[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 1.83

1.83    ! schwarze    1: /*     $Id: mdoc.c,v 1.82 2011/04/21 22:59:54 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:
1.83    ! schwarze   27: #include "mdoc.h"
1.54      schwarze   28: #include "mandoc.h"
1.1       kristaps   29: #include "libmdoc.h"
1.32      schwarze   30: #include "libmandoc.h"
1.1       kristaps   31:
                     32: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.7       schwarze   33:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   34:        "Sh",           "Ss",           "Pp",           "D1",
                     35:        "Dl",           "Bd",           "Ed",           "Bl",
                     36:        "El",           "It",           "Ad",           "An",
                     37:        "Ar",           "Cd",           "Cm",           "Dv",
                     38:        "Er",           "Ev",           "Ex",           "Fa",
                     39:        "Fd",           "Fl",           "Fn",           "Ft",
                     40:        "Ic",           "In",           "Li",           "Nd",
                     41:        "Nm",           "Op",           "Ot",           "Pa",
                     42:        "Rv",           "St",           "Va",           "Vt",
                     43:        /* LINTED */
1.33      schwarze   44:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   45:        /* LINTED */
1.33      schwarze   46:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps   47:        /* LINTED */
1.33      schwarze   48:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   49:        "Ac",           "Ao",           "Aq",           "At",
                     50:        "Bc",           "Bf",           "Bo",           "Bq",
                     51:        "Bsx",          "Bx",           "Db",           "Dc",
                     52:        "Do",           "Dq",           "Ec",           "Ef",
                     53:        "Em",           "Eo",           "Fx",           "Ms",
                     54:        "No",           "Ns",           "Nx",           "Ox",
                     55:        "Pc",           "Pf",           "Po",           "Pq",
                     56:        "Qc",           "Ql",           "Qo",           "Qq",
                     57:        "Re",           "Rs",           "Sc",           "So",
                     58:        "Sq",           "Sm",           "Sx",           "Sy",
                     59:        "Tn",           "Ux",           "Xc",           "Xo",
                     60:        "Fo",           "Fc",           "Oo",           "Oc",
                     61:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze   62:        "Fr",           "Ud",           "Lb",           "Lp",
                     63:        "Lk",           "Mt",           "Brq",          "Bro",
1.1       kristaps   64:        /* LINTED */
1.33      schwarze   65:        "Brc",          "%C",           "Es",           "En",
1.1       kristaps   66:        /* LINTED */
1.33      schwarze   67:        "Dx",           "%Q",           "br",           "sp",
1.31      schwarze   68:        /* LINTED */
1.77      schwarze   69:        "%U",           "Ta"
1.1       kristaps   70:        };
                     71:
                     72: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                     73:        "split",                "nosplit",              "ragged",
                     74:        "unfilled",             "literal",              "file",
                     75:        "offset",               "bullet",               "dash",
                     76:        "hyphen",               "item",                 "enum",
                     77:        "tag",                  "diag",                 "hang",
                     78:        "ohang",                "inset",                "column",
                     79:        "width",                "compact",              "std",
                     80:        "filled",               "words",                "emphasis",
1.30      schwarze   81:        "symbolic",             "nested",               "centered"
1.1       kristaps   82:        };
                     83:
                     84: const  char * const *mdoc_macronames = __mdoc_macronames;
                     85: const  char * const *mdoc_argnames = __mdoc_argnames;
                     86:
1.40      schwarze   87: static void              mdoc_node_free(struct mdoc_node *);
                     88: static void              mdoc_node_unlink(struct mdoc *,
                     89:                                struct mdoc_node *);
1.1       kristaps   90: static void              mdoc_free1(struct mdoc *);
1.32      schwarze   91: static void              mdoc_alloc1(struct mdoc *);
1.1       kristaps   92: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.37      schwarze   93:                                enum mdoct, enum mdoc_type);
1.1       kristaps   94: static int               node_append(struct mdoc *,
                     95:                                struct mdoc_node *);
1.53      schwarze   96: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     97: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.1       kristaps   98:
                     99: const struct mdoc_node *
                    100: mdoc_node(const struct mdoc *m)
                    101: {
                    102:
1.78      schwarze  103:        assert( ! (MDOC_HALT & m->flags));
                    104:        return(m->first);
1.1       kristaps  105: }
                    106:
                    107:
                    108: const struct mdoc_meta *
                    109: mdoc_meta(const struct mdoc *m)
                    110: {
                    111:
1.78      schwarze  112:        assert( ! (MDOC_HALT & m->flags));
                    113:        return(&m->meta);
1.1       kristaps  114: }
                    115:
                    116:
1.8       schwarze  117: /*
                    118:  * Frees volatile resources (parse tree, meta-data, fields).
