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