[BACK]Return to mandoc.c CVS log [TXT][DIR] Up to [cvsweb.bsd.lv] / mandoc

Annotation of mandoc/mandoc.c, Revision 1.101

1.101   ! schwarze    1: /*     $Id: mandoc.c,v 1.100 2017/06/02 19:21:23 schwarze Exp $ */
1.1       kristaps    2: /*
1.90      schwarze    3:  * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.100     schwarze    4:  * Copyright (c) 2011-2015, 2017 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    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:  *
1.36      schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.1       kristaps   11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.36      schwarze   12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.1       kristaps   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:  */
1.9       kristaps   18: #include "config.h"
1.7       kristaps   19:
1.2       kristaps   20: #include <sys/types.h>
                     21:
1.1       kristaps   22: #include <assert.h>
                     23: #include <ctype.h>
1.50      kristaps   24: #include <errno.h>
                     25: #include <limits.h>
1.1       kristaps   26: #include <stdlib.h>
1.4       kristaps   27: #include <stdio.h>
                     28: #include <string.h>
1.7       kristaps   29: #include <time.h>
1.1       kristaps   30:
1.101   ! schwarze   31: #include "mandoc_aux.h"
1.18      kristaps   32: #include "mandoc.h"
1.101   ! schwarze   33: #include "roff.h"
1.1       kristaps   34: #include "libmandoc.h"
1.37      schwarze   35:
1.18      kristaps   36: static int      a2time(time_t *, const char *, const char *);
1.37      schwarze   37: static char    *time2a(time_t);
1.7       kristaps   38:
1.45      kristaps   39:
                     40: enum mandoc_esc
1.74      schwarze   41: mandoc_escape(const char **end, const char **start, int *sz)
1.1       kristaps   42: {
1.65      schwarze   43:        const char      *local_start;
                     44:        int              local_sz;
                     45:        char             term;
1.79      schwarze   46:        enum mandoc_esc  gly;
1.45      kristaps   47:
1.65      schwarze   48:        /*
                     49:         * When the caller doesn't provide return storage,
                     50:         * use local storage.
                     51:         */
                     52:
                     53:        if (NULL == start)
                     54:                start = &local_start;
                     55:        if (NULL == sz)
                     56:                sz = &local_sz;
                     57:
                     58:        /*
                     59:         * Beyond the backslash, at least one input character
                     60:         * is part of the escape sequence.  With one exception
                     61:         * (see below), that character won't be returned.
                     62:         */
                     63:
1.45      kristaps   64:        gly = ESCAPE_ERROR;
1.65      schwarze   65:        *start = ++*end;
                     66:        *sz = 0;
1.64      schwarze   67:        term = '\0';
1.18      kristaps   68:
1.65      schwarze   69:        switch ((*start)[-1]) {
1.45      kristaps   70:        /*
                     71:         * First the glyphs.  There are several different forms of
                     72:         * these, but each eventually returns a substring of the glyph
                     73:         * name.
                     74:         */
1.79      schwarze   75:        case '(':
1.45      kristaps   76:                gly = ESCAPE_SPECIAL;
1.65      schwarze   77:                *sz = 2;
1.45      kristaps   78:                break;
1.79      schwarze   79:        case '[':
1.45      kristaps   80:                gly = ESCAPE_SPECIAL;
                     81:                term = ']';
                     82:                break;
1.79      schwarze   83:        case 'C':
1.65      schwarze   84:                if ('\'' != **start)
1.94      schwarze   85:                        return ESCAPE_ERROR;
1.65      schwarze   86:                *start = ++*end;
1.87      schwarze   87:                gly = ESCAPE_SPECIAL;
1.45      kristaps   88:                term = '\'';
                     89:                break;
1.72      schwarze   90:
                     91:        /*
                     92:         * Escapes taking no arguments at all.
