1 /* quotearg.c - quote arguments for output
3 Copyright (C) 1998-2002, 2004-2012 Free Software Foundation, Inc.
5 This program is free software: you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation; either version 3 of the License, or
8 (at your option) any later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>. */
18 /* Written by Paul Eggert <eggert@twinsun.com> */
20 /* Without this pragma, gcc 4.7.0 20111124 mistakenly suggests that
21 the quoting_options_from_style function might be candidate for
23 #if (__GNUC__ == 4 && 6 <= __GNUC_MINOR__) || 4 < __GNUC__
24 # pragma GCC diagnostic ignored "-Wsuggest-attribute=pure"
32 #include "c-strcaseeq.h"
33 #include "localcharset.h"
45 #define _(msgid) gettext (msgid)
46 #define N_(msgid) msgid
49 # define SIZE_MAX ((size_t) -1)
52 #define INT_BITS (sizeof (int) * CHAR_BIT)
54 struct quoting_options
56 /* Basic quoting style. */
57 enum quoting_style style;
59 /* Additional flags. Bitwise combination of enum quoting_flags. */
62 /* Quote the characters indicated by this bit vector even if the
63 quoting style would not normally require them to be quoted. */
64 unsigned int quote_these_too[(UCHAR_MAX / INT_BITS) + 1];
66 /* The left quote for custom_quoting_style. */
67 char const *left_quote;
69 /* The right quote for custom_quoting_style. */
70 char const *right_quote;
73 /* Names of quoting styles. */
74 char const *const quoting_style_args[] =
87 /* Correspondences to quoting style names. */
88 enum quoting_style const quoting_style_vals[] =
90 literal_quoting_style,
92 shell_always_quoting_style,
94 c_maybe_quoting_style,
100 /* The default quoting options. */
101 static struct quoting_options default_quoting_options;
103 /* Allocate a new set of quoting options, with contents initially identical
104 to O if O is not null, or to the default if O is null.
105 It is the caller's responsibility to free the result. */
106 struct quoting_options *
107 clone_quoting_options (struct quoting_options *o)
110 struct quoting_options *p = xmemdup (o ? o : &default_quoting_options,
116 /* Get the value of O's quoting style. If O is null, use the default. */
118 get_quoting_style (struct quoting_options *o)
120 return (o ? o : &default_quoting_options)->style;
123 /* In O (or in the default if O is null),
124 set the value of the quoting style to S. */
126 set_quoting_style (struct quoting_options *o, enum quoting_style s)
128 (o ? o : &default_quoting_options)->style = s;
131 /* In O (or in the default if O is null),
132 set the value of the quoting options for character C to I.
133 Return the old value. Currently, the only values defined for I are
134 0 (the default) and 1 (which means to quote the character even if
135 it would not otherwise be quoted). */
137 set_char_quoting (struct quoting_options *o, char c, int i)
139 unsigned char uc = c;
141 (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS;
142 int shift = uc % INT_BITS;
143 int r = (*p >> shift) & 1;
144 *p ^= ((i & 1) ^ r) << shift;
148 /* In O (or in the default if O is null),
149 set the value of the quoting options flag to I, which can be a
150 bitwise combination of enum quoting_flags, or 0 for default
151 behavior. Return the old value. */
153 set_quoting_flags (struct quoting_options *o, int i)
157 o = &default_quoting_options;
164 set_custom_quoting (struct quoting_options *o,
165 char const *left_quote, char const *right_quote)
168 o = &default_quoting_options;
169 o->style = custom_quoting_style;
170 if (!left_quote || !right_quote)
172 o->left_quote = left_quote;
173 o->right_quote = right_quote;
176 /* Return quoting options for STYLE, with no extra quoting. */
177 static struct quoting_options /* NOT PURE!! */
178 quoting_options_from_style (enum quoting_style style)
180 struct quoting_options o = { 0 };
181 if (style == custom_quoting_style)
187 /* MSGID approximates a quotation mark. Return its translation if it
188 has one; otherwise, return either it or "\"", depending on S.
190 S is either clocale_quoting_style or locale_quoting_style. */
192 gettext_quote (char const *msgid, enum quoting_style s)
194 char const *translation = _(msgid);
195 char const *locale_code;
197 if (translation != msgid)
200 /* For UTF-8 and GB-18030, use single quotes U+2018 and U+2019.
