[BACK]Return to mdoc.c CVS log [TXT][DIR] Up to [local] / src / usr.bin / mandoc

Annotation of src/usr.bin/mandoc/mdoc.c, Revision 1.47

1.47    ! schwarze    1: /*     $Id: mdoc.c,v 1.46 2010/05/14 01:54:37 schwarze Exp $ */
1.1       kristaps    2: /*
1.3       schwarze    3:  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
1.1       kristaps    4:  *
                      5:  * Permission to use, copy, modify, and distribute this software for any
1.3       schwarze    6:  * purpose with or without fee is hereby granted, provided that the above
                      7:  * copyright notice and this permission notice appear in all copies.
1.1       kristaps    8:  *
1.3       schwarze    9:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
                     10:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
                     11:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
                     12:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     13:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     14:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     15:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
1.1       kristaps   16:  */
1.28      schwarze   17: #include <sys/types.h>
                     18:
1.1       kristaps   19: #include <assert.h>
                     20: #include <ctype.h>
                     21: #include <stdarg.h>
                     22: #include <stdio.h>
                     23: #include <stdlib.h>
                     24: #include <string.h>
1.39      schwarze   25: #include <time.h>
1.1       kristaps   26:
                     27: #include "libmdoc.h"
1.32      schwarze   28: #include "libmandoc.h"
1.1       kristaps   29:
1.14      schwarze   30: const  char *const __mdoc_merrnames[MERRMAX] = {
                     31:        "trailing whitespace", /* ETAILWS */
                     32:        "unexpected quoted parameter", /* EQUOTPARM */
                     33:        "unterminated quoted parameter", /* EQUOTTERM */
                     34:        "argument parameter suggested", /* EARGVAL */
                     35:        "macro disallowed in prologue", /* EBODYPROL */
                     36:        "macro disallowed in body", /* EPROLBODY */
                     37:        "text disallowed in prologue", /* ETEXTPROL */
                     38:        "blank line disallowed", /* ENOBLANK */
                     39:        "text parameter too long", /* ETOOLONG */
                     40:        "invalid escape sequence", /* EESCAPE */
                     41:        "invalid character", /* EPRINT */
                     42:        "document has no body", /* ENODAT */
                     43:        "document has no prologue", /* ENOPROLOGUE */
                     44:        "expected line arguments", /* ELINE */
                     45:        "invalid AT&T argument", /* EATT */
                     46:        "default name not yet set", /* ENAME */
                     47:        "missing list type", /* ELISTTYPE */
                     48:        "missing display type", /* EDISPTYPE */
                     49:        "too many display types", /* EMULTIDISP */
                     50:        "too many list types", /* EMULTILIST */
                     51:        "NAME section must be first", /* ESECNAME */
                     52:        "badly-formed NAME section", /* ENAMESECINC */
                     53:        "argument repeated", /* EARGREP */
                     54:        "expected boolean parameter", /* EBOOL */
                     55:        "inconsistent column syntax", /* ECOLMIS */
                     56:        "nested display invalid", /* ENESTDISP */
                     57:        "width argument missing", /* EMISSWIDTH */
                     58:        "invalid section for this manual section", /* EWRONGMSEC */
                     59:        "section out of conventional order", /* ESECOOO */
                     60:        "section repeated", /* ESECREP */
                     61:        "invalid standard argument", /* EBADSTAND */
                     62:        "multi-line arguments discouraged", /* ENOMULTILINE */
                     63:        "multi-line arguments suggested", /* EMULTILINE */
                     64:        "line arguments discouraged", /* ENOLINE */
                     65:        "prologue macro out of conventional order", /* EPROLOOO */
                     66:        "prologue macro repeated", /* EPROLREP */
                     67:        "invalid manual section", /* EBADMSEC */
                     68:        "invalid section", /* EBADSEC */
                     69:        "invalid font mode", /* EFONT */
                     70:        "invalid date syntax", /* EBADDATE */
                     71:        "invalid number format", /* ENUMFMT */
                     72:        "superfluous width argument", /* ENOWIDTH */
                     73:        "system: utsname error", /* EUTSNAME */
                     74:        "obsolete macro", /* EOBS */
                     75:        "end-of-line scope violation", /* EIMPBRK */
                     76:        "empty macro ignored", /* EIGNE */
                     77:        "unclosed explicit scope", /* EOPEN */
                     78:        "unterminated quoted phrase", /* EQUOTPHR */
                     79:        "closure macro without prior context", /* ENOCTX */
1.29      schwarze   80:        "no description found for library", /* ELIB */
                     81:        "bad child for parent context", /* EBADCHILD */
1.30      schwarze   82:        "list arguments preceding type", /* ENOTYPE */
1.44      schwarze   83:        "deprecated comment style", /* EBADCOMMENT */
1.1       kristaps   84: };
                     85:
                     86: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.7       schwarze   87:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   88:        "Sh",           "Ss",           "Pp",           "D1",
                     89:        "Dl",           "Bd",           "Ed",           "Bl",
                     90:        "El",           "It",           "Ad",           "An",
                     91:        "Ar",           "Cd",           "Cm",           "Dv",
                     92:        "Er",           "Ev",           "Ex",           "Fa",
                     93:        "Fd",           "Fl",           "Fn",           "Ft",
                     94:        "Ic",           "In",           "Li",           "Nd",
                     95:        "Nm",           "Op",           "Ot",           "Pa",
                     96:        "Rv",           "St",           "Va",           "Vt",
                     97:        /* LINTED */
1.33      schwarze   98:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   99:        /* LINTED */
1.33      schwarze  100:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps  101:        /* LINTED */
1.33      schwarze  102:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps  103:        "Ac",           "Ao",           "Aq",           "At",
                    104:        "Bc",           "Bf",           "Bo",           "Bq",
                    105:        "Bsx",          "Bx",           "Db",           "Dc",
                    106:        "Do",           "Dq",           "Ec",           "Ef",
                    107:        "Em",           "Eo",           "Fx",           "Ms",
                    108:        "No",           "Ns",           "Nx",           "Ox",
                    109:        "Pc",           "Pf",           "Po",           "Pq",
                    110:        "Qc",           "Ql",           "Qo",           "Qq",
                    111:        "Re",           "Rs",           "Sc",           "So",
                    112:        "Sq",           "Sm",           "Sx",           "Sy",
                    113:        "Tn",           "Ux",           "Xc",           "Xo",
                    114:        "Fo",           "Fc",           "Oo",           "Oc",
                    115:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze  116:        "Fr",           "Ud",           "Lb",           "Lp",
                    117:        "Lk",           "Mt",           "Brq",          "Bro",
1.1       kristaps  118:        /* LINTED */
1.33      schwarze  119:        "Brc",          "%C",           "Es",           "En",
1.1       kristaps  120:        /* LINTED */
1.33      schwarze  121:        "Dx",           "%Q",           "br",           "sp",
1.31      schwarze  122:        /* LINTED */
1.35      schwarze  123:        "%U",           "eos"
1.1       kristaps  124:        };
                    125:
                    126: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                    127:        "split",                "nosplit",              "ragged",
                    128:        "unfilled",             "literal",              "file",
                    129:        "offset",               "bullet",               "dash",
                    130:        "hyphen",               "item",                 "enum",
                    131:        "tag",                  "diag",                 "hang",
                    132:        "ohang",                "inset",                "column",
                    133:        "width",                "compact",              "std",
                    134:        "filled",               "words",                "emphasis",
1.30      schwarze  135:        "symbolic",             "nested",               "centered"
1.1       kristaps  136:        };
                    137:
                    138: const  char * const *mdoc_macronames = __mdoc_macronames;
                    139: const  char * const *mdoc_argnames = __mdoc_argnames;
                    140:
1.40      schwarze  141: static void              mdoc_node_free(struct mdoc_node *);
                    142: static void              mdoc_node_unlink(struct mdoc *,
                    143:                                struct mdoc_node *);
1.1       kristaps  144: static void              mdoc_free1(struct mdoc *);
1.32      schwarze  145: static void              mdoc_alloc1(struct mdoc *);
1.1       kristaps  146: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.37      schwarze  147:                                enum mdoct, enum mdoc_type);
1.1       kristaps  148: static int               node_append(struct mdoc *,
                    149:                                struct mdoc_node *);
1.45      schwarze  150: static int               mdoc_ptext(struct mdoc *, int, char *);
                    151: static int               mdoc_pmacro(struct mdoc *, int, char *);
1.1       kristaps  152: static int               macrowarn(struct mdoc *, int, const char *);
1.46      schwarze  153:
1.1       kristaps  154:
                    155: const struct mdoc_node *
                    156: mdoc_node(const struct mdoc *m)
                    157: {
                    158:
                    159:        return(MDOC_HALT & m->flags ? NULL : m->first);
                    160: }
                    161:
                    162:
                    163: const struct mdoc_meta *
                    164: mdoc_meta(const struct mdoc *m)
                    165: {
                    166:
                    167:        return(MDOC_HALT & m->flags ? NULL : &m->meta);
                    168: }
                    169:
                    170:
1.8       schwarze  171: /*
                    172:  * Frees volatile resources (parse tree, meta-data, fields).
