[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 1.81

1.81    ! schwarze    1: /*     $Id: mdoc.c,v 1.80 2011/03/07 01:35:33 schwarze Exp $ */
1.1       kristaps    2: /*
1.77      schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.61      schwarze    4:  * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    5:  *
                      6:  * Permission to use, copy, modify, and distribute this software for any
1.3       schwarze    7:  * purpose with or without fee is hereby granted, provided that the above
                      8:  * copyright notice and this permission notice appear in all copies.
1.1       kristaps    9:  *
1.3       schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
                     11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
                     12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR 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.
1.1       kristaps   17:  */
1.28      schwarze   18: #include <sys/types.h>
                     19:
1.1       kristaps   20: #include <assert.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:
1.54      schwarze   27: #include "mandoc.h"
1.1       kristaps   28: #include "libmdoc.h"
1.32      schwarze   29: #include "libmandoc.h"
1.1       kristaps   30:
                     31: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.7       schwarze   32:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   33:        "Sh",           "Ss",           "Pp",           "D1",
                     34:        "Dl",           "Bd",           "Ed",           "Bl",
                     35:        "El",           "It",           "Ad",           "An",
                     36:        "Ar",           "Cd",           "Cm",           "Dv",
                     37:        "Er",           "Ev",           "Ex",           "Fa",
                     38:        "Fd",           "Fl",           "Fn",           "Ft",
                     39:        "Ic",           "In",           "Li",           "Nd",
                     40:        "Nm",           "Op",           "Ot",           "Pa",
                     41:        "Rv",           "St",           "Va",           "Vt",
                     42:        /* LINTED */
1.33      schwarze   43:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   44:        /* LINTED */
1.33      schwarze   45:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps   46:        /* LINTED */
1.33      schwarze   47:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   48:        "Ac",           "Ao",           "Aq",           "At",
                     49:        "Bc",           "Bf",           "Bo",           "Bq",
                     50:        "Bsx",          "Bx",           "Db",           "Dc",
                     51:        "Do",           "Dq",           "Ec",           "Ef",
                     52:        "Em",           "Eo",           "Fx",           "Ms",
                     53:        "No",           "Ns",           "Nx",           "Ox",
                     54:        "Pc",           "Pf",           "Po",           "Pq",
                     55:        "Qc",           "Ql",           "Qo",           "Qq",
                     56:        "Re",           "Rs",           "Sc",           "So",
                     57:        "Sq",           "Sm",           "Sx",           "Sy",
                     58:        "Tn",           "Ux",           "Xc",           "Xo",
                     59:        "Fo",           "Fc",           "Oo",           "Oc",
                     60:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze   61:        "Fr",           "Ud",           "Lb",           "Lp",
                     62:        "Lk",           "Mt",           "Brq",          "Bro",
1.1       kristaps   63:        /* LINTED */
1.33      schwarze   64:        "Brc",          "%C",           "Es",           "En",
1.1       kristaps   65:        /* LINTED */
1.33      schwarze   66:        "Dx",           "%Q",           "br",           "sp",
1.31      schwarze   67:        /* LINTED */
1.77      schwarze   68:        "%U",           "Ta"
1.1       kristaps   69:        };
                     70:
                     71: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                     72:        "split",                "nosplit",              "ragged",
                     73:        "unfilled",             "literal",              "file",
                     74:        "offset",               "bullet",               "dash",
                     75:        "hyphen",               "item",                 "enum",
                     76:        "tag",                  "diag",                 "hang",
                     77:        "ohang",                "inset",                "column",
                     78:        "width",                "compact",              "std",
                     79:        "filled",               "words",                "emphasis",
1.30      schwarze   80:        "symbolic",             "nested",               "centered"
1.1       kristaps   81:        };
                     82:
                     83: const  char * const *mdoc_macronames = __mdoc_macronames;
                     84: const  char * const *mdoc_argnames = __mdoc_argnames;
                     85:
1.40      schwarze   86: static void              mdoc_node_free(struct mdoc_node *);
                     87: static void              mdoc_node_unlink(struct mdoc *,
                     88:                                struct mdoc_node *);
1.1       kristaps   89: static void              mdoc_free1(struct mdoc *);
1.32      schwarze   90: static void              mdoc_alloc1(struct mdoc *);
1.1       kristaps   91: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.37      schwarze   92:                                enum mdoct, enum mdoc_type);
1.1       kristaps   93: static int               node_append(struct mdoc *,
                     94:                                struct mdoc_node *);
1.53      schwarze   95: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     96: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.1       kristaps   97:
                     98: const struct mdoc_node *
                     99: mdoc_node(const struct mdoc *m)
                    100: {
                    101:
1.78      schwarze  102:        assert( ! (MDOC_HALT & m->flags));
                    103:        return(m->first);
1.1       kristaps  104: }
                    105:
                    106:
                    107: const struct mdoc_meta *
                    108: mdoc_meta(const struct mdoc *m)
                    109: {
                    110:
1.78      schwarze  111:        assert( ! (MDOC_HALT & m->flags));
                    112:        return(&m->meta);
1.1       kristaps  113: }
                    114:
                    115:
1.8       schwarze  116: /*
                    117:  * Frees volatile resources (parse tree, meta-data, fields).