201 Here is a list of other locales that include U+2018 and U+2019:
203 ISO-8859-7 0xA1 KOI8-T 0x91
204 CP869 0x8B CP874 0x91
205 CP932 0x81 0x65 CP936 0xA1 0xAE
206 CP949 0xA1 0xAE CP950 0xA1 0xA5
207 CP1250 0x91 CP1251 0x91
208 CP1252 0x91 CP1253 0x91
209 CP1254 0x91 CP1255 0x91
210 CP1256 0x91 CP1257 0x91
211 EUC-JP 0xA1 0xC6 EUC-KR 0xA1 0xAE
212 EUC-TW 0xA1 0xE4 BIG5 0xA1 0xA5
213 BIG5-HKSCS 0xA1 0xA5 EUC-CN 0xA1 0xAE
214 GBK 0xA1 0xAE Georgian-PS 0x91
217 None of these is still in wide use; using iconv is overkill. */
218 locale_code = locale_charset ();
219 if (STRCASEEQ (locale_code, "UTF-8", 'U','T','F','-','8',0,0,0,0))
220 return msgid[0] == '`' ? "\xe2\x80\x98": "\xe2\x80\x99";
221 if (STRCASEEQ (locale_code, "GB18030", 'G','B','1','8','0','3','0',0,0))
222 return msgid[0] == '`' ? "\xa1\ae": "\xa1\xaf";
224 if (s == clocale_quoting_style)
230 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
231 argument ARG (of size ARGSIZE), using QUOTING_STYLE, FLAGS, and
232 QUOTE_THESE_TOO to control quoting.
233 Terminate the output with a null character, and return the written
234 size of the output, not counting the terminating null.
235 If BUFFERSIZE is too small to store the output string, return the
236 value that would have been returned had BUFFERSIZE been large enough.
237 If ARGSIZE is SIZE_MAX, use the string length of the argument for ARGSIZE.
239 This function acts like quotearg_buffer (BUFFER, BUFFERSIZE, ARG,
240 ARGSIZE, O), except it breaks O into its component pieces and is
241 not careful about errno. */
244 quotearg_buffer_restyled (char *buffer, size_t buffersize,
245 char const *arg, size_t argsize,
246 enum quoting_style quoting_style, int flags,
247 unsigned int const *quote_these_too,
248 char const *left_quote,
249 char const *right_quote)
253 char const *quote_string = 0;
254 size_t quote_string_len = 0;
255 bool backslash_escapes = false;
256 bool unibyte_locale = MB_CUR_MAX == 1;
257 bool elide_outer_quotes = (flags & QA_ELIDE_OUTER_QUOTES) != 0;
262 if (len < buffersize) \
268 switch (quoting_style)
270 case c_maybe_quoting_style:
271 quoting_style = c_quoting_style;
272 elide_outer_quotes = true;
274 case c_quoting_style:
275 if (!elide_outer_quotes)
277 backslash_escapes = true;
279 quote_string_len = 1;
282 case escape_quoting_style:
283 backslash_escapes = true;
284 elide_outer_quotes = false;
287 case locale_quoting_style:
288 case clocale_quoting_style:
289 case custom_quoting_style:
291 if (quoting_style != custom_quoting_style)
