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

Annotation of mandoc/mdoc.c, Revision 1.221

1.221   ! schwarze    1: /*     $Id: mdoc.c,v 1.220 2014/07/09 11:31:43 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>
1.211     schwarze   25: #include <ctype.h>
1.1       kristaps   26: #include <stdarg.h>
1.73      kristaps   27: #include <stdio.h>
1.1       kristaps   28: #include <stdlib.h>
                     29: #include <string.h>
1.120     kristaps   30: #include <time.h>
1.1       kristaps   31:
1.187     kristaps   32: #include "mdoc.h"
1.136     kristaps   33: #include "mandoc.h"
1.209     schwarze   34: #include "mandoc_aux.h"
1.70      kristaps   35: #include "libmdoc.h"
1.113     kristaps   36: #include "libmandoc.h"
1.1       kristaps   37:
1.216     schwarze   38: const  char *const __mdoc_macronames[MDOC_MAX + 1] = {
1.82      kristaps   39:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   40:        "Sh",           "Ss",           "Pp",           "D1",
                     41:        "Dl",           "Bd",           "Ed",           "Bl",
                     42:        "El",           "It",           "Ad",           "An",
                     43:        "Ar",           "Cd",           "Cm",           "Dv",
                     44:        "Er",           "Ev",           "Ex",           "Fa",
                     45:        "Fd",           "Fl",           "Fn",           "Ft",
                     46:        "Ic",           "In",           "Li",           "Nd",
                     47:        "Nm",           "Op",           "Ot",           "Pa",
                     48:        "Rv",           "St",           "Va",           "Vt",
1.114     kristaps   49:        "Xr",           "%A",           "%B",           "%D",
                     50:        "%I",           "%J",           "%N",           "%O",
                     51:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   52:        "Ac",           "Ao",           "Aq",           "At",
                     53:        "Bc",           "Bf",           "Bo",           "Bq",
                     54:        "Bsx",          "Bx",           "Db",           "Dc",
                     55:        "Do",           "Dq",           "Ec",           "Ef",
                     56:        "Em",           "Eo",           "Fx",           "Ms",
                     57:        "No",           "Ns",           "Nx",           "Ox",
                     58:        "Pc",           "Pf",           "Po",           "Pq",
                     59:        "Qc",           "Ql",           "Qo",           "Qq",
                     60:        "Re",           "Rs",           "Sc",           "So",
                     61:        "Sq",           "Sm",           "Sx",           "Sy",
                     62:        "Tn",           "Ux",           "Xc",           "Xo",
                     63:        "Fo",           "Fc",           "Oo",           "Oc",
                     64:        "Bk",           "Ek",           "Bt",           "Hf",
1.82      kristaps   65:        "Fr",           "Ud",           "Lb",           "Lp",
                     66:        "Lk",           "Mt",           "Brq",          "Bro",
1.114     kristaps   67:        "Brc",          "%C",           "Es",           "En",
                     68:        "Dx",           "%Q",           "br",           "sp",
1.216     schwarze   69:        "%U",           "Ta",           "ll",           "text",
1.1       kristaps   70:        };
                     71:
1.213     schwarze   72: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
1.1       kristaps   73:        "split",                "nosplit",              "ragged",
1.213     schwarze   74:        "unfilled",             "literal",              "file",
                     75:        "offset",               "bullet",               "dash",
                     76:        "hyphen",               "item",                 "enum",
                     77:        "tag",                  "diag",                 "hang",
                     78:        "ohang",                "inset",                "column",
                     79:        "width",                "compact",              "std",
1.52      kristaps   80:        "filled",               "words",                "emphasis",
1.108     kristaps   81:        "symbolic",             "nested",               "centered"
1.1       kristaps   82:        };
                     83:
                     84: const  char * const *mdoc_macronames = __mdoc_macronames;
                     85: const  char * const *mdoc_argnames = __mdoc_argnames;
                     86:
1.121     kristaps   87: static void              mdoc_node_free(struct mdoc_node *);
1.213     schwarze   88: static void              mdoc_node_unlink(struct mdoc *,
1.121     kristaps   89:                                struct mdoc_node *);
1.73      kristaps   90: static void              mdoc_free1(struct mdoc *);
1.113     kristaps   91: static void              mdoc_alloc1(struct mdoc *);
1.213     schwarze   92: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.117     kristaps   93:                                enum mdoct, enum mdoc_type);
1.213     schwarze   94: static int               node_append(struct mdoc *,
1.71      kristaps   95:                                struct mdoc_node *);
1.193     kristaps   96: #if 0
1.191     kristaps   97: static int               mdoc_preptext(struct mdoc *, int, char *, int);
1.193     kristaps   98: #endif
1.149     kristaps   99: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                    100: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.88      kristaps  101:
1.213     schwarze  102:
1.1       kristaps  103: const struct mdoc_node *
1.203     schwarze  104: mdoc_node(const struct mdoc *mdoc)
1.1       kristaps  105: {
                    106:
1.203     schwarze  107:        return(mdoc->first);
1.1       kristaps  108: }
                    109:
1.37      kristaps  110: const struct mdoc_meta *
1.203     schwarze  111: mdoc_meta(const struct mdoc *mdoc)
1.37      kristaps  112: {
                    113:
1.203     schwarze  114:        return(&mdoc->meta);
1.37      kristaps  115: }
                    116:
1.85      kristaps  117: /*
                    118:  * Frees volatile resources (parse tree, meta-data, fields).
                    119:  */
1.73      kristaps  120: static void
                    121: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  122: {
                    123:
                    124:        if (mdoc->first)
1.121     kristaps  125:                mdoc_node_delete(mdoc, mdoc->first);
1.67      kristaps  126:        if (mdoc->meta.title)
                    127:                free(mdoc->meta.title);
                    128:        if (mdoc->meta.os)
                    129:                free(mdoc->meta.os);
                    130:        if (mdoc->meta.name)
                    131:                free(mdoc->meta.name);
                    132:        if (mdoc->meta.arch)
                    133:                free(mdoc->meta.arch);
                    134:        if (mdoc->meta.vol)
                    135:                free(mdoc->meta.vol);
1.133     kristaps  136:        if (mdoc->meta.msec)
                    137:                free(mdoc->meta.msec);
1.183     kristaps  138:        if (mdoc->meta.date)
                    139:                free(mdoc->meta.date);
1.73      kristaps  140: }
                    141:
1.85      kristaps  142: /*
                    143:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    144:  */
1.113     kristaps  145: static void
1.73      kristaps  146: mdoc_alloc1(struct mdoc *mdoc)
                    147: {
1.67      kristaps  148:
1.112     kristaps  149:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  150:        mdoc->flags = 0;
1.85      kristaps  151:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  152:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  153:        mdoc->first = mdoc->last;
1.67      kristaps  154:        mdoc->last->type = MDOC_ROOT;
1.196     schwarze  155:        mdoc->last->tok = MDOC_MAX;
1.67      kristaps  156:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  157: }
                    158:
                    159: /*
1.85      kristaps  160:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    161:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    162:  * and the parser is ready for re-invocation on a new tree; however,
                    163:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  164:  */
1.113     kristaps  165: void
1.73      kristaps  166: mdoc_reset(struct mdoc *mdoc)
                    167: {
                    168:
                    169:        mdoc_free1(mdoc);
1.113     kristaps  170:        mdoc_alloc1(mdoc);
1.67      kristaps  171: }
                    172:
1.68      kristaps  173: /*
1.85      kristaps  174:  * Completely free up all volatile and non-volatile parse resources.
                    175:  * After invocation, the pointer is no longer usable.
1.68      kristaps  176:  */
1.67      kristaps  177: void
1.38      kristaps  178: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  179: {
                    180:
1.73      kristaps  181:        mdoc_free1(mdoc);
1.1       kristaps  182:        free(mdoc);
                    183: }
                    184:
1.85      kristaps  185: /*
1.213     schwarze  186:  * Allocate volatile and non-volatile parse resources.
1.85      kristaps  187:  */
1.1       kristaps  188: struct mdoc *
1.208     schwarze  189: mdoc_alloc(struct roff *roff, struct mparse *parse,
1.220     schwarze  190:        const char *defos, int quick)
1.1       kristaps  191: {
                    192:        struct mdoc     *p;
                    193:
1.113     kristaps  194:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    195:
1.185     kristaps  196:        p->parse = parse;
1.197     schwarze  197:        p->defos = defos;
1.208     schwarze  198:        p->quick = quick;
1.189     kristaps  199:        p->roff = roff;
1.73      kristaps  200:
1.113     kristaps  201:        mdoc_hash_init();
                    202:        mdoc_alloc1(p);
                    203:        return(p);
1.1       kristaps  204: }
                    205:
                    206: int
1.203     schwarze  207: mdoc_endparse(struct mdoc *mdoc)
1.20      kristaps  208: {
                    209:
1.221   ! schwarze  210:        return(mdoc_macroend(mdoc));
1.181     kristaps  211: }
                    212:
                    213: int
1.203     schwarze  214: mdoc_addeqn(struct mdoc *mdoc, const struct eqn *ep)
1.181     kristaps  215: {
                    216:        struct mdoc_node *n;
                    217:
1.203     schwarze  218:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  219:        n->eqn = ep;
                    220:
1.203     schwarze  221:        if ( ! node_append(mdoc, n))
1.181     kristaps  222:                return(0);
                    223:
1.203     schwarze  224:        mdoc->next = MDOC_NEXT_SIBLING;
1.181     kristaps  225:        return(1);
1.20      kristaps  226: }
                    227:
1.175     kristaps  228: int
1.203     schwarze  229: mdoc_addspan(struct mdoc *mdoc, const struct tbl_span *sp)
1.175     kristaps  230: {
1.180     kristaps  231:        struct mdoc_node *n;
1.175     kristaps  232:
1.203     schwarze  233:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.180     kristaps  234:        n->span = sp;
                    235:
1.203     schwarze  236:        if ( ! node_append(mdoc, n))
1.180     kristaps  237:                return(0);
                    238:
1.203     schwarze  239:        mdoc->next = MDOC_NEXT_SIBLING;
1.180     kristaps  240:        return(1);
1.175     kristaps  241: }
                    242:
1.50      kristaps  243: /*
1.53      kristaps  244:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  245:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  246:  */
1.20      kristaps  247: int
1.203     schwarze  248: mdoc_parseln(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  249: {
                    250:
1.203     schwarze  251:        mdoc->flags |= MDOC_NEWLINE;
1.153     schwarze  252:
                    253:        /*
                    254:         * Let the roff nS register switch SYNOPSIS mode early,
                    255:         * such that the parser knows at all times
                    256:         * whether this mode is on or off.
                    257:         * Note that this mode is also switched by the Sh macro.
                    258:         */
1.204     schwarze  259:        if (roff_getreg(mdoc->roff, "nS"))
                    260:                mdoc->flags |= MDOC_SYNOPSIS;
                    261:        else
                    262:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.153     schwarze  263:
1.203     schwarze  264:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
1.213     schwarze  265:            mdoc_pmacro(mdoc, ln, buf, offs) :
                    266:            mdoc_ptext(mdoc, ln, buf, offs));
1.1       kristaps  267: }
                    268:
1.88      kristaps  269: int
1.148     kristaps  270: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  271: {
1.122     kristaps  272:        assert(tok < MDOC_MAX);
                    273:
                    274:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  275:
1.213     schwarze  276:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
                    277:            MDOC_PBODY & mdoc->flags) {
1.215     schwarze  278:                mandoc_vmsg(MANDOCERR_PROLOG_ONLY, mdoc->parse,
                    279:                    line, ppos, "%s", mdoc_macronames[tok]);
1.174     kristaps  280:                return(1);
                    281:        }
1.122     kristaps  282:
                    283:        /* If we're in the prologue, deny "body" macros.  */
                    284:
1.213     schwarze  285:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    286:             ! (MDOC_PBODY & mdoc->flags)) {
1.215     schwarze  287:                mandoc_vmsg(MANDOCERR_PROLOG_BAD, mdoc->parse,
                    288:                    line, ppos, "%s", mdoc_macronames[tok]);
1.203     schwarze  289:                if (NULL == mdoc->meta.msec)
                    290:                        mdoc->meta.msec = mandoc_strdup("1");
                    291:                if (NULL == mdoc->meta.title)
                    292:                        mdoc->meta.title = mandoc_strdup("UNKNOWN");
                    293:                if (NULL == mdoc->meta.vol)
                    294:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    295:                if (NULL == mdoc->meta.os)
                    296:                        mdoc->meta.os = mandoc_strdup("LOCAL");
                    297:                if (NULL == mdoc->meta.date)
                    298:                        mdoc->meta.date = mandoc_normdate
                    299:                                (mdoc->parse, NULL, line, ppos);
                    300:                mdoc->flags |= MDOC_PBODY;
1.120     kristaps  301:        }
1.88      kristaps  302:
1.203     schwarze  303:        return((*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf));
1.73      kristaps  304: }
                    305:
                    306:
                    307: static int
                    308: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  309: {
                    310:
1.25      kristaps  311:        assert(mdoc->last);
                    312:        assert(mdoc->first);
                    313:        assert(MDOC_ROOT != p->type);
1.1       kristaps  314:
1.13      kristaps  315:        switch (mdoc->next) {
1.213     schwarze  316:        case MDOC_NEXT_SIBLING:
1.6       kristaps  317:                mdoc->last->next = p;
                    318:                p->prev = mdoc->last;
1.13      kristaps  319:                p->parent = mdoc->last->parent;
1.1       kristaps  320:                break;
1.213     schwarze  321:        case MDOC_NEXT_CHILD:
1.13      kristaps  322:                mdoc->last->child = p;
1.1       kristaps  323:                p->parent = mdoc->last;
                    324:                break;
                    325:        default:
1.13      kristaps  326:                abort();
                    327:                /* NOTREACHED */
1.1       kristaps  328:        }
                    329:
1.86      kristaps  330:        p->parent->nchild++;
                    331:
1.172     kristaps  332:        /*
                    333:         * Copy over the normalised-data pointer of our parent.  Not
                    334:         * everybody has one, but copying a null pointer is fine.
                    335:         */
                    336:
                    337:        switch (p->type) {
1.213     schwarze  338:        case MDOC_BODY:
1.202     schwarze  339:                if (ENDBODY_NOT != p->end)
                    340:                        break;
1.172     kristaps  341:                /* FALLTHROUGH */
1.213     schwarze  342:        case MDOC_TAIL:
1.172     kristaps  343:                /* FALLTHROUGH */
1.213     schwarze  344:        case MDOC_HEAD:
1.172     kristaps  345:                p->norm = p->parent->norm;
                    346:                break;
                    347:        default:
                    348:                break;
                    349:        }
                    350:
1.23      kristaps  351:        if ( ! mdoc_valid_pre(mdoc, p))
                    352:                return(0);
1.27      kristaps  353:
                    354:        switch (p->type) {
1.213     schwarze  355:        case MDOC_HEAD:
1.27      kristaps  356:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  357:                p->parent->head = p;
1.27      kristaps  358:                break;
1.213     schwarze  359:        case MDOC_TAIL:
1.27      kristaps  360:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  361:                p->parent->tail = p;
1.27      kristaps  362:                break;
1.213     schwarze  363:        case MDOC_BODY:
1.152     schwarze  364:                if (p->end)
                    365:                        break;
1.27      kristaps  366:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  367:                p->parent->body = p;
1.27      kristaps  368:                break;
                    369:        default:
                    370:                break;
                    371:        }
                    372:
1.1       kristaps  373:        mdoc->last = p;
1.71      kristaps  374:
                    375:        switch (p->type) {
1.213     schwarze  376:        case MDOC_TBL:
1.176     kristaps  377:                /* FALLTHROUGH */
1.213     schwarze  378:        case MDOC_TEXT:
1.71      kristaps  379:                if ( ! mdoc_valid_post(mdoc))
                    380:                        return(0);
                    381:                break;
                    382:        default:
                    383:                break;
                    384:        }
                    385:
1.23      kristaps  386:        return(1);
1.1       kristaps  387: }
                    388:
1.45      kristaps  389: static struct mdoc_node *
1.213     schwarze  390: node_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  391:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  392: {
1.46      kristaps  393:        struct mdoc_node *p;
                    394:
1.113     kristaps  395:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.203     schwarze  396:        p->sec = mdoc->lastsec;
1.73      kristaps  397:        p->line = line;
                    398:        p->pos = pos;
1.206     schwarze  399:        p->lastline = line;
1.73      kristaps  400:        p->tok = tok;
1.118     kristaps  401:        p->type = type;
1.150     kristaps  402:
                    403:        /* Flag analysis. */
                    404:
1.203     schwarze  405:        if (MDOC_SYNOPSIS & mdoc->flags)
1.153     schwarze  406:                p->flags |= MDOC_SYNPRETTY;
                    407:        else
                    408:                p->flags &= ~MDOC_SYNPRETTY;
1.203     schwarze  409:        if (MDOC_NEWLINE & mdoc->flags)
1.130     kristaps  410:                p->flags |= MDOC_LINE;
1.203     schwarze  411:        mdoc->flags &= ~MDOC_NEWLINE;
1.150     kristaps  412:
1.46      kristaps  413:        return(p);
1.45      kristaps  414: }
                    415:
1.23      kristaps  416: int
1.203     schwarze  417: mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.17      kristaps  418: {
                    419:        struct mdoc_node *p;
                    420:
1.203     schwarze  421:        p = node_alloc(mdoc, line, pos, tok, MDOC_TAIL);
                    422:        if ( ! node_append(mdoc, p))
1.102     kristaps  423:                return(0);
1.203     schwarze  424:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  425:        return(1);
1.17      kristaps  426: }
                    427:
1.23      kristaps  428: int
1.203     schwarze  429: mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  430: {
                    431:        struct mdoc_node *p;
                    432:
1.203     schwarze  433:        assert(mdoc->first);
                    434:        assert(mdoc->last);
1.1       kristaps  435:
1.203     schwarze  436:        p = node_alloc(mdoc, line, pos, tok, MDOC_HEAD);
                    437:        if ( ! node_append(mdoc, p))
1.102     kristaps  438:                return(0);
1.203     schwarze  439:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  440:        return(1);
1.1       kristaps  441: }
                    442:
1.23      kristaps  443: int
1.203     schwarze  444: mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  445: {
                    446:        struct mdoc_node *p;
                    447:
1.203     schwarze  448:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
                    449:        if ( ! node_append(mdoc, p))
1.102     kristaps  450:                return(0);
1.203     schwarze  451:        mdoc->next = MDOC_NEXT_CHILD;
1.152     schwarze  452:        return(1);
                    453: }
                    454:
                    455: int
1.203     schwarze  456: mdoc_endbody_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok,
1.152     schwarze  457:                struct mdoc_node *body, enum mdoc_endbody end)
                    458: {
                    459:        struct mdoc_node *p;
                    460:
1.203     schwarze  461:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.152     schwarze  462:        p->pending = body;
1.202     schwarze  463:        p->norm = body->norm;
1.152     schwarze  464:        p->end = end;
1.203     schwarze  465:        if ( ! node_append(mdoc, p))
1.152     schwarze  466:                return(0);
1.203     schwarze  467:        mdoc->next = MDOC_NEXT_SIBLING;
1.102     kristaps  468:        return(1);
1.1       kristaps  469: }
                    470:
1.23      kristaps  471: int
1.213     schwarze  472: mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  473:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  474: {
                    475:        struct mdoc_node *p;
                    476:
1.203     schwarze  477:        p = node_alloc(mdoc, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  478:        p->args = args;
                    479:        if (p->args)
1.53      kristaps  480:                (args->refcnt)++;
1.172     kristaps  481:
                    482:        switch (tok) {
1.213     schwarze  483:        case MDOC_Bd:
1.172     kristaps  484:                /* FALLTHROUGH */
1.213     schwarze  485:        case MDOC_Bf:
1.172     kristaps  486:                /* FALLTHROUGH */
1.213     schwarze  487:        case MDOC_Bl:
1.217     schwarze  488:                /* FALLTHROUGH */
                    489:        case MDOC_En:
1.173     kristaps  490:                /* FALLTHROUGH */
1.213     schwarze  491:        case MDOC_Rs:
1.172     kristaps  492:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    493:                break;
                    494:        default:
                    495:                break;
                    496:        }
                    497:
1.203     schwarze  498:        if ( ! node_append(mdoc, p))
1.102     kristaps  499:                return(0);
1.203     schwarze  500:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  501:        return(1);
1.1       kristaps  502: }
                    503:
1.23      kristaps  504: int
1.213     schwarze  505: mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  506:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  507: {
                    508:        struct mdoc_node *p;
                    509:
1.203     schwarze  510:        p = node_alloc(mdoc, line, pos, tok, MDOC_ELEM);
1.77      kristaps  511:        p->args = args;
                    512:        if (p->args)
1.53      kristaps  513:                (args->refcnt)++;
1.172     kristaps  514:
                    515:        switch (tok) {
1.213     schwarze  516:        case MDOC_An:
1.172     kristaps  517:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    518:                break;
                    519:        default:
                    520:                break;
                    521:        }
                    522:
1.203     schwarze  523:        if ( ! node_append(mdoc, p))
1.102     kristaps  524:                return(0);
1.203     schwarze  525:        mdoc->next = MDOC_NEXT_CHILD;
1.175     kristaps  526:        return(1);
                    527: }
1.1       kristaps  528:
1.124     kristaps  529: int
1.203     schwarze  530: mdoc_word_alloc(struct mdoc *mdoc, int line, int pos, const char *p)
1.1       kristaps  531: {
1.91      kristaps  532:        struct mdoc_node *n;
1.1       kristaps  533:
1.203     schwarze  534:        n = node_alloc(mdoc, line, pos, MDOC_MAX, MDOC_TEXT);
                    535:        n->string = roff_strdup(mdoc->roff, p);
1.91      kristaps  536:
1.203     schwarze  537:        if ( ! node_append(mdoc, n))
1.101     kristaps  538:                return(0);
1.124     kristaps  539:
1.203     schwarze  540:        mdoc->next = MDOC_NEXT_SIBLING;
1.101     kristaps  541:        return(1);
1.91      kristaps  542: }
                    543:
1.205     schwarze  544: void
                    545: mdoc_word_append(struct mdoc *mdoc, const char *p)
                    546: {
                    547:        struct mdoc_node        *n;
                    548:        char                    *addstr, *newstr;
                    549:
                    550:        n = mdoc->last;
                    551:        addstr = roff_strdup(mdoc->roff, p);
1.210     schwarze  552:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.205     schwarze  553:        free(addstr);
                    554:        free(n->string);
                    555:        n->string = newstr;
                    556:        mdoc->next = MDOC_NEXT_SIBLING;
                    557: }
1.91      kristaps  558:
1.155     kristaps  559: static void
1.53      kristaps  560: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  561: {
                    562:
1.172     kristaps  563:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  564:                free(p->norm);
1.53      kristaps  565:        if (p->string)
                    566:                free(p->string);
                    567:        if (p->args)
                    568:                mdoc_argv_free(p->args);
1.1       kristaps  569:        free(p);
                    570: }
                    571:
1.121     kristaps  572: static void
1.203     schwarze  573: mdoc_node_unlink(struct mdoc *mdoc, struct mdoc_node *n)
1.121     kristaps  574: {
                    575:
                    576:        /* Adjust siblings. */
                    577:
                    578:        if (n->prev)
                    579:                n->prev->next = n->next;
                    580:        if (n->next)
                    581:                n->next->prev = n->prev;
                    582:
                    583:        /* Adjust parent. */
                    584:
                    585:        if (n->parent) {
                    586:                n->parent->nchild--;
                    587:                if (n->parent->child == n)
                    588:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  589:                if (n->parent->last == n)
                    590:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  591:        }
                    592:
                    593:        /* Adjust parse point, if applicable. */
                    594:
1.203     schwarze  595:        if (mdoc && mdoc->last == n) {
1.121     kristaps  596:                if (n->prev) {
1.203     schwarze  597:                        mdoc->last = n->prev;
                    598:                        mdoc->next = MDOC_NEXT_SIBLING;
1.121     kristaps  599:                } else {
1.203     schwarze  600:                        mdoc->last = n->parent;
                    601:                        mdoc->next = MDOC_NEXT_CHILD;
1.121     kristaps  602:                }
                    603:        }
                    604:
1.203     schwarze  605:        if (mdoc && mdoc->first == n)
                    606:                mdoc->first = NULL;
1.121     kristaps  607: }
                    608:
1.53      kristaps  609: void
1.203     schwarze  610: mdoc_node_delete(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  611: {
                    612:
1.121     kristaps  613:        while (p->child) {
                    614:                assert(p->nchild);
1.203     schwarze  615:                mdoc_node_delete(mdoc, p->child);
1.121     kristaps  616:        }
                    617:        assert(0 == p->nchild);
1.1       kristaps  618:
1.203     schwarze  619:        mdoc_node_unlink(mdoc, p);
1.53      kristaps  620:        mdoc_node_free(p);
1.201     schwarze  621: }
                    622:
                    623: int
1.203     schwarze  624: mdoc_node_relink(struct mdoc *mdoc, struct mdoc_node *p)
1.201     schwarze  625: {
                    626:
1.203     schwarze  627:        mdoc_node_unlink(mdoc, p);
                    628:        return(node_append(mdoc, p));
1.1       kristaps  629: }
                    630:
1.193     kristaps  631: #if 0
1.191     kristaps  632: /*
                    633:  * Pre-treat a text line.
                    634:  * Text lines can consist of equations, which must be handled apart from
                    635:  * the regular text.
                    636:  * Thus, use this function to step through a line checking if it has any
                    637:  * equations embedded in it.
                    638:  * This must handle multiple equations AND equations that do not end at
                    639:  * the end-of-line, i.e., will re-enter in the next roff parse.
                    640:  */
                    641: static int
1.203     schwarze  642: mdoc_preptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.191     kristaps  643: {
                    644:        char            *start, *end;
                    645:        char             delim;
                    646:
                    647:        while ('\0' != buf[offs]) {
                    648:                /* Mark starting position if eqn is set. */
                    649:                start = NULL;
1.203     schwarze  650:                if ('\0' != (delim = roff_eqndelim(mdoc->roff)))
1.191     kristaps  651:                        if (NULL != (start = strchr(buf + offs, delim)))
                    652:                                *start++ = '\0';
                    653:
                    654:                /* Parse text as normal. */
1.203     schwarze  655:                if ( ! mdoc_ptext(mdoc, line, buf, offs))
1.191     kristaps  656:                        return(0);
                    657:
                    658:                /* Continue only if an equation exists. */
                    659:                if (NULL == start)
                    660:                        break;
                    661:
                    662:                /* Read past the end of the equation. */
                    663:                offs += start - (buf + offs);
                    664:                assert(start == &buf[offs]);
                    665:                if (NULL != (end = strchr(buf + offs, delim))) {
                    666:                        *end++ = '\0';
                    667:                        while (' ' == *end)
                    668:                                end++;
                    669:                }
                    670:
                    671:                /* Parse the equation itself. */
1.203     schwarze  672:                roff_openeqn(mdoc->roff, NULL, line, offs, buf);
1.191     kristaps  673:
                    674:                /* Process a finished equation? */
1.203     schwarze  675:                if (roff_closeeqn(mdoc->roff))
                    676:                        if ( ! mdoc_addeqn(mdoc, roff_eqn(mdoc->roff)))
1.191     kristaps  677:                                return(0);
                    678:                offs += (end - (buf + offs));
1.213     schwarze  679:        }
1.191     kristaps  680:
                    681:        return(1);
                    682: }
1.193     kristaps  683: #endif
1.1       kristaps  684:
1.53      kristaps  685: /*
                    686:  * Parse free-form text, that is, a line that does not begin with the
                    687:  * control character.
                    688:  */
                    689: static int
1.203     schwarze  690: mdoc_ptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.1       kristaps  691: {
1.142     kristaps  692:        char             *c, *ws, *end;
                    693:        struct mdoc_node *n;
                    694:
1.203     schwarze  695:        assert(mdoc->last);
                    696:        n = mdoc->last;
1.142     kristaps  697:
                    698:        /*
1.144     kristaps  699:         * Divert directly to list processing if we're encountering a
1.142     kristaps  700:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  701:         * (a MDOC_BODY means it's already open, in which case we should
                    702:         * process within its context in the normal way).
1.142     kristaps  703:         */
                    704:
1.143     kristaps  705:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  706:            LIST_column == n->norm->Bl.type) {
1.144     kristaps  707:                /* `Bl' is open without any children. */
1.203     schwarze  708:                mdoc->flags |= MDOC_FREECOL;
                    709:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  710:        }
                    711:
                    712:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  713:            NULL != n->parent &&
                    714:            MDOC_Bl == n->parent->tok &&
                    715:            LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  716:                /* `Bl' has block-level `It' children. */
1.203     schwarze  717:                mdoc->flags |= MDOC_FREECOL;
                    718:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  719:        }
1.124     kristaps  720:
1.137     schwarze  721:        /*
                    722:         * Search for the beginning of unescaped trailing whitespace (ws)
                    723:         * and for the first character not to be output (end).
                    724:         */
1.139     kristaps  725:
                    726:        /* FIXME: replace with strcspn(). */
1.137     schwarze  727:        ws = NULL;
                    728:        for (c = end = buf + offs; *c; c++) {
                    729:                switch (*c) {
                    730:                case ' ':
                    731:                        if (NULL == ws)
                    732:                                ws = c;
                    733:                        continue;
                    734:                case '\t':
                    735:                        /*
                    736:                         * Always warn about trailing tabs,
                    737:                         * even outside literal context,
                    738:                         * where they should be put on the next line.
                    739:                         */
                    740:                        if (NULL == ws)
                    741:                                ws = c;
                    742:                        /*
                    743:                         * Strip trailing tabs in literal context only;
                    744:                         * outside, they affect the next line.
                    745:                         */
1.203     schwarze  746:                        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  747:                                continue;
                    748:                        break;
                    749:                case '\\':
                    750:                        /* Skip the escaped character, too, if any. */
                    751:                        if (c[1])
                    752:                                c++;
                    753:                        /* FALLTHROUGH */
                    754:                default:
                    755:                        ws = NULL;
                    756:                        break;
                    757:                }
                    758:                end = c + 1;
                    759:        }
                    760:        *end = '\0';
1.91      kristaps  761:
1.137     schwarze  762:        if (ws)
1.218     schwarze  763:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    764:                    line, (int)(ws-buf), NULL);
1.115     kristaps  765:
1.203     schwarze  766:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & mdoc->flags)) {
1.218     schwarze  767:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    768:                    line, (int)(c - buf), NULL);
1.124     kristaps  769:
1.119     kristaps  770:                /*
1.165     schwarze  771:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  772:                 * blank lines aren't allowed, but enough manuals assume this
                    773:                 * behaviour that we want to work around it.
1.119     kristaps  774:                 */
1.203     schwarze  775:                if ( ! mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL))
1.119     kristaps  776:                        return(0);
1.124     kristaps  777:
1.203     schwarze  778:                mdoc->next = MDOC_NEXT_SIBLING;
1.199     schwarze  779:
1.203     schwarze  780:                return(mdoc_valid_post(mdoc));
1.119     kristaps  781:        }
1.68      kristaps  782:
1.203     schwarze  783:        if ( ! mdoc_word_alloc(mdoc, line, offs, buf+offs))
1.137     schwarze  784:                return(0);
1.91      kristaps  785:
1.203     schwarze  786:        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  787:                return(1);
1.128     kristaps  788:
                    789:        /*
                    790:         * End-of-sentence check.  If the last character is an unescaped
                    791:         * EOS character, then flag the node as being the end of a
                    792:         * sentence.  The front-end will know how to interpret this.
                    793:         */
1.132     kristaps  794:
1.137     schwarze  795:        assert(buf < end);
                    796:
1.207     schwarze  797:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.203     schwarze  798:                mdoc->last->flags |= MDOC_EOS;
1.128     kristaps  799:
                    800:        return(1);
1.1       kristaps  801: }
                    802:
1.53      kristaps  803: /*
                    804:  * Parse a macro line, that is, a line beginning with the control
                    805:  * character.
                    806:  */
1.155     kristaps  807: static int
1.203     schwarze  808: mdoc_pmacro(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  809: {
1.144     kristaps  810:        enum mdoct        tok;
1.188     kristaps  811:        int               i, sv;
1.144     kristaps  812:        char              mac[5];
                    813:        struct mdoc_node *n;
1.1       kristaps  814:
1.188     kristaps  815:        /* Empty post-control lines are ignored. */
1.63      kristaps  816:
1.188     kristaps  817:        if ('"' == buf[offs]) {
1.218     schwarze  818:                mandoc_msg(MANDOCERR_COMMENT_BAD, mdoc->parse,
                    819:                    ln, offs, NULL);
1.188     kristaps  820:                return(1);
                    821:        } else if ('\0' == buf[offs])
1.63      kristaps  822:                return(1);
                    823:
1.188     kristaps  824:        sv = offs;
1.130     kristaps  825:
1.213     schwarze  826:        /*
1.162     schwarze  827:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  828:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  829:         */
1.1       kristaps  830:
1.188     kristaps  831:        i = 0;
1.213     schwarze  832:        while (i < 4 && '\0' != buf[offs] && ' ' != buf[offs] &&
                    833:            '\t' != buf[offs])
1.188     kristaps  834:                mac[i++] = buf[offs++];
                    835:
                    836:        mac[i] = '\0';
                    837:
1.214     schwarze  838:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  839:
1.163     schwarze  840:        if (MDOC_MAX == tok) {
1.213     schwarze  841:                mandoc_vmsg(MANDOCERR_MACRO, mdoc->parse,
                    842:                    ln, sv, "%s", buf + sv - 1);
1.58      kristaps  843:                return(1);
1.53      kristaps  844:        }
1.1       kristaps  845:
1.160     kristaps  846:        /* Disregard the first trailing tab, if applicable. */
                    847:
1.188     kristaps  848:        if ('\t' == buf[offs])
                    849:                offs++;
1.160     kristaps  850:
                    851:        /* Jump to the next non-whitespace word. */
1.1       kristaps  852:
1.188     kristaps  853:        while (buf[offs] && ' ' == buf[offs])
                    854:                offs++;
1.1       kristaps  855:
1.213     schwarze  856:        /*
1.125     kristaps  857:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    858:         * into the parser as "text", so we only warn about spaces here.
                    859:         */
1.115     kristaps  860:
1.188     kristaps  861:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.218     schwarze  862:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    863:                    ln, offs - 1, NULL);
1.115     kristaps  864:
1.144     kristaps  865:        /*
                    866:         * If an initial macro or a list invocation, divert directly
                    867:         * into macro processing.
                    868:         */
                    869:
1.221   ! schwarze  870:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok)
        !           871:                return(mdoc_macro(mdoc, tok, ln, sv, &offs, buf));
1.144     kristaps  872:
1.203     schwarze  873:        n = mdoc->last;
                    874:        assert(mdoc->last);
1.144     kristaps  875:
                    876:        /*
                    877:         * If the first macro of a `Bl -column', open an `It' block
                    878:         * context around the parsed macro.
                    879:         */
                    880:
                    881:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  882:            LIST_column == n->norm->Bl.type) {
1.203     schwarze  883:                mdoc->flags |= MDOC_FREECOL;
1.221   ! schwarze  884:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.144     kristaps  885:        }
                    886:
                    887:        /*
                    888:         * If we're following a block-level `It' within a `Bl -column'
                    889:         * context (perhaps opened in the above block or in ptext()),
                    890:         * then open an `It' block context around the parsed macro.
1.98      kristaps  891:         */
1.144     kristaps  892:
                    893:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  894:            NULL != n->parent &&
                    895:            MDOC_Bl == n->parent->tok &&
                    896:            LIST_column == n->parent->norm->Bl.type) {
1.203     schwarze  897:                mdoc->flags |= MDOC_FREECOL;
1.221   ! schwarze  898:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.144     kristaps  899:        }
                    900:
                    901:        /* Normal processing of a macro. */
                    902:
1.213     schwarze  903:        if ( ! mdoc_macro(mdoc, tok, ln, sv, &offs, buf))
1.221   ! schwarze  904:                return(0);
1.208     schwarze  905:
                    906:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    907:
                    908:        if (mdoc->quick && MDOC_Sh == tok &&
                    909:            SEC_NAME != mdoc->last->sec)
                    910:                return(2);
1.1       kristaps  911:
1.53      kristaps  912:        return(1);
1.1       kristaps  913: }
1.100     kristaps  914:
1.186     kristaps  915: enum mdelim
                    916: mdoc_isdelim(const char *p)
                    917: {
                    918:
                    919:        if ('\0' == p[0])
                    920:                return(DELIM_NONE);
                    921:
                    922:        if ('\0' == p[1])
                    923:                switch (p[0]) {
1.213     schwarze  924:                case '(':
1.186     kristaps  925:                        /* FALLTHROUGH */
1.213     schwarze  926:                case '[':
1.186     kristaps  927:                        return(DELIM_OPEN);
1.213     schwarze  928:                case '|':
1.186     kristaps  929:                        return(DELIM_MIDDLE);
1.213     schwarze  930:                case '.':
1.186     kristaps  931:                        /* FALLTHROUGH */
1.213     schwarze  932:                case ',':
1.186     kristaps  933:                        /* FALLTHROUGH */
1.213     schwarze  934:                case ';':
1.186     kristaps  935:                        /* FALLTHROUGH */
1.213     schwarze  936:                case ':':
1.186     kristaps  937:                        /* FALLTHROUGH */
1.213     schwarze  938:                case '?':
1.186     kristaps  939:                        /* FALLTHROUGH */
1.213     schwarze  940:                case '!':
1.186     kristaps  941:                        /* FALLTHROUGH */
1.213     schwarze  942:                case ')':
1.186     kristaps  943:                        /* FALLTHROUGH */
1.213     schwarze  944:                case ']':
1.186     kristaps  945:                        return(DELIM_CLOSE);
                    946:                default:
                    947:                        return(DELIM_NONE);
                    948:                }
                    949:
                    950:        if ('\\' != p[0])
                    951:                return(DELIM_NONE);
1.100     kristaps  952:
1.186     kristaps  953:        if (0 == strcmp(p + 1, "."))
                    954:                return(DELIM_CLOSE);
1.200     schwarze  955:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.186     kristaps  956:                return(DELIM_MIDDLE);
                    957:
                    958:        return(DELIM_NONE);
1.211     schwarze  959: }
                    960:
                    961: void
                    962: mdoc_deroff(char **dest, const struct mdoc_node *n)
                    963: {
                    964:        char    *cp;
                    965:        size_t   sz;
                    966:
                    967:        if (MDOC_TEXT != n->type) {
                    968:                for (n = n->child; n; n = n->next)
                    969:                        mdoc_deroff(dest, n);
                    970:                return;
                    971:        }
                    972:
                    973:        /* Skip leading whitespace. */
                    974:
                    975:        for (cp = n->string; '\0' != *cp; cp++)
                    976:                if (0 == isspace((unsigned char)*cp))
                    977:                        break;
                    978:
                    979:        /* Skip trailing whitespace. */
                    980:
                    981:        for (sz = strlen(cp); sz; sz--)
                    982:                if (0 == isspace((unsigned char)cp[sz-1]))
                    983:                        break;
                    984:
                    985:        /* Skip empty strings. */
                    986:
                    987:        if (0 == sz)
                    988:                return;
                    989:
                    990:        if (NULL == *dest) {
                    991:                *dest = mandoc_strndup(cp, sz);
                    992:                return;
                    993:        }
                    994:
                    995:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    996:        free(*dest);
                    997:        *dest = cp;
1.186     kristaps  998: }

CVSweb