| 1 | /* |
| 2 | * info backend for Halibut |
| 3 | * |
| 4 | * Possible future work: |
| 5 | * |
| 6 | * - configurable choice of how to allocate node names? |
| 7 | * + possibly a template-like approach, choosing node names to |
| 8 | * be the full section title or perhaps the internal keyword? |
| 9 | * + neither of those seems quite right. Perhaps instead a |
| 10 | * Windows Help-like mechanism, where a magic config |
| 11 | * directive allows user choice of name for every node. |
| 12 | * + Only trouble with that is, now what happens to the section |
| 13 | * numbers? Do they become completely vestigial and just sit |
| 14 | * in the title text of each node? Or do we keep them in the |
| 15 | * menus somehow? I think people might occasionally want to |
| 16 | * go to a section by number, if only because all the _other_ |
| 17 | * formats of the same document will reference the numbers |
| 18 | * all the time. So our menu lines could look like one of |
| 19 | * these: |
| 20 | * * Nodename: Section 1.2. Title of section. |
| 21 | * * Section 1.2: Nodename. Title of section. |
| 22 | * |
| 23 | * - might be helpful to diagnose duplicate node names! |
| 24 | */ |
| 25 | |
| 26 | #include <stdio.h> |
| 27 | #include <stdlib.h> |
| 28 | #include <assert.h> |
| 29 | #include "halibut.h" |
| 30 | |
| 31 | typedef struct { |
| 32 | char *filename; |
| 33 | int maxfilesize; |
| 34 | int charset; |
| 35 | int listindentbefore, listindentafter; |
| 36 | int indent_code, width, index_width; |
| 37 | wchar_t *bullet, *listsuffix; |
| 38 | wchar_t *startemph, *endemph; |
| 39 | wchar_t *lquote, *rquote; |
| 40 | wchar_t *sectsuffix, *underline; |
| 41 | wchar_t *rule; |
| 42 | wchar_t *index_text; |
| 43 | } infoconfig; |
| 44 | |
| 45 | typedef struct { |
| 46 | rdstringc output; |
| 47 | int charset; |
| 48 | charset_state state; |
| 49 | int wcmode; |
| 50 | } info_data; |
| 51 | #define EMPTY_INFO_DATA { { 0, 0, NULL }, 0, CHARSET_INIT_STATE, FALSE } |
| 52 | static const info_data empty_info_data = EMPTY_INFO_DATA; |
| 53 | |
| 54 | typedef struct node_tag node; |
| 55 | struct node_tag { |
| 56 | node *listnext; |
| 57 | node *up, *prev, *next, *lastchild; |
| 58 | int pos, started_menu, filenum; |
| 59 | char *name; |
| 60 | info_data text; |
| 61 | }; |
| 62 | |
| 63 | typedef struct { |
| 64 | char *text; |
| 65 | int length; |
| 66 | int nnodes, nodesize; |
| 67 | node **nodes; |
| 68 | } info_idx; |
| 69 | |
| 70 | static int info_rdadd(info_data *, wchar_t); |
| 71 | static int info_rdadds(info_data *, wchar_t const *); |
| 72 | static int info_rdaddc(info_data *, char); |
| 73 | static int info_rdaddsc(info_data *, char const *); |
| 74 | |
| 75 | static void info_heading(info_data *, word *, word *, int, infoconfig *); |
| 76 | static void info_rule(info_data *, int, int, infoconfig *); |
| 77 | static void info_para(info_data *, word *, wchar_t *, word *, keywordlist *, |
| 78 | int, int, int, infoconfig *); |
| 79 | static void info_codepara(info_data *, word *, int, int); |
| 80 | static void info_versionid(info_data *, word *, infoconfig *); |
| 81 | static void info_menu_item(info_data *, node *, paragraph *, infoconfig *); |
| 82 | static word *info_transform_wordlist(word *, keywordlist *); |
| 83 | static int info_check_index(word *, node *, indexdata *); |
| 84 | |
| 85 | static int info_rdaddwc(info_data *, word *, word *, int, infoconfig *); |
| 86 | |
| 87 | static node *info_node_new(char *name, int charset); |
| 88 | static char *info_node_name_for_para(paragraph *p, infoconfig *); |
| 89 | static char *info_node_name_for_text(wchar_t *text, infoconfig *); |
| 90 | |
| 91 | static infoconfig info_configure(paragraph *source) { |
| 92 | infoconfig ret; |
| 93 | paragraph *p; |
| 94 | |
| 95 | /* |
| 96 | * Defaults. |
| 97 | */ |
| 98 | ret.filename = dupstr("output.info"); |
| 99 | ret.maxfilesize = 64 << 10; |
| 100 | ret.charset = CS_ASCII; |
| 101 | ret.width = 70; |
| 102 | ret.listindentbefore = 1; |
| 103 | ret.listindentafter = 3; |
| 104 | ret.indent_code = 2; |
| 105 | ret.index_width = 40; |
| 106 | ret.listsuffix = L"."; |
| 107 | ret.bullet = L"\x2022\0-\0\0"; |
| 108 | ret.rule = L"\x2500\0-\0\0"; |
| 109 | ret.startemph = L"_\0_\0\0"; |
| 110 | ret.endemph = uadv(ret.startemph); |
| 111 | ret.lquote = L"\x2018\0\x2019\0`\0'\0\0"; |
| 112 | ret.rquote = uadv(ret.lquote); |
| 113 | ret.sectsuffix = L": "; |
| 114 | ret.underline = L"\x203E\0-\0\0"; |
| 115 | ret.index_text = L"Index"; |
| 116 | |
| 117 | /* |
| 118 | * Two-pass configuration so that we can pick up global config |
| 119 | * (e.g. `quotes') before having it overridden by specific |
| 120 | * config (`info-quotes'), irrespective of the order in which |
| 121 | * they occur. |
| 122 | */ |
| 123 | for (p = source; p; p = p->next) { |
| 124 | if (p->type == para_Config) { |
| 125 | if (!ustricmp(p->keyword, L"quotes")) { |
| 126 | if (*uadv(p->keyword) && *uadv(uadv(p->keyword))) { |
| 127 | ret.lquote = uadv(p->keyword); |
| 128 | ret.rquote = uadv(ret.lquote); |
| 129 | } |
| 130 | } else if (!ustricmp(p->keyword, L"index")) { |
| 131 | ret.index_text = uadv(p->keyword); |