294 Get translations for open and closing quotation marks.
295 The message catalog should translate "`" to a left
296 quotation mark suitable for the locale, and similarly for
297 "'". For example, a French Unicode local should translate
298 these to U+00AB (LEFT-POINTING DOUBLE ANGLE
299 QUOTATION MARK), and U+00BB (RIGHT-POINTING DOUBLE ANGLE
300 QUOTATION MARK), respectively.
302 If the catalog has no translation, we will try to
303 use Unicode U+2018 (LEFT SINGLE QUOTATION MARK) and
304 Unicode U+2019 (RIGHT SINGLE QUOTATION MARK). If the
305 current locale is not Unicode, locale_quoting_style
306 will quote `like this', and clocale_quoting_style will
307 quote "like this". You should always include translations
308 for "`" and "'" even if U+2018 and U+2019 are appropriate
311 If you don't know what to put here, please see
312 <http://en.wikipedia.org/wiki/Quotation_marks_in_other_languages>
313 and use glyphs suitable for your language. */
314 left_quote = gettext_quote (N_("`"), quoting_style);
315 right_quote = gettext_quote (N_("'"), quoting_style);
317 if (!elide_outer_quotes)
318 for (quote_string = left_quote; *quote_string; quote_string++)
319 STORE (*quote_string);
320 backslash_escapes = true;
321 quote_string = right_quote;
322 quote_string_len = strlen (quote_string);
326 case shell_quoting_style:
327 quoting_style = shell_always_quoting_style;
328 elide_outer_quotes = true;
330 case shell_always_quoting_style:
331 if (!elide_outer_quotes)
334 quote_string_len = 1;
337 case literal_quoting_style:
338 elide_outer_quotes = false;
345 for (i = 0; ! (argsize == SIZE_MAX ? arg[i] == '\0' : i == argsize); i++)
349 bool is_right_quote = false;
351 if (backslash_escapes
353 && i + quote_string_len <= argsize
354 && memcmp (arg + i, quote_string, quote_string_len) == 0)
356 if (elide_outer_quotes)
357 goto force_outer_quoting_style;
358 is_right_quote = true;
365 if (backslash_escapes)
367 if (elide_outer_quotes)
368 goto force_outer_quoting_style;
370 /* If quote_string were to begin with digits, we'd need to
371 test for the end of the arg as well. However, it's
372 hard to imagine any locale that would use digits in
373 quotes, and set_custom_quoting is documented not to
375 if (i + 1 < argsize && '0' <= arg[i + 1] && arg[i + 1] <= '9')
381 /* We don't have to worry that this last '0' will be
382 backslash-escaped because, again, quote_string should
383 not start with it and because quote_these_too is
384 documented as not accepting it. */
386 else if (flags & QA_ELIDE_NULL_BYTES)
391 switch (quoting_style)
393 case shell_always_quoting_style:
394 if (elide_outer_quotes)
395 goto force_outer_quoting_style;
398 case c_quoting_style:
399 if ((flags & QA_SPLIT_TRIGRAPHS)
400 && i + 2 < argsize && arg[i + 1] == '?')
404 case '(': case ')': case '-': case '/':
405 case '<': case '=': case '>':
406 /* Escape the second '?' in what would otherwise be
408 if (elide_outer_quotes)
409 goto force_outer_quoting_style;
428 case '\a': esc = 'a'; goto c_escape;
429 case '\b': esc = 'b'; goto c_escape;
430 case '\f': esc = 'f'; goto c_escape;
431 case '\n': esc = 'n'; goto c_and_shell_escape;
432 case '\r': esc = 'r'; goto c_and_shell_escape;
433 case '\t': esc = 't'; goto c_and_shell_escape;
434 case '\v': esc = 'v'; goto c_escape;
436 /* No need to escape the escape if we are trying to elide
437 outer quotes and nothing else is problematic. */
438 if (backslash_escapes && elide_outer_quotes && quote_string_len)
442 if (quoting_style == shell_always_quoting_style
443 && elide_outer_quotes)
444 goto force_outer_quoting_style;
447 if (backslash_escapes)
454 case '{': case '}': /* sometimes special if isolated */
455 if (! (argsize == SIZE_MAX ? arg[1] == '\0' : argsize == 1))
463 case '!': /* special in bash */
464 case '"': case '$': case '&':
465 case '(': case ')': case '*': case ';':
467 case '=': /* sometimes special in 0th or (with "set -k") later args */
469 case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */
471 /* A shell special character. In theory, '$' and '`' could
472 be the first bytes of multibyte characters, which means
473 we should check them with mbrtowc, but in practice this
474 doesn't happen so it's not worth worrying about. */
475 if (quoting_style == shell_always_quoting_style
476 && elide_outer_quotes)
477 goto force_outer_quoting_style;
481 if (quoting_style == shell_always_quoting_style)
483 if (elide_outer_quotes)
484 goto force_outer_quoting_style;
491 case '%': case '+': case ',': case '-': case '.': case '/':
492 case '0': case '1': case '2': case '3': case '4': case '5':
493 case '6': case '7': case '8': case '9': case ':':
494 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
495 case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
496 case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
497 case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
498 case 'Y': case 'Z': case ']': case '_': case 'a': case 'b':
499 case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
500 case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
501 case 'o': case 'p': case 'q': case 'r': case 's': case 't':
502 case 'u': case 'v': case 'w': case 'x': case 'y': case 'z':
503 /* These characters don't cause problems, no matter what the
504 quoting style is. They cannot start multibyte sequences.
505 A digit or a special letter would cause trouble if it
506 appeared at the beginning of quote_string because we'd then
507 escape by prepending a backslash. However, it's hard to
508 imagine any locale that would use digits or letters as
509 quotes, and set_custom_quoting is documented not to accept
510 them. Also, a digit or a special letter would cause
511 trouble if it appeared in quote_these_too, but that's also
512 documented as not accepting them. */
516 /* If we have a multibyte sequence, copy it until we reach
517 its end, find an error, or come back to the initial shift
518 state. For C-like styles, if the sequence has
519 unprintable characters, escape the whole sequence, since
520 we can't easily escape single characters within it. */
522 /* Length of multibyte sequence found so far. */
530 printable = isprint (c) != 0;
535 memset (&mbstate, 0, sizeof mbstate);
539 if (argsize == SIZE_MAX)
540 argsize = strlen (arg);
545 size_t bytes = mbrtowc (&w, &arg[i + m],
546 argsize - (i + m), &mbstate);
549 else if (bytes == (size_t) -1)
554 else if (bytes == (size_t) -2)
557 while (i + m < argsize && arg[i + m])
563 /* Work around a bug with older shells that "see" a '\'
564 that is really the 2nd byte of a multibyte character.