                     93:         */
1.79      schwarze   94:        case 'd':
                     95:        case 'u':
1.93      schwarze   96:        case ',':
                     97:        case '/':
1.94      schwarze   98:                return ESCAPE_IGNORE;
1.63      schwarze   99:
                    100:        /*
                    101:         * The \z escape is supposed to output the following
1.79      schwarze  102:         * character without advancing the cursor position.
1.63      schwarze  103:         * Since we are mostly dealing with terminal mode,
                    104:         * let us just skip the next character.
                    105:         */
1.79      schwarze  106:        case 'z':
1.94      schwarze  107:                return ESCAPE_SKIPCHAR;
1.1       kristaps  108:
1.45      kristaps  109:        /*
                    110:         * Handle all triggers matching \X(xy, \Xx, and \X[xxxx], where
                    111:         * 'X' is the trigger.  These have opaque sub-strings.
                    112:         */
1.79      schwarze  113:        case 'F':
                    114:        case 'g':
                    115:        case 'k':
                    116:        case 'M':
                    117:        case 'm':
                    118:        case 'n':
                    119:        case 'V':
                    120:        case 'Y':
1.60      schwarze  121:                gly = ESCAPE_IGNORE;
1.24      kristaps  122:                /* FALLTHROUGH */
1.79      schwarze  123:        case 'f':
1.45      kristaps  124:                if (ESCAPE_ERROR == gly)
                    125:                        gly = ESCAPE_FONT;
1.65      schwarze  126:                switch (**start) {
1.79      schwarze  127:                case '(':
1.65      schwarze  128:                        *start = ++*end;
                    129:                        *sz = 2;
1.45      kristaps  130:                        break;
1.79      schwarze  131:                case '[':
1.65      schwarze  132:                        *start = ++*end;
1.45      kristaps  133:                        term = ']';
                    134:                        break;
                    135:                default:
1.65      schwarze  136:                        *sz = 1;
1.45      kristaps  137:                        break;
                    138:                }
                    139:                break;
                    140:
                    141:        /*
                    142:         * These escapes are of the form \X'Y', where 'X' is the trigger
                    143:         * and 'Y' is any string.  These have opaque sub-strings.
1.78      schwarze  144:         * The \B and \w escapes are handled in roff.c, roff_res().
1.45      kristaps  145:         */
1.79      schwarze  146:        case 'A':
                    147:        case 'b':
                    148:        case 'D':
                    149:        case 'R':
                    150:        case 'X':
                    151:        case 'Z':
1.91      schwarze  152:                gly = ESCAPE_IGNORE;
                    153:                /* FALLTHROUGH */
                    154:        case 'o':
                    155:                if (**start == '\0')
1.94      schwarze  156:                        return ESCAPE_ERROR;
1.91      schwarze  157:                if (gly == ESCAPE_ERROR)
                    158:                        gly = ESCAPE_OVERSTRIKE;
1.77      schwarze  159:                term = **start;
1.65      schwarze  160:                *start = ++*end;
1.24      kristaps  161:                break;
1.45      kristaps  162:
                    163:        /*
                    164:         * These escapes are of the form \X'N', where 'X' is the trigger
                    165:         * and 'N' resolves to a numerical expression.
                    166:         */
1.79      schwarze  167:        case 'h':
                    168:        case 'H':
                    169:        case 'L':
                    170:        case 'l':
                    171:        case 'S':
                    172:        case 'v':
                    173:        case 'x':
1.82      schwarze  174:                if (strchr(" %&()*+-./0123456789:<=>", **start)) {
1.86      kristaps  175:                        if ('\0' != **start)
                    176:                                ++*end;
1.94      schwarze  177:                        return ESCAPE_ERROR;
1.82      schwarze  178:                }
1.100     schwarze  179:                switch ((*start)[-1]) {
                    180:                case 'h':
                    181:                        gly = ESCAPE_HORIZ;
                    182:                        break;
                    183:                case 'l':
                    184:                        gly = ESCAPE_HLINE;
                    185:                        break;
                    186:                default:
                    187:                        gly = ESCAPE_IGNORE;
                    188:                        break;
                    189:                }
1.77      schwarze  190:                term = **start;
1.65      schwarze  191:                *start = ++*end;
1.45      kristaps  192:                break;
1.60      schwarze  193:
                    194:        /*
                    195:         * Special handling for the numbered character escape.