                    173:  */
1.1       kristaps  174: static void
                    175: mdoc_free1(struct mdoc *mdoc)
                    176: {
                    177:
                    178:        if (mdoc->first)
1.40      schwarze  179:                mdoc_node_delete(mdoc, mdoc->first);
1.1       kristaps  180:        if (mdoc->meta.title)
                    181:                free(mdoc->meta.title);
                    182:        if (mdoc->meta.os)
                    183:                free(mdoc->meta.os);
                    184:        if (mdoc->meta.name)
                    185:                free(mdoc->meta.name);
                    186:        if (mdoc->meta.arch)
                    187:                free(mdoc->meta.arch);
                    188:        if (mdoc->meta.vol)
                    189:                free(mdoc->meta.vol);
                    190: }
                    191:
                    192:
1.8       schwarze  193: /*
                    194:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    195:  */
1.32      schwarze  196: static void
1.1       kristaps  197: mdoc_alloc1(struct mdoc *mdoc)
                    198: {
                    199:
1.32      schwarze  200:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.1       kristaps  201:        mdoc->flags = 0;
1.9       schwarze  202:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.32      schwarze  203:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.1       kristaps  204:        mdoc->first = mdoc->last;
                    205:        mdoc->last->type = MDOC_ROOT;
                    206:        mdoc->next = MDOC_NEXT_CHILD;
                    207: }
                    208:
                    209:
                    210: /*
1.8       schwarze  211:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    212:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    213:  * and the parser is ready for re-invocation on a new tree; however,
                    214:  * cross-parse non-volatile data is kept intact.
1.1       kristaps  215:  */
1.32      schwarze  216: void
1.1       kristaps  217: mdoc_reset(struct mdoc *mdoc)
                    218: {
                    219:
                    220:        mdoc_free1(mdoc);
1.32      schwarze  221:        mdoc_alloc1(mdoc);
1.1       kristaps  222: }
                    223:
                    224:
                    225: /*
1.8       schwarze  226:  * Completely free up all volatile and non-volatile parse resources.
                    227:  * After invocation, the pointer is no longer usable.
1.1       kristaps  228:  */
                    229: void
                    230: mdoc_free(struct mdoc *mdoc)
                    231: {
                    232:
                    233:        mdoc_free1(mdoc);
                    234:        free(mdoc);
                    235: }
                    236:
                    237:
1.8       schwarze  238: /*
                    239:  * Allocate volatile and non-volatile parse resources.
                    240:  */
1.1       kristaps  241: struct mdoc *
                    242: mdoc_alloc(void *data, int pflags, const struct mdoc_cb *cb)
                    243: {
                    244:        struct mdoc     *p;
                    245:
1.32      schwarze  246:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    247:
1.1       kristaps  248:        if (cb)
1.32      schwarze  249:                memcpy(&p->cb, cb, sizeof(struct mdoc_cb));
1.27      schwarze  250:
1.1       kristaps  251:        p->data = data;
                    252:        p->pflags = pflags;
                    253:
1.32      schwarze  254:        mdoc_hash_init();
                    255:        mdoc_alloc1(p);
                    256:        return(p);
1.1       kristaps  257: }
                    258:
                    259:
                    260: /*
                    261:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.8       schwarze  262:  * through to macro_end() in macro.c.
1.1       kristaps  263:  */
                    264: int
                    265: mdoc_endparse(struct mdoc *m)
                    266: {
                    267:
                    268:        if (MDOC_HALT & m->flags)
                    269:                return(0);
                    270:        else if (mdoc_macroend(m))
                    271:                return(1);
                    272:        m->flags |= MDOC_HALT;
                    273:        return(0);
                    274: }
                    275:
                    276:
                    277: /*
                    278:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  279:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  280:  */
                    281: int
                    282: mdoc_parseln(struct mdoc *m, int ln, char *buf)
                    283: {
                    284:
                    285:        if (MDOC_HALT & m->flags)
                    286:                return(0);
                    287:
1.47    ! schwarze  288:        m->flags |= MDOC_NEWLINE;
        !           289:        return('.' == *buf ?
        !           290:                        mdoc_pmacro(m, ln, buf) :
1.45      schwarze  291:                        mdoc_ptext(m, ln, buf));
1.1       kristaps  292: }
                    293:
                    294:
                    295: int
1.4       schwarze  296: mdoc_verr(struct mdoc *mdoc, int ln, int pos,
                    297:                const char *fmt, ...)
1.1       kristaps  298: {
                    299:        char             buf[256];
                    300:        va_list          ap;
                    301:
                    302:        if (NULL == mdoc->cb.mdoc_err)
                    303:                return(0);
                    304:
                    305:        va_start(ap, fmt);
                    306:        (void)vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
                    307:        va_end(ap);
1.12      schwarze  308:
1.1       kristaps  309:        return((*mdoc->cb.mdoc_err)(mdoc->data, ln, pos, buf));
                    310: }
                    311:
                    312:
                    313: int
1.12      schwarze  314: mdoc_vwarn(struct mdoc *mdoc, int ln, int pos, const char *fmt, ...)
1.1       kristaps  315: {
                    316:        char             buf[256];
                    317:        va_list          ap;
                    318:
                    319:        if (NULL == mdoc->cb.mdoc_warn)
                    320:                return(0);
                    321:
                    322:        va_start(ap, fmt);
                    323:        (void)vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
                    324:        va_end(ap);
1.17      schwarze  325:
1.11      schwarze  326:        return((*mdoc->cb.mdoc_warn)(mdoc->data, ln, pos, buf));
1.5       schwarze  327: }
                    328:
                    329:
                    330: int
1.14      schwarze  331: mdoc_err(struct mdoc *m, int line, int pos, int iserr, enum merr type)
1.5       schwarze  332: {
1.14      schwarze  333:        const char      *p;
1.5       schwarze  334:
1.14      schwarze  335:        p = __mdoc_merrnames[(int)type];
                    336:        assert(p);
1.5       schwarze  337:
1.14      schwarze  338:        if (iserr)
                    339:                return(mdoc_verr(m, line, pos, p));
1.16      schwarze  340:
1.14      schwarze  341:        return(mdoc_vwarn(m, line, pos, p));
1.5       schwarze  342: }
                    343:
                    344:
                    345: int
1.37      schwarze  346: mdoc_macro(struct mdoc *m, enum mdoct tok,
1.1       kristaps  347:                int ln, int pp, int *pos, char *buf)
                    348: {
1.46      schwarze  349:        assert(tok < MDOC_MAX);
                    350:
                    351:        /* If we're in the body, deny prologue calls. */
1.37      schwarze  352:
1.1       kristaps  353:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.9       schwarze  354:                        MDOC_PBODY & m->flags)
1.16      schwarze  355:                return(mdoc_perr(m, ln, pp, EPROLBODY));
1.46      schwarze  356:
                    357:        /* If we're in the prologue, deny "body" macros.  */
                    358:
1.1       kristaps  359:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
1.39      schwarze  360:                        ! (MDOC_PBODY & m->flags)) {
                    361:                if ( ! mdoc_pwarn(m, ln, pp, EBODYPROL))
                    362:                        return(0);
                    363:                if (NULL == m->meta.title)
                    364:                        m->meta.title = mandoc_strdup("unknown");
                    365:                if (NULL == m->meta.vol)
                    366:                        m->meta.vol = mandoc_strdup("local");
                    367:                if (NULL == m->meta.os)
                    368:                        m->meta.os = mandoc_strdup("local");
                    369:                if (0 == m->meta.date)
                    370:                        m->meta.date = time(NULL);
                    371:                m->flags |= MDOC_PBODY;
                    372:        }
1.1       kristaps  373:
                    374:        return((*mdoc_macros[tok].fp)(m, tok, ln, pp, pos, buf));
                    375: }
                    376:
                    377:
                    378: static int
                    379: node_append(struct mdoc *mdoc, struct mdoc_node *p)
                    380: {
                    381:
                    382:        assert(mdoc->last);
                    383:        assert(mdoc->first);
                    384:        assert(MDOC_ROOT != p->type);
                    385:
                    386:        switch (mdoc->next) {
                    387:        case (MDOC_NEXT_SIBLING):
                    388:                mdoc->last->next = p;
                    389:                p->prev = mdoc->last;
                    390:                p->parent = mdoc->last->parent;
                    391:                break;
                    392:        case (MDOC_NEXT_CHILD):
                    393:                mdoc->last->child = p;
                    394:                p->parent = mdoc->last;
                    395:                break;
                    396:        default:
                    397:                abort();
                    398:                /* NOTREACHED */
                    399:        }
                    400:
1.10      schwarze  401:        p->parent->nchild++;
                    402:
1.1       kristaps  403:        if ( ! mdoc_valid_pre(mdoc, p))
                    404:                return(0);
                    405:        if ( ! mdoc_action_pre(mdoc, p))
                    406:                return(0);
                    407:
                    408:        switch (p->type) {
                    409:        case (MDOC_HEAD):
                    410:                assert(MDOC_BLOCK == p->parent->type);
                    411:                p->parent->head = p;
                    412:                break;
                    413:        case (MDOC_TAIL):
                    414:                assert(MDOC_BLOCK == p->parent->type);
                    415:                p->parent->tail = p;
                    416:                break;
                    417:        case (MDOC_BODY):
                    418:                assert(MDOC_BLOCK == p->parent->type);
                    419:                p->parent->body = p;
                    420:                break;
                    421:        default:
                    422:                break;
                    423:        }
                    424:
                    425:        mdoc->last = p;
                    426:
                    427:        switch (p->type) {
                    428:        case (MDOC_TEXT):
                    429:                if ( ! mdoc_valid_post(mdoc))
                    430:                        return(0);
                    431:                if ( ! mdoc_action_post(mdoc))
                    432:                        return(0);
                    433:                break;
                    434:        default:
                    435:                break;
                    436:        }
                    437:
                    438:        return(1);
                    439: }
                    440:
                    441:
                    442: static struct mdoc_node *
1.37      schwarze  443: node_alloc(struct mdoc *m, int line, int pos,
                    444:                enum mdoct tok, enum mdoc_type type)
1.1       kristaps  445: {
                    446:        struct mdoc_node *p;
                    447:
1.32      schwarze  448:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.19      schwarze  449:        p->sec = m->lastsec;
1.1       kristaps  450:        p->line = line;
                    451:        p->pos = pos;
                    452:        p->tok = tok;
1.37      schwarze  453:        p->type = type;
1.47    ! schwarze  454:        if (MDOC_NEWLINE & m->flags)
        !           455:                p->flags |= MDOC_LINE;
        !           456:        m->flags &= ~MDOC_NEWLINE;
1.1       kristaps  457:        return(p);
                    458: }
                    459:
                    460:
                    461: int
1.37      schwarze  462: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  463: {
                    464:        struct mdoc_node *p;
                    465:
1.19      schwarze  466:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.25      schwarze  467:        if ( ! node_append(m, p))
                    468:                return(0);
                    469:        m->next = MDOC_NEXT_CHILD;
                    470:        return(1);
1.1       kristaps  471: }
                    472:
                    473:
                    474: int
1.37      schwarze  475: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  476: {
                    477:        struct mdoc_node *p;
                    478:
1.19      schwarze  479:        assert(m->first);
                    480:        assert(m->last);
1.1       kristaps  481:
1.19      schwarze  482:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.25      schwarze  483:        if ( ! node_append(m, p))
                    484:                return(0);
                    485:        m->next = MDOC_NEXT_CHILD;
                    486:        return(1);
1.1       kristaps  487: }
                    488:
                    489:
                    490: int
1.37      schwarze  491: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  492: {
                    493:        struct mdoc_node *p;
                    494:
1.19      schwarze  495:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.25      schwarze  496:        if ( ! node_append(m, p))
                    497:                return(0);
                    498:        m->next = MDOC_NEXT_CHILD;
                    499:        return(1);
1.1       kristaps  500: }
                    501:
                    502:
                    503: int
1.19      schwarze  504: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  505:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  506: {
                    507:        struct mdoc_node *p;
                    508:
1.19      schwarze  509:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.4       schwarze  510:        p->args = args;
                    511:        if (p->args)
1.1       kristaps  512:                (args->refcnt)++;
1.25      schwarze  513:        if ( ! node_append(m, p))
                    514:                return(0);
                    515:        m->next = MDOC_NEXT_CHILD;
                    516:        return(1);
1.1       kristaps  517: }
                    518:
                    519:
                    520: int
1.19      schwarze  521: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  522:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  523: {
                    524:        struct mdoc_node *p;
                    525:
1.19      schwarze  526:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.4       schwarze  527:        p->args = args;
                    528:        if (p->args)
1.1       kristaps  529:                (args->refcnt)++;
1.25      schwarze  530:        if ( ! node_append(m, p))
                    531:                return(0);
                    532:        m->next = MDOC_NEXT_CHILD;
                    533:        return(1);
1.1       kristaps  534: }
                    535:
                    536:
1.46      schwarze  537: int
                    538: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  539: {
1.19      schwarze  540:        struct mdoc_node *n;
1.46      schwarze  541:        size_t            sv, len;
                    542:
                    543:        len = strlen(p);
1.1       kristaps  544:
1.46      schwarze  545:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.32      schwarze  546:        n->string = mandoc_malloc(len + 1);
1.19      schwarze  547:        sv = strlcpy(n->string, p, len + 1);
                    548:
                    549:        /* Prohibit truncation. */
                    550:        assert(sv < len + 1);
                    551:
1.25      schwarze  552:        if ( ! node_append(m, n))
                    553:                return(0);
1.46      schwarze  554:
1.25      schwarze  555:        m->next = MDOC_NEXT_SIBLING;
                    556:        return(1);
1.19      schwarze  557: }
                    558:
                    559:
1.1       kristaps  560: void
                    561: mdoc_node_free(struct mdoc_node *p)
                    562: {
                    563:
                    564:        if (p->string)
                    565:                free(p->string);
                    566:        if (p->args)
                    567:                mdoc_argv_free(p->args);
                    568:        free(p);
                    569: }
                    570:
                    571:
1.40      schwarze  572: static void
                    573: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    574: {
                    575:
                    576:        /* Adjust siblings. */
                    577:
                    578:        if (n->prev)
                    579:                n->prev->next = n->next;
                    580:        if (n->next)
                    581:                n->next->prev = n->prev;
                    582:
                    583:        /* Adjust parent. */
                    584:
                    585:        if (n->parent) {
                    586:                n->parent->nchild--;
                    587:                if (n->parent->child == n)
                    588:                        n->parent->child = n->prev ? n->prev : n->next;
                    589:        }
                    590:
                    591:        /* Adjust parse point, if applicable. */
                    592:
                    593:        if (m && m->last == n) {
                    594:                if (n->prev) {
                    595:                        m->last = n->prev;
                    596:                        m->next = MDOC_NEXT_SIBLING;
                    597:                } else {
                    598:                        m->last = n->parent;
                    599:                        m->next = MDOC_NEXT_CHILD;
                    600:                }
                    601:        }
                    602:
                    603:        if (m && m->first == n)
                    604:                m->first = NULL;
                    605: }
                    606:
                    607:
1.1       kristaps  608: void
1.40      schwarze  609: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  610: {
                    611:
1.40      schwarze  612:        while (p->child) {
                    613:                assert(p->nchild);
                    614:                mdoc_node_delete(m, p->child);
                    615:        }
                    616:        assert(0 == p->nchild);
1.1       kristaps  617:
1.40      schwarze  618:        mdoc_node_unlink(m, p);
1.1       kristaps  619:        mdoc_node_free(p);
                    620: }
                    621:
                    622:
                    623: /*
                    624:  * Parse free-form text, that is, a line that does not begin with the
                    625:  * control character.