565 In practice the problem is limited to ASCII
566 chars >= '@' that are shell special chars. */
567 if ('[' == 0x5b && elide_outer_quotes
568 && quoting_style == shell_always_quoting_style)
571 for (j = 1; j < bytes; j++)
572 switch (arg[i + m + j])
574 case '[': case '\\': case '^':
576 goto force_outer_quoting_style;
588 while (! mbsinit (&mbstate));
591 if (1 < m || (backslash_escapes && ! printable))
593 /* Output a multibyte sequence, or an escaped
594 unprintable unibyte character. */
599 if (backslash_escapes && ! printable)
601 if (elide_outer_quotes)
602 goto force_outer_quoting_style;
604 STORE ('0' + (c >> 6));
605 STORE ('0' + ((c >> 3) & 7));
608 else if (is_right_quote)
611 is_right_quote = false;
624 if (! ((backslash_escapes || elide_outer_quotes)
626 && quote_these_too[c / INT_BITS] & (1 << (c % INT_BITS)))
631 if (elide_outer_quotes)
632 goto force_outer_quoting_style;
639 if (len == 0 && quoting_style == shell_always_quoting_style
640 && elide_outer_quotes)
641 goto force_outer_quoting_style;
643 if (quote_string && !elide_outer_quotes)
644 for (; *quote_string; quote_string++)
645 STORE (*quote_string);
647 if (len < buffersize)
651 force_outer_quoting_style:
652 /* Don't reuse quote_these_too, since the addition of outer quotes
653 sufficiently quotes the specified characters. */
654 return quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
656 flags & ~QA_ELIDE_OUTER_QUOTES, NULL,
657 left_quote, right_quote);
660 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
661 argument ARG (of size ARGSIZE), using O to control quoting.
662 If O is null, use the default.
663 Terminate the output with a null character, and return the written
664 size of the output, not counting the terminating null.
665 If BUFFERSIZE is too small to store the output string, return the
666 value that would have been returned had BUFFERSIZE been large enough.
667 If ARGSIZE is SIZE_MAX, use the string length of the argument for
670 quotearg_buffer (char *buffer, size_t buffersize,
671 char const *arg, size_t argsize,
672 struct quoting_options const *o)
674 struct quoting_options const *p = o ? o : &default_quoting_options;
676 size_t r = quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
677 p->style, p->flags, p->quote_these_too,
678 p->left_quote, p->right_quote);
683 /* Equivalent to quotearg_alloc (ARG, ARGSIZE, NULL, O). */
685 quotearg_alloc (char const *arg, size_t argsize,
686 struct quoting_options const *o)
688 return quotearg_alloc_mem (arg, argsize, NULL, o);
691 /* Like quotearg_buffer (..., ARG, ARGSIZE, O), except return newly
692 allocated storage containing the quoted string, and store the
693 resulting size into *SIZE, if non-NULL. The result can contain
694 embedded null bytes only if ARGSIZE is not SIZE_MAX, SIZE is not
695 NULL, and set_quoting_flags has not set the null byte elision
698 quotearg_alloc_mem (char const *arg, size_t argsize, size_t *size,
699 struct quoting_options const *o)
701 struct quoting_options const *p = o ? o : &default_quoting_options;
703 /* Elide embedded null bytes if we can't return a size. */
704 int flags = p->flags | (size ? 0 : QA_ELIDE_NULL_BYTES);
705 size_t bufsize = quotearg_buffer_restyled (0, 0, arg, argsize, p->style,
706 flags, p->quote_these_too,
709 char *buf = xcharalloc (bufsize);
710 quotearg_buffer_restyled (buf, bufsize, arg, argsize, p->style, flags,
712 p->left_quote, p->right_quote);
719 /* A storage slot with size and pointer to a value. */
726 /* Preallocate a slot 0 buffer, so that the caller can always quote
727 one small component of a "memory exhausted" message in slot 0. */
728 static char slot0[256];
729 static unsigned int nslots = 1;
730 static struct slotvec slotvec0 = {sizeof slot0, slot0};
731 static struct slotvec *slotvec = &slotvec0;
736 struct slotvec *sv = slotvec;
738 for (i = 1; i < nslots; i++)
740 if (sv[0].val != slot0)
743 slotvec0.size = sizeof slot0;
744 slotvec0.val = slot0;
754 /* Use storage slot N to return a quoted version of argument ARG.
