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/* Strip trailing LF or CR/LF */ 194voidstrbuf_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 */ 200intstrbuf_reencode(struct strbuf *sb,const char*from,const char*to); 201 202/** 203 * Lowercase each character in the buffer using `tolower`. 204 */ 205voidstrbuf_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 */ 212intstrbuf_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 */ 228staticinlinevoidstrbuf_addch(struct strbuf *sb,int c) 229{ 230if(!strbuf_avail(sb)) 231strbuf_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 */ 239voidstrbuf_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 */ 245voidstrbuf_insert(struct strbuf *sb,size_t pos,const void*,size_t); 246 247/** 248 * Insert data to the given position of the buffer giving a printf format 249 * string. The contents will be shifted, not overwritten. 250 */ 251voidstrbuf_vinsertf(struct strbuf *sb,size_t pos,const char*fmt, 252va_list ap); 253 254voidstrbuf_insertf(struct strbuf *sb,size_t pos,const char*fmt, ...); 255 256/** 257 * Remove given amount of data from a given position of the buffer. 258 */ 259voidstrbuf_remove(struct strbuf *sb,size_t pos,size_t len); 260 261/** 262 * Remove the bytes between `pos..pos+len` and replace it with the given 263 * data. 264 */ 265voidstrbuf_splice(struct strbuf *sb,size_t pos,size_t len, 266const void*data,size_t data_len); 267 268/** 269 * Add a NUL-terminated string to the buffer. Each line will be prepended 270 * by a comment character and a blank. 271 */ 272voidstrbuf_add_commented_lines(struct strbuf *out, 273const char*buf,size_t size); 274 275 276/** 277 * Add data of given length to the buffer. 278 */ 279voidstrbuf_add(struct strbuf *sb,const void*data,size_t len); 280 281/** 282 * Add a NUL-terminated string to the buffer. 283 * 284 * NOTE: This function will *always* be implemented as an inline or a macro 285 * using strlen, meaning that this is efficient to write things like: 286 * 287 * strbuf_addstr(sb, "immediate string"); 288 * 289 */ 290staticinlinevoidstrbuf_addstr(struct strbuf *sb,const char*s) 291{ 292strbuf_add(sb, s,strlen(s)); 293} 294 295/** 296 * Copy the contents of another buffer at the end of the current one. 297 */ 298voidstrbuf_addbuf(struct strbuf *sb,const struct strbuf *sb2); 299 300/** 301 * Join the arguments into a buffer. `delim` is put between every 302 * two arguments. 303 */ 304const char*strbuf_join_argv(struct strbuf *buf,int argc, 305const char**argv,char delim); 306 307/** 308 * This function can be used to expand a format string containing 309 * placeholders. To that end, it parses the string and calls the specified 310 * function for every percent sign found. 311 * 312 * The callback function is given a pointer to the character after the `%` 313 * and a pointer to the struct strbuf. It is expected to add the expanded 314 * version of the placeholder to the strbuf, e.g. to add a newline 315 * character if the letter `n` appears after a `%`. The function returns 316 * the length of the placeholder recognized and `strbuf_expand()` skips 317 * over it. 318 * 319 * The format `%%` is automatically expanded to a single `%` as a quoting 320 * mechanism; callers do not need to handle the `%` placeholder themselves, 321 * and the callback function will not be invoked for this placeholder. 322 * 323 * All other characters (non-percent and not skipped ones) are copied 324 * verbatim to the strbuf. If the callback returned zero, meaning that the 325 * placeholder is unknown, then the percent sign is copied, too. 326 * 327 * In order to facilitate caching and to make it possible to give 328 * parameters to the callback, `strbuf_expand()` passes a context pointer, 329 * which can be used by the programmer of the callback as she sees fit. 330 */ 331typedefsize_t(*expand_fn_t) (struct strbuf *sb, 332const char*placeholder, 333void*context); 334voidstrbuf_expand(struct strbuf *sb, 335const char*format, 336 expand_fn_t fn, 337void*context); 338 339/** 340 * Used as callback for `strbuf_expand()`, expects