                    119:  */
1.1       kristaps  120: static void
                    121: mdoc_free1(struct mdoc *mdoc)
                    122: {
                    123:
                    124:        if (mdoc->first)
1.40      schwarze  125:                mdoc_node_delete(mdoc, mdoc->first);
1.1       kristaps  126:        if (mdoc->meta.title)
                    127:                free(mdoc->meta.title);
                    128:        if (mdoc->meta.os)
                    129:                free(mdoc->meta.os);
                    130:        if (mdoc->meta.name)
                    131:                free(mdoc->meta.name);
                    132:        if (mdoc->meta.arch)
                    133:                free(mdoc->meta.arch);
                    134:        if (mdoc->meta.vol)
                    135:                free(mdoc->meta.vol);
1.50      schwarze  136:        if (mdoc->meta.msec)
                    137:                free(mdoc->meta.msec);
1.82      schwarze  138:        if (mdoc->meta.date)
                    139:                free(mdoc->meta.date);
1.1       kristaps  140: }
                    141:
                    142:
1.8       schwarze  143: /*
                    144:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    145:  */
1.32      schwarze  146: static void
1.1       kristaps  147: mdoc_alloc1(struct mdoc *mdoc)
                    148: {
                    149:
1.32      schwarze  150:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.1       kristaps  151:        mdoc->flags = 0;
1.9       schwarze  152:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.32      schwarze  153:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.1       kristaps  154:        mdoc->first = mdoc->last;
                    155:        mdoc->last->type = MDOC_ROOT;
                    156:        mdoc->next = MDOC_NEXT_CHILD;
                    157: }
                    158:
                    159:
                    160: /*
1.8       schwarze  161:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    162:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    163:  * and the parser is ready for re-invocation on a new tree; however,
                    164:  * cross-parse non-volatile data is kept intact.
1.1       kristaps  165:  */
1.32      schwarze  166: void
1.1       kristaps  167: mdoc_reset(struct mdoc *mdoc)
                    168: {
                    169:
                    170:        mdoc_free1(mdoc);
1.32      schwarze  171:        mdoc_alloc1(mdoc);
1.1       kristaps  172: }
                    173:
                    174:
                    175: /*
1.8       schwarze  176:  * Completely free up all volatile and non-volatile parse resources.
                    177:  * After invocation, the pointer is no longer usable.
1.1       kristaps  178:  */
                    179: void
                    180: mdoc_free(struct mdoc *mdoc)
                    181: {
                    182:
                    183:        mdoc_free1(mdoc);
                    184:        free(mdoc);
                    185: }
                    186:
                    187:
1.8       schwarze  188: /*
                    189:  * Allocate volatile and non-volatile parse resources.
                    190:  */
1.1       kristaps  191: struct mdoc *
1.83    ! schwarze  192: mdoc_alloc(struct regset *regs, struct mparse *parse)
1.1       kristaps  193: {
                    194:        struct mdoc     *p;
                    195:
1.32      schwarze  196:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    197:
1.83    ! schwarze  198:        p->parse = parse;
1.58      schwarze  199:        p->regs = regs;
1.1       kristaps  200:
1.32      schwarze  201:        mdoc_hash_init();
                    202:        mdoc_alloc1(p);
                    203:        return(p);
1.1       kristaps  204: }
                    205:
                    206:
                    207: /*
                    208:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.8       schwarze  209:  * through to macro_end() in macro.c.