755 ARG is of size ARGSIZE, but if that is SIZE_MAX, ARG is a
756 null-terminated string.
757 OPTIONS specifies the quoting options.
758 The returned value points to static storage that can be
759 reused by the next call to this function with the same value of N.
760 N must be nonnegative. N is deliberately declared with type "int"
761 to allow for future extensions (using negative values). */
763 quotearg_n_options (int n, char const *arg, size_t argsize,
764 struct quoting_options const *options)
769 struct slotvec *sv = slotvec;
776 /* FIXME: technically, the type of n1 should be 'unsigned int',
777 but that evokes an unsuppressible warning from gcc-4.0.1 and
778 older. If gcc ever provides an option to suppress that warning,
779 revert to the original type, so that the test in xalloc_oversized
780 is once again performed only at compile time. */
782 bool preallocated = (sv == &slotvec0);
784 if (xalloc_oversized (n1, sizeof *sv))
787 slotvec = sv = xrealloc (preallocated ? NULL : sv, n1 * sizeof *sv);
790 memset (sv + nslots, 0, (n1 - nslots) * sizeof *sv);
795 size_t size = sv[n].size;
796 char *val = sv[n].val;
797 /* Elide embedded null bytes since we don't return a size. */
798 int flags = options->flags | QA_ELIDE_NULL_BYTES;
799 size_t qsize = quotearg_buffer_restyled (val, size, arg, argsize,
800 options->style, flags,
801 options->quote_these_too,
803 options->right_quote);
807 sv[n].size = size = qsize + 1;
810 sv[n].val = val = xcharalloc (size);
811 quotearg_buffer_restyled (val, size, arg, argsize, options->style,
812 flags, options->quote_these_too,
814 options->right_quote);
823 quotearg_n (int n, char const *arg)
825 return quotearg_n_options (n, arg, SIZE_MAX, &default_quoting_options);
829 quotearg_n_mem (int n, char const *arg, size_t argsize)
831 return quotearg_n_options (n, arg, argsize, &default_quoting_options);
835 quotearg (char const *arg)
837 return quotearg_n (0, arg);
841 quotearg_mem (char const *arg, size_t argsize)
843 return quotearg_n_mem (0, arg, argsize);
847 quotearg_n_style (int n, enum quoting_style s, char const *arg)
849 struct quoting_options const o = quoting_options_from_style (s);
850 return quotearg_n_options (n, arg, SIZE_MAX, &o);
854 quotearg_n_style_mem (int n, enum quoting_style s,
855 char const *arg, size_t argsize)
857 struct quoting_options const o = quoting_options_from_style (s);
858 return quotearg_n_options (n, arg, argsize, &o);
862 quotearg_style (enum quoting_style s, char const *arg)
864 return quotearg_n_style (0, s, arg);
868 quotearg_style_mem (enum quoting_style s, char const *arg, size_t argsize)
870 return quotearg_n_style_mem (0, s, arg, argsize);
874 quotearg_char_mem (char const *arg, size_t argsize, char ch)
876 struct quoting_options options;
877 options = default_quoting_options;
878 set_char_quoting (&options, ch, 1);
879 return quotearg_n_options (0, arg, argsize, &options);
883 quotearg_char (char const *arg, char ch)
885 return quotearg_char_mem (arg, SIZE_MAX, ch);
889 quotearg_colon (char const *arg)
891 return quotearg_char (arg, ':');
895 quotearg_colon_mem (char const *arg, size_t argsize)
897 return quotearg_char_mem (arg, argsize, ':');
901 quotearg_n_custom (int n, char const *left_quote,
902 char const *right_quote, char const *arg)
904 return quotearg_n_custom_mem (n, left_quote, right_quote, arg,
909 quotearg_n_custom_mem (int n, char const *left_quote,
910 char const *right_quote,
911 char const *arg, size_t argsize)
913 struct quoting_options o = default_quoting_options;
914 set_custom_quoting (&o, left_quote, right_quote);
915 return quotearg_n_options (n, arg, argsize, &o);
919 quotearg_custom (char const *left_quote, char const *right_quote,
922 return quotearg_n_custom (0, left_quote, right_quote, arg);
926 quotearg_custom_mem (char const *left_quote, char const *right_quote,
927 char const *arg, size_t argsize)
929 return quotearg_n_custom_mem (0, left_quote, right_quote, arg,