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

Annotation of mandoc/mdoc.c, Revision 1.198

1.198   ! kristaps    1: /*     $Id: mdoc.c,v 1.197 2012/05/27 17:48:57 schwarze 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;
1.196     schwarze  163:        mdoc->last->tok = MDOC_MAX;
1.67      kristaps  164:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  165: }
                    166:
                    167:
                    168: /*
1.85      kristaps  169:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    170:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    171:  * and the parser is ready for re-invocation on a new tree; however,
                    172:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  173:  */
1.113     kristaps  174: void
1.73      kristaps  175: mdoc_reset(struct mdoc *mdoc)
                    176: {
                    177:
                    178:        mdoc_free1(mdoc);
1.113     kristaps  179:        mdoc_alloc1(mdoc);
1.67      kristaps  180: }
                    181:
                    182:
1.68      kristaps  183: /*
1.85      kristaps  184:  * Completely free up all volatile and non-volatile parse resources.
                    185:  * After invocation, the pointer is no longer usable.
1.68      kristaps  186:  */
1.67      kristaps  187: void
1.38      kristaps  188: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  189: {
                    190:
1.73      kristaps  191:        mdoc_free1(mdoc);
1.1       kristaps  192:        free(mdoc);
                    193: }
                    194:
                    195:
1.85      kristaps  196: /*
                    197:  * Allocate volatile and non-volatile parse resources.
                    198:  */
1.1       kristaps  199: struct mdoc *
1.197     schwarze  200: mdoc_alloc(struct roff *roff, struct mparse *parse, char *defos)
1.1       kristaps  201: {
                    202:        struct mdoc     *p;
                    203:
1.113     kristaps  204:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    205:
1.185     kristaps  206:        p->parse = parse;
1.197     schwarze  207:        p->defos = defos;
1.189     kristaps  208:        p->roff = roff;
1.73      kristaps  209:
1.113     kristaps  210:        mdoc_hash_init();
                    211:        mdoc_alloc1(p);
                    212:        return(p);
1.1       kristaps  213: }
                    214:
                    215:
1.68      kristaps  216: /*
                    217:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  218:  * through to macro_end() in macro.c.
1.68      kristaps  219:  */
1.1       kristaps  220: int
1.72      kristaps  221: mdoc_endparse(struct mdoc *m)
1.20      kristaps  222: {
                    223:
1.177     kristaps  224:        assert( ! (MDOC_HALT & m->flags));
                    225:        if (mdoc_macroend(m))
1.20      kristaps  226:                return(1);
1.72      kristaps  227:        m->flags |= MDOC_HALT;
                    228:        return(0);
1.181     kristaps  229: }
                    230:
                    231: int
                    232: mdoc_addeqn(struct mdoc *m, const struct eqn *ep)
                    233: {
                    234:        struct mdoc_node *n;
                    235:
                    236:        assert( ! (MDOC_HALT & m->flags));
                    237:
                    238:        /* No text before an initial macro. */
                    239:
                    240:        if (SEC_NONE == m->lastnamed) {
1.190     kristaps  241:                mdoc_pmsg(m, ep->ln, ep->pos, MANDOCERR_NOTEXT);
1.181     kristaps  242:                return(1);
                    243:        }
                    244:
1.190     kristaps  245:        n = node_alloc(m, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  246:        n->eqn = ep;
                    247:
                    248:        if ( ! node_append(m, n))
                    249:                return(0);
                    250:
                    251:        m->next = MDOC_NEXT_SIBLING;
                    252:        return(1);
1.20      kristaps  253: }
                    254:
1.175     kristaps  255: int
                    256: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    257: {
1.180     kristaps  258:        struct mdoc_node *n;
1.175     kristaps  259:
1.177     kristaps  260:        assert( ! (MDOC_HALT & m->flags));
1.175     kristaps  261:
                    262:        /* No text before an initial macro. */
                    263:
                    264:        if (SEC_NONE == m->lastnamed) {
1.179     kristaps  265:                mdoc_pmsg(m, sp->line, 0, MANDOCERR_NOTEXT);
1.175     kristaps  266:                return(1);
                    267:        }
                    268:
1.180     kristaps  269:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
                    270:        n->span = sp;
                    271:
                    272:        if ( ! node_append(m, n))
                    273:                return(0);
                    274:
                    275:        m->next = MDOC_NEXT_SIBLING;
                    276:        return(1);
1.175     kristaps  277: }
                    278:
1.20      kristaps  279:
1.50      kristaps  280: /*
1.53      kristaps  281:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  282:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  283:  */
1.20      kristaps  284: int
1.149     kristaps  285: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  286: {
                    287:
1.177     kristaps  288:        assert( ! (MDOC_HALT & m->flags));
1.50      kristaps  289:
1.130     kristaps  290:        m->flags |= MDOC_NEWLINE;
1.153     schwarze  291:
                    292:        /*
                    293:         * Let the roff nS register switch SYNOPSIS mode early,
                    294:         * such that the parser knows at all times
                    295:         * whether this mode is on or off.