an array of 341 * struct strbuf_expand_dict_entry as context, i.e. pairs of 342 * placeholder and replacement string. The array needs to be 343 * terminated by an entry with placeholder set to NULL. 344 */ 345struct strbuf_expand_dict_entry { 346const char*placeholder; 347const char*value; 348}; 349size_tstrbuf_expand_dict_cb(struct strbuf *sb, 350const char*placeholder, 351void*context); 352 353/** 354 * Append the contents of one strbuf to another, quoting any 355 * percent signs ("%") into double-percents ("%%") in the 356 * destination. This is useful for literal data to be fed to either 357 * strbuf_expand or to the *printf family of functions. 358 */ 359voidstrbuf_addbuf_percentquote(struct strbuf *dst,const struct strbuf *src); 360 361/** 362 * Append the given byte size as a human-readable string (i.e. 12.23 KiB, 363 * 3.50 MiB). 364 */ 365voidstrbuf_humanise_bytes(struct strbuf *buf, off_t bytes); 366 367/** 368 * Add a formatted string to the buffer. 369 */ 370__attribute__((format(printf,2,3))) 371voidstrbuf_addf(struct strbuf *sb,const char*fmt, ...); 372 373/** 374 * Add a formatted string prepended by a comment character and a 375 * blank to the buffer. 376 */ 377__attribute__((format(printf,2,3))) 378voidstrbuf_commented_addf(struct strbuf *sb,const char*fmt, ...); 379 380__attribute__((format(printf,2,0))) 381voidstrbuf_vaddf(struct strbuf *sb,const char*fmt,va_list ap); 382 383/** 384 * Add the time specified by `tm`, as formatted by `strftime`. 385 * `tz_offset` is in decimal hhmm format, e.g. -600 means six hours west 386 * of Greenwich, and it's used to expand %z internally. However, tokens 387 * with modifiers (e.g. %Ez) are passed to `strftime`. 388 * `suppress_tz_name`, when set, expands %Z internally to the empty 389 * string rather than passing it to `strftime`. 390 */ 391voidstrbuf_addftime(struct strbuf *sb,const char*fmt, 392const struct tm *tm,int tz_offset, 393int suppress_tz_name); 394 395/** 396 * Read a given size of data from a FILE* pointer to the buffer. 397 * 398 * NOTE: The buffer is rewound if the read fails. If -1 is returned, 399 * `errno` must be consulted, like you would do for `read(3)`. 400 * `strbuf_read()`, `strbuf_read_file()` and `strbuf_getline_*()` 401 * family of functions have the same behaviour as well. 402 */ 403size_tstrbuf_fread(struct strbuf *sb,size_t size,FILE*file); 404 405/** 406 * Read the contents of a given file descriptor. The third argument can be 407 * used to give a hint about the file size, to avoid reallocs. If read fails, 408 * any partial read is undone. 409 */ 410ssize_t strbuf_read(struct strbuf *sb,int fd,size_t hint); 411 412/** 413 * Read the contents of a given file descriptor partially by using only one 414 * attempt of xread. The third argument can be used to give a hint about the 415 * file size, to avoid reallocs. Returns the number of new bytes appended to 416 * the sb. 417 */ 418ssize_t strbuf_read_once(struct strbuf *sb,int fd,size_t hint); 419 420/** 421 * Read the contents of a file, specified by its path. The third argument 422 * can be used to give a hint about the file size, to avoid reallocs. 423 * Return the number of bytes read or a negative value if some error 424 * occurred while opening or reading the file. 425 */ 426ssize_t strbuf_read_file(struct strbuf *sb,const char*path,size_t hint); 427 428/** 429 * Read the target of a symbolic link, specified by its path. The third 430 * argument can be used to give a hint about the size, to avoid reallocs. 431 */ 432intstrbuf_readlink(struct strbuf *sb,const char*path,size_t hint); 433 434/** 435 * Write the whole content of the strbuf to the stream not stopping at 436 * NUL bytes. 437 */ 438ssize_t strbuf_write(struct strbuf *sb,FILE*stream); 439 440/** 441 * Read a line from a FILE *, overwriting the existing contents of 442 * the strbuf. The strbuf_getline*() family of functions share 443 * this signature, but have different line termination conventions. 