1 /* -*- buffer-read-only: t -*- vi: set ro: */
2 /* DO NOT EDIT! GENERATED AUTOMATICALLY! */
3 /* quotearg.c - quote arguments for output
5 Copyright (C) 1998-2002, 2004-2014 Free Software Foundation, Inc.
7 This program is free software: you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 3 of the License, or
10 (at your option) any later version.
12 This program is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with this program. If not, see <http://www.gnu.org/licenses/>. */
20 /* Written by Paul Eggert <eggert@twinsun.com> */
22 /* Without this pragma, gcc 4.7.0 20111124 mistakenly suggests that
23 the quoting_options_from_style function might be candidate for
25 #if (__GNUC__ == 4 && 6 <= __GNUC_MINOR__) || 4 < __GNUC__
26 # pragma GCC diagnostic ignored "-Wsuggest-attribute=pure"
35 #include "c-strcaseeq.h"
36 #include "localcharset.h"
48 #define _(msgid) gettext (msgid)
49 #define N_(msgid) msgid
52 # define SIZE_MAX ((size_t) -1)
55 #define INT_BITS (sizeof (int) * CHAR_BIT)
57 struct quoting_options
59 /* Basic quoting style. */
60 enum quoting_style style;
62 /* Additional flags. Bitwise combination of enum quoting_flags. */
65 /* Quote the characters indicated by this bit vector even if the
66 quoting style would not normally require them to be quoted. */
67 unsigned int quote_these_too[(UCHAR_MAX / INT_BITS) + 1];
69 /* The left quote for custom_quoting_style. */
70 char const *left_quote;
72 /* The right quote for custom_quoting_style. */
73 char const *right_quote;
76 /* Names of quoting styles. */
77 char const *const quoting_style_args[] =
90 /* Correspondences to quoting style names. */
91 enum quoting_style const quoting_style_vals[] =
93 literal_quoting_style,
95 shell_always_quoting_style,
97 c_maybe_quoting_style,
100 clocale_quoting_style
103 /* The default quoting options. */
104 static struct quoting_options default_quoting_options;
106 /* Allocate a new set of quoting options, with contents initially identical
107 to O if O is not null, or to the default if O is null.
108 It is the caller's responsibility to free the result. */
109 struct quoting_options *
110 clone_quoting_options (struct quoting_options *o)
113 struct quoting_options *p = xmemdup (o ? o : &default_quoting_options,
119 /* Get the value of O's quoting style. If O is null, use the default. */
121 get_quoting_style (struct quoting_options *o)
123 return (o ? o : &default_quoting_options)->style;
126 /* In O (or in the default if O is null),
127 set the value of the quoting style to S. */
129 set_quoting_style (struct quoting_options *o, enum quoting_style s)
131 (o ? o : &default_quoting_options)->style = s;
134 /* In O (or in the default if O is null),
135 set the value of the quoting options for character C to I.
136 Return the old value. Currently, the only values defined for I are
137 0 (the default) and 1 (which means to quote the character even if
138 it would not otherwise be quoted). */
140 set_char_quoting (struct quoting_options *o, char c, int i)
142 unsigned char uc = c;
144 (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS;
145 int shift = uc % INT_BITS;
146 int r = (*p >> shift) & 1;
147 *p ^= ((i & 1) ^ r) << shift;
151 /* In O (or in the default if O is null),
152 set the value of the quoting options flag to I, which can be a
153 bitwise combination of enum quoting_flags, or 0 for default
154 behavior. Return the old value. */
156 set_quoting_flags (struct quoting_options *o, int i)
160 o = &default_quoting_options;
167 set_custom_quoting (struct quoting_options *o,
168 char const *left_quote, char const *right_quote)
171 o = &default_quoting_options;
172 o->style = custom_quoting_style;
173 if (!left_quote || !right_quote)
175 o->left_quote = left_quote;
176 o->right_quote = right_quote;
179 /* Return quoting options for STYLE, with no extra quoting. */
180 static struct quoting_options /* NOT PURE!! */
181 quoting_options_from_style (enum quoting_style style)
183 struct quoting_options o = { 0, 0, { 0 }, NULL, NULL };
184 if (style == custom_quoting_style)
190 /* MSGID approximates a quotation mark. Return its translation if it
191 has one; otherwise, return either it or "\"", depending on S.
193 S is either clocale_quoting_style or locale_quoting_style. */
195 gettext_quote (char const *msgid, enum quoting_style s)
197 char const *translation = _(msgid);
198 char const *locale_code;
200 if (translation != msgid)
203 /* For UTF-8 and GB-18030, use single quotes U+2018 and U+2019.
204 Here is a list of other locales that include U+2018 and U+2019:
206 ISO-8859-7 0xA1 KOI8-T 0x91
207 CP869 0x8B CP874 0x91
208 CP932 0x81 0x65 CP936 0xA1 0xAE
209 CP949 0xA1 0xAE CP950 0xA1 0xA5
210 CP1250 0x91 CP1251 0x91
211 CP1252 0x91 CP1253 0x91
212 CP1254 0x91 CP1255 0x91
213 CP1256 0x91 CP1257 0x91
214 EUC-JP 0xA1 0xC6 EUC-KR 0xA1 0xAE
215 EUC-TW 0xA1 0xE4 BIG5 0xA1 0xA5
216 BIG5-HKSCS 0xA1 0xA5 EUC-CN 0xA1 0xAE
217 GBK 0xA1 0xAE Georgian-PS 0x91
220 None of these is still in wide use; using iconv is overkill. */
221 locale_code = locale_charset ();
222 if (STRCASEEQ (locale_code, "UTF-8", 'U','T','F','-','8',0,0,0,0))
223 return msgid[0] == '`' ? "\xe2\x80\x98": "\xe2\x80\x99";
224 if (STRCASEEQ (locale_code, "GB18030", 'G','B','1','8','0','3','0',0,0))
225 return msgid[0] == '`' ? "\xa1\ae": "\xa1\xaf";
227 return (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
354 <= (argsize == SIZE_MAX && 1 < quote_string_len
355 /* Use strlen only if we must: when argsize is SIZE_MAX,
356 and when the quote string is more than 1 byte long.
357 If we do call strlen, save the result. */
358 ? (argsize = strlen (arg)) : argsize))
359 && memcmp (arg + i, quote_string, quote_string_len) == 0)
361 if (elide_outer_quotes)
362 goto force_outer_quoting_style;
363 is_right_quote = true;
370 if (backslash_escapes)
372 if (elide_outer_quotes)
373 goto force_outer_quoting_style;
375 /* If quote_string were to begin with digits, we'd need to
376 test for the end of the arg as well. However, it's
377 hard to imagine any locale that would use digits in
378 quotes, and set_custom_quoting is documented not to
380 if (i + 1 < argsize && '0' <= arg[i + 1] && arg[i + 1] <= '9')
386 /* We don't have to worry that this last '0' will be
387 backslash-escaped because, again, quote_string should
388 not start with it and because quote_these_too is
389 documented as not accepting it. */
391 else if (flags & QA_ELIDE_NULL_BYTES)
396 switch (quoting_style)
398 case shell_always_quoting_style:
399 if (elide_outer_quotes)
400 goto force_outer_quoting_style;
403 case c_quoting_style:
404 if ((flags & QA_SPLIT_TRIGRAPHS)
405 && i + 2 < argsize && arg[i + 1] == '?')
