1#ifndef STRBUF_H 2#define STRBUF_H 3 4struct 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 */ 66struct strbuf { 67size_t alloc; 68size_t len; 69char*buf; 70}; 71 72externchar 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 */ 79struct 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 */ 90voidstrbuf_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 */ 100voidstrbuf_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 */ 110char*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 */ 120voidstrbuf_attach(struct strbuf *sb,void*str,size_t len,size_t mem); 121 122/** 123 * Swap the contents of two string buffers. 124 */ 125staticinlinevoidstrbuf_swap(struct strbuf *a,struct strbuf *b) 126{ 127SWAP(*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 */ 139staticinlinesize_tstrbuf_avail(const struct strbuf *sb) 140{ 141return 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 */ 151voidstrbuf_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 */ 160staticinlinevoidstrbuf_setlen(struct strbuf *sb,size_t len) 161{ 162if(len > (sb->alloc ? sb->alloc -1:0)) 163die("BUG: strbuf_setlen() beyond buffer"); 164 sb->len = len; 165if(sb->buf != strbuf_slopbuf) 166 sb->buf[len] ='\0'; 167else 168assert(!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 */ 186voidstrbuf_trim(struct strbuf *sb); 187voidstrbuf_rtrim(struct strbuf *sb); 188voidstrbuf_ltrim(struct strbuf *sb); 189 190/* Strip trailing directory separators */ 191voidstrbuf_trim_trailing_dir_sep(struct strbuf *sb); 192 193/** 194 * Replace the contents of the strbuf with a reencoded form. Returns -1 195 * on error, 0 on success. 196 */ 197intstrbuf_reencode(struct strbuf *sb,const char*from,const char*to); 198 199/** 200 * Lowercase each character in the buffer using `tolower`. 201 */ 202voidstrbuf_tolower(struct strbuf *sb); 203 204/** 205 * Compare two buffers. Returns an integer less than, equal to, or greater 206 * than zero if the first buffer is found, respectively, to be less than, 207 * to match, or be greater than the second buffer. 208 */ 209intstrbuf_cmp(const struct strbuf *first,const struct strbuf *second); 210 211 212/** 213 * Adding data to the buffer 214 * ------------------------- 215 * 216 * NOTE: All of the functions in this section will grow the buffer as 217 * necessary. If they fail for some reason other than memory shortage and the 218 * buffer hadn't been allocated before (i.e. the `struct strbuf` was set to 219 * `STRBUF_INIT`), then they will free() it. 220 */ 221 222/** 223 * Add a single character to the buffer. 224 */ 225staticinlinevoidstrbuf_addch(struct strbuf *sb,int c) 226{ 227if(!strbuf_avail(sb)) 228strbuf_grow(sb,1); 229 sb->buf[sb->len++] = c; 230 sb->buf[sb->len] ='\0'; 231} 232 233/** 234 * Add a character the specified number of times to the buffer. 235 */ 236voidstrbuf_addchars(struct strbuf *sb,int c,size_t n); 237 238/** 239 * Insert data to the given position of the buffer. The remaining contents 240 * will be shifted, not overwritten. 241 */ 242voidstrbuf_insert(struct strbuf *sb,size_t pos,const void*,size_t); 243 244/** 245 * Remove given amount of data from a given position of the buffer. 246 */ 247voidstrbuf_remove(struct strbuf *sb,size_t pos,size_t len); 248 249/** 250 * Remove the bytes between `pos..pos+len` and replace it with the given 251 * data. 252 */ 253voidstrbuf_splice(struct strbuf *sb,size_t pos,size_t len, 254const void*data,size_t data_len); 255 256/** 257 * Add a NUL-terminated string to the buffer. Each line will be prepended 258 * by a comment character and a blank. 259 */ 260voidstrbuf_add_commented_lines(struct strbuf *out, 261const char*buf,size_t size); 262 263 264/** 265 * Add data of given length to the buffer. 266 */ 267voidstrbuf_add(struct strbuf *sb,const void*data,size_t len); 268 269/** 270 * Add a NUL-terminated string to the buffer. 271 * 272 * NOTE: This function will *always* be implemented as an inline or a macro 273 * using strlen, meaning that this is efficient to write things like: 274 * 275 * strbuf_addstr(sb, "immediate string"); 276 * 277 */ 278staticinlinevoidstrbuf_addstr(struct strbuf *sb,const char*s) 279{ 280strbuf_add(sb, s,strlen(s)); 281} 282 283/** 284 * Copy the contents of another buffer at the end of the current one. 285 */ 286voidstrbuf_addbuf(struct strbuf *sb,const struct strbuf *sb2); 287 288/** 289 * This function can be used to expand a format string containing 290 * placeholders. To that end, it parses the string and calls the specified 291 * function for every percent sign found. 