1 /* $Id: term_ascii.c,v 1.69 2023/11/13 19:13:01 schwarze Exp $ */ 2 /* 3 * Copyright (c) 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv> 4 * Copyright (c) 2014,2015,2017,2018,2020 Ingo Schwarze <schwarze@openbsd.org> 5 * 6 * Permission to use, copy, modify, and distribute this software for any 7 * purpose with or without fee is hereby granted, provided that the above 8 * copyright notice and this permission notice appear in all copies. 9 * 10 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES 11 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF 12 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR 13 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES 14 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN 15 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF 16 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. 17 */ 18 #include "config.h" 19 20 #include <sys/types.h> 21 22 #include <assert.h> 23 #if HAVE_WCHAR 24 #include <langinfo.h> 25 #include <locale.h> 26 #endif 27 #include <stdint.h> 28 #include <stdio.h> 29 #include <stdlib.h> 30 #include <string.h> 31 #include <unistd.h> 32 #if HAVE_WCHAR 33 #include <wchar.h> 34 #endif 35 36 #include "mandoc.h" 37 #include "mandoc_aux.h" 38 #include "out.h" 39 #include "term.h" 40 #include "manconf.h" 41 #include "main.h" 42 43 static struct termp *ascii_init(enum termenc, const struct manoutput *); 44 static int ascii_hspan(const struct termp *, 45 const struct roffsu *); 46 static size_t ascii_width(const struct termp *, int); 47 static void ascii_advance(struct termp *, size_t); 48 static void ascii_begin(struct termp *); 49 static void ascii_end(struct termp *); 50 static void ascii_endline(struct termp *); 51 static void ascii_letter(struct termp *, int); 52 static void ascii_setwidth(struct termp *, int, int); 53 54 #if HAVE_WCHAR 55 static void locale_advance(struct termp *, size_t); 56 static void locale_endline(struct termp *); 57 static void locale_letter(struct termp *, int); 58 static size_t locale_width(const struct termp *, int); 59 #endif 60 61 62 static struct termp * 63 ascii_init(enum termenc enc, const struct manoutput *outopts) 64 { 65 #if HAVE_WCHAR 66 char *v; 67 #endif 68 struct termp *p; 69 70 p = mandoc_calloc(1, sizeof(*p)); 71 p->tcol = p->tcols = mandoc_calloc(1, sizeof(*p->tcol)); 72 p->maxtcol = 1; 73 74 p->line = 1; 75 p->defindent = 5; 76 p->defrmargin = p->lastrmargin = 78; 77 p->fontq = mandoc_reallocarray(NULL, 78 (p->fontsz = 8), sizeof(*p->fontq)); 79 p->fontq[0] = p->fontl = TERMFONT_NONE; 80 81 p->begin = ascii_begin; 82 p->end = ascii_end; 83 p->hspan = ascii_hspan; 84 p->type = TERMTYPE_CHAR; 85 86 p->enc = TERMENC_ASCII; 87 p->advance = ascii_advance; 88 p->endline = ascii_endline; 89 p->letter = ascii_letter; 90 p->setwidth = ascii_setwidth; 91 p->width = ascii_width; 92 93 #if HAVE_WCHAR 94 if (enc != TERMENC_ASCII) { 95 96 /* 97 * Do not change any of this to LC_ALL. It might break 98 * the formatting by subtly changing the behaviour of 99 * various functions, for example strftime(3). As a 100 * worst case, it might even cause buffer overflows. 