1.1       kristaps  210:  */
                    211: int
                    212: mdoc_endparse(struct mdoc *m)
                    213: {
                    214:
1.78      schwarze  215:        assert( ! (MDOC_HALT & m->flags));
                    216:        if (mdoc_macroend(m))
1.1       kristaps  217:                return(1);
                    218:        m->flags |= MDOC_HALT;
                    219:        return(0);
1.81      schwarze  220: }
                    221:
                    222: int
                    223: mdoc_addeqn(struct mdoc *m, const struct eqn *ep)
                    224: {
                    225:        struct mdoc_node *n;
                    226:
                    227:        assert( ! (MDOC_HALT & m->flags));
                    228:
                    229:        /* No text before an initial macro. */
                    230:
                    231:        if (SEC_NONE == m->lastnamed) {
                    232:                mdoc_pmsg(m, ep->line, ep->pos, MANDOCERR_NOTEXT);
                    233:                return(1);
                    234:        }
                    235:
                    236:        n = node_alloc(m, ep->line, ep->pos, MDOC_MAX, MDOC_EQN);
                    237:        n->eqn = ep;
                    238:
                    239:        if ( ! node_append(m, n))
                    240:                return(0);
                    241:
                    242:        m->next = MDOC_NEXT_SIBLING;
                    243:        return(1);
1.1       kristaps  244: }
                    245:
1.77      schwarze  246: int
                    247: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    248: {
1.79      schwarze  249:        struct mdoc_node *n;
1.77      schwarze  250:
1.78      schwarze  251:        assert( ! (MDOC_HALT & m->flags));
1.77      schwarze  252:
                    253:        /* No text before an initial macro. */
                    254:
                    255:        if (SEC_NONE == m->lastnamed) {
1.79      schwarze  256:                mdoc_pmsg(m, sp->line, 0, MANDOCERR_NOTEXT);
1.77      schwarze  257:                return(1);
                    258:        }
                    259:
1.79      schwarze  260:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
                    261:        n->span = sp;
                    262:
                    263:        if ( ! node_append(m, n))
                    264:                return(0);
                    265:
                    266:        m->next = MDOC_NEXT_SIBLING;
                    267:        return(1);
1.77      schwarze  268: }
                    269:
1.1       kristaps  270:
                    271: /*
                    272:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  273:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  274:  */
                    275: int
1.53      schwarze  276: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  277: {
                    278:
1.78      schwarze  279:        assert( ! (MDOC_HALT & m->flags));
1.1       kristaps  280:
1.47      schwarze  281:        m->flags |= MDOC_NEWLINE;
1.60      schwarze  282:
                    283:        /*
                    284:         * Let the roff nS register switch SYNOPSIS mode early,
                    285:         * such that the parser knows at all times
                    286:         * whether this mode is on or off.
                    287:         * Note that this mode is also switched by the Sh macro.
                    288:         */
                    289:        if (m->regs->regs[(int)REG_nS].set) {
                    290:                if (m->regs->regs[(int)REG_nS].v.u)
                    291:                        m->flags |= MDOC_SYNOPSIS;
                    292:                else
                    293:                        m->flags &= ~MDOC_SYNOPSIS;
                    294:        }
                    295:
1.83    ! schwarze  296:        return(mandoc_getcontrol(buf, &offs) ?
1.53      schwarze  297:                        mdoc_pmacro(m, ln, buf, offs) :
                    298:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  299: }
                    300:
1.5       schwarze  301: int
1.58      schwarze  302: mdoc_macro(MACRO_PROT_ARGS)
1.1       kristaps  303: {
1.46      schwarze  304:        assert(tok < MDOC_MAX);
                    305:
                    306:        /* If we're in the body, deny prologue calls. */
1.37      schwarze  307:
1.1       kristaps  308:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.76      schwarze  309:                        MDOC_PBODY & m->flags) {
                    310:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    311:                return(1);
                    312:        }
1.46      schwarze  313:
                    314:        /* If we're in the prologue, deny "body" macros.  */
                    315:
1.1       kristaps  316:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
1.39      schwarze  317:                        ! (MDOC_PBODY & m->flags)) {
1.76      schwarze  318:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.71      schwarze  319:                if (NULL == m->meta.msec)
                    320:                        m->meta.msec = mandoc_strdup("1");
1.39      schwarze  321:                if (NULL == m->meta.title)
1.56      schwarze  322:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.39      schwarze  323:                if (NULL == m->meta.vol)
1.56      schwarze  324:                        m->meta.vol = mandoc_strdup("LOCAL");
1.39      schwarze  325:                if (NULL == m->meta.os)
1.56      schwarze  326:                        m->meta.os = mandoc_strdup("LOCAL");
1.80      schwarze  327:                if (NULL == m->meta.date)
1.83    ! schwarze  328:                        m->meta.date = mandoc_normdate
