The fifth batch
[git/git.git] / strbuf.h
1 #ifndef STRBUF_H
2 #define STRBUF_H
3
4 struct string_list;
5
6 /**
7 * strbuf's are meant to be used with all the usual C string and memory
8 * APIs. Given that the length of the buffer is known, it's often better to
9 * use the mem* functions than a str* one (memchr vs. strchr e.g.).
10 * Though, one has to be careful about the fact that str* functions often
11 * stop on NULs and that strbufs may have embedded NULs.
12 *
13 * A strbuf is NUL terminated for convenience, but no function in the
14 * strbuf API actually relies on the string being free of NULs.
15 *
16 * strbufs have some invariants that are very important to keep in mind:
17 *
18 * - The `buf` member is never NULL, so it can be used in any usual C
19 * string operations safely. strbuf's _have_ to be initialized either by
20 * `strbuf_init()` or by `= STRBUF_INIT` before the invariants, though.
21 *
22 * Do *not* assume anything on what `buf` really is (e.g. if it is
23 * allocated memory or not), use `strbuf_detach()` to unwrap a memory
24 * buffer from its strbuf shell in a safe way. That is the sole supported
25 * way. This will give you a malloced buffer that you can later `free()`.
26 *
27 * However, it is totally safe to modify anything in the string pointed by
28 * the `buf` member, between the indices `0` and `len-1` (inclusive).
29 *
30 * - The `buf` member is a byte array that has at least `len + 1` bytes
31 * allocated. The extra byte is used to store a `'\0'`, allowing the
32 * `buf` member to be a valid C-string. Every strbuf function ensure this
33 * invariant is preserved.
34 *
35 * NOTE: It is OK to "play" with the buffer directly if you work it this
36 * way:
37 *
38 * strbuf_grow(sb, SOME_SIZE); <1>
39 * strbuf_setlen(sb, sb->len + SOME_OTHER_SIZE);
40 *
41 * <1> Here, the memory array starting at `sb->buf`, and of length
42 * `strbuf_avail(sb)` is all yours, and you can be sure that
43 * `strbuf_avail(sb)` is at least `SOME_SIZE`.
44 *
45 * NOTE: `SOME_OTHER_SIZE` must be smaller or equal to `strbuf_avail(sb)`.
46 *
47 * Doing so is safe, though if it has to be done in many places, adding the
48 * missing API to the strbuf module is the way to go.
49 *
50 * WARNING: Do _not_ assume that the area that is yours is of size `alloc
51 * - 1` even if it's true in the current implementation. Alloc is somehow a
52 * "private" member that should not be messed with. Use `strbuf_avail()`
53 * instead.
54 */
55
56 /**
57 * Data Structures
58 * ---------------
59 */
60
61 /**
62 * This is the string buffer structure. The `len` member can be used to
63 * determine the current length of the string, and `buf` member provides
64 * access to the string itself.
65 */
66 struct strbuf {
67 size_t alloc;
68 size_t len;
69 char *buf;
70 };
71
72 extern char strbuf_slopbuf[];
73 #define STRBUF_INIT { .alloc = 0, .len = 0, .buf = strbuf_slopbuf }
74
75 /*
76 * Predeclare this here, since cache.h includes this file before it defines the
77 * struct.
78 */
79 struct object_id;
80
81 /**
82 * Life Cycle Functions
83 * --------------------
84 */
85
86 /**
87 * Initialize the structure. The second parameter can be zero or a bigger
88 * number to allocate memory, in case you want to prevent further reallocs.
89 */
90 void strbuf_init(struct strbuf *sb, size_t alloc);
91
92 /**
93 * Release a string buffer and the memory it used. After this call, the
94 * strbuf points to an empty string that does not need to be free()ed, as
95 * if it had been set to `STRBUF_INIT` and never modified.
96 *
97 * To clear a strbuf in preparation for further use without the overhead
98 * of free()ing and malloc()ing again, use strbuf_reset() instead.
99 */
100 void strbuf_release(struct strbuf *sb);
101
102 /**
103 * Detach the string from the strbuf and returns it; you now own the
104 * storage the string occupies and it is your responsibility from then on
105 * to release it with `free(3)` when you are done with it.