                    196:         * XXX Do any other escapes need similar handling?
                    197:         */
1.79      schwarze  198:        case 'N':
1.65      schwarze  199:                if ('\0' == **start)
1.94      schwarze  200:                        return ESCAPE_ERROR;
1.65      schwarze  201:                (*end)++;
                    202:                if (isdigit((unsigned char)**start)) {
                    203:                        *sz = 1;
1.94      schwarze  204:                        return ESCAPE_IGNORE;
1.65      schwarze  205:                }
                    206:                (*start)++;
1.60      schwarze  207:                while (isdigit((unsigned char)**end))
                    208:                        (*end)++;
1.65      schwarze  209:                *sz = *end - *start;
1.60      schwarze  210:                if ('\0' != **end)
                    211:                        (*end)++;
1.94      schwarze  212:                return ESCAPE_NUMBERED;
1.45      kristaps  213:
1.79      schwarze  214:        /*
1.45      kristaps  215:         * Sizes get a special category of their own.
                    216:         */
1.79      schwarze  217:        case 's':
1.45      kristaps  218:                gly = ESCAPE_IGNORE;
1.28      kristaps  219:
1.45      kristaps  220:                /* See +/- counts as a sign. */
1.65      schwarze  221:                if ('+' == **end || '-' == **end || ASCII_HYPH == **end)
1.90      schwarze  222:                        *start = ++*end;
1.8       kristaps  223:
1.65      schwarze  224:                switch (**end) {
1.79      schwarze  225:                case '(':
1.65      schwarze  226:                        *start = ++*end;
                    227:                        *sz = 2;
1.22      kristaps  228:                        break;
1.79      schwarze  229:                case '[':
1.65      schwarze  230:                        *start = ++*end;
1.64      schwarze  231:                        term = ']';
1.22      kristaps  232:                        break;
1.79      schwarze  233:                case '\'':
1.65      schwarze  234:                        *start = ++*end;
1.64      schwarze  235:                        term = '\'';
1.92      schwarze  236:                        break;
                    237:                case '3':
                    238:                case '2':
                    239:                case '1':
                    240:                        *sz = (*end)[-1] == 's' &&
                    241:                            isdigit((unsigned char)(*end)[1]) ? 2 : 1;
1.22      kristaps  242:                        break;
                    243:                default:
1.65      schwarze  244:                        *sz = 1;
1.22      kristaps  245:                        break;
1.8       kristaps  246:                }
                    247:
1.45      kristaps  248:                break;
1.33      kristaps  249:
1.45      kristaps  250:        /*
                    251:         * Anything else is assumed to be a glyph.
1.65      schwarze  252:         * In this case, pass back the character after the backslash.
1.45      kristaps  253:         */
                    254:        default:
                    255:                gly = ESCAPE_SPECIAL;
1.65      schwarze  256:                *start = --*end;
                    257:                *sz = 1;
1.22      kristaps  258:                break;
1.45      kristaps  259:        }
                    260:
                    261:        assert(ESCAPE_ERROR != gly);
                    262:
                    263:        /*
1.64      schwarze  264:         * Read up to the terminating character,
                    265:         * paying attention to nested escapes.