444 * 445 * Reading stops after the terminator or at EOF. The terminator 446 * is removed from the buffer before returning. Returns 0 unless 447 * there was nothing left before EOF, in which case it returns `EOF`. 448 */ 449typedefint(*strbuf_getline_fn)(struct strbuf *,FILE*); 450 451/* Uses LF as the line terminator */ 452intstrbuf_getline_lf(struct strbuf *sb,FILE*fp); 453 454/* Uses NUL as the line terminator */ 455intstrbuf_getline_nul(struct strbuf *sb,FILE*fp); 456 457/* 458 * Similar to strbuf_getline_lf(), but additionally treats a CR that 459 * comes immediately before the LF as part of the terminator. 460 * This is the most friendly version to be used to read "text" files 461 * that can come from platforms whose native text format is CRLF 462 * terminated. 463 */ 464intstrbuf_getline(struct strbuf *sb,FILE*file); 465 466 467/** 468 * Like `strbuf_getline`, but keeps the trailing terminator (if 469 * any) in the buffer. 470 */ 471intstrbuf_getwholeline(struct strbuf *sb,FILE*file,int term); 472 473/** 474 * Like `strbuf_getwholeline`, but operates on a file descriptor. 475 * It reads one character at a time, so it is very slow. Do not 476 * use it unless you need the correct position in the file 477 * descriptor. 478 */ 479intstrbuf_getwholeline_fd(struct strbuf *sb,int fd,int term); 480 481/** 482 * Set the buffer to the path of the current working directory. 483 */ 484intstrbuf_getcwd(struct strbuf *sb); 485 486/** 487 * Add a path to a buffer, converting a relative path to an 488 * absolute one in the process. Symbolic links are not 489 * resolved. 490 */ 491voidstrbuf_add_absolute_path(struct strbuf *sb,const char*path); 492 493/** 494 * Canonize `path` (make it absolute, resolve symlinks, remove extra 495 * slashes) and append it to `sb`. Die with an informative error 496 * message if there is a problem. 497 * 498 * The directory part of `path` (i.e., everything up to the last 499 * dir_sep) must denote a valid, existing directory, but the last 500 * component need not exist. 501 * 502 * Callers that don't mind links should use the more lightweight 503 * strbuf_add_absolute_path() instead. 504 */ 505voidstrbuf_add_real_path(struct strbuf *sb,const char*path); 506 507 508/** 509 * Normalize in-place the path contained in the strbuf. See 510 * normalize_path_copy() for details. If an error occurs, the contents of "sb" 511 * are left untouched, and -1 is returned. 512 */ 513intstrbuf_normalize_path(struct strbuf *sb); 514 515/** 516 * Strip whitespace from a buffer. The second parameter controls if 517 * comments are considered contents to be removed or not. 518 */ 519voidstrbuf_stripspace(struct strbuf *buf,int skip_comments); 520 521staticinlineintstrbuf_strip_suffix(struct strbuf *sb,const char*suffix) 522{ 523if(strip_suffix_mem(sb->buf, &sb->len, suffix)) { 524strbuf_setlen(sb, sb->len); 525return1; 526}else 527return0; 528} 529 530/** 531 * Split str (of length slen) at the specified terminator character. 532 * Return a null-terminated array of pointers to strbuf objects 533 * holding the substrings. The substrings include the terminator, 534 * except for the last substring, which might be unterminated if the 535 * original string did not end with a terminator. If max is positive, 536 * then split the string into at most max substrings (with the last 537 * substring containing everything following the (max-1)th terminator 538 * character). 539 * 540 * The most generic form is `strbuf_split_buf`, which takes an arbitrary 541 * pointer/len buffer. The `_str` variant takes a NUL-terminated string, 542 * the `_max` variant takes a strbuf, and just `strbuf_split` is a convenience 543 * wrapper to drop the `max` parameter. 544 * 545 * For lighter-weight alternatives, see string_list_split() and 546 * string_list_split_in_place(). 