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

Annotation of mandoc/mdoc.c, Revision 1.187

1.187   ! kristaps    1: /*     $Id: mdoc.c,v 1.186 2011/03/22 14:05:45 kristaps Exp $ */
1.1       kristaps    2: /*
1.182     schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.157     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
1.75      kristaps    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:  *
1.75      kristaps   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:  */
1.114     kristaps   18: #ifdef HAVE_CONFIG_H
                     19: #include "config.h"
                     20: #endif
                     21:
1.106     kristaps   22: #include <sys/types.h>
                     23:
1.1       kristaps   24: #include <assert.h>
                     25: #include <stdarg.h>
1.73      kristaps   26: #include <stdio.h>
1.1       kristaps   27: #include <stdlib.h>
                     28: #include <string.h>
1.120     kristaps   29: #include <time.h>
1.1       kristaps   30:
1.187   ! kristaps   31: #include "mdoc.h"
1.136     kristaps   32: #include "mandoc.h"
1.70      kristaps   33: #include "libmdoc.h"
1.113     kristaps   34: #include "libmandoc.h"
1.1       kristaps   35:
                     36: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.82      kristaps   37:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   38:        "Sh",           "Ss",           "Pp",           "D1",
                     39:        "Dl",           "Bd",           "Ed",           "Bl",
                     40:        "El",           "It",           "Ad",           "An",
                     41:        "Ar",           "Cd",           "Cm",           "Dv",
                     42:        "Er",           "Ev",           "Ex",           "Fa",
                     43:        "Fd",           "Fl",           "Fn",           "Ft",
                     44:        "Ic",           "In",           "Li",           "Nd",
                     45:        "Nm",           "Op",           "Ot",           "Pa",
                     46:        "Rv",           "St",           "Va",           "Vt",
                     47:        /* LINTED */
1.114     kristaps   48:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   49:        /* LINTED */
1.114     kristaps   50:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps   51:        /* LINTED */
1.114     kristaps   52:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   53:        "Ac",           "Ao",           "Aq",           "At",
                     54:        "Bc",           "Bf",           "Bo",           "Bq",
                     55:        "Bsx",          "Bx",           "Db",           "Dc",
                     56:        "Do",           "Dq",           "Ec",           "Ef",
                     57:        "Em",           "Eo",           "Fx",           "Ms",
                     58:        "No",           "Ns",           "Nx",           "Ox",
                     59:        "Pc",           "Pf",           "Po",           "Pq",
                     60:        "Qc",           "Ql",           "Qo",           "Qq",
                     61:        "Re",           "Rs",           "Sc",           "So",
                     62:        "Sq",           "Sm",           "Sx",           "Sy",
                     63:        "Tn",           "Ux",           "Xc",           "Xo",
                     64:        "Fo",           "Fc",           "Oo",           "Oc",
                     65:        "Bk",           "Ek",           "Bt",           "Hf",
1.82      kristaps   66:        "Fr",           "Ud",           "Lb",           "Lp",
                     67:        "Lk",           "Mt",           "Brq",          "Bro",
1.64      kristaps   68:        /* LINTED */
1.114     kristaps   69:        "Brc",          "%C",           "Es",           "En",
1.69      kristaps   70:        /* LINTED */
1.114     kristaps   71:        "Dx",           "%Q",           "br",           "sp",
1.110     kristaps   72:        /* LINTED */
1.141     kristaps   73:        "%U",           "Ta"
1.1       kristaps   74:        };
                     75:
                     76: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                     77:        "split",                "nosplit",              "ragged",
                     78:        "unfilled",             "literal",              "file",
                     79:        "offset",               "bullet",               "dash",
                     80:        "hyphen",               "item",                 "enum",
                     81:        "tag",                  "diag",                 "hang",
                     82:        "ohang",                "inset",                "column",
                     83:        "width",                "compact",              "std",
1.52      kristaps   84:        "filled",               "words",                "emphasis",
1.108     kristaps   85:        "symbolic",             "nested",               "centered"
1.1       kristaps   86:        };
                     87:
                     88: const  char * const *mdoc_macronames = __mdoc_macronames;
                     89: const  char * const *mdoc_argnames = __mdoc_argnames;
                     90:
1.121     kristaps   91: static void              mdoc_node_free(struct mdoc_node *);
                     92: static void              mdoc_node_unlink(struct mdoc *,
                     93:                                struct mdoc_node *);
1.73      kristaps   94: static void              mdoc_free1(struct mdoc *);
1.113     kristaps   95: static void              mdoc_alloc1(struct mdoc *);
1.73      kristaps   96: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.117     kristaps   97:                                enum mdoct, enum mdoc_type);
1.73      kristaps   98: static int               node_append(struct mdoc *,
1.71      kristaps   99:                                struct mdoc_node *);
1.149     kristaps  100: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                    101: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.88      kristaps  102:
1.1       kristaps  103: const struct mdoc_node *
1.71      kristaps  104: mdoc_node(const struct mdoc *m)
1.1       kristaps  105: {
                    106:
1.177     kristaps  107:        assert( ! (MDOC_HALT & m->flags));
                    108:        return(m->first);
1.1       kristaps  109: }
                    110:
                    111:
1.37      kristaps  112: const struct mdoc_meta *
1.71      kristaps  113: mdoc_meta(const struct mdoc *m)
1.37      kristaps  114: {
                    115:
1.177     kristaps  116:        assert( ! (MDOC_HALT & m->flags));
                    117:        return(&m->meta);
1.37      kristaps  118: }
                    119:
                    120:
1.85      kristaps  121: /*
                    122:  * Frees volatile resources (parse tree, meta-data, fields).