292 * 293 * The callback function is given a pointer to the character after the `%` 294 * and a pointer to the struct strbuf. It is expected to add the expanded 295 * version of the placeholder to the strbuf, e.g. to add a newline 296 * character if the letter `n` appears after a `%`. The function returns 297 * the length of the placeholder recognized and `strbuf_expand()` skips 298 * over it. 299 * 300 * The format `%%` is automatically expanded to a single `%` as a quoting 301 * mechanism; callers do not need to handle the `%` placeholder themselves, 302 * and the callback function will not be invoked for this placeholder. 303 * 304 * All other characters (non-percent and not skipped ones) are copied 305 * verbatim to the strbuf. If the callback returned zero, meaning that the 306 * placeholder is unknown, then the percent sign is copied, too. 307 * 308 * In order to facilitate caching and to make it possible to give 309 * parameters to the callback, `strbuf_expand()` passes a context pointer, 310 * which can be used by the programmer of the callback as she sees fit. 311 */ 312typedefsize_t(*expand_fn_t) (struct strbuf *sb, 313const char*placeholder, 314void*context); 315voidstrbuf_expand(struct strbuf *sb, 316const char*format, 317 expand_fn_t fn, 318void*context); 319 320/** 321 * Used as callback for `strbuf_expand()`, expects an array of 322 * struct strbuf_expand_dict_entry as context, i.e. pairs of 323 * placeholder and replacement string. The array needs to be 324 * terminated by an entry with placeholder set to NULL. 325 */ 326struct strbuf_expand_dict_entry { 327const char*placeholder; 328const char*value; 329}; 330size_tstrbuf_expand_dict_cb(struct strbuf *sb, 331const char*placeholder, 332void*context); 333 334/** 335 * Append the contents of one strbuf to another, quoting any 336 * percent signs ("%") into double-percents ("%%") in the 337 * destination. This is useful for literal data to be fed to either 338 * strbuf_expand or to the *printf family of functions. 339 */ 340voidstrbuf_addbuf_percentquote(struct strbuf *dst,const struct strbuf *src); 341 342/** 343 * Append the given byte size as a human-readable string (i.e. 12.23 KiB, 344 * 3.50 MiB). 345 */ 346voidstrbuf_humanise_bytes(struct strbuf *buf, off_t bytes); 347 348/** 349 * Add a formatted string to the buffer. 350 */ 351__attribute__((format(printf,2,3))) 352voidstrbuf_addf(struct strbuf *sb,const char*fmt, ...); 353 354/** 355 * Add a formatted string prepended by a comment character and a 356 * blank to the buffer. 357 */ 358__attribute__((format(printf,2,3))) 359voidstrbuf_commented_addf(struct strbuf *sb,const char*fmt, ...); 360 361__attribute__((format(printf,2,0))) 362voidstrbuf_vaddf(struct strbuf *sb,const char*fmt,va_list ap); 363 364/** 365 * Add the time specified by `tm`, as formatted by `strftime`. 366 * `tz_offset` is in decimal hhmm format, e.g. -600 means six hours west 367 * of Greenwich, and it's used to expand %z internally. However, tokens 368 * with modifiers (e.g. %Ez) are passed to `strftime`. 369 * `suppress_tz_name`, when set, expands %Z internally to the empty 370 * string rather than passing it to `strftime`. 371 */ 372voidstrbuf_addftime(struct strbuf *sb,const char*fmt, 373const struct tm *tm,int tz_offset, 374int suppress_tz_name); 375 376/** 377 * Read a given size of data from a FILE* pointer to the buffer. 378 * 379 * NOTE: The buffer is rewound if the read fails. If -1 is returned, 380 * `errno` must be consulted, like you would do for `read(3)`. 381 * `strbuf_read()`, `strbuf_read_file()` and `strbuf_getline_*()` 382 * family of functions have the same behaviour as well. 383 */ 384size_tstrbuf_fread(struct strbuf *sb,size_t size,FILE*file); 385 386/** 387 * Read the contents of a given file descriptor. The third argument can be 388 * used to give a hint about the file size, to avoid reallocs. If read fails, 389 * any partial read is undone. 390 */ 391ssize_t strbuf_read(struct strbuf *sb,int fd,size_t hint); 392 393/** 394 * Read the contents of a given file descriptor partially by using only one 395 * attempt of xread. The third argument can be used to give a hint about the 396 * file size, to avoid reallocs. Returns the number of new bytes appended to 397 * the sb. 398 */ 399ssize_t strbuf_read_once(struct strbuf *sb,int fd,size_t hint); 400 401/** 402 * Read the contents of a file, specified by its path. The third argument 403 * can be used to give a hint about the file size, to avoid reallocs. 404 * Return the number of bytes read or a negative value if some error 405 * occurred while opening or reading the file. 406 */ 407ssize_t strbuf_read_file(struct strbuf *sb,const char*path,size_t hint); 408 409/** 410 * Read the target of a symbolic link, specified by its path. The third 411 * argument can be used to give a hint about the size, to avoid reallocs. 412 */ 413intstrbuf_readlink(struct strbuf *sb,const char*path,size_t hint); 414 415/** 416 * Write the whole content of the strbuf to the stream not stopping at 417 * NUL bytes. 418 */ 419ssize_t strbuf_write(struct strbuf *sb,FILE*stream); 420 421/** 422 * Read a line from a FILE *, overwriting the existing contents of 423 * the strbuf. The strbuf_getline*() family of functions share 424 * this signature, but have different line termination conventions. 425 * 426 * Reading stops after the terminator or at EOF. The terminator 427 * is removed from the buffer before returning. Returns 0 unless 428 * there was nothing left before EOF, in which case it returns `EOF`. 429 */ 430typedefint(*strbuf_getline_fn)(struct strbuf *,FILE*); 431 432/* Uses LF as the line terminator */ 433intstrbuf_getline_lf(struct strbuf *sb,FILE*fp); 434 435/* Uses NUL as the line terminator */ 436intstrbuf_getline_nul(struct strbuf *sb,FILE*fp); 437 438/* 439 * Similar to strbuf_getline_lf(), but additionally treats a CR that 440 * comes immediately before the LF as part of the terminator. 441 * This is the most friendly version to be used to read "text" files 442 * that can come from platforms whose native text format is CRLF 443 * terminated. 444 */ 445intstrbuf_getline(struct strbuf *sb,FILE*file); 446 447 448/** 449 * Like `strbuf_getline`, but keeps the trailing terminator (if 450 * any) in the buffer. 451 */ 452intstrbuf_getwholeline(struct strbuf *sb,FILE*file,int term); 453 454/** 455 * Like `strbuf_getwholeline`, but operates on a file descriptor. 456 * It reads one character at a time, so it is very slow. Do not 457 * use it unless you need the correct position in the file 458 * descriptor. 459 */ 460intstrbuf_getwholeline_fd(struct strbuf *sb,int fd,int term); 461 462/** 463 * Set the buffer to the path of the current working directory. 464 */ 465intstrbuf_getcwd(struct strbuf *sb); 466 467/** 468 * Add a path to a buffer, converting a relative path to an 469 * absolute one in the process. Symbolic links are not 470 * resolved. 471 */ 472voidstrbuf_add_absolute_path(struct strbuf *sb,const char*path); 473 474/** 475 * Canonize `path` (make it absolute, resolve symlinks, remove extra 476 * slashes) and append it to `sb`. Die with an informative error 477 * message if there is a problem. 478 * 479 * The directory part of `path` (i.e., everything up to the last 480 * dir_sep) must denote a valid, existing directory, but the last 481 * component need not exist. 482 * 483 * Callers that don't mind links should use the more lightweight 484 * strbuf_add_absolute_path() instead. 485 */ 486voidstrbuf_add_real_path(struct strbuf *sb,const char*path); 487 488 489/** 490 * Normalize in-place the path contained in the strbuf. See 491 * normalize_path_copy() for details. If an error occurs, the contents of "sb" 492 * are left untouched, and -1 is returned. 493 */ 494intstrbuf_normalize_path(struct strbuf *sb); 495 496/** 497 * Strip whitespace from a buffer. The second parameter controls if 498 * comments are considered contents to be removed or not. 499 */ 500voidstrbuf_stripspace(struct strbuf *buf,int skip_comments); 501 502staticinlineintstrbuf_strip_suffix(struct strbuf *sb,const char*suffix) 503{ 504if(strip_suffix_mem(sb->buf, &sb->len, suffix)) { 505strbuf_setlen(sb, sb->len); 506return1; 507}else 508return0; 509} 510 511/** 512 * Split str (of length slen) at the specified terminator character. 513 * Return a null-terminated array of pointers to strbuf objects 514 * holding the substrings. The substrings include the terminator, 515 * except for the last substring, which might be unterminated if the 516 * original string did not end with a terminator. If max is positive, 517 * then split the string into at most max substrings (with the last 518 * substring containing everything following the (max-1)th terminator 519 * character). 520 * 521 * The most generic form is `strbuf_split_buf`, which takes an arbitrary 522 * pointer/len buffer. The `_str` variant takes a NUL-terminated string, 523 * the `_max` variant takes a strbuf, and just `strbuf_split` is a convenience 524 * wrapper to drop the `max` parameter. 525 * 526 * For lighter-weight alternatives, see string_list_split() and 527 * string_list_split_in_place(). 528 */ 529struct strbuf **strbuf_split_buf(const char*str,size_t len, 530int terminator,int max); 531 532staticinlinestruct strbuf **strbuf_split_str(const char*str, 533int terminator,int max) 534{ 535returnstrbuf_split_buf(str,strlen(str), terminator, max); 536} 537 538staticinlinestruct strbuf **strbuf_split_max(const struct strbuf *sb, 539int terminator,int max) 540{ 541returnstrbuf_split_buf(sb->buf, sb->len, terminator, max); 542} 543 544staticinlinestruct strbuf **strbuf_split(const struct strbuf *sb, 545int terminator) 546{ 547returnstrbuf_split_max(sb, terminator,0); 548} 549 550/* 551 * Adds all strings of a string list to the strbuf, separated by the given 552 * separator. For example, if sep is 553 * ', ' 554 * and slist contains 555 * ['element1', 'element2', ..., 'elementN'], 556 * then write: 557 * 'element1, element2, ..., elementN' 558 * to str. If only one element, just write "element1" to str. 559 */ 560voidstrbuf_add_separated_string_list(struct strbuf *str, 561const char*sep, 562struct string_list *slist); 563 564/** 565 * Free a NULL-terminated list of strbufs (for example, the return 566 * values of the strbuf_split*() functions). 567 */ 568voidstrbuf_list_free(struct strbuf **list); 569 570/** 571 * Add the abbreviation, as generated by find_unique_abbrev, of `sha1` to 572 * the strbuf `sb`. 573 */ 574voidstrbuf_add_unique_abbrev(struct strbuf *sb, 575const struct object_id *oid, 576int abbrev_len); 577 578/** 579 * Launch the user preferred editor to edit a file and fill the buffer 580 * with the file's contents upon the user completing their editing. The 581 * third argument can be used to set the environment which the editor is 582 * run in. If the buffer is NULL the editor is launched as usual but the 583 * file's contents are not read into the buffer upon completion. 584 */ 585intlaunch_editor(const char*path, 586struct strbuf *buffer, 587const char*const*env); 588 589voidstrbuf_add_lines(struct strbuf *sb, 590const char*prefix, 591const char*buf, 592size_t size); 593 594/** 595 * Append s to sb, with the characters '<', '>', '&' and '"' converted 596 * into XML entities. 597 */ 598voidstrbuf_addstr_xml_quoted(struct strbuf *sb, 599const char*s); 600 601/** 602 * "Complete" the contents of `sb` by ensuring that either it ends with the 603 * character `term`, or it is empty. This can be used, for example, 604 * to ensure that text ends with a newline, but without creating an empty 605 * blank line if there is no content in the first place. 606 */ 607staticinlinevoidstrbuf_complete(struct strbuf *sb,char term) 608{ 609if(sb->len && sb->buf[sb->len -1] != term) 610strbuf_addch(sb, term); 611} 612 613staticinlinevoidstrbuf_complete_line(struct strbuf *sb) 614{ 615strbuf_complete(sb,'\n'); 616} 617 618/* 619 * Copy "name" to "sb", expanding any special @-marks as handled by 620 * interpret_branch_name(). The result is a non-qualified branch name 621 * (so "foo" or "origin/master" instead of "refs/heads/foo" or 622 * "refs/remotes/origin/master"). 623 * 624 * Note that the resulting name may not be a syntactically valid refname. 625 * 626 * If "allowed" is non-zero, restrict the set of allowed expansions. See 627 * interpret_branch_name() for details. 628 */ 629voidstrbuf_branchname(struct strbuf *sb,const char*name, 630unsigned allowed); 631 632/* 633 * Like strbuf_branchname() above, but confirm that the result is 634 * syntactically valid to be used as a local branch name in refs/heads/. 635 * 636 * The return value is "0" if the result is valid, and "-1" otherwise. 637 */ 638intstrbuf_check_branch_ref(struct strbuf *sb,const char*name); 639 640voidstrbuf_addstr_urlencode(struct strbuf *sb,const char*name, 641int reserved); 642 643__attribute__((format(printf,1,2))) 644intprintf_ln(const char*fmt, ...); 645__attribute__((format(printf,2,3))) 646intfprintf_ln(FILE*fp,const char*fmt, ...); 647 648char*xstrdup_tolower(const char*); 649char*xstrdup_toupper(const char*); 650 651/** 652 * Create a newly allocated string using printf format. You can do this easily 653 * with a strbuf, but this provides a shortcut to save a few lines. 654 */ 655__attribute__((format(printf,1,0))) 656char*xstrvfmt(const char*fmt,va_list ap); 657__attribute__((format(printf,1,2))) 658char*xstrfmt(const char*fmt, ...); 659 660#endif/* STRBUF_H */