1.45      kristaps  266:         */
                    267:
                    268:        if ('\0' != term) {
1.64      schwarze  269:                while (**end != term) {
                    270:                        switch (**end) {
1.79      schwarze  271:                        case '\0':
1.94      schwarze  272:                                return ESCAPE_ERROR;
1.79      schwarze  273:                        case '\\':
1.64      schwarze  274:                                (*end)++;
                    275:                                if (ESCAPE_ERROR ==
                    276:                                    mandoc_escape(end, NULL, NULL))
1.94      schwarze  277:                                        return ESCAPE_ERROR;
1.64      schwarze  278:                                break;
                    279:                        default:
                    280:                                (*end)++;
                    281:                                break;
                    282:                        }
                    283:                }
1.65      schwarze  284:                *sz = (*end)++ - *start;
1.64      schwarze  285:        } else {
1.65      schwarze  286:                assert(*sz > 0);
                    287:                if ((size_t)*sz > strlen(*start))
1.94      schwarze  288:                        return ESCAPE_ERROR;
1.65      schwarze  289:                *end += *sz;
1.45      kristaps  290:        }
                    291:
                    292:        /* Run post-processors. */
                    293:
                    294:        switch (gly) {
1.79      schwarze  295:        case ESCAPE_FONT:
1.68      schwarze  296:                if (2 == *sz) {
                    297:                        if ('C' == **start) {
                    298:                                /*
                    299:                                 * Treat constant-width font modes
                    300:                                 * just like regular font modes.
                    301:                                 */
                    302:                                (*start)++;
                    303:                                (*sz)--;
                    304:                        } else {
                    305:                                if ('B' == (*start)[0] && 'I' == (*start)[1])
                    306:                                        gly = ESCAPE_FONTBI;
                    307:                                break;
                    308:                        }
1.65      schwarze  309:                } else if (1 != *sz)
1.45      kristaps  310:                        break;
1.61      kristaps  311:
1.65      schwarze  312:                switch (**start) {
1.79      schwarze  313:                case '3':
                    314:                case 'B':
1.45      kristaps  315:                        gly = ESCAPE_FONTBOLD;
                    316:                        break;
1.79      schwarze  317:                case '2':
                    318:                case 'I':
1.45      kristaps  319:                        gly = ESCAPE_FONTITALIC;
1.22      kristaps  320:                        break;
1.79      schwarze  321:                case 'P':
1.45      kristaps  322:                        gly = ESCAPE_FONTPREV;
1.22      kristaps  323:                        break;
1.79      schwarze  324:                case '1':
                    325:                case 'R':
1.45      kristaps  326:                        gly = ESCAPE_FONTROMAN;
1.1       kristaps  327:                        break;
                    328:                }
1.46      kristaps  329:                break;
1.79      schwarze  330:        case ESCAPE_SPECIAL:
1.65      schwarze  331:                if (1 == *sz && 'c' == **start)
1.45      kristaps  332:                        gly = ESCAPE_NOSPACE;
1.87      schwarze  333:                /*
1.88      schwarze  334:                 * Unicode escapes are defined in groff as \[u0000]
1.87      schwarze  335:                 * to \[u10FFFF], where the contained value must be
                    336:                 * a valid Unicode codepoint.  Here, however, only
1.88      schwarze  337:                 * check the length and range.
1.87      schwarze  338:                 */
1.88      schwarze  339:                if (**start != 'u' || *sz < 5 || *sz > 7)
                    340:                        break;
                    341:                if (*sz == 7 && ((*start)[1] != '1' || (*start)[2] != '0'))
                    342:                        break;
                    343:                if (*sz == 6 && (*start)[1] == '0')
1.96      schwarze  344:                        break;
                    345:                if (*sz == 5 && (*start)[1] == 'D' &&
                    346:                    strchr("89ABCDEF", (*start)[2]) != NULL)
1.88      schwarze  347:                        break;
                    348:                if ((int)strspn(*start + 1, "0123456789ABCDEFabcdef")
1.87      schwarze  349:                    + 1 == *sz)
                    350:                        gly = ESCAPE_UNICODE;
1.22      kristaps  351:                break;
1.1       kristaps  352:        default:
1.22      kristaps  353:                break;
1.1       kristaps  354:        }
                    355:
1.94      schwarze  356:        return gly;
1.36      schwarze  357: }
                    358:
                    359: /*
                    360:  * Parse a quoted or unquoted roff-style request or macro argument.
                    361:  * Return a pointer to the parsed argument, which is either the original
                    362:  * pointer or advanced by one byte in case the argument is quoted.
1.71      schwarze  363:  * NUL-terminate the argument in place.
1.36      schwarze  364:  * Collapse pairs of quotes inside quoted arguments.
                    365:  * Advance the argument pointer to the next argument,
1.71      schwarze  366:  * or to the NUL byte terminating the argument line.
1.36      schwarze  367:  */
                    368: char *
1.48      kristaps  369: mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
1.36      schwarze  370: {
                    371:        char     *start, *cp;
                    372:        int       quoted, pairs, white;
                    373:
                    374:        /* Quoting can only start with a new word. */
                    375:        start = *cpp;
1.47      kristaps  376:        quoted = 0;
1.36      schwarze  377:        if ('"' == *start) {
                    378:                quoted = 1;
                    379:                start++;
1.79      schwarze  380:        }
1.36      schwarze  381:
                    382:        pairs = 0;
                    383:        white = 0;
                    384:        for (cp = start; '\0' != *cp; cp++) {
1.67      schwarze  385:
                    386:                /*
                    387:                 * Move the following text left
                    388:                 * after quoted quotes and after "\\" and "\t".
                    389:                 */
1.36      schwarze  390:                if (pairs)
                    391:                        cp[-pairs] = cp[0];
1.67      schwarze  392:
1.36      schwarze  393:                if ('\\' == cp[0]) {
1.67      schwarze  394:                        /*
                    395:                         * In copy mode, translate double to single
                    396:                         * backslashes and backslash-t to literal tabs.
                    397:                         */
                    398:                        switch (cp[1]) {
1.79      schwarze  399:                        case 't':
1.67      schwarze  400:                                cp[0] = '\t';
                    401:                                /* FALLTHROUGH */
1.79      schwarze  402:                        case '\\':
1.36      schwarze  403:                                pairs++;
                    404:                                cp++;
1.67      schwarze  405:                                break;
1.79      schwarze  406:                        case ' ':
1.36      schwarze  407:                                /* Skip escaped blanks. */
1.67      schwarze  408:                                if (0 == quoted)
                    409:                                        cp++;
                    410:                                break;
                    411:                        default:
                    412:                                break;
                    413:                        }
1.36      schwarze  414:                } else if (0 == quoted) {
                    415:                        if (' ' == cp[0]) {
                    416:                                /* Unescaped blanks end unquoted args. */
                    417:                                white = 1;
                    418:                                break;
                    419:                        }
                    420:                } else if ('"' == cp[0]) {
                    421:                        if ('"' == cp[1]) {
                    422:                                /* Quoted quotes collapse. */
                    423:                                pairs++;
                    424:                                cp++;
                    425:                        } else {
                    426:                                /* Unquoted quotes end quoted args. */
                    427:                                quoted = 2;
                    428:                                break;
                    429:                        }
                    430:                }
                    431:        }
                    432:
                    433:        /* Quoted argument without a closing quote. */
1.48      kristaps  434:        if (1 == quoted)
1.83      schwarze  435:                mandoc_msg(MANDOCERR_ARG_QUOTE, parse, ln, *pos, NULL);
1.36      schwarze  436:
1.71      schwarze  437:        /* NUL-terminate this argument and move to the next one. */
1.36      schwarze  438:        if (pairs)
                    439:                cp[-pairs] = '\0';
                    440:        if ('\0' != *cp) {
                    441:                *cp++ = '\0';
                    442:                while (' ' == *cp)
                    443:                        cp++;
                    444:        }
1.39      kristaps  445:        *pos += (int)(cp - start) + (quoted ? 1 : 0);
1.36      schwarze  446:        *cpp = cp;
                    447:
1.48      kristaps  448:        if ('\0' == *cp && (white || ' ' == cp[-1]))
1.83      schwarze  449:                mandoc_msg(MANDOCERR_SPACE_EOL, parse, ln, *pos, NULL);
1.36      schwarze  450:
1.94      schwarze  451:        return start;
1.4       kristaps  452: }
1.7       kristaps  453:
                    454: static int
                    455: a2time(time_t *t, const char *fmt, const char *p)
                    456: {
                    457:        struct tm        tm;
                    458:        char            *pp;
                    459:
                    460:        memset(&tm, 0, sizeof(struct tm));
                    461:
1.56      kristaps  462:        pp = NULL;
1.85      schwarze  463: #if HAVE_STRPTIME
1.7       kristaps  464:        pp = strptime(p, fmt, &tm);
1.56      kristaps  465: #endif
1.7       kristaps  466:        if (NULL != pp && '\0' == *pp) {
                    467:                *t = mktime(&tm);
1.94      schwarze  468:                return 1;
1.7       kristaps  469:        }
                    470:
1.94      schwarze  471:        return 0;
1.7       kristaps  472: }
                    473:
1.37      schwarze  474: static char *
                    475: time2a(time_t t)
                    476: {
1.56      kristaps  477:        struct tm       *tm;
1.38      schwarze  478:        char            *buf, *p;
                    479:        size_t           ssz;
1.37      schwarze  480:        int              isz;
                    481:
1.56      kristaps  482:        tm = localtime(&t);
1.89      schwarze  483:        if (tm == NULL)
1.94      schwarze  484:                return NULL;
1.37      schwarze  485:
1.38      schwarze  486:        /*
                    487:         * Reserve space:
                    488:         * up to 9 characters for the month (September) + blank
                    489:         * up to 2 characters for the day + comma + blank
                    490:         * 4 characters for the year and a terminating '\0'
                    491:         */
1.98      schwarze  492:
1.38      schwarze  493:        p = buf = mandoc_malloc(10 + 4 + 4 + 1);
                    494:
1.98      schwarze  495:        if ((ssz = strftime(p, 10 + 1, "%B ", tm)) == 0)
1.38      schwarze  496:                goto fail;
                    497:        p += (int)ssz;
1.37      schwarze  498:
1.98      schwarze  499:        /*
                    500:         * The output format is just "%d" here, not "%2d" or "%02d".
                    501:         * That's also the reason why we can't just format the
                    502:         * date as a whole with "%B %e, %Y" or "%B %d, %Y".
                    503:         * Besides, the present approach is less prone to buffer
                    504:         * overflows, in case anybody should ever introduce the bug
                    505:         * of looking at LC_TIME.
                    506:         */
                    507:
                    508:        if ((isz = snprintf(p, 4 + 1, "%d, ", tm->tm_mday)) == -1)
1.38      schwarze  509:                goto fail;
1.37      schwarze  510:        p += isz;
                    511:
1.98      schwarze  512:        if (strftime(p, 4 + 1, "%Y", tm) == 0)
1.38      schwarze  513:                goto fail;
1.94      schwarze  514:        return buf;
1.38      schwarze  515:
                    516: fail:
                    517:        free(buf);
1.94      schwarze  518:        return NULL;
1.37      schwarze  519: }
                    520:
                    521: char *
1.101   ! schwarze  522: mandoc_normdate(struct roff_man *man, char *in, int ln, int pos)
1.7       kristaps  523: {
                    524:        time_t           t;
                    525:
1.98      schwarze  526:        /* No date specified: use today's date. */
                    527:
                    528:        if (in == NULL || *in == '\0' || strcmp(in, "$" "Mdocdate$") == 0) {