        !           329:                                (m->parse, NULL, line, ppos);
1.39      schwarze  330:                m->flags |= MDOC_PBODY;
                    331:        }
1.1       kristaps  332:
1.58      schwarze  333:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.1       kristaps  334: }
                    335:
                    336:
                    337: static int
                    338: node_append(struct mdoc *mdoc, struct mdoc_node *p)
                    339: {
                    340:
                    341:        assert(mdoc->last);
                    342:        assert(mdoc->first);
                    343:        assert(MDOC_ROOT != p->type);
                    344:
                    345:        switch (mdoc->next) {
                    346:        case (MDOC_NEXT_SIBLING):
                    347:                mdoc->last->next = p;
                    348:                p->prev = mdoc->last;
                    349:                p->parent = mdoc->last->parent;
                    350:                break;
                    351:        case (MDOC_NEXT_CHILD):
                    352:                mdoc->last->child = p;
                    353:                p->parent = mdoc->last;
                    354:                break;
                    355:        default:
                    356:                abort();
                    357:                /* NOTREACHED */
                    358:        }
                    359:
1.10      schwarze  360:        p->parent->nchild++;
                    361:
1.75      schwarze  362:        /*
                    363:         * Copy over the normalised-data pointer of our parent.  Not
                    364:         * everybody has one, but copying a null pointer is fine.
                    365:         */
                    366:
                    367:        switch (p->type) {
                    368:        case (MDOC_BODY):
                    369:                /* FALLTHROUGH */
                    370:        case (MDOC_TAIL):
                    371:                /* FALLTHROUGH */
                    372:        case (MDOC_HEAD):
                    373:                p->norm = p->parent->norm;
                    374:                break;
                    375:        default:
                    376:                break;
                    377:        }
                    378:
1.1       kristaps  379:        if ( ! mdoc_valid_pre(mdoc, p))
                    380:                return(0);
                    381:
                    382:        switch (p->type) {
                    383:        case (MDOC_HEAD):
                    384:                assert(MDOC_BLOCK == p->parent->type);
                    385:                p->parent->head = p;
                    386:                break;
                    387:        case (MDOC_TAIL):
                    388:                assert(MDOC_BLOCK == p->parent->type);
                    389:                p->parent->tail = p;
                    390:                break;
                    391:        case (MDOC_BODY):
1.59      schwarze  392:                if (p->end)
                    393:                        break;
1.1       kristaps  394:                assert(MDOC_BLOCK == p->parent->type);
                    395:                p->parent->body = p;
                    396:                break;
                    397:        default:
                    398:                break;
                    399:        }
                    400:
                    401:        mdoc->last = p;
                    402:
                    403:        switch (p->type) {
1.77      schwarze  404:        case (MDOC_TBL):
                    405:                /* FALLTHROUGH */
1.1       kristaps  406:        case (MDOC_TEXT):
                    407:                if ( ! mdoc_valid_post(mdoc))
                    408:                        return(0);
                    409:                break;
                    410:        default:
                    411:                break;
                    412:        }
                    413:
                    414:        return(1);
                    415: }
                    416:
                    417:
                    418: static struct mdoc_node *
1.37      schwarze  419: node_alloc(struct mdoc *m, int line, int pos,
                    420:                enum mdoct tok, enum mdoc_type type)
1.1       kristaps  421: {
                    422:        struct mdoc_node *p;
                    423:
1.32      schwarze  424:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.19      schwarze  425:        p->sec = m->lastsec;
1.1       kristaps  426:        p->line = line;
                    427:        p->pos = pos;
                    428:        p->tok = tok;
1.37      schwarze  429:        p->type = type;
1.58      schwarze  430:
                    431:        /* Flag analysis. */
                    432:
1.60      schwarze  433:        if (MDOC_SYNOPSIS & m->flags)
                    434:                p->flags |= MDOC_SYNPRETTY;
                    435:        else
                    436:                p->flags &= ~MDOC_SYNPRETTY;
1.47      schwarze  437:        if (MDOC_NEWLINE & m->flags)
                    438:                p->flags |= MDOC_LINE;
                    439:        m->flags &= ~MDOC_NEWLINE;
1.58      schwarze  440:
1.1       kristaps  441:        return(p);
                    442: }
                    443:
                    444:
                    445: int
1.37      schwarze  446: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  447: {
                    448:        struct mdoc_node *p;
                    449:
1.19      schwarze  450:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.25      schwarze  451:        if ( ! node_append(m, p))
                    452:                return(0);
                    453:        m->next = MDOC_NEXT_CHILD;
                    454:        return(1);
1.1       kristaps  455: }
                    456:
                    457:
                    458: int
1.37      schwarze  459: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  460: {
                    461:        struct mdoc_node *p;
                    462:
1.19      schwarze  463:        assert(m->first);
                    464:        assert(m->last);
1.1       kristaps  465:
1.19      schwarze  466:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
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_body_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:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.25      schwarze  480:        if ( ! node_append(m, p))
                    481:                return(0);
                    482:        m->next = MDOC_NEXT_CHILD;
1.59      schwarze  483:        return(1);
                    484: }
                    485:
                    486:
                    487: int
                    488: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    489:                struct mdoc_node *body, enum mdoc_endbody end)
                    490: {
                    491:        struct mdoc_node *p;
                    492:
                    493:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    494:        p->pending = body;
                    495:        p->end = end;
                    496:        if ( ! node_append(m, p))
                    497:                return(0);
                    498:        m->next = MDOC_NEXT_SIBLING;
1.25      schwarze  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)++;
1.75      schwarze  513:
                    514:        switch (tok) {
                    515:        case (MDOC_Bd):
                    516:                /* FALLTHROUGH */
                    517:        case (MDOC_Bf):
                    518:                /* FALLTHROUGH */
                    519:        case (MDOC_Bl):
                    520:                /* FALLTHROUGH */
                    521:        case (MDOC_Rs):
                    522:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    523:                break;
                    524:        default:
                    525:                break;
                    526:        }
                    527:
1.25      schwarze  528:        if ( ! node_append(m, p))
                    529:                return(0);
                    530:        m->next = MDOC_NEXT_CHILD;
                    531:        return(1);
1.1       kristaps  532: }
                    533:
                    534:
                    535: int
1.19      schwarze  536: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  537:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  538: {
                    539:        struct mdoc_node *p;
                    540:
1.19      schwarze  541:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.4       schwarze  542:        p->args = args;
                    543:        if (p->args)
1.1       kristaps  544:                (args->refcnt)++;
1.75      schwarze  545:
                    546:        switch (tok) {
                    547:        case (MDOC_An):
                    548:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    549:                break;
                    550:        default:
                    551:                break;
                    552:        }
                    553:
1.25      schwarze  554:        if ( ! node_append(m, p))
                    555:                return(0);
                    556:        m->next = MDOC_NEXT_CHILD;
                    557:        return(1);
1.1       kristaps  558: }
                    559:
1.46      schwarze  560: int
                    561: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  562: {
1.19      schwarze  563:        struct mdoc_node *n;
1.46      schwarze  564:        size_t            sv, len;
                    565:
                    566:        len = strlen(p);
1.1       kristaps  567:
1.46      schwarze  568:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.32      schwarze  569:        n->string = mandoc_malloc(len + 1);
1.19      schwarze  570:        sv = strlcpy(n->string, p, len + 1);
                    571:
                    572:        /* Prohibit truncation. */
                    573:        assert(sv < len + 1);
                    574:
1.25      schwarze  575:        if ( ! node_append(m, n))
                    576:                return(0);
1.46      schwarze  577:
1.25      schwarze  578:        m->next = MDOC_NEXT_SIBLING;
                    579:        return(1);
1.19      schwarze  580: }
                    581:
                    582:
1.61      schwarze  583: static void
1.1       kristaps  584: mdoc_node_free(struct mdoc_node *p)
                    585: {
1.61      schwarze  586:
1.75      schwarze  587:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
                    588:                free(p->norm);
1.1       kristaps  589:        if (p->string)
                    590:                free(p->string);
                    591:        if (p->args)
                    592:                mdoc_argv_free(p->args);
                    593:        free(p);
                    594: }
                    595:
                    596:
1.40      schwarze  597: static void
                    598: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    599: {
                    600:
                    601:        /* Adjust siblings. */
                    602:
                    603:        if (n->prev)
                    604:                n->prev->next = n->next;
                    605:        if (n->next)
                    606:                n->next->prev = n->prev;
                    607:
                    608:        /* Adjust parent. */
                    609:
                    610:        if (n->parent) {
                    611:                n->parent->nchild--;
                    612:                if (n->parent->child == n)
                    613:                        n->parent->child = n->prev ? n->prev : n->next;
1.72      schwarze  614:                if (n->parent->last == n)
                    615:                        n->parent->last = n->prev ? n->prev : NULL;
1.40      schwarze  616:        }
                    617:
                    618:        /* Adjust parse point, if applicable. */
                    619:
                    620:        if (m && m->last == n) {
                    621:                if (n->prev) {
                    622:                        m->last = n->prev;
                    623:                        m->next = MDOC_NEXT_SIBLING;
                    624:                } else {
                    625:                        m->last = n->parent;
                    626:                        m->next = MDOC_NEXT_CHILD;
                    627:                }
                    628:        }
                    629:
                    630:        if (m && m->first == n)
                    631:                m->first = NULL;
                    632: }
                    633:
                    634:
1.1       kristaps  635: void
1.40      schwarze  636: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  637: {
                    638:
1.40      schwarze  639:        while (p->child) {
                    640:                assert(p->nchild);
                    641:                mdoc_node_delete(m, p->child);
                    642:        }
                    643:        assert(0 == p->nchild);
1.1       kristaps  644:
1.40      schwarze  645:        mdoc_node_unlink(m, p);
1.1       kristaps  646:        mdoc_node_free(p);
                    647: }
                    648:
                    649:
                    650: /*
                    651:  * Parse free-form text, that is, a line that does not begin with the
                    652:  * control character.
                    653:  */
                    654: static int
1.53      schwarze  655: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  656: {
1.56      schwarze  657:        char             *c, *ws, *end;
                    658:        struct mdoc_node *n;
1.44      schwarze  659:
1.46      schwarze  660:        /* No text before an initial macro. */
                    661:
1.76      schwarze  662:        if (SEC_NONE == m->lastnamed) {
                    663:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    664:                return(1);
                    665:        }
1.46      schwarze  666:
1.56      schwarze  667:        assert(m->last);
                    668:        n = m->last;
                    669:
                    670:        /*
                    671:         * Divert directly to list processing if we're encountering a
                    672:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
                    673:         * (a MDOC_BODY means it's already open, in which case we should
                    674:         * process within its context in the normal way).
                    675:         */
                    676:
                    677:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.75      schwarze  678:                        LIST_column == n->norm->Bl.type) {
1.56      schwarze  679:                /* `Bl' is open without any children. */
                    680:                m->flags |= MDOC_FREECOL;
                    681:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    682:        }
                    683:
                    684:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    685:                        NULL != n->parent &&
                    686:                        MDOC_Bl == n->parent->tok &&
1.75      schwarze  687:                        LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  688:                /* `Bl' has block-level `It' children. */
                    689:                m->flags |= MDOC_FREECOL;
                    690:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    691:        }
                    692:
1.52      schwarze  693:        /*
                    694:         * Search for the beginning of unescaped trailing whitespace (ws)
                    695:         * and for the first character not to be output (end).
                    696:         */
1.56      schwarze  697:
                    698:        /* FIXME: replace with strcspn(). */
1.52      schwarze  699:        ws = NULL;
1.53      schwarze  700:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  701:                switch (*c) {
1.55      schwarze  702:                case '-':
                    703:                        if (mandoc_hyph(buf + offs, c))
                    704:                                *c = ASCII_HYPH;
1.56      schwarze  705:                        ws = NULL;
1.55      schwarze  706:                        break;
1.52      schwarze  707:                case ' ':
                    708:                        if (NULL == ws)
                    709:                                ws = c;
                    710:                        continue;
                    711:                case '\t':
                    712:                        /*
                    713:                         * Always warn about trailing tabs,
                    714:                         * even outside literal context,
                    715:                         * where they should be put on the next line.
                    716:                         */
                    717:                        if (NULL == ws)
                    718:                                ws = c;
                    719:                        /*
                    720:                         * Strip trailing tabs in literal context only;
                    721:                         * outside, they affect the next line.
                    722:                         */
                    723:                        if (MDOC_LITERAL & m->flags)
                    724:                                continue;
                    725:                        break;
                    726:                case '\\':
                    727:                        /* Skip the escaped character, too, if any. */
                    728:                        if (c[1])
                    729:                                c++;
                    730:                        /* FALLTHROUGH */
                    731:                default:
                    732:                        ws = NULL;
                    733:                        break;
                    734:                }
                    735:                end = c + 1;
                    736:        }
                    737:        *end = '\0';
1.19      schwarze  738:
1.52      schwarze  739:        if (ws)
1.76      schwarze  740:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.34      schwarze  741:
1.53      schwarze  742:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.76      schwarze  743:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.46      schwarze  744:
1.40      schwarze  745:                /*
1.66      schwarze  746:                 * Insert a `sp' in the case of a blank line.  Technically,
1.46      schwarze  747:                 * blank lines aren't allowed, but enough manuals assume this
                    748:                 * behaviour that we want to work around it.
1.40      schwarze  749:                 */
1.66      schwarze  750:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.42      schwarze  751:                        return(0);
1.46      schwarze  752:
1.42      schwarze  753:                m->next = MDOC_NEXT_SIBLING;
                    754:                return(1);
1.38      schwarze  755:        }
1.1       kristaps  756:
1.53      schwarze  757:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
1.1       kristaps  758:                return(0);
                    759:
1.52      schwarze  760:        if (MDOC_LITERAL & m->flags)
                    761:                return(1);
                    762:
1.35      schwarze  763:        /*
1.48      schwarze  764:         * End-of-sentence check.  If the last character is an unescaped
                    765:         * EOS character, then flag the node as being the end of a
                    766:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  767:         */
                    768:
1.52      schwarze  769:        assert(buf < end);
1.48      schwarze  770:
1.62      schwarze  771:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.48      schwarze  772:                m->last->flags |= MDOC_EOS;
                    773:
1.1       kristaps  774:        return(1);
                    775: }
1.19      schwarze  776:
1.1       kristaps  777:
                    778: /*
                    779:  * Parse a macro line, that is, a line beginning with the control
                    780:  * character.
                    781:  */
1.61      schwarze  782: static int
1.53      schwarze  783: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  784: {
1.56      schwarze  785:        enum mdoct        tok;
1.83    ! schwarze  786:        int               i, sv;
1.56      schwarze  787:        char              mac[5];
                    788:        struct mdoc_node *n;
1.1       kristaps  789:
1.83    ! schwarze  790:        /* Empty post-control lines are ignored. */
1.1       kristaps  791:
1.83    ! schwarze  792:        if ('"' == buf[offs]) {
        !           793:                mdoc_pmsg(m, ln, offs, MANDOCERR_BADCOMMENT);
        !           794:                return(1);
        !           795:        } else if ('\0' == buf[offs])
1.1       kristaps  796:                return(1);
                    797:
1.83    ! schwarze  798:        sv = offs;
1.47      schwarze  799:
1.63      schwarze  800:        /*
1.64      schwarze  801:         * Copy the first word into a nil-terminated buffer.
                    802:         * Stop copying when a tab, space, or eoln is encountered.
1.63      schwarze  803:         */
1.1       kristaps  804:
1.83    ! schwarze  805:        i = 0;
        !           806:        while (i < 4 && '\0' != buf[offs] &&
        !           807:                        ' ' != buf[offs] && '\t' != buf[offs])
        !           808:                mac[i++] = buf[offs++];
        !           809:
        !           810:        mac[i] = '\0';
        !           811:
        !           812:        tok = (i > 1 || i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  813:
1.65      schwarze  814:        if (MDOC_MAX == tok) {
1.83    ! schwarze  815:                mandoc_vmsg(MANDOCERR_MACRO, m->parse,
        !           816:                                ln, sv, "%s", buf + sv - 1);
1.1       kristaps  817:                return(1);
                    818:        }
                    819:
1.63      schwarze  820:        /* Disregard the first trailing tab, if applicable. */
                    821:
1.83    ! schwarze  822:        if ('\t' == buf[offs])
        !           823:                offs++;
1.63      schwarze  824:
                    825:        /* Jump to the next non-whitespace word. */
1.1       kristaps  826:
1.83    ! schwarze  827:        while (buf[offs] && ' ' == buf[offs])
        !           828:                offs++;
1.34      schwarze  829:
1.46      schwarze  830:        /*
                    831:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    832:         * into the parser as "text", so we only warn about spaces here.