                    296:         * Note that this mode is also switched by the Sh macro.
                    297:         */
1.189     kristaps  298:        if (roff_regisset(m->roff, REG_nS)) {
                    299:                if (roff_regget(m->roff, REG_nS))
1.153     schwarze  300:                        m->flags |= MDOC_SYNOPSIS;
                    301:                else
                    302:                        m->flags &= ~MDOC_SYNOPSIS;
                    303:        }
                    304:
1.198   ! kristaps  305:        return(roff_getcontrol(m->roff, buf, &offs) ?
1.149     kristaps  306:                        mdoc_pmacro(m, ln, buf, offs) :
1.193     kristaps  307:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  308: }
                    309:
1.88      kristaps  310: int
1.148     kristaps  311: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  312: {
1.122     kristaps  313:        assert(tok < MDOC_MAX);
                    314:
                    315:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  316:
1.88      kristaps  317:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.174     kristaps  318:                        MDOC_PBODY & m->flags) {
                    319:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    320:                return(1);
                    321:        }
1.122     kristaps  322:
                    323:        /* If we're in the prologue, deny "body" macros.  */
                    324:
                    325:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    326:                        ! (MDOC_PBODY & m->flags)) {
1.174     kristaps  327:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.167     kristaps  328:                if (NULL == m->meta.msec)
                    329:                        m->meta.msec = mandoc_strdup("1");
1.120     kristaps  330:                if (NULL == m->meta.title)
1.140     kristaps  331:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.120     kristaps  332:                if (NULL == m->meta.vol)
1.140     kristaps  333:                        m->meta.vol = mandoc_strdup("LOCAL");
1.120     kristaps  334:                if (NULL == m->meta.os)
1.140     kristaps  335:                        m->meta.os = mandoc_strdup("LOCAL");
1.182     schwarze  336:                if (NULL == m->meta.date)
1.185     kristaps  337:                        m->meta.date = mandoc_normdate
                    338:                                (m->parse, NULL, line, ppos);
1.120     kristaps  339:                m->flags |= MDOC_PBODY;
                    340:        }
1.88      kristaps  341:
1.149     kristaps  342:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.73      kristaps  343: }
                    344:
                    345:
                    346: static int
                    347: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  348: {
                    349:
1.25      kristaps  350:        assert(mdoc->last);
                    351:        assert(mdoc->first);
                    352:        assert(MDOC_ROOT != p->type);
1.1       kristaps  353:
1.13      kristaps  354:        switch (mdoc->next) {
                    355:        case (MDOC_NEXT_SIBLING):
1.6       kristaps  356:                mdoc->last->next = p;
                    357:                p->prev = mdoc->last;
1.13      kristaps  358:                p->parent = mdoc->last->parent;
1.1       kristaps  359:                break;
1.13      kristaps  360:        case (MDOC_NEXT_CHILD):
                    361:                mdoc->last->child = p;
1.1       kristaps  362:                p->parent = mdoc->last;
                    363:                break;
                    364:        default:
1.13      kristaps  365:                abort();
                    366:                /* NOTREACHED */
1.1       kristaps  367:        }
                    368:
1.86      kristaps  369:        p->parent->nchild++;
                    370:
1.172     kristaps  371:        /*
                    372:         * Copy over the normalised-data pointer of our parent.  Not
                    373:         * everybody has one, but copying a null pointer is fine.
                    374:         */
                    375:
                    376:        switch (p->type) {
                    377:        case (MDOC_BODY):
                    378:                /* FALLTHROUGH */
                    379:        case (MDOC_TAIL):
                    380:                /* FALLTHROUGH */
                    381:        case (MDOC_HEAD):
                    382:                p->norm = p->parent->norm;
                    383:                break;
                    384:        default:
                    385:                break;
                    386:        }
                    387:
1.23      kristaps  388:        if ( ! mdoc_valid_pre(mdoc, p))
                    389:                return(0);
1.27      kristaps  390:
                    391:        switch (p->type) {
                    392:        case (MDOC_HEAD):
                    393:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  394:                p->parent->head = p;
1.27      kristaps  395:                break;
                    396:        case (MDOC_TAIL):
                    397:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  398:                p->parent->tail = p;
1.27      kristaps  399:                break;
                    400:        case (MDOC_BODY):
1.152     schwarze  401:                if (p->end)
                    402:                        break;
1.27      kristaps  403:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  404:                p->parent->body = p;
1.27      kristaps  405:                break;
                    406:        default:
                    407:                break;
                    408:        }
                    409:
1.1       kristaps  410:        mdoc->last = p;
1.71      kristaps  411:
                    412:        switch (p->type) {
1.176     kristaps  413:        case (MDOC_TBL):
                    414:                /* FALLTHROUGH */
1.71      kristaps  415:        case (MDOC_TEXT):
                    416:                if ( ! mdoc_valid_post(mdoc))
                    417:                        return(0);
                    418:                break;
                    419:        default:
                    420:                break;
                    421:        }
                    422:
1.23      kristaps  423:        return(1);
1.1       kristaps  424: }
                    425:
                    426:
1.45      kristaps  427: static struct mdoc_node *
1.117     kristaps  428: node_alloc(struct mdoc *m, int line, int pos,
                    429:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  430: {
1.46      kristaps  431:        struct mdoc_node *p;
                    432:
1.113     kristaps  433:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.91      kristaps  434:        p->sec = m->lastsec;
1.73      kristaps  435:        p->line = line;
                    436:        p->pos = pos;
                    437:        p->tok = tok;
1.118     kristaps  438:        p->type = type;
1.150     kristaps  439:
                    440:        /* Flag analysis. */
                    441:
1.153     schwarze  442:        if (MDOC_SYNOPSIS & m->flags)
                    443:                p->flags |= MDOC_SYNPRETTY;
                    444:        else
                    445:                p->flags &= ~MDOC_SYNPRETTY;
1.130     kristaps  446:        if (MDOC_NEWLINE & m->flags)
                    447:                p->flags |= MDOC_LINE;
                    448:        m->flags &= ~MDOC_NEWLINE;
1.150     kristaps  449:
1.46      kristaps  450:        return(p);
1.45      kristaps  451: }
                    452:
                    453:
1.23      kristaps  454: int
1.117     kristaps  455: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.17      kristaps  456: {
                    457:        struct mdoc_node *p;
                    458:
1.91      kristaps  459:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.102     kristaps  460:        if ( ! node_append(m, p))
                    461:                return(0);
                    462:        m->next = MDOC_NEXT_CHILD;
                    463:        return(1);
1.17      kristaps  464: }
                    465:
                    466:
1.23      kristaps  467: int
1.117     kristaps  468: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  469: {
                    470:        struct mdoc_node *p;
                    471:
1.91      kristaps  472:        assert(m->first);
                    473:        assert(m->last);
1.1       kristaps  474:
1.91      kristaps  475:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.102     kristaps  476:        if ( ! node_append(m, p))
                    477:                return(0);
                    478:        m->next = MDOC_NEXT_CHILD;
                    479:        return(1);
1.1       kristaps  480: }
                    481:
                    482:
1.23      kristaps  483: int
1.117     kristaps  484: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  485: {
                    486:        struct mdoc_node *p;
                    487:
1.91      kristaps  488:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.102     kristaps  489:        if ( ! node_append(m, p))
                    490:                return(0);
                    491:        m->next = MDOC_NEXT_CHILD;
1.152     schwarze  492:        return(1);
                    493: }
                    494:
                    495:
                    496: int
                    497: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    498:                struct mdoc_node *body, enum mdoc_endbody end)
                    499: {
                    500:        struct mdoc_node *p;
                    501:
                    502:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    503:        p->pending = body;
                    504:        p->end = end;
                    505:        if ( ! node_append(m, p))
                    506:                return(0);
                    507:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  508:        return(1);
1.1       kristaps  509: }
                    510:
                    511:
1.23      kristaps  512: int
1.91      kristaps  513: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  514:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  515: {
                    516:        struct mdoc_node *p;
                    517:
1.91      kristaps  518:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  519:        p->args = args;
                    520:        if (p->args)
1.53      kristaps  521:                (args->refcnt)++;
1.172     kristaps  522:
                    523:        switch (tok) {
                    524:        case (MDOC_Bd):
                    525:                /* FALLTHROUGH */
                    526:        case (MDOC_Bf):
                    527:                /* FALLTHROUGH */
                    528:        case (MDOC_Bl):
1.173     kristaps  529:                /* FALLTHROUGH */
                    530:        case (MDOC_Rs):
1.172     kristaps  531:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    532:                break;
                    533:        default:
                    534:                break;
                    535:        }
                    536:
1.102     kristaps  537:        if ( ! node_append(m, p))
                    538:                return(0);
                    539:        m->next = MDOC_NEXT_CHILD;
                    540:        return(1);
1.1       kristaps  541: }
                    542:
                    543:
1.23      kristaps  544: int
1.91      kristaps  545: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  546:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  547: {
                    548:        struct mdoc_node *p;
                    549:
1.91      kristaps  550:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.77      kristaps  551:        p->args = args;
                    552:        if (p->args)
1.53      kristaps  553:                (args->refcnt)++;
1.172     kristaps  554:
                    555:        switch (tok) {
                    556:        case (MDOC_An):
                    557:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    558:                break;
                    559:        default:
                    560:                break;
                    561:        }
                    562:
1.102     kristaps  563:        if ( ! node_append(m, p))
                    564:                return(0);
                    565:        m->next = MDOC_NEXT_CHILD;
1.175     kristaps  566:        return(1);
                    567: }
1.1       kristaps  568:
1.124     kristaps  569: int
                    570: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  571: {
1.91      kristaps  572:        struct mdoc_node *n;
1.1       kristaps  573:
1.125     kristaps  574:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.195     kristaps  575:        n->string = roff_strdup(m->roff, p);
1.91      kristaps  576:
1.101     kristaps  577:        if ( ! node_append(m, n))
                    578:                return(0);
1.124     kristaps  579:
1.101     kristaps  580:        m->next = MDOC_NEXT_SIBLING;
                    581:        return(1);
1.91      kristaps  582: }
                    583:
                    584:
1.155     kristaps  585: static void
1.53      kristaps  586: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  587: {
                    588:
1.172     kristaps  589:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  590:                free(p->norm);
1.53      kristaps  591:        if (p->string)
                    592:                free(p->string);
                    593:        if (p->args)
                    594:                mdoc_argv_free(p->args);
1.1       kristaps  595:        free(p);
                    596: }
                    597:
                    598:
1.121     kristaps  599: static void
                    600: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    601: {
                    602:
                    603:        /* Adjust siblings. */
                    604:
                    605:        if (n->prev)
                    606:                n->prev->next = n->next;
                    607:        if (n->next)
                    608:                n->next->prev = n->prev;
                    609:
                    610:        /* Adjust parent. */
                    611:
                    612:        if (n->parent) {
                    613:                n->parent->nchild--;
                    614:                if (n->parent->child == n)
                    615:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  616:                if (n->parent->last == n)
                    617:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  618:        }
                    619:
                    620:        /* Adjust parse point, if applicable. */
                    621:
                    622:        if (m && m->last == n) {
                    623:                if (n->prev) {
                    624:                        m->last = n->prev;
                    625:                        m->next = MDOC_NEXT_SIBLING;
                    626:                } else {
                    627:                        m->last = n->parent;
                    628:                        m->next = MDOC_NEXT_CHILD;
                    629:                }
                    630:        }
                    631:
                    632:        if (m && m->first == n)
                    633:                m->first = NULL;
                    634: }
                    635:
                    636:
1.53      kristaps  637: void
1.121     kristaps  638: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  639: {
                    640:
1.121     kristaps  641:        while (p->child) {
                    642:                assert(p->nchild);
                    643:                mdoc_node_delete(m, p->child);
                    644:        }
                    645:        assert(0 == p->nchild);
1.1       kristaps  646:
1.121     kristaps  647:        mdoc_node_unlink(m, p);
1.53      kristaps  648:        mdoc_node_free(p);
1.1       kristaps  649: }
                    650:
1.193     kristaps  651: #if 0
1.191     kristaps  652: /*
                    653:  * Pre-treat a text line.
                    654:  * Text lines can consist of equations, which must be handled apart from
                    655:  * the regular text.
                    656:  * Thus, use this function to step through a line checking if it has any
                    657:  * equations embedded in it.
                    658:  * This must handle multiple equations AND equations that do not end at
                    659:  * the end-of-line, i.e., will re-enter in the next roff parse.
                    660:  */
                    661: static int
                    662: mdoc_preptext(struct mdoc *m, int line, char *buf, int offs)
                    663: {
                    664:        char            *start, *end;
                    665:        char             delim;
                    666:
                    667:        while ('\0' != buf[offs]) {
                    668:                /* Mark starting position if eqn is set. */
                    669:                start = NULL;
                    670:                if ('\0' != (delim = roff_eqndelim(m->roff)))
                    671:                        if (NULL != (start = strchr(buf + offs, delim)))
                    672:                                *start++ = '\0';
                    673:
                    674:                /* Parse text as normal. */
                    675:                if ( ! mdoc_ptext(m, line, buf, offs))
                    676:                        return(0);
                    677:
                    678:                /* Continue only if an equation exists. */
                    679:                if (NULL == start)
                    680:                        break;
                    681:
                    682:                /* Read past the end of the equation. */
                    683:                offs += start - (buf + offs);
                    684:                assert(start == &buf[offs]);
                    685:                if (NULL != (end = strchr(buf + offs, delim))) {
                    686:                        *end++ = '\0';
                    687:                        while (' ' == *end)
                    688:                                end++;
                    689:                }
                    690:
                    691:                /* Parse the equation itself. */
                    692:                roff_openeqn(m->roff, NULL, line, offs, buf);
                    693:
                    694:                /* Process a finished equation? */
                    695:                if (roff_closeeqn(m->roff))
                    696:                        if ( ! mdoc_addeqn(m, roff_eqn(m->roff)))
                    697:                                return(0);
                    698:                offs += (end - (buf + offs));
                    699:        }
                    700:
                    701:        return(1);
                    702: }
1.193     kristaps  703: #endif
1.1       kristaps  704:
1.53      kristaps  705: /*
                    706:  * Parse free-form text, that is, a line that does not begin with the
                    707:  * control character.
                    708:  */
                    709: static int
1.149     kristaps  710: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  711: {
1.142     kristaps  712:        char             *c, *ws, *end;
                    713:        struct mdoc_node *n;
1.1       kristaps  714:
1.124     kristaps  715:        /* No text before an initial macro. */
                    716:
1.174     kristaps  717:        if (SEC_NONE == m->lastnamed) {
                    718:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    719:                return(1);
                    720:        }
1.142     kristaps  721:
                    722:        assert(m->last);
                    723:        n = m->last;
                    724:
                    725:        /*
1.144     kristaps  726:         * Divert directly to list processing if we're encountering a
1.142     kristaps  727:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  728:         * (a MDOC_BODY means it's already open, in which case we should
                    729:         * process within its context in the normal way).
1.142     kristaps  730:         */
                    731:
1.143     kristaps  732:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  733:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  734:                /* `Bl' is open without any children. */
1.142     kristaps  735:                m->flags |= MDOC_FREECOL;
1.149     kristaps  736:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  737:        }
                    738:
                    739:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    740:                        NULL != n->parent &&
                    741:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  742:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  743:                /* `Bl' has block-level `It' children. */
1.142     kristaps  744:                m->flags |= MDOC_FREECOL;
1.149     kristaps  745:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  746:        }
1.124     kristaps  747:
1.137     schwarze  748:        /*
                    749:         * Search for the beginning of unescaped trailing whitespace (ws)
                    750:         * and for the first character not to be output (end).
                    751:         */
1.139     kristaps  752:
                    753:        /* FIXME: replace with strcspn(). */
1.137     schwarze  754:        ws = NULL;
                    755:        for (c = end = buf + offs; *c; c++) {
                    756:                switch (*c) {
                    757:                case ' ':
                    758:                        if (NULL == ws)
                    759:                                ws = c;
                    760:                        continue;
                    761:                case '\t':
                    762:                        /*
                    763:                         * Always warn about trailing tabs,
                    764:                         * even outside literal context,
                    765:                         * where they should be put on the next line.
                    766:                         */
                    767:                        if (NULL == ws)
                    768:                                ws = c;
                    769:                        /*
                    770:                         * Strip trailing tabs in literal context only;
                    771:                         * outside, they affect the next line.
                    772:                         */
                    773:                        if (MDOC_LITERAL & m->flags)
                    774:                                continue;
                    775:                        break;
                    776:                case '\\':
                    777:                        /* Skip the escaped character, too, if any. */
                    778:                        if (c[1])
                    779:                                c++;
                    780:                        /* FALLTHROUGH */
                    781:                default:
                    782:                        ws = NULL;
                    783:                        break;
                    784:                }
                    785:                end = c + 1;
                    786:        }
                    787:        *end = '\0';
1.91      kristaps  788:
1.137     schwarze  789:        if (ws)
1.174     kristaps  790:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.115     kristaps  791:
1.137     schwarze  792:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.174     kristaps  793:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.124     kristaps  794:
1.119     kristaps  795:                /*
1.165     schwarze  796:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  797:                 * blank lines aren't allowed, but enough manuals assume this
                    798:                 * behaviour that we want to work around it.
1.119     kristaps  799:                 */
1.165     schwarze  800:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.119     kristaps  801:                        return(0);
1.124     kristaps  802:
                    803:                m->next = MDOC_NEXT_SIBLING;
                    804:                return(1);
1.119     kristaps  805:        }
1.68      kristaps  806:
1.137     schwarze  807:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
                    808:                return(0);
1.91      kristaps  809:
1.137     schwarze  810:        if (MDOC_LITERAL & m->flags)
                    811:                return(1);
1.128     kristaps  812:
                    813:        /*
                    814:         * End-of-sentence check.  If the last character is an unescaped
                    815:         * EOS character, then flag the node as being the end of a
                    816:         * sentence.  The front-end will know how to interpret this.
                    817:         */
1.132     kristaps  818:
1.137     schwarze  819:        assert(buf < end);
                    820:
1.159     schwarze  821:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.128     kristaps  822:                m->last->flags |= MDOC_EOS;
                    823:
                    824:        return(1);
1.1       kristaps  825: }
                    826:
                    827:
1.53      kristaps  828: /*
                    829:  * Parse a macro line, that is, a line beginning with the control
                    830:  * character.
                    831:  */
1.155     kristaps  832: static int
1.149     kristaps  833: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  834: {
1.144     kristaps  835:        enum mdoct        tok;
1.188     kristaps  836:        int               i, sv;
1.144     kristaps  837:        char              mac[5];
                    838:        struct mdoc_node *n;
1.1       kristaps  839:
1.188     kristaps  840:        /* Empty post-control lines are ignored. */
1.63      kristaps  841:
1.188     kristaps  842:        if ('"' == buf[offs]) {
                    843:                mdoc_pmsg(m, ln, offs, MANDOCERR_BADCOMMENT);
                    844:                return(1);
                    845:        } else if ('\0' == buf[offs])
1.63      kristaps  846:                return(1);
                    847:
1.188     kristaps  848:        sv = offs;
1.130     kristaps  849:
1.160     kristaps  850:        /*
1.162     schwarze  851:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  852:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  853:         */
1.1       kristaps  854:
1.188     kristaps  855:        i = 0;
                    856:        while (i < 4 && '\0' != buf[offs] &&
                    857:                        ' ' != buf[offs] && '\t' != buf[offs])
                    858:                mac[i++] = buf[offs++];
                    859:
                    860:        mac[i] = '\0';
                    861:
                    862:        tok = (i > 1 || i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  863:
1.163     schwarze  864:        if (MDOC_MAX == tok) {
1.185     kristaps  865:                mandoc_vmsg(MANDOCERR_MACRO, m->parse,
                    866:                                ln, sv, "%s", buf + sv - 1);
1.58      kristaps  867:                return(1);
1.53      kristaps  868:        }
1.1       kristaps  869:
1.160     kristaps  870:        /* Disregard the first trailing tab, if applicable. */
                    871:
1.188     kristaps  872:        if ('\t' == buf[offs])
                    873:                offs++;
1.160     kristaps  874:
                    875:        /* Jump to the next non-whitespace word. */
1.1       kristaps  876:
1.188     kristaps  877:        while (buf[offs] && ' ' == buf[offs])
                    878:                offs++;
1.1       kristaps  879:
1.125     kristaps  880:        /*
                    881:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    882:         * into the parser as "text", so we only warn about spaces here.
                    883:         */
1.115     kristaps  884:
1.188     kristaps  885:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
                    886:                mdoc_pmsg(m, ln, offs - 1, MANDOCERR_EOLNSPACE);
1.115     kristaps  887:
1.144     kristaps  888:        /*
                    889:         * If an initial macro or a list invocation, divert directly
                    890:         * into macro processing.
                    891:         */
                    892:
                    893:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.188     kristaps  894:                if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.144     kristaps  895:                        goto err;
                    896:                return(1);
                    897:        }
                    898:
                    899:        n = m->last;
                    900:        assert(m->last);
                    901:
                    902:        /*
                    903:         * If the first macro of a `Bl -column', open an `It' block
                    904:         * context around the parsed macro.
                    905:         */
                    906:
                    907:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  908:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  909:                m->flags |= MDOC_FREECOL;
1.149     kristaps  910:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  911:                        goto err;
                    912:                return(1);
                    913:        }
                    914:
                    915:        /*
                    916:         * If we're following a block-level `It' within a `Bl -column'
                    917:         * context (perhaps opened in the above block or in ptext()),
                    918:         * then open an `It' block context around the parsed macro.
1.98      kristaps  919:         */
1.144     kristaps  920:
                    921:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    922:                        NULL != n->parent &&
                    923:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  924:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  925:                m->flags |= MDOC_FREECOL;
1.149     kristaps  926:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  927:                        goto err;
                    928:                return(1);
                    929:        }
                    930:
                    931:        /* Normal processing of a macro. */
                    932:
1.188     kristaps  933:        if ( ! mdoc_macro(m, tok, ln, sv, &offs, buf))
1.53      kristaps  934:                goto err;
1.1       kristaps  935:
1.53      kristaps  936:        return(1);
1.1       kristaps  937:
1.53      kristaps  938: err:   /* Error out. */
1.1       kristaps  939:
1.53      kristaps  940:        m->flags |= MDOC_HALT;
                    941:        return(0);
1.1       kristaps  942: }
1.100     kristaps  943:
1.186     kristaps  944: enum mdelim
                    945: mdoc_isdelim(const char *p)
                    946: {
                    947:
                    948:        if ('\0' == p[0])
                    949:                return(DELIM_NONE);
                    950:
                    951:        if ('\0' == p[1])
                    952:                switch (p[0]) {
                    953:                case('('):
                    954:                        /* FALLTHROUGH */
                    955:                case('['):
                    956:                        return(DELIM_OPEN);
                    957:                case('|'):
                    958:                        return(DELIM_MIDDLE);
                    959:                case('.'):
                    960:                        /* FALLTHROUGH */
                    961:                case(','):
                    962:                        /* FALLTHROUGH */
                    963:                case(';'):
                    964:                        /* FALLTHROUGH */
                    965:                case(':'):
                    966:                        /* FALLTHROUGH */
                    967:                case('?'):
                    968:                        /* FALLTHROUGH */
                    969:                case('!'):
                    970:                        /* FALLTHROUGH */
                    971:                case(')'):
                    972:                        /* FALLTHROUGH */
                    973:                case(']'):
                    974:                        return(DELIM_CLOSE);
                    975:                default:
                    976:                        return(DELIM_NONE);
                    977:                }
                    978:
                    979:        if ('\\' != p[0])
                    980:                return(DELIM_NONE);
1.100     kristaps  981:
1.186     kristaps  982:        if (0 == strcmp(p + 1, "."))
                    983:                return(DELIM_CLOSE);
                    984:        if (0 == strcmp(p + 1, "*(Ba"))
                    985:                return(DELIM_MIDDLE);
                    986:
                    987:        return(DELIM_NONE);
                    988: }

CVSweb