                    626:  */
                    627: static int
1.45      schwarze  628: mdoc_ptext(struct mdoc *m, int line, char *buf)
1.1       kristaps  629: {
1.46      schwarze  630:        int              i;
1.44      schwarze  631:
                    632:        /* Ignore bogus comments. */
                    633:
                    634:        if ('\\' == buf[0] && '.' == buf[1] && '\"' == buf[2])
                    635:                return(mdoc_pwarn(m, line, 0, EBADCOMMENT));
1.1       kristaps  636:
1.46      schwarze  637:        /* No text before an initial macro. */
                    638:
1.9       schwarze  639:        if (SEC_NONE == m->lastnamed)
1.16      schwarze  640:                return(mdoc_perr(m, line, 0, ETEXTPROL));
1.46      schwarze  641:
                    642:        /* Literal just gets pulled in as-is. */
1.19      schwarze  643:
1.25      schwarze  644:        if (MDOC_LITERAL & m->flags)
                    645:                return(mdoc_word_alloc(m, line, 0, buf));
1.19      schwarze  646:
1.46      schwarze  647:        /* Check for a blank line, which may also consist of spaces. */
1.19      schwarze  648:
                    649:        for (i = 0; ' ' == buf[i]; i++)
1.46      schwarze  650:                /* Skip to first non-space. */ ;
1.34      schwarze  651:
1.38      schwarze  652:        if ('\0' == buf[i]) {
                    653:                if ( ! mdoc_pwarn(m, line, 0, ENOBLANK))
                    654:                        return(0);
1.46      schwarze  655:
1.40      schwarze  656:                /*
1.46      schwarze  657:                 * Insert a `Pp' in the case of a blank line.  Technically,
                    658:                 * blank lines aren't allowed, but enough manuals assume this
                    659:                 * behaviour that we want to work around it.
1.40      schwarze  660:                 */
1.42      schwarze  661:                if ( ! mdoc_elem_alloc(m, line, 0, MDOC_Pp, NULL))
                    662:                        return(0);
1.46      schwarze  663:
1.42      schwarze  664:                m->next = MDOC_NEXT_SIBLING;
                    665:                return(1);
1.38      schwarze  666:        }
1.1       kristaps  667:
1.46      schwarze  668:        /*
                    669:         * Warn if the last un-escaped character is whitespace. Then
                    670:         * strip away the remaining spaces (tabs stay!).
1.19      schwarze  671:         */
                    672:
1.46      schwarze  673:        i = (int)strlen(buf);
                    674:        assert(i);
1.34      schwarze  675:
1.46      schwarze  676:        if (' ' == buf[i - 1] || '\t' == buf[i - 1]) {
                    677:                if (i > 1 && '\\' != buf[i - 2])
1.34      schwarze  678:                        if ( ! mdoc_pwarn(m, line, i - 1, ETAILWS))
                    679:                                return(0);
                    680:
1.46      schwarze  681:                for (--i; i && ' ' == buf[i]; i--)
                    682:                        /* Spin back to non-space. */ ;
1.34      schwarze  683:
1.46      schwarze  684:                /* Jump ahead of escaped whitespace. */
                    685:                i += '\\' == buf[i] ? 2 : 1;
1.34      schwarze  686:
1.46      schwarze  687:                buf[i] = '\0';
1.19      schwarze  688:        }
                    689:
1.46      schwarze  690:        /* Allocate the whole word. */
                    691:        if ( ! mdoc_word_alloc(m, line, 0, buf))
1.1       kristaps  692:                return(0);
                    693:
1.35      schwarze  694:        /*
                    695:         * Mark the end of a sentence.  Only works when you respect
                    696:         * Jason's rule: "new sentence, new line".
                    697:         */
                    698:        if ('.' == buf[i-1] || '!' == buf[i-1] || '?' == buf[i-1]) {
                    699:                m->next = MDOC_NEXT_SIBLING;
                    700:                if ( ! mdoc_elem_alloc(m, line, i, MDOC_eos, NULL))
                    701:                        return(0);
                    702:        }
                    703:
1.1       kristaps  704:        m->next = MDOC_NEXT_SIBLING;
                    705:        return(1);
                    706: }
1.19      schwarze  707:
1.1       kristaps  708:
                    709: static int
                    710: macrowarn(struct mdoc *m, int ln, const char *buf)
                    711: {
                    712:        if ( ! (MDOC_IGN_MACRO & m->pflags))
1.46      schwarze  713:                return(mdoc_verr(m, ln, 0, "unknown macro: %s%s",
1.1       kristaps  714:                                buf, strlen(buf) > 3 ? "..." : ""));
1.23      schwarze  715:        return(mdoc_vwarn(m, ln, 0, "unknown macro: %s%s",
1.1       kristaps  716:                                buf, strlen(buf) > 3 ? "..." : ""));
                    717: }
                    718:
                    719:
                    720: /*
                    721:  * Parse a macro line, that is, a line beginning with the control
                    722:  * character.
