[BACK]Return to mdoc.c CVS log [TXT][DIR] Up to [local] / src / usr.bin / mandoc

Annotation of src/usr.bin/mandoc/mdoc.c, Revision 1.132

1.132   ! schwarze    1: /*     $OpenBSD: mdoc.c,v 1.131 2015/04/18 16:04:40 schwarze Exp $ */
1.1       kristaps    2: /*
1.77      schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.126     schwarze    4:  * Copyright (c) 2010, 2012-2015 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    5:  *
                      6:  * Permission to use, copy, modify, and distribute this software for any
1.3       schwarze    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:  *
1.130     schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.3       schwarze   11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.130     schwarze   12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.3       schwarze   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:  */
1.28      schwarze   18: #include <sys/types.h>
                     19:
1.1       kristaps   20: #include <assert.h>
1.101     schwarze   21: #include <ctype.h>
1.1       kristaps   22: #include <stdarg.h>
                     23: #include <stdio.h>
                     24: #include <stdlib.h>
                     25: #include <string.h>
1.39      schwarze   26: #include <time.h>
1.1       kristaps   27:
1.128     schwarze   28: #include "mandoc_aux.h"
                     29: #include "mandoc.h"
                     30: #include "roff.h"
1.83      schwarze   31: #include "mdoc.h"
1.128     schwarze   32: #include "libmandoc.h"
1.1       kristaps   33: #include "libmdoc.h"
                     34:
1.106     schwarze   35: const  char *const __mdoc_macronames[MDOC_MAX + 1] = {
1.7       schwarze   36:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   37:        "Sh",           "Ss",           "Pp",           "D1",
                     38:        "Dl",           "Bd",           "Ed",           "Bl",
                     39:        "El",           "It",           "Ad",           "An",
                     40:        "Ar",           "Cd",           "Cm",           "Dv",
                     41:        "Er",           "Ev",           "Ex",           "Fa",
                     42:        "Fd",           "Fl",           "Fn",           "Ft",
                     43:        "Ic",           "In",           "Li",           "Nd",
                     44:        "Nm",           "Op",           "Ot",           "Pa",
                     45:        "Rv",           "St",           "Va",           "Vt",
1.33      schwarze   46:        "Xr",           "%A",           "%B",           "%D",
                     47:        "%I",           "%J",           "%N",           "%O",
                     48:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   49:        "Ac",           "Ao",           "Aq",           "At",
                     50:        "Bc",           "Bf",           "Bo",           "Bq",
                     51:        "Bsx",          "Bx",           "Db",           "Dc",
                     52:        "Do",           "Dq",           "Ec",           "Ef",
                     53:        "Em",           "Eo",           "Fx",           "Ms",
                     54:        "No",           "Ns",           "Nx",           "Ox",
                     55:        "Pc",           "Pf",           "Po",           "Pq",
                     56:        "Qc",           "Ql",           "Qo",           "Qq",
                     57:        "Re",           "Rs",           "Sc",           "So",
                     58:        "Sq",           "Sm",           "Sx",           "Sy",
                     59:        "Tn",           "Ux",           "Xc",           "Xo",
                     60:        "Fo",           "Fc",           "Oo",           "Oc",
                     61:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze   62:        "Fr",           "Ud",           "Lb",           "Lp",
                     63:        "Lk",           "Mt",           "Brq",          "Bro",
1.33      schwarze   64:        "Brc",          "%C",           "Es",           "En",
                     65:        "Dx",           "%Q",           "br",           "sp",
1.106     schwarze   66:        "%U",           "Ta",           "ll",           "text",
1.1       kristaps   67:        };
                     68:
1.103     schwarze   69: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
1.1       kristaps   70:        "split",                "nosplit",              "ragged",
1.103     schwarze   71:        "unfilled",             "literal",              "file",
                     72:        "offset",               "bullet",               "dash",
                     73:        "hyphen",               "item",                 "enum",
                     74:        "tag",                  "diag",                 "hang",
                     75:        "ohang",                "inset",                "column",
                     76:        "width",                "compact",              "std",
1.1       kristaps   77:        "filled",               "words",                "emphasis",
1.30      schwarze   78:        "symbolic",             "nested",               "centered"
1.1       kristaps   79:        };
                     80:
                     81: const  char * const *mdoc_macronames = __mdoc_macronames;
                     82: const  char * const *mdoc_argnames = __mdoc_argnames;
                     83:
1.129     schwarze   84: static void              mdoc_node_free(struct roff_node *);
1.131     schwarze   85: static void              mdoc_node_unlink(struct roff_man *,
1.129     schwarze   86:                                struct roff_node *);
1.131     schwarze   87: static void              mdoc_free1(struct roff_man *);
                     88: static void              mdoc_alloc1(struct roff_man *);
                     89: static struct roff_node *node_alloc(struct roff_man *, int, int,
1.129     schwarze   90:                                int, enum roff_type);
1.131     schwarze   91: static void              node_append(struct roff_man *, struct roff_node *);
                     92: static int               mdoc_ptext(struct roff_man *, int, char *, int);
                     93: static int               mdoc_pmacro(struct roff_man *, int, char *, int);
1.1       kristaps   94:
1.103     schwarze   95:
1.129     schwarze   96: const struct roff_node *
1.131     schwarze   97: mdoc_node(const struct roff_man *mdoc)
1.1       kristaps   98: {
                     99:
1.93      schwarze  100:        return(mdoc->first);
1.1       kristaps  101: }
                    102:
1.130     schwarze  103: const struct roff_meta *
1.131     schwarze  104: mdoc_meta(const struct roff_man *mdoc)
1.1       kristaps  105: {
                    106:
1.93      schwarze  107:        return(&mdoc->meta);
1.1       kristaps  108: }
                    109:
1.8       schwarze  110: /*
                    111:  * Frees volatile resources (parse tree, meta-data, fields).
                    112:  */
1.1       kristaps  113: static void
1.131     schwarze  114: mdoc_free1(struct roff_man *mdoc)
1.1       kristaps  115: {
                    116:
                    117:        if (mdoc->first)
1.40      schwarze  118:                mdoc_node_delete(mdoc, mdoc->first);
1.113     schwarze  119:        free(mdoc->meta.msec);
                    120:        free(mdoc->meta.vol);
                    121:        free(mdoc->meta.arch);
                    122:        free(mdoc->meta.date);
                    123:        free(mdoc->meta.title);
                    124:        free(mdoc->meta.os);
                    125:        free(mdoc->meta.name);
1.1       kristaps  126: }
                    127:
1.8       schwarze  128: /*
                    129:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    130:  */
1.32      schwarze  131: static void
1.131     schwarze  132: mdoc_alloc1(struct roff_man *mdoc)
1.1       kristaps  133: {
                    134:
1.130     schwarze  135:        memset(&mdoc->meta, 0, sizeof(mdoc->meta));
1.132   ! schwarze  136:        mdoc->macroset = MACROSET_MDOC;
1.1       kristaps  137:        mdoc->flags = 0;
1.9       schwarze  138:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.129     schwarze  139:        mdoc->last = mandoc_calloc(1, sizeof(*mdoc->last));
1.1       kristaps  140:        mdoc->first = mdoc->last;
1.128     schwarze  141:        mdoc->last->type = ROFFT_ROOT;
1.86      schwarze  142:        mdoc->last->tok = MDOC_MAX;
1.131     schwarze  143:        mdoc->next = ROFF_NEXT_CHILD;
1.1       kristaps  144: }
                    145:
                    146: /*
1.8       schwarze  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.
1.1       kristaps  151:  */
1.32      schwarze  152: void
1.131     schwarze  153: mdoc_reset(struct roff_man *mdoc)
1.1       kristaps  154: {
                    155:
                    156:        mdoc_free1(mdoc);
1.32      schwarze  157:        mdoc_alloc1(mdoc);
1.1       kristaps  158: }
                    159:
                    160: /*
1.8       schwarze  161:  * Completely free up all volatile and non-volatile parse resources.
                    162:  * After invocation, the pointer is no longer usable.
1.1       kristaps  163:  */
                    164: void
1.131     schwarze  165: mdoc_free(struct roff_man *mdoc)
1.1       kristaps  166: {
                    167:
                    168:        mdoc_free1(mdoc);
                    169:        free(mdoc);
                    170: }
                    171:
1.8       schwarze  172: /*
1.103     schwarze  173:  * Allocate volatile and non-volatile parse resources.
1.8       schwarze  174:  */
1.131     schwarze  175: struct roff_man *
1.98      schwarze  176: mdoc_alloc(struct roff *roff, struct mparse *parse,
1.110     schwarze  177:        const char *defos, int quick)
1.1       kristaps  178: {
1.131     schwarze  179:        struct roff_man *p;
1.1       kristaps  180:
1.131     schwarze  181:        p = mandoc_calloc(1, sizeof(*p));
1.32      schwarze  182:
1.83      schwarze  183:        p->parse = parse;
1.87      schwarze  184:        p->defos = defos;
1.98      schwarze  185:        p->quick = quick;
1.84      schwarze  186:        p->roff = roff;
1.1       kristaps  187:
1.32      schwarze  188:        mdoc_hash_init();
                    189:        mdoc_alloc1(p);
                    190:        return(p);
1.1       kristaps  191: }
                    192:
1.123     schwarze  193: void
1.131     schwarze  194: mdoc_endparse(struct roff_man *mdoc)
1.1       kristaps  195: {
                    196:
1.121     schwarze  197:        mdoc_macroend(mdoc);
1.81      schwarze  198: }
                    199:
1.122     schwarze  200: void
1.131     schwarze  201: mdoc_addeqn(struct roff_man *mdoc, const struct eqn *ep)
1.81      schwarze  202: {
1.129     schwarze  203:        struct roff_node *n;
1.81      schwarze  204:
1.128     schwarze  205:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, ROFFT_EQN);
1.81      schwarze  206:        n->eqn = ep;
1.116     schwarze  207:        if (ep->ln > mdoc->last->line)
                    208:                n->flags |= MDOC_LINE;
1.119     schwarze  209:        node_append(mdoc, n);
1.131     schwarze  210:        mdoc->next = ROFF_NEXT_SIBLING;
1.1       kristaps  211: }
                    212:
1.122     schwarze  213: void
1.131     schwarze  214: mdoc_addspan(struct roff_man *mdoc, const struct tbl_span *sp)
1.77      schwarze  215: {
1.129     schwarze  216:        struct roff_node *n;
1.77      schwarze  217:
1.128     schwarze  218:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, ROFFT_TBL);
1.79      schwarze  219:        n->span = sp;
1.119     schwarze  220:        node_append(mdoc, n);
1.131     schwarze  221:        mdoc->next = ROFF_NEXT_SIBLING;
1.77      schwarze  222: }
                    223:
1.1       kristaps  224: /*
                    225:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  226:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  227:  */
                    228: int
1.131     schwarze  229: mdoc_parseln(struct roff_man *mdoc, int ln, char *buf, int offs)
1.1       kristaps  230: {
                    231:
1.128     schwarze  232:        if (mdoc->last->type != ROFFT_EQN || ln > mdoc->last->line)
1.117     schwarze  233:                mdoc->flags |= MDOC_NEWLINE;
1.60      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:         */
1.94      schwarze  241:        if (roff_getreg(mdoc->roff, "nS"))
                    242:                mdoc->flags |= MDOC_SYNOPSIS;
                    243:        else
                    244:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.60      schwarze  245:
1.93      schwarze  246:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
1.103     schwarze  247:            mdoc_pmacro(mdoc, ln, buf, offs) :
                    248:            mdoc_ptext(mdoc, ln, buf, offs));
1.1       kristaps  249: }
                    250:
1.121     schwarze  251: void
1.58      schwarze  252: mdoc_macro(MACRO_PROT_ARGS)
1.1       kristaps  253: {
1.46      schwarze  254:        assert(tok < MDOC_MAX);
                    255:
1.113     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);
1.121     schwarze  261:                        return;
1.113     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:                }
1.93      schwarze  270:                if (NULL == mdoc->meta.vol)
                    271:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    272:                mdoc->flags |= MDOC_PBODY;
1.39      schwarze  273:        }
1.121     schwarze  274:        (*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf);
1.1       kristaps  275: }
                    276:
                    277:
1.119     schwarze  278: static void
1.131     schwarze  279: node_append(struct roff_man *mdoc, struct roff_node *p)
1.1       kristaps  280: {
                    281:
                    282:        assert(mdoc->last);
                    283:        assert(mdoc->first);
1.128     schwarze  284:        assert(p->type != ROFFT_ROOT);
1.1       kristaps  285:
                    286:        switch (mdoc->next) {
1.131     schwarze  287:        case ROFF_NEXT_SIBLING:
1.1       kristaps  288:                mdoc->last->next = p;
                    289:                p->prev = mdoc->last;
                    290:                p->parent = mdoc->last->parent;
                    291:                break;
1.131     schwarze  292:        case ROFF_NEXT_CHILD:
1.1       kristaps  293:                mdoc->last->child = p;
                    294:                p->parent = mdoc->last;
                    295:                break;
                    296:        default:
                    297:                abort();
                    298:                /* NOTREACHED */
                    299:        }
                    300:
1.10      schwarze  301:        p->parent->nchild++;
                    302:
1.75      schwarze  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) {
1.128     schwarze  309:        case ROFFT_BODY:
1.92      schwarze  310:                if (ENDBODY_NOT != p->end)
                    311:                        break;
1.75      schwarze  312:                /* FALLTHROUGH */
1.128     schwarze  313:        case ROFFT_TAIL:
1.75      schwarze  314:                /* FALLTHROUGH */
1.128     schwarze  315:        case ROFFT_HEAD:
1.75      schwarze  316:                p->norm = p->parent->norm;
                    317:                break;
                    318:        default:
                    319:                break;
                    320:        }
                    321:
1.119     schwarze  322:        mdoc_valid_pre(mdoc, p);
1.1       kristaps  323:
                    324:        switch (p->type) {
1.128     schwarze  325:        case ROFFT_HEAD:
                    326:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  327:                p->parent->head = p;
                    328:                break;
1.128     schwarze  329:        case ROFFT_TAIL:
                    330:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  331:                p->parent->tail = p;
                    332:                break;
1.128     schwarze  333:        case ROFFT_BODY:
1.59      schwarze  334:                if (p->end)
                    335:                        break;
1.128     schwarze  336:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  337:                p->parent->body = p;
                    338:                break;
                    339:        default:
                    340:                break;
                    341:        }
                    342:
                    343:        mdoc->last = p;
                    344:
                    345:        switch (p->type) {
1.128     schwarze  346:        case ROFFT_TBL:
1.77      schwarze  347:                /* FALLTHROUGH */
1.128     schwarze  348:        case ROFFT_TEXT:
1.119     schwarze  349:                mdoc_valid_post(mdoc);
1.1       kristaps  350:                break;
                    351:        default:
                    352:                break;
                    353:        }
                    354: }
                    355:
1.129     schwarze  356: static struct roff_node *
1.131     schwarze  357: node_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  358:        int tok, enum roff_type type)
1.1       kristaps  359: {
1.129     schwarze  360:        struct roff_node *p;
1.1       kristaps  361:
1.129     schwarze  362:        p = mandoc_calloc(1, sizeof(*p));
1.93      schwarze  363:        p->sec = mdoc->lastsec;
1.1       kristaps  364:        p->line = line;
                    365:        p->pos = pos;
                    366:        p->tok = tok;
1.37      schwarze  367:        p->type = type;
1.58      schwarze  368:
                    369:        /* Flag analysis. */
                    370:
1.93      schwarze  371:        if (MDOC_SYNOPSIS & mdoc->flags)
1.60      schwarze  372:                p->flags |= MDOC_SYNPRETTY;
                    373:        else
                    374:                p->flags &= ~MDOC_SYNPRETTY;
1.93      schwarze  375:        if (MDOC_NEWLINE & mdoc->flags)
1.47      schwarze  376:                p->flags |= MDOC_LINE;
1.93      schwarze  377:        mdoc->flags &= ~MDOC_NEWLINE;
1.58      schwarze  378:
1.1       kristaps  379:        return(p);
                    380: }
                    381:
1.120     schwarze  382: void
1.131     schwarze  383: mdoc_tail_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  384: {
1.129     schwarze  385:        struct roff_node *p;
1.1       kristaps  386:
1.128     schwarze  387:        p = node_alloc(mdoc, line, pos, tok, ROFFT_TAIL);
1.119     schwarze  388:        node_append(mdoc, p);
1.131     schwarze  389:        mdoc->next = ROFF_NEXT_CHILD;
1.1       kristaps  390: }
                    391:
1.129     schwarze  392: struct roff_node *
1.131     schwarze  393: mdoc_head_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  394: {
1.129     schwarze  395:        struct roff_node *p;
1.1       kristaps  396:
1.93      schwarze  397:        assert(mdoc->first);
                    398:        assert(mdoc->last);
1.128     schwarze  399:        p = node_alloc(mdoc, line, pos, tok, ROFFT_HEAD);
1.119     schwarze  400:        node_append(mdoc, p);
1.131     schwarze  401:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  402:        return(p);
1.1       kristaps  403: }
                    404:
1.129     schwarze  405: struct roff_node *
1.131     schwarze  406: mdoc_body_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  407: {
1.129     schwarze  408:        struct roff_node *p;
1.1       kristaps  409:
1.128     schwarze  410:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BODY);
1.119     schwarze  411:        node_append(mdoc, p);
1.131     schwarze  412:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  413:        return(p);
1.59      schwarze  414: }
                    415:
1.129     schwarze  416: struct roff_node *
1.131     schwarze  417: mdoc_endbody_alloc(struct roff_man *mdoc, int line, int pos, int tok,
1.129     schwarze  418:                struct roff_node *body, enum mdoc_endbody end)
1.59      schwarze  419: {
1.129     schwarze  420:        struct roff_node *p;
1.59      schwarze  421:
1.126     schwarze  422:        body->flags |= MDOC_ENDED;
                    423:        body->parent->flags |= MDOC_ENDED;
1.128     schwarze  424:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BODY);
1.126     schwarze  425:        p->body = body;
1.92      schwarze  426:        p->norm = body->norm;
1.59      schwarze  427:        p->end = end;
1.119     schwarze  428:        node_append(mdoc, p);
1.131     schwarze  429:        mdoc->next = ROFF_NEXT_SIBLING;
1.124     schwarze  430:        return(p);
1.1       kristaps  431: }
                    432:
1.129     schwarze  433: struct roff_node *
1.131     schwarze  434: mdoc_block_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  435:        int tok, struct mdoc_arg *args)
1.1       kristaps  436: {
1.129     schwarze  437:        struct roff_node *p;
1.1       kristaps  438:
1.128     schwarze  439:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BLOCK);
1.4       schwarze  440:        p->args = args;
                    441:        if (p->args)
1.1       kristaps  442:                (args->refcnt)++;
1.75      schwarze  443:
                    444:        switch (tok) {
1.103     schwarze  445:        case MDOC_Bd:
1.75      schwarze  446:                /* FALLTHROUGH */
1.103     schwarze  447:        case MDOC_Bf:
1.75      schwarze  448:                /* FALLTHROUGH */
1.103     schwarze  449:        case MDOC_Bl:
1.107     schwarze  450:                /* FALLTHROUGH */
                    451:        case MDOC_En:
1.75      schwarze  452:                /* FALLTHROUGH */
1.103     schwarze  453:        case MDOC_Rs:
1.75      schwarze  454:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    455:                break;
                    456:        default:
                    457:                break;
                    458:        }
1.119     schwarze  459:        node_append(mdoc, p);
1.131     schwarze  460:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  461:        return(p);
1.1       kristaps  462: }
                    463:
1.120     schwarze  464: void
1.131     schwarze  465: mdoc_elem_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  466:        int tok, struct mdoc_arg *args)
1.1       kristaps  467: {
1.129     schwarze  468:        struct roff_node *p;
1.1       kristaps  469:
1.128     schwarze  470:        p = node_alloc(mdoc, line, pos, tok, ROFFT_ELEM);
1.4       schwarze  471:        p->args = args;
                    472:        if (p->args)
1.1       kristaps  473:                (args->refcnt)++;
1.75      schwarze  474:
                    475:        switch (tok) {
1.103     schwarze  476:        case MDOC_An:
1.75      schwarze  477:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    478:                break;
                    479:        default:
                    480:                break;
                    481:        }
1.119     schwarze  482:        node_append(mdoc, p);
1.131     schwarze  483:        mdoc->next = ROFF_NEXT_CHILD;
1.1       kristaps  484: }
                    485:
1.120     schwarze  486: void
1.131     schwarze  487: mdoc_word_alloc(struct roff_man *mdoc, int line, int pos, const char *p)
1.1       kristaps  488: {
1.129     schwarze  489:        struct roff_node *n;
1.1       kristaps  490:
1.128     schwarze  491:        n = node_alloc(mdoc, line, pos, MDOC_MAX, ROFFT_TEXT);
1.93      schwarze  492:        n->string = roff_strdup(mdoc->roff, p);
1.119     schwarze  493:        node_append(mdoc, n);
1.131     schwarze  494:        mdoc->next = ROFF_NEXT_SIBLING;
1.19      schwarze  495: }
                    496:
1.95      schwarze  497: void
1.131     schwarze  498: mdoc_word_append(struct roff_man *mdoc, const char *p)
1.95      schwarze  499: {
1.129     schwarze  500:        struct roff_node        *n;
1.95      schwarze  501:        char                    *addstr, *newstr;
                    502:
                    503:        n = mdoc->last;
                    504:        addstr = roff_strdup(mdoc->roff, p);
1.100     schwarze  505:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.95      schwarze  506:        free(addstr);
                    507:        free(n->string);
                    508:        n->string = newstr;
1.131     schwarze  509:        mdoc->next = ROFF_NEXT_SIBLING;
1.95      schwarze  510: }
1.19      schwarze  511:
1.61      schwarze  512: static void
1.129     schwarze  513: mdoc_node_free(struct roff_node *p)
1.1       kristaps  514: {
1.61      schwarze  515:
1.128     schwarze  516:        if (p->type == ROFFT_BLOCK || p->type == ROFFT_ELEM)
1.75      schwarze  517:                free(p->norm);
1.1       kristaps  518:        if (p->string)
                    519:                free(p->string);
                    520:        if (p->args)
                    521:                mdoc_argv_free(p->args);
                    522:        free(p);
                    523: }
                    524:
1.40      schwarze  525: static void
1.131     schwarze  526: mdoc_node_unlink(struct roff_man *mdoc, struct roff_node *n)
1.40      schwarze  527: {
                    528:
                    529:        /* Adjust siblings. */
                    530:
                    531:        if (n->prev)
                    532:                n->prev->next = n->next;
                    533:        if (n->next)
                    534:                n->next->prev = n->prev;
                    535:
                    536:        /* Adjust parent. */
                    537:
                    538:        if (n->parent) {
                    539:                n->parent->nchild--;
                    540:                if (n->parent->child == n)
                    541:                        n->parent->child = n->prev ? n->prev : n->next;
1.72      schwarze  542:                if (n->parent->last == n)
                    543:                        n->parent->last = n->prev ? n->prev : NULL;
1.40      schwarze  544:        }
                    545:
                    546:        /* Adjust parse point, if applicable. */
                    547:
1.93      schwarze  548:        if (mdoc && mdoc->last == n) {
1.40      schwarze  549:                if (n->prev) {
1.93      schwarze  550:                        mdoc->last = n->prev;
1.131     schwarze  551:                        mdoc->next = ROFF_NEXT_SIBLING;
1.40      schwarze  552:                } else {
1.93      schwarze  553:                        mdoc->last = n->parent;
1.131     schwarze  554:                        mdoc->next = ROFF_NEXT_CHILD;
1.40      schwarze  555:                }
                    556:        }
                    557:
1.93      schwarze  558:        if (mdoc && mdoc->first == n)
                    559:                mdoc->first = NULL;
1.40      schwarze  560: }
                    561:
1.1       kristaps  562: void
1.131     schwarze  563: mdoc_node_delete(struct roff_man *mdoc, struct roff_node *p)
1.1       kristaps  564: {
                    565:
1.40      schwarze  566:        while (p->child) {
                    567:                assert(p->nchild);
1.93      schwarze  568:                mdoc_node_delete(mdoc, p->child);
1.40      schwarze  569:        }
                    570:        assert(0 == p->nchild);
1.1       kristaps  571:
1.93      schwarze  572:        mdoc_node_unlink(mdoc, p);
1.1       kristaps  573:        mdoc_node_free(p);
1.91      schwarze  574: }
                    575:
1.120     schwarze  576: void
1.131     schwarze  577: mdoc_node_relink(struct roff_man *mdoc, struct roff_node *p)
1.91      schwarze  578: {
                    579:
1.93      schwarze  580:        mdoc_node_unlink(mdoc, p);
1.119     schwarze  581:        node_append(mdoc, p);
1.1       kristaps  582: }
                    583:
                    584: /*
                    585:  * Parse free-form text, that is, a line that does not begin with the
                    586:  * control character.
                    587:  */
                    588: static int
1.131     schwarze  589: mdoc_ptext(struct roff_man *mdoc, int line, char *buf, int offs)
1.1       kristaps  590: {
1.129     schwarze  591:        struct roff_node *n;
1.56      schwarze  592:        char             *c, *ws, *end;
1.46      schwarze  593:
1.93      schwarze  594:        assert(mdoc->last);
                    595:        n = mdoc->last;
1.56      schwarze  596:
                    597:        /*
                    598:         * Divert directly to list processing if we're encountering a
1.128     schwarze  599:         * columnar ROFFT_BLOCK with or without a prior ROFFT_BLOCK entry
                    600:         * (a ROFFT_BODY means it's already open, in which case we should
1.56      schwarze  601:         * process within its context in the normal way).
                    602:         */
                    603:
1.128     schwarze  604:        if (n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
1.127     schwarze  605:            n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) {
1.56      schwarze  606:                /* `Bl' is open without any children. */
1.93      schwarze  607:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  608:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    609:                return(1);
1.56      schwarze  610:        }
                    611:
1.128     schwarze  612:        if (n->tok == MDOC_It && n->type == ROFFT_BLOCK &&
1.103     schwarze  613:            NULL != n->parent &&
                    614:            MDOC_Bl == n->parent->tok &&
                    615:            LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  616:                /* `Bl' has block-level `It' children. */
1.93      schwarze  617:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  618:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    619:                return(1);
1.56      schwarze  620:        }
                    621:
1.52      schwarze  622:        /*
                    623:         * Search for the beginning of unescaped trailing whitespace (ws)
                    624:         * and for the first character not to be output (end).
                    625:         */
1.56      schwarze  626:
                    627:        /* FIXME: replace with strcspn(). */
1.52      schwarze  628:        ws = NULL;
1.53      schwarze  629:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  630:                switch (*c) {
                    631:                case ' ':
                    632:                        if (NULL == ws)
                    633:                                ws = c;
                    634:                        continue;
                    635:                case '\t':
                    636:                        /*
                    637:                         * Always warn about trailing tabs,
                    638:                         * even outside literal context,
                    639:                         * where they should be put on the next line.
                    640:                         */
                    641:                        if (NULL == ws)
                    642:                                ws = c;
                    643:                        /*
                    644:                         * Strip trailing tabs in literal context only;
                    645:                         * outside, they affect the next line.
                    646:                         */
1.93      schwarze  647:                        if (MDOC_LITERAL & mdoc->flags)
1.52      schwarze  648:                                continue;
                    649:                        break;
                    650:                case '\\':
                    651:                        /* Skip the escaped character, too, if any. */
                    652:                        if (c[1])
                    653:                                c++;
                    654:                        /* FALLTHROUGH */
                    655:                default:
                    656:                        ws = NULL;
                    657:                        break;
                    658:                }
                    659:                end = c + 1;
                    660:        }
                    661:        *end = '\0';
1.19      schwarze  662:
1.52      schwarze  663:        if (ws)
1.108     schwarze  664:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    665:                    line, (int)(ws-buf), NULL);
1.34      schwarze  666:
1.120     schwarze  667:        if (buf[offs] == '\0' && ! (mdoc->flags & MDOC_LITERAL)) {
1.108     schwarze  668:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    669:                    line, (int)(c - buf), NULL);
1.46      schwarze  670:
1.40      schwarze  671:                /*
1.66      schwarze  672:                 * Insert a `sp' in the case of a blank line.  Technically,
1.46      schwarze  673:                 * blank lines aren't allowed, but enough manuals assume this
                    674:                 * behaviour that we want to work around it.
1.40      schwarze  675:                 */
1.120     schwarze  676:                mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL);
1.131     schwarze  677:                mdoc->next = ROFF_NEXT_SIBLING;
1.119     schwarze  678:                mdoc_valid_post(mdoc);
                    679:                return(1);
1.38      schwarze  680:        }
1.1       kristaps  681:
1.120     schwarze  682:        mdoc_word_alloc(mdoc, line, offs, buf+offs);
1.1       kristaps  683:
1.120     schwarze  684:        if (mdoc->flags & MDOC_LITERAL)
1.52      schwarze  685:                return(1);
                    686:
1.35      schwarze  687:        /*
1.48      schwarze  688:         * End-of-sentence check.  If the last character is an unescaped
                    689:         * EOS character, then flag the node as being the end of a
                    690:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  691:         */
                    692:
1.52      schwarze  693:        assert(buf < end);
1.48      schwarze  694:
1.97      schwarze  695:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.93      schwarze  696:                mdoc->last->flags |= MDOC_EOS;
1.1       kristaps  697:        return(1);
                    698: }
1.19      schwarze  699:
1.1       kristaps  700: /*
                    701:  * Parse a macro line, that is, a line beginning with the control
                    702:  * character.
                    703:  */
1.61      schwarze  704: static int
1.131     schwarze  705: mdoc_pmacro(struct roff_man *mdoc, int ln, char *buf, int offs)
1.1       kristaps  706: {
1.129     schwarze  707:        struct roff_node *n;
1.118     schwarze  708:        const char       *cp;
1.129     schwarze  709:        int               tok;
1.83      schwarze  710:        int               i, sv;
1.56      schwarze  711:        char              mac[5];
1.1       kristaps  712:
1.83      schwarze  713:        sv = offs;
1.47      schwarze  714:
1.103     schwarze  715:        /*
1.64      schwarze  716:         * Copy the first word into a nil-terminated buffer.
1.118     schwarze  717:         * Stop when a space, tab, escape, or eoln is encountered.
1.63      schwarze  718:         */
1.1       kristaps  719:
1.83      schwarze  720:        i = 0;
1.118     schwarze  721:        while (i < 4 && strchr(" \t\\", buf[offs]) == NULL)
1.83      schwarze  722:                mac[i++] = buf[offs++];
                    723:
                    724:        mac[i] = '\0';
                    725:
1.104     schwarze  726:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  727:
1.118     schwarze  728:        if (tok == MDOC_MAX) {
1.112     schwarze  729:                mandoc_msg(MANDOCERR_MACRO, mdoc->parse,
                    730:                    ln, sv, buf + sv - 1);
1.1       kristaps  731:                return(1);
                    732:        }
                    733:
1.118     schwarze  734:        /* Skip a leading escape sequence or tab. */
1.63      schwarze  735:
1.118     schwarze  736:        switch (buf[offs]) {
                    737:        case '\\':
                    738:                cp = buf + offs + 1;
                    739:                mandoc_escape(&cp, NULL, NULL);
                    740:                offs = cp - buf;
                    741:                break;
                    742:        case '\t':
1.83      schwarze  743:                offs++;
1.118     schwarze  744:                break;
                    745:        default:
                    746:                break;
                    747:        }
1.63      schwarze  748:
                    749:        /* Jump to the next non-whitespace word. */
1.1       kristaps  750:
1.83      schwarze  751:        while (buf[offs] && ' ' == buf[offs])
                    752:                offs++;
1.34      schwarze  753:
1.103     schwarze  754:        /*
1.46      schwarze  755:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    756:         * into the parser as "text", so we only warn about spaces here.
                    757:         */
1.34      schwarze  758:
1.83      schwarze  759:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.108     schwarze  760:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    761:                    ln, offs - 1, NULL);
1.1       kristaps  762:
1.56      schwarze  763:        /*
                    764:         * If an initial macro or a list invocation, divert directly
                    765:         * into macro processing.
                    766:         */
                    767:
1.121     schwarze  768:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok) {
                    769:                mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
                    770:                return(1);
                    771:        }
1.56      schwarze  772:
1.93      schwarze  773:        n = mdoc->last;
                    774:        assert(mdoc->last);
1.56      schwarze  775:
                    776:        /*
                    777:         * If the first macro of a `Bl -column', open an `It' block
                    778:         * context around the parsed macro.
                    779:         */
                    780:
1.128     schwarze  781:        if (n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
1.127     schwarze  782:            n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) {
1.93      schwarze  783:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  784:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    785:                return(1);
1.56      schwarze  786:        }
                    787:
                    788:        /*
                    789:         * If we're following a block-level `It' within a `Bl -column'
                    790:         * context (perhaps opened in the above block or in ptext()),
                    791:         * then open an `It' block context around the parsed macro.
1.24      schwarze  792:         */
1.56      schwarze  793:
1.128     schwarze  794:        if (n->tok == MDOC_It && n->type == ROFFT_BLOCK &&
1.103     schwarze  795:            NULL != n->parent &&
                    796:            MDOC_Bl == n->parent->tok &&
                    797:            LIST_column == n->parent->norm->Bl.type) {
1.93      schwarze  798:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  799:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    800:                return(1);
1.56      schwarze  801:        }
                    802:
                    803:        /* Normal processing of a macro. */
                    804:
1.121     schwarze  805:        mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
1.98      schwarze  806:
                    807:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    808:
                    809:        if (mdoc->quick && MDOC_Sh == tok &&
                    810:            SEC_NAME != mdoc->last->sec)
                    811:                return(2);
1.1       kristaps  812:
                    813:        return(1);
                    814: }
1.24      schwarze  815:
1.83      schwarze  816: enum mdelim
                    817: mdoc_isdelim(const char *p)
                    818: {
                    819:
                    820:        if ('\0' == p[0])
                    821:                return(DELIM_NONE);
                    822:
                    823:        if ('\0' == p[1])
                    824:                switch (p[0]) {
1.103     schwarze  825:                case '(':
1.83      schwarze  826:                        /* FALLTHROUGH */
1.103     schwarze  827:                case '[':
1.83      schwarze  828:                        return(DELIM_OPEN);
1.103     schwarze  829:                case '|':
1.83      schwarze  830:                        return(DELIM_MIDDLE);
1.103     schwarze  831:                case '.':
1.83      schwarze  832:                        /* FALLTHROUGH */
1.103     schwarze  833:                case ',':
1.83      schwarze  834:                        /* FALLTHROUGH */
1.103     schwarze  835:                case ';':
1.83      schwarze  836:                        /* FALLTHROUGH */
1.103     schwarze  837:                case ':':
1.83      schwarze  838:                        /* FALLTHROUGH */
1.103     schwarze  839:                case '?':
1.83      schwarze  840:                        /* FALLTHROUGH */
1.103     schwarze  841:                case '!':
1.83      schwarze  842:                        /* FALLTHROUGH */
1.103     schwarze  843:                case ')':
1.83      schwarze  844:                        /* FALLTHROUGH */
1.103     schwarze  845:                case ']':
1.83      schwarze  846:                        return(DELIM_CLOSE);
                    847:                default:
                    848:                        return(DELIM_NONE);
                    849:                }
                    850:
                    851:        if ('\\' != p[0])
                    852:                return(DELIM_NONE);
1.24      schwarze  853:
1.83      schwarze  854:        if (0 == strcmp(p + 1, "."))
                    855:                return(DELIM_CLOSE);
1.90      schwarze  856:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.83      schwarze  857:                return(DELIM_MIDDLE);
                    858:
                    859:        return(DELIM_NONE);
1.101     schwarze  860: }
                    861:
                    862: void
1.129     schwarze  863: mdoc_deroff(char **dest, const struct roff_node *n)
1.101     schwarze  864: {
                    865:        char    *cp;
                    866:        size_t   sz;
                    867:
1.128     schwarze  868:        if (n->type != ROFFT_TEXT) {
1.101     schwarze  869:                for (n = n->child; n; n = n->next)
                    870:                        mdoc_deroff(dest, n);
                    871:                return;
                    872:        }
                    873:
                    874:        /* Skip leading whitespace. */
                    875:
                    876:        for (cp = n->string; '\0' != *cp; cp++)
                    877:                if (0 == isspace((unsigned char)*cp))
                    878:                        break;
                    879:
                    880:        /* Skip trailing whitespace. */
                    881:
                    882:        for (sz = strlen(cp); sz; sz--)
                    883:                if (0 == isspace((unsigned char)cp[sz-1]))
                    884:                        break;
                    885:
                    886:        /* Skip empty strings. */
                    887:
                    888:        if (0 == sz)
                    889:                return;
                    890:
                    891:        if (NULL == *dest) {
                    892:                *dest = mandoc_strndup(cp, sz);
                    893:                return;
                    894:        }
                    895:
                    896:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    897:        free(*dest);
                    898:        *dest = cp;
1.83      schwarze  899: }