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

Annotation of mandoc/mdoc.c, Revision 1.234

1.234   ! schwarze    1: /*     $Id: mdoc.c,v 1.233 2014/11/28 06:27:05 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);
1.230     schwarze   92: static void              node_append(struct mdoc *, struct mdoc_node *);
1.149     kristaps   93: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     94: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.88      kristaps   95:
1.213     schwarze   96:
1.1       kristaps   97: const struct mdoc_node *
1.203     schwarze   98: mdoc_node(const struct mdoc *mdoc)
1.1       kristaps   99: {
                    100:
1.203     schwarze  101:        return(mdoc->first);
1.1       kristaps  102: }
                    103:
1.37      kristaps  104: const struct mdoc_meta *
1.203     schwarze  105: mdoc_meta(const struct mdoc *mdoc)
1.37      kristaps  106: {
                    107:
1.203     schwarze  108:        return(&mdoc->meta);
1.37      kristaps  109: }
                    110:
1.85      kristaps  111: /*
                    112:  * Frees volatile resources (parse tree, meta-data, fields).
                    113:  */
1.73      kristaps  114: static void
                    115: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  116: {
                    117:
                    118:        if (mdoc->first)
1.121     kristaps  119:                mdoc_node_delete(mdoc, mdoc->first);
1.223     schwarze  120:        free(mdoc->meta.msec);
                    121:        free(mdoc->meta.vol);
                    122:        free(mdoc->meta.arch);
                    123:        free(mdoc->meta.date);
                    124:        free(mdoc->meta.title);
                    125:        free(mdoc->meta.os);
                    126:        free(mdoc->meta.name);
1.73      kristaps  127: }
                    128:
1.85      kristaps  129: /*
                    130:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    131:  */
1.113     kristaps  132: static void
1.73      kristaps  133: mdoc_alloc1(struct mdoc *mdoc)
                    134: {
1.67      kristaps  135:
1.112     kristaps  136:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  137:        mdoc->flags = 0;
1.85      kristaps  138:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  139:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  140:        mdoc->first = mdoc->last;
1.67      kristaps  141:        mdoc->last->type = MDOC_ROOT;
1.196     schwarze  142:        mdoc->last->tok = MDOC_MAX;
1.67      kristaps  143:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  144: }
                    145:
                    146: /*
1.85      kristaps  147:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    148:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    149:  * and the parser is ready for re-invocation on a new tree; however,
                    150:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  151:  */
1.113     kristaps  152: void
1.73      kristaps  153: mdoc_reset(struct mdoc *mdoc)
                    154: {
                    155:
                    156:        mdoc_free1(mdoc);
1.113     kristaps  157:        mdoc_alloc1(mdoc);
1.67      kristaps  158: }
                    159:
1.68      kristaps  160: /*
1.85      kristaps  161:  * Completely free up all volatile and non-volatile parse resources.
                    162:  * After invocation, the pointer is no longer usable.
1.68      kristaps  163:  */
1.67      kristaps  164: void
1.38      kristaps  165: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  166: {
                    167:
1.73      kristaps  168:        mdoc_free1(mdoc);
1.1       kristaps  169:        free(mdoc);
                    170: }
                    171:
1.85      kristaps  172: /*
1.213     schwarze  173:  * Allocate volatile and non-volatile parse resources.
1.85      kristaps  174:  */
1.1       kristaps  175: struct mdoc *
1.208     schwarze  176: mdoc_alloc(struct roff *roff, struct mparse *parse,
1.220     schwarze  177:        const char *defos, int quick)
1.1       kristaps  178: {
                    179:        struct mdoc     *p;
                    180:
1.113     kristaps  181:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    182:
1.185     kristaps  183:        p->parse = parse;
1.197     schwarze  184:        p->defos = defos;
1.208     schwarze  185:        p->quick = quick;
1.189     kristaps  186:        p->roff = roff;
1.73      kristaps  187:
1.113     kristaps  188:        mdoc_hash_init();
                    189:        mdoc_alloc1(p);
                    190:        return(p);
1.1       kristaps  191: }
                    192:
1.234   ! schwarze  193: void
1.203     schwarze  194: mdoc_endparse(struct mdoc *mdoc)
1.20      kristaps  195: {
                    196:
1.232     schwarze  197:        mdoc_macroend(mdoc);
1.181     kristaps  198: }
                    199:
1.233     schwarze  200: void
1.203     schwarze  201: mdoc_addeqn(struct mdoc *mdoc, const struct eqn *ep)
1.181     kristaps  202: {
                    203:        struct mdoc_node *n;
                    204:
1.203     schwarze  205:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.181     kristaps  206:        n->eqn = ep;
1.227     schwarze  207:        if (ep->ln > mdoc->last->line)
                    208:                n->flags |= MDOC_LINE;
1.230     schwarze  209:        node_append(mdoc, n);
1.203     schwarze  210:        mdoc->next = MDOC_NEXT_SIBLING;
1.20      kristaps  211: }
                    212:
1.233     schwarze  213: void
1.203     schwarze  214: mdoc_addspan(struct mdoc *mdoc, const struct tbl_span *sp)
1.175     kristaps  215: {
1.180     kristaps  216:        struct mdoc_node *n;
1.175     kristaps  217:
1.203     schwarze  218:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.180     kristaps  219:        n->span = sp;
1.230     schwarze  220:        node_append(mdoc, n);
1.203     schwarze  221:        mdoc->next = MDOC_NEXT_SIBLING;
1.175     kristaps  222: }
                    223:
1.50      kristaps  224: /*
1.53      kristaps  225:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  226:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  227:  */
1.20      kristaps  228: int
1.203     schwarze  229: mdoc_parseln(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  230: {
                    231:
1.228     schwarze  232:        if (mdoc->last->type != MDOC_EQN || ln > mdoc->last->line)
                    233:                mdoc->flags |= MDOC_NEWLINE;
1.153     schwarze  234:
                    235:        /*
                    236:         * Let the roff nS register switch SYNOPSIS mode early,
                    237:         * such that the parser knows at all times
                    238:         * whether this mode is on or off.
                    239:         * Note that this mode is also switched by the Sh macro.
                    240:         */
1.204     schwarze  241:        if (roff_getreg(mdoc->roff, "nS"))
                    242:                mdoc->flags |= MDOC_SYNOPSIS;
                    243:        else
                    244:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.153     schwarze  245:
1.203     schwarze  246:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
1.213     schwarze  247:            mdoc_pmacro(mdoc, ln, buf, offs) :
                    248:            mdoc_ptext(mdoc, ln, buf, offs));
1.1       kristaps  249: }
                    250:
1.232     schwarze  251: void
1.148     kristaps  252: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  253: {
1.122     kristaps  254:        assert(tok < MDOC_MAX);
                    255:
1.223     schwarze  256:        if (mdoc->flags & MDOC_PBODY) {
                    257:                if (tok == MDOC_Dt) {
                    258:                        mandoc_vmsg(MANDOCERR_DT_LATE,
                    259:                            mdoc->parse, line, ppos,
                    260:                            "Dt %s", buf + *pos);
1.232     schwarze  261:                        return;
1.223     schwarze  262:                }
                    263:        } else if ( ! (mdoc_macros[tok].flags & MDOC_PROLOGUE)) {
                    264:                if (mdoc->meta.title == NULL) {
                    265:                        mandoc_vmsg(MANDOCERR_DT_NOTITLE,
                    266:                            mdoc->parse, line, ppos, "%s %s",
                    267:                            mdoc_macronames[tok], buf + *pos);
                    268:                        mdoc->meta.title = mandoc_strdup("UNTITLED");
                    269:                }
1.203     schwarze  270:                if (NULL == mdoc->meta.vol)
                    271:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    272:                mdoc->flags |= MDOC_PBODY;
1.120     kristaps  273:        }
1.232     schwarze  274:        (*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf);
1.73      kristaps  275: }
                    276:
                    277:
1.230     schwarze  278: static void
1.73      kristaps  279: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  280: {
                    281:
1.25      kristaps  282:        assert(mdoc->last);
                    283:        assert(mdoc->first);
                    284:        assert(MDOC_ROOT != p->type);
1.1       kristaps  285:
1.13      kristaps  286:        switch (mdoc->next) {
1.213     schwarze  287:        case MDOC_NEXT_SIBLING:
1.6       kristaps  288:                mdoc->last->next = p;
                    289:                p->prev = mdoc->last;
1.13      kristaps  290:                p->parent = mdoc->last->parent;
1.1       kristaps  291:                break;
1.213     schwarze  292:        case MDOC_NEXT_CHILD:
1.13      kristaps  293:                mdoc->last->child = p;
1.1       kristaps  294:                p->parent = mdoc->last;
                    295:                break;
                    296:        default:
1.13      kristaps  297:                abort();
                    298:                /* NOTREACHED */
1.1       kristaps  299:        }
                    300:
1.86      kristaps  301:        p->parent->nchild++;
                    302:
1.172     kristaps  303:        /*
                    304:         * Copy over the normalised-data pointer of our parent.  Not
                    305:         * everybody has one, but copying a null pointer is fine.
                    306:         */
                    307:
                    308:        switch (p->type) {
1.213     schwarze  309:        case MDOC_BODY:
1.202     schwarze  310:                if (ENDBODY_NOT != p->end)
                    311:                        break;
1.172     kristaps  312:                /* FALLTHROUGH */
1.213     schwarze  313:        case MDOC_TAIL:
1.172     kristaps  314:                /* FALLTHROUGH */
1.213     schwarze  315:        case MDOC_HEAD:
1.172     kristaps  316:                p->norm = p->parent->norm;
                    317:                break;
                    318:        default:
                    319:                break;
                    320:        }
                    321:
1.230     schwarze  322:        mdoc_valid_pre(mdoc, p);
1.27      kristaps  323:
                    324:        switch (p->type) {
1.213     schwarze  325:        case MDOC_HEAD:
1.27      kristaps  326:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  327:                p->parent->head = p;
1.27      kristaps  328:                break;
1.213     schwarze  329:        case MDOC_TAIL:
1.27      kristaps  330:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  331:                p->parent->tail = p;
1.27      kristaps  332:                break;
1.213     schwarze  333:        case MDOC_BODY:
1.152     schwarze  334:                if (p->end)
                    335:                        break;
1.27      kristaps  336:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  337:                p->parent->body = p;
1.27      kristaps  338:                break;
                    339:        default:
                    340:                break;
                    341:        }
                    342:
1.1       kristaps  343:        mdoc->last = p;
1.71      kristaps  344:
                    345:        switch (p->type) {
1.213     schwarze  346:        case MDOC_TBL:
1.176     kristaps  347:                /* FALLTHROUGH */
1.213     schwarze  348:        case MDOC_TEXT:
1.230     schwarze  349:                mdoc_valid_post(mdoc);
1.71      kristaps  350:                break;
                    351:        default:
                    352:                break;
                    353:        }
1.1       kristaps  354: }
                    355:
1.45      kristaps  356: static struct mdoc_node *
1.213     schwarze  357: node_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  358:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  359: {
1.46      kristaps  360:        struct mdoc_node *p;
                    361:
1.113     kristaps  362:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.203     schwarze  363:        p->sec = mdoc->lastsec;
1.73      kristaps  364:        p->line = line;
                    365:        p->pos = pos;
1.206     schwarze  366:        p->lastline = line;
1.73      kristaps  367:        p->tok = tok;
1.118     kristaps  368:        p->type = type;
1.150     kristaps  369:
                    370:        /* Flag analysis. */
                    371:
1.203     schwarze  372:        if (MDOC_SYNOPSIS & mdoc->flags)
1.153     schwarze  373:                p->flags |= MDOC_SYNPRETTY;
                    374:        else
                    375:                p->flags &= ~MDOC_SYNPRETTY;
1.203     schwarze  376:        if (MDOC_NEWLINE & mdoc->flags)
1.130     kristaps  377:                p->flags |= MDOC_LINE;
1.203     schwarze  378:        mdoc->flags &= ~MDOC_NEWLINE;
1.150     kristaps  379:
1.46      kristaps  380:        return(p);
1.45      kristaps  381: }
                    382:
1.231     schwarze  383: void
1.203     schwarze  384: mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.17      kristaps  385: {
                    386:        struct mdoc_node *p;
                    387:
1.203     schwarze  388:        p = node_alloc(mdoc, line, pos, tok, MDOC_TAIL);
1.230     schwarze  389:        node_append(mdoc, p);
1.203     schwarze  390:        mdoc->next = MDOC_NEXT_CHILD;
1.17      kristaps  391: }
                    392:
1.231     schwarze  393: struct mdoc_node *
1.203     schwarze  394: mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  395: {
                    396:        struct mdoc_node *p;
                    397:
1.203     schwarze  398:        assert(mdoc->first);
                    399:        assert(mdoc->last);
                    400:        p = node_alloc(mdoc, line, pos, tok, MDOC_HEAD);
1.230     schwarze  401:        node_append(mdoc, p);
1.203     schwarze  402:        mdoc->next = MDOC_NEXT_CHILD;
1.231     schwarze  403:        return(p);
1.1       kristaps  404: }
                    405:
1.231     schwarze  406: struct mdoc_node *
1.203     schwarze  407: mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  408: {
                    409:        struct mdoc_node *p;
                    410:
1.203     schwarze  411:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.230     schwarze  412:        node_append(mdoc, p);
1.203     schwarze  413:        mdoc->next = MDOC_NEXT_CHILD;
1.231     schwarze  414:        return(p);
1.152     schwarze  415: }
                    416:
1.231     schwarze  417: void
1.203     schwarze  418: mdoc_endbody_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok,
1.152     schwarze  419:                struct mdoc_node *body, enum mdoc_endbody end)
                    420: {
                    421:        struct mdoc_node *p;
                    422:
1.203     schwarze  423:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.152     schwarze  424:        p->pending = body;
1.202     schwarze  425:        p->norm = body->norm;
1.152     schwarze  426:        p->end = end;
1.230     schwarze  427:        node_append(mdoc, p);
1.203     schwarze  428:        mdoc->next = MDOC_NEXT_SIBLING;
1.1       kristaps  429: }
                    430:
1.231     schwarze  431: struct mdoc_node *
1.213     schwarze  432: mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  433:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  434: {
                    435:        struct mdoc_node *p;
                    436:
1.203     schwarze  437:        p = node_alloc(mdoc, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  438:        p->args = args;
                    439:        if (p->args)
1.53      kristaps  440:                (args->refcnt)++;
1.172     kristaps  441:
                    442:        switch (tok) {
1.213     schwarze  443:        case MDOC_Bd:
1.172     kristaps  444:                /* FALLTHROUGH */
1.213     schwarze  445:        case MDOC_Bf:
1.172     kristaps  446:                /* FALLTHROUGH */
1.213     schwarze  447:        case MDOC_Bl:
1.217     schwarze  448:                /* FALLTHROUGH */
                    449:        case MDOC_En:
1.173     kristaps  450:                /* FALLTHROUGH */
1.213     schwarze  451:        case MDOC_Rs:
1.172     kristaps  452:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    453:                break;
                    454:        default:
                    455:                break;
                    456:        }
1.230     schwarze  457:        node_append(mdoc, p);
1.203     schwarze  458:        mdoc->next = MDOC_NEXT_CHILD;
1.231     schwarze  459:        return(p);
1.1       kristaps  460: }
                    461:
1.231     schwarze  462: void
1.213     schwarze  463: mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
1.117     kristaps  464:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  465: {
                    466:        struct mdoc_node *p;
                    467:
1.203     schwarze  468:        p = node_alloc(mdoc, line, pos, tok, MDOC_ELEM);
1.77      kristaps  469:        p->args = args;
                    470:        if (p->args)
1.53      kristaps  471:                (args->refcnt)++;
1.172     kristaps  472:
                    473:        switch (tok) {
1.213     schwarze  474:        case MDOC_An:
1.172     kristaps  475:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    476:                break;
                    477:        default:
                    478:                break;
                    479:        }
1.230     schwarze  480:        node_append(mdoc, p);
1.203     schwarze  481:        mdoc->next = MDOC_NEXT_CHILD;
1.175     kristaps  482: }
1.1       kristaps  483:
1.231     schwarze  484: void
1.203     schwarze  485: mdoc_word_alloc(struct mdoc *mdoc, int line, int pos, const char *p)
1.1       kristaps  486: {
1.91      kristaps  487:        struct mdoc_node *n;
1.1       kristaps  488:
1.203     schwarze  489:        n = node_alloc(mdoc, line, pos, MDOC_MAX, MDOC_TEXT);
                    490:        n->string = roff_strdup(mdoc->roff, p);
1.230     schwarze  491:        node_append(mdoc, n);
1.203     schwarze  492:        mdoc->next = MDOC_NEXT_SIBLING;
1.91      kristaps  493: }
                    494:
1.205     schwarze  495: void
                    496: mdoc_word_append(struct mdoc *mdoc, const char *p)
                    497: {
                    498:        struct mdoc_node        *n;
                    499:        char                    *addstr, *newstr;
                    500:
                    501:        n = mdoc->last;
                    502:        addstr = roff_strdup(mdoc->roff, p);
1.210     schwarze  503:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.205     schwarze  504:        free(addstr);
                    505:        free(n->string);
                    506:        n->string = newstr;
                    507:        mdoc->next = MDOC_NEXT_SIBLING;
                    508: }
1.91      kristaps  509:
1.155     kristaps  510: static void
1.53      kristaps  511: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  512: {
                    513:
1.172     kristaps  514:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  515:                free(p->norm);
1.53      kristaps  516:        if (p->string)
                    517:                free(p->string);
                    518:        if (p->args)
                    519:                mdoc_argv_free(p->args);
1.1       kristaps  520:        free(p);
                    521: }
                    522:
1.121     kristaps  523: static void
1.203     schwarze  524: mdoc_node_unlink(struct mdoc *mdoc, struct mdoc_node *n)
1.121     kristaps  525: {
                    526:
                    527:        /* Adjust siblings. */
                    528:
                    529:        if (n->prev)
                    530:                n->prev->next = n->next;
                    531:        if (n->next)
                    532:                n->next->prev = n->prev;
                    533:
                    534:        /* Adjust parent. */
                    535:
                    536:        if (n->parent) {
                    537:                n->parent->nchild--;
                    538:                if (n->parent->child == n)
                    539:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  540:                if (n->parent->last == n)
                    541:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  542:        }
                    543:
                    544:        /* Adjust parse point, if applicable. */
                    545:
1.203     schwarze  546:        if (mdoc && mdoc->last == n) {
1.121     kristaps  547:                if (n->prev) {
1.203     schwarze  548:                        mdoc->last = n->prev;
                    549:                        mdoc->next = MDOC_NEXT_SIBLING;
1.121     kristaps  550:                } else {
1.203     schwarze  551:                        mdoc->last = n->parent;
                    552:                        mdoc->next = MDOC_NEXT_CHILD;
1.121     kristaps  553:                }
                    554:        }
                    555:
1.203     schwarze  556:        if (mdoc && mdoc->first == n)
                    557:                mdoc->first = NULL;
1.121     kristaps  558: }
                    559:
1.53      kristaps  560: void
1.203     schwarze  561: mdoc_node_delete(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  562: {
                    563:
1.121     kristaps  564:        while (p->child) {
                    565:                assert(p->nchild);
1.203     schwarze  566:                mdoc_node_delete(mdoc, p->child);
1.121     kristaps  567:        }
                    568:        assert(0 == p->nchild);
1.1       kristaps  569:
1.203     schwarze  570:        mdoc_node_unlink(mdoc, p);
1.53      kristaps  571:        mdoc_node_free(p);
1.201     schwarze  572: }
                    573:
1.231     schwarze  574: void
1.203     schwarze  575: mdoc_node_relink(struct mdoc *mdoc, struct mdoc_node *p)
1.201     schwarze  576: {
                    577:
1.203     schwarze  578:        mdoc_node_unlink(mdoc, p);
1.230     schwarze  579:        node_append(mdoc, p);
1.1       kristaps  580: }
                    581:
1.53      kristaps  582: /*
                    583:  * Parse free-form text, that is, a line that does not begin with the
                    584:  * control character.
                    585:  */
                    586: static int
1.203     schwarze  587: mdoc_ptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.1       kristaps  588: {
1.142     kristaps  589:        char             *c, *ws, *end;
                    590:        struct mdoc_node *n;
                    591:
1.203     schwarze  592:        assert(mdoc->last);
                    593:        n = mdoc->last;
1.142     kristaps  594:
                    595:        /*
1.144     kristaps  596:         * Divert directly to list processing if we're encountering a
1.142     kristaps  597:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  598:         * (a MDOC_BODY means it's already open, in which case we should
                    599:         * process within its context in the normal way).
1.142     kristaps  600:         */
                    601:
1.143     kristaps  602:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  603:            LIST_column == n->norm->Bl.type) {
1.144     kristaps  604:                /* `Bl' is open without any children. */
1.203     schwarze  605:                mdoc->flags |= MDOC_FREECOL;
1.232     schwarze  606:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    607:                return(1);
1.142     kristaps  608:        }
                    609:
                    610:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  611:            NULL != n->parent &&
                    612:            MDOC_Bl == n->parent->tok &&
                    613:            LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  614:                /* `Bl' has block-level `It' children. */
1.203     schwarze  615:                mdoc->flags |= MDOC_FREECOL;
1.232     schwarze  616:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    617:                return(1);
1.142     kristaps  618:        }
1.124     kristaps  619:
1.137     schwarze  620:        /*
                    621:         * Search for the beginning of unescaped trailing whitespace (ws)
                    622:         * and for the first character not to be output (end).
                    623:         */
1.139     kristaps  624:
                    625:        /* FIXME: replace with strcspn(). */
1.137     schwarze  626:        ws = NULL;
                    627:        for (c = end = buf + offs; *c; c++) {
                    628:                switch (*c) {
                    629:                case ' ':
                    630:                        if (NULL == ws)
                    631:                                ws = c;
                    632:                        continue;
                    633:                case '\t':
                    634:                        /*
                    635:                         * Always warn about trailing tabs,
                    636:                         * even outside literal context,
                    637:                         * where they should be put on the next line.
                    638:                         */
                    639:                        if (NULL == ws)
                    640:                                ws = c;
                    641:                        /*
                    642:                         * Strip trailing tabs in literal context only;
                    643:                         * outside, they affect the next line.
                    644:                         */
1.203     schwarze  645:                        if (MDOC_LITERAL & mdoc->flags)
1.137     schwarze  646:                                continue;
                    647:                        break;
                    648:                case '\\':
                    649:                        /* Skip the escaped character, too, if any. */
                    650:                        if (c[1])
                    651:                                c++;
                    652:                        /* FALLTHROUGH */
                    653:                default:
                    654:                        ws = NULL;
                    655:                        break;
                    656:                }
                    657:                end = c + 1;
                    658:        }
                    659:        *end = '\0';
1.91      kristaps  660:
1.137     schwarze  661:        if (ws)
1.218     schwarze  662:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    663:                    line, (int)(ws-buf), NULL);
1.115     kristaps  664:
1.231     schwarze  665:        if (buf[offs] == '\0' && ! (mdoc->flags & MDOC_LITERAL)) {
1.218     schwarze  666:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    667:                    line, (int)(c - buf), NULL);
1.124     kristaps  668:
1.119     kristaps  669:                /*
1.165     schwarze  670:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  671:                 * blank lines aren't allowed, but enough manuals assume this
                    672:                 * behaviour that we want to work around it.
1.119     kristaps  673:                 */
1.231     schwarze  674:                mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL);
1.203     schwarze  675:                mdoc->next = MDOC_NEXT_SIBLING;
1.230     schwarze  676:                mdoc_valid_post(mdoc);
                    677:                return(1);
1.119     kristaps  678:        }
1.68      kristaps  679:
1.231     schwarze  680:        mdoc_word_alloc(mdoc, line, offs, buf+offs);
1.91      kristaps  681:
1.231     schwarze  682:        if (mdoc->flags & MDOC_LITERAL)
1.137     schwarze  683:                return(1);
1.128     kristaps  684:
                    685:        /*
                    686:         * End-of-sentence check.  If the last character is an unescaped
                    687:         * EOS character, then flag the node as being the end of a
                    688:         * sentence.  The front-end will know how to interpret this.
                    689:         */
1.132     kristaps  690:
1.137     schwarze  691:        assert(buf < end);
                    692:
1.207     schwarze  693:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.203     schwarze  694:                mdoc->last->flags |= MDOC_EOS;
1.128     kristaps  695:        return(1);
1.1       kristaps  696: }
                    697:
1.53      kristaps  698: /*
                    699:  * Parse a macro line, that is, a line beginning with the control
                    700:  * character.
                    701:  */
1.155     kristaps  702: static int
1.203     schwarze  703: mdoc_pmacro(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  704: {
1.229     schwarze  705:        struct mdoc_node *n;
                    706:        const char       *cp;
1.144     kristaps  707:        enum mdoct        tok;
1.188     kristaps  708:        int               i, sv;
1.144     kristaps  709:        char              mac[5];
1.63      kristaps  710:
1.188     kristaps  711:        sv = offs;
1.130     kristaps  712:
1.213     schwarze  713:        /*
1.162     schwarze  714:         * Copy the first word into a nil-terminated buffer.
1.229     schwarze  715:         * Stop when a space, tab, escape, or eoln is encountered.
1.160     kristaps  716:         */
1.1       kristaps  717:
1.188     kristaps  718:        i = 0;
1.229     schwarze  719:        while (i < 4 && strchr(" \t\\", buf[offs]) == NULL)
1.188     kristaps  720:                mac[i++] = buf[offs++];
                    721:
                    722:        mac[i] = '\0';
                    723:
1.214     schwarze  724:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  725:
1.229     schwarze  726:        if (tok == MDOC_MAX) {
1.222     schwarze  727:                mandoc_msg(MANDOCERR_MACRO, mdoc->parse,
                    728:                    ln, sv, buf + sv - 1);
1.58      kristaps  729:                return(1);
1.53      kristaps  730:        }
1.1       kristaps  731:
1.229     schwarze  732:        /* Skip a leading escape sequence or tab. */
1.160     kristaps  733:
1.229     schwarze  734:        switch (buf[offs]) {
                    735:        case '\\':
                    736:                cp = buf + offs + 1;
                    737:                mandoc_escape(&cp, NULL, NULL);
                    738:                offs = cp - buf;
                    739:                break;
                    740:        case '\t':
1.188     kristaps  741:                offs++;
1.229     schwarze  742:                break;
                    743:        default:
                    744:                break;
                    745:        }
1.160     kristaps  746:
                    747:        /* Jump to the next non-whitespace word. */
1.1       kristaps  748:
1.188     kristaps  749:        while (buf[offs] && ' ' == buf[offs])
                    750:                offs++;
1.1       kristaps  751:
1.213     schwarze  752:        /*
1.125     kristaps  753:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    754:         * into the parser as "text", so we only warn about spaces here.
                    755:         */
1.115     kristaps  756:
1.188     kristaps  757:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.218     schwarze  758:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    759:                    ln, offs - 1, NULL);
1.115     kristaps  760:
1.144     kristaps  761:        /*
                    762:         * If an initial macro or a list invocation, divert directly
                    763:         * into macro processing.
                    764:         */
                    765:
1.232     schwarze  766:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok) {
                    767:                mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
                    768:                return(1);
                    769:        }
1.144     kristaps  770:
1.203     schwarze  771:        n = mdoc->last;
                    772:        assert(mdoc->last);
1.144     kristaps  773:
                    774:        /*
                    775:         * If the first macro of a `Bl -column', open an `It' block
                    776:         * context around the parsed macro.
                    777:         */
                    778:
                    779:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.213     schwarze  780:            LIST_column == n->norm->Bl.type) {
1.203     schwarze  781:                mdoc->flags |= MDOC_FREECOL;
1.232     schwarze  782:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    783:                return(1);
1.144     kristaps  784:        }
                    785:
                    786:        /*
                    787:         * If we're following a block-level `It' within a `Bl -column'
                    788:         * context (perhaps opened in the above block or in ptext()),
                    789:         * then open an `It' block context around the parsed macro.
1.98      kristaps  790:         */
1.144     kristaps  791:
                    792:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.213     schwarze  793:            NULL != n->parent &&
                    794:            MDOC_Bl == n->parent->tok &&
                    795:            LIST_column == n->parent->norm->Bl.type) {
1.203     schwarze  796:                mdoc->flags |= MDOC_FREECOL;
1.232     schwarze  797:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    798:                return(1);
1.144     kristaps  799:        }
                    800:
                    801:        /* Normal processing of a macro. */
                    802:
1.232     schwarze  803:        mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
1.208     schwarze  804:
                    805:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    806:
                    807:        if (mdoc->quick && MDOC_Sh == tok &&
                    808:            SEC_NAME != mdoc->last->sec)
                    809:                return(2);
1.1       kristaps  810:
1.53      kristaps  811:        return(1);
1.1       kristaps  812: }
1.100     kristaps  813:
1.186     kristaps  814: enum mdelim
                    815: mdoc_isdelim(const char *p)
                    816: {
                    817:
                    818:        if ('\0' == p[0])
                    819:                return(DELIM_NONE);
                    820:
                    821:        if ('\0' == p[1])
                    822:                switch (p[0]) {
1.213     schwarze  823:                case '(':
1.186     kristaps  824:                        /* FALLTHROUGH */
1.213     schwarze  825:                case '[':
1.186     kristaps  826:                        return(DELIM_OPEN);
1.213     schwarze  827:                case '|':
1.186     kristaps  828:                        return(DELIM_MIDDLE);
1.213     schwarze  829:                case '.':
1.186     kristaps  830:                        /* FALLTHROUGH */
1.213     schwarze  831:                case ',':
1.186     kristaps  832:                        /* FALLTHROUGH */
1.213     schwarze  833:                case ';':
1.186     kristaps  834:                        /* FALLTHROUGH */
1.213     schwarze  835:                case ':':
1.186     kristaps  836:                        /* FALLTHROUGH */
1.213     schwarze  837:                case '?':
1.186     kristaps  838:                        /* FALLTHROUGH */
1.213     schwarze  839:                case '!':
1.186     kristaps  840:                        /* FALLTHROUGH */
1.213     schwarze  841:                case ')':
1.186     kristaps  842:                        /* FALLTHROUGH */
1.213     schwarze  843:                case ']':
1.186     kristaps  844:                        return(DELIM_CLOSE);
                    845:                default:
                    846:                        return(DELIM_NONE);
                    847:                }
                    848:
                    849:        if ('\\' != p[0])
                    850:                return(DELIM_NONE);
1.100     kristaps  851:
1.186     kristaps  852:        if (0 == strcmp(p + 1, "."))
                    853:                return(DELIM_CLOSE);
1.200     schwarze  854:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.186     kristaps  855:                return(DELIM_MIDDLE);
                    856:
                    857:        return(DELIM_NONE);
1.211     schwarze  858: }
                    859:
                    860: void
                    861: mdoc_deroff(char **dest, const struct mdoc_node *n)
                    862: {
                    863:        char    *cp;
                    864:        size_t   sz;
                    865:
                    866:        if (MDOC_TEXT != n->type) {
                    867:                for (n = n->child; n; n = n->next)
                    868:                        mdoc_deroff(dest, n);
                    869:                return;
                    870:        }
                    871:
                    872:        /* Skip leading whitespace. */
                    873:
                    874:        for (cp = n->string; '\0' != *cp; cp++)
                    875:                if (0 == isspace((unsigned char)*cp))
                    876:                        break;
                    877:
                    878:        /* Skip trailing whitespace. */
                    879:
                    880:        for (sz = strlen(cp); sz; sz--)
                    881:                if (0 == isspace((unsigned char)cp[sz-1]))
                    882:                        break;
                    883:
                    884:        /* Skip empty strings. */
                    885:
                    886:        if (0 == sz)
                    887:                return;
                    888:
                    889:        if (NULL == *dest) {
                    890:                *dest = mandoc_strndup(cp, sz);
                    891:                return;
                    892:        }
                    893:
                    894:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    895:        free(*dest);
                    896:        *dest = cp;
1.186     kristaps  897: }

CVSweb