101 */ 102 103 v = enc == TERMENC_LOCALE ? 104 setlocale(LC_CTYPE, "") : 105 setlocale(LC_CTYPE, UTF8_LOCALE); 106 107 /* 108 * We only support UTF-8, 109 * so revert to ASCII for anything else. 110 */ 111 112 if (v != NULL && 113 strcmp(nl_langinfo(CODESET), "UTF-8") != 0) 114 v = setlocale(LC_CTYPE, "C"); 115 116 if (v != NULL && MB_CUR_MAX > 1) { 117 p->enc = TERMENC_UTF8; 118 p->advance = locale_advance; 119 p->endline = locale_endline; 120 p->letter = locale_letter; 121 p->width = locale_width; 122 } 123 } 124 #endif 125 126 if (outopts->mdoc) 127 p->mdocstyle = 1; 128 if (outopts->indent) 129 p->defindent = outopts->indent; 130 if (outopts->width) 131 p->defrmargin = outopts->width; 132 if (outopts->synopsisonly) 133 p->synopsisonly = 1; 134 135 assert(p->defindent < UINT16_MAX); 136 assert(p->defrmargin < UINT16_MAX); 137 return p; 138 } 139 140 void * 141 ascii_alloc(const struct manoutput *outopts) 142 { 143 144 return ascii_init(TERMENC_ASCII, outopts); 145 } 146 147 void * 148 utf8_alloc(const struct manoutput *outopts) 149 { 150 151 return ascii_init(TERMENC_UTF8, outopts); 152 } 153 154 void * 155 locale_alloc(const struct manoutput *outopts) 156 { 157 158 return ascii_init(TERMENC_LOCALE, outopts); 159 } 160 161 static void 162 ascii_setwidth(struct termp *p, int iop, int width) 163 { 164 165 width /= 24; 166 p->tcol->rmargin = p->defrmargin; 167 if (iop > 0) 168 p->defrmargin += width; 169 else if (iop == 0) 170 p->defrmargin = width ? (size_t)width : p->lastrmargin; 171 else if (p->defrmargin > (size_t)width) 172 p->defrmargin -= width; 173 else 174 p->defrmargin = 0; 175 if (p->defrmargin > 1000) 176 p->defrmargin = 1000; 177 p->lastrmargin = p->tcol->rmargin; 178 p->tcol->rmargin = p->maxrmargin = p->defrmargin; 179 } 180 181 void 182 terminal_sepline(void *arg) 183 { 184 struct termp *p; 185 size_t i; 186 187 p = (struct termp *)arg; 188 (*p->endline)(p); 189 for (i = 0; i < p->defrmargin; i++) 190 (*p->letter)(p, '-'); 191 (*p->endline)(p); 192 (*p->endline)(p); 193 } 194 195 static size_t 196 ascii_width(const struct termp *p, int c) 197 { 198 return c != ASCII_BREAK && c != ASCII_NBRZW && c != ASCII_TABREF; 199 } 200 201 void 202 ascii_free(void *arg) 203 { 204 205 term_free((struct termp *)arg); 206 } 207 208 static void 209 ascii_letter(struct termp *p, int c) 210 { 211 212 putchar(c); 213 } 214 215 static void 216 ascii_begin(struct termp *p) 217 { 218 219 (*p->headf)(p, p->argf); 220 } 221 222 static void 223 ascii_end(struct termp *p) 224 { 225 226 (*p->footf)(p, p->argf); 227 } 228 229 static void 230 ascii_endline(struct termp *p) 231 { 232 233 p->line++; 234 if ((int)p->tcol->offset > p->ti) 235 p->tcol->offset -= p->ti; 236 else 237 p->tcol->offset = 0; 238 p->ti = 0; 239 putchar('\n'); 240 } 241 242 static void 243 ascii_advance(struct termp *p, size_t len) 244 { 245 size_t i; 246 247 /* 248 * XXX We used to have "assert(len < UINT16_MAX)" here. 249 * that is not quite right because the input document 250 * can trigger that by merely providing large input. 251 * For now, simply truncate. 252 */ 253 if (len > 256) 254 len = 256; 255 for (i = 0; i < len; i++) 256 putchar(' '); 257 } 258 259 static int 260 ascii_hspan(const struct termp *p, const struct roffsu *su) 261 { 262 double r; 263 264 switch (su->unit) { 265 case SCALE_BU: 266 r = su->scale; 267 break; 268 case SCALE_CM: 269 r = su->scale * 240.0 / 2.54; 270 break; 271 case SCALE_FS: 272 r = su->scale * 65536.0; 273 break; 274 case SCALE_IN: 275 r = su->scale * 240.0; 276 break; 277 case SCALE_MM: 278 r = su->scale * 0.24; 279 break; 280 case SCALE_VS: 281 case SCALE_PC: 282 r = su->scale * 40.0; 283 break; 284 case SCALE_PT: 285 r = su->scale * 10.0 / 3.0; 286 break; 287 case SCALE_EN: 288 case SCALE_EM: 289 r = su->scale * 24.0; 290 break; 291 default: 292 abort(); 293 } 294 return r > 0.0 ? r + 0.01 : r - 0.01; 295 } 296 297 const char * 298 ascii_uc2str(int uc) 299 { 300 static const char nbrsp[2] = { ASCII_NBRSP, '\0' }; 301 static const char *tab[] = { 302 "<NUL>","<SOH>","<STX>","<ETX>","<EOT>","<ENQ>","<ACK>","<BEL>", 303 "<BS>", "\t", "<LF>", "<VT>", "<FF>", "<CR>", "<SO>", "<SI>", 304 "<DLE>","<DC1>","<DC2>","<DC3>","<DC4>","<NAK>","<SYN>","<ETB>", 305 "<CAN>","<EM>", "<SUB>","<ESC>","<FS>", "<GS>", "<RS>", "<US>", 306 " ", "!", "\"", "#", "$", "%", "&", "'", 307 "(", ")", "*", "+", ",", "-", ".", "/", 308 "0", "1", "2", "3", "4", "5", "6", "7", 309 "8", "9", ":", ";", "<", "=", ">", "?", 310 "@", "A", "B", "C", "D", "E", "F", "G", 311 "H", "I", "J", "K", "L", "M", "N", "O", 312 "P", "Q", "R", "S", "T", "U", "V", "W", 313 "X", "Y", "Z", "[", "\\", "]", "^", "_", 314 "`", "a", "b", "c", "d", "e", "f", "g", 315 "h", "i", "j", "k", "l", "m", "n", "o", 316 "p", "q", "r", "s", "t", "u", "v", "w", 317 "x", "y", "z", "{", "|", "}", "~", "<DEL>", 318 "<80>", "<81>", "<82>", "<83>", "<84>", "<85>", "<86>", "<87>", 319 "<88>", "<89>", "<8A>", "<8B>", "<8C>", "<8D>", "<8E>", "<8F>", 320 "<90>", "<91>", "<92>", "<93>", "<94>", "<95>", "<96>", "<97>", 321 "<98>", "<99>", "<9A>", "<9B>", "<9C>", "<9D>", "<9E>", "<9F>", 322 nbrsp, "!", "/\bc", "-\bL", "o\bx", "=\bY", "|", "<section>", 323 "\"", "(C)", "_\ba", "<<", "~", "", "(R)", "-", 324 "<degree>","+-","^2", "^3", "'","<micro>","<paragraph>",".", 325 ",", "^1", "_\bo", ">>", "1/4", "1/2", "3/4", "?", 326 "`\bA", "'\bA", "^\bA", "~\bA", "\"\bA","o\bA", "AE", ",\bC", 327 "`\bE", "'\bE", "^\bE", "\"\bE","`\bI", "'\bI", "^\bI", "\"\bI", 328 "Dh", "~\bN", "`\bO", "'\bO", "^\bO", "~\bO", "\"\bO","x", 329 "/\bO", "`\bU", "'\bU", "^\bU", "\"\bU","'\bY", "Th", "ss", 330 "`\ba", "'\ba", "^\ba", "~\ba", "\"\ba","o\ba", "ae", ",\bc", 