[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 1.134

1.134   ! schwarze    1: /*     $OpenBSD: mdoc.c,v 1.133 2015/04/18 17:01:28 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 struct roff_node *node_alloc(struct roff_man *, int, int,
1.129     schwarze   88:                                int, enum roff_type);
1.131     schwarze   89: static void              node_append(struct roff_man *, struct roff_node *);
                     90: static int               mdoc_ptext(struct roff_man *, int, char *, int);
                     91: static int               mdoc_pmacro(struct roff_man *, int, char *, int);
1.1       kristaps   92:
1.103     schwarze   93:
1.129     schwarze   94: const struct roff_node *
1.131     schwarze   95: mdoc_node(const struct roff_man *mdoc)
1.1       kristaps   96: {
                     97:
1.93      schwarze   98:        return(mdoc->first);
1.1       kristaps   99: }
                    100:
1.130     schwarze  101: const struct roff_meta *
1.131     schwarze  102: mdoc_meta(const struct roff_man *mdoc)
1.1       kristaps  103: {
                    104:
1.93      schwarze  105:        return(&mdoc->meta);
1.1       kristaps  106: }
                    107:
1.123     schwarze  108: void
1.131     schwarze  109: mdoc_endparse(struct roff_man *mdoc)
1.1       kristaps  110: {
                    111:
1.121     schwarze  112:        mdoc_macroend(mdoc);
1.81      schwarze  113: }
                    114:
1.122     schwarze  115: void
1.131     schwarze  116: mdoc_addeqn(struct roff_man *mdoc, const struct eqn *ep)
1.81      schwarze  117: {
1.129     schwarze  118:        struct roff_node *n;
1.81      schwarze  119:
1.128     schwarze  120:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, ROFFT_EQN);
1.81      schwarze  121:        n->eqn = ep;
1.116     schwarze  122:        if (ep->ln > mdoc->last->line)
                    123:                n->flags |= MDOC_LINE;
1.119     schwarze  124:        node_append(mdoc, n);
1.131     schwarze  125:        mdoc->next = ROFF_NEXT_SIBLING;
1.1       kristaps  126: }
                    127:
1.122     schwarze  128: void
1.131     schwarze  129: mdoc_addspan(struct roff_man *mdoc, const struct tbl_span *sp)
1.77      schwarze  130: {
1.129     schwarze  131:        struct roff_node *n;
1.77      schwarze  132:
1.128     schwarze  133:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, ROFFT_TBL);
1.79      schwarze  134:        n->span = sp;
1.119     schwarze  135:        node_append(mdoc, n);
1.131     schwarze  136:        mdoc->next = ROFF_NEXT_SIBLING;
1.77      schwarze  137: }
                    138:
1.1       kristaps  139: /*
                    140:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  141:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  142:  */
                    143: int
1.131     schwarze  144: mdoc_parseln(struct roff_man *mdoc, int ln, char *buf, int offs)
1.1       kristaps  145: {
                    146:
1.128     schwarze  147:        if (mdoc->last->type != ROFFT_EQN || ln > mdoc->last->line)
1.117     schwarze  148:                mdoc->flags |= MDOC_NEWLINE;
1.60      schwarze  149:
                    150:        /*
                    151:         * Let the roff nS register switch SYNOPSIS mode early,
                    152:         * such that the parser knows at all times
                    153:         * whether this mode is on or off.
                    154:         * Note that this mode is also switched by the Sh macro.
                    155:         */
1.94      schwarze  156:        if (roff_getreg(mdoc->roff, "nS"))
                    157:                mdoc->flags |= MDOC_SYNOPSIS;
                    158:        else
                    159:                mdoc->flags &= ~MDOC_SYNOPSIS;
1.60      schwarze  160:
1.93      schwarze  161:        return(roff_getcontrol(mdoc->roff, buf, &offs) ?
1.103     schwarze  162:            mdoc_pmacro(mdoc, ln, buf, offs) :
                    163:            mdoc_ptext(mdoc, ln, buf, offs));
1.1       kristaps  164: }
                    165:
1.121     schwarze  166: void
1.58      schwarze  167: mdoc_macro(MACRO_PROT_ARGS)
1.1       kristaps  168: {
1.46      schwarze  169:        assert(tok < MDOC_MAX);
                    170:
1.113     schwarze  171:        if (mdoc->flags & MDOC_PBODY) {
                    172:                if (tok == MDOC_Dt) {
                    173:                        mandoc_vmsg(MANDOCERR_DT_LATE,
                    174:                            mdoc->parse, line, ppos,
                    175:                            "Dt %s", buf + *pos);
1.121     schwarze  176:                        return;
1.113     schwarze  177:                }
                    178:        } else if ( ! (mdoc_macros[tok].flags & MDOC_PROLOGUE)) {
                    179:                if (mdoc->meta.title == NULL) {
                    180:                        mandoc_vmsg(MANDOCERR_DT_NOTITLE,
                    181:                            mdoc->parse, line, ppos, "%s %s",
                    182:                            mdoc_macronames[tok], buf + *pos);
                    183:                        mdoc->meta.title = mandoc_strdup("UNTITLED");
                    184:                }
1.93      schwarze  185:                if (NULL == mdoc->meta.vol)
                    186:                        mdoc->meta.vol = mandoc_strdup("LOCAL");
                    187:                mdoc->flags |= MDOC_PBODY;
1.39      schwarze  188:        }
1.121     schwarze  189:        (*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf);
1.1       kristaps  190: }
                    191:
                    192:
1.119     schwarze  193: static void
1.131     schwarze  194: node_append(struct roff_man *mdoc, struct roff_node *p)
1.1       kristaps  195: {
                    196:
                    197:        assert(mdoc->last);
                    198:        assert(mdoc->first);
1.128     schwarze  199:        assert(p->type != ROFFT_ROOT);
1.1       kristaps  200:
                    201:        switch (mdoc->next) {
1.131     schwarze  202:        case ROFF_NEXT_SIBLING:
1.1       kristaps  203:                mdoc->last->next = p;
                    204:                p->prev = mdoc->last;
                    205:                p->parent = mdoc->last->parent;
                    206:                break;
1.131     schwarze  207:        case ROFF_NEXT_CHILD:
1.1       kristaps  208:                mdoc->last->child = p;
                    209:                p->parent = mdoc->last;
                    210:                break;
                    211:        default:
                    212:                abort();
                    213:                /* NOTREACHED */
                    214:        }
                    215:
1.10      schwarze  216:        p->parent->nchild++;
                    217:
1.75      schwarze  218:        /*
                    219:         * Copy over the normalised-data pointer of our parent.  Not
                    220:         * everybody has one, but copying a null pointer is fine.
                    221:         */
                    222:
                    223:        switch (p->type) {
1.128     schwarze  224:        case ROFFT_BODY:
1.92      schwarze  225:                if (ENDBODY_NOT != p->end)
                    226:                        break;
1.75      schwarze  227:                /* FALLTHROUGH */
1.128     schwarze  228:        case ROFFT_TAIL:
1.75      schwarze  229:                /* FALLTHROUGH */
1.128     schwarze  230:        case ROFFT_HEAD:
1.75      schwarze  231:                p->norm = p->parent->norm;
                    232:                break;
                    233:        default:
                    234:                break;
                    235:        }
                    236:
1.119     schwarze  237:        mdoc_valid_pre(mdoc, p);
1.1       kristaps  238:
                    239:        switch (p->type) {
1.128     schwarze  240:        case ROFFT_HEAD:
                    241:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  242:                p->parent->head = p;
                    243:                break;
1.128     schwarze  244:        case ROFFT_TAIL:
                    245:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  246:                p->parent->tail = p;
                    247:                break;
1.128     schwarze  248:        case ROFFT_BODY:
1.59      schwarze  249:                if (p->end)
                    250:                        break;
1.128     schwarze  251:                assert(p->parent->type == ROFFT_BLOCK);
1.1       kristaps  252:                p->parent->body = p;
                    253:                break;
                    254:        default:
                    255:                break;
                    256:        }
                    257:
                    258:        mdoc->last = p;
                    259:
                    260:        switch (p->type) {
1.128     schwarze  261:        case ROFFT_TBL:
1.77      schwarze  262:                /* FALLTHROUGH */
1.128     schwarze  263:        case ROFFT_TEXT:
1.119     schwarze  264:                mdoc_valid_post(mdoc);
1.1       kristaps  265:                break;
                    266:        default:
                    267:                break;
                    268:        }
                    269: }
                    270:
1.129     schwarze  271: static struct roff_node *
1.131     schwarze  272: node_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  273:        int tok, enum roff_type type)
1.1       kristaps  274: {
1.129     schwarze  275:        struct roff_node *p;
1.1       kristaps  276:
1.129     schwarze  277:        p = mandoc_calloc(1, sizeof(*p));
1.93      schwarze  278:        p->sec = mdoc->lastsec;
1.1       kristaps  279:        p->line = line;
                    280:        p->pos = pos;
                    281:        p->tok = tok;
1.37      schwarze  282:        p->type = type;
1.58      schwarze  283:
                    284:        /* Flag analysis. */
                    285:
1.93      schwarze  286:        if (MDOC_SYNOPSIS & mdoc->flags)
1.60      schwarze  287:                p->flags |= MDOC_SYNPRETTY;
                    288:        else
                    289:                p->flags &= ~MDOC_SYNPRETTY;
1.93      schwarze  290:        if (MDOC_NEWLINE & mdoc->flags)
1.47      schwarze  291:                p->flags |= MDOC_LINE;
1.93      schwarze  292:        mdoc->flags &= ~MDOC_NEWLINE;
1.58      schwarze  293:
1.1       kristaps  294:        return(p);
                    295: }
                    296:
1.120     schwarze  297: void
1.131     schwarze  298: mdoc_tail_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  299: {
1.129     schwarze  300:        struct roff_node *p;
1.1       kristaps  301:
1.128     schwarze  302:        p = node_alloc(mdoc, line, pos, tok, ROFFT_TAIL);
1.119     schwarze  303:        node_append(mdoc, p);
1.131     schwarze  304:        mdoc->next = ROFF_NEXT_CHILD;
1.1       kristaps  305: }
                    306:
1.129     schwarze  307: struct roff_node *
1.131     schwarze  308: mdoc_head_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  309: {
1.129     schwarze  310:        struct roff_node *p;
1.1       kristaps  311:
1.93      schwarze  312:        assert(mdoc->first);
                    313:        assert(mdoc->last);
1.128     schwarze  314:        p = node_alloc(mdoc, line, pos, tok, ROFFT_HEAD);
1.119     schwarze  315:        node_append(mdoc, p);
1.131     schwarze  316:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  317:        return(p);
1.1       kristaps  318: }
                    319:
1.129     schwarze  320: struct roff_node *
1.131     schwarze  321: mdoc_body_alloc(struct roff_man *mdoc, int line, int pos, int tok)
1.1       kristaps  322: {
1.129     schwarze  323:        struct roff_node *p;
1.1       kristaps  324:
1.128     schwarze  325:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BODY);
1.119     schwarze  326:        node_append(mdoc, p);
1.131     schwarze  327:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  328:        return(p);
1.59      schwarze  329: }
                    330:
1.129     schwarze  331: struct roff_node *
1.131     schwarze  332: mdoc_endbody_alloc(struct roff_man *mdoc, int line, int pos, int tok,
1.129     schwarze  333:                struct roff_node *body, enum mdoc_endbody end)
1.59      schwarze  334: {
1.129     schwarze  335:        struct roff_node *p;
1.59      schwarze  336:
1.126     schwarze  337:        body->flags |= MDOC_ENDED;
                    338:        body->parent->flags |= MDOC_ENDED;
1.128     schwarze  339:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BODY);
1.126     schwarze  340:        p->body = body;
1.92      schwarze  341:        p->norm = body->norm;
1.59      schwarze  342:        p->end = end;
1.119     schwarze  343:        node_append(mdoc, p);
1.131     schwarze  344:        mdoc->next = ROFF_NEXT_SIBLING;
1.124     schwarze  345:        return(p);
1.1       kristaps  346: }
                    347:
1.129     schwarze  348: struct roff_node *
1.131     schwarze  349: mdoc_block_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  350:        int tok, struct mdoc_arg *args)
1.1       kristaps  351: {
1.129     schwarze  352:        struct roff_node *p;
1.1       kristaps  353:
1.128     schwarze  354:        p = node_alloc(mdoc, line, pos, tok, ROFFT_BLOCK);
1.4       schwarze  355:        p->args = args;
                    356:        if (p->args)
1.1       kristaps  357:                (args->refcnt)++;
1.75      schwarze  358:
                    359:        switch (tok) {
1.103     schwarze  360:        case MDOC_Bd:
1.75      schwarze  361:                /* FALLTHROUGH */
1.103     schwarze  362:        case MDOC_Bf:
1.75      schwarze  363:                /* FALLTHROUGH */
1.103     schwarze  364:        case MDOC_Bl:
1.107     schwarze  365:                /* FALLTHROUGH */
                    366:        case MDOC_En:
1.75      schwarze  367:                /* FALLTHROUGH */
1.103     schwarze  368:        case MDOC_Rs:
1.75      schwarze  369:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    370:                break;
                    371:        default:
                    372:                break;
                    373:        }
1.119     schwarze  374:        node_append(mdoc, p);
1.131     schwarze  375:        mdoc->next = ROFF_NEXT_CHILD;
1.120     schwarze  376:        return(p);
1.1       kristaps  377: }
                    378:
1.120     schwarze  379: void
1.131     schwarze  380: mdoc_elem_alloc(struct roff_man *mdoc, int line, int pos,
1.129     schwarze  381:        int tok, struct mdoc_arg *args)
1.1       kristaps  382: {
1.129     schwarze  383:        struct roff_node *p;
1.1       kristaps  384:
1.128     schwarze  385:        p = node_alloc(mdoc, line, pos, tok, ROFFT_ELEM);
1.4       schwarze  386:        p->args = args;
                    387:        if (p->args)
1.1       kristaps  388:                (args->refcnt)++;
1.75      schwarze  389:
                    390:        switch (tok) {
1.103     schwarze  391:        case MDOC_An:
1.75      schwarze  392:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    393:                break;
                    394:        default:
                    395:                break;
                    396:        }
1.119     schwarze  397:        node_append(mdoc, p);
1.131     schwarze  398:        mdoc->next = ROFF_NEXT_CHILD;
1.1       kristaps  399: }
                    400:
1.120     schwarze  401: void
1.131     schwarze  402: mdoc_word_alloc(struct roff_man *mdoc, int line, int pos, const char *p)
1.1       kristaps  403: {
1.129     schwarze  404:        struct roff_node *n;
1.1       kristaps  405:
1.128     schwarze  406:        n = node_alloc(mdoc, line, pos, MDOC_MAX, ROFFT_TEXT);
1.93      schwarze  407:        n->string = roff_strdup(mdoc->roff, p);
1.119     schwarze  408:        node_append(mdoc, n);
1.131     schwarze  409:        mdoc->next = ROFF_NEXT_SIBLING;
1.19      schwarze  410: }
                    411:
1.95      schwarze  412: void
1.131     schwarze  413: mdoc_word_append(struct roff_man *mdoc, const char *p)
1.95      schwarze  414: {
1.129     schwarze  415:        struct roff_node        *n;
1.95      schwarze  416:        char                    *addstr, *newstr;
                    417:
                    418:        n = mdoc->last;
                    419:        addstr = roff_strdup(mdoc->roff, p);
1.100     schwarze  420:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.95      schwarze  421:        free(addstr);
                    422:        free(n->string);
                    423:        n->string = newstr;
1.131     schwarze  424:        mdoc->next = ROFF_NEXT_SIBLING;
1.95      schwarze  425: }
1.19      schwarze  426:
1.61      schwarze  427: static void
1.129     schwarze  428: mdoc_node_free(struct roff_node *p)
1.1       kristaps  429: {
1.61      schwarze  430:
1.128     schwarze  431:        if (p->type == ROFFT_BLOCK || p->type == ROFFT_ELEM)
1.75      schwarze  432:                free(p->norm);
1.1       kristaps  433:        if (p->string)
                    434:                free(p->string);
                    435:        if (p->args)
                    436:                mdoc_argv_free(p->args);
                    437:        free(p);
                    438: }
                    439:
1.40      schwarze  440: static void
1.131     schwarze  441: mdoc_node_unlink(struct roff_man *mdoc, struct roff_node *n)
1.40      schwarze  442: {
                    443:
                    444:        /* Adjust siblings. */
                    445:
                    446:        if (n->prev)
                    447:                n->prev->next = n->next;
                    448:        if (n->next)
                    449:                n->next->prev = n->prev;
                    450:
                    451:        /* Adjust parent. */
                    452:
                    453:        if (n->parent) {
                    454:                n->parent->nchild--;
                    455:                if (n->parent->child == n)
                    456:                        n->parent->child = n->prev ? n->prev : n->next;
1.72      schwarze  457:                if (n->parent->last == n)
                    458:                        n->parent->last = n->prev ? n->prev : NULL;
1.40      schwarze  459:        }
                    460:
                    461:        /* Adjust parse point, if applicable. */
                    462:
1.93      schwarze  463:        if (mdoc && mdoc->last == n) {
1.40      schwarze  464:                if (n->prev) {
1.93      schwarze  465:                        mdoc->last = n->prev;
1.131     schwarze  466:                        mdoc->next = ROFF_NEXT_SIBLING;
1.40      schwarze  467:                } else {
1.93      schwarze  468:                        mdoc->last = n->parent;
1.131     schwarze  469:                        mdoc->next = ROFF_NEXT_CHILD;
1.40      schwarze  470:                }
                    471:        }
                    472:
1.93      schwarze  473:        if (mdoc && mdoc->first == n)
                    474:                mdoc->first = NULL;
1.40      schwarze  475: }
                    476:
1.1       kristaps  477: void
1.131     schwarze  478: mdoc_node_delete(struct roff_man *mdoc, struct roff_node *p)
1.1       kristaps  479: {
                    480:
1.40      schwarze  481:        while (p->child) {
                    482:                assert(p->nchild);
1.93      schwarze  483:                mdoc_node_delete(mdoc, p->child);
1.40      schwarze  484:        }
                    485:        assert(0 == p->nchild);
1.1       kristaps  486:
1.93      schwarze  487:        mdoc_node_unlink(mdoc, p);
1.1       kristaps  488:        mdoc_node_free(p);
1.91      schwarze  489: }
                    490:
1.120     schwarze  491: void
1.131     schwarze  492: mdoc_node_relink(struct roff_man *mdoc, struct roff_node *p)
1.91      schwarze  493: {
                    494:
1.93      schwarze  495:        mdoc_node_unlink(mdoc, p);
1.119     schwarze  496:        node_append(mdoc, p);
1.1       kristaps  497: }
                    498:
                    499: /*
                    500:  * Parse free-form text, that is, a line that does not begin with the
                    501:  * control character.
                    502:  */
                    503: static int
1.131     schwarze  504: mdoc_ptext(struct roff_man *mdoc, int line, char *buf, int offs)
1.1       kristaps  505: {
1.129     schwarze  506:        struct roff_node *n;
1.56      schwarze  507:        char             *c, *ws, *end;
1.46      schwarze  508:
1.93      schwarze  509:        assert(mdoc->last);
                    510:        n = mdoc->last;
1.56      schwarze  511:
                    512:        /*
                    513:         * Divert directly to list processing if we're encountering a
1.128     schwarze  514:         * columnar ROFFT_BLOCK with or without a prior ROFFT_BLOCK entry
                    515:         * (a ROFFT_BODY means it's already open, in which case we should
1.56      schwarze  516:         * process within its context in the normal way).
                    517:         */
                    518:
1.128     schwarze  519:        if (n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
1.127     schwarze  520:            n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) {
1.56      schwarze  521:                /* `Bl' is open without any children. */
1.93      schwarze  522:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  523:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    524:                return(1);
1.56      schwarze  525:        }
                    526:
1.128     schwarze  527:        if (n->tok == MDOC_It && n->type == ROFFT_BLOCK &&
1.103     schwarze  528:            NULL != n->parent &&
                    529:            MDOC_Bl == n->parent->tok &&
                    530:            LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  531:                /* `Bl' has block-level `It' children. */
1.93      schwarze  532:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  533:                mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf);
                    534:                return(1);
1.56      schwarze  535:        }
                    536:
1.52      schwarze  537:        /*
                    538:         * Search for the beginning of unescaped trailing whitespace (ws)
                    539:         * and for the first character not to be output (end).
                    540:         */
1.56      schwarze  541:
                    542:        /* FIXME: replace with strcspn(). */
1.52      schwarze  543:        ws = NULL;
1.53      schwarze  544:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  545:                switch (*c) {
                    546:                case ' ':
                    547:                        if (NULL == ws)
                    548:                                ws = c;
                    549:                        continue;
                    550:                case '\t':
                    551:                        /*
                    552:                         * Always warn about trailing tabs,
                    553:                         * even outside literal context,
                    554:                         * where they should be put on the next line.
                    555:                         */
                    556:                        if (NULL == ws)
                    557:                                ws = c;
                    558:                        /*
                    559:                         * Strip trailing tabs in literal context only;
                    560:                         * outside, they affect the next line.
                    561:                         */
1.93      schwarze  562:                        if (MDOC_LITERAL & mdoc->flags)
1.52      schwarze  563:                                continue;
                    564:                        break;
                    565:                case '\\':
                    566:                        /* Skip the escaped character, too, if any. */
                    567:                        if (c[1])
                    568:                                c++;
                    569:                        /* FALLTHROUGH */
                    570:                default:
                    571:                        ws = NULL;
                    572:                        break;
                    573:                }
                    574:                end = c + 1;
                    575:        }
                    576:        *end = '\0';
1.19      schwarze  577:
1.52      schwarze  578:        if (ws)
1.108     schwarze  579:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    580:                    line, (int)(ws-buf), NULL);
1.34      schwarze  581:
1.120     schwarze  582:        if (buf[offs] == '\0' && ! (mdoc->flags & MDOC_LITERAL)) {
1.108     schwarze  583:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    584:                    line, (int)(c - buf), NULL);
1.46      schwarze  585:
1.40      schwarze  586:                /*
1.66      schwarze  587:                 * Insert a `sp' in the case of a blank line.  Technically,
1.46      schwarze  588:                 * blank lines aren't allowed, but enough manuals assume this
                    589:                 * behaviour that we want to work around it.
1.40      schwarze  590:                 */
1.120     schwarze  591:                mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL);
1.131     schwarze  592:                mdoc->next = ROFF_NEXT_SIBLING;
1.119     schwarze  593:                mdoc_valid_post(mdoc);
                    594:                return(1);
1.38      schwarze  595:        }
1.1       kristaps  596:
1.120     schwarze  597:        mdoc_word_alloc(mdoc, line, offs, buf+offs);
1.1       kristaps  598:
1.120     schwarze  599:        if (mdoc->flags & MDOC_LITERAL)
1.52      schwarze  600:                return(1);
                    601:
1.35      schwarze  602:        /*
1.48      schwarze  603:         * End-of-sentence check.  If the last character is an unescaped
                    604:         * EOS character, then flag the node as being the end of a
                    605:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  606:         */
                    607:
1.52      schwarze  608:        assert(buf < end);
1.48      schwarze  609:
1.97      schwarze  610:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.93      schwarze  611:                mdoc->last->flags |= MDOC_EOS;
1.1       kristaps  612:        return(1);
                    613: }
1.19      schwarze  614:
1.1       kristaps  615: /*
                    616:  * Parse a macro line, that is, a line beginning with the control
                    617:  * character.
                    618:  */
1.61      schwarze  619: static int
1.131     schwarze  620: mdoc_pmacro(struct roff_man *mdoc, int ln, char *buf, int offs)
1.1       kristaps  621: {
1.129     schwarze  622:        struct roff_node *n;
1.118     schwarze  623:        const char       *cp;
1.129     schwarze  624:        int               tok;
1.83      schwarze  625:        int               i, sv;
1.56      schwarze  626:        char              mac[5];
1.1       kristaps  627:
1.83      schwarze  628:        sv = offs;
1.47      schwarze  629:
1.103     schwarze  630:        /*
1.64      schwarze  631:         * Copy the first word into a nil-terminated buffer.
1.118     schwarze  632:         * Stop when a space, tab, escape, or eoln is encountered.
1.63      schwarze  633:         */
1.1       kristaps  634:
1.83      schwarze  635:        i = 0;
1.118     schwarze  636:        while (i < 4 && strchr(" \t\\", buf[offs]) == NULL)
1.83      schwarze  637:                mac[i++] = buf[offs++];
                    638:
                    639:        mac[i] = '\0';
                    640:
1.104     schwarze  641:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  642:
1.118     schwarze  643:        if (tok == MDOC_MAX) {
1.112     schwarze  644:                mandoc_msg(MANDOCERR_MACRO, mdoc->parse,
                    645:                    ln, sv, buf + sv - 1);
1.1       kristaps  646:                return(1);
                    647:        }
                    648:
1.118     schwarze  649:        /* Skip a leading escape sequence or tab. */
1.63      schwarze  650:
1.118     schwarze  651:        switch (buf[offs]) {
                    652:        case '\\':
                    653:                cp = buf + offs + 1;
                    654:                mandoc_escape(&cp, NULL, NULL);
                    655:                offs = cp - buf;
                    656:                break;
                    657:        case '\t':
1.83      schwarze  658:                offs++;
1.118     schwarze  659:                break;
                    660:        default:
                    661:                break;
                    662:        }
1.63      schwarze  663:
                    664:        /* Jump to the next non-whitespace word. */
1.1       kristaps  665:
1.83      schwarze  666:        while (buf[offs] && ' ' == buf[offs])
                    667:                offs++;
1.34      schwarze  668:
1.103     schwarze  669:        /*
1.46      schwarze  670:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    671:         * into the parser as "text", so we only warn about spaces here.
                    672:         */
1.34      schwarze  673:
1.83      schwarze  674:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.108     schwarze  675:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    676:                    ln, offs - 1, NULL);
1.1       kristaps  677:
1.56      schwarze  678:        /*
                    679:         * If an initial macro or a list invocation, divert directly
                    680:         * into macro processing.
                    681:         */
                    682:
1.121     schwarze  683:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok) {
                    684:                mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
                    685:                return(1);
                    686:        }
1.56      schwarze  687:
1.93      schwarze  688:        n = mdoc->last;
                    689:        assert(mdoc->last);
1.56      schwarze  690:
                    691:        /*
                    692:         * If the first macro of a `Bl -column', open an `It' block
                    693:         * context around the parsed macro.
                    694:         */
                    695:
1.128     schwarze  696:        if (n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
1.127     schwarze  697:            n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) {
1.93      schwarze  698:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  699:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    700:                return(1);
1.56      schwarze  701:        }
                    702:
                    703:        /*
                    704:         * If we're following a block-level `It' within a `Bl -column'
                    705:         * context (perhaps opened in the above block or in ptext()),
                    706:         * then open an `It' block context around the parsed macro.
1.24      schwarze  707:         */
1.56      schwarze  708:
1.128     schwarze  709:        if (n->tok == MDOC_It && n->type == ROFFT_BLOCK &&
1.103     schwarze  710:            NULL != n->parent &&
                    711:            MDOC_Bl == n->parent->tok &&
                    712:            LIST_column == n->parent->norm->Bl.type) {
1.93      schwarze  713:                mdoc->flags |= MDOC_FREECOL;
1.121     schwarze  714:                mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf);
                    715:                return(1);
1.56      schwarze  716:        }
                    717:
                    718:        /* Normal processing of a macro. */
                    719:
1.121     schwarze  720:        mdoc_macro(mdoc, tok, ln, sv, &offs, buf);
1.98      schwarze  721:
                    722:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    723:
                    724:        if (mdoc->quick && MDOC_Sh == tok &&
                    725:            SEC_NAME != mdoc->last->sec)
                    726:                return(2);
1.1       kristaps  727:
                    728:        return(1);
                    729: }
1.24      schwarze  730:
1.83      schwarze  731: enum mdelim
                    732: mdoc_isdelim(const char *p)
                    733: {
                    734:
                    735:        if ('\0' == p[0])
                    736:                return(DELIM_NONE);
                    737:
                    738:        if ('\0' == p[1])
                    739:                switch (p[0]) {
1.103     schwarze  740:                case '(':
1.83      schwarze  741:                        /* FALLTHROUGH */
1.103     schwarze  742:                case '[':
1.83      schwarze  743:                        return(DELIM_OPEN);
1.103     schwarze  744:                case '|':
1.83      schwarze  745:                        return(DELIM_MIDDLE);
1.103     schwarze  746:                case '.':
1.83      schwarze  747:                        /* FALLTHROUGH */
1.103     schwarze  748:                case ',':
1.83      schwarze  749:                        /* FALLTHROUGH */
1.103     schwarze  750:                case ';':
1.83      schwarze  751:                        /* FALLTHROUGH */
1.103     schwarze  752:                case ':':
1.83      schwarze  753:                        /* FALLTHROUGH */
1.103     schwarze  754:                case '?':
1.83      schwarze  755:                        /* FALLTHROUGH */
1.103     schwarze  756:                case '!':
1.83      schwarze  757:                        /* FALLTHROUGH */
1.103     schwarze  758:                case ')':
1.83      schwarze  759:                        /* FALLTHROUGH */
1.103     schwarze  760:                case ']':
1.83      schwarze  761:                        return(DELIM_CLOSE);
                    762:                default:
                    763:                        return(DELIM_NONE);
                    764:                }
                    765:
                    766:        if ('\\' != p[0])
                    767:                return(DELIM_NONE);
1.24      schwarze  768:
1.83      schwarze  769:        if (0 == strcmp(p + 1, "."))
                    770:                return(DELIM_CLOSE);
1.90      schwarze  771:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.83      schwarze  772:                return(DELIM_MIDDLE);
                    773:
                    774:        return(DELIM_NONE);
1.101     schwarze  775: }
                    776:
                    777: void
1.129     schwarze  778: mdoc_deroff(char **dest, const struct roff_node *n)
1.101     schwarze  779: {
                    780:        char    *cp;
                    781:        size_t   sz;
                    782:
1.128     schwarze  783:        if (n->type != ROFFT_TEXT) {
1.101     schwarze  784:                for (n = n->child; n; n = n->next)
                    785:                        mdoc_deroff(dest, n);
                    786:                return;
                    787:        }
                    788:
                    789:        /* Skip leading whitespace. */
                    790:
                    791:        for (cp = n->string; '\0' != *cp; cp++)
                    792:                if (0 == isspace((unsigned char)*cp))
                    793:                        break;
                    794:
                    795:        /* Skip trailing whitespace. */
                    796:
                    797:        for (sz = strlen(cp); sz; sz--)
                    798:                if (0 == isspace((unsigned char)cp[sz-1]))
                    799:                        break;
                    800:
                    801:        /* Skip empty strings. */
                    802:
                    803:        if (0 == sz)
                    804:                return;
                    805:
                    806:        if (NULL == *dest) {
                    807:                *dest = mandoc_strndup(cp, sz);
                    808:                return;
                    809:        }
                    810:
                    811:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    812:        free(*dest);
                    813:        *dest = cp;
1.83      schwarze  814: }