                    833:         */
1.34      schwarze  834:
1.83    ! schwarze  835:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
        !           836:                mdoc_pmsg(m, ln, offs - 1, MANDOCERR_EOLNSPACE);
1.1       kristaps  837:
1.56      schwarze  838:        /*
                    839:         * If an initial macro or a list invocation, divert directly
                    840:         * into macro processing.
                    841:         */
                    842:
                    843:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.83    ! schwarze  844:                if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.56      schwarze  845:                        goto err;
                    846:                return(1);
                    847:        }
                    848:
                    849:        n = m->last;
                    850:        assert(m->last);
                    851:
                    852:        /*
                    853:         * If the first macro of a `Bl -column', open an `It' block
                    854:         * context around the parsed macro.
                    855:         */
                    856:
                    857:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.75      schwarze  858:                        LIST_column == n->norm->Bl.type) {
1.56      schwarze  859:                m->flags |= MDOC_FREECOL;
1.58      schwarze  860:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.56      schwarze  861:                        goto err;
                    862:                return(1);
                    863:        }
                    864:
                    865:        /*
                    866:         * If we're following a block-level `It' within a `Bl -column'
                    867:         * context (perhaps opened in the above block or in ptext()),
                    868:         * then open an `It' block context around the parsed macro.
1.24      schwarze  869:         */
1.56      schwarze  870:
                    871:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    872:                        NULL != n->parent &&
                    873:                        MDOC_Bl == n->parent->tok &&
1.75      schwarze  874:                        LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  875:                m->flags |= MDOC_FREECOL;
                    876:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
                    877:                        goto err;
                    878:                return(1);
                    879:        }
                    880:
                    881:        /* Normal processing of a macro. */
                    882:
1.83    ! schwarze  883:        if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.1       kristaps  884:                goto err;
                    885:
                    886:        return(1);
                    887:
                    888: err:   /* Error out. */
                    889:
                    890:        m->flags |= MDOC_HALT;
                    891:        return(0);
                    892: }
1.24      schwarze  893:
1.83    ! schwarze  894: enum mdelim
        !           895: mdoc_isdelim(const char *p)
        !           896: {
        !           897:
        !           898:        if ('\0' == p[0])
        !           899:                return(DELIM_NONE);
        !           900:
        !           901:        if ('\0' == p[1])
        !           902:                switch (p[0]) {
        !           903:                case('('):
        !           904:                        /* FALLTHROUGH */
        !           905:                case('['):
        !           906:                        return(DELIM_OPEN);
        !           907:                case('|'):
        !           908:                        return(DELIM_MIDDLE);
        !           909:                case('.'):
        !           910:                        /* FALLTHROUGH */
        !           911:                case(','):
        !           912:                        /* FALLTHROUGH */
        !           913:                case(';'):
        !           914:                        /* FALLTHROUGH */
        !           915:                case(':'):
        !           916:                        /* FALLTHROUGH */
        !           917:                case('?'):
        !           918:                        /* FALLTHROUGH */
        !           919:                case('!'):
        !           920:                        /* FALLTHROUGH */
        !           921:                case(')'):
        !           922:                        /* FALLTHROUGH */
        !           923:                case(']'):
        !           924:                        return(DELIM_CLOSE);
        !           925:                default:
        !           926:                        return(DELIM_NONE);
        !           927:                }
        !           928:
        !           929:        if ('\\' != p[0])
        !           930:                return(DELIM_NONE);
1.24      schwarze  931:
1.83    ! schwarze  932:        if (0 == strcmp(p + 1, "."))
        !           933:                return(DELIM_CLOSE);
        !           934:        if (0 == strcmp(p + 1, "*(Ba"))
        !           935:                return(DELIM_MIDDLE);
        !           936:
        !           937:        return(DELIM_NONE);
        !           938: }