                    123:  */
1.73      kristaps  124: static void
                    125: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  126: {
                    127:
                    128:        if (mdoc->first)
1.121     kristaps  129:                mdoc_node_delete(mdoc, mdoc->first);
1.67      kristaps  130:        if (mdoc->meta.title)
                    131:                free(mdoc->meta.title);
                    132:        if (mdoc->meta.os)
                    133:                free(mdoc->meta.os);
                    134:        if (mdoc->meta.name)
                    135:                free(mdoc->meta.name);
                    136:        if (mdoc->meta.arch)
                    137:                free(mdoc->meta.arch);
                    138:        if (mdoc->meta.vol)
                    139:                free(mdoc->meta.vol);
1.133     kristaps  140:        if (mdoc->meta.msec)
                    141:                free(mdoc->meta.msec);
1.183     kristaps  142:        if (mdoc->meta.date)
                    143:                free(mdoc->meta.date);
1.73      kristaps  144: }
                    145:
                    146:
1.85      kristaps  147: /*
                    148:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    149:  */
1.113     kristaps  150: static void
1.73      kristaps  151: mdoc_alloc1(struct mdoc *mdoc)
                    152: {
1.67      kristaps  153:
1.112     kristaps  154:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  155:        mdoc->flags = 0;
1.85      kristaps  156:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  157:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  158:        mdoc->first = mdoc->last;
1.67      kristaps  159:        mdoc->last->type = MDOC_ROOT;
                    160:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  161: }
                    162:
                    163:
                    164: /*
1.85      kristaps  165:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    166:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    167:  * and the parser is ready for re-invocation on a new tree; however,
                    168:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  169:  */
1.113     kristaps  170: void
1.73      kristaps  171: mdoc_reset(struct mdoc *mdoc)
                    172: {
                    173:
                    174:        mdoc_free1(mdoc);
1.113     kristaps  175:        mdoc_alloc1(mdoc);
1.67      kristaps  176: }
                    177:
                    178:
1.68      kristaps  179: /*
1.85      kristaps  180:  * Completely free up all volatile and non-volatile parse resources.
                    181:  * After invocation, the pointer is no longer usable.
1.68      kristaps  182:  */
1.67      kristaps  183: void
1.38      kristaps  184: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  185: {
                    186:
1.73      kristaps  187:        mdoc_free1(mdoc);
1.1       kristaps  188:        free(mdoc);
                    189: }
                    190:
                    191:
1.85      kristaps  192: /*
                    193:  * Allocate volatile and non-volatile parse resources.
                    194:  */
1.1       kristaps  195: struct mdoc *
1.185     kristaps  196: mdoc_alloc(struct regset *regs, struct mparse *parse)
1.1       kristaps  197: {
                    198:        struct mdoc     *p;
                    199:
1.113     kristaps  200:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    201:
1.185     kristaps  202:        p->parse = parse;
1.149     kristaps  203:        p->regs = regs;
1.73      kristaps  204:
1.113     kristaps  205:        mdoc_hash_init();
                    206:        mdoc_alloc1(p);
                    207:        return(p);
1.1       kristaps  208: }
                    209:
                    210:
1.68      kristaps  211: /*
                    212:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  213:  * through to macro_end() in macro.c.
1.68      kristaps  214:  */
1.1       kristaps  215: int
1.72      kristaps  216: mdoc_endparse(struct mdoc *m)
1.20      kristaps  217: {
                    218:
1.177     kristaps  219:        assert( ! (MDOC_HALT & m->flags));
                    220:        if (mdoc_macroend(m))
1.20      kristaps  221:                return(1);
1.72      kristaps  222:        m->flags |= MDOC_HALT;
                    223:        return(0);
1.181     kristaps  224: }
                    225:
                    226: int
                    227: mdoc_addeqn(struct mdoc *m, const struct eqn *ep)
                    228: {
                    229:        struct mdoc_node *n;
                    230:
                    231:        assert( ! (MDOC_HALT & m->flags));
                    232:
                    233:        /* No text before an initial macro. */
                    234:
                    235:        if (SEC_NONE == m->lastnamed) {
                    236:                mdoc_pmsg(m, ep->line, ep->pos, MANDOCERR_NOTEXT);
                    237:                return(1);
                    238:        }
                    239:
                    240:        n = node_alloc(m, ep->line, ep->pos, MDOC_MAX, MDOC_EQN);
                    241:        n->eqn = ep;
                    242:
                    243:        if ( ! node_append(m, n))
                    244:                return(0);
                    245:
                    246:        m->next = MDOC_NEXT_SIBLING;
                    247:        return(1);
1.20      kristaps  248: }
                    249:
1.175     kristaps  250: int
                    251: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    252: {
1.180     kristaps  253:        struct mdoc_node *n;
1.175     kristaps  254:
1.177     kristaps  255:        assert( ! (MDOC_HALT & m->flags));
1.175     kristaps  256:
                    257:        /* No text before an initial macro. */
                    258:
                    259:        if (SEC_NONE == m->lastnamed) {
1.179     kristaps  260:                mdoc_pmsg(m, sp->line, 0, MANDOCERR_NOTEXT);
1.175     kristaps  261:                return(1);
                    262:        }
                    263:
1.180     kristaps  264:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
                    265:        n->span = sp;
                    266:
                    267:        if ( ! node_append(m, n))
                    268:                return(0);
                    269:
                    270:        m->next = MDOC_NEXT_SIBLING;
                    271:        return(1);
1.175     kristaps  272: }
                    273:
1.20      kristaps  274:
1.50      kristaps  275: /*
1.53      kristaps  276:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  277:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  278:  */
1.20      kristaps  279: int
1.149     kristaps  280: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  281: {
                    282:
1.177     kristaps  283:        assert( ! (MDOC_HALT & m->flags));
1.50      kristaps  284:
1.130     kristaps  285:        m->flags |= MDOC_NEWLINE;
1.153     schwarze  286:
                    287:        /*
                    288:         * Let the roff nS register switch SYNOPSIS mode early,
                    289:         * such that the parser knows at all times
                    290:         * whether this mode is on or off.
                    291:         * Note that this mode is also switched by the Sh macro.
                    292:         */
                    293:        if (m->regs->regs[(int)REG_nS].set) {
                    294:                if (m->regs->regs[(int)REG_nS].v.u)
                    295:                        m->flags |= MDOC_SYNOPSIS;
                    296:                else
                    297:                        m->flags &= ~MDOC_SYNOPSIS;
                    298:        }
                    299:
1.135     kristaps  300:        return(('.' == buf[offs] || '\'' == buf[offs]) ?
1.149     kristaps  301:                        mdoc_pmacro(m, ln, buf, offs) :
                    302:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  303: }
                    304:
1.88      kristaps  305: int
1.148     kristaps  306: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  307: {
1.122     kristaps  308:        assert(tok < MDOC_MAX);
                    309:
                    310:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  311:
1.88      kristaps  312:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.174     kristaps  313:                        MDOC_PBODY & m->flags) {
                    314:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    315:                return(1);
                    316:        }
1.122     kristaps  317:
                    318:        /* If we're in the prologue, deny "body" macros.  */
                    319:
                    320:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    321:                        ! (MDOC_PBODY & m->flags)) {
1.174     kristaps  322:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.167     kristaps  323:                if (NULL == m->meta.msec)
                    324:                        m->meta.msec = mandoc_strdup("1");
1.120     kristaps  325:                if (NULL == m->meta.title)
1.140     kristaps  326:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.120     kristaps  327:                if (NULL == m->meta.vol)
1.140     kristaps  328:                        m->meta.vol = mandoc_strdup("LOCAL");
1.120     kristaps  329:                if (NULL == m->meta.os)
1.140     kristaps  330:                        m->meta.os = mandoc_strdup("LOCAL");
1.182     schwarze  331:                if (NULL == m->meta.date)
1.185     kristaps  332:                        m->meta.date = mandoc_normdate
                    333:                                (m->parse, NULL, line, ppos);
1.120     kristaps  334:                m->flags |= MDOC_PBODY;
                    335:        }
1.88      kristaps  336:
1.149     kristaps  337:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.73      kristaps  338: }
                    339:
                    340:
                    341: static int
                    342: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  343: {
                    344:
1.25      kristaps  345:        assert(mdoc->last);
                    346:        assert(mdoc->first);
                    347:        assert(MDOC_ROOT != p->type);
1.1       kristaps  348:
1.13      kristaps  349:        switch (mdoc->next) {
                    350:        case (MDOC_NEXT_SIBLING):
1.6       kristaps  351:                mdoc->last->next = p;
                    352:                p->prev = mdoc->last;
1.13      kristaps  353:                p->parent = mdoc->last->parent;
1.1       kristaps  354:                break;
1.13      kristaps  355:        case (MDOC_NEXT_CHILD):
                    356:                mdoc->last->child = p;
1.1       kristaps  357:                p->parent = mdoc->last;
                    358:                break;
                    359:        default:
1.13      kristaps  360:                abort();
                    361:                /* NOTREACHED */
1.1       kristaps  362:        }
                    363:
1.86      kristaps  364:        p->parent->nchild++;
                    365:
1.172     kristaps  366:        /*
                    367:         * Copy over the normalised-data pointer of our parent.  Not
                    368:         * everybody has one, but copying a null pointer is fine.
                    369:         */
                    370:
                    371:        switch (p->type) {
                    372:        case (MDOC_BODY):
                    373:                /* FALLTHROUGH */
                    374:        case (MDOC_TAIL):
                    375:                /* FALLTHROUGH */
                    376:        case (MDOC_HEAD):
                    377:                p->norm = p->parent->norm;
                    378:                break;
                    379:        default:
                    380:                break;
                    381:        }
                    382:
1.23      kristaps  383:        if ( ! mdoc_valid_pre(mdoc, p))
                    384:                return(0);
1.27      kristaps  385:
                    386:        switch (p->type) {
                    387:        case (MDOC_HEAD):
                    388:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  389:                p->parent->head = p;
1.27      kristaps  390:                break;
                    391:        case (MDOC_TAIL):
                    392:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  393:                p->parent->tail = p;
1.27      kristaps  394:                break;
                    395:        case (MDOC_BODY):
1.152     schwarze  396:                if (p->end)
                    397:                        break;
1.27      kristaps  398:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  399:                p->parent->body = p;
1.27      kristaps  400:                break;
                    401:        default:
                    402:                break;
                    403:        }
                    404:
1.1       kristaps  405:        mdoc->last = p;
1.71      kristaps  406:
                    407:        switch (p->type) {
1.176     kristaps  408:        case (MDOC_TBL):
                    409:                /* FALLTHROUGH */
1.71      kristaps  410:        case (MDOC_TEXT):
                    411:                if ( ! mdoc_valid_post(mdoc))
                    412:                        return(0);
                    413:                break;
                    414:        default:
                    415:                break;
                    416:        }
                    417:
1.23      kristaps  418:        return(1);
1.1       kristaps  419: }
                    420:
                    421:
1.45      kristaps  422: static struct mdoc_node *
1.117     kristaps  423: node_alloc(struct mdoc *m, int line, int pos,
                    424:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  425: {
1.46      kristaps  426:        struct mdoc_node *p;
                    427:
1.113     kristaps  428:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.91      kristaps  429:        p->sec = m->lastsec;
1.73      kristaps  430:        p->line = line;
                    431:        p->pos = pos;
                    432:        p->tok = tok;
1.118     kristaps  433:        p->type = type;
1.150     kristaps  434:
                    435:        /* Flag analysis. */
                    436:
1.153     schwarze  437:        if (MDOC_SYNOPSIS & m->flags)
                    438:                p->flags |= MDOC_SYNPRETTY;
                    439:        else
                    440:                p->flags &= ~MDOC_SYNPRETTY;
1.130     kristaps  441:        if (MDOC_NEWLINE & m->flags)
                    442:                p->flags |= MDOC_LINE;
                    443:        m->flags &= ~MDOC_NEWLINE;
1.150     kristaps  444:
1.46      kristaps  445:        return(p);
1.45      kristaps  446: }
                    447:
                    448:
1.23      kristaps  449: int
1.117     kristaps  450: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.17      kristaps  451: {
                    452:        struct mdoc_node *p;
                    453:
1.91      kristaps  454:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.102     kristaps  455:        if ( ! node_append(m, p))
                    456:                return(0);
                    457:        m->next = MDOC_NEXT_CHILD;
                    458:        return(1);
1.17      kristaps  459: }
                    460:
                    461:
1.23      kristaps  462: int
1.117     kristaps  463: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  464: {
                    465:        struct mdoc_node *p;
                    466:
1.91      kristaps  467:        assert(m->first);
                    468:        assert(m->last);
1.1       kristaps  469:
1.91      kristaps  470:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.102     kristaps  471:        if ( ! node_append(m, p))
                    472:                return(0);
                    473:        m->next = MDOC_NEXT_CHILD;
                    474:        return(1);
1.1       kristaps  475: }
                    476:
                    477:
1.23      kristaps  478: int
1.117     kristaps  479: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  480: {
                    481:        struct mdoc_node *p;
                    482:
1.91      kristaps  483:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.102     kristaps  484:        if ( ! node_append(m, p))
                    485:                return(0);
                    486:        m->next = MDOC_NEXT_CHILD;
1.152     schwarze  487:        return(1);
                    488: }
                    489:
                    490:
                    491: int
                    492: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    493:                struct mdoc_node *body, enum mdoc_endbody end)
                    494: {
                    495:        struct mdoc_node *p;
                    496:
                    497:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    498:        p->pending = body;
                    499:        p->end = end;
                    500:        if ( ! node_append(m, p))
                    501:                return(0);
                    502:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  503:        return(1);
1.1       kristaps  504: }
                    505:
                    506:
1.23      kristaps  507: int
1.91      kristaps  508: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  509:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  510: {
                    511:        struct mdoc_node *p;
                    512:
1.91      kristaps  513:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  514:        p->args = args;
                    515:        if (p->args)
1.53      kristaps  516:                (args->refcnt)++;
1.172     kristaps  517:
                    518:        switch (tok) {
                    519:        case (MDOC_Bd):
                    520:                /* FALLTHROUGH */
                    521:        case (MDOC_Bf):
                    522:                /* FALLTHROUGH */
                    523:        case (MDOC_Bl):
1.173     kristaps  524:                /* FALLTHROUGH */
                    525:        case (MDOC_Rs):
1.172     kristaps  526:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    527:                break;
                    528:        default:
                    529:                break;
                    530:        }
                    531:
1.102     kristaps  532:        if ( ! node_append(m, p))
                    533:                return(0);
                    534:        m->next = MDOC_NEXT_CHILD;
                    535:        return(1);
1.1       kristaps  536: }
                    537:
                    538:
1.23      kristaps  539: int
1.91      kristaps  540: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  541:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  542: {
                    543:        struct mdoc_node *p;
                    544:
1.91      kristaps  545:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.77      kristaps  546:        p->args = args;
                    547:        if (p->args)
1.53      kristaps  548:                (args->refcnt)++;
1.172     kristaps  549:
                    550:        switch (tok) {
                    551:        case (MDOC_An):
                    552:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    553:                break;
                    554:        default:
                    555:                break;
                    556:        }
                    557:
1.102     kristaps  558:        if ( ! node_append(m, p))
                    559:                return(0);
                    560:        m->next = MDOC_NEXT_CHILD;
1.175     kristaps  561:        return(1);
                    562: }
1.1       kristaps  563:
1.124     kristaps  564: int
                    565: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  566: {
1.91      kristaps  567:        struct mdoc_node *n;
1.124     kristaps  568:        size_t            sv, len;
                    569:
                    570:        len = strlen(p);
1.1       kristaps  571:
1.125     kristaps  572:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.113     kristaps  573:        n->string = mandoc_malloc(len + 1);
1.91      kristaps  574:        sv = strlcpy(n->string, p, len + 1);
                    575:
                    576:        /* Prohibit truncation. */
                    577:        assert(sv < len + 1);
                    578:
1.101     kristaps  579:        if ( ! node_append(m, n))
                    580:                return(0);
1.124     kristaps  581:
1.101     kristaps  582:        m->next = MDOC_NEXT_SIBLING;
                    583:        return(1);
1.91      kristaps  584: }
                    585:
                    586:
1.155     kristaps  587: static void
1.53      kristaps  588: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  589: {
                    590:
1.172     kristaps  591:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  592:                free(p->norm);
1.53      kristaps  593:        if (p->string)
                    594:                free(p->string);
                    595:        if (p->args)
                    596:                mdoc_argv_free(p->args);
1.1       kristaps  597:        free(p);
                    598: }
                    599:
                    600:
1.121     kristaps  601: static void
                    602: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    603: {
                    604:
                    605:        /* Adjust siblings. */
                    606:
                    607:        if (n->prev)
                    608:                n->prev->next = n->next;
                    609:        if (n->next)
                    610:                n->next->prev = n->prev;
                    611:
                    612:        /* Adjust parent. */
                    613:
                    614:        if (n->parent) {
                    615:                n->parent->nchild--;
                    616:                if (n->parent->child == n)
                    617:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  618:                if (n->parent->last == n)
                    619:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  620:        }
                    621:
                    622:        /* Adjust parse point, if applicable. */
                    623:
                    624:        if (m && m->last == n) {
                    625:                if (n->prev) {
                    626:                        m->last = n->prev;
                    627:                        m->next = MDOC_NEXT_SIBLING;
                    628:                } else {
                    629:                        m->last = n->parent;
                    630:                        m->next = MDOC_NEXT_CHILD;
                    631:                }
                    632:        }
                    633:
                    634:        if (m && m->first == n)
                    635:                m->first = NULL;
                    636: }
                    637:
                    638:
1.53      kristaps  639: void
1.121     kristaps  640: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  641: {
                    642:
1.121     kristaps  643:        while (p->child) {
                    644:                assert(p->nchild);
                    645:                mdoc_node_delete(m, p->child);
                    646:        }
                    647:        assert(0 == p->nchild);
1.1       kristaps  648:
1.121     kristaps  649:        mdoc_node_unlink(m, p);
1.53      kristaps  650:        mdoc_node_free(p);
1.1       kristaps  651: }
                    652:
                    653:
1.53      kristaps  654: /*
                    655:  * Parse free-form text, that is, a line that does not begin with the
                    656:  * control character.
                    657:  */
                    658: static int
1.149     kristaps  659: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  660: {
1.142     kristaps  661:        char             *c, *ws, *end;
                    662:        struct mdoc_node *n;
1.1       kristaps  663:
1.123     kristaps  664:        /* Ignore bogus comments. */
                    665:
1.135     kristaps  666:        if ('\\' == buf[offs] &&
                    667:                        '.' == buf[offs + 1] &&
1.174     kristaps  668:                        '"' == buf[offs + 2]) {
                    669:                mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT);
                    670:                return(1);
                    671:        }
1.123     kristaps  672:
1.124     kristaps  673:        /* No text before an initial macro. */
                    674:
1.174     kristaps  675:        if (SEC_NONE == m->lastnamed) {
                    676:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    677:                return(1);
                    678:        }
1.142     kristaps  679:
                    680:        assert(m->last);
                    681:        n = m->last;
                    682:
                    683:        /*
1.144     kristaps  684:         * Divert directly to list processing if we're encountering a
1.142     kristaps  685:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  686:         * (a MDOC_BODY means it's already open, in which case we should
                    687:         * process within its context in the normal way).
1.142     kristaps  688:         */
                    689:
1.143     kristaps  690:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  691:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  692:                /* `Bl' is open without any children. */
1.142     kristaps  693:                m->flags |= MDOC_FREECOL;
1.149     kristaps  694:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  695:        }
                    696:
                    697:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    698:                        NULL != n->parent &&
                    699:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  700:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  701:                /* `Bl' has block-level `It' children. */
1.142     kristaps  702:                m->flags |= MDOC_FREECOL;
1.149     kristaps  703:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  704:        }
1.124     kristaps  705:
1.137     schwarze  706:        /*
                    707:         * Search for the beginning of unescaped trailing whitespace (ws)
                    708:         * and for the first character not to be output (end).
                    709:         */
1.139     kristaps  710:
                    711:        /* FIXME: replace with strcspn(). */
1.137     schwarze  712:        ws = NULL;
                    713:        for (c = end = buf + offs; *c; c++) {
                    714:                switch (*c) {
1.138     kristaps  715:                case '-':
                    716:                        if (mandoc_hyph(buf + offs, c))
                    717:                                *c = ASCII_HYPH;
1.145     kristaps  718:                        ws = NULL;
1.138     kristaps  719:                        break;
1.137     schwarze  720:                case ' ':
                    721:                        if (NULL == ws)
                    722:                                ws = c;
                    723:                        continue;
                    724:                case '\t':
                    725:                        /*
                    726:                         * Always warn about trailing tabs,
                    727:                         * even outside literal context,
                    728:                         * where they should be put on the next line.
                    729:                         */
                    730:                        if (NULL == ws)
                    731:                                ws = c;
                    732:                        /*
                    733:                         * Strip trailing tabs in literal context only;
                    734:                         * outside, they affect the next line.
                    735:                         */
                    736:                        if (MDOC_LITERAL & m->flags)
                    737:                                continue;
                    738:                        break;
                    739:                case '\\':
                    740:                        /* Skip the escaped character, too, if any. */
                    741:                        if (c[1])
                    742:                                c++;
                    743:                        /* FALLTHROUGH */
                    744:                default:
                    745:                        ws = NULL;
                    746:                        break;
                    747:                }
                    748:                end = c + 1;
                    749:        }
                    750:        *end = '\0';
1.91      kristaps  751:
1.137     schwarze  752:        if (ws)
1.174     kristaps  753:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.115     kristaps  754:
1.137     schwarze  755:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.174     kristaps  756:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.124     kristaps  757:
1.119     kristaps  758:                /*
1.165     schwarze  759:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  760:                 * blank lines aren't allowed, but enough manuals assume this
                    761:                 * behaviour that we want to work around it.
1.119     kristaps  762:                 */
1.165     schwarze  763:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.119     kristaps  764:                        return(0);
1.124     kristaps  765:
                    766:                m->next = MDOC_NEXT_SIBLING;
                    767:                return(1);
1.119     kristaps  768:        }
1.68      kristaps  769:
1.137     schwarze  770:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
                    771:                return(0);
1.91      kristaps  772:
1.137     schwarze  773:        if (MDOC_LITERAL & m->flags)
                    774:                return(1);
1.128     kristaps  775:
                    776:        /*
                    777:         * End-of-sentence check.  If the last character is an unescaped
                    778:         * EOS character, then flag the node as being the end of a
                    779:         * sentence.  The front-end will know how to interpret this.
                    780:         */
1.132     kristaps  781:
1.137     schwarze  782:        assert(buf < end);
                    783:
1.159     schwarze  784:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.128     kristaps  785:                m->last->flags |= MDOC_EOS;
                    786:
                    787:        return(1);
1.1       kristaps  788: }
                    789:
                    790:
1.53      kristaps  791: /*
                    792:  * Parse a macro line, that is, a line beginning with the control
                    793:  * character.
                    794:  */
1.155     kristaps  795: static int
1.149     kristaps  796: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  797: {
1.144     kristaps  798:        enum mdoct        tok;
                    799:        int               i, j, sv;
                    800:        char              mac[5];
                    801:        struct mdoc_node *n;
1.1       kristaps  802:
1.81      kristaps  803:        /* Empty lines are ignored. */
1.63      kristaps  804:
1.135     kristaps  805:        offs++;
                    806:
                    807:        if ('\0' == buf[offs])
1.63      kristaps  808:                return(1);
                    809:
1.135     kristaps  810:        i = offs;
1.100     kristaps  811:
1.160     kristaps  812:        /* Accept tabs/whitespace after the initial control char. */
1.100     kristaps  813:
1.160     kristaps  814:        if (' ' == buf[i] || '\t' == buf[i]) {
1.100     kristaps  815:                i++;
1.160     kristaps  816:                while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
1.63      kristaps  817:                        i++;
1.115     kristaps  818:                if ('\0' == buf[i])
1.63      kristaps  819:                        return(1);
                    820:        }
1.1       kristaps  821:
1.130     kristaps  822:        sv = i;
                    823:
1.160     kristaps  824:        /*
1.162     schwarze  825:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  826:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  827:         */
1.1       kristaps  828:
1.162     schwarze  829:        j = 0;
1.165     schwarze  830:        while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
1.162     schwarze  831:                mac[j++] = buf[i++];
1.135     kristaps  832:        mac[j] = '\0';
1.1       kristaps  833:
1.163     schwarze  834:        tok = (j > 1 || j < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
                    835:        if (MDOC_MAX == tok) {
1.185     kristaps  836:                mandoc_vmsg(MANDOCERR_MACRO, m->parse,
                    837:                                ln, sv, "%s", buf + sv - 1);
1.58      kristaps  838:                return(1);
1.53      kristaps  839:        }
1.1       kristaps  840:
1.160     kristaps  841:        /* Disregard the first trailing tab, if applicable. */
                    842:
                    843:        if ('\t' == buf[i])
                    844:                i++;
                    845:
                    846:        /* Jump to the next non-whitespace word. */
1.1       kristaps  847:
1.65      kristaps  848:        while (buf[i] && ' ' == buf[i])
1.53      kristaps  849:                i++;
1.1       kristaps  850:
1.125     kristaps  851:        /*
                    852:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    853:         * into the parser as "text", so we only warn about spaces here.
                    854:         */
1.115     kristaps  855:
                    856:        if ('\0' == buf[i] && ' ' == buf[i - 1])
1.174     kristaps  857:                mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE);
1.115     kristaps  858:
1.144     kristaps  859:        /*
                    860:         * If an initial macro or a list invocation, divert directly
                    861:         * into macro processing.
                    862:         */
                    863:
                    864:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.149     kristaps  865:                if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.144     kristaps  866:                        goto err;
                    867:                return(1);
                    868:        }
                    869:
                    870:        n = m->last;
                    871:        assert(m->last);
                    872:
                    873:        /*
                    874:         * If the first macro of a `Bl -column', open an `It' block
                    875:         * context around the parsed macro.
                    876:         */
                    877:
                    878:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  879:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  880:                m->flags |= MDOC_FREECOL;
1.149     kristaps  881:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  882:                        goto err;
                    883:                return(1);
                    884:        }
                    885:
                    886:        /*
                    887:         * If we're following a block-level `It' within a `Bl -column'
                    888:         * context (perhaps opened in the above block or in ptext()),
                    889:         * then open an `It' block context around the parsed macro.
1.98      kristaps  890:         */
1.144     kristaps  891:
                    892:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    893:                        NULL != n->parent &&
                    894:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  895:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  896:                m->flags |= MDOC_FREECOL;
1.149     kristaps  897:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  898:                        goto err;
                    899:                return(1);
                    900:        }
                    901:
                    902:        /* Normal processing of a macro. */
                    903:
1.149     kristaps  904:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.53      kristaps  905:                goto err;
1.1       kristaps  906:
1.53      kristaps  907:        return(1);
1.1       kristaps  908:
1.53      kristaps  909: err:   /* Error out. */
1.1       kristaps  910:
1.53      kristaps  911:        m->flags |= MDOC_HALT;
                    912:        return(0);
1.1       kristaps  913: }
1.100     kristaps  914:
1.186     kristaps  915: enum mdelim
                    916: mdoc_isdelim(const char *p)
                    917: {
                    918:
                    919:        if ('\0' == p[0])
                    920:                return(DELIM_NONE);
                    921:
                    922:        if ('\0' == p[1])
                    923:                switch (p[0]) {
                    924:                case('('):
                    925:                        /* FALLTHROUGH */
                    926:                case('['):
                    927:                        return(DELIM_OPEN);
                    928:                case('|'):
                    929:                        return(DELIM_MIDDLE);
                    930:                case('.'):
                    931:                        /* FALLTHROUGH */
                    932:                case(','):
                    933:                        /* FALLTHROUGH */
                    934:                case(';'):
                    935:                        /* FALLTHROUGH */
                    936:                case(':'):
                    937:                        /* FALLTHROUGH */
                    938:                case('?'):
                    939:                        /* FALLTHROUGH */
                    940:                case('!'):
                    941:                        /* FALLTHROUGH */
                    942:                case(')'):
                    943:                        /* FALLTHROUGH */
                    944:                case(']'):
                    945:                        return(DELIM_CLOSE);
                    946:                default:
                    947:                        return(DELIM_NONE);
                    948:                }
                    949:
                    950:        if ('\\' != p[0])
                    951:                return(DELIM_NONE);
1.100     kristaps  952:
1.186     kristaps  953:        if (0 == strcmp(p + 1, "."))
                    954:                return(DELIM_CLOSE);
                    955:        if (0 == strcmp(p + 1, "*(Ba"))
                    956:                return(DELIM_MIDDLE);
                    957:
                    958:        return(DELIM_NONE);
                    959: }

CVSweb