547 */ 548struct strbuf **strbuf_split_buf(const char*str,size_t len, 549int terminator,int max); 550 551staticinlinestruct strbuf **strbuf_split_str(const char*str, 552int terminator,int max) 553{ 554returnstrbuf_split_buf(str,strlen(str), terminator, max); 555} 556 557staticinlinestruct strbuf **strbuf_split_max(const struct strbuf *sb, 558int terminator,int max) 559{ 560returnstrbuf_split_buf(sb->buf, sb->len, terminator, max); 561} 562 563staticinlinestruct strbuf **strbuf_split(const struct strbuf *sb, 564int terminator) 565{ 566returnstrbuf_split_max(sb, terminator,0); 567} 568 569/* 570 * Adds all strings of a string list to the strbuf, separated by the given 571 * separator. For example, if sep is 572 * ', ' 573 * and slist contains 574 * ['element1', 'element2', ..., 'elementN'], 575 * then write: 576 * 'element1, element2, ..., elementN' 577 * to str. If only one element, just write "element1" to str. 578 */ 579voidstrbuf_add_separated_string_list(struct strbuf *str, 580const char*sep, 581struct string_list *slist); 582 583/** 584 * Free a NULL-terminated list of strbufs (for example, the return 585 * values of the strbuf_split*() functions). 586 */ 587voidstrbuf_list_free(struct strbuf **list); 588 589/** 590 * Add the abbreviation, as generated by find_unique_abbrev, of `sha1` to 591 * the strbuf `sb`. 592 */ 593voidstrbuf_add_unique_abbrev(struct strbuf *sb, 594const struct object_id *oid, 595int abbrev_len); 596 597/** 598 * Launch the user preferred editor to edit a file and fill the buffer 599 * with the file's contents upon the user completing their editing. The 600 * third argument can be used to set the environment which the editor is 601 * run in. If the buffer is NULL the editor is launched as usual but the 602 * file's contents are not read into the buffer upon completion. 603 */ 604intlaunch_editor(const char*path,struct strbuf *buffer, 605const char*const*env); 606 607intlaunch_sequence_editor(const char*path,struct strbuf *buffer, 608const char*const*env); 609 610voidstrbuf_add_lines(struct strbuf *sb, 611const char*prefix, 612const char*buf, 613size_t size); 614 615/** 616 * Append s to sb, with the characters '<', '>', '&' and '"' converted 617 * into XML entities. 618 */ 619voidstrbuf_addstr_xml_quoted(struct strbuf *sb, 620const char*s); 621 622/** 623 * "Complete" the contents of `sb` by ensuring that either it ends with the 624 * character `term`, or it is empty. This can be used, for example, 625 * to ensure that text ends with a newline, but without creating an empty 626 * blank line if there is no content in the first place. 627 */ 628staticinlinevoidstrbuf_complete(struct strbuf *sb,char term) 629{ 630if(sb->len && sb->buf[sb->len -1] != term) 631strbuf_addch(sb, term); 632} 633 634staticinlinevoidstrbuf_complete_line(struct strbuf *sb) 635{ 636strbuf_complete(sb,'\n'); 637} 638 639/* 640 * Copy "name" to "sb", expanding any special @-marks as handled by 641 * interpret_branch_name(). The result is a non-qualified branch name 642 * (so "foo" or "origin/master" instead of "refs/heads/foo" or 643 * "refs/remotes/origin/master"). 644 * 645 * Note that the resulting name may not be a syntactically valid refname. 646 * 647 * If "allowed" is non-zero, restrict the set of allowed expansions. See 648 * interpret_branch_name() for details. 649 */ 650voidstrbuf_branchname(struct strbuf *sb,const char*name, 651unsigned allowed); 652 653/* 654 * Like strbuf_branchname() above, but confirm that the result is 655 * syntactically valid to be used as a local branch name in refs/heads/. 656 * 657 * The return value is "0" if the result is valid, and "-1" otherwise. 658 */ 659intstrbuf_check_branch_ref(struct strbuf *sb,const char*name); 660 661voidstrbuf_addstr_urlencode(struct strbuf *sb,const char*name, 662int reserved); 663 664__attribute__((format(printf,1,2))) 665intprintf_ln(const char*fmt, ...); 666__attribute__((format(printf,2,3))) 667intfprintf_ln(FILE*fp,const char*fmt, ...); 668 669char*xstrdup_tolower(const char*); 670char*xstrdup_toupper(const char*); 671 672/** 673 * Create a newly allocated string using printf format. You can do this easily 674 * with a strbuf, but this provides a shortcut to save a few lines. 675 */ 676__attribute__((format(printf,1,0))) 677char*xstrvfmt(const char*fmt,va_list ap); 678__attribute__((format(printf,1,2))) 679char*xstrfmt(const char*fmt, ...); 680 681#endif/* STRBUF_H */