                    118:  */
1.1       kristaps  119: static void
                    120: mdoc_free1(struct mdoc *mdoc)
                    121: {
                    122:
                    123:        if (mdoc->first)
1.40      schwarze  124:                mdoc_node_delete(mdoc, mdoc->first);
1.1       kristaps  125:        if (mdoc->meta.title)
                    126:                free(mdoc->meta.title);
                    127:        if (mdoc->meta.os)
                    128:                free(mdoc->meta.os);
                    129:        if (mdoc->meta.name)
                    130:                free(mdoc->meta.name);
                    131:        if (mdoc->meta.arch)
                    132:                free(mdoc->meta.arch);
                    133:        if (mdoc->meta.vol)
                    134:                free(mdoc->meta.vol);
1.50      schwarze  135:        if (mdoc->meta.msec)
                    136:                free(mdoc->meta.msec);
1.1       kristaps  137: }
                    138:
                    139:
1.8       schwarze  140: /*
                    141:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    142:  */
1.32      schwarze  143: static void
1.1       kristaps  144: mdoc_alloc1(struct mdoc *mdoc)
                    145: {
                    146:
1.32      schwarze  147:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.1       kristaps  148:        mdoc->flags = 0;
1.9       schwarze  149:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.32      schwarze  150:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.1       kristaps  151:        mdoc->first = mdoc->last;
                    152:        mdoc->last->type = MDOC_ROOT;
                    153:        mdoc->next = MDOC_NEXT_CHILD;
                    154: }
                    155:
                    156:
                    157: /*
1.8       schwarze  158:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    159:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    160:  * and the parser is ready for re-invocation on a new tree; however,
                    161:  * cross-parse non-volatile data is kept intact.
1.1       kristaps  162:  */
1.32      schwarze  163: void
1.1       kristaps  164: mdoc_reset(struct mdoc *mdoc)
                    165: {
                    166:
                    167:        mdoc_free1(mdoc);
1.32      schwarze  168:        mdoc_alloc1(mdoc);
1.1       kristaps  169: }
                    170:
                    171:
                    172: /*
1.8       schwarze  173:  * Completely free up all volatile and non-volatile parse resources.
                    174:  * After invocation, the pointer is no longer usable.
1.1       kristaps  175:  */
                    176: void
                    177: mdoc_free(struct mdoc *mdoc)
                    178: {
                    179:
                    180:        mdoc_free1(mdoc);
                    181:        free(mdoc);
                    182: }
                    183:
                    184:
1.8       schwarze  185: /*
                    186:  * Allocate volatile and non-volatile parse resources.
                    187:  */
1.1       kristaps  188: struct mdoc *
1.65      schwarze  189: mdoc_alloc(struct regset *regs, void *data, mandocmsg msg)
1.1       kristaps  190: {
                    191:        struct mdoc     *p;
                    192:
1.32      schwarze  193:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    194:
1.54      schwarze  195:        p->msg = msg;
1.1       kristaps  196:        p->data = data;
1.58      schwarze  197:        p->regs = regs;
1.1       kristaps  198:
1.32      schwarze  199:        mdoc_hash_init();
                    200:        mdoc_alloc1(p);
                    201:        return(p);
1.1       kristaps  202: }
                    203:
                    204:
                    205: /*
                    206:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.8       schwarze  207:  * through to macro_end() in macro.c.