                    723:  */
                    724: int
1.45      schwarze  725: mdoc_pmacro(struct mdoc *m, int ln, char *buf)
1.1       kristaps  726: {
1.46      schwarze  727:        enum mdoct      tok;
1.47    ! schwarze  728:        int             i, j, sv;
1.46      schwarze  729:        char            mac[5];
1.35      schwarze  730:        struct mdoc_node *n;
                    731:        char             *t;
1.1       kristaps  732:
1.7       schwarze  733:        /* Empty lines are ignored. */
1.1       kristaps  734:
1.34      schwarze  735:        if ('\0' == buf[1])
1.1       kristaps  736:                return(1);
                    737:
1.24      schwarze  738:        i = 1;
                    739:
                    740:        /* Accept whitespace after the initial control char. */
                    741:
                    742:        if (' ' == buf[i]) {
                    743:                i++;
1.1       kristaps  744:                while (buf[i] && ' ' == buf[i])
                    745:                        i++;
1.34      schwarze  746:                if ('\0' == buf[i])
1.1       kristaps  747:                        return(1);
                    748:        }
                    749:
1.47    ! schwarze  750:        sv = i;
        !           751:
1.1       kristaps  752:        /* Copy the first word into a nil-terminated buffer. */
                    753:
1.24      schwarze  754:        for (j = 0; j < 4; j++, i++) {
1.34      schwarze  755:                if ('\0' == (mac[j] = buf[i]))
1.1       kristaps  756:                        break;
                    757:                else if (' ' == buf[i])
                    758:                        break;
1.26      schwarze  759:
                    760:                /* Check for invalid characters. */
                    761:
                    762:                if (isgraph((u_char)buf[i]))
                    763:                        continue;
                    764:                return(mdoc_perr(m, ln, i, EPRINT));
1.1       kristaps  765:        }
                    766:
1.24      schwarze  767:        mac[j] = 0;
1.1       kristaps  768:
1.24      schwarze  769:        if (j == 4 || j < 2) {
1.1       kristaps  770:                if ( ! macrowarn(m, ln, mac))
                    771:                        goto err;
                    772:                return(1);
                    773:        }
                    774:
1.46      schwarze  775:        if (MDOC_MAX == (tok = mdoc_hash_find(mac))) {
1.1       kristaps  776:                if ( ! macrowarn(m, ln, mac))
                    777:                        goto err;
                    778:                return(1);
                    779:        }
                    780:
                    781:        /* The macro is sane.  Jump to the next word. */
                    782:
                    783:        while (buf[i] && ' ' == buf[i])
                    784:                i++;
1.34      schwarze  785:
1.46      schwarze  786:        /*
                    787:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    788:         * into the parser as "text", so we only warn about spaces here.
                    789:         */
1.34      schwarze  790:
                    791:        if ('\0' == buf[i] && ' ' == buf[i - 1])
                    792:                if ( ! mdoc_pwarn(m, ln, i - 1, ETAILWS))
                    793:                        goto err;
1.1       kristaps  794:
1.24      schwarze  795:        /*
                    796:         * Begin recursive parse sequence.  Since we're at the start of
                    797:         * the line, we don't need to do callable/parseable checks.
                    798:         */
1.47    ! schwarze  799:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.1       kristaps  800:                goto err;
1.35      schwarze  801:
                    802:        /*
                    803:         * Mark the end of a sentence, but be careful not to insert
1.43      schwarze  804:         * markers into reference blocks and after ellipses in
                    805:         * function definitions.
1.35      schwarze  806:         */
                    807:        n = m->last;
                    808:        if (n->child)
                    809:                n = n->child;
                    810:        while (n->next)
                    811:                n = n->next;
1.43      schwarze  812:        if (MDOC_TEXT == n->type &&
                    813:            MDOC_Fn != n->parent->tok &&
                    814:            MDOC_Rs != m->last->parent->tok) {
1.35      schwarze  815:                t = n->string;
1.36      schwarze  816:                while (t[0] && t[1])
1.35      schwarze  817:                        t++;
                    818:                if ('.' == *t || '!' == *t || '?' == *t) {
                    819:                        if ( ! mdoc_elem_alloc(m, ln, i, MDOC_eos, NULL))
                    820:                                return(0);
                    821:                        m->next = MDOC_NEXT_SIBLING;
                    822:                }
                    823:        }
1.1       kristaps  824:
                    825:        return(1);
                    826:
                    827: err:   /* Error out. */
                    828:
                    829:        m->flags |= MDOC_HALT;
                    830:        return(0);
                    831: }
1.24      schwarze  832:
                    833: