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

Annotation of mandoc/mdoc.c, Revision 1.224

1.224   ! schwarze    1: /*     $Id: mdoc.c,v 1.223 2014/08/06 15:09:05 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: #include "config.h"
                     19:
1.106     kristaps   20: #include <sys/types.h>
                     21:
1.1       kristaps   22: #include <assert.h>
1.211     schwarze   23: #include <ctype.h>
1.1       kristaps   24: #include <stdarg.h>
1.73      kristaps   25: #include <stdio.h>
1.1       kristaps   26: #include <stdlib.h>
                     27: #include <string.h>
1.120     kristaps   28: #include <time.h>
1.1       kristaps   29:
1.187     kristaps   30: #include "mdoc.h"
1.136     kristaps   31: #include "mandoc.h"
1.209     schwarze   32: #include "mandoc_aux.h"
1.70      kristaps   33: #include "libmdoc.h"
1.113     kristaps   34: #include "libmandoc.h"
1.1       kristaps   35:
1.216     schwarze   36: const  char *const __mdoc_macronames[MDOC_MAX + 1] = {
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",
1.114     kristaps   47:        "Xr",           "%A",           "%B",           "%D",
                     48:        "%I",           "%J",           "%N",           "%O",
                     49:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   50:        "Ac",           "Ao",           "Aq",           "At",
                     51:        "Bc",           "Bf",           "Bo",           "Bq",
                     52:        "Bsx",          "Bx",           "Db",           "Dc",
                     53:        "Do",           "Dq",           "Ec",           "Ef",
                     54:        "Em",           "Eo",           "Fx",           "Ms",
                     55:        "No",           "Ns",           "Nx",           "Ox",
                     56:        "Pc",           "Pf",           "Po",           "Pq",
                     57:        "Qc",           "Ql",           "Qo",           "Qq",
                     58:        "Re",           "Rs",           "Sc",           "So",
                     59:        "Sq",           "Sm",           "Sx",           "Sy",
                     60:        "Tn",           "Ux",           "Xc",           "Xo",
                     61:        "Fo",           "Fc",           "Oo",           "Oc",
                     62:        "Bk",           "Ek",           "Bt",           "Hf",
1.82      kristaps   63:        "Fr",           "Ud",           "Lb",           "Lp",
                     64:        "Lk",           "Mt",           "Brq",          "Bro",
1.114     kristaps   65:        "Brc",          "%C",           "Es",           "En",
                     66:        "Dx",           "%Q",           "br",           "sp",
1.216     schwarze   67:        "%U",           "Ta",           "ll",           "text",
1.1       kristaps   68:        };
                     69:
1.213     schwarze   70: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
1.1       kristaps   71:        "split",                "nosplit",              "ragged",
1.213     schwarze   72:        "unfilled",             "literal",              "file",
                     73:        "offset",               "bullet",               "dash",
                     74:        "hyphen",               "item",                 "enum",
                     75:        "tag",                  "diag",                 "hang",
                     76:        "ohang",                "inset",                "column",
                     77:        "width",                "compact",              "std",
1.52      kristaps   78:        "filled",               "words",                "emphasis",
1.108     kristaps   79:        "symbolic",             "nested",               "centered"
1.1       kristaps   80:        };
                     81:
                     82: const  char * const *mdoc_macronames = __mdoc_macronames;
                     83: const  char * const *mdoc_argnames = __mdoc_argnames;
                     84:
1.121     kristaps   85: static void              mdoc_node_free(struct mdoc_node *);
1.213     schwarze   86: static void              mdoc_node_unlink(struct mdoc *,
1.121     kristaps   87:                                struct mdoc_node *);
1.73      kristaps   88: static void              mdoc_free1(struct mdoc *);
1.113     kristaps   89: static void              mdoc_alloc1(struct mdoc *);
1.213     schwarze   90: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.117     kristaps   91:                                enum mdoct, enum mdoc_type);
1.213     schwarze   92: static int               node_append(struct mdoc *,
1.71      kristaps   93:                                struct mdoc_node *);
1.193     kristaps   94: #if 0
1.191     kristaps   95: static int               mdoc_preptext(struct mdoc *, int, char *, int);
1.193     kristaps   96: #endif
1.149     kristaps   97: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     98: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.88      kristaps   99:
1.213     schwarze  100:
1.1       kristaps  101: const struct mdoc_node *
1.203     schwarze  102: mdoc_node(const struct mdoc *mdoc)
1.1       kristaps  103: {
                    104:
1.203     schwarze  105:        return(mdoc->first);
1.1       kristaps  106: }
                    107:
1.37      kristaps  108: const struct mdoc_meta *
1.203     schwarze  109: mdoc_meta(const struct mdoc *mdoc)
1.37      kristaps  110: {
                    111:
1.203     schwarze  112:        return(&mdoc->meta);
1.37      kristaps  113: }
                    114:
1.85      kristaps  115: /*
                    116:  * Frees volatile resources (parse tree, meta-data, fields).
                    117:  */
1.73      kristaps  118: static void
                    119: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  120: {
                    121:
                    122:        if (mdoc->first)
1.121     kristaps  123:                mdoc_node_delete(mdoc, mdoc->first);
1.223     schwarze  124:        free(mdoc->meta.msec);
                    125:        free(mdoc->meta.vol);
                    126:        free(mdoc->meta.arch);
                    127:        free(mdoc->meta.date);
                    128:        free(mdoc->meta.title);
                    129:        free(mdoc->meta.os);
                    130:        free(mdoc->meta.name);
1.73      kristaps  131: }
                    132:
1.85      kristaps  133: /*
                    134:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    135:  */
1.113     kristaps  136: static void
1.73      kristaps  137: mdoc_alloc1(struct mdoc *mdoc)
                    138: {
1.67      kristaps  139:
1.112     kristaps  140:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  141:        mdoc->flags = 0;
1.85      kristaps  142:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  143:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  144:        mdoc->first = mdoc->last;
1.67      kristaps  145:        mdoc->last->type = MDOC_ROOT;
1.196     schwarze  146:        mdoc->last->tok = MDOC_MAX;
1.67      kristaps  147:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  148: }
                    149:
                    150: /*
1.85      kristaps  151:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    152:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    153:  * and the parser is ready for re-invocation on a new tree; however,
                    154:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  155:  */
1.113     kristaps  156: void
1.73      kristaps  157: mdoc_reset(struct mdoc *mdoc)
                    158: {
                    159:
                    160:        mdoc_free1(mdoc);
1.113     kristaps  161:        mdoc_alloc1(mdoc);
1.67      kristaps  162: }
                    163:
1.68      kristaps  164: /*
1.85      kristaps  165:  * Completely free up all volatile and non-volatile parse resources.
                    166:  * After invocation, the pointer is no longer usable.
1.68      kristaps  167:  */
1.67      kristaps  168: void
1.38      kristaps  169: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  170: {
                    171:
1.73      kristaps  172:        mdoc_free1(mdoc);
1.1       kristaps  173:        free(mdoc);
                    174: }
                    175:
1.85      kristaps  176: /*
1.213     schwarze  177:  * Allocate volatile and non-volatile parse resources.
1.85      kristaps  178:  */
1.1       kristaps  179: struct mdoc *
1.208     schwarze  180: mdoc_alloc(struct roff *roff, struct mparse *parse,
1.220     schwarze  181:        const char *defos, int quick)
1.1       kristaps  182: {
                    183:        struct mdoc     *p;
                    184:
1.113     kristaps  185:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    186:
1.185     kristaps  187:        p->parse = parse;
1.197     schwarze  188:        p->defos = defos;
1.208     schwarze  189:        p->quick = quick;
1.189     kristaps  190:        p->roff = roff;
1.73      kristaps  191:
1.113     kristaps  192:        mdoc_hash_init();
                    193:        mdoc_alloc1(p);
                    194:        return(p);
1.1       kristaps  195: }
                    196:
                    197: int
1.203     schwarze  198: mdoc_endparse(struct mdoc *mdoc)
1.20      kristaps  199: {
                    200:
1.221     schwarze  201:        return(mdoc_macroend(mdoc));
1.181     kristaps  202: }
                    203:
                    204: int
1.203     schwarze  205: mdoc_addeqn(struct mdoc *mdoc, const struct eqn *ep)
1.181     kristaps  206: {
                    207:        struct mdoc_node *n;
                    208:
1.203     schwarze  209:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  210:        n->eqn = ep;
                    211:
1.203     schwarze  212:        if ( ! node_append(mdoc, n))
1.181     kristaps  213:                return(0);
                    214:
1.203     schwarze  215:        mdoc->next = MDOC_NEXT_SIBLING;
1.181     kristaps  216:        return(1);
1.20      kristaps  217: }
                    218:
1.175     kristaps  219: int
1.203     schwarze  220: mdoc_addspan(struct mdoc *mdoc, const struct tbl_span *sp)
1.175     kristaps  221: {
1.180     kristaps  222:        struct mdoc_node *n;
1.175     kristaps  223:
1.203     schwarze  224:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.180     kristaps  225:        n->span = sp;
                    226:
1.203     schwarze  227:        if ( ! node_append(mdoc, n))
1.180     kristaps  228:                return(0);
                    229:
1.203     schwarze  230:        mdoc->next = MDOC_NEXT_SIBLING;
1.180     kristaps  231:        return(1);
1.175     kristaps  232: }
                    233:
1.50      kristaps  234: /*
1.53      kristaps  235:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  236:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  237:  */
1.20      kristaps  238: int
1.203     schwarze  239: mdoc_parseln(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  240: {
                    241:
1.203     schwarze  242:        mdoc->flags |= MDOC_NEWLINE;
1.153     schwarze  243:
                    244:        /*
                    245:         * Let the roff nS register switch SYNOPSIS mode early,
                    246:         * such that the parser knows at all times
                    247:         * whether this mode is on or off.
                    248:         * Note that this mode is also switched by the Sh macro.
                    249:         */
1.204     schwarze  250:        if (roff_getreg(mdoc->roff, "nS"))
                    251:                mdoc->flags |= MDOC_SYNOPSIS;
                    252:        else
                    253:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.153     schwarze  254:
1.203     schwarze  255:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
1.213     schwarze  256:            mdoc_pmacro(mdoc, ln, buf, offs) :
                    257:            mdoc_ptext(mdoc, ln, buf, offs));
1.1       kristaps  258: }
                    259:
1.88      kristaps  260: int
1.148     kristaps  261: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  262: {
1.122     kristaps  263:        assert(tok < MDOC_MAX);
                    264:
1.223     schwarze  265:        if (mdoc->flags & MDOC_PBODY) {
                    266:                if (tok == MDOC_Dt) {
                    267:                        mandoc_vmsg(MANDOCERR_DT_LATE,
                    268:                            mdoc->parse, line, ppos,
                    269:                            "Dt %s", buf + *pos);
                    270:                        return(1);
                    271:                }
                    272:        } else if ( ! (mdoc_macros[tok].flags & MDOC_PROLOGUE)) {
                    273:                if (mdoc->meta.title == NULL) {
                    274:                        mandoc_vmsg(MANDOCERR_DT_NOTITLE,
                    275:                            mdoc->parse, line, ppos, "%s %s",
                    276:                            mdoc_macronames[tok], buf + *pos);
                    277:                        mdoc->meta.title = mandoc_strdup("UNTITLED");
                    278:                }
1.203     schwarze  279:                if (NULL == mdoc->meta.vol)
                    280:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    281:                mdoc->flags |= MDOC_PBODY;
1.120     kristaps  282:        }
1.88      kristaps  283:
1.203     schwarze  284:        return((*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf));
1.73      kristaps  285: }
                    286:
                    287:
                    288: static int
                    289: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  290: {
                    291:
1.25      kristaps  292:        assert(mdoc->last);
                    293:        assert(mdoc->first);
                    294:        assert(MDOC_ROOT != p->type);
1.1       kristaps  295:
1.13      kristaps  296:        switch (mdoc->next) {
1.213     schwarze  297:        case MDOC_NEXT_SIBLING:
1.6       kristaps  298:                mdoc->last->next = p;
                    299:                p->prev = mdoc->last;
1.13      kristaps  300:                p->parent = mdoc->last->parent;
1.1       kristaps  301:                break;
1.213     schwarze  302:        case MDOC_NEXT_CHILD:
1.13      kristaps  303:                mdoc->last->child = p;
1.1       kristaps  304:                p->parent = mdoc->last;
                    305:                break;
                    306:        default:
1.13      kristaps  307:                abort();
                    308:                /* NOTREACHED */
1.1       kristaps  309:        }
                    310:
1.86      kristaps  311:        p->parent->nchild++;
                    312:
1.172     kristaps  313:        /*
                    314:         * Copy over the normalised-data pointer of our parent.  Not
                    315:         * everybody has one, but copying a null pointer is fine.
                    316:         */
                    317:
                    318:        switch (p->type) {
1.213     schwarze  319:        case MDOC_BODY:
1.202     schwarze  320:                if (ENDBODY_NOT != p->end)
                    321:                        break;
1.172     kristaps  322:                /* FALLTHROUGH */
1.213     schwarze  323:        case MDOC_TAIL:
1.172     kristaps  324:                /* FALLTHROUGH */
1.213     schwarze  325:        case MDOC_HEAD:
1.172     kristaps  326:                p->norm = p->parent->norm;
                    327:                break;
                    328:        default:
                    329:                break;
                    330:        }
                    331:
1.23      kristaps  332:        if ( ! mdoc_valid_pre(mdoc, p))
                    333:                return(0);
1.27      kristaps  334:
                    335:        switch (p->type) {
1.213     schwarze  336:        case MDOC_HEAD:
1.27      kristaps  337:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  338:                p->parent->head = p;
1.27      kristaps  339:                break;
1.213     schwarze  340:        case MDOC_TAIL:
1.27      kristaps  341:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  342:                p->parent->tail = p;
1.27      kristaps  343:                break;
1.213     schwarze  344:        case MDOC_BODY:
1.152     schwarze  345:                if (p->end)
                    346:                        break;
1.27      kristaps  347:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  348:                p->parent->body = p;
1.27      kristaps  349:                break;
                    350:        default:
                    351:                break;
                    352:        }
                    353:
1.1       kristaps  354:        mdoc->last = p;
1.71      kristaps  355:
                    356:        switch (p->type) {
1.213     schwarze  357:        case MDOC_TBL:
1.176     kristaps  358:                /* FALLTHROUGH */
1.213     schwarze  359:        case MDOC_TEXT:
1.71      kristaps  360:                if ( ! mdoc_valid_post(mdoc))
                    361:                        return(0);
                    362:                break;
                    363:        default:
                    364:                break;
                    365:        }
                    366:
1.23      kristaps  367:        return(1);
1.1       kristaps  368: }
                    369:
1.45      kristaps  370: static struct mdoc_node *
1.213     schwarze  371: node_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  372:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  373: {
1.46      kristaps  374:        struct mdoc_node *p;
                    375:
1.113     kristaps  376:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.203     schwarze  377:        p->sec = mdoc->lastsec;
1.73      kristaps  378:        p->line = line;
                    379:        p->pos = pos;
1.206     schwarze  380:        p->lastline = line;
1.73      kristaps  381:        p->tok = tok;
1.118     kristaps  382:        p->type = type;
1.150     kristaps  383:
                    384:        /* Flag analysis. */
                    385:
1.203     schwarze  386:        if (MDOC_SYNOPSIS & mdoc->flags)
1.153     schwarze  387:                p->flags |= MDOC_SYNPRETTY;
                    388:        else
                    389:                p->flags &= ~MDOC_SYNPRETTY;
1.203     schwarze  390:        if (MDOC_NEWLINE & mdoc->flags)
1.130     kristaps  391:                p->flags |= MDOC_LINE;
1.203     schwarze  392:        mdoc->flags &= ~MDOC_NEWLINE;
1.150     kristaps  393:
1.46      kristaps  394:        return(p);
1.45      kristaps  395: }
                    396:
1.23      kristaps  397: int
1.203     schwarze  398: mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.17      kristaps  399: {
                    400:        struct mdoc_node *p;
                    401:
1.203     schwarze  402:        p = node_alloc(mdoc, line, pos, tok, MDOC_TAIL);
                    403:        if ( ! node_append(mdoc, p))
1.102     kristaps  404:                return(0);
1.203     schwarze  405:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  406:        return(1);
1.17      kristaps  407: }
                    408:
1.23      kristaps  409: int
1.203     schwarze  410: mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  411: {
                    412:        struct mdoc_node *p;
                    413:
1.203     schwarze  414:        assert(mdoc->first);
                    415:        assert(mdoc->last);
1.1       kristaps  416:
1.203     schwarze  417:        p = node_alloc(mdoc, line, pos, tok, MDOC_HEAD);
                    418:        if ( ! node_append(mdoc, p))
1.102     kristaps  419:                return(0);
1.203     schwarze  420:        mdoc->next = MDOC_NEXT_CHILD;
1.102     kristaps  421:        return(1);
1.1       kristaps  422: }
                    423:
1.23      kristaps  424: int
1.203     schwarze  425: mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  426: {
                    427:        struct mdoc_node *p;
                    428:
1.203     schwarze  429:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
                    430:        if ( ! node_append(mdoc, p))
1.102     kristaps  431:                return(0);
1.203     schwarze  432:        mdoc->next = MDOC_NEXT_CHILD;
1.152     schwarze  433:        return(1);
                    434: }
                    435:
                    436: int
1.203     schwarze  437: mdoc_endbody_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok,
1.152     schwarze  438:                struct mdoc_node *body, enum mdoc_endbody end)
                    439: {
                    440:        struct mdoc_node *p;
                    441:
1.203     schwarze  442:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.152     schwarze  443:        p->pending = body;
1.202     schwarze  444:        p->norm = body->norm;
1.152     schwarze  445:        p->end = end;
1.203     schwarze  446:        if ( ! node_append(mdoc, p))
1.152     schwarze  447:                return(0);
1.203     schwarze  448:        mdoc->next = MDOC_NEXT_SIBLING;
1.102     kristaps  449:        return(1);
1.1       kristaps  450: }
                    451:
1.23      kristaps  452: int
1.213     schwarze  453: mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  454:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  455: {
                    456:        struct mdoc_node *p;
                    457:
1.203     schwarze  458:        p = node_alloc(mdoc, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  459:        p->args = args;
                    460:        if (p->args)
1.53      kristaps  461:                (args->refcnt)++;
1.172     kristaps  462:
                    463:        switch (tok) {
1.213     schwarze  464:        case MDOC_Bd:
1.172     kristaps  465:                /* FALLTHROUGH */
1.213     schwarze  466:        case MDOC_Bf:
1.172     kristaps  467:                /* FALLTHROUGH */
1.213     schwarze  468:        case MDOC_Bl:
1.217     schwarze  469:                /* FALLTHROUGH */
                    470:        case MDOC_En:
1.173     kristaps  471:                /* FALLTHROUGH */
1.213     schwarze  472:        case MDOC_Rs:
1.172     kristaps  473:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    474:                break;
                    475:        default:
                    476:                break;
                    477:        }
                    478:
1.203     schwarze  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:
1.23      kristaps  485: int
1.213     schwarze  486: mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  487:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  488: {
                    489:        struct mdoc_node *p;
                    490:
1.203     schwarze  491:        p = node_alloc(mdoc, line, pos, tok, MDOC_ELEM);
1.77      kristaps  492:        p->args = args;
                    493:        if (p->args)
1.53      kristaps  494:                (args->refcnt)++;
1.172     kristaps  495:
                    496:        switch (tok) {
1.213     schwarze  497:        case MDOC_An:
1.172     kristaps  498:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    499:                break;
                    500:        default:
                    501:                break;
                    502:        }
                    503:
1.203     schwarze  504:        if ( ! node_append(mdoc, p))
1.102     kristaps  505:                return(0);
1.203     schwarze  506:        mdoc->next = MDOC_NEXT_CHILD;
1.175     kristaps  507:        return(1);
                    508: }
1.1       kristaps  509:
1.124     kristaps  510: int
1.203     schwarze  511: mdoc_word_alloc(struct mdoc *mdoc, int line, int pos, const char *p)
1.1       kristaps  512: {
1.91      kristaps  513:        struct mdoc_node *n;
1.1       kristaps  514:
1.203     schwarze  515:        n = node_alloc(mdoc, line, pos, MDOC_MAX, MDOC_TEXT);
                    516:        n->string = roff_strdup(mdoc->roff, p);
1.91      kristaps  517:
1.203     schwarze  518:        if ( ! node_append(mdoc, n))
1.101     kristaps  519:                return(0);
1.124     kristaps  520:
1.203     schwarze  521:        mdoc->next = MDOC_NEXT_SIBLING;
1.101     kristaps  522:        return(1);
1.91      kristaps  523: }
                    524:
1.205     schwarze  525: void
                    526: mdoc_word_append(struct mdoc *mdoc, const char *p)
                    527: {
                    528:        struct mdoc_node        *n;
                    529:        char                    *addstr, *newstr;
                    530:
                    531:        n = mdoc->last;
                    532:        addstr = roff_strdup(mdoc->roff, p);
1.210     schwarze  533:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.205     schwarze  534:        free(addstr);
                    535:        free(n->string);
                    536:        n->string = newstr;
                    537:        mdoc->next = MDOC_NEXT_SIBLING;
                    538: }
1.91      kristaps  539:
1.155     kristaps  540: static void
1.53      kristaps  541: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  542: {
                    543:
1.172     kristaps  544:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  545:                free(p->norm);
1.53      kristaps  546:        if (p->string)
                    547:                free(p->string);
                    548:        if (p->args)
                    549:                mdoc_argv_free(p->args);
1.1       kristaps  550:        free(p);
                    551: }
                    552:
1.121     kristaps  553: static void
1.203     schwarze  554: mdoc_node_unlink(struct mdoc *mdoc, struct mdoc_node *n)
1.121     kristaps  555: {
                    556:
                    557:        /* Adjust siblings. */
                    558:
                    559:        if (n->prev)
                    560:                n->prev->next = n->next;
                    561:        if (n->next)
                    562:                n->next->prev = n->prev;
                    563:
                    564:        /* Adjust parent. */
                    565:
                    566:        if (n->parent) {
                    567:                n->parent->nchild--;
                    568:                if (n->parent->child == n)
                    569:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  570:                if (n->parent->last == n)
                    571:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  572:        }
                    573:
                    574:        /* Adjust parse point, if applicable. */
                    575:
1.203     schwarze  576:        if (mdoc && mdoc->last == n) {
1.121     kristaps  577:                if (n->prev) {
1.203     schwarze  578:                        mdoc->last = n->prev;
                    579:                        mdoc->next = MDOC_NEXT_SIBLING;
1.121     kristaps  580:                } else {
1.203     schwarze  581:                        mdoc->last = n->parent;
                    582:                        mdoc->next = MDOC_NEXT_CHILD;
1.121     kristaps  583:                }
                    584:        }
                    585:
1.203     schwarze  586:        if (mdoc && mdoc->first == n)
                    587:                mdoc->first = NULL;
1.121     kristaps  588: }
                    589:
1.53      kristaps  590: void
1.203     schwarze  591: mdoc_node_delete(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  592: {
                    593:
1.121     kristaps  594:        while (p->child) {
                    595:                assert(p->nchild);
1.203     schwarze  596:                mdoc_node_delete(mdoc, p->child);
1.121     kristaps  597:        }
                    598:        assert(0 == p->nchild);
1.1       kristaps  599:
1.203     schwarze  600:        mdoc_node_unlink(mdoc, p);
1.53      kristaps  601:        mdoc_node_free(p);
1.201     schwarze  602: }
                    603:
                    604: int
1.203     schwarze  605: mdoc_node_relink(struct mdoc *mdoc, struct mdoc_node *p)
1.201     schwarze  606: {
                    607:
1.203     schwarze  608:        mdoc_node_unlink(mdoc, p);
                    609:        return(node_append(mdoc, p));
1.1       kristaps  610: }
                    611:
1.193     kristaps  612: #if 0
1.191     kristaps  613: /*
                    614:  * Pre-treat a text line.
                    615:  * Text lines can consist of equations, which must be handled apart from
                    616:  * the regular text.
                    617:  * Thus, use this function to step through a line checking if it has any
                    618:  * equations embedded in it.
                    619:  * This must handle multiple equations AND equations that do not end at
                    620:  * the end-of-line, i.e., will re-enter in the next roff parse.
                    621:  */
                    622: static int
1.203     schwarze  623: mdoc_preptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.191     kristaps  624: {
                    625:        char            *start, *end;
                    626:        char             delim;
                    627:
                    628:        while ('\0' != buf[offs]) {
                    629:                /* Mark starting position if eqn is set. */
                    630:                start = NULL;
1.203     schwarze  631:                if ('\0' != (delim = roff_eqndelim(mdoc->roff)))
1.191     kristaps  632:                        if (NULL != (start = strchr(buf + offs, delim)))
                    633:                                *start++ = '\0';
                    634:
                    635:                /* Parse text as normal. */
1.203     schwarze  636:                if ( ! mdoc_ptext(mdoc, line, buf, offs))
1.191     kristaps  637:                        return(0);
                    638:
                    639:                /* Continue only if an equation exists. */
                    640:                if (NULL == start)
                    641:                        break;
                    642:
                    643:                /* Read past the end of the equation. */
                    644:                offs += start - (buf + offs);
                    645:                assert(start == &buf[offs]);
                    646:                if (NULL != (end = strchr(buf + offs, delim))) {
                    647:                        *end++ = '\0';
                    648:                        while (' ' == *end)
                    649:                                end++;
                    650:                }
                    651:
                    652:                /* Parse the equation itself. */
1.203     schwarze  653:                roff_openeqn(mdoc->roff, NULL, line, offs, buf);
1.191     kristaps  654:
                    655:                /* Process a finished equation? */
1.203     schwarze  656:                if (roff_closeeqn(mdoc->roff))
                    657:                        if ( ! mdoc_addeqn(mdoc, roff_eqn(mdoc->roff)))
1.191     kristaps  658:                                return(0);
                    659:                offs += (end - (buf + offs));
1.213     schwarze  660:        }
1.191     kristaps  661:
                    662:        return(1);
                    663: }
1.193     kristaps  664: #endif
1.1       kristaps  665:
1.53      kristaps  666: /*
                    667:  * Parse free-form text, that is, a line that does not begin with the
                    668:  * control character.
                    669:  */
                    670: static int
1.203     schwarze  671: mdoc_ptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.1       kristaps  672: {
1.142     kristaps  673:        char             *c, *ws, *end;
                    674:        struct mdoc_node *n;
                    675:
1.203     schwarze  676:        assert(mdoc->last);
                    677:        n = mdoc->last;
1.142     kristaps  678:
                    679:        /*
1.144     kristaps  680:         * Divert directly to list processing if we're encountering a
1.142     kristaps  681:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  682:         * (a MDOC_BODY means it's already open, in which case we should
                    683:         * process within its context in the normal way).
1.142     kristaps  684:         */
                    685:
1.143     kristaps  686:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  687:            LIST_column == n->norm->Bl.type) {
1.144     kristaps  688:                /* `Bl' is open without any children. */
1.203     schwarze  689:                mdoc->flags |= MDOC_FREECOL;
                    690:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  691:        }
                    692:
                    693:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  694:            NULL != n->parent &&
                    695:            MDOC_Bl == n->parent->tok &&
                    696:            LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  697:                /* `Bl' has block-level `It' children. */
1.203     schwarze  698:                mdoc->flags |= MDOC_FREECOL;
                    699:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  700:        }
1.124     kristaps  701:
1.137     schwarze  702:        /*
                    703:         * Search for the beginning of unescaped trailing whitespace (ws)
                    704:         * and for the first character not to be output (end).
                    705:         */
1.139     kristaps  706:
                    707:        /* FIXME: replace with strcspn(). */
1.137     schwarze  708:        ws = NULL;
                    709:        for (c = end = buf + offs; *c; c++) {
                    710:                switch (*c) {
                    711:                case ' ':
                    712:                        if (NULL == ws)
                    713:                                ws = c;
                    714:                        continue;
                    715:                case '\t':
                    716:                        /*
                    717:                         * Always warn about trailing tabs,
                    718:                         * even outside literal context,
                    719:                         * where they should be put on the next line.
                    720:                         */
                    721:                        if (NULL == ws)
                    722:                                ws = c;
                    723:                        /*
                    724:                         * Strip trailing tabs in literal context only;
                    725:                         * outside, they affect the next line.
                    726:                         */
1.203     schwarze  727:                        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  728:                                continue;
                    729:                        break;
                    730:                case '\\':
                    731:                        /* Skip the escaped character, too, if any. */
                    732:                        if (c[1])
                    733:                                c++;
                    734:                        /* FALLTHROUGH */
                    735:                default:
                    736:                        ws = NULL;
                    737:                        break;
                    738:                }
                    739:                end = c + 1;
                    740:        }
                    741:        *end = '\0';
1.91      kristaps  742:
1.137     schwarze  743:        if (ws)
1.218     schwarze  744:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    745:                    line, (int)(ws-buf), NULL);
1.115     kristaps  746:
1.203     schwarze  747:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & mdoc->flags)) {
1.218     schwarze  748:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    749:                    line, (int)(c - buf), NULL);
1.124     kristaps  750:
1.119     kristaps  751:                /*
1.165     schwarze  752:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  753:                 * blank lines aren't allowed, but enough manuals assume this
                    754:                 * behaviour that we want to work around it.
1.119     kristaps  755:                 */
1.203     schwarze  756:                if ( ! mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL))
1.119     kristaps  757:                        return(0);
1.124     kristaps  758:
1.203     schwarze  759:                mdoc->next = MDOC_NEXT_SIBLING;
1.199     schwarze  760:
1.203     schwarze  761:                return(mdoc_valid_post(mdoc));
1.119     kristaps  762:        }
1.68      kristaps  763:
1.203     schwarze  764:        if ( ! mdoc_word_alloc(mdoc, line, offs, buf+offs))
1.137     schwarze  765:                return(0);
1.91      kristaps  766:
1.203     schwarze  767:        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  768:                return(1);
1.128     kristaps  769:
                    770:        /*
                    771:         * End-of-sentence check.  If the last character is an unescaped
                    772:         * EOS character, then flag the node as being the end of a
                    773:         * sentence.  The front-end will know how to interpret this.
                    774:         */
1.132     kristaps  775:
1.137     schwarze  776:        assert(buf < end);
                    777:
1.207     schwarze  778:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.203     schwarze  779:                mdoc->last->flags |= MDOC_EOS;
1.128     kristaps  780:
                    781:        return(1);
1.1       kristaps  782: }
                    783:
1.53      kristaps  784: /*
                    785:  * Parse a macro line, that is, a line beginning with the control
                    786:  * character.
                    787:  */
1.155     kristaps  788: static int
1.203     schwarze  789: mdoc_pmacro(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  790: {
1.144     kristaps  791:        enum mdoct        tok;
1.188     kristaps  792:        int               i, sv;
1.144     kristaps  793:        char              mac[5];
                    794:        struct mdoc_node *n;
1.1       kristaps  795:
1.188     kristaps  796:        /* Empty post-control lines are ignored. */
1.63      kristaps  797:
1.188     kristaps  798:        if ('"' == buf[offs]) {
1.218     schwarze  799:                mandoc_msg(MANDOCERR_COMMENT_BAD, mdoc->parse,
                    800:                    ln, offs, NULL);
1.188     kristaps  801:                return(1);
                    802:        } else if ('\0' == buf[offs])
1.63      kristaps  803:                return(1);
                    804:
1.188     kristaps  805:        sv = offs;
1.130     kristaps  806:
1.213     schwarze  807:        /*
1.162     schwarze  808:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  809:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  810:         */
1.1       kristaps  811:
1.188     kristaps  812:        i = 0;
1.213     schwarze  813:        while (i < 4 && '\0' != buf[offs] && ' ' != buf[offs] &&
                    814:            '\t' != buf[offs])
1.188     kristaps  815:                mac[i++] = buf[offs++];
                    816:
                    817:        mac[i] = '\0';
                    818:
1.214     schwarze  819:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  820:
1.163     schwarze  821:        if (MDOC_MAX == tok) {
1.222     schwarze  822:                mandoc_msg(MANDOCERR_MACRO, mdoc->parse,
                    823:                    ln, sv, buf + sv - 1);
1.58      kristaps  824:                return(1);
1.53      kristaps  825:        }
1.1       kristaps  826:
1.160     kristaps  827:        /* Disregard the first trailing tab, if applicable. */
                    828:
1.188     kristaps  829:        if ('\t' == buf[offs])
                    830:                offs++;
1.160     kristaps  831:
                    832:        /* Jump to the next non-whitespace word. */
1.1       kristaps  833:
1.188     kristaps  834:        while (buf[offs] && ' ' == buf[offs])
                    835:                offs++;
1.1       kristaps  836:
1.213     schwarze  837:        /*
1.125     kristaps  838:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    839:         * into the parser as "text", so we only warn about spaces here.
                    840:         */
1.115     kristaps  841:
1.188     kristaps  842:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.218     schwarze  843:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    844:                    ln, offs - 1, NULL);
1.115     kristaps  845:
1.144     kristaps  846:        /*
                    847:         * If an initial macro or a list invocation, divert directly
                    848:         * into macro processing.
                    849:         */
                    850:
1.221     schwarze  851:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok)
                    852:                return(mdoc_macro(mdoc, tok, ln, sv, &offs, buf));
1.144     kristaps  853:
1.203     schwarze  854:        n = mdoc->last;
                    855:        assert(mdoc->last);
1.144     kristaps  856:
                    857:        /*
                    858:         * If the first macro of a `Bl -column', open an `It' block
                    859:         * context around the parsed macro.
                    860:         */
                    861:
                    862:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  863:            LIST_column == n->norm->Bl.type) {
1.203     schwarze  864:                mdoc->flags |= MDOC_FREECOL;
1.221     schwarze  865:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.144     kristaps  866:        }
                    867:
                    868:        /*
                    869:         * If we're following a block-level `It' within a `Bl -column'
                    870:         * context (perhaps opened in the above block or in ptext()),
                    871:         * then open an `It' block context around the parsed macro.
1.98      kristaps  872:         */
1.144     kristaps  873:
                    874:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  875:            NULL != n->parent &&
                    876:            MDOC_Bl == n->parent->tok &&
                    877:            LIST_column == n->parent->norm->Bl.type) {
1.203     schwarze  878:                mdoc->flags |= MDOC_FREECOL;
1.221     schwarze  879:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.144     kristaps  880:        }
                    881:
                    882:        /* Normal processing of a macro. */
                    883:
1.213     schwarze  884:        if ( ! mdoc_macro(mdoc, tok, ln, sv, &offs, buf))
1.221     schwarze  885:                return(0);
1.208     schwarze  886:
                    887:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    888:
                    889:        if (mdoc->quick && MDOC_Sh == tok &&
                    890:            SEC_NAME != mdoc->last->sec)
                    891:                return(2);
1.1       kristaps  892:
1.53      kristaps  893:        return(1);
1.1       kristaps  894: }
1.100     kristaps  895:
1.186     kristaps  896: enum mdelim
                    897: mdoc_isdelim(const char *p)
                    898: {
                    899:
                    900:        if ('\0' == p[0])
                    901:                return(DELIM_NONE);
                    902:
                    903:        if ('\0' == p[1])
                    904:                switch (p[0]) {
1.213     schwarze  905:                case '(':
1.186     kristaps  906:                        /* FALLTHROUGH */
1.213     schwarze  907:                case '[':
1.186     kristaps  908:                        return(DELIM_OPEN);
1.213     schwarze  909:                case '|':
1.186     kristaps  910:                        return(DELIM_MIDDLE);
1.213     schwarze  911:                case '.':
1.186     kristaps  912:                        /* FALLTHROUGH */
1.213     schwarze  913:                case ',':
1.186     kristaps  914:                        /* FALLTHROUGH */
1.213     schwarze  915:                case ';':
1.186     kristaps  916:                        /* FALLTHROUGH */
1.213     schwarze  917:                case ':':
1.186     kristaps  918:                        /* FALLTHROUGH */
1.213     schwarze  919:                case '?':
1.186     kristaps  920:                        /* FALLTHROUGH */
1.213     schwarze  921:                case '!':
1.186     kristaps  922:                        /* FALLTHROUGH */
1.213     schwarze  923:                case ')':
1.186     kristaps  924:                        /* FALLTHROUGH */
1.213     schwarze  925:                case ']':
1.186     kristaps  926:                        return(DELIM_CLOSE);
                    927:                default:
                    928:                        return(DELIM_NONE);
                    929:                }
                    930:
                    931:        if ('\\' != p[0])
                    932:                return(DELIM_NONE);
1.100     kristaps  933:
1.186     kristaps  934:        if (0 == strcmp(p + 1, "."))
                    935:                return(DELIM_CLOSE);
1.200     schwarze  936:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.186     kristaps  937:                return(DELIM_MIDDLE);
                    938:
                    939:        return(DELIM_NONE);
1.211     schwarze  940: }
                    941:
                    942: void
                    943: mdoc_deroff(char **dest, const struct mdoc_node *n)
                    944: {
                    945:        char    *cp;
                    946:        size_t   sz;
                    947:
                    948:        if (MDOC_TEXT != n->type) {
                    949:                for (n = n->child; n; n = n->next)
                    950:                        mdoc_deroff(dest, n);
                    951:                return;
                    952:        }
                    953:
                    954:        /* Skip leading whitespace. */
                    955:
                    956:        for (cp = n->string; '\0' != *cp; cp++)
                    957:                if (0 == isspace((unsigned char)*cp))
                    958:                        break;
                    959:
                    960:        /* Skip trailing whitespace. */
                    961:
                    962:        for (sz = strlen(cp); sz; sz--)
                    963:                if (0 == isspace((unsigned char)cp[sz-1]))
                    964:                        break;
                    965:
                    966:        /* Skip empty strings. */
                    967:
                    968:        if (0 == sz)
                    969:                return;
                    970:
                    971:        if (NULL == *dest) {
                    972:                *dest = mandoc_strndup(cp, sz);
                    973:                return;
                    974:        }
                    975:
                    976:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    977:        free(*dest);
                    978:        *dest = cp;
1.186     kristaps  979: }

CVSweb