409 case '(': case ')': case '-': case '/':
410 case '<': case '=': case '>':
411 /* Escape the second '?' in what would otherwise be
413 if (elide_outer_quotes)
414 goto force_outer_quoting_style;
433 case '\a': esc = 'a'; goto c_escape;
434 case '\b': esc = 'b'; goto c_escape;
435 case '\f': esc = 'f'; goto c_escape;
436 case '\n': esc = 'n'; goto c_and_shell_escape;
437 case '\r': esc = 'r'; goto c_and_shell_escape;
438 case '\t': esc = 't'; goto c_and_shell_escape;
439 case '\v': esc = 'v'; goto c_escape;
441 /* No need to escape the escape if we are trying to elide
442 outer quotes and nothing else is problematic. */
443 if (backslash_escapes && elide_outer_quotes && quote_string_len)
447 if (quoting_style == shell_always_quoting_style
448 && elide_outer_quotes)
449 goto force_outer_quoting_style;
452 if (backslash_escapes)
459 case '{': case '}': /* sometimes special if isolated */
460 if (! (argsize == SIZE_MAX ? arg[1] == '\0' : argsize == 1))
468 case '!': /* special in bash */
469 case '"': case '$': case '&':
470 case '(': case ')': case '*': case ';':
472 case '=': /* sometimes special in 0th or (with "set -k") later args */
474 case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */
476 /* A shell special character. In theory, '$' and '`' could
477 be the first bytes of multibyte characters, which means
478 we should check them with mbrtowc, but in practice this
479 doesn't happen so it's not worth worrying about. */
480 if (quoting_style == shell_always_quoting_style
481 && elide_outer_quotes)
482 goto force_outer_quoting_style;
486 if (quoting_style == shell_always_quoting_style)
488 if (elide_outer_quotes)
489 goto force_outer_quoting_style;
496 case '%': case '+': case ',': case '-': case '.': case '/':
497 case '0': case '1': case '2': case '3': case '4': case '5':
498 case '6': case '7': case '8': case '9': case ':':
499 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
500 case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
501 case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
502 case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
503 case 'Y': case 'Z': case ']': case '_': case 'a': case 'b':
504 case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
505 case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
506 case 'o': case 'p': case 'q': case 'r': case 's': case 't':
507 case 'u': case 'v': case 'w': case 'x': case 'y': case 'z':
508 /* These characters don't cause problems, no matter what the
509 quoting style is. They cannot start multibyte sequences.
510 A digit or a special letter would cause trouble if it
511 appeared at the beginning of quote_string because we'd then
512 escape by prepending a backslash. However, it's hard to
513 imagine any locale that would use digits or letters as
514 quotes, and set_custom_quoting is documented not to accept
515 them. Also, a digit or a special letter would cause
516 trouble if it appeared in quote_these_too, but that's also
517 documented as not accepting them. */
521 /* If we have a multibyte sequence, copy it until we reach
522 its end, find an error, or come back to the initial shift
523 state. For C-like styles, if the sequence has
524 unprintable characters, escape the whole sequence, since
525 we can't easily escape single characters within it. */
527 /* Length of multibyte sequence found so far. */
535 printable = isprint (c) != 0;
540 memset (&mbstate, 0, sizeof mbstate);
544 if (argsize == SIZE_MAX)
545 argsize = strlen (arg);
550 size_t bytes = mbrtowc (&w, &arg[i + m],
551 argsize - (i + m), &mbstate);
554 else if (bytes == (size_t) -1)
559 else if (bytes == (size_t) -2)
562 while (i + m < argsize && arg[i + m])
568 /* Work around a bug with older shells that "see" a '\'
569 that is really the 2nd byte of a multibyte character.
570 In practice the problem is limited to ASCII
571 chars >= '@' that are shell special chars. */
572 if ('[' == 0x5b && elide_outer_quotes
573 && quoting_style == shell_always_quoting_style)
576 for (j = 1; j < bytes; j++)
577 switch (arg[i + m + j])
579 case '[': case '\\': case '^':
581 goto force_outer_quoting_style;
593 while (! mbsinit (&mbstate));
596 if (1 < m || (backslash_escapes && ! printable))
598 /* Output a multibyte sequence, or an escaped
599 unprintable unibyte character. */
604 if (backslash_escapes && ! printable)
606 if (elide_outer_quotes)
607 goto force_outer_quoting_style;
609 STORE ('0' + (c >> 6));
610 STORE ('0' + ((c >> 3) & 7));
613 else if (is_right_quote)
616 is_right_quote = false;
629 if (! ((backslash_escapes || elide_outer_quotes)
631 && quote_these_too[c / INT_BITS] >> (c % INT_BITS) & 1)
636 if (elide_outer_quotes)
637 goto force_outer_quoting_style;
644 if (len == 0 && quoting_style == shell_always_quoting_style
645 && elide_outer_quotes)
646 goto force_outer_quoting_style;
648 if (quote_string && !elide_outer_quotes)
649 for (; *quote_string; quote_string++)
650 STORE (*quote_string);
652 if (len < buffersize)
656 force_outer_quoting_style:
657 /* Don't reuse quote_these_too, since the addition of outer quotes
658 sufficiently quotes the specified characters. */
659 return quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
661 flags & ~QA_ELIDE_OUTER_QUOTES, NULL,
662 left_quote, right_quote);
665 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
666 argument ARG (of size ARGSIZE), using O to control quoting.