331 "`\be", "'\be", "^\be", "\"\be","`\bi", "'\bi", "^\bi", "\"\bi", 332 "dh", "~\bn", "`\bo", "'\bo", "^\bo", "~\bo", "\"\bo","/", 333 "/\bo", "`\bu", "'\bu", "^\bu", "\"\bu","'\by", "th", "\"\by", 334 "A", "a", "A", "a", "A", "a", "'\bC", "'\bc", 335 "^\bC", "^\bc", "C", "c", "C", "c", "D", "d", 336 "/\bD", "/\bd", "E", "e", "E", "e", "E", "e", 337 "E", "e", "E", "e", "^\bG", "^\bg", "G", "g", 338 "G", "g", ",\bG", ",\bg", "^\bH", "^\bh", "/\bH", "/\bh", 339 "~\bI", "~\bi", "I", "i", "I", "i", "I", "i", 340 "I", "i", "IJ", "ij", "^\bJ", "^\bj", ",\bK", ",\bk", 341 "q", "'\bL", "'\bl", ",\bL", ",\bl", "L", "l", "L", 342 "l", "/\bL", "/\bl", "'\bN", "'\bn", ",\bN", ",\bn", "N", 343 "n", "'n", "Ng", "ng", "O", "o", "O", "o", 344 "O", "o", "OE", "oe", "'\bR", "'\br", ",\bR", ",\br", 345 "R", "r", "'\bS", "'\bs", "^\bS", "^\bs", ",\bS", ",\bs", 346 "S", "s", ",\bT", ",\bt", "T", "t", "/\bT", "/\bt", 347 "~\bU", "~\bu", "U", "u", "U", "u", "U", "u", 348 "U", "u", "U", "u", "^\bW", "^\bw", "^\bY", "^\by", 349 "\"\bY","'\bZ", "'\bz", "Z", "z", "Z", "z", "s", 350 "b", "B", "B", "b", "6", "6", "O", "C", 351 "c", "D", "D", "D", "d", "d", "3", "@", 352 "E", "F", ",\bf", "G", "G", "hv", "I", "/\bI", 353 "K", "k", "/\bl", "l", "W", "N", "n", "~\bO", 354 "O", "o", "OI", "oi", "P", "p", "YR", "2", 355 "2", "SH", "sh", "t", "T", "t", "T", "U", 356 "u", "Y", "V", "Y", "y", "/\bZ", "/\bz", "ZH", 357 "ZH", "zh", "zh", "/\b2", "5", "5", "ts", "w", 358 "|", "||", "|=", "!", "DZ", "Dz", "dz", "LJ", 359 "Lj", "lj", "NJ", "Nj", "nj", "A", "a", "I", 360 "i", "O", "o", "U", "u", "U", "u", "U", 361 "u", "U", "u", "U", "u", "@", "A", "a", 362 "A", "a", "AE", "ae", "/\bG", "/\bg", "G", "g", 363 "K", "k", "O", "o", "O", "o", "ZH", "zh", 364 "j", "DZ", "Dz", "dz", "'\bG", "'\bg", "HV", "W", 365 "`\bN", "`\bn", "A", "a", "'\bAE","'\bae","O", "o"}; 366 367 assert(uc >= 0); 368 if ((size_t)uc < sizeof(tab)/sizeof(tab[0])) 369 return tab[uc]; 370 return mchars_uc2str(uc); 371 } 372 373 #if HAVE_WCHAR 374 static size_t 375 locale_width(const struct termp *p, int c) 376 { 377 int rc; 378 379 if (c == ASCII_NBRSP) 380 c = ' '; 381 rc = wcwidth(c); 382 if (rc < 0) 383 rc = 0; 384 return rc; 385 } 386 387 static void 388 locale_advance(struct termp *p, size_t len) 389 { 390 size_t i; 391 392 /* 393 * XXX We used to have "assert(len < UINT16_MAX)" here. 394 * that is not quite right because the input document 395 * can trigger that by merely providing large input. 396 * For now, simply truncate. 397 */ 398 if (len > 256) 399 len = 256; 400 for (i = 0; i < len; i++) 401 putwchar(L' '); 402 } 403 404 static void 405 locale_endline(struct termp *p) 406 { 407 408 p->line++; 409 if ((int)p->tcol->offset > p->ti) 410 p->tcol->offset -= p->ti; 411 else 412 p->tcol->offset = 0; 413 p->ti = 0; 414 putwchar(L'\n'); 415 } 416 417 static void 418 locale_letter(struct termp *p, int c) 419 { 420 421 putwchar(c); 422 } 423 #endif 424