| 132 | } |
| 133 | } |
| 134 | } |
| 135 | |
| 136 | for (p = source; p; p = p->next) { |
| 137 | if (p->type == para_Config) { |
| 138 | if (!ustricmp(p->keyword, L"info-filename")) { |
| 139 | sfree(ret.filename); |
| 140 | ret.filename = dupstr(adv(p->origkeyword)); |
| 141 | } else if (!ustricmp(p->keyword, L"info-charset")) { |
| 142 | ret.charset = charset_from_ustr(&p->fpos, uadv(p->keyword)); |
| 143 | } else if (!ustricmp(p->keyword, L"info-max-file-size")) { |
| 144 | ret.maxfilesize = utoi(uadv(p->keyword)); |
| 145 | } else if (!ustricmp(p->keyword, L"info-width")) { |
| 146 | ret.width = utoi(uadv(p->keyword)); |
| 147 | } else if (!ustricmp(p->keyword, L"info-indent-code")) { |
| 148 | ret.indent_code = utoi(uadv(p->keyword)); |
| 149 | } else if (!ustricmp(p->keyword, L"info-index-width")) { |
| 150 | ret.index_width = utoi(uadv(p->keyword)); |
| 151 | } else if (!ustricmp(p->keyword, L"info-list-indent")) { |
| 152 | ret.listindentbefore = utoi(uadv(p->keyword)); |
| 153 | } else if (!ustricmp(p->keyword, L"info-listitem-indent")) { |
| 154 | ret.listindentafter = utoi(uadv(p->keyword)); |
| 155 | } else if (!ustricmp(p->keyword, L"info-section-suffix")) { |
| 156 | ret.sectsuffix = uadv(p->keyword); |
| 157 | } else if (!ustricmp(p->keyword, L"info-underline")) { |
| 158 | ret.underline = uadv(p->keyword); |
| 159 | } else if (!ustricmp(p->keyword, L"info-bullet")) { |
| 160 | ret.bullet = uadv(p->keyword); |
| 161 | } else if (!ustricmp(p->keyword, L"info-rule")) { |
| 162 | ret.rule = uadv(p->keyword); |
| 163 | } else if (!ustricmp(p->keyword, L"info-list-suffix")) { |
| 164 | ret.listsuffix = uadv(p->keyword); |
| 165 | } else if (!ustricmp(p->keyword, L"info-emphasis")) { |
| 166 | if (*uadv(p->keyword) && *uadv(uadv(p->keyword))) { |
| 167 | ret.startemph = uadv(p->keyword); |
| 168 | ret.endemph = uadv(ret.startemph); |
| 169 | } |
| 170 | } else if (!ustricmp(p->keyword, L"info-quotes")) { |
| 171 | if (*uadv(p->keyword) && *uadv(uadv(p->keyword))) { |
| 172 | ret.lquote = uadv(p->keyword); |
| 173 | ret.rquote = uadv(ret.lquote); |
| 174 | } |
| 175 | } |
| 176 | } |
| 177 | } |
| 178 | |
| 179 | /* |
| 180 | * Now process fallbacks on quote characters, underlines, the |
| 181 | * rule character, the emphasis characters, and bullets. |
| 182 | */ |
| 183 | while (*uadv(ret.rquote) && *uadv(uadv(ret.rquote)) && |
| 184 | (!cvt_ok(ret.charset, ret.lquote) || |
| 185 | !cvt_ok(ret.charset, ret.rquote))) { |
| 186 | ret.lquote = uadv(ret.rquote); |
| 187 | ret.rquote = uadv(ret.lquote); |
| 188 | } |
| 189 | |
| 190 | while (*uadv(ret.endemph) && *uadv(uadv(ret.endemph)) && |
| 191 | (!cvt_ok(ret.charset, ret.startemph) || |
| 192 | !cvt_ok(ret.charset, ret.endemph))) { |
| 193 | ret.startemph = uadv(ret.endemph); |
| 194 | ret.endemph = uadv(ret.startemph); |
| 195 | } |
| 196 | |
| 197 | while (*ret.underline && *uadv(ret.underline) && |
| 198 | !cvt_ok(ret.charset, ret.underline)) |
| 199 | ret.underline = uadv(ret.underline); |
| 200 | |
| 201 | while (*ret.bullet && *uadv(ret.bullet) && |
| 202 | !cvt_ok(ret.charset, ret.bullet)) |
| 203 | ret.bullet = uadv(ret.bullet); |
| 204 | |
| 205 | while (*ret.rule && *uadv(ret.rule) && |
| 206 | !cvt_ok(ret.charset, ret.rule)) |
| 207 | ret.rule = uadv(ret.rule); |
| 208 | |
| 209 | return ret; |
| 210 | } |
| 211 | |
| 212 | paragraph *info_config_filename(char *filename) |
| 213 | { |
| 214 | return cmdline_cfg_simple("info-filename", filename, NULL); |
| 215 | } |
| 216 | |
| 217 | void info_backend(paragraph *sourceform, keywordlist *keywords, |
| 218 | indexdata *idx, void *unused) { |
| 219 | paragraph *p; |
| 220 | infoconfig conf; |
| 221 | word *prefix, *body, *wp; |
| 222 | word spaceword; |
| 223 | wchar_t *prefixextra; |
| 224 | int nesting, nestindent; |
| 225 | int indentb, indenta; |
| 226 | int filepos; |
| 227 | int has_index; |
| 228 | info_data intro_text = EMPTY_INFO_DATA; |
| 229 | node *topnode, *currnode; |
| 230 | word bullet; |
| 231 | FILE *fp; |
| 232 | |
| 233 | IGNORE(unused); |
| 234 | |
| 235 | conf = info_configure(sourceform); |
| 236 | |
| 237 | /* |
| 238 | * Go through and create a node for each section. |
| 239 | */ |
| 240 | topnode = info_node_new("Top", conf.charset); |
| 241 | currnode = topnode; |
| 242 | for (p = sourceform; p; p = p->next) switch (p->type) { |
| 243 | /* |
| 244 | * Chapter titles. |
| 245 | */ |
| 246 | case para_Chapter: |
| 247 | case para_Appendix: |
| 248 | case para_UnnumberedChapter: |
| 249 | case para_Heading: |
| 250 | case para_Subsect: |
| 251 | { |
| 252 | node *newnode, *upnode; |
| 253 | char *nodename; |
| 254 | |
| 255 | nodename = info_node_name_for_para(p, &conf); |
| 256 | newnode = info_node_new(nodename, conf.charset); |
| 257 | sfree(nodename); |
| 258 | |
| 259 | p->private_data = newnode; |
| 260 | |
| 261 | if (p->parent) |
| 262 | upnode = (node *)p->parent->private_data; |
| 263 | else |
| 264 | upnode = topnode; |
| 265 | assert(upnode); |
| 266 | newnode->up = upnode; |
| 267 | |
| 268 | currnode->next = newnode; |
| 269 | newnode->prev = currnode; |
| 270 | |
| 271 | currnode->listnext = newnode; |