667 If O is null, use the default.
668 Terminate the output with a null character, and return the written
669 size of the output, not counting the terminating null.
670 If BUFFERSIZE is too small to store the output string, return the
671 value that would have been returned had BUFFERSIZE been large enough.
672 If ARGSIZE is SIZE_MAX, use the string length of the argument for
675 quotearg_buffer (char *buffer, size_t buffersize,
676 char const *arg, size_t argsize,
677 struct quoting_options const *o)
679 struct quoting_options const *p = o ? o : &default_quoting_options;
681 size_t r = quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
682 p->style, p->flags, p->quote_these_too,
683 p->left_quote, p->right_quote);
688 /* Equivalent to quotearg_alloc (ARG, ARGSIZE, NULL, O). */
690 quotearg_alloc (char const *arg, size_t argsize,
691 struct quoting_options const *o)
693 return quotearg_alloc_mem (arg, argsize, NULL, o);
696 /* Like quotearg_buffer (..., ARG, ARGSIZE, O), except return newly
697 allocated storage containing the quoted string, and store the
698 resulting size into *SIZE, if non-NULL. The result can contain
699 embedded null bytes only if ARGSIZE is not SIZE_MAX, SIZE is not
700 NULL, and set_quoting_flags has not set the null byte elision
703 quotearg_alloc_mem (char const *arg, size_t argsize, size_t *size,
704 struct quoting_options const *o)
706 struct quoting_options const *p = o ? o : &default_quoting_options;
708 /* Elide embedded null bytes if we can't return a size. */
709 int flags = p->flags | (size ? 0 : QA_ELIDE_NULL_BYTES);
710 size_t bufsize = quotearg_buffer_restyled (0, 0, arg, argsize, p->style,
711 flags, p->quote_these_too,
714 char *buf = xcharalloc (bufsize);
715 quotearg_buffer_restyled (buf, bufsize, arg, argsize, p->style, flags,
717 p->left_quote, p->right_quote);
724 /* A storage slot with size and pointer to a value. */
731 /* Preallocate a slot 0 buffer, so that the caller can always quote
732 one small component of a "memory exhausted" message in slot 0. */
733 static char slot0[256];
734 static unsigned int nslots = 1;
735 static struct slotvec slotvec0 = {sizeof slot0, slot0};
736 static struct slotvec *slotvec = &slotvec0;
741 struct slotvec *sv = slotvec;
743 for (i = 1; i < nslots; i++)
745 if (sv[0].val != slot0)
748 slotvec0.size = sizeof slot0;
749 slotvec0.val = slot0;
759 /* Use storage slot N to return a quoted version of argument ARG.
760 ARG is of size ARGSIZE, but if that is SIZE_MAX, ARG is a
761 null-terminated string.
762 OPTIONS specifies the quoting options.
763 The returned value points to static storage that can be
764 reused by the next call to this function with the same value of N.
765 N must be nonnegative. N is deliberately declared with type "int"
766 to allow for future extensions (using negative values). */
768 quotearg_n_options (int n, char const *arg, size_t argsize,
769 struct quoting_options const *options)
774 struct slotvec *sv = slotvec;
781 /* FIXME: technically, the type of n1 should be 'unsigned int',
782 but that evokes an unsuppressible warning from gcc-4.0.1 and
783 older. If gcc ever provides an option to suppress that warning,
784 revert to the original type, so that the test in xalloc_oversized
785 is once again performed only at compile time. */
787 bool preallocated = (sv == &slotvec0);
789 if (xalloc_oversized (n1, sizeof *sv))
792 slotvec = sv = xrealloc (preallocated ? NULL : sv, n1 * sizeof *sv);
795 memset (sv + nslots, 0, (n1 - nslots) * sizeof *sv);
800 size_t size = sv[n].size;
801 char *val = sv[n].val;
802 /* Elide embedded null bytes since we don't return a size. */
803 int flags = options->flags | QA_ELIDE_NULL_BYTES;
804 size_t qsize = quotearg_buffer_restyled (val, size, arg, argsize,
805 options->style, flags,
806 options->quote_these_too,
808 options->right_quote);
812 sv[n].size = size = qsize + 1;
815 sv[n].val = val = xcharalloc (size);
816 quotearg_buffer_restyled (val, size, arg, argsize, options->style,
817 flags, options->quote_these_too,
819 options->right_quote);
828 quotearg_n (int n, char const *arg)
830 return quotearg_n_options (n, arg, SIZE_MAX, &default_quoting_options);
834 quotearg_n_mem (int n, char const *arg, size_t argsize)
836 return quotearg_n_options (n, arg, argsize, &default_quoting_options);
840 quotearg (char const *arg)
842 return quotearg_n (0, arg);
846 quotearg_mem (char const *arg, size_t argsize)
848 return quotearg_n_mem (0, arg, argsize);
852 quotearg_n_style (int n, enum quoting_style s, char const *arg)
854 struct quoting_options const o = quoting_options_from_style (s);
855 return quotearg_n_options (n, arg, SIZE_MAX, &o);
859 quotearg_n_style_mem (int n, enum quoting_style s,
860 char const *arg, size_t argsize)
862 struct quoting_options const o = quoting_options_from_style (s);
863 return quotearg_n_options (n, arg, argsize, &o);
867 quotearg_style (enum quoting_style s, char const *arg)
869 return quotearg_n_style (0, s, arg);
873 quotearg_style_mem (enum quoting_style s, char const *arg, size_t argsize)
875 return quotearg_n_style_mem (0, s, arg, argsize);
879 quotearg_char_mem (char const *arg, size_t argsize, char ch)
881 struct quoting_options options;
882 options = default_quoting_options;
883 set_char_quoting (&options, ch, 1);
884 return quotearg_n_options (0, arg, argsize, &options);
888 quotearg_char (char const *arg, char ch)
890 return quotearg_char_mem (arg, SIZE_MAX, ch);
894 quotearg_colon (char const *arg)
896 return quotearg_char (arg, ':');
900 quotearg_colon_mem (char const *arg, size_t argsize)
902 return quotearg_char_mem (arg, argsize, ':');
906 quotearg_n_custom (int n, char const *left_quote,
907 char const *right_quote, char const *arg)
909 return quotearg_n_custom_mem (n, left_quote, right_quote, arg,
914 quotearg_n_custom_mem (int n, char const *left_quote,
915 char const *right_quote,
916 char const *arg, size_t argsize)
918 struct quoting_options o = default_quoting_options;
919 set_custom_quoting (&o, left_quote, right_quote);
920 return quotearg_n_options (n, arg, argsize, &o);
924 quotearg_custom (char const *left_quote, char const *right_quote,
927 return quotearg_n_custom (0, left_quote, right_quote, arg);
931 quotearg_custom_mem (char const *left_quote, char const *right_quote,
932 char const *arg, size_t argsize)
934 return quotearg_n_custom_mem (0, left_quote, right_quote, arg,
939 /* The quoting option used by the functions of quote.h. */
940 struct quoting_options quote_quoting_options =
942 locale_quoting_style,
949 quote_n_mem (int n, char const *arg, size_t argsize)
951 return quotearg_n_options (n, arg, argsize, "e_quoting_options);
955 quote_mem (char const *arg, size_t argsize)
957 return quote_n_mem (0, arg, argsize);
961 quote_n (int n, char const *arg)
963 return quote_n_mem (n, arg, SIZE_MAX);
967 quote (char const *arg)
969 return quote_n (0, arg);