1.1       kristaps  208:  */
                    209: int
                    210: mdoc_endparse(struct mdoc *m)
                    211: {
                    212:
1.78      schwarze  213:        assert( ! (MDOC_HALT & m->flags));
                    214:        if (mdoc_macroend(m))
1.1       kristaps  215:                return(1);
                    216:        m->flags |= MDOC_HALT;
                    217:        return(0);
1.81    ! schwarze  218: }
        !           219:
        !           220: int
        !           221: mdoc_addeqn(struct mdoc *m, const struct eqn *ep)
        !           222: {
        !           223:        struct mdoc_node *n;
        !           224:
        !           225:        assert( ! (MDOC_HALT & m->flags));
        !           226:
        !           227:        /* No text before an initial macro. */
        !           228:
        !           229:        if (SEC_NONE == m->lastnamed) {
        !           230:                mdoc_pmsg(m, ep->line, ep->pos, MANDOCERR_NOTEXT);
        !           231:                return(1);
        !           232:        }
        !           233:
        !           234:        n = node_alloc(m, ep->line, ep->pos, MDOC_MAX, MDOC_EQN);
        !           235:        n->eqn = ep;
        !           236:
        !           237:        if ( ! node_append(m, n))
        !           238:                return(0);
        !           239:
        !           240:        m->next = MDOC_NEXT_SIBLING;
        !           241:        return(1);
1.1       kristaps  242: }
                    243:
1.77      schwarze  244: int
                    245: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    246: {
1.79      schwarze  247:        struct mdoc_node *n;
1.77      schwarze  248:
1.78      schwarze  249:        assert( ! (MDOC_HALT & m->flags));
1.77      schwarze  250:
                    251:        /* No text before an initial macro. */
                    252:
                    253:        if (SEC_NONE == m->lastnamed) {
1.79      schwarze  254:                mdoc_pmsg(m, sp->line, 0, MANDOCERR_NOTEXT);
1.77      schwarze  255:                return(1);
                    256:        }
                    257:
1.79      schwarze  258:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
                    259:        n->span = sp;
                    260:
                    261:        if ( ! node_append(m, n))
                    262:                return(0);
                    263:
                    264:        m->next = MDOC_NEXT_SIBLING;
                    265:        return(1);
1.77      schwarze  266: }
                    267:
1.1       kristaps  268:
                    269: /*
                    270:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  271:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  272:  */
                    273: int
1.53      schwarze  274: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  275: {
                    276:
1.78      schwarze  277:        assert( ! (MDOC_HALT & m->flags));
1.1       kristaps  278:
1.47      schwarze  279:        m->flags |= MDOC_NEWLINE;
1.60      schwarze  280:
                    281:        /*
                    282:         * Let the roff nS register switch SYNOPSIS mode early,
                    283:         * such that the parser knows at all times
                    284:         * whether this mode is on or off.
                    285:         * Note that this mode is also switched by the Sh macro.
                    286:         */
                    287:        if (m->regs->regs[(int)REG_nS].set) {
                    288:                if (m->regs->regs[(int)REG_nS].v.u)
                    289:                        m->flags |= MDOC_SYNOPSIS;
                    290:                else
                    291:                        m->flags &= ~MDOC_SYNOPSIS;
                    292:        }
                    293:
1.53      schwarze  294:        return(('.' == buf[offs] || '\'' == buf[offs]) ?
                    295:                        mdoc_pmacro(m, ln, buf, offs) :
                    296:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  297: }
                    298:
                    299:
                    300: int
1.54      schwarze  301: mdoc_vmsg(struct mdoc *mdoc, enum mandocerr t,
                    302:                int ln, int pos, const char *fmt, ...)
1.1       kristaps  303: {
                    304:        char             buf[256];
                    305:        va_list          ap;
                    306:
                    307:        va_start(ap, fmt);
1.54      schwarze  308:        vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
1.1       kristaps  309:        va_end(ap);
1.17      schwarze  310:
1.54      schwarze  311:        return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
1.5       schwarze  312: }
                    313:
                    314:
                    315: int
1.58      schwarze  316: mdoc_macro(MACRO_PROT_ARGS)
1.1       kristaps  317: {
1.46      schwarze  318:        assert(tok < MDOC_MAX);
                    319:
                    320:        /* If we're in the body, deny prologue calls. */
1.37      schwarze  321:
1.1       kristaps  322:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.76      schwarze  323:                        MDOC_PBODY & m->flags) {
                    324:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    325:                return(1);
                    326:        }
1.46      schwarze  327:
                    328:        /* If we're in the prologue, deny "body" macros.  */
                    329:
1.1       kristaps  330:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
1.39      schwarze  331:                        ! (MDOC_PBODY & m->flags)) {
1.76      schwarze  332:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.71      schwarze  333:                if (NULL == m->meta.msec)
                    334:                        m->meta.msec = mandoc_strdup("1");
1.39      schwarze  335:                if (NULL == m->meta.title)
1.56      schwarze  336:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.39      schwarze  337:                if (NULL == m->meta.vol)
1.56      schwarze  338:                        m->meta.vol = mandoc_strdup("LOCAL");
1.39      schwarze  339:                if (NULL == m->meta.os)
1.56      schwarze  340:                        m->meta.os = mandoc_strdup("LOCAL");
1.80      schwarze  341:                if (NULL == m->meta.date)
                    342:                        m->meta.date = mandoc_normdate(NULL,
                    343:                            m->msg, m->data, line, ppos);
1.39      schwarze  344:                m->flags |= MDOC_PBODY;
                    345:        }
1.1       kristaps  346:
1.58      schwarze  347:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.1       kristaps  348: }
                    349:
                    350:
                    351: static int
                    352: node_append(struct mdoc *mdoc, struct mdoc_node *p)
                    353: {
                    354:
                    355:        assert(mdoc->last);
                    356:        assert(mdoc->first);
                    357:        assert(MDOC_ROOT != p->type);
                    358:
                    359:        switch (mdoc->next) {
                    360:        case (MDOC_NEXT_SIBLING):
                    361:                mdoc->last->next = p;
                    362:                p->prev = mdoc->last;
                    363:                p->parent = mdoc->last->parent;
                    364:                break;
                    365:        case (MDOC_NEXT_CHILD):
                    366:                mdoc->last->child = p;
                    367:                p->parent = mdoc->last;
                    368:                break;
                    369:        default:
                    370:                abort();
                    371:                /* NOTREACHED */
                    372:        }
                    373:
1.10      schwarze  374:        p->parent->nchild++;
                    375:
1.75      schwarze  376:        /*
                    377:         * Copy over the normalised-data pointer of our parent.  Not
                    378:         * everybody has one, but copying a null pointer is fine.
                    379:         */
                    380:
                    381:        switch (p->type) {
                    382:        case (MDOC_BODY):
                    383:                /* FALLTHROUGH */
                    384:        case (MDOC_TAIL):
                    385:                /* FALLTHROUGH */
                    386:        case (MDOC_HEAD):
                    387:                p->norm = p->parent->norm;
                    388:                break;
                    389:        default:
                    390:                break;
                    391:        }
                    392:
1.1       kristaps  393:        if ( ! mdoc_valid_pre(mdoc, p))
                    394:                return(0);
                    395:
                    396:        switch (p->type) {
                    397:        case (MDOC_HEAD):
                    398:                assert(MDOC_BLOCK == p->parent->type);
                    399:                p->parent->head = p;
                    400:                break;
                    401:        case (MDOC_TAIL):
                    402:                assert(MDOC_BLOCK == p->parent->type);
                    403:                p->parent->tail = p;
                    404:                break;
                    405:        case (MDOC_BODY):
1.59      schwarze  406:                if (p->end)
                    407:                        break;
1.1       kristaps  408:                assert(MDOC_BLOCK == p->parent->type);
                    409:                p->parent->body = p;
                    410:                break;
                    411:        default:
                    412:                break;
                    413:        }
                    414:
                    415:        mdoc->last = p;
                    416:
                    417:        switch (p->type) {
1.77      schwarze  418:        case (MDOC_TBL):
                    419:                /* FALLTHROUGH */
1.1       kristaps  420:        case (MDOC_TEXT):
                    421:                if ( ! mdoc_valid_post(mdoc))
                    422:                        return(0);
                    423:                break;