| 272 | currnode = newnode; |
| 273 | } |
| 274 | break; |
| 275 | default: |
| 276 | p->private_data = NULL; |
| 277 | break; |
| 278 | } |
| 279 | |
| 280 | /* |
| 281 | * Set up the display form of each index entry. |
| 282 | */ |
| 283 | { |
| 284 | int i; |
| 285 | indexentry *entry; |
| 286 | |
| 287 | for (i = 0; (entry = index234(idx->entries, i)) != NULL; i++) { |
| 288 | info_idx *ii = snew(info_idx); |
| 289 | info_data id = EMPTY_INFO_DATA; |
| 290 | |
| 291 | id.charset = conf.charset; |
| 292 | |
| 293 | ii->nnodes = ii->nodesize = 0; |
| 294 | ii->nodes = NULL; |
| 295 | |
| 296 | ii->length = info_rdaddwc(&id, entry->text, NULL, FALSE, &conf); |
| 297 | |
| 298 | ii->text = id.output.text; |
| 299 | |
| 300 | entry->backend_data = ii; |
| 301 | } |
| 302 | } |
| 303 | |
| 304 | /* |
| 305 | * An Info file begins with a piece of introductory text which |
| 306 | * is apparently never shown anywhere. This seems to me to be a |
| 307 | * good place to put the copyright notice and the version IDs. |
| 308 | * Also, Info directory entries are expected to go here. |
| 309 | */ |
| 310 | intro_text.charset = conf.charset; |
| 311 | |
| 312 | info_rdaddsc(&intro_text, |
| 313 | "This Info file generated by Halibut, "); |
| 314 | info_rdaddsc(&intro_text, version); |
| 315 | info_rdaddsc(&intro_text, "\n\n"); |
| 316 | |
| 317 | for (p = sourceform; p; p = p->next) |
| 318 | if (p->type == para_Config && |
| 319 | !ustricmp(p->keyword, L"info-dir-entry")) { |
| 320 | wchar_t *section, *shortname, *longname, *kw; |
| 321 | char *s; |
| 322 | |
| 323 | section = uadv(p->keyword); |
| 324 | shortname = *section ? uadv(section) : L""; |
| 325 | longname = *shortname ? uadv(shortname) : L""; |
| 326 | kw = *longname ? uadv(longname) : L""; |
| 327 | |
| 328 | if (!*longname) { |
| 329 | error(err_cfginsufarg, &p->fpos, p->origkeyword, 3); |
| 330 | continue; |
| 331 | } |
| 332 | |
| 333 | info_rdaddsc(&intro_text, "INFO-DIR-SECTION "); |
| 334 | info_rdadds(&intro_text, section); |
| 335 | info_rdaddsc(&intro_text, "\nSTART-INFO-DIR-ENTRY\n* "); |
| 336 | info_rdadds(&intro_text, shortname); |
| 337 | info_rdaddsc(&intro_text, ": ("); |
| 338 | s = dupstr(conf.filename); |
| 339 | if (strlen(s) > 5 && !strcmp(s+strlen(s)-5, ".info")) |
| 340 | s[strlen(s)-5] = '\0'; |
| 341 | info_rdaddsc(&intro_text, s); |
| 342 | sfree(s); |
| 343 | info_rdaddsc(&intro_text, ")"); |
| 344 | if (*kw) { |
| 345 | keyword *kwl = kw_lookup(keywords, kw); |
| 346 | if (kwl && kwl->para->private_data) { |
| 347 | node *n = (node *)kwl->para->private_data; |
| 348 | info_rdaddsc(&intro_text, n->name); |
| 349 | } |
| 350 | } |
| 351 | info_rdaddsc(&intro_text, ". "); |
| 352 | info_rdadds(&intro_text, longname); |
| 353 | info_rdaddsc(&intro_text, "\nEND-INFO-DIR-ENTRY\n\n"); |
| 354 | } |
| 355 | |
| 356 | for (p = sourceform; p; p = p->next) |
| 357 | if (p->type == para_Copyright) |
| 358 | info_para(&intro_text, NULL, NULL, p->words, keywords, |
| 359 | 0, 0, conf.width, &conf); |
| 360 | |
| 361 | for (p = sourceform; p; p = p->next) |
| 362 | if (p->type == para_VersionID) |
| 363 | info_versionid(&intro_text, p->words, &conf); |
| 364 | |
| 365 | if (intro_text.output.text[intro_text.output.pos-1] != '\n') |
| 366 | info_rdaddc(&intro_text, '\n'); |
| 367 | |
| 368 | /* Do the title */ |
| 369 | for (p = sourceform; p; p = p->next) |
| 370 | if (p->type == para_Title) |
| 371 | info_heading(&topnode->text, NULL, p->words, conf.width, &conf); |
| 372 | |
| 373 | nestindent = conf.listindentbefore + conf.listindentafter; |
| 374 | nesting = 0; |
| 375 | |
| 376 | currnode = topnode; |
| 377 | |
| 378 | /* Do the main document */ |
| 379 | for (p = sourceform; p; p = p->next) switch (p->type) { |
| 380 | |
| 381 | case para_QuotePush: |
| 382 | nesting += 2; |
| 383 | break; |
| 384 | case para_QuotePop: |
| 385 | nesting -= 2; |
| 386 | assert(nesting >= 0); |
| 387 | break; |
| 388 | |
| 389 | case para_LcontPush: |
| 390 | nesting += nestindent; |
| 391 | break; |
| 392 | case para_LcontPop: |
| 393 | nesting -= nestindent; |
| 394 | assert(nesting >= 0); |
| 395 | break; |
| 396 | |
| 397 | /* |
| 398 | * Things we ignore because we've already processed them or |
| 399 | * aren't going to touch them in this pass. |
| 400 | */ |
| 401 | case para_IM: |
| 402 | case para_BR: |
| 403 | case para_Biblio: /* only touch BiblioCited */ |
| 404 | case para_VersionID: |
| 405 | case para_NoCite: |
| 406 | case para_Title: |
| 407 | break; |
| 408 | |
| 409 | /* |
| 410 | * Chapter titles. |
| 411 | */ |
| 412 | case para_Chapter: |
| 413 | case para_Appendix: |
| 414 | case para_UnnumberedChapter: |
| 415 | case para_Heading: |
| 416 | case para_Subsect: |
| 417 | currnode = p->private_data; |
| 418 | assert(currnode); |
| 419 | assert(currnode->up); |
| 420 | |
| 421 | if (!currnode->up->started_menu) { |
| 422 | info_rdaddsc(&currnode->up->text, "* Menu:\n\n"); |
| 423 | currnode->up->started_menu = TRUE; |
| 424 | } |
| 425 | info_menu_item(&currnode->up->text, currnode, p, &conf); |
| 426 | |
| 427 | has_index |= info_check_index(p->words, currnode, idx); |
| 428 | info_heading(&currnode->text, p->kwtext, p->words, conf.width, &conf); |
| 429 | nesting = 0; |
| 430 | break; |
| 431 | |
| 432 | case para_Rule: |
| 433 | info_rule(&currnode->text, nesting, conf.width - nesting, &conf); |
| 434 | break; |
| 435 | |
| 436 | case para_Normal: |
| 437 | case para_Copyright: |
| 438 | case para_DescribedThing: |
| 439 | case para_Description: |
| 440 | case para_BiblioCited: |
| 441 | case para_Bullet: |
| 442 | case para_NumberedList: |
| 443 | has_index |= info_check_index(p->words, currnode, idx); |
| 444 | if (p->type == para_Bullet) { |
| 445 | bullet.next = NULL; |
| 446 | bullet.alt = NULL; |
| 447 | bullet.type = word_Normal; |
| 448 | bullet.text = conf.bullet; |
| 449 | prefix = • |
| 450 | prefixextra = NULL; |
| 451 | indentb = conf.listindentbefore; |
| 452 | indenta = conf.listindentafter; |
| 453 | } else if (p->type == para_NumberedList) { |
| 454 | prefix = p->kwtext; |
| 455 | prefixextra = conf.listsuffix; |
| 456 | indentb = conf.listindentbefore; |
| 457 | indenta = conf.listindentafter; |
| 458 | } else if (p->type == para_Description) { |
| 459 | prefix = NULL; |
| 460 | prefixextra = NULL; |
| 461 | indentb = conf.listindentbefore; |
| 462 | indenta = conf.listindentafter; |
| 463 | } else { |
| 464 | prefix = NULL; |
| 465 | prefixextra = NULL; |
| 466 | indentb = indenta = 0; |
| 467 | } |
| 468 | if (p->type == para_BiblioCited) { |
| 469 | body = dup_word_list(p->kwtext); |
| 470 | for (wp = body; wp->next; wp = wp->next); |
| 471 | wp->next = &spaceword; |
| 472 | spaceword.next = p->words; |
| 473 | spaceword.alt = NULL; |
| 474 | spaceword.type = word_WhiteSpace; |
| 475 | spaceword.text = NULL; |
| 476 | } else { |
| 477 | wp = NULL; |
| 478 | body = p->words; |
| 479 | } |
| 480 | info_para(&currnode->text, prefix, prefixextra, body, keywords, |
| 481 | nesting + indentb, indenta, |
| 482 | conf.width - nesting - indentb - indenta, &conf); |
| 483 | if (wp) { |
| 484 | wp->next = NULL; |
| 485 | free_word_list(body); |
| 486 | } |
| 487 | break; |
| 488 | |
| 489 | case para_Code: |
| 490 | info_codepara(&currnode->text, p->words, |
| 491 | nesting + conf.indent_code, |
| 492 | conf.width - nesting - 2 * conf.indent_code); |
| 493 | break; |
| 494 | } |
| 495 | |
| 496 | /* |
| 497 | * Create an index node if required. |
| 498 | */ |
| 499 | if (has_index) { |
| 500 | node *newnode; |
| 501 | int i, j, k; |
| 502 | indexentry *entry; |
| 503 | char *nodename; |
| 504 | |
| 505 | nodename = info_node_name_for_text(conf.index_text, &conf); |
| 506 | newnode = info_node_new(nodename, conf.charset); |
| 507 | sfree(nodename); |
| 508 | |
| 509 | newnode->up = topnode; |
| 510 | |
| 511 | currnode->next = newnode; |
| 512 | newnode->prev = currnode; |
| 513 | currnode->listnext = newnode; |
| 514 | |
| 515 | k = info_rdadds(&newnode->text, conf.index_text); |
| 516 | info_rdaddsc(&newnode->text, "\n"); |
| 517 | while (k > 0) { |
| 518 | info_rdadds(&newnode->text, conf.underline); |
| 519 | k -= ustrwid(conf.underline, conf.charset); |
| 520 | } |
| 521 | info_rdaddsc(&newnode->text, "\n\n"); |
| 522 | |
| 523 | info_menu_item(&topnode->text, newnode, NULL, &conf); |
| 524 | |
| 525 | for (i = 0; (entry = index234(idx->entries, i)) != NULL; i++) { |
| 526 | info_idx *ii = (info_idx *)entry->backend_data; |
| 527 | |
| 528 | for (j = 0; j < ii->nnodes; j++) { |
| 529 | /* |
| 530 | * When we have multiple references for a single |
| 531 | * index term, we only display the actual term on |
| 532 | * the first line, to make it clear that the terms |
| 533 | * really are the same. |
| 534 | */ |
| 535 | if (j == 0) |
| 536 | info_rdaddsc(&newnode->text, ii->text); |
| 537 | for (k = (j ? 0 : ii->length); k < conf.index_width-2; k++) |
| 538 | info_rdaddc(&newnode->text, ' '); |
| 539 | info_rdaddsc(&newnode->text, " *Note "); |
| 540 | info_rdaddsc(&newnode->text, ii->nodes[j]->name); |
| 541 | info_rdaddsc(&newnode->text, "::\n"); |
| 542 | } |
| 543 | } |
| 544 | } |
| 545 | |
| 546 | /* |
| 547 | * Finalise the text of each node, by adding the ^_ delimiter |
| 548 | * and the node line at the top. |
| 549 | */ |
| 550 | for (currnode = topnode; currnode; currnode = currnode->listnext) { |
| 551 | char *origtext = currnode->text.output.text; |
| 552 | currnode->text = empty_info_data; |
| 553 | currnode->text.charset = conf.charset; |
| 554 | info_rdaddsc(&currnode->text, "\037\nFile: "); |
| 555 | info_rdaddsc(&currnode->text, conf.filename); |
| 556 | info_rdaddsc(&currnode->text, ", Node: "); |
| 557 | info_rdaddsc(&currnode->text, currnode->name); |
| 558 | if (currnode->prev) { |
| 559 | info_rdaddsc(&currnode->text, ", Prev: "); |
| 560 | info_rdaddsc(&currnode->text, currnode->prev->name); |
| 561 | } |
| 562 | info_rdaddsc(&currnode->text, ", Up: "); |
| 563 | info_rdaddsc(&currnode->text, (currnode->up ? |
| 564 | currnode->up->name : "(dir)")); |
| 565 | if (currnode->next) { |
| 566 | info_rdaddsc(&currnode->text, ", Next: "); |
| 567 | info_rdaddsc(&currnode->text, currnode->next->name); |
| 568 | } |
| 569 | info_rdaddsc(&currnode->text, "\n\n"); |
| 570 | info_rdaddsc(&currnode->text, origtext); |
| 571 | /* |
| 572 | * Just make _absolutely_ sure we end with a newline. |
| 573 | */ |
| 574 | if (currnode->text.output.text[currnode->text.output.pos-1] != '\n') |
| 575 | info_rdaddc(&currnode->text, '\n'); |
| 576 | |
| 577 | sfree(origtext); |
| 578 | } |
| 579 | |
| 580 | /* |
| 581 | * Compute the offsets for the tag table. |
| 582 | */ |
| 583 | filepos = intro_text.output.pos; |
| 584 | for (currnode = topnode; currnode; currnode = currnode->listnext) { |
| 585 | currnode->pos = filepos; |
| 586 | filepos += currnode->text.output.pos; |
| 587 | } |
| 588 | |
| 589 | /* |
| 590 | * Split into sub-files. |
| 591 | */ |
| 592 | if (conf.maxfilesize > 0) { |
| 593 | int currfilesize = intro_text.output.pos, currfilenum = 1; |
| 594 | for (currnode = topnode; currnode; currnode = currnode->listnext) { |
| 595 | if (currfilesize > intro_text.output.pos && |
| 596 | currfilesize + currnode->text.output.pos > conf.maxfilesize) { |
| 597 | currfilenum++; |
| 598 | currfilesize = intro_text.output.pos; |
| 599 | } |
| 600 | currnode->filenum = currfilenum; |
| 601 | currfilesize += currnode->text.output.pos; |
| 602 | } |
| 603 | } |
| 604 | |
| 605 | /* |
| 606 | * Write the primary output file. |
| 607 | */ |
| 608 | fp = fopen(conf.filename, "w"); |
| 609 | if (!fp) { |
| 610 | error(err_cantopenw, conf.filename); |
| 611 | return; |
| 612 | } |
| 613 | fputs(intro_text.output.text, fp); |
| 614 | if (conf.maxfilesize == 0) { |
| 615 | for (currnode = topnode; currnode; currnode = currnode->listnext) |
| 616 | fputs(currnode->text.output.text, fp); |
| 617 | } else { |
| 618 | int filenum = 0; |
| 619 | fprintf(fp, "\037\nIndirect:\n"); |
| 620 | for (currnode = topnode; currnode; currnode = currnode->listnext) |
| 621 | if (filenum != currnode->filenum) { |
| 622 | filenum = currnode->filenum; |
| 623 | fprintf(fp, "%s-%d: %d\n", conf.filename, filenum, |
| 624 | currnode->pos); |
| 625 | } |
| 626 | } |
| 627 | fprintf(fp, "\037\nTag Table:\n"); |
| 628 | if (conf.maxfilesize > 0) |
| 629 | fprintf(fp, "(Indirect)\n"); |
| 630 | for (currnode = topnode; currnode; currnode = currnode->listnext) |
| 631 | fprintf(fp, "Node: %s\177%d\n", currnode->name, currnode->pos); |
| 632 | fprintf(fp, "\037\nEnd Tag Table\n"); |
| 633 | fclose(fp); |
| 634 | |
| 635 | /* |
| 636 | * Write the subfiles. |
| 637 | */ |
| 638 | if (conf.maxfilesize > 0) { |
| 639 | int filenum = 0; |
| 640 | fp = NULL; |
| 641 | |
| 642 | for (currnode = topnode; currnode; currnode = currnode->listnext) { |
| 643 | if (filenum != currnode->filenum) { |
| 644 | char *fname; |
| 645 | |
| 646 | filenum = currnode->filenum; |
| 647 | |
| 648 | if (fp) |
| 649 | fclose(fp); |
| 650 | fname = snewn(strlen(conf.filename) + 40, char); |
| 651 | sprintf(fname, "%s-%d", conf.filename, filenum); |
| 652 | fp = fopen(fname, "w"); |
| 653 | if (!fp) { |
| 654 | error(err_cantopenw, fname); |
| 655 | return; |
| 656 | } |
| 657 | sfree(fname); |
| 658 | fputs(intro_text.output.text, fp); |
| 659 | } |
| 660 | fputs(currnode->text.output.text, fp); |
| 661 | } |
| 662 | |
| 663 | if (fp) |
| 664 | fclose(fp); |
| 665 | } |
| 666 | } |
| 667 | |
| 668 | static int info_check_index(word *w, node *n, indexdata *idx) |
| 669 | { |
| 670 | int ret = 0; |
| 671 | |
| 672 | for (; w; w = w->next) { |
| 673 | if (w->type == word_IndexRef) { |
| 674 | indextag *tag; |
| 675 | int i; |
| 676 | |
| 677 | tag = index_findtag(idx, w->text); |
| 678 | if (!tag) |
| 679 | break; |
| 680 | |
| 681 | for (i = 0; i < tag->nrefs; i++) { |
| 682 | indexentry *entry = tag->refs[i]; |
| 683 | info_idx *ii = (info_idx *)entry->backend_data; |
| 684 | |
| 685 | if (ii->nnodes > 0 && ii->nodes[ii->nnodes-1] == n) { |
| 686 | /* |
| 687 | * If the same index term is indexed twice |
| 688 | * within the same section, we only want to |
| 689 | * mention it once in the index. So do nothing |
| 690 | * here. |
| 691 | */ |
| 692 | continue; |
| 693 | } |
| 694 | |
| 695 | if (ii->nnodes >= ii->nodesize) { |
| 696 | ii->nodesize += 32; |
| 697 | ii->nodes = sresize(ii->nodes, ii->nodesize, node *); |
| 698 | } |
| 699 | |
| 700 | ii->nodes[ii->nnodes++] = n; |
| 701 | |
| 702 | ret = 1; |
| 703 | } |
| 704 | } |
| 705 | } |
| 706 | |
| 707 | return ret; |
| 708 | } |
| 709 | |
| 710 | static word *info_transform_wordlist(word *words, keywordlist *keywords) |
| 711 | { |
| 712 | word *ret = dup_word_list(words); |
| 713 | word *w; |
| 714 | keyword *kwl; |
| 715 | |
| 716 | for (w = ret; w; w = w->next) { |
| 717 | w->private_data = NULL; |
| 718 | if (w->type == word_UpperXref || w->type == word_LowerXref) { |
| 719 | kwl = kw_lookup(keywords, w->text); |
| 720 | if (kwl) { |
| 721 | if (kwl->para->type == para_NumberedList || |
| 722 | kwl->para->type == para_BiblioCited) { |
| 723 | /* |
| 724 | * In Info, we do nothing special for xrefs to |
| 725 | * numbered list items or bibliography entries. |
| 726 | */ |
| 727 | continue; |
| 728 | } else { |
| 729 | /* |
| 730 | * An xref to a different section has its text |
| 731 | * completely replaced. |
| 732 | */ |
| 733 | word *w2, *w3, *w4; |
| 734 | w2 = w3 = w->next; |
| 735 | w4 = NULL; |
| 736 | while (w2) { |
| 737 | if (w2->type == word_XrefEnd) { |
| 738 | w4 = w2->next; |
| 739 | w2->next = NULL; |
| 740 | break; |
| 741 | } |
| 742 | w2 = w2->next; |
| 743 | } |
| 744 | free_word_list(w3); |
| 745 | |
| 746 | /* |
| 747 | * Now w is the UpperXref / LowerXref we |
| 748 | * started with, and w4 is the next word after |
| 749 | * the corresponding XrefEnd (if any). The |
| 750 | * simplest thing is just to stick a pointer to |
| 751 | * the target node structure in the private |
| 752 | * data field of the xref word, and let |
| 753 | * info_rdaddwc and friends read the node name |
| 754 | * out from there. |
| 755 | */ |
| 756 | w->next = w4; |
| 757 | w->private_data = kwl->para->private_data; |
| 758 | assert(w->private_data); |
| 759 | } |
| 760 | } |
| 761 | } |
| 762 | } |
| 763 | |
| 764 | return ret; |
| 765 | } |
| 766 | |
| 767 | static int info_rdaddwc(info_data *id, word *words, word *end, int xrefs, |
| 768 | infoconfig *cfg) { |
| 769 | int ret = 0; |
| 770 | |
| 771 | for (; words && words != end; words = words->next) switch (words->type) { |
| 772 | case word_HyperLink: |
| 773 | case word_HyperEnd: |
| 774 | case word_XrefEnd: |
| 775 | case word_IndexRef: |
| 776 | break; |
| 777 | |
| 778 | case word_Normal: |
| 779 | case word_Emph: |
| 780 | case word_Code: |
| 781 | case word_WeakCode: |
| 782 | case word_WhiteSpace: |
| 783 | case word_EmphSpace: |
| 784 | case word_CodeSpace: |
| 785 | case word_WkCodeSpace: |
| 786 | case word_Quote: |
| 787 | case word_EmphQuote: |
| 788 | case word_CodeQuote: |
| 789 | case word_WkCodeQuote: |
| 790 | assert(words->type != word_CodeQuote && |
| 791 | words->type != word_WkCodeQuote); |
| 792 | if (towordstyle(words->type) == word_Emph && |
| 793 | (attraux(words->aux) == attr_First || |
| 794 | attraux(words->aux) == attr_Only)) |
| 795 | ret += info_rdadds(id, cfg->startemph); |
| 796 | else if (towordstyle(words->type) == word_Code && |
| 797 | (attraux(words->aux) == attr_First || |
| 798 | attraux(words->aux) == attr_Only)) |
| 799 | ret += info_rdadds(id, cfg->lquote); |
| 800 | if (removeattr(words->type) == word_Normal) { |
| 801 | if (cvt_ok(id->charset, words->text) || !words->alt) |
| 802 | ret += info_rdadds(id, words->text); |
| 803 | else |
| 804 | ret += info_rdaddwc(id, words->alt, NULL, FALSE, cfg); |
| 805 | } else if (removeattr(words->type) == word_WhiteSpace) { |
| 806 | ret += info_rdadd(id, L' '); |
| 807 | } else if (removeattr(words->type) == word_Quote) { |
| 808 | ret += info_rdadds(id, quoteaux(words->aux) == quote_Open ? |
| 809 | cfg->lquote : cfg->rquote); |
| 810 | } |
| 811 | if (towordstyle(words->type) == word_Emph && |
| 812 | (attraux(words->aux) == attr_Last || |
| 813 | attraux(words->aux) == attr_Only)) |
| 814 | ret += info_rdadds(id, cfg->endemph); |
| 815 | else if (towordstyle(words->type) == word_Code && |
| 816 | (attraux(words->aux) == attr_Last || |
| 817 | attraux(words->aux) == attr_Only)) |
| 818 | ret += info_rdadds(id, cfg->rquote); |
| 819 | break; |
| 820 | |
| 821 | case word_UpperXref: |
| 822 | case word_LowerXref: |
| 823 | if (xrefs && words->private_data) { |
| 824 | /* |
| 825 | * This bit is structural and so must be done in char |
| 826 | * rather than wchar_t. |
| 827 | */ |
| 828 | ret += info_rdaddsc(id, "*Note "); |
| 829 | ret += info_rdaddsc(id, ((node *)words->private_data)->name); |
| 830 | ret += info_rdaddsc(id, "::"); |
| 831 | } |
| 832 | break; |
| 833 | } |
| 834 | |
| 835 | return ret; |
| 836 | } |
| 837 | |
| 838 | static int info_width_internal(word *words, int xrefs, infoconfig *cfg); |
| 839 | |
| 840 | static int info_width_internal_list(word *words, int xrefs, infoconfig *cfg) { |
| 841 | int w = 0; |
| 842 | while (words) { |
| 843 | w += info_width_internal(words, xrefs, cfg); |
| 844 | words = words->next; |
| 845 | } |
| 846 | return w; |
| 847 | } |
| 848 | |
| 849 | static int info_width_internal(word *words, int xrefs, infoconfig *cfg) { |
| 850 | int wid; |
| 851 | int attr; |
| 852 | |
| 853 | switch (words->type) { |
| 854 | case word_HyperLink: |
| 855 | case word_HyperEnd: |
| 856 | case word_XrefEnd: |
| 857 | case word_IndexRef: |
| 858 | return 0; |
| 859 | |
| 860 | case word_UpperXref: |
| 861 | case word_LowerXref: |
| 862 | if (xrefs && words->private_data) { |
| 863 | /* "*Note " plus "::" comes to 8 characters */ |
| 864 | return 8 + strwid(((node *)words->private_data)->name, |
| 865 | cfg->charset); |
| 866 | } else |
| 867 | return 0; |
| 868 | } |
| 869 | |
| 870 | assert(words->type < word_internal_endattrs); |
| 871 | |
| 872 | wid = 0; |
| 873 | attr = towordstyle(words->type); |
| 874 | |
| 875 | if (attr == word_Emph || attr == word_Code) { |
| 876 | if (attraux(words->aux) == attr_Only || |
| 877 | attraux(words->aux) == attr_First) |
| 878 | wid += ustrwid(attr == word_Emph ? cfg->startemph : cfg->lquote, |
| 879 | cfg->charset); |
| 880 | } |
| 881 | if (attr == word_Emph || attr == word_Code) { |
| 882 | if (attraux(words->aux) == attr_Only || |
| 883 | attraux(words->aux) == attr_Last) |
| 884 | wid += ustrwid(attr == word_Emph ? cfg->startemph : cfg->lquote, |
| 885 | cfg->charset); |
| 886 | } |
| 887 | |
| 888 | switch (words->type) { |
| 889 | case word_Normal: |
| 890 | case word_Emph: |
| 891 | case word_Code: |
| 892 | case word_WeakCode: |
| 893 | if (cvt_ok(cfg->charset, words->text) || !words->alt) |
| 894 | wid += ustrwid(words->text, cfg->charset); |
| 895 | else |
| 896 | wid += info_width_internal_list(words->alt, xrefs, cfg); |
| 897 | return wid; |
| 898 | |
| 899 | case word_WhiteSpace: |
| 900 | case word_EmphSpace: |
| 901 | case word_CodeSpace: |
| 902 | case word_WkCodeSpace: |
| 903 | case word_Quote: |
| 904 | case word_EmphQuote: |
| 905 | case word_CodeQuote: |
| 906 | case word_WkCodeQuote: |
| 907 | assert(words->type != word_CodeQuote && |
| 908 | words->type != word_WkCodeQuote); |
| 909 | if (removeattr(words->type) == word_Quote) { |
| 910 | if (quoteaux(words->aux) == quote_Open) |
| 911 | wid += ustrwid(cfg->lquote, cfg->charset); |
| 912 | else |
| 913 | wid += ustrwid(cfg->rquote, cfg->charset); |
| 914 | } else |
| 915 | wid++; /* space */ |
| 916 | } |
| 917 | return wid; |
| 918 | } |
| 919 | |
| 920 | static int info_width_noxrefs(void *ctx, word *words) |
| 921 | { |
| 922 | return info_width_internal(words, FALSE, (infoconfig *)ctx); |
| 923 | } |
| 924 | static int info_width_xrefs(void *ctx, word *words) |
| 925 | { |
| 926 | return info_width_internal(words, TRUE, (infoconfig *)ctx); |
| 927 | } |
| 928 | |
| 929 | static void info_heading(info_data *text, word *tprefix, |
| 930 | word *words, int width, infoconfig *cfg) { |
| 931 | int length; |
| 932 | int firstlinewidth, wrapwidth; |
| 933 | wrappedline *wrapping, *p; |
| 934 | |
| 935 | length = 0; |
| 936 | if (tprefix) { |
| 937 | length += info_rdaddwc(text, tprefix, NULL, FALSE, cfg); |
| 938 | length += info_rdadds(text, cfg->sectsuffix); |
| 939 | } |
| 940 | |
| 941 | wrapwidth = width; |
| 942 | firstlinewidth = width - length; |
| 943 | |
| 944 | wrapping = wrap_para(words, firstlinewidth, wrapwidth, |
| 945 | info_width_noxrefs, cfg, 0); |
| 946 | for (p = wrapping; p; p = p->next) { |
| 947 | length += info_rdaddwc(text, p->begin, p->end, FALSE, cfg); |
| 948 | info_rdadd(text, L'\n'); |
| 949 | while (length > 0) { |
| 950 | info_rdadds(text, cfg->underline); |
| 951 | length -= ustrwid(cfg->underline, cfg->charset); |
| 952 | } |
| 953 | info_rdadd(text, L'\n'); |
| 954 | length = 0; |
| 955 | } |
| 956 | wrap_free(wrapping); |
| 957 | info_rdadd(text, L'\n'); |
| 958 | } |
| 959 | |
| 960 | static void info_rule(info_data *text, int indent, int width, infoconfig *cfg) |
| 961 | { |
| 962 | while (indent--) info_rdadd(text, L' '); |
| 963 | while (width > 0) { |
| 964 | info_rdadds(text, cfg->rule); |
| 965 | width -= ustrwid(cfg->rule, cfg->charset); |
| 966 | } |
| 967 | info_rdadd(text, L'\n'); |
| 968 | info_rdadd(text, L'\n'); |
| 969 | } |
| 970 | |
| 971 | static void info_para(info_data *text, word *prefix, wchar_t *prefixextra, |
| 972 | word *input, keywordlist *keywords, int indent, |
| 973 | int extraindent, int width, infoconfig *cfg) { |
| 974 | wrappedline *wrapping, *p; |
| 975 | word *words; |
| 976 | int e; |
| 977 | int i; |
| 978 | int firstlinewidth = width; |
| 979 | |
| 980 | words = info_transform_wordlist(input, keywords); |
| 981 | |
| 982 | if (prefix) { |
| 983 | for (i = 0; i < indent; i++) |
| 984 | info_rdadd(text, L' '); |
| 985 | e = info_rdaddwc(text, prefix, NULL, FALSE, cfg); |
| 986 | if (prefixextra) |
| 987 | e += info_rdadds(text, prefixextra); |
| 988 | /* If the prefix is too long, shorten the first line to fit. */ |
| 989 | e = extraindent - e; |
| 990 | if (e < 0) { |
| 991 | firstlinewidth += e; /* this decreases it, since e < 0 */ |
| 992 | if (firstlinewidth < 0) { |
| 993 | e = indent + extraindent; |
| 994 | firstlinewidth = width; |
| 995 | info_rdadd(text, L'\n'); |
| 996 | } else |
| 997 | e = 0; |
| 998 | } |
| 999 | } else |
| 1000 | e = indent + extraindent; |
| 1001 | |
| 1002 | wrapping = wrap_para(words, firstlinewidth, width, info_width_xrefs, |
| 1003 | cfg, 0); |
| 1004 | for (p = wrapping; p; p = p->next) { |
| 1005 | for (i = 0; i < e; i++) |
| 1006 | info_rdadd(text, L' '); |
| 1007 | info_rdaddwc(text, p->begin, p->end, TRUE, cfg); |
| 1008 | info_rdadd(text, L'\n'); |
| 1009 | e = indent + extraindent; |
| 1010 | } |
| 1011 | wrap_free(wrapping); |
| 1012 | info_rdadd(text, L'\n'); |
| 1013 | |
| 1014 | free_word_list(words); |
| 1015 | } |
| 1016 | |
| 1017 | static void info_codepara(info_data *text, word *words, |
| 1018 | int indent, int width) { |
| 1019 | int i; |
| 1020 | |
| 1021 | for (; words; words = words->next) if (words->type == word_WeakCode) { |
| 1022 | for (i = 0; i < indent; i++) |
| 1023 | info_rdadd(text, L' '); |
| 1024 | if (info_rdadds(text, words->text) > width) { |
| 1025 | /* FIXME: warn */ |
| 1026 | } |
| 1027 | info_rdadd(text, L'\n'); |
| 1028 | } |
| 1029 | |
| 1030 | info_rdadd(text, L'\n'); |
| 1031 | } |
| 1032 | |
| 1033 | static void info_versionid(info_data *text, word *words, infoconfig *cfg) { |
| 1034 | info_rdadd(text, L'['); |