106 *
107 * The strbuf that previously held the string is reset to `STRBUF_INIT` so
108 * it can be reused after calling this function.
109 */
110 char *strbuf_detach(struct strbuf *sb, size_t *sz);
111
112 /**
113 * Attach a string to a buffer. You should specify the string to attach,
114 * the current length of the string and the amount of allocated memory.
115 * The amount must be larger than the string length, because the string you
116 * pass is supposed to be a NUL-terminated string. This string _must_ be
117 * malloc()ed, and after attaching, the pointer cannot be relied upon
118 * anymore, and neither be free()d directly.
119 */
120 void strbuf_attach(struct strbuf *sb, void *str, size_t len, size_t mem);
121
122 /**
123 * Swap the contents of two string buffers.
124 */
125 static inline void strbuf_swap(struct strbuf *a, struct strbuf *b)
126 {
127 SWAP(*a, *b);
128 }
129
130
131 /**
132 * Functions related to the size of the buffer
133 * -------------------------------------------
134 */
135
136 /**
137 * Determine the amount of allocated but unused memory.
138 */
139 static inline size_t strbuf_avail(const struct strbuf *sb)
140 {
141 return sb->alloc ? sb->alloc - sb->len - 1 : 0;
142 }
143
144 /**
145 * Ensure that at least this amount of unused memory is available after
146 * `len`. This is used when you know a typical size for what you will add
147 * and want to avoid repetitive automatic resizing of the underlying buffer.
148 * This is never a needed operation, but can be critical for performance in
149 * some cases.
150 */
151 void strbuf_grow(struct strbuf *sb, size_t amount);
152
153 /**
154 * Set the length of the buffer to a given value. This function does *not*
155 * allocate new memory, so you should not perform a `strbuf_setlen()` to a
156 * length that is larger than `len + strbuf_avail()`. `strbuf_setlen()` is
157 * just meant as a 'please fix invariants from this strbuf I just messed
158 * with'.
159 */
160 static inline void strbuf_setlen(struct strbuf *sb, size_t len)
161 {
162 if (len > (sb->alloc ? sb->alloc - 1 : 0))
163 die("BUG: strbuf_setlen() beyond buffer");
164 sb->len = len;
165 if (sb->buf != strbuf_slopbuf)
166 sb->buf[len] = '\0';
167 else
168 assert(!strbuf_slopbuf[0]);
169 }
170
171 /**
172 * Empty the buffer by setting the size of it to zero.
173 */
174 #define strbuf_reset(sb) strbuf_setlen(sb, 0)
175
176
177 /**
178 * Functions related to the contents of the buffer
179 * -----------------------------------------------
180 */
181
182 /**
183 * Strip whitespace from the beginning (`ltrim`), end (`rtrim`), or both side
184 * (`trim`) of a string.
185 */
186 void strbuf_trim(struct strbuf *sb);
187 void strbuf_rtrim(struct strbuf *sb);
188 void strbuf_ltrim(struct strbuf *sb);
189
190 /* Strip trailing directory separators */
191 void strbuf_trim_trailing_dir_sep(struct strbuf *sb);
192
193 /* Strip trailing LF or CR/LF */
194 void strbuf_trim_trailing_newline(struct strbuf *sb);
195
196 /**
197 * Replace the contents of the strbuf with a reencoded form. Returns -1
198 * on error, 0 on success.
199 */
200 int strbuf_reencode(struct strbuf *sb, const char *from, const char *to);
201
202 /**
203 * Lowercase each character in the buffer using `tolower`.
204 */
205 void strbuf_tolower(struct strbuf *sb);
206
207 /**
208 * Compare two buffers. Returns an integer less than, equal to, or greater
209 * than zero if the first buffer is found, respectively, to be less than,
210 * to match, or be greater than the second buffer.
211 */
212 int strbuf_cmp(const struct strbuf *first, const struct strbuf *second);
213
214
215 /**
216 * Adding data to the buffer
217 * -------------------------
218 *
219 * NOTE: All of the functions in this section will grow the buffer as
220 * necessary. If they fail for some reason other than memory shortage and the
221 * buffer hadn't been allocated before (i.e. the `struct strbuf` was set to
222 * `STRBUF_INIT`), then they will free() it.
223 */
224
225 /**
226 * Add a single character to the buffer.
227 */
228 static inline void strbuf_addch(struct strbuf *sb, int c)
229 {
230 if (!strbuf_avail(sb))
231 strbuf_grow(sb, 1);
232 sb->buf[sb->len++] = c;
233 sb->buf[sb->len] = '\0';
234 }
235
236 /**
237 * Add a character the specified number of times to the buffer.
238 */
239 void strbuf_addchars(struct strbuf *sb, int c, size_t n);
240
241 /**
242 * Insert data to the given position of the buffer. The remaining contents
243 * will be shifted, not overwritten.
244 */
245 void strbuf_insert(struct strbuf *sb, size_t pos, const void *, size_t);
246
247 /**
248 * Remove given amount of data from a given position of the buffer.
249 */
250 void strbuf_remove(struct strbuf *sb, size_t pos, size_t len);
251
252 /**
253 * Remove the bytes between `pos..pos+len` and replace it with the given
254 * data.
255 */
256 void strbuf_splice(struct strbuf *sb, size_t pos, size_t len,
257 const void *data, size_t data_len);
258
259 /**
260 * Add a NUL-terminated string to the buffer. Each line will be prepended
261 * by a comment character and a blank.
262 */
263 void strbuf_add_commented_lines(struct strbuf *out,
264 const char *buf, size_t size);
265
266
267 /**
268 * Add data of given length to the buffer.
269 */
270 void strbuf_add(struct strbuf *sb, const void *data, size_t len);
271
272 /**
273 * Add a NUL-terminated string to the buffer.
274 *
275 * NOTE: This function will *always* be implemented as an inline or a macro
276 * using strlen, meaning that this is efficient to write things like:
277 *
278 * strbuf_addstr(sb, "immediate string");
279 *
280 */
281 static inline void strbuf_addstr(struct strbuf *sb, const char *s)
282 {
283 strbuf_add(sb, s, strlen(s));
284 }
285
286 /**
287 * Copy the contents of another buffer at the end of the current one.
288 */
289 void strbuf_addbuf(struct strbuf *sb, const struct strbuf *sb2);
290
291 /**
292 * This function can be used to expand a format string containing
293 * placeholders. To that end, it parses the string and calls the specified
294 * function for every percent sign found.
295 *
296 * The callback function is given a pointer to the character after the `%`
297 * and a pointer to the struct strbuf. It is expected to add the expanded
298 * version of the placeholder to the strbuf, e.g. to add a newline
299 * character if the letter `n` appears after a `%`. The function returns
300 * the length of the placeholder recognized and `strbuf_expand()` skips
301 * over it.
302 *
303 * The format `%%` is automatically expanded to a single `%` as a quoting
304 * mechanism; callers do not need to handle the `%` placeholder themselves,
305 * and the callback function will not be invoked for this placeholder.
306 *
307 * All other characters (non-percent and not skipped ones) are copied
308 * verbatim to the strbuf. If the callback returned zero, meaning that the
309 * placeholder is unknown, then the percent sign is copied, too.
310 *
311 * In order to facilitate caching and to make it possible to give
312 * parameters to the callback, `strbuf_expand()` passes a context pointer,
313 * which can be used by the programmer of the callback as she sees fit.
314 */
315 typedef size_t (*expand_fn_t) (struct strbuf *sb,
316 const char *placeholder,
317 void *context);
318 void strbuf_expand(struct strbuf *sb,
319 const char *format,
320 expand_fn_t fn,
321 void *context);
322
323 /**
324 * Used as callback for `strbuf_expand` to only expand literals
325 * (i.e. %n and %xNN). The context argument is ignored.
326 */
327 size_t strbuf_expand_literal_cb(struct strbuf *sb,
328 const char *placeholder,
329 void *context);
330
331 /**
332 * Used as callback for `strbuf_expand()`, expects an array of
333 * struct strbuf_expand_dict_entry as context, i.e. pairs of
334 * placeholder and replacement string. The array needs to be
335 * terminated by an entry with placeholder set to NULL.
336 */
337 struct strbuf_expand_dict_entry {
338 const char *placeholder;
339 const char *value;
340 };
341 size_t strbuf_expand_dict_cb(struct strbuf *sb,
342 const char *placeholder,
343 void *context);
344
345 /**
346 * Append the contents of one strbuf to another, quoting any
347 * percent signs ("%") into double-percents ("%%") in the
348 * destination. This is useful for literal data to be fed to either
349 * strbuf_expand or to the *printf family of functions.
350 */
351 void strbuf_addbuf_percentquote(struct strbuf *dst, const struct strbuf *src);
352
353 /**
354 * Append the given byte size as a human-readable string (i.e. 12.23 KiB,
355 * 3.50 MiB).
356 */
357 void strbuf_humanise_bytes(struct strbuf *buf, off_t bytes);
358
359 /**
360 * Add a formatted string to the buffer.
361 */
362 __attribute__((format (printf,2,3)))
363 void strbuf_addf(struct strbuf *sb, const char *fmt, ...);
364
365 /**
366 * Add a formatted string prepended by a comment character and a
367 * blank to the buffer.
368 */
369 __attribute__((format (printf, 2, 3)))
370 void strbuf_commented_addf(struct strbuf *sb, const char *fmt, ...);
371
372 __attribute__((format (printf,2,0)))
373 void strbuf_vaddf(struct strbuf *sb, const char *fmt, va_list ap);
374
375 /**
376 * Add the time specified by `tm`, as formatted by `strftime`.
377 * `tz_offset` is in decimal hhmm format, e.g. -600 means six hours west
378 * of Greenwich, and it's used to expand %z internally. However, tokens
379 * with modifiers (e.g. %Ez) are passed to `strftime`.
380 * `suppress_tz_name`, when set, expands %Z internally to the empty
381 * string rather than passing it to `strftime`.
382 */
383 void strbuf_addftime(struct strbuf *sb, const char *fmt,
384 const struct tm *tm, int tz_offset,
385 int suppress_tz_name);
386
387 /**
388 * Read a given size of data from a FILE* pointer to the buffer.
389 *
390 * NOTE: The buffer is rewound if the read fails. If -1 is returned,
391 * `errno` must be consulted, like you would do for `read(3)`.
392 * `strbuf_read()`, `strbuf_read_file()` and `strbuf_getline_*()`
393 * family of functions have the same behaviour as well.
394 */
395 size_t strbuf_fread(struct strbuf *sb, size_t size, FILE *file);
396
397 /**
398 * Read the contents of a given file descriptor. The third argument can be
399 * used to give a hint about the file size, to avoid reallocs. If read fails,
400 * any partial read is undone.
401 */
402 ssize_t strbuf_read(struct strbuf *sb, int fd, size_t hint);
403
404 /**
405 * Read the contents of a given file descriptor partially by using only one
406 * attempt of xread. The third argument can be used to give a hint about the
407 * file size, to avoid reallocs. Returns the number of new bytes appended to
408 * the sb.
409 */
410 ssize_t strbuf_read_once(struct strbuf *sb, int fd, size_t hint);
411
412 /**
413 * Read the contents of a file, specified by its path. The third argument
414 * can be used to give a hint about the file size, to avoid reallocs.
415 * Return the number of bytes read or a negative value if some error
416 * occurred while opening or reading the file.
417 */
418 ssize_t strbuf_read_file(struct strbuf *sb, const char *path, size_t hint);
419
420 /**
421 * Read the target of a symbolic link, specified by its path. The third
422 * argument can be used to give a hint about the size, to avoid reallocs.
423 */
424 int strbuf_readlink(struct strbuf *sb, const char *path, size_t hint);
425
426 /**
427 * Write the whole content of the strbuf to the stream not stopping at
428 * NUL bytes.
429 */
430 ssize_t strbuf_write(struct strbuf *sb, FILE *stream);
431
432 /**
433 * Read a line from a FILE *, overwriting the existing contents of
434 * the strbuf. The strbuf_getline*() family of functions share
435 * this signature, but have different line termination conventions.
436 *
437 * Reading stops after the terminator or at EOF. The terminator
438 * is removed from the buffer before returning. Returns 0 unless
439 * there was nothing left before EOF, in which case it returns `EOF`.
440 */
441 typedef int (*strbuf_getline_fn)(struct strbuf *, FILE *);
442
443 /* Uses LF as the line terminator */
444 int strbuf_getline_lf(struct strbuf *sb, FILE *fp);
445
446 /* Uses NUL as the line terminator */
447 int strbuf_getline_nul(struct strbuf *sb, FILE *fp);
448
449 /*
450 * Similar to strbuf_getline_lf(), but additionally treats a CR that
451 * comes immediately before the LF as part of the terminator.
452 * This is the most friendly version to be used to read "text" files
453 * that can come from platforms whose native text format is CRLF
454 * terminated.
455 */
456 int strbuf_getline(struct strbuf *sb, FILE *file);
457
458
459 /**
460 * Like `strbuf_getline`, but keeps the trailing terminator (if
461 * any) in the buffer.
462 */
463 int strbuf_getwholeline(struct strbuf *sb, FILE *file, int term);
464
465 /**
466 * Like `strbuf_getwholeline`, but operates on a file descriptor.
467 * It reads one character at a time, so it is very slow. Do not
468 * use it unless you need the correct position in the file
469 * descriptor.
470 */
471 int strbuf_getwholeline_fd(struct strbuf *sb, int fd, int term);
472
473 /**
474 * Set the buffer to the path of the current working directory.
475 */
476 int strbuf_getcwd(struct strbuf *sb);
477
478 /**
479 * Add a path to a buffer, converting a relative path to an
480 * absolute one in the process. Symbolic links are not
481 * resolved.
482 */
483 void strbuf_add_absolute_path(struct strbuf *sb, const char *path);
484
485 /**
486 * Canonize `path` (make it absolute, resolve symlinks, remove extra
487 * slashes) and append it to `sb`. Die with an informative error
488 * message if there is a problem.
489 *
490 * The directory part of `path` (i.e., everything up to the last
491 * dir_sep) must denote a valid, existing directory, but the last
492 * component need not exist.
493 *
494 * Callers that don't mind links should use the more lightweight
495 * strbuf_add_absolute_path() instead.
496 */
497 void strbuf_add_real_path(struct strbuf *sb, const char *path);
498
499
500 /**
501 * Normalize in-place the path contained in the strbuf. See
502 * normalize_path_copy() for details. If an error occurs, the contents of "sb"
503 * are left untouched, and -1 is returned.
504 */
505 int strbuf_normalize_path(struct strbuf *sb);
506
507 /**
508 * Strip whitespace from a buffer. The second parameter controls if
509 * comments are considered contents to be removed or not.
510 */
511 void strbuf_stripspace(struct strbuf *buf, int skip_comments);
512
513 static inline int strbuf_strip_suffix(struct strbuf *sb, const char *suffix)
514 {
515 if (strip_suffix_mem(sb->buf, &sb->len, suffix)) {
516 strbuf_setlen(sb, sb->len);
517 return 1;
518 } else
519 return 0;
520 }
521
522 /**
523 * Split str (of length slen) at the specified terminator character.
524 * Return a null-terminated array of pointers to strbuf objects
525 * holding the substrings. The substrings include the terminator,
526 * except for the last substring, which might be unterminated if the
527 * original string did not end with a terminator. If max is positive,
528 * then split the string into at most max substrings (with the last
529 * substring containing everything following the (max-1)th terminator
530 * character).
531 *
532 * The most generic form is `strbuf_split_buf`, which takes an arbitrary
533 * pointer/len buffer. The `_str` variant takes a NUL-terminated string,
534 * the `_max` variant takes a strbuf, and just `strbuf_split` is a convenience
535 * wrapper to drop the `max` parameter.
536 *
537 * For lighter-weight alternatives, see string_list_split() and
538 * string_list_split_in_place().
539 */
540 struct strbuf **strbuf_split_buf(const char *str, size_t len,
541 int terminator, int max);
542
543 static inline struct strbuf **strbuf_split_str(const char *str,
544 int terminator, int max)
545 {
546 return strbuf_split_buf(str, strlen(str), terminator, max);
547 }
548
549 static inline struct strbuf **strbuf_split_max(const struct strbuf *sb,
550 int terminator, int max)
551 {
552 return strbuf_split_buf(sb->buf, sb->len, terminator, max);
553 }
554
555 static inline struct strbuf **strbuf_split(const struct strbuf *sb,
556 int terminator)
557 {
558 return strbuf_split_max(sb, terminator, 0);
559 }
560
561 /*
562 * Adds all strings of a string list to the strbuf, separated by the given
563 * separator. For example, if sep is
564 * ', '
565 * and slist contains
566 * ['element1', 'element2', ..., 'elementN'],
567 * then write:
568 * 'element1, element2, ..., elementN'
569 * to str. If only one element, just write "element1" to str.
570 */
571 void strbuf_add_separated_string_list(struct strbuf *str,
572 const char *sep,
573 struct string_list *slist);
574
575 /**
576 * Free a NULL-terminated list of strbufs (for example, the return
577 * values of the strbuf_split*() functions).
578 */
579 void strbuf_list_free(struct strbuf **list);
580
581 /**
582 * Add the abbreviation, as generated by find_unique_abbrev, of `sha1` to
583 * the strbuf `sb`.
584 */
585 void strbuf_add_unique_abbrev(struct strbuf *sb,
586 const struct object_id *oid,
587 int abbrev_len);
588
589 /**
590 * Launch the user preferred editor to edit a file and fill the buffer
591 * with the file's contents upon the user completing their editing. The
592 * third argument can be used to set the environment which the editor is
593 * run in. If the buffer is NULL the editor is launched as usual but the
594 * file's contents are not read into the buffer upon completion.
595 */
596 int launch_editor(const char *path, struct strbuf *buffer,
597 const char *const *env);
598
599 int launch_sequence_editor(const char *path, struct strbuf *buffer,
600 const char *const *env);
601
602 void strbuf_add_lines(struct strbuf *sb,
603 const char *prefix,
604 const char *buf,
605 size_t size);
606
607 /**
608 * Append s to sb, with the characters '<', '>', '&' and '"' converted
609 * into XML entities.
610 */
611 void strbuf_addstr_xml_quoted(struct strbuf *sb,
612 const char *s);
613
614 /**
615 * "Complete" the contents of `sb` by ensuring that either it ends with the
616 * character `term`, or it is empty. This can be used, for example,
617 * to ensure that text ends with a newline, but without creating an empty
618 * blank line if there is no content in the first place.
619 */
620 static inline void strbuf_complete(struct strbuf *sb, char term)
621 {
622 if (sb->len && sb->buf[sb->len - 1] != term)
623 strbuf_addch(sb, term);
624 }
625
626 static inline void strbuf_complete_line(struct strbuf *sb)
627 {
628 strbuf_complete(sb, '\n');
629 }
630
631 /*
632 * Copy "name" to "sb", expanding any special @-marks as handled by
633 * interpret_branch_name(). The result is a non-qualified branch name
634 * (so "foo" or "origin/master" instead of "refs/heads/foo" or
635 * "refs/remotes/origin/master").
636 *
637 * Note that the resulting name may not be a syntactically valid refname.
638 *
639 * If "allowed" is non-zero, restrict the set of allowed expansions. See
640 * interpret_branch_name() for details.
641 */
642 void strbuf_branchname(struct strbuf *sb, const char *name,
643 unsigned allowed);
644
645 /*
646 * Like strbuf_branchname() above, but confirm that the result is
647 * syntactically valid to be used as a local branch name in refs/heads/.
648 *
649 * The return value is "0" if the result is valid, and "-1" otherwise.
650 */
651 int strbuf_check_branch_ref(struct strbuf *sb, const char *name);
652
653 void strbuf_addstr_urlencode(struct strbuf *sb, const char *name,
654 int reserved);
655
656 __attribute__((format (printf,1,2)))
657 int printf_ln(const char *fmt, ...);
658 __attribute__((format (printf,2,3)))
659 int fprintf_ln(FILE *fp, const char *fmt, ...);
660
661 char *xstrdup_tolower(const char *);
662 char *xstrdup_toupper(const char *);
663
664 /**
665 * Create a newly allocated string using printf format. You can do this easily
666 * with a strbuf, but this provides a shortcut to save a few lines.
667 */
668 __attribute__((format (printf, 1, 0)))
669 char *xstrvfmt(const char *fmt, va_list ap);
670 __attribute__((format (printf, 1, 2)))
671 char *xstrfmt(const char *fmt, ...);
672
673 #endif /* STRBUF_H */