                    424:        default:
                    425:                break;
                    426:        }
                    427:
                    428:        return(1);
                    429: }
                    430:
                    431:
                    432: static struct mdoc_node *
1.37      schwarze  433: node_alloc(struct mdoc *m, int line, int pos,
                    434:                enum mdoct tok, enum mdoc_type type)
1.1       kristaps  435: {
                    436:        struct mdoc_node *p;
                    437:
1.32      schwarze  438:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.19      schwarze  439:        p->sec = m->lastsec;
1.1       kristaps  440:        p->line = line;
                    441:        p->pos = pos;
                    442:        p->tok = tok;
1.37      schwarze  443:        p->type = type;
1.58      schwarze  444:
                    445:        /* Flag analysis. */
                    446:
1.60      schwarze  447:        if (MDOC_SYNOPSIS & m->flags)
                    448:                p->flags |= MDOC_SYNPRETTY;
                    449:        else
                    450:                p->flags &= ~MDOC_SYNPRETTY;
1.47      schwarze  451:        if (MDOC_NEWLINE & m->flags)
                    452:                p->flags |= MDOC_LINE;
                    453:        m->flags &= ~MDOC_NEWLINE;
1.58      schwarze  454:
1.1       kristaps  455:        return(p);
                    456: }
                    457:
                    458:
                    459: int
1.37      schwarze  460: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  461: {
                    462:        struct mdoc_node *p;
                    463:
1.19      schwarze  464:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.25      schwarze  465:        if ( ! node_append(m, p))
                    466:                return(0);
                    467:        m->next = MDOC_NEXT_CHILD;
                    468:        return(1);
1.1       kristaps  469: }
                    470:
                    471:
                    472: int
1.37      schwarze  473: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  474: {
                    475:        struct mdoc_node *p;
                    476:
1.19      schwarze  477:        assert(m->first);
                    478:        assert(m->last);
1.1       kristaps  479:
1.19      schwarze  480:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.25      schwarze  481:        if ( ! node_append(m, p))
                    482:                return(0);
                    483:        m->next = MDOC_NEXT_CHILD;
                    484:        return(1);
1.1       kristaps  485: }
                    486:
                    487:
                    488: int
1.37      schwarze  489: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  490: {
                    491:        struct mdoc_node *p;
                    492:
1.19      schwarze  493:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.25      schwarze  494:        if ( ! node_append(m, p))
                    495:                return(0);
                    496:        m->next = MDOC_NEXT_CHILD;
1.59      schwarze  497:        return(1);
                    498: }
                    499:
                    500:
                    501: int
                    502: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    503:                struct mdoc_node *body, enum mdoc_endbody end)
                    504: {
                    505:        struct mdoc_node *p;
                    506:
                    507:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    508:        p->pending = body;
                    509:        p->end = end;
                    510:        if ( ! node_append(m, p))
                    511:                return(0);
                    512:        m->next = MDOC_NEXT_SIBLING;
1.25      schwarze  513:        return(1);
1.1       kristaps  514: }
                    515:
                    516:
                    517: int
1.19      schwarze  518: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  519:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  520: {
                    521:        struct mdoc_node *p;
                    522:
1.19      schwarze  523:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.4       schwarze  524:        p->args = args;
                    525:        if (p->args)
1.1       kristaps  526:                (args->refcnt)++;
1.75      schwarze  527:
                    528:        switch (tok) {
                    529:        case (MDOC_Bd):
                    530:                /* FALLTHROUGH */
                    531:        case (MDOC_Bf):
                    532:                /* FALLTHROUGH */
                    533:        case (MDOC_Bl):
                    534:                /* FALLTHROUGH */
                    535:        case (MDOC_Rs):
                    536:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    537:                break;
                    538:        default:
                    539:                break;
                    540:        }
                    541:
1.25      schwarze  542:        if ( ! node_append(m, p))
                    543:                return(0);
                    544:        m->next = MDOC_NEXT_CHILD;
                    545:        return(1);
1.1       kristaps  546: }
                    547:
                    548:
                    549: int
1.19      schwarze  550: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  551:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  552: {
                    553:        struct mdoc_node *p;
                    554:
1.19      schwarze  555:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.4       schwarze  556:        p->args = args;
                    557:        if (p->args)
1.1       kristaps  558:                (args->refcnt)++;
1.75      schwarze  559:
                    560:        switch (tok) {
                    561:        case (MDOC_An):
                    562:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    563:                break;
                    564:        default:
                    565:                break;
                    566:        }
                    567:
1.25      schwarze  568:        if ( ! node_append(m, p))
                    569:                return(0);
                    570:        m->next = MDOC_NEXT_CHILD;
                    571:        return(1);
1.1       kristaps  572: }
                    573:
1.46      schwarze  574: int
                    575: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  576: {
1.19      schwarze  577:        struct mdoc_node *n;
1.46      schwarze  578:        size_t            sv, len;
                    579:
                    580:        len = strlen(p);
1.1       kristaps  581:
1.46      schwarze  582:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.32      schwarze  583:        n->string = mandoc_malloc(len + 1);
1.19      schwarze  584:        sv = strlcpy(n->string, p, len + 1);
                    585:
                    586:        /* Prohibit truncation. */
                    587:        assert(sv < len + 1);
                    588:
1.25      schwarze  589:        if ( ! node_append(m, n))
                    590:                return(0);
1.46      schwarze  591:
1.25      schwarze  592:        m->next = MDOC_NEXT_SIBLING;
                    593:        return(1);
1.19      schwarze  594: }
                    595:
                    596:
1.61      schwarze  597: static void
1.1       kristaps  598: mdoc_node_free(struct mdoc_node *p)
                    599: {
1.61      schwarze  600:
1.75      schwarze  601:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
                    602:                free(p->norm);
1.1       kristaps  603:        if (p->string)
                    604:                free(p->string);
                    605:        if (p->args)
                    606:                mdoc_argv_free(p->args);
                    607:        free(p);
                    608: }
                    609:
                    610:
1.40      schwarze  611: static void
                    612: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    613: {
                    614:
                    615:        /* Adjust siblings. */
                    616:
                    617:        if (n->prev)
                    618:                n->prev->next = n->next;
                    619:        if (n->next)
                    620:                n->next->prev = n->prev;
                    621:
                    622:        /* Adjust parent. */
                    623:
                    624:        if (n->parent) {
                    625:                n->parent->nchild--;
                    626:                if (n->parent->child == n)
                    627:                        n->parent->child = n->prev ? n->prev : n->next;
1.72      schwarze  628:                if (n->parent->last == n)
                    629:                        n->parent->last = n->prev ? n->prev : NULL;
1.40      schwarze  630:        }
                    631:
                    632:        /* Adjust parse point, if applicable. */
                    633:
                    634:        if (m && m->last == n) {
                    635:                if (n->prev) {
                    636:                        m->last = n->prev;
                    637:                        m->next = MDOC_NEXT_SIBLING;
                    638:                } else {
                    639:                        m->last = n->parent;
                    640:                        m->next = MDOC_NEXT_CHILD;
                    641:                }
                    642:        }
                    643:
                    644:        if (m && m->first == n)
                    645:                m->first = NULL;
                    646: }
                    647:
                    648:
1.1       kristaps  649: void
1.40      schwarze  650: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  651: {
                    652:
1.40      schwarze  653:        while (p->child) {
                    654:                assert(p->nchild);
                    655:                mdoc_node_delete(m, p->child);
                    656:        }
                    657:        assert(0 == p->nchild);
1.1       kristaps  658:
1.40      schwarze  659:        mdoc_node_unlink(m, p);
1.1       kristaps  660:        mdoc_node_free(p);
                    661: }
                    662:
                    663:
                    664: /*
                    665:  * Parse free-form text, that is, a line that does not begin with the
                    666:  * control character.
                    667:  */
                    668: static int
1.53      schwarze  669: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  670: {
1.56      schwarze  671:        char             *c, *ws, *end;
                    672:        struct mdoc_node *n;
1.44      schwarze  673:
                    674:        /* Ignore bogus comments. */
                    675:
1.53      schwarze  676:        if ('\\' == buf[offs] &&
                    677:                        '.' == buf[offs + 1] &&
1.76      schwarze  678:                        '"' == buf[offs + 2]) {
                    679:                mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT);
                    680:                return(1);
                    681:        }
1.1       kristaps  682:
1.46      schwarze  683:        /* No text before an initial macro. */
                    684:
1.76      schwarze  685:        if (SEC_NONE == m->lastnamed) {
                    686:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    687:                return(1);
                    688:        }
1.46      schwarze  689:
1.56      schwarze  690:        assert(m->last);
                    691:        n = m->last;
                    692:
                    693:        /*
                    694:         * Divert directly to list processing if we're encountering a
                    695:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
                    696:         * (a MDOC_BODY means it's already open, in which case we should
                    697:         * process within its context in the normal way).
