + 1;
713 sv[n].val = val = xcharalloc (size);
714 quotearg_buffer_restyled (val, size, arg, argsize, options->style,
715 flags, options->quote_these_too);
724 quotearg_n (int n, char const *arg)
726 return quotearg_n_options (n, arg, SIZE_MAX, &default_quoting_options);
730 quotearg_n_mem (int n, char const *arg, size_t argsize)
732 return quotearg_n_options (n, arg, argsize, &default_quoting_options);
736 quotearg (char const *arg)
738 return quotearg_n (0, arg);
742 quotearg_mem (char const *arg, size_t argsize)
744 return quotearg_n_mem (0, arg, argsize);
748 quotearg_n_style (int n, enum quoting_style s, char const *arg)
750 struct quoting_options const o = quoting_options_from_style (s);
751 return quotearg_n_options (n, arg, SIZE_MAX, &o);
755 quotearg_n_style_mem (int n, enum quoting_style s,
756 char const *arg, size_t argsize)
758 struct quoting_options const o = quoting_options_from_style (s);
759 return quotearg_n_options (n, arg, argsize, &o);
763 quotearg_style (enum quoting_style s, char const *arg)
765 return quotearg_n_style (0, s, arg);
769 quotearg_style_mem (enum quoting_style s, char const *arg, size_t argsize)
771 return quotearg_n_style_mem (0, s, arg, argsize);
775 quotearg_char_mem (char const *arg, size_t argsize, char ch)
777 struct quoting_options options;
778 options = default_quoting_options;
779 set_char_quoting (&options, ch, 1);
780 return quotearg_n_options (0, arg, argsize, &options);
784 quotearg_char (char const *arg, char ch)
786 return quotearg_char_mem (arg, SIZE_MAX, ch);
790 quotearg_colon (char const *arg)
792 return quotearg_char (arg, ':');
796 quotearg_colon_mem (char const *arg, size_t argsize)
798 return quotearg_char_mem (arg, argsize, ':');