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

Annotation of mandoc/mdoc.c, Revision 1.193

1.193   ! kristaps    1: /*     $Id: mdoc.c,v 1.192 2011/07/27 07:09:41 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 *);
1.193   ! kristaps  100: #if 0
1.191     kristaps  101: static int               mdoc_preptext(struct mdoc *, int, char *, int);
1.193   ! kristaps  102: #endif
1.149     kristaps  103: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                    104: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.88      kristaps  105:
1.1       kristaps  106: const struct mdoc_node *
1.71      kristaps  107: mdoc_node(const struct mdoc *m)
1.1       kristaps  108: {
                    109:
1.177     kristaps  110:        assert( ! (MDOC_HALT & m->flags));
                    111:        return(m->first);
1.1       kristaps  112: }
                    113:
                    114:
1.37      kristaps  115: const struct mdoc_meta *
1.71      kristaps  116: mdoc_meta(const struct mdoc *m)
1.37      kristaps  117: {
                    118:
1.177     kristaps  119:        assert( ! (MDOC_HALT & m->flags));
                    120:        return(&m->meta);
1.37      kristaps  121: }
                    122:
                    123:
1.85      kristaps  124: /*
                    125:  * Frees volatile resources (parse tree, meta-data, fields).
                    126:  */
1.73      kristaps  127: static void
                    128: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  129: {
                    130:
                    131:        if (mdoc->first)
1.121     kristaps  132:                mdoc_node_delete(mdoc, mdoc->first);
1.67      kristaps  133:        if (mdoc->meta.title)
                    134:                free(mdoc->meta.title);
                    135:        if (mdoc->meta.os)
                    136:                free(mdoc->meta.os);
                    137:        if (mdoc->meta.name)
                    138:                free(mdoc->meta.name);
                    139:        if (mdoc->meta.arch)
                    140:                free(mdoc->meta.arch);
                    141:        if (mdoc->meta.vol)
                    142:                free(mdoc->meta.vol);
1.133     kristaps  143:        if (mdoc->meta.msec)
                    144:                free(mdoc->meta.msec);
1.183     kristaps  145:        if (mdoc->meta.date)
                    146:                free(mdoc->meta.date);
1.73      kristaps  147: }
                    148:
                    149:
1.85      kristaps  150: /*
                    151:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    152:  */
1.113     kristaps  153: static void
1.73      kristaps  154: mdoc_alloc1(struct mdoc *mdoc)
                    155: {
1.67      kristaps  156:
1.112     kristaps  157:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  158:        mdoc->flags = 0;
1.85      kristaps  159:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  160:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  161:        mdoc->first = mdoc->last;
1.67      kristaps  162:        mdoc->last->type = MDOC_ROOT;
                    163:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  164: }
                    165:
                    166:
                    167: /*
1.85      kristaps  168:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    169:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    170:  * and the parser is ready for re-invocation on a new tree; however,
                    171:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  172:  */
1.113     kristaps  173: void
1.73      kristaps  174: mdoc_reset(struct mdoc *mdoc)
                    175: {
                    176:
                    177:        mdoc_free1(mdoc);
1.113     kristaps  178:        mdoc_alloc1(mdoc);
1.67      kristaps  179: }
                    180:
                    181:
1.68      kristaps  182: /*
1.85      kristaps  183:  * Completely free up all volatile and non-volatile parse resources.
                    184:  * After invocation, the pointer is no longer usable.
1.68      kristaps  185:  */
1.67      kristaps  186: void
1.38      kristaps  187: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  188: {
                    189:
1.73      kristaps  190:        mdoc_free1(mdoc);
1.1       kristaps  191:        free(mdoc);
                    192: }
                    193:
                    194:
1.85      kristaps  195: /*
                    196:  * Allocate volatile and non-volatile parse resources.
                    197:  */
1.1       kristaps  198: struct mdoc *
1.189     kristaps  199: mdoc_alloc(struct roff *roff, struct mparse *parse)
1.1       kristaps  200: {
                    201:        struct mdoc     *p;
                    202:
1.113     kristaps  203:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    204:
1.185     kristaps  205:        p->parse = parse;
1.189     kristaps  206:        p->roff = roff;
1.73      kristaps  207:
1.113     kristaps  208:        mdoc_hash_init();
                    209:        mdoc_alloc1(p);
                    210:        return(p);
1.1       kristaps  211: }
                    212:
                    213:
1.68      kristaps  214: /*
                    215:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  216:  * through to macro_end() in macro.c.
1.68      kristaps  217:  */
1.1       kristaps  218: int
1.72      kristaps  219: mdoc_endparse(struct mdoc *m)
1.20      kristaps  220: {
                    221:
1.177     kristaps  222:        assert( ! (MDOC_HALT & m->flags));
                    223:        if (mdoc_macroend(m))
1.20      kristaps  224:                return(1);
1.72      kristaps  225:        m->flags |= MDOC_HALT;
                    226:        return(0);
1.181     kristaps  227: }
                    228:
                    229: int
                    230: mdoc_addeqn(struct mdoc *m, const struct eqn *ep)
                    231: {
                    232:        struct mdoc_node *n;
                    233:
                    234:        assert( ! (MDOC_HALT & m->flags));
                    235:
                    236:        /* No text before an initial macro. */
                    237:
                    238:        if (SEC_NONE == m->lastnamed) {
1.190     kristaps  239:                mdoc_pmsg(m, ep->ln, ep->pos, MANDOCERR_NOTEXT);
1.181     kristaps  240:                return(1);
                    241:        }
                    242:
1.190     kristaps  243:        n = node_alloc(m, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  244:        n->eqn = ep;
                    245:
                    246:        if ( ! node_append(m, n))
                    247:                return(0);
                    248:
                    249:        m->next = MDOC_NEXT_SIBLING;
                    250:        return(1);
1.20      kristaps  251: }
                    252:
1.175     kristaps  253: int
                    254: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    255: {
1.180     kristaps  256:        struct mdoc_node *n;
1.175     kristaps  257:
1.177     kristaps  258:        assert( ! (MDOC_HALT & m->flags));
1.175     kristaps  259:
                    260:        /* No text before an initial macro. */
                    261:
                    262:        if (SEC_NONE == m->lastnamed) {
1.179     kristaps  263:                mdoc_pmsg(m, sp->line, 0, MANDOCERR_NOTEXT);
1.175     kristaps  264:                return(1);
                    265:        }
                    266:
1.180     kristaps  267:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
                    268:        n->span = sp;
                    269:
                    270:        if ( ! node_append(m, n))
                    271:                return(0);
                    272:
                    273:        m->next = MDOC_NEXT_SIBLING;
                    274:        return(1);
1.175     kristaps  275: }
                    276:
1.20      kristaps  277:
1.50      kristaps  278: /*
1.53      kristaps  279:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  280:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  281:  */
1.20      kristaps  282: int
1.149     kristaps  283: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  284: {
                    285:
1.177     kristaps  286:        assert( ! (MDOC_HALT & m->flags));
1.50      kristaps  287:
1.130     kristaps  288:        m->flags |= MDOC_NEWLINE;
1.153     schwarze  289:
                    290:        /*
                    291:         * Let the roff nS register switch SYNOPSIS mode early,
                    292:         * such that the parser knows at all times
                    293:         * whether this mode is on or off.
                    294:         * Note that this mode is also switched by the Sh macro.
                    295:         */
1.189     kristaps  296:        if (roff_regisset(m->roff, REG_nS)) {
                    297:                if (roff_regget(m->roff, REG_nS))
1.153     schwarze  298:                        m->flags |= MDOC_SYNOPSIS;
                    299:                else
                    300:                        m->flags &= ~MDOC_SYNOPSIS;
                    301:        }
                    302:
1.188     kristaps  303:        return(mandoc_getcontrol(buf, &offs) ?
1.149     kristaps  304:                        mdoc_pmacro(m, ln, buf, offs) :
1.193   ! kristaps  305:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  306: }
                    307:
1.88      kristaps  308: int
1.148     kristaps  309: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  310: {
1.122     kristaps  311:        assert(tok < MDOC_MAX);
                    312:
                    313:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  314:
1.88      kristaps  315:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.174     kristaps  316:                        MDOC_PBODY & m->flags) {
                    317:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    318:                return(1);
                    319:        }
1.122     kristaps  320:
                    321:        /* If we're in the prologue, deny "body" macros.  */
                    322:
                    323:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    324:                        ! (MDOC_PBODY & m->flags)) {
1.174     kristaps  325:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.167     kristaps  326:                if (NULL == m->meta.msec)
                    327:                        m->meta.msec = mandoc_strdup("1");
1.120     kristaps  328:                if (NULL == m->meta.title)
1.140     kristaps  329:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.120     kristaps  330:                if (NULL == m->meta.vol)
1.140     kristaps  331:                        m->meta.vol = mandoc_strdup("LOCAL");
1.120     kristaps  332:                if (NULL == m->meta.os)
1.140     kristaps  333:                        m->meta.os = mandoc_strdup("LOCAL");
1.182     schwarze  334:                if (NULL == m->meta.date)
1.185     kristaps  335:                        m->meta.date = mandoc_normdate
                    336:                                (m->parse, NULL, line, ppos);
1.120     kristaps  337:                m->flags |= MDOC_PBODY;
                    338:        }
1.88      kristaps  339:
1.149     kristaps  340:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.73      kristaps  341: }
                    342:
                    343:
                    344: static int
                    345: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  346: {
                    347:
1.25      kristaps  348:        assert(mdoc->last);
                    349:        assert(mdoc->first);
                    350:        assert(MDOC_ROOT != p->type);
1.1       kristaps  351:
1.13      kristaps  352:        switch (mdoc->next) {
                    353:        case (MDOC_NEXT_SIBLING):
1.6       kristaps  354:                mdoc->last->next = p;
                    355:                p->prev = mdoc->last;
1.13      kristaps  356:                p->parent = mdoc->last->parent;
1.1       kristaps  357:                break;
1.13      kristaps  358:        case (MDOC_NEXT_CHILD):
                    359:                mdoc->last->child = p;
1.1       kristaps  360:                p->parent = mdoc->last;
                    361:                break;
                    362:        default:
1.13      kristaps  363:                abort();
                    364:                /* NOTREACHED */
1.1       kristaps  365:        }
                    366:
1.86      kristaps  367:        p->parent->nchild++;
                    368:
1.172     kristaps  369:        /*
                    370:         * Copy over the normalised-data pointer of our parent.  Not
                    371:         * everybody has one, but copying a null pointer is fine.
                    372:         */
                    373:
                    374:        switch (p->type) {
                    375:        case (MDOC_BODY):
                    376:                /* FALLTHROUGH */
                    377:        case (MDOC_TAIL):
                    378:                /* FALLTHROUGH */
                    379:        case (MDOC_HEAD):
                    380:                p->norm = p->parent->norm;
                    381:                break;
                    382:        default:
                    383:                break;
                    384:        }
                    385:
1.23      kristaps  386:        if ( ! mdoc_valid_pre(mdoc, p))
                    387:                return(0);
1.27      kristaps  388:
                    389:        switch (p->type) {
                    390:        case (MDOC_HEAD):
                    391:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  392:                p->parent->head = p;
1.27      kristaps  393:                break;
                    394:        case (MDOC_TAIL):
                    395:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  396:                p->parent->tail = p;
1.27      kristaps  397:                break;
                    398:        case (MDOC_BODY):
1.152     schwarze  399:                if (p->end)
                    400:                        break;
1.27      kristaps  401:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  402:                p->parent->body = p;
1.27      kristaps  403:                break;
                    404:        default:
                    405:                break;
                    406:        }
                    407:
1.1       kristaps  408:        mdoc->last = p;
1.71      kristaps  409:
                    410:        switch (p->type) {
1.176     kristaps  411:        case (MDOC_TBL):
                    412:                /* FALLTHROUGH */
1.71      kristaps  413:        case (MDOC_TEXT):
                    414:                if ( ! mdoc_valid_post(mdoc))
                    415:                        return(0);
                    416:                break;
                    417:        default:
                    418:                break;
                    419:        }
                    420:
1.23      kristaps  421:        return(1);
1.1       kristaps  422: }
                    423:
                    424:
1.45      kristaps  425: static struct mdoc_node *
1.117     kristaps  426: node_alloc(struct mdoc *m, int line, int pos,
                    427:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  428: {
1.46      kristaps  429:        struct mdoc_node *p;
                    430:
1.113     kristaps  431:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.91      kristaps  432:        p->sec = m->lastsec;
1.73      kristaps  433:        p->line = line;
                    434:        p->pos = pos;
                    435:        p->tok = tok;
1.118     kristaps  436:        p->type = type;
1.150     kristaps  437:
                    438:        /* Flag analysis. */
                    439:
1.153     schwarze  440:        if (MDOC_SYNOPSIS & m->flags)
                    441:                p->flags |= MDOC_SYNPRETTY;
                    442:        else
                    443:                p->flags &= ~MDOC_SYNPRETTY;
1.130     kristaps  444:        if (MDOC_NEWLINE & m->flags)
                    445:                p->flags |= MDOC_LINE;
                    446:        m->flags &= ~MDOC_NEWLINE;
1.150     kristaps  447:
1.46      kristaps  448:        return(p);
1.45      kristaps  449: }
                    450:
                    451:
1.23      kristaps  452: int
1.117     kristaps  453: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.17      kristaps  454: {
                    455:        struct mdoc_node *p;
                    456:
1.91      kristaps  457:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.102     kristaps  458:        if ( ! node_append(m, p))
                    459:                return(0);
                    460:        m->next = MDOC_NEXT_CHILD;
                    461:        return(1);
1.17      kristaps  462: }
                    463:
                    464:
1.23      kristaps  465: int
1.117     kristaps  466: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  467: {
                    468:        struct mdoc_node *p;
                    469:
1.91      kristaps  470:        assert(m->first);
                    471:        assert(m->last);
1.1       kristaps  472:
1.91      kristaps  473:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.102     kristaps  474:        if ( ! node_append(m, p))
                    475:                return(0);
                    476:        m->next = MDOC_NEXT_CHILD;
                    477:        return(1);
1.1       kristaps  478: }
                    479:
                    480:
1.23      kristaps  481: int
1.117     kristaps  482: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  483: {
                    484:        struct mdoc_node *p;
                    485:
1.91      kristaps  486:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.102     kristaps  487:        if ( ! node_append(m, p))
                    488:                return(0);
                    489:        m->next = MDOC_NEXT_CHILD;
1.152     schwarze  490:        return(1);
                    491: }
                    492:
                    493:
                    494: int
                    495: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    496:                struct mdoc_node *body, enum mdoc_endbody end)
                    497: {
                    498:        struct mdoc_node *p;
                    499:
                    500:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    501:        p->pending = body;
                    502:        p->end = end;
                    503:        if ( ! node_append(m, p))
                    504:                return(0);
                    505:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  506:        return(1);
1.1       kristaps  507: }
                    508:
                    509:
1.23      kristaps  510: int
1.91      kristaps  511: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  512:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  513: {
                    514:        struct mdoc_node *p;
                    515:
1.91      kristaps  516:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  517:        p->args = args;
                    518:        if (p->args)
1.53      kristaps  519:                (args->refcnt)++;
1.172     kristaps  520:
                    521:        switch (tok) {
                    522:        case (MDOC_Bd):
                    523:                /* FALLTHROUGH */
                    524:        case (MDOC_Bf):
                    525:                /* FALLTHROUGH */
                    526:        case (MDOC_Bl):
1.173     kristaps  527:                /* FALLTHROUGH */
                    528:        case (MDOC_Rs):
1.172     kristaps  529:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    530:                break;
                    531:        default:
                    532:                break;
                    533:        }
                    534:
1.102     kristaps  535:        if ( ! node_append(m, p))
                    536:                return(0);
                    537:        m->next = MDOC_NEXT_CHILD;
                    538:        return(1);
1.1       kristaps  539: }
                    540:
                    541:
1.23      kristaps  542: int
1.91      kristaps  543: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  544:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  545: {
                    546:        struct mdoc_node *p;
                    547:
1.91      kristaps  548:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.77      kristaps  549:        p->args = args;
                    550:        if (p->args)
1.53      kristaps  551:                (args->refcnt)++;
1.172     kristaps  552:
                    553:        switch (tok) {
                    554:        case (MDOC_An):
                    555:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    556:                break;
                    557:        default:
                    558:                break;
                    559:        }
                    560:
1.102     kristaps  561:        if ( ! node_append(m, p))
                    562:                return(0);
                    563:        m->next = MDOC_NEXT_CHILD;
1.175     kristaps  564:        return(1);
                    565: }
1.1       kristaps  566:
1.124     kristaps  567: int
                    568: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  569: {
1.91      kristaps  570:        struct mdoc_node *n;
1.124     kristaps  571:        size_t            sv, len;
                    572:
                    573:        len = strlen(p);
1.1       kristaps  574:
1.125     kristaps  575:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.113     kristaps  576:        n->string = mandoc_malloc(len + 1);
1.91      kristaps  577:        sv = strlcpy(n->string, p, len + 1);
                    578:
                    579:        /* Prohibit truncation. */
                    580:        assert(sv < len + 1);
                    581:
1.101     kristaps  582:        if ( ! node_append(m, n))
                    583:                return(0);
1.124     kristaps  584:
1.101     kristaps  585:        m->next = MDOC_NEXT_SIBLING;
                    586:        return(1);
1.91      kristaps  587: }
                    588:
                    589:
1.155     kristaps  590: static void
1.53      kristaps  591: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  592: {
                    593:
1.172     kristaps  594:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  595:                free(p->norm);
1.53      kristaps  596:        if (p->string)
                    597:                free(p->string);
                    598:        if (p->args)
                    599:                mdoc_argv_free(p->args);
1.1       kristaps  600:        free(p);
                    601: }
                    602:
                    603:
1.121     kristaps  604: static void
                    605: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    606: {
                    607:
                    608:        /* Adjust siblings. */
                    609:
                    610:        if (n->prev)
                    611:                n->prev->next = n->next;
                    612:        if (n->next)
                    613:                n->next->prev = n->prev;
                    614:
                    615:        /* Adjust parent. */
                    616:
                    617:        if (n->parent) {
                    618:                n->parent->nchild--;
                    619:                if (n->parent->child == n)
                    620:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  621:                if (n->parent->last == n)
                    622:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  623:        }
                    624:
                    625:        /* Adjust parse point, if applicable. */
                    626:
                    627:        if (m && m->last == n) {
                    628:                if (n->prev) {
                    629:                        m->last = n->prev;
                    630:                        m->next = MDOC_NEXT_SIBLING;
                    631:                } else {
                    632:                        m->last = n->parent;
                    633:                        m->next = MDOC_NEXT_CHILD;
                    634:                }
                    635:        }
                    636:
                    637:        if (m && m->first == n)
                    638:                m->first = NULL;
                    639: }
                    640:
                    641:
1.53      kristaps  642: void
1.121     kristaps  643: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  644: {
                    645:
1.121     kristaps  646:        while (p->child) {
                    647:                assert(p->nchild);
                    648:                mdoc_node_delete(m, p->child);
                    649:        }
                    650:        assert(0 == p->nchild);
1.1       kristaps  651:
1.121     kristaps  652:        mdoc_node_unlink(m, p);
1.53      kristaps  653:        mdoc_node_free(p);
1.1       kristaps  654: }
                    655:
1.193   ! kristaps  656: #if 0
1.191     kristaps  657: /*
                    658:  * Pre-treat a text line.
                    659:  * Text lines can consist of equations, which must be handled apart from
                    660:  * the regular text.
                    661:  * Thus, use this function to step through a line checking if it has any
                    662:  * equations embedded in it.
                    663:  * This must handle multiple equations AND equations that do not end at
                    664:  * the end-of-line, i.e., will re-enter in the next roff parse.
                    665:  */
                    666: static int
                    667: mdoc_preptext(struct mdoc *m, int line, char *buf, int offs)
                    668: {
                    669:        char            *start, *end;
                    670:        char             delim;
                    671:
                    672:        while ('\0' != buf[offs]) {
                    673:                /* Mark starting position if eqn is set. */
                    674:                start = NULL;
                    675:                if ('\0' != (delim = roff_eqndelim(m->roff)))
                    676:                        if (NULL != (start = strchr(buf + offs, delim)))
                    677:                                *start++ = '\0';
                    678:
                    679:                /* Parse text as normal. */
                    680:                if ( ! mdoc_ptext(m, line, buf, offs))
                    681:                        return(0);
                    682:
                    683:                /* Continue only if an equation exists. */
                    684:                if (NULL == start)
                    685:                        break;
                    686:
                    687:                /* Read past the end of the equation. */
                    688:                offs += start - (buf + offs);
                    689:                assert(start == &buf[offs]);
                    690:                if (NULL != (end = strchr(buf + offs, delim))) {
                    691:                        *end++ = '\0';
                    692:                        while (' ' == *end)
                    693:                                end++;
                    694:                }
                    695:
                    696:                /* Parse the equation itself. */
                    697:                roff_openeqn(m->roff, NULL, line, offs, buf);
                    698:
                    699:                /* Process a finished equation? */
                    700:                if (roff_closeeqn(m->roff))
                    701:                        if ( ! mdoc_addeqn(m, roff_eqn(m->roff)))
                    702:                                return(0);
                    703:                offs += (end - (buf + offs));
                    704:        }
                    705:
                    706:        return(1);
                    707: }
1.193   ! kristaps  708: #endif
1.1       kristaps  709:
1.53      kristaps  710: /*
                    711:  * Parse free-form text, that is, a line that does not begin with the
                    712:  * control character.
                    713:  */
                    714: static int
1.149     kristaps  715: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  716: {
1.142     kristaps  717:        char             *c, *ws, *end;
                    718:        struct mdoc_node *n;
1.1       kristaps  719:
1.124     kristaps  720:        /* No text before an initial macro. */
                    721:
1.174     kristaps  722:        if (SEC_NONE == m->lastnamed) {
                    723:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    724:                return(1);
                    725:        }
1.142     kristaps  726:
                    727:        assert(m->last);
                    728:        n = m->last;
                    729:
                    730:        /*
1.144     kristaps  731:         * Divert directly to list processing if we're encountering a
1.142     kristaps  732:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  733:         * (a MDOC_BODY means it's already open, in which case we should
                    734:         * process within its context in the normal way).
1.142     kristaps  735:         */
                    736:
1.143     kristaps  737:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  738:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  739:                /* `Bl' is open without any children. */
1.142     kristaps  740:                m->flags |= MDOC_FREECOL;
1.149     kristaps  741:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  742:        }
                    743:
                    744:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    745:                        NULL != n->parent &&
                    746:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  747:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  748:                /* `Bl' has block-level `It' children. */
1.142     kristaps  749:                m->flags |= MDOC_FREECOL;
1.149     kristaps  750:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  751:        }
1.124     kristaps  752:
1.137     schwarze  753:        /*
                    754:         * Search for the beginning of unescaped trailing whitespace (ws)
                    755:         * and for the first character not to be output (end).
                    756:         */
1.139     kristaps  757:
                    758:        /* FIXME: replace with strcspn(). */
1.137     schwarze  759:        ws = NULL;
                    760:        for (c = end = buf + offs; *c; c++) {
                    761:                switch (*c) {
                    762:                case ' ':
                    763:                        if (NULL == ws)
                    764:                                ws = c;
                    765:                        continue;
                    766:                case '\t':
                    767:                        /*
                    768:                         * Always warn about trailing tabs,
                    769:                         * even outside literal context,
                    770:                         * where they should be put on the next line.
                    771:                         */
                    772:                        if (NULL == ws)
                    773:                                ws = c;
                    774:                        /*
                    775:                         * Strip trailing tabs in literal context only;
                    776:                         * outside, they affect the next line.
                    777:                         */
                    778:                        if (MDOC_LITERAL & m->flags)
                    779:                                continue;
                    780:                        break;
                    781:                case '\\':
                    782:                        /* Skip the escaped character, too, if any. */
                    783:                        if (c[1])
                    784:                                c++;
                    785:                        /* FALLTHROUGH */
                    786:                default:
                    787:                        ws = NULL;
                    788:                        break;
                    789:                }
                    790:                end = c + 1;
                    791:        }
                    792:        *end = '\0';
1.91      kristaps  793:
1.137     schwarze  794:        if (ws)
1.174     kristaps  795:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.115     kristaps  796:
1.137     schwarze  797:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.174     kristaps  798:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.124     kristaps  799:
1.119     kristaps  800:                /*
1.165     schwarze  801:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  802:                 * blank lines aren't allowed, but enough manuals assume this
                    803:                 * behaviour that we want to work around it.
1.119     kristaps  804:                 */
1.165     schwarze  805:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.119     kristaps  806:                        return(0);
1.124     kristaps  807:
                    808:                m->next = MDOC_NEXT_SIBLING;
                    809:                return(1);
1.119     kristaps  810:        }
1.68      kristaps  811:
1.137     schwarze  812:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
                    813:                return(0);
1.91      kristaps  814:
1.137     schwarze  815:        if (MDOC_LITERAL & m->flags)
                    816:                return(1);
1.128     kristaps  817:
                    818:        /*
                    819:         * End-of-sentence check.  If the last character is an unescaped
                    820:         * EOS character, then flag the node as being the end of a
                    821:         * sentence.  The front-end will know how to interpret this.
                    822:         */
1.132     kristaps  823:
1.137     schwarze  824:        assert(buf < end);
                    825:
1.159     schwarze  826:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.128     kristaps  827:                m->last->flags |= MDOC_EOS;
                    828:
                    829:        return(1);
1.1       kristaps  830: }
                    831:
                    832:
1.53      kristaps  833: /*
                    834:  * Parse a macro line, that is, a line beginning with the control
                    835:  * character.
                    836:  */
1.155     kristaps  837: static int
1.149     kristaps  838: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  839: {
1.144     kristaps  840:        enum mdoct        tok;
1.188     kristaps  841:        int               i, sv;
1.144     kristaps  842:        char              mac[5];
                    843:        struct mdoc_node *n;
1.1       kristaps  844:
1.188     kristaps  845:        /* Empty post-control lines are ignored. */
1.63      kristaps  846:
1.188     kristaps  847:        if ('"' == buf[offs]) {
                    848:                mdoc_pmsg(m, ln, offs, MANDOCERR_BADCOMMENT);
                    849:                return(1);
                    850:        } else if ('\0' == buf[offs])
1.63      kristaps  851:                return(1);
                    852:
1.188     kristaps  853:        sv = offs;
1.130     kristaps  854:
1.160     kristaps  855:        /*
1.162     schwarze  856:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  857:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  858:         */
1.1       kristaps  859:
1.188     kristaps  860:        i = 0;
                    861:        while (i < 4 && '\0' != buf[offs] &&
                    862:                        ' ' != buf[offs] && '\t' != buf[offs])
                    863:                mac[i++] = buf[offs++];
                    864:
                    865:        mac[i] = '\0';
                    866:
                    867:        tok = (i > 1 || i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  868:
1.163     schwarze  869:        if (MDOC_MAX == tok) {
1.185     kristaps  870:                mandoc_vmsg(MANDOCERR_MACRO, m->parse,
                    871:                                ln, sv, "%s", buf + sv - 1);
1.58      kristaps  872:                return(1);
1.53      kristaps  873:        }
1.1       kristaps  874:
1.160     kristaps  875:        /* Disregard the first trailing tab, if applicable. */
                    876:
1.188     kristaps  877:        if ('\t' == buf[offs])
                    878:                offs++;
1.160     kristaps  879:
                    880:        /* Jump to the next non-whitespace word. */
1.1       kristaps  881:
1.188     kristaps  882:        while (buf[offs] && ' ' == buf[offs])
                    883:                offs++;
1.1       kristaps  884:
1.125     kristaps  885:        /*
                    886:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    887:         * into the parser as "text", so we only warn about spaces here.
                    888:         */
1.115     kristaps  889:
1.188     kristaps  890:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
                    891:                mdoc_pmsg(m, ln, offs - 1, MANDOCERR_EOLNSPACE);
1.115     kristaps  892:
1.144     kristaps  893:        /*
                    894:         * If an initial macro or a list invocation, divert directly
                    895:         * into macro processing.
                    896:         */
                    897:
                    898:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.188     kristaps  899:                if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.144     kristaps  900:                        goto err;
                    901:                return(1);
                    902:        }
                    903:
                    904:        n = m->last;
                    905:        assert(m->last);
                    906:
                    907:        /*
                    908:         * If the first macro of a `Bl -column', open an `It' block
                    909:         * context around the parsed macro.
                    910:         */
                    911:
                    912:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  913:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  914:                m->flags |= MDOC_FREECOL;
1.149     kristaps  915:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  916:                        goto err;
                    917:                return(1);
                    918:        }
                    919:
                    920:        /*
                    921:         * If we're following a block-level `It' within a `Bl -column'
                    922:         * context (perhaps opened in the above block or in ptext()),
                    923:         * then open an `It' block context around the parsed macro.
1.98      kristaps  924:         */
1.144     kristaps  925:
                    926:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    927:                        NULL != n->parent &&
                    928:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  929:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  930:                m->flags |= MDOC_FREECOL;
1.149     kristaps  931:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  932:                        goto err;
                    933:                return(1);
                    934:        }
                    935:
                    936:        /* Normal processing of a macro. */
                    937:
1.188     kristaps  938:        if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.53      kristaps  939:                goto err;
1.1       kristaps  940:
1.53      kristaps  941:        return(1);
1.1       kristaps  942:
1.53      kristaps  943: err:   /* Error out. */
1.1       kristaps  944:
1.53      kristaps  945:        m->flags |= MDOC_HALT;
                    946:        return(0);
1.1       kristaps  947: }
1.100     kristaps  948:
1.186     kristaps  949: enum mdelim
                    950: mdoc_isdelim(const char *p)
                    951: {
                    952:
                    953:        if ('\0' == p[0])
                    954:                return(DELIM_NONE);
                    955:
                    956:        if ('\0' == p[1])
                    957:                switch (p[0]) {
                    958:                case('('):
                    959:                        /* FALLTHROUGH */
                    960:                case('['):
                    961:                        return(DELIM_OPEN);
                    962:                case('|'):
                    963:                        return(DELIM_MIDDLE);
                    964:                case('.'):
                    965:                        /* FALLTHROUGH */
                    966:                case(','):
                    967:                        /* FALLTHROUGH */
                    968:                case(';'):
                    969:                        /* FALLTHROUGH */
                    970:                case(':'):
                    971:                        /* FALLTHROUGH */
                    972:                case('?'):
                    973:                        /* FALLTHROUGH */
                    974:                case('!'):
                    975:                        /* FALLTHROUGH */
                    976:                case(')'):
                    977:                        /* FALLTHROUGH */
                    978:                case(']'):
                    979:                        return(DELIM_CLOSE);
                    980:                default:
                    981:                        return(DELIM_NONE);
                    982:                }
                    983:
                    984:        if ('\\' != p[0])
                    985:                return(DELIM_NONE);
1.100     kristaps  986:
1.186     kristaps  987:        if (0 == strcmp(p + 1, "."))
                    988:                return(DELIM_CLOSE);
                    989:        if (0 == strcmp(p + 1, "*(Ba"))
                    990:                return(DELIM_MIDDLE);
                    991:
                    992:        return(DELIM_NONE);
                    993: }

CVSweb