                    698:         */
                    699:
                    700:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.75      schwarze  701:                        LIST_column == n->norm->Bl.type) {
1.56      schwarze  702:                /* `Bl' is open without any children. */
                    703:                m->flags |= MDOC_FREECOL;
                    704:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    705:        }
                    706:
                    707:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    708:                        NULL != n->parent &&
                    709:                        MDOC_Bl == n->parent->tok &&
1.75      schwarze  710:                        LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  711:                /* `Bl' has block-level `It' children. */
                    712:                m->flags |= MDOC_FREECOL;
                    713:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    714:        }
                    715:
1.52      schwarze  716:        /*
                    717:         * Search for the beginning of unescaped trailing whitespace (ws)
                    718:         * and for the first character not to be output (end).
                    719:         */
1.56      schwarze  720:
                    721:        /* FIXME: replace with strcspn(). */
1.52      schwarze  722:        ws = NULL;
1.53      schwarze  723:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  724:                switch (*c) {
1.55      schwarze  725:                case '-':
                    726:                        if (mandoc_hyph(buf + offs, c))
                    727:                                *c = ASCII_HYPH;
1.56      schwarze  728:                        ws = NULL;
1.55      schwarze  729:                        break;
1.52      schwarze  730:                case ' ':
                    731:                        if (NULL == ws)
                    732:                                ws = c;
                    733:                        continue;
                    734:                case '\t':
                    735:                        /*
                    736:                         * Always warn about trailing tabs,
                    737:                         * even outside literal context,
                    738:                         * where they should be put on the next line.
                    739:                         */
                    740:                        if (NULL == ws)
                    741:                                ws = c;
                    742:                        /*
                    743:                         * Strip trailing tabs in literal context only;
                    744:                         * outside, they affect the next line.
                    745:                         */
                    746:                        if (MDOC_LITERAL & m->flags)
                    747:                                continue;
                    748:                        break;
                    749:                case '\\':
                    750:                        /* Skip the escaped character, too, if any. */
                    751:                        if (c[1])
                    752:                                c++;
                    753:                        /* FALLTHROUGH */
                    754:                default:
                    755:                        ws = NULL;
                    756:                        break;
                    757:                }
                    758:                end = c + 1;
                    759:        }
                    760:        *end = '\0';
1.19      schwarze  761:
1.52      schwarze  762:        if (ws)
1.76      schwarze  763:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.34      schwarze  764:
1.53      schwarze  765:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.76      schwarze  766:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.46      schwarze  767:
1.40      schwarze  768:                /*
1.66      schwarze  769:                 * Insert a `sp' in the case of a blank line.  Technically,
1.46      schwarze  770:                 * blank lines aren't allowed, but enough manuals assume this
                    771:                 * behaviour that we want to work around it.
1.40      schwarze  772:                 */
1.66      schwarze  773:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.42      schwarze  774:                        return(0);
1.46      schwarze  775:
1.42      schwarze  776:                m->next = MDOC_NEXT_SIBLING;
                    777:                return(1);
1.38      schwarze  778:        }
1.1       kristaps  779:
1.53      schwarze  780:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
1.1       kristaps  781:                return(0);
                    782:
1.52      schwarze  783:        if (MDOC_LITERAL & m->flags)
                    784:                return(1);
                    785:
1.35      schwarze  786:        /*
1.48      schwarze  787:         * End-of-sentence check.  If the last character is an unescaped
                    788:         * EOS character, then flag the node as being the end of a
                    789:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  790:         */
                    791:
1.52      schwarze  792:        assert(buf < end);
1.48      schwarze  793:
1.62      schwarze  794:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.48      schwarze  795:                m->last->flags |= MDOC_EOS;
                    796:
1.1       kristaps  797:        return(1);
                    798: }
1.19      schwarze  799:
1.1       kristaps  800:
                    801: /*
                    802:  * Parse a macro line, that is, a line beginning with the control
                    803:  * character.