| 1035 | info_rdaddwc(text, words, NULL, FALSE, cfg); |
| 1036 | info_rdadds(text, L"]\n"); |
| 1037 | } |
| 1038 | |
| 1039 | static node *info_node_new(char *name, int charset) |
| 1040 | { |
| 1041 | node *n; |
| 1042 | |
| 1043 | n = snew(node); |
| 1044 | n->text = empty_info_data; |
| 1045 | n->text.charset = charset; |
| 1046 | n->up = n->next = n->prev = n->lastchild = n->listnext = NULL; |
| 1047 | n->name = dupstr(name); |
| 1048 | n->started_menu = FALSE; |
| 1049 | |
| 1050 | return n; |
| 1051 | } |
| 1052 | |
| 1053 | static char *info_node_name_core(info_data *id, filepos *fpos) |
| 1054 | { |
| 1055 | char *p, *q; |
| 1056 | |
| 1057 | /* |
| 1058 | * We cannot have commas, colons or parentheses in a node name. |
| 1059 | * Remove any that we find, with a warning. |
| 1060 | */ |
| 1061 | p = q = id->output.text; |
| 1062 | while (*p) { |
| 1063 | if (*p == ':' || *p == ',' || *p == '(' || *p == ')') { |
| 1064 | error(err_infonodechar, fpos, *p); |
| 1065 | } else { |
| 1066 | *q++ = *p; |
| 1067 | } |
| 1068 | p++; |
| 1069 | } |
| 1070 | *q = '\0'; |
| 1071 | |
| 1072 | return id->output.text; |
| 1073 | } |
| 1074 | |
| 1075 | static char *info_node_name_for_para(paragraph *par, infoconfig *cfg) |
| 1076 | { |
| 1077 | info_data id = EMPTY_INFO_DATA; |
| 1078 | |
| 1079 | id.charset = cfg->charset; |
| 1080 | info_rdaddwc(&id, par->kwtext ? par->kwtext : par->words, |
| 1081 | NULL, FALSE, cfg); |
| 1082 | info_rdaddsc(&id, NULL); |
| 1083 | |
| 1084 | return info_node_name_core(&id, &par->fpos); |
| 1085 | } |
| 1086 | |
| 1087 | static char *info_node_name_for_text(wchar_t *text, infoconfig *cfg) |
| 1088 | { |
| 1089 | info_data id = EMPTY_INFO_DATA; |
| 1090 | |
| 1091 | id.charset = cfg->charset; |
| 1092 | info_rdadds(&id, text); |
| 1093 | info_rdaddsc(&id, NULL); |
| 1094 | |
| 1095 | return info_node_name_core(&id, NULL); |
| 1096 | } |
| 1097 | |
| 1098 | static void info_menu_item(info_data *text, node *n, paragraph *p, |
| 1099 | infoconfig *cfg) |
| 1100 | { |
| 1101 | /* |
| 1102 | * FIXME: Depending on how we're doing node names in this info |
| 1103 | * file, we might want to do |
| 1104 | * |
| 1105 | * * Node name:: Chapter title |
| 1106 | * |
| 1107 | * _or_ |
| 1108 | * |
| 1109 | * * Chapter number: Node name. |
| 1110 | * |
| 1111 | * This function mostly works in char rather than wchar_t, |
| 1112 | * because a menu item is a structural component. |
| 1113 | */ |
| 1114 | info_rdaddsc(text, "* "); |
| 1115 | info_rdaddsc(text, n->name); |
| 1116 | info_rdaddsc(text, "::"); |
| 1117 | if (p) { |
| 1118 | info_rdaddc(text, ' '); |
| 1119 | info_rdaddwc(text, p->words, NULL, FALSE, cfg); |
| 1120 | } |
| 1121 | info_rdaddc(text, '\n'); |
| 1122 | } |
| 1123 | |
| 1124 | /* |
| 1125 | * These functions implement my wrapper on the rdadd* calls which |
| 1126 | * allows me to switch arbitrarily between literal octet-string |
| 1127 | * text and charset-translated Unicode. (Because no matter what |
| 1128 | * character set I write the actual text in, I expect info readers |
| 1129 | * to treat node names and file names literally and to expect |
| 1130 | * keywords like `*Note' in their canonical form, so I have to take |
| 1131 | * steps to ensure that those structural elements of the file |
| 1132 | * aren't messed with.) |
| 1133 | */ |
| 1134 | static int info_rdadds(info_data *d, wchar_t const *wcs) |
| 1135 | { |
| 1136 | if (!d->wcmode) { |
| 1137 | d->state = charset_init_state; |
| 1138 | d->wcmode = TRUE; |
| 1139 | } |
| 1140 | |
| 1141 | if (wcs) { |
| 1142 | char buf[256]; |
| 1143 | int len, width, ret; |
| 1144 | |
| 1145 | width = ustrwid(wcs, d->charset); |
| 1146 | |
| 1147 | len = ustrlen(wcs); |
| 1148 | while (len > 0) { |
| 1149 | int prevlen = len; |
| 1150 | |
| 1151 | ret = charset_from_unicode(&wcs, &len, buf, lenof(buf), |
| 1152 | d->charset, &d->state, NULL); |
| 1153 | |
| 1154 | assert(len < prevlen); |
| 1155 | |
| 1156 | if (ret > 0) { |
| 1157 | buf[ret] = '\0'; |
| 1158 | rdaddsc(&d->output, buf); |
| 1159 | } |
| 1160 | } |
| 1161 | |
| 1162 | return width; |
| 1163 | } else |
| 1164 | return 0; |
| 1165 | } |
| 1166 | |
| 1167 | static int info_rdaddsc(info_data *d, char const *cs) |
| 1168 | { |
| 1169 | if (d->wcmode) { |
| 1170 | char buf[256]; |
| 1171 | int ret; |
| 1172 | |
| 1173 | ret = charset_from_unicode(NULL, 0, buf, lenof(buf), |
| 1174 | d->charset, &d->state, NULL); |
| 1175 | if (ret > 0) { |
| 1176 | buf[ret] = '\0'; |
| 1177 | rdaddsc(&d->output, buf); |
| 1178 | } |
| 1179 | |
| 1180 | d->wcmode = FALSE; |
| 1181 | } |
| 1182 | |
| 1183 | if (cs) { |
| 1184 | rdaddsc(&d->output, cs); |
| 1185 | return strwid(cs, d->charset); |
| 1186 | } else |
| 1187 | return 0; |
| 1188 | } |
| 1189 | |
| 1190 | static int info_rdadd(info_data *d, wchar_t wc) |
| 1191 | { |
| 1192 | wchar_t wcs[2]; |
| 1193 | wcs[0] = wc; |
| 1194 | wcs[1] = L'\0'; |
| 1195 | return info_rdadds(d, wcs); |
| 1196 | } |
| 1197 | |
| 1198 | static int info_rdaddc(info_data *d, char c) |
| 1199 | { |
| 1200 | char cs[2]; |
| 1201 | cs[0] = c; |
| 1202 | cs[1] = '\0'; |
| 1203 | return info_rdaddsc(d, cs); |
| 1204 | } |