1.101   ! schwarze  529:                mandoc_msg(MANDOCERR_DATE_MISSING, man->parse, ln, pos, NULL);
1.98      schwarze  530:                return time2a(time(NULL));
1.37      schwarze  531:        }
1.98      schwarze  532:
                    533:        /* Valid mdoc(7) date format. */
                    534:
                    535:        if (a2time(&t, "$" "Mdocdate: %b %d %Y $", in) ||
                    536:            a2time(&t, "%b %d, %Y", in))
                    537:                return time2a(t);
                    538:
1.101   ! schwarze  539:        /* In man(7), do not warn about the legacy format. */
1.98      schwarze  540:
1.101   ! schwarze  541:        if (a2time(&t, "%Y-%m-%d", in) == 0)
        !           542:                mandoc_msg(MANDOCERR_DATE_BAD, man->parse, ln, pos, in);
        !           543:        else if (man->macroset == MACROSET_MDOC)
        !           544:                mandoc_vmsg(MANDOCERR_DATE_LEGACY, man->parse,
        !           545:                    ln, pos, "Dd %s", in);
1.98      schwarze  546:
                    547:        /* Use any non-mdoc(7) date verbatim. */
                    548:
                    549:        return mandoc_strdup(in);
1.7       kristaps  550: }
                    551:
1.12      kristaps  552: int
1.75      schwarze  553: mandoc_eos(const char *p, size_t sz)
1.12      kristaps  554: {
1.75      schwarze  555:        const char      *q;
                    556:        int              enclosed, found;
1.12      kristaps  557:
1.13      kristaps  558:        if (0 == sz)
1.94      schwarze  559:                return 0;
1.12      kristaps  560:
1.14      kristaps  561:        /*
                    562:         * End-of-sentence recognition must include situations where
                    563:         * some symbols, such as `)', allow prior EOS punctuation to
1.49      kristaps  564:         * propagate outward.
1.14      kristaps  565:         */
                    566:
1.75      schwarze  567:        enclosed = found = 0;
1.25      kristaps  568:        for (q = p + (int)sz - 1; q >= p; q--) {
1.23      schwarze  569:                switch (*q) {
1.79      schwarze  570:                case '\"':
                    571:                case '\'':
                    572:                case ']':
                    573:                case ')':
1.23      schwarze  574:                        if (0 == found)
                    575:                                enclosed = 1;
1.14      kristaps  576:                        break;
1.79      schwarze  577:                case '.':
                    578:                case '!':
                    579:                case '?':
1.23      schwarze  580:                        found = 1;
                    581:                        break;
1.14      kristaps  582:                default:
1.94      schwarze  583:                        return found &&
                    584:                            (!enclosed || isalnum((unsigned char)*q));
1.14      kristaps  585:                }
1.12      kristaps  586:        }
                    587:
1.94      schwarze  588:        return found && !enclosed;
1.44      kristaps  589: }
1.50      kristaps  590:
                    591: /*
                    592:  * Convert a string to a long that may not be <0.
                    593:  * If the string is invalid, or is less than 0, return -1.
                    594:  */
                    595: int
1.54      kristaps  596: mandoc_strntoi(const char *p, size_t sz, int base)
1.50      kristaps  597: {
                    598:        char             buf[32];
                    599:        char            *ep;
                    600:        long             v;
                    601:
                    602:        if (sz > 31)
1.94      schwarze  603:                return -1;
1.50      kristaps  604:
                    605:        memcpy(buf, p, sz);
1.51      kristaps  606:        buf[(int)sz] = '\0';
1.50      kristaps  607:
                    608:        errno = 0;
                    609:        v = strtol(buf, &ep, base);
                    610:
                    611:        if (buf[0] == '\0' || *ep != '\0')
1.94      schwarze  612:                return -1;
1.50      kristaps  613:
1.54      kristaps  614:        if (v > INT_MAX)
                    615:                v = INT_MAX;
                    616:        if (v < INT_MIN)
                    617:                v = INT_MIN;
1.50      kristaps  618:
1.94      schwarze  619:        return (int)v;
1.50      kristaps  620: }

CVSweb