                    804:  */
1.61      schwarze  805: static int
1.53      schwarze  806: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  807: {
1.56      schwarze  808:        enum mdoct        tok;
                    809:        int               i, j, sv;
                    810:        char              mac[5];
                    811:        struct mdoc_node *n;
1.1       kristaps  812:
1.7       schwarze  813:        /* Empty lines are ignored. */
1.1       kristaps  814:
1.53      schwarze  815:        offs++;
                    816:
                    817:        if ('\0' == buf[offs])
1.1       kristaps  818:                return(1);
                    819:
1.53      schwarze  820:        i = offs;
1.24      schwarze  821:
1.63      schwarze  822:        /* Accept tabs/whitespace after the initial control char. */
1.24      schwarze  823:
1.63      schwarze  824:        if (' ' == buf[i] || '\t' == buf[i]) {
1.24      schwarze  825:                i++;
1.63      schwarze  826:                while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
1.1       kristaps  827:                        i++;
1.34      schwarze  828:                if ('\0' == buf[i])
1.1       kristaps  829:                        return(1);
                    830:        }
                    831:
1.47      schwarze  832:        sv = i;
                    833:
1.63      schwarze  834:        /*
1.64      schwarze  835:         * Copy the first word into a nil-terminated buffer.
                    836:         * Stop copying when a tab, space, or eoln is encountered.
1.63      schwarze  837:         */
1.1       kristaps  838:
1.64      schwarze  839:        j = 0;
                    840:        while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
                    841:                mac[j++] = buf[i++];
1.53      schwarze  842:        mac[j] = '\0';
1.1       kristaps  843:
1.65      schwarze  844:        tok = (j > 1 || j < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
                    845:        if (MDOC_MAX == tok) {
1.69      schwarze  846:                mdoc_vmsg(m, MANDOCERR_MACRO, ln, sv, "%s", buf + sv - 1);
1.1       kristaps  847:                return(1);
                    848:        }
                    849:
1.63      schwarze  850:        /* Disregard the first trailing tab, if applicable. */
                    851:
                    852:        if ('\t' == buf[i])
                    853:                i++;
                    854:
                    855:        /* Jump to the next non-whitespace word. */
1.1       kristaps  856:
                    857:        while (buf[i] && ' ' == buf[i])
                    858:                i++;
1.34      schwarze  859:
1.46      schwarze  860:        /*
                    861:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    862:         * into the parser as "text", so we only warn about spaces here.
                    863:         */
1.34      schwarze  864:
                    865:        if ('\0' == buf[i] && ' ' == buf[i - 1])
1.76      schwarze  866:                mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE);
1.1       kristaps  867:
1.56      schwarze  868:        /*
                    869:         * If an initial macro or a list invocation, divert directly
                    870:         * into macro processing.
                    871:         */
                    872:
                    873:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
                    874:                if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
                    875:                        goto err;
                    876:                return(1);
                    877:        }
                    878:
                    879:        n = m->last;
                    880:        assert(m->last);
                    881:
                    882:        /*
                    883:         * If the first macro of a `Bl -column', open an `It' block
                    884:         * context around the parsed macro.
                    885:         */
                    886:
                    887:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.75      schwarze  888:                        LIST_column == n->norm->Bl.type) {
1.56      schwarze  889:                m->flags |= MDOC_FREECOL;
1.58      schwarze  890:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.56      schwarze  891:                        goto err;
                    892:                return(1);
                    893:        }
                    894:
                    895:        /*
                    896:         * If we're following a block-level `It' within a `Bl -column'
                    897:         * context (perhaps opened in the above block or in ptext()),
                    898:         * then open an `It' block context around the parsed macro.
1.24      schwarze  899:         */
1.56      schwarze  900:
                    901:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    902:                        NULL != n->parent &&
                    903:                        MDOC_Bl == n->parent->tok &&
1.75      schwarze  904:                        LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  905:                m->flags |= MDOC_FREECOL;
                    906:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
                    907:                        goto err;
                    908:                return(1);
                    909:        }
                    910:
                    911:        /* Normal processing of a macro. */
                    912:
1.47      schwarze  913:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.1       kristaps  914:                goto err;
                    915:
                    916:        return(1);
                    917:
                    918: err:   /* Error out. */
                    919:
                    920:        m->flags |= MDOC_HALT;
                    921:        return(0);
                    922: }
1.24      schwarze  923:
                    924: