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

Annotation of mandoc/mdoc.c, Revision 1.208

1.208   ! schwarze    1: /*     $Id: mdoc.c,v 1.207 2013/12/31 23:23:11 schwarze Exp $ */
1.1       kristaps    2: /*
1.182     schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.208   ! schwarze    4:  * Copyright (c) 2010, 2012, 2013, 2014 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 *
1.203     schwarze  107: mdoc_node(const struct mdoc *mdoc)
1.1       kristaps  108: {
                    109:
1.203     schwarze  110:        assert( ! (MDOC_HALT & mdoc->flags));
                    111:        return(mdoc->first);
1.1       kristaps  112: }
                    113:
                    114:
1.37      kristaps  115: const struct mdoc_meta *
1.203     schwarze  116: mdoc_meta(const struct mdoc *mdoc)
1.37      kristaps  117: {
                    118:
1.203     schwarze  119:        assert( ! (MDOC_HALT & mdoc->flags));
                    120:        return(&mdoc->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 *
1.208   ! schwarze  200: mdoc_alloc(struct roff *roff, struct mparse *parse,
        !           201:        char *defos, int quick)
1.1       kristaps  202: {
                    203:        struct mdoc     *p;
                    204:
1.113     kristaps  205:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    206:
1.185     kristaps  207:        p->parse = parse;
1.197     schwarze  208:        p->defos = defos;
1.208   ! schwarze  209:        p->quick = quick;
1.189     kristaps  210:        p->roff = roff;
1.73      kristaps  211:
1.113     kristaps  212:        mdoc_hash_init();
                    213:        mdoc_alloc1(p);
                    214:        return(p);
1.1       kristaps  215: }
                    216:
                    217:
1.68      kristaps  218: /*
                    219:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  220:  * through to macro_end() in macro.c.
1.68      kristaps  221:  */
1.1       kristaps  222: int
1.203     schwarze  223: mdoc_endparse(struct mdoc *mdoc)
1.20      kristaps  224: {
                    225:
1.203     schwarze  226:        assert( ! (MDOC_HALT & mdoc->flags));
                    227:        if (mdoc_macroend(mdoc))
1.20      kristaps  228:                return(1);
1.203     schwarze  229:        mdoc->flags |= MDOC_HALT;
1.72      kristaps  230:        return(0);
1.181     kristaps  231: }
                    232:
                    233: int
1.203     schwarze  234: mdoc_addeqn(struct mdoc *mdoc, const struct eqn *ep)
1.181     kristaps  235: {
                    236:        struct mdoc_node *n;
                    237:
1.203     schwarze  238:        assert( ! (MDOC_HALT & mdoc->flags));
1.181     kristaps  239:
                    240:        /* No text before an initial macro. */
                    241:
1.203     schwarze  242:        if (SEC_NONE == mdoc->lastnamed) {
                    243:                mdoc_pmsg(mdoc, ep->ln, ep->pos, MANDOCERR_NOTEXT);
1.181     kristaps  244:                return(1);
                    245:        }
                    246:
1.203     schwarze  247:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  248:        n->eqn = ep;
                    249:
1.203     schwarze  250:        if ( ! node_append(mdoc, n))
1.181     kristaps  251:                return(0);
                    252:
1.203     schwarze  253:        mdoc->next = MDOC_NEXT_SIBLING;
1.181     kristaps  254:        return(1);
1.20      kristaps  255: }
                    256:
1.175     kristaps  257: int
1.203     schwarze  258: mdoc_addspan(struct mdoc *mdoc, const struct tbl_span *sp)
1.175     kristaps  259: {
1.180     kristaps  260:        struct mdoc_node *n;
1.175     kristaps  261:
1.203     schwarze  262:        assert( ! (MDOC_HALT & mdoc->flags));
1.175     kristaps  263:
                    264:        /* No text before an initial macro. */
                    265:
1.203     schwarze  266:        if (SEC_NONE == mdoc->lastnamed) {
                    267:                mdoc_pmsg(mdoc, sp->line, 0, MANDOCERR_NOTEXT);
1.175     kristaps  268:                return(1);
                    269:        }
                    270:
1.203     schwarze  271:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.180     kristaps  272:        n->span = sp;
                    273:
1.203     schwarze  274:        if ( ! node_append(mdoc, n))
1.180     kristaps  275:                return(0);
                    276:
1.203     schwarze  277:        mdoc->next = MDOC_NEXT_SIBLING;
1.180     kristaps  278:        return(1);
1.175     kristaps  279: }
                    280:
1.20      kristaps  281:
1.50      kristaps  282: /*
1.53      kristaps  283:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  284:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  285:  */
1.20      kristaps  286: int
1.203     schwarze  287: mdoc_parseln(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  288: {
                    289:
1.203     schwarze  290:        assert( ! (MDOC_HALT & mdoc->flags));
1.50      kristaps  291:
1.203     schwarze  292:        mdoc->flags |= MDOC_NEWLINE;
1.153     schwarze  293:
                    294:        /*
                    295:         * Let the roff nS register switch SYNOPSIS mode early,
                    296:         * such that the parser knows at all times
                    297:         * whether this mode is on or off.
                    298:         * Note that this mode is also switched by the Sh macro.
                    299:         */
1.204     schwarze  300:        if (roff_getreg(mdoc->roff, "nS"))
                    301:                mdoc->flags |= MDOC_SYNOPSIS;
                    302:        else
                    303:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.153     schwarze  304:
1.203     schwarze  305:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
                    306:                        mdoc_pmacro(mdoc, ln, buf, offs) :
                    307:                        mdoc_ptext(mdoc, 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 &&
1.203     schwarze  318:                        MDOC_PBODY & mdoc->flags) {
                    319:                mdoc_pmsg(mdoc, line, ppos, MANDOCERR_BADBODY);
1.174     kristaps  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) &&
1.203     schwarze  326:                        ! (MDOC_PBODY & mdoc->flags)) {
                    327:                mdoc_pmsg(mdoc, line, ppos, MANDOCERR_BADPROLOG);
                    328:                if (NULL == mdoc->meta.msec)
                    329:                        mdoc->meta.msec = mandoc_strdup("1");
                    330:                if (NULL == mdoc->meta.title)
                    331:                        mdoc->meta.title = mandoc_strdup("UNKNOWN");
                    332:                if (NULL == mdoc->meta.vol)
                    333:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    334:                if (NULL == mdoc->meta.os)
                    335:                        mdoc->meta.os = mandoc_strdup("LOCAL");
                    336:                if (NULL == mdoc->meta.date)
                    337:                        mdoc->meta.date = mandoc_normdate
                    338:                                (mdoc->parse, NULL, line, ppos);
                    339:                mdoc->flags |= MDOC_PBODY;
1.120     kristaps  340:        }
1.88      kristaps  341:
1.203     schwarze  342:        return((*mdoc_macros[tok].fp)(mdoc, 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):
1.202     schwarze  378:                if (ENDBODY_NOT != p->end)
                    379:                        break;
1.172     kristaps  380:                /* FALLTHROUGH */
                    381:        case (MDOC_TAIL):
                    382:                /* FALLTHROUGH */
                    383:        case (MDOC_HEAD):
                    384:                p->norm = p->parent->norm;
                    385:                break;
                    386:        default:
                    387:                break;
                    388:        }
                    389:
1.23      kristaps  390:        if ( ! mdoc_valid_pre(mdoc, p))
                    391:                return(0);
1.27      kristaps  392:
                    393:        switch (p->type) {
                    394:        case (MDOC_HEAD):
                    395:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  396:                p->parent->head = p;
1.27      kristaps  397:                break;
                    398:        case (MDOC_TAIL):
                    399:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  400:                p->parent->tail = p;
1.27      kristaps  401:                break;
                    402:        case (MDOC_BODY):
1.152     schwarze  403:                if (p->end)
                    404:                        break;
1.27      kristaps  405:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  406:                p->parent->body = p;
1.27      kristaps  407:                break;
                    408:        default:
                    409:                break;
                    410:        }
                    411:
1.1       kristaps  412:        mdoc->last = p;
1.71      kristaps  413:
                    414:        switch (p->type) {
1.176     kristaps  415:        case (MDOC_TBL):
                    416:                /* FALLTHROUGH */
1.71      kristaps  417:        case (MDOC_TEXT):
                    418:                if ( ! mdoc_valid_post(mdoc))
                    419:                        return(0);
                    420:                break;
                    421:        default:
                    422:                break;
                    423:        }
                    424:
1.23      kristaps  425:        return(1);
1.1       kristaps  426: }
                    427:
                    428:
1.45      kristaps  429: static struct mdoc_node *
1.203     schwarze  430: node_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  431:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  432: {
1.46      kristaps  433:        struct mdoc_node *p;
                    434:
1.113     kristaps  435:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.203     schwarze  436:        p->sec = mdoc->lastsec;
1.73      kristaps  437:        p->line = line;
                    438:        p->pos = pos;
1.206     schwarze  439:        p->lastline = line;
1.73      kristaps  440:        p->tok = tok;
1.118     kristaps  441:        p->type = type;
1.150     kristaps  442:
                    443:        /* Flag analysis. */
                    444:
1.203     schwarze  445:        if (MDOC_SYNOPSIS & mdoc->flags)
1.153     schwarze  446:                p->flags |= MDOC_SYNPRETTY;
                    447:        else
                    448:                p->flags &= ~MDOC_SYNPRETTY;
1.203     schwarze  449:        if (MDOC_NEWLINE & mdoc->flags)
1.130     kristaps  450:                p->flags |= MDOC_LINE;
1.203     schwarze  451:        mdoc->flags &= ~MDOC_NEWLINE;
1.150     kristaps  452:
1.46      kristaps  453:        return(p);
1.45      kristaps  454: }
                    455:
                    456:
1.23      kristaps  457: int
1.203     schwarze  458: mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.17      kristaps  459: {
                    460:        struct mdoc_node *p;
                    461:
1.203     schwarze  462:        p = node_alloc(mdoc, line, pos, tok, MDOC_TAIL);
                    463:        if ( ! node_append(mdoc, p))
1.102     kristaps  464:                return(0);
1.203     schwarze  465:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  466:        return(1);
1.17      kristaps  467: }
                    468:
                    469:
1.23      kristaps  470: int
1.203     schwarze  471: mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  472: {
                    473:        struct mdoc_node *p;
                    474:
1.203     schwarze  475:        assert(mdoc->first);
                    476:        assert(mdoc->last);
1.1       kristaps  477:
1.203     schwarze  478:        p = node_alloc(mdoc, line, pos, tok, MDOC_HEAD);
                    479:        if ( ! node_append(mdoc, p))
1.102     kristaps  480:                return(0);
1.203     schwarze  481:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  482:        return(1);
1.1       kristaps  483: }
                    484:
                    485:
1.23      kristaps  486: int
1.203     schwarze  487: mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  488: {
                    489:        struct mdoc_node *p;
                    490:
1.203     schwarze  491:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
                    492:        if ( ! node_append(mdoc, p))
1.102     kristaps  493:                return(0);
1.203     schwarze  494:        mdoc->next = MDOC_NEXT_CHILD;
1.152     schwarze  495:        return(1);
                    496: }
                    497:
                    498:
                    499: int
1.203     schwarze  500: mdoc_endbody_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok,
1.152     schwarze  501:                struct mdoc_node *body, enum mdoc_endbody end)
                    502: {
                    503:        struct mdoc_node *p;
                    504:
1.203     schwarze  505:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.152     schwarze  506:        p->pending = body;
1.202     schwarze  507:        p->norm = body->norm;
1.152     schwarze  508:        p->end = end;
1.203     schwarze  509:        if ( ! node_append(mdoc, p))
1.152     schwarze  510:                return(0);
1.203     schwarze  511:        mdoc->next = MDOC_NEXT_SIBLING;
1.102     kristaps  512:        return(1);
1.1       kristaps  513: }
                    514:
                    515:
1.23      kristaps  516: int
1.203     schwarze  517: mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  518:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  519: {
                    520:        struct mdoc_node *p;
                    521:
1.203     schwarze  522:        p = node_alloc(mdoc, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  523:        p->args = args;
                    524:        if (p->args)
1.53      kristaps  525:                (args->refcnt)++;
1.172     kristaps  526:
                    527:        switch (tok) {
                    528:        case (MDOC_Bd):
                    529:                /* FALLTHROUGH */
                    530:        case (MDOC_Bf):
                    531:                /* FALLTHROUGH */
                    532:        case (MDOC_Bl):
1.173     kristaps  533:                /* FALLTHROUGH */
                    534:        case (MDOC_Rs):
1.172     kristaps  535:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    536:                break;
                    537:        default:
                    538:                break;
                    539:        }
                    540:
1.203     schwarze  541:        if ( ! node_append(mdoc, p))
1.102     kristaps  542:                return(0);
1.203     schwarze  543:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  544:        return(1);
1.1       kristaps  545: }
                    546:
                    547:
1.23      kristaps  548: int
1.203     schwarze  549: mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  550:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  551: {
                    552:        struct mdoc_node *p;
                    553:
1.203     schwarze  554:        p = node_alloc(mdoc, line, pos, tok, MDOC_ELEM);
1.77      kristaps  555:        p->args = args;
                    556:        if (p->args)
1.53      kristaps  557:                (args->refcnt)++;
1.172     kristaps  558:
                    559:        switch (tok) {
                    560:        case (MDOC_An):
                    561:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    562:                break;
                    563:        default:
                    564:                break;
                    565:        }
                    566:
1.203     schwarze  567:        if ( ! node_append(mdoc, p))
1.102     kristaps  568:                return(0);
1.203     schwarze  569:        mdoc->next = MDOC_NEXT_CHILD;
1.175     kristaps  570:        return(1);
                    571: }
1.1       kristaps  572:
1.124     kristaps  573: int
1.203     schwarze  574: mdoc_word_alloc(struct mdoc *mdoc, int line, int pos, const char *p)
1.1       kristaps  575: {
1.91      kristaps  576:        struct mdoc_node *n;
1.1       kristaps  577:
1.203     schwarze  578:        n = node_alloc(mdoc, line, pos, MDOC_MAX, MDOC_TEXT);
                    579:        n->string = roff_strdup(mdoc->roff, p);
1.91      kristaps  580:
1.203     schwarze  581:        if ( ! node_append(mdoc, n))
1.101     kristaps  582:                return(0);
1.124     kristaps  583:
1.203     schwarze  584:        mdoc->next = MDOC_NEXT_SIBLING;
1.101     kristaps  585:        return(1);
1.91      kristaps  586: }
                    587:
1.205     schwarze  588: void
                    589: mdoc_word_append(struct mdoc *mdoc, const char *p)
                    590: {
                    591:        struct mdoc_node        *n;
                    592:        char                    *addstr, *newstr;
                    593:
                    594:        n = mdoc->last;
                    595:        addstr = roff_strdup(mdoc->roff, p);
                    596:        if (-1 == asprintf(&newstr, "%s %s", n->string, addstr)) {
                    597:                perror(NULL);
                    598:                exit((int)MANDOCLEVEL_SYSERR);
                    599:        }
                    600:        free(addstr);
                    601:        free(n->string);
                    602:        n->string = newstr;
                    603:        mdoc->next = MDOC_NEXT_SIBLING;
                    604: }
1.91      kristaps  605:
1.155     kristaps  606: static void
1.53      kristaps  607: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  608: {
                    609:
1.172     kristaps  610:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  611:                free(p->norm);
1.53      kristaps  612:        if (p->string)
                    613:                free(p->string);
                    614:        if (p->args)
                    615:                mdoc_argv_free(p->args);
1.1       kristaps  616:        free(p);
                    617: }
                    618:
                    619:
1.121     kristaps  620: static void
1.203     schwarze  621: mdoc_node_unlink(struct mdoc *mdoc, struct mdoc_node *n)
1.121     kristaps  622: {
                    623:
                    624:        /* Adjust siblings. */
                    625:
                    626:        if (n->prev)
                    627:                n->prev->next = n->next;
                    628:        if (n->next)
                    629:                n->next->prev = n->prev;
                    630:
                    631:        /* Adjust parent. */
                    632:
                    633:        if (n->parent) {
                    634:                n->parent->nchild--;
                    635:                if (n->parent->child == n)
                    636:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  637:                if (n->parent->last == n)
                    638:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  639:        }
                    640:
                    641:        /* Adjust parse point, if applicable. */
                    642:
1.203     schwarze  643:        if (mdoc && mdoc->last == n) {
1.121     kristaps  644:                if (n->prev) {
1.203     schwarze  645:                        mdoc->last = n->prev;
                    646:                        mdoc->next = MDOC_NEXT_SIBLING;
1.121     kristaps  647:                } else {
1.203     schwarze  648:                        mdoc->last = n->parent;
                    649:                        mdoc->next = MDOC_NEXT_CHILD;
1.121     kristaps  650:                }
                    651:        }
                    652:
1.203     schwarze  653:        if (mdoc && mdoc->first == n)
                    654:                mdoc->first = NULL;
1.121     kristaps  655: }
                    656:
                    657:
1.53      kristaps  658: void
1.203     schwarze  659: mdoc_node_delete(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  660: {
                    661:
1.121     kristaps  662:        while (p->child) {
                    663:                assert(p->nchild);
1.203     schwarze  664:                mdoc_node_delete(mdoc, p->child);
1.121     kristaps  665:        }
                    666:        assert(0 == p->nchild);
1.1       kristaps  667:
1.203     schwarze  668:        mdoc_node_unlink(mdoc, p);
1.53      kristaps  669:        mdoc_node_free(p);
1.201     schwarze  670: }
                    671:
                    672: int
1.203     schwarze  673: mdoc_node_relink(struct mdoc *mdoc, struct mdoc_node *p)
1.201     schwarze  674: {
                    675:
1.203     schwarze  676:        mdoc_node_unlink(mdoc, p);
                    677:        return(node_append(mdoc, p));
1.1       kristaps  678: }
                    679:
1.193     kristaps  680: #if 0
1.191     kristaps  681: /*
                    682:  * Pre-treat a text line.
                    683:  * Text lines can consist of equations, which must be handled apart from
                    684:  * the regular text.
                    685:  * Thus, use this function to step through a line checking if it has any
                    686:  * equations embedded in it.
                    687:  * This must handle multiple equations AND equations that do not end at
                    688:  * the end-of-line, i.e., will re-enter in the next roff parse.
                    689:  */
                    690: static int
1.203     schwarze  691: mdoc_preptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.191     kristaps  692: {
                    693:        char            *start, *end;
                    694:        char             delim;
                    695:
                    696:        while ('\0' != buf[offs]) {
                    697:                /* Mark starting position if eqn is set. */
                    698:                start = NULL;
1.203     schwarze  699:                if ('\0' != (delim = roff_eqndelim(mdoc->roff)))
1.191     kristaps  700:                        if (NULL != (start = strchr(buf + offs, delim)))
                    701:                                *start++ = '\0';
                    702:
                    703:                /* Parse text as normal. */
1.203     schwarze  704:                if ( ! mdoc_ptext(mdoc, line, buf, offs))
1.191     kristaps  705:                        return(0);
                    706:
                    707:                /* Continue only if an equation exists. */
                    708:                if (NULL == start)
                    709:                        break;
                    710:
                    711:                /* Read past the end of the equation. */
                    712:                offs += start - (buf + offs);
                    713:                assert(start == &buf[offs]);
                    714:                if (NULL != (end = strchr(buf + offs, delim))) {
                    715:                        *end++ = '\0';
                    716:                        while (' ' == *end)
                    717:                                end++;
                    718:                }
                    719:
                    720:                /* Parse the equation itself. */
1.203     schwarze  721:                roff_openeqn(mdoc->roff, NULL, line, offs, buf);
1.191     kristaps  722:
                    723:                /* Process a finished equation? */
1.203     schwarze  724:                if (roff_closeeqn(mdoc->roff))
                    725:                        if ( ! mdoc_addeqn(mdoc, roff_eqn(mdoc->roff)))
1.191     kristaps  726:                                return(0);
                    727:                offs += (end - (buf + offs));
                    728:        }
                    729:
                    730:        return(1);
                    731: }
1.193     kristaps  732: #endif
1.1       kristaps  733:
1.53      kristaps  734: /*
                    735:  * Parse free-form text, that is, a line that does not begin with the
                    736:  * control character.
                    737:  */
                    738: static int
1.203     schwarze  739: mdoc_ptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.1       kristaps  740: {
1.142     kristaps  741:        char             *c, *ws, *end;
                    742:        struct mdoc_node *n;
1.1       kristaps  743:
1.124     kristaps  744:        /* No text before an initial macro. */
                    745:
1.203     schwarze  746:        if (SEC_NONE == mdoc->lastnamed) {
                    747:                mdoc_pmsg(mdoc, line, offs, MANDOCERR_NOTEXT);
1.174     kristaps  748:                return(1);
                    749:        }
1.142     kristaps  750:
1.203     schwarze  751:        assert(mdoc->last);
                    752:        n = mdoc->last;
1.142     kristaps  753:
                    754:        /*
1.144     kristaps  755:         * Divert directly to list processing if we're encountering a
1.142     kristaps  756:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  757:         * (a MDOC_BODY means it's already open, in which case we should
                    758:         * process within its context in the normal way).
1.142     kristaps  759:         */
                    760:
1.143     kristaps  761:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  762:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  763:                /* `Bl' is open without any children. */
1.203     schwarze  764:                mdoc->flags |= MDOC_FREECOL;
                    765:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  766:        }
                    767:
                    768:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    769:                        NULL != n->parent &&
                    770:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  771:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  772:                /* `Bl' has block-level `It' children. */
1.203     schwarze  773:                mdoc->flags |= MDOC_FREECOL;
                    774:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  775:        }
1.124     kristaps  776:
1.137     schwarze  777:        /*
                    778:         * Search for the beginning of unescaped trailing whitespace (ws)
                    779:         * and for the first character not to be output (end).
                    780:         */
1.139     kristaps  781:
                    782:        /* FIXME: replace with strcspn(). */
1.137     schwarze  783:        ws = NULL;
                    784:        for (c = end = buf + offs; *c; c++) {
                    785:                switch (*c) {
                    786:                case ' ':
                    787:                        if (NULL == ws)
                    788:                                ws = c;
                    789:                        continue;
                    790:                case '\t':
                    791:                        /*
                    792:                         * Always warn about trailing tabs,
                    793:                         * even outside literal context,
                    794:                         * where they should be put on the next line.
                    795:                         */
                    796:                        if (NULL == ws)
                    797:                                ws = c;
                    798:                        /*
                    799:                         * Strip trailing tabs in literal context only;
                    800:                         * outside, they affect the next line.
                    801:                         */
1.203     schwarze  802:                        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  803:                                continue;
                    804:                        break;
                    805:                case '\\':
                    806:                        /* Skip the escaped character, too, if any. */
                    807:                        if (c[1])
                    808:                                c++;
                    809:                        /* FALLTHROUGH */
                    810:                default:
                    811:                        ws = NULL;
                    812:                        break;
                    813:                }
                    814:                end = c + 1;
                    815:        }
                    816:        *end = '\0';
1.91      kristaps  817:
1.137     schwarze  818:        if (ws)
1.203     schwarze  819:                mdoc_pmsg(mdoc, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.115     kristaps  820:
1.203     schwarze  821:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & mdoc->flags)) {
                    822:                mdoc_pmsg(mdoc, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.124     kristaps  823:
1.119     kristaps  824:                /*
1.165     schwarze  825:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  826:                 * blank lines aren't allowed, but enough manuals assume this
                    827:                 * behaviour that we want to work around it.
1.119     kristaps  828:                 */
1.203     schwarze  829:                if ( ! mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL))
1.119     kristaps  830:                        return(0);
1.124     kristaps  831:
1.203     schwarze  832:                mdoc->next = MDOC_NEXT_SIBLING;
1.199     schwarze  833:
1.203     schwarze  834:                return(mdoc_valid_post(mdoc));
1.119     kristaps  835:        }
1.68      kristaps  836:
1.203     schwarze  837:        if ( ! mdoc_word_alloc(mdoc, line, offs, buf+offs))
1.137     schwarze  838:                return(0);
1.91      kristaps  839:
1.203     schwarze  840:        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  841:                return(1);
1.128     kristaps  842:
                    843:        /*
                    844:         * End-of-sentence check.  If the last character is an unescaped
                    845:         * EOS character, then flag the node as being the end of a
                    846:         * sentence.  The front-end will know how to interpret this.
                    847:         */
1.132     kristaps  848:
1.137     schwarze  849:        assert(buf < end);
                    850:
1.207     schwarze  851:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.203     schwarze  852:                mdoc->last->flags |= MDOC_EOS;
1.128     kristaps  853:
                    854:        return(1);
1.1       kristaps  855: }
                    856:
                    857:
1.53      kristaps  858: /*
                    859:  * Parse a macro line, that is, a line beginning with the control
                    860:  * character.
                    861:  */
1.155     kristaps  862: static int
1.203     schwarze  863: mdoc_pmacro(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  864: {
1.144     kristaps  865:        enum mdoct        tok;
1.188     kristaps  866:        int               i, sv;
1.144     kristaps  867:        char              mac[5];
                    868:        struct mdoc_node *n;
1.1       kristaps  869:
1.188     kristaps  870:        /* Empty post-control lines are ignored. */
1.63      kristaps  871:
1.188     kristaps  872:        if ('"' == buf[offs]) {
1.203     schwarze  873:                mdoc_pmsg(mdoc, ln, offs, MANDOCERR_BADCOMMENT);
1.188     kristaps  874:                return(1);
                    875:        } else if ('\0' == buf[offs])
1.63      kristaps  876:                return(1);
                    877:
1.188     kristaps  878:        sv = offs;
1.130     kristaps  879:
1.160     kristaps  880:        /*
1.162     schwarze  881:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  882:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  883:         */
1.1       kristaps  884:
1.188     kristaps  885:        i = 0;
                    886:        while (i < 4 && '\0' != buf[offs] &&
                    887:                        ' ' != buf[offs] && '\t' != buf[offs])
                    888:                mac[i++] = buf[offs++];
                    889:
                    890:        mac[i] = '\0';
                    891:
                    892:        tok = (i > 1 || i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  893:
1.163     schwarze  894:        if (MDOC_MAX == tok) {
1.203     schwarze  895:                mandoc_vmsg(MANDOCERR_MACRO, mdoc->parse,
1.185     kristaps  896:                                ln, sv, "%s", buf + sv - 1);
1.58      kristaps  897:                return(1);
1.53      kristaps  898:        }
1.1       kristaps  899:
1.160     kristaps  900:        /* Disregard the first trailing tab, if applicable. */
                    901:
1.188     kristaps  902:        if ('\t' == buf[offs])
                    903:                offs++;
1.160     kristaps  904:
                    905:        /* Jump to the next non-whitespace word. */
1.1       kristaps  906:
1.188     kristaps  907:        while (buf[offs] && ' ' == buf[offs])
                    908:                offs++;
1.1       kristaps  909:
1.125     kristaps  910:        /*
                    911:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    912:         * into the parser as "text", so we only warn about spaces here.
                    913:         */
1.115     kristaps  914:
1.188     kristaps  915:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.203     schwarze  916:                mdoc_pmsg(mdoc, ln, offs - 1, MANDOCERR_EOLNSPACE);
1.115     kristaps  917:
1.144     kristaps  918:        /*
                    919:         * If an initial macro or a list invocation, divert directly
                    920:         * into macro processing.
                    921:         */
                    922:
1.203     schwarze  923:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok) {
                    924:                if ( ! mdoc_macro(mdoc, tok, ln, sv, &offs, buf))
1.144     kristaps  925:                        goto err;
                    926:                return(1);
                    927:        }
                    928:
1.203     schwarze  929:        n = mdoc->last;
                    930:        assert(mdoc->last);
1.144     kristaps  931:
                    932:        /*
                    933:         * If the first macro of a `Bl -column', open an `It' block
                    934:         * context around the parsed macro.
                    935:         */
                    936:
                    937:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  938:                        LIST_column == n->norm->Bl.type) {
1.203     schwarze  939:                mdoc->flags |= MDOC_FREECOL;
                    940:                if ( ! mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  941:                        goto err;
                    942:                return(1);
                    943:        }
                    944:
                    945:        /*
                    946:         * If we're following a block-level `It' within a `Bl -column'
                    947:         * context (perhaps opened in the above block or in ptext()),
                    948:         * then open an `It' block context around the parsed macro.
1.98      kristaps  949:         */
1.144     kristaps  950:
                    951:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    952:                        NULL != n->parent &&
                    953:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  954:                        LIST_column == n->parent->norm->Bl.type) {
1.203     schwarze  955:                mdoc->flags |= MDOC_FREECOL;
                    956:                if ( ! mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  957:                        goto err;
                    958:                return(1);
                    959:        }
                    960:
                    961:        /* Normal processing of a macro. */
                    962:
1.203     schwarze  963:        if ( ! mdoc_macro(mdoc, tok, ln, sv, &offs, buf))
1.53      kristaps  964:                goto err;
1.208   ! schwarze  965:
        !           966:        /* In quick mode (for mandocdb), abort after the NAME section. */
        !           967:
        !           968:        if (mdoc->quick && MDOC_Sh == tok &&
        !           969:            SEC_NAME != mdoc->last->sec)
        !           970:                return(2);
1.1       kristaps  971:
1.53      kristaps  972:        return(1);
1.1       kristaps  973:
1.53      kristaps  974: err:   /* Error out. */
1.1       kristaps  975:
1.203     schwarze  976:        mdoc->flags |= MDOC_HALT;
1.53      kristaps  977:        return(0);
1.1       kristaps  978: }
1.100     kristaps  979:
1.186     kristaps  980: enum mdelim
                    981: mdoc_isdelim(const char *p)
                    982: {
                    983:
                    984:        if ('\0' == p[0])
                    985:                return(DELIM_NONE);
                    986:
                    987:        if ('\0' == p[1])
                    988:                switch (p[0]) {
                    989:                case('('):
                    990:                        /* FALLTHROUGH */
                    991:                case('['):
                    992:                        return(DELIM_OPEN);
                    993:                case('|'):
                    994:                        return(DELIM_MIDDLE);
                    995:                case('.'):
                    996:                        /* FALLTHROUGH */
                    997:                case(','):
                    998:                        /* FALLTHROUGH */
                    999:                case(';'):
                   1000:                        /* FALLTHROUGH */
                   1001:                case(':'):
                   1002:                        /* FALLTHROUGH */
                   1003:                case('?'):
                   1004:                        /* FALLTHROUGH */
                   1005:                case('!'):
                   1006:                        /* FALLTHROUGH */
                   1007:                case(')'):
                   1008:                        /* FALLTHROUGH */
                   1009:                case(']'):
                   1010:                        return(DELIM_CLOSE);
                   1011:                default:
                   1012:                        return(DELIM_NONE);
                   1013:                }
                   1014:
                   1015:        if ('\\' != p[0])
                   1016:                return(DELIM_NONE);
1.100     kristaps 1017:
1.186     kristaps 1018:        if (0 == strcmp(p + 1, "."))
                   1019:                return(DELIM_CLOSE);
1.200     schwarze 1020:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.186     kristaps 1021:                return(DELIM_MIDDLE);
                   1022:
                   1023:        return(DELIM_NONE);
                   1024: }

CVSweb