[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 1.120

1.120   ! schwarze    1: /*     $OpenBSD: mdoc.c,v 1.119 2014/11/28 01:05:40 schwarze Exp $ */
1.1       kristaps    2: /*
1.77      schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.98      schwarze    4:  * Copyright (c) 2010, 2012, 2013, 2014 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    5:  *
                      6:  * Permission to use, copy, modify, and distribute this software for any
1.3       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:  *
1.3       schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
                     11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
                     12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
                     13:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     14:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     15:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     16:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
1.1       kristaps   17:  */
1.28      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:
1.83      schwarze   28: #include "mdoc.h"
1.54      schwarze   29: #include "mandoc.h"
1.99      schwarze   30: #include "mandoc_aux.h"
1.1       kristaps   31: #include "libmdoc.h"
1.32      schwarze   32: #include "libmandoc.h"
1.1       kristaps   33:
1.106     schwarze   34: const  char *const __mdoc_macronames[MDOC_MAX + 1] = {
1.7       schwarze   35:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   36:        "Sh",           "Ss",           "Pp",           "D1",
                     37:        "Dl",           "Bd",           "Ed",           "Bl",
                     38:        "El",           "It",           "Ad",           "An",
                     39:        "Ar",           "Cd",           "Cm",           "Dv",
                     40:        "Er",           "Ev",           "Ex",           "Fa",
                     41:        "Fd",           "Fl",           "Fn",           "Ft",
                     42:        "Ic",           "In",           "Li",           "Nd",
                     43:        "Nm",           "Op",           "Ot",           "Pa",
                     44:        "Rv",           "St",           "Va",           "Vt",
1.33      schwarze   45:        "Xr",           "%A",           "%B",           "%D",
                     46:        "%I",           "%J",           "%N",           "%O",
                     47:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   48:        "Ac",           "Ao",           "Aq",           "At",
                     49:        "Bc",           "Bf",           "Bo",           "Bq",
                     50:        "Bsx",          "Bx",           "Db",           "Dc",
                     51:        "Do",           "Dq",           "Ec",           "Ef",
                     52:        "Em",           "Eo",           "Fx",           "Ms",
                     53:        "No",           "Ns",           "Nx",           "Ox",
                     54:        "Pc",           "Pf",           "Po",           "Pq",
                     55:        "Qc",           "Ql",           "Qo",           "Qq",
                     56:        "Re",           "Rs",           "Sc",           "So",
                     57:        "Sq",           "Sm",           "Sx",           "Sy",
                     58:        "Tn",           "Ux",           "Xc",           "Xo",
                     59:        "Fo",           "Fc",           "Oo",           "Oc",
                     60:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze   61:        "Fr",           "Ud",           "Lb",           "Lp",
                     62:        "Lk",           "Mt",           "Brq",          "Bro",
1.33      schwarze   63:        "Brc",          "%C",           "Es",           "En",
                     64:        "Dx",           "%Q",           "br",           "sp",
1.106     schwarze   65:        "%U",           "Ta",           "ll",           "text",
1.1       kristaps   66:        };
                     67:
1.103     schwarze   68: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
1.1       kristaps   69:        "split",                "nosplit",              "ragged",
1.103     schwarze   70:        "unfilled",             "literal",              "file",
                     71:        "offset",               "bullet",               "dash",
                     72:        "hyphen",               "item",                 "enum",
                     73:        "tag",                  "diag",                 "hang",
                     74:        "ohang",                "inset",                "column",
                     75:        "width",                "compact",              "std",
1.1       kristaps   76:        "filled",               "words",                "emphasis",
1.30      schwarze   77:        "symbolic",             "nested",               "centered"
1.1       kristaps   78:        };
                     79:
                     80: const  char * const *mdoc_macronames = __mdoc_macronames;
                     81: const  char * const *mdoc_argnames = __mdoc_argnames;
                     82:
1.40      schwarze   83: static void              mdoc_node_free(struct mdoc_node *);
1.103     schwarze   84: static void              mdoc_node_unlink(struct mdoc *,
1.40      schwarze   85:                                struct mdoc_node *);
1.1       kristaps   86: static void              mdoc_free1(struct mdoc *);
1.32      schwarze   87: static void              mdoc_alloc1(struct mdoc *);
1.103     schwarze   88: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.37      schwarze   89:                                enum mdoct, enum mdoc_type);
1.119     schwarze   90: static void              node_append(struct mdoc *, struct mdoc_node *);
1.53      schwarze   91: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     92: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.1       kristaps   93:
1.103     schwarze   94:
1.1       kristaps   95: const struct mdoc_node *
1.93      schwarze   96: mdoc_node(const struct mdoc *mdoc)
1.1       kristaps   97: {
                     98:
1.93      schwarze   99:        return(mdoc->first);
1.1       kristaps  100: }
                    101:
                    102: const struct mdoc_meta *
1.93      schwarze  103: mdoc_meta(const struct mdoc *mdoc)
1.1       kristaps  104: {
                    105:
1.93      schwarze  106:        return(&mdoc->meta);
1.1       kristaps  107: }
                    108:
1.8       schwarze  109: /*
                    110:  * Frees volatile resources (parse tree, meta-data, fields).
                    111:  */
1.1       kristaps  112: static void
                    113: mdoc_free1(struct mdoc *mdoc)
                    114: {
                    115:
                    116:        if (mdoc->first)
1.40      schwarze  117:                mdoc_node_delete(mdoc, mdoc->first);
1.113     schwarze  118:        free(mdoc->meta.msec);
                    119:        free(mdoc->meta.vol);
                    120:        free(mdoc->meta.arch);
                    121:        free(mdoc->meta.date);
                    122:        free(mdoc->meta.title);
                    123:        free(mdoc->meta.os);
                    124:        free(mdoc->meta.name);
1.1       kristaps  125: }
                    126:
1.8       schwarze  127: /*
                    128:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    129:  */
1.32      schwarze  130: static void
1.1       kristaps  131: mdoc_alloc1(struct mdoc *mdoc)
                    132: {
                    133:
1.32      schwarze  134:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.1       kristaps  135:        mdoc->flags = 0;
1.9       schwarze  136:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.32      schwarze  137:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.1       kristaps  138:        mdoc->first = mdoc->last;
                    139:        mdoc->last->type = MDOC_ROOT;
1.86      schwarze  140:        mdoc->last->tok = MDOC_MAX;
1.1       kristaps  141:        mdoc->next = MDOC_NEXT_CHILD;
                    142: }
                    143:
                    144: /*
1.8       schwarze  145:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    146:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    147:  * and the parser is ready for re-invocation on a new tree; however,
                    148:  * cross-parse non-volatile data is kept intact.
1.1       kristaps  149:  */
1.32      schwarze  150: void
1.1       kristaps  151: mdoc_reset(struct mdoc *mdoc)
                    152: {
                    153:
                    154:        mdoc_free1(mdoc);
1.32      schwarze  155:        mdoc_alloc1(mdoc);
1.1       kristaps  156: }
                    157:
                    158: /*
1.8       schwarze  159:  * Completely free up all volatile and non-volatile parse resources.
                    160:  * After invocation, the pointer is no longer usable.
1.1       kristaps  161:  */
                    162: void
                    163: mdoc_free(struct mdoc *mdoc)
                    164: {
                    165:
                    166:        mdoc_free1(mdoc);
                    167:        free(mdoc);
                    168: }
                    169:
1.8       schwarze  170: /*
1.103     schwarze  171:  * Allocate volatile and non-volatile parse resources.
1.8       schwarze  172:  */
1.1       kristaps  173: struct mdoc *
1.98      schwarze  174: mdoc_alloc(struct roff *roff, struct mparse *parse,
1.110     schwarze  175:        const char *defos, int quick)
1.1       kristaps  176: {
                    177:        struct mdoc     *p;
                    178:
1.32      schwarze  179:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    180:
1.83      schwarze  181:        p->parse = parse;
1.87      schwarze  182:        p->defos = defos;
1.98      schwarze  183:        p->quick = quick;
1.84      schwarze  184:        p->roff = roff;
1.1       kristaps  185:
1.32      schwarze  186:        mdoc_hash_init();
                    187:        mdoc_alloc1(p);
                    188:        return(p);
1.1       kristaps  189: }
                    190:
                    191: int
1.93      schwarze  192: mdoc_endparse(struct mdoc *mdoc)
1.1       kristaps  193: {
                    194:
1.111     schwarze  195:        return(mdoc_macroend(mdoc));
1.81      schwarze  196: }
                    197:
                    198: int
1.93      schwarze  199: mdoc_addeqn(struct mdoc *mdoc, const struct eqn *ep)
1.81      schwarze  200: {
                    201:        struct mdoc_node *n;
                    202:
1.93      schwarze  203:        n = node_alloc(mdoc, ep->ln, ep->pos, MDOC_MAX, MDOC_EQN);
1.81      schwarze  204:        n->eqn = ep;
1.116     schwarze  205:        if (ep->ln > mdoc->last->line)
                    206:                n->flags |= MDOC_LINE;
1.119     schwarze  207:        node_append(mdoc, n);
1.93      schwarze  208:        mdoc->next = MDOC_NEXT_SIBLING;
1.81      schwarze  209:        return(1);
1.1       kristaps  210: }
                    211:
1.77      schwarze  212: int
1.93      schwarze  213: mdoc_addspan(struct mdoc *mdoc, const struct tbl_span *sp)
1.77      schwarze  214: {
1.79      schwarze  215:        struct mdoc_node *n;
1.77      schwarze  216:
1.93      schwarze  217:        n = node_alloc(mdoc, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.79      schwarze  218:        n->span = sp;
1.119     schwarze  219:        node_append(mdoc, n);
1.93      schwarze  220:        mdoc->next = MDOC_NEXT_SIBLING;
1.79      schwarze  221:        return(1);
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
1.93      schwarze  229: mdoc_parseln(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  230: {
                    231:
1.117     schwarze  232:        if (mdoc->last->type != MDOC_EQN || ln > mdoc->last->line)
                    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:
1.5       schwarze  251: int
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);
                    261:                        return(1);
                    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:        }
1.1       kristaps  274:
1.93      schwarze  275:        return((*mdoc_macros[tok].fp)(mdoc, tok, line, ppos, pos, buf));
1.1       kristaps  276: }
                    277:
                    278:
1.119     schwarze  279: static void
1.1       kristaps  280: node_append(struct mdoc *mdoc, struct mdoc_node *p)
                    281: {
                    282:
                    283:        assert(mdoc->last);
                    284:        assert(mdoc->first);
                    285:        assert(MDOC_ROOT != p->type);
                    286:
                    287:        switch (mdoc->next) {
1.103     schwarze  288:        case MDOC_NEXT_SIBLING:
1.1       kristaps  289:                mdoc->last->next = p;
                    290:                p->prev = mdoc->last;
                    291:                p->parent = mdoc->last->parent;
                    292:                break;
1.103     schwarze  293:        case MDOC_NEXT_CHILD:
1.1       kristaps  294:                mdoc->last->child = p;
                    295:                p->parent = mdoc->last;
                    296:                break;
                    297:        default:
                    298:                abort();
                    299:                /* NOTREACHED */
                    300:        }
                    301:
1.10      schwarze  302:        p->parent->nchild++;
                    303:
1.75      schwarze  304:        /*
                    305:         * Copy over the normalised-data pointer of our parent.  Not
                    306:         * everybody has one, but copying a null pointer is fine.
                    307:         */
                    308:
                    309:        switch (p->type) {
1.103     schwarze  310:        case MDOC_BODY:
1.92      schwarze  311:                if (ENDBODY_NOT != p->end)
                    312:                        break;
1.75      schwarze  313:                /* FALLTHROUGH */
1.103     schwarze  314:        case MDOC_TAIL:
1.75      schwarze  315:                /* FALLTHROUGH */
1.103     schwarze  316:        case MDOC_HEAD:
1.75      schwarze  317:                p->norm = p->parent->norm;
                    318:                break;
                    319:        default:
                    320:                break;
                    321:        }
                    322:
1.119     schwarze  323:        mdoc_valid_pre(mdoc, p);
1.1       kristaps  324:
                    325:        switch (p->type) {
1.103     schwarze  326:        case MDOC_HEAD:
1.1       kristaps  327:                assert(MDOC_BLOCK == p->parent->type);
                    328:                p->parent->head = p;
                    329:                break;
1.103     schwarze  330:        case MDOC_TAIL:
1.1       kristaps  331:                assert(MDOC_BLOCK == p->parent->type);
                    332:                p->parent->tail = p;
                    333:                break;
1.103     schwarze  334:        case MDOC_BODY:
1.59      schwarze  335:                if (p->end)
                    336:                        break;
1.1       kristaps  337:                assert(MDOC_BLOCK == p->parent->type);
                    338:                p->parent->body = p;
                    339:                break;
                    340:        default:
                    341:                break;
                    342:        }
                    343:
                    344:        mdoc->last = p;
                    345:
                    346:        switch (p->type) {
1.103     schwarze  347:        case MDOC_TBL:
1.77      schwarze  348:                /* FALLTHROUGH */
1.103     schwarze  349:        case MDOC_TEXT:
1.119     schwarze  350:                mdoc_valid_post(mdoc);
1.1       kristaps  351:                break;
                    352:        default:
                    353:                break;
                    354:        }
                    355: }
                    356:
                    357: static struct mdoc_node *
1.103     schwarze  358: node_alloc(struct mdoc *mdoc, int line, int pos,
1.37      schwarze  359:                enum mdoct tok, enum mdoc_type type)
1.1       kristaps  360: {
                    361:        struct mdoc_node *p;
                    362:
1.32      schwarze  363:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.93      schwarze  364:        p->sec = mdoc->lastsec;
1.1       kristaps  365:        p->line = line;
                    366:        p->pos = pos;
1.96      schwarze  367:        p->lastline = line;
1.1       kristaps  368:        p->tok = tok;
1.37      schwarze  369:        p->type = type;
1.58      schwarze  370:
                    371:        /* Flag analysis. */
                    372:
1.93      schwarze  373:        if (MDOC_SYNOPSIS & mdoc->flags)
1.60      schwarze  374:                p->flags |= MDOC_SYNPRETTY;
                    375:        else
                    376:                p->flags &= ~MDOC_SYNPRETTY;
1.93      schwarze  377:        if (MDOC_NEWLINE & mdoc->flags)
1.47      schwarze  378:                p->flags |= MDOC_LINE;
1.93      schwarze  379:        mdoc->flags &= ~MDOC_NEWLINE;
1.58      schwarze  380:
1.1       kristaps  381:        return(p);
                    382: }
                    383:
1.120   ! schwarze  384: void
1.93      schwarze  385: mdoc_tail_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  386: {
                    387:        struct mdoc_node *p;
                    388:
1.93      schwarze  389:        p = node_alloc(mdoc, line, pos, tok, MDOC_TAIL);
1.119     schwarze  390:        node_append(mdoc, p);
1.93      schwarze  391:        mdoc->next = MDOC_NEXT_CHILD;
1.1       kristaps  392: }
                    393:
1.120   ! schwarze  394: struct mdoc_node *
1.93      schwarze  395: mdoc_head_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  396: {
                    397:        struct mdoc_node *p;
                    398:
1.93      schwarze  399:        assert(mdoc->first);
                    400:        assert(mdoc->last);
                    401:        p = node_alloc(mdoc, line, pos, tok, MDOC_HEAD);
1.119     schwarze  402:        node_append(mdoc, p);
1.93      schwarze  403:        mdoc->next = MDOC_NEXT_CHILD;
1.120   ! schwarze  404:        return(p);
1.1       kristaps  405: }
                    406:
1.120   ! schwarze  407: struct mdoc_node *
1.93      schwarze  408: mdoc_body_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok)
1.1       kristaps  409: {
                    410:        struct mdoc_node *p;
                    411:
1.93      schwarze  412:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.119     schwarze  413:        node_append(mdoc, p);
1.93      schwarze  414:        mdoc->next = MDOC_NEXT_CHILD;
1.120   ! schwarze  415:        return(p);
1.59      schwarze  416: }
                    417:
1.120   ! schwarze  418: void
1.93      schwarze  419: mdoc_endbody_alloc(struct mdoc *mdoc, int line, int pos, enum mdoct tok,
1.59      schwarze  420:                struct mdoc_node *body, enum mdoc_endbody end)
                    421: {
                    422:        struct mdoc_node *p;
                    423:
1.93      schwarze  424:        p = node_alloc(mdoc, line, pos, tok, MDOC_BODY);
1.59      schwarze  425:        p->pending = body;
1.92      schwarze  426:        p->norm = body->norm;
1.59      schwarze  427:        p->end = end;
1.119     schwarze  428:        node_append(mdoc, p);
1.93      schwarze  429:        mdoc->next = MDOC_NEXT_SIBLING;
1.1       kristaps  430: }
                    431:
1.120   ! schwarze  432: struct mdoc_node *
1.103     schwarze  433: mdoc_block_alloc(struct mdoc *mdoc, int line, int pos,
1.37      schwarze  434:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  435: {
                    436:        struct mdoc_node *p;
                    437:
1.93      schwarze  438:        p = node_alloc(mdoc, line, pos, tok, MDOC_BLOCK);
1.4       schwarze  439:        p->args = args;
                    440:        if (p->args)
1.1       kristaps  441:                (args->refcnt)++;
1.75      schwarze  442:
                    443:        switch (tok) {
1.103     schwarze  444:        case MDOC_Bd:
1.75      schwarze  445:                /* FALLTHROUGH */
1.103     schwarze  446:        case MDOC_Bf:
1.75      schwarze  447:                /* FALLTHROUGH */
1.103     schwarze  448:        case MDOC_Bl:
1.107     schwarze  449:                /* FALLTHROUGH */
                    450:        case MDOC_En:
1.75      schwarze  451:                /* FALLTHROUGH */
1.103     schwarze  452:        case MDOC_Rs:
1.75      schwarze  453:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    454:                break;
                    455:        default:
                    456:                break;
                    457:        }
1.119     schwarze  458:        node_append(mdoc, p);
1.93      schwarze  459:        mdoc->next = MDOC_NEXT_CHILD;
1.120   ! schwarze  460:        return(p);
1.1       kristaps  461: }
                    462:
1.120   ! schwarze  463: void
1.103     schwarze  464: mdoc_elem_alloc(struct mdoc *mdoc, int line, int pos,
1.37      schwarze  465:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  466: {
                    467:        struct mdoc_node *p;
                    468:
1.93      schwarze  469:        p = node_alloc(mdoc, line, pos, tok, MDOC_ELEM);
1.4       schwarze  470:        p->args = args;
                    471:        if (p->args)
1.1       kristaps  472:                (args->refcnt)++;
1.75      schwarze  473:
                    474:        switch (tok) {
1.103     schwarze  475:        case MDOC_An:
1.75      schwarze  476:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    477:                break;
                    478:        default:
                    479:                break;
                    480:        }
1.119     schwarze  481:        node_append(mdoc, p);
1.93      schwarze  482:        mdoc->next = MDOC_NEXT_CHILD;
1.1       kristaps  483: }
                    484:
1.120   ! schwarze  485: void
1.93      schwarze  486: mdoc_word_alloc(struct mdoc *mdoc, int line, int pos, const char *p)
1.1       kristaps  487: {
1.19      schwarze  488:        struct mdoc_node *n;
1.1       kristaps  489:
1.93      schwarze  490:        n = node_alloc(mdoc, line, pos, MDOC_MAX, MDOC_TEXT);
                    491:        n->string = roff_strdup(mdoc->roff, p);
1.119     schwarze  492:        node_append(mdoc, n);
1.93      schwarze  493:        mdoc->next = MDOC_NEXT_SIBLING;
1.19      schwarze  494: }
                    495:
1.95      schwarze  496: void
                    497: mdoc_word_append(struct mdoc *mdoc, const char *p)
                    498: {
                    499:        struct mdoc_node        *n;
                    500:        char                    *addstr, *newstr;
                    501:
                    502:        n = mdoc->last;
                    503:        addstr = roff_strdup(mdoc->roff, p);
1.100     schwarze  504:        mandoc_asprintf(&newstr, "%s %s", n->string, addstr);
1.95      schwarze  505:        free(addstr);
                    506:        free(n->string);
                    507:        n->string = newstr;
                    508:        mdoc->next = MDOC_NEXT_SIBLING;
                    509: }
1.19      schwarze  510:
1.61      schwarze  511: static void
1.1       kristaps  512: mdoc_node_free(struct mdoc_node *p)
                    513: {
1.61      schwarze  514:
1.75      schwarze  515:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
                    516:                free(p->norm);
1.1       kristaps  517:        if (p->string)
                    518:                free(p->string);
                    519:        if (p->args)
                    520:                mdoc_argv_free(p->args);
                    521:        free(p);
                    522: }
                    523:
1.40      schwarze  524: static void
1.93      schwarze  525: mdoc_node_unlink(struct mdoc *mdoc, struct mdoc_node *n)
1.40      schwarze  526: {
                    527:
                    528:        /* Adjust siblings. */
                    529:
                    530:        if (n->prev)
                    531:                n->prev->next = n->next;
                    532:        if (n->next)
                    533:                n->next->prev = n->prev;
                    534:
                    535:        /* Adjust parent. */
                    536:
                    537:        if (n->parent) {
                    538:                n->parent->nchild--;
                    539:                if (n->parent->child == n)
                    540:                        n->parent->child = n->prev ? n->prev : n->next;
1.72      schwarze  541:                if (n->parent->last == n)
                    542:                        n->parent->last = n->prev ? n->prev : NULL;
1.40      schwarze  543:        }
                    544:
                    545:        /* Adjust parse point, if applicable. */
                    546:
1.93      schwarze  547:        if (mdoc && mdoc->last == n) {
1.40      schwarze  548:                if (n->prev) {
1.93      schwarze  549:                        mdoc->last = n->prev;
                    550:                        mdoc->next = MDOC_NEXT_SIBLING;
1.40      schwarze  551:                } else {
1.93      schwarze  552:                        mdoc->last = n->parent;
                    553:                        mdoc->next = MDOC_NEXT_CHILD;
1.40      schwarze  554:                }
                    555:        }
                    556:
1.93      schwarze  557:        if (mdoc && mdoc->first == n)
                    558:                mdoc->first = NULL;
1.40      schwarze  559: }
                    560:
1.1       kristaps  561: void
1.93      schwarze  562: mdoc_node_delete(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  563: {
                    564:
1.40      schwarze  565:        while (p->child) {
                    566:                assert(p->nchild);
1.93      schwarze  567:                mdoc_node_delete(mdoc, p->child);
1.40      schwarze  568:        }
                    569:        assert(0 == p->nchild);
1.1       kristaps  570:
1.93      schwarze  571:        mdoc_node_unlink(mdoc, p);
1.1       kristaps  572:        mdoc_node_free(p);
1.91      schwarze  573: }
                    574:
1.120   ! schwarze  575: void
1.93      schwarze  576: mdoc_node_relink(struct mdoc *mdoc, struct mdoc_node *p)
1.91      schwarze  577: {
                    578:
1.93      schwarze  579:        mdoc_node_unlink(mdoc, p);
1.119     schwarze  580:        node_append(mdoc, p);
1.1       kristaps  581: }
                    582:
                    583: /*
                    584:  * Parse free-form text, that is, a line that does not begin with the
                    585:  * control character.
                    586:  */
                    587: static int
1.93      schwarze  588: mdoc_ptext(struct mdoc *mdoc, int line, char *buf, int offs)
1.1       kristaps  589: {
1.56      schwarze  590:        char             *c, *ws, *end;
                    591:        struct mdoc_node *n;
1.46      schwarze  592:
1.93      schwarze  593:        assert(mdoc->last);
                    594:        n = mdoc->last;
1.56      schwarze  595:
                    596:        /*
                    597:         * Divert directly to list processing if we're encountering a
                    598:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
                    599:         * (a MDOC_BODY means it's already open, in which case we should
                    600:         * process within its context in the normal way).
                    601:         */
                    602:
                    603:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.103     schwarze  604:            LIST_column == n->norm->Bl.type) {
1.56      schwarze  605:                /* `Bl' is open without any children. */
1.93      schwarze  606:                mdoc->flags |= MDOC_FREECOL;
                    607:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.56      schwarze  608:        }
                    609:
                    610:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.103     schwarze  611:            NULL != n->parent &&
                    612:            MDOC_Bl == n->parent->tok &&
                    613:            LIST_column == n->parent->norm->Bl.type) {
1.56      schwarze  614:                /* `Bl' has block-level `It' children. */
1.93      schwarze  615:                mdoc->flags |= MDOC_FREECOL;
                    616:                return(mdoc_macro(mdoc, MDOC_It, line, offs, &offs, buf));
1.56      schwarze  617:        }
                    618:
1.52      schwarze  619:        /*
                    620:         * Search for the beginning of unescaped trailing whitespace (ws)
                    621:         * and for the first character not to be output (end).
                    622:         */
1.56      schwarze  623:
                    624:        /* FIXME: replace with strcspn(). */
1.52      schwarze  625:        ws = NULL;
1.53      schwarze  626:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  627:                switch (*c) {
                    628:                case ' ':
                    629:                        if (NULL == ws)
                    630:                                ws = c;
                    631:                        continue;
                    632:                case '\t':
                    633:                        /*
                    634:                         * Always warn about trailing tabs,
                    635:                         * even outside literal context,
                    636:                         * where they should be put on the next line.
                    637:                         */
                    638:                        if (NULL == ws)
                    639:                                ws = c;
                    640:                        /*
                    641:                         * Strip trailing tabs in literal context only;
                    642:                         * outside, they affect the next line.
                    643:                         */
1.93      schwarze  644:                        if (MDOC_LITERAL & mdoc->flags)
1.52      schwarze  645:                                continue;
                    646:                        break;
                    647:                case '\\':
                    648:                        /* Skip the escaped character, too, if any. */
                    649:                        if (c[1])
                    650:                                c++;
                    651:                        /* FALLTHROUGH */
                    652:                default:
                    653:                        ws = NULL;
                    654:                        break;
                    655:                }
                    656:                end = c + 1;
                    657:        }
                    658:        *end = '\0';
1.19      schwarze  659:
1.52      schwarze  660:        if (ws)
1.108     schwarze  661:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    662:                    line, (int)(ws-buf), NULL);
1.34      schwarze  663:
1.120   ! schwarze  664:        if (buf[offs] == '\0' && ! (mdoc->flags & MDOC_LITERAL)) {
1.108     schwarze  665:                mandoc_msg(MANDOCERR_FI_BLANK, mdoc->parse,
                    666:                    line, (int)(c - buf), NULL);
1.46      schwarze  667:
1.40      schwarze  668:                /*
1.66      schwarze  669:                 * Insert a `sp' in the case of a blank line.  Technically,
1.46      schwarze  670:                 * blank lines aren't allowed, but enough manuals assume this
                    671:                 * behaviour that we want to work around it.
1.40      schwarze  672:                 */
1.120   ! schwarze  673:                mdoc_elem_alloc(mdoc, line, offs, MDOC_sp, NULL);
1.93      schwarze  674:                mdoc->next = MDOC_NEXT_SIBLING;
1.119     schwarze  675:                mdoc_valid_post(mdoc);
                    676:                return(1);
1.38      schwarze  677:        }
1.1       kristaps  678:
1.120   ! schwarze  679:        mdoc_word_alloc(mdoc, line, offs, buf+offs);
1.1       kristaps  680:
1.120   ! schwarze  681:        if (mdoc->flags & MDOC_LITERAL)
1.52      schwarze  682:                return(1);
                    683:
1.35      schwarze  684:        /*
1.48      schwarze  685:         * End-of-sentence check.  If the last character is an unescaped
                    686:         * EOS character, then flag the node as being the end of a
                    687:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  688:         */
                    689:
1.52      schwarze  690:        assert(buf < end);
1.48      schwarze  691:
1.97      schwarze  692:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
1.93      schwarze  693:                mdoc->last->flags |= MDOC_EOS;
1.48      schwarze  694:
1.1       kristaps  695:        return(1);
                    696: }
1.19      schwarze  697:
1.1       kristaps  698: /*
                    699:  * Parse a macro line, that is, a line beginning with the control
                    700:  * character.
                    701:  */
1.61      schwarze  702: static int
1.93      schwarze  703: mdoc_pmacro(struct mdoc *mdoc, int ln, char *buf, int offs)
1.1       kristaps  704: {
1.118     schwarze  705:        struct mdoc_node *n;
                    706:        const char       *cp;
1.56      schwarze  707:        enum mdoct        tok;
1.83      schwarze  708:        int               i, sv;
1.56      schwarze  709:        char              mac[5];
1.1       kristaps  710:
1.83      schwarze  711:        sv = offs;
1.47      schwarze  712:
1.103     schwarze  713:        /*
1.64      schwarze  714:         * Copy the first word into a nil-terminated buffer.
1.118     schwarze  715:         * Stop when a space, tab, escape, or eoln is encountered.
1.63      schwarze  716:         */
1.1       kristaps  717:
1.83      schwarze  718:        i = 0;
1.118     schwarze  719:        while (i < 4 && strchr(" \t\\", buf[offs]) == NULL)
1.83      schwarze  720:                mac[i++] = buf[offs++];
                    721:
                    722:        mac[i] = '\0';
                    723:
1.104     schwarze  724:        tok = (i > 1 && i < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
1.1       kristaps  725:
1.118     schwarze  726:        if (tok == MDOC_MAX) {
1.112     schwarze  727:                mandoc_msg(MANDOCERR_MACRO, mdoc->parse,
                    728:                    ln, sv, buf + sv - 1);
1.1       kristaps  729:                return(1);
                    730:        }
                    731:
1.118     schwarze  732:        /* Skip a leading escape sequence or tab. */
1.63      schwarze  733:
1.118     schwarze  734:        switch (buf[offs]) {
                    735:        case '\\':
                    736:                cp = buf + offs + 1;
                    737:                mandoc_escape(&cp, NULL, NULL);
                    738:                offs = cp - buf;
                    739:                break;
                    740:        case '\t':
1.83      schwarze  741:                offs++;
1.118     schwarze  742:                break;
                    743:        default:
                    744:                break;
                    745:        }
1.63      schwarze  746:
                    747:        /* Jump to the next non-whitespace word. */
1.1       kristaps  748:
1.83      schwarze  749:        while (buf[offs] && ' ' == buf[offs])
                    750:                offs++;
1.34      schwarze  751:
1.103     schwarze  752:        /*
1.46      schwarze  753:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    754:         * into the parser as "text", so we only warn about spaces here.
                    755:         */
1.34      schwarze  756:
1.83      schwarze  757:        if ('\0' == buf[offs] && ' ' == buf[offs - 1])
1.108     schwarze  758:                mandoc_msg(MANDOCERR_SPACE_EOL, mdoc->parse,
                    759:                    ln, offs - 1, NULL);
1.1       kristaps  760:
1.56      schwarze  761:        /*
                    762:         * If an initial macro or a list invocation, divert directly
                    763:         * into macro processing.
                    764:         */
                    765:
1.111     schwarze  766:        if (NULL == mdoc->last || MDOC_It == tok || MDOC_El == tok)
                    767:                return(mdoc_macro(mdoc, tok, ln, sv, &offs, buf));
1.56      schwarze  768:
1.93      schwarze  769:        n = mdoc->last;
                    770:        assert(mdoc->last);
1.56      schwarze  771:
                    772:        /*
                    773:         * If the first macro of a `Bl -column', open an `It' block
                    774:         * context around the parsed macro.
                    775:         */
                    776:
                    777:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.103     schwarze  778:            LIST_column == n->norm->Bl.type) {
1.93      schwarze  779:                mdoc->flags |= MDOC_FREECOL;
1.111     schwarze  780:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.56      schwarze  781:        }
                    782:
                    783:        /*
                    784:         * If we're following a block-level `It' within a `Bl -column'
                    785:         * context (perhaps opened in the above block or in ptext()),
                    786:         * then open an `It' block context around the parsed macro.
1.24      schwarze  787:         */
1.56      schwarze  788:
                    789:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
1.103     schwarze  790:            NULL != n->parent &&
                    791:            MDOC_Bl == n->parent->tok &&
                    792:            LIST_column == n->parent->norm->Bl.type) {
1.93      schwarze  793:                mdoc->flags |= MDOC_FREECOL;
1.111     schwarze  794:                return(mdoc_macro(mdoc, MDOC_It, ln, sv, &sv, buf));
1.56      schwarze  795:        }
                    796:
                    797:        /* Normal processing of a macro. */
                    798:
1.103     schwarze  799:        if ( ! mdoc_macro(mdoc, tok, ln, sv, &offs, buf))
1.111     schwarze  800:                return(0);
1.98      schwarze  801:
                    802:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    803:
                    804:        if (mdoc->quick && MDOC_Sh == tok &&
                    805:            SEC_NAME != mdoc->last->sec)
                    806:                return(2);
1.1       kristaps  807:
                    808:        return(1);
                    809: }
1.24      schwarze  810:
1.83      schwarze  811: enum mdelim
                    812: mdoc_isdelim(const char *p)
                    813: {
                    814:
                    815:        if ('\0' == p[0])
                    816:                return(DELIM_NONE);
                    817:
                    818:        if ('\0' == p[1])
                    819:                switch (p[0]) {
1.103     schwarze  820:                case '(':
1.83      schwarze  821:                        /* FALLTHROUGH */
1.103     schwarze  822:                case '[':
1.83      schwarze  823:                        return(DELIM_OPEN);
1.103     schwarze  824:                case '|':
1.83      schwarze  825:                        return(DELIM_MIDDLE);
1.103     schwarze  826:                case '.':
1.83      schwarze  827:                        /* FALLTHROUGH */
1.103     schwarze  828:                case ',':
1.83      schwarze  829:                        /* FALLTHROUGH */
1.103     schwarze  830:                case ';':
1.83      schwarze  831:                        /* FALLTHROUGH */
1.103     schwarze  832:                case ':':
1.83      schwarze  833:                        /* FALLTHROUGH */
1.103     schwarze  834:                case '?':
1.83      schwarze  835:                        /* FALLTHROUGH */
1.103     schwarze  836:                case '!':
1.83      schwarze  837:                        /* FALLTHROUGH */
1.103     schwarze  838:                case ')':
1.83      schwarze  839:                        /* FALLTHROUGH */
1.103     schwarze  840:                case ']':
1.83      schwarze  841:                        return(DELIM_CLOSE);
                    842:                default:
                    843:                        return(DELIM_NONE);
                    844:                }
                    845:
                    846:        if ('\\' != p[0])
                    847:                return(DELIM_NONE);
1.24      schwarze  848:
1.83      schwarze  849:        if (0 == strcmp(p + 1, "."))
                    850:                return(DELIM_CLOSE);
1.90      schwarze  851:        if (0 == strcmp(p + 1, "fR|\\fP"))
1.83      schwarze  852:                return(DELIM_MIDDLE);
                    853:
                    854:        return(DELIM_NONE);
1.101     schwarze  855: }
                    856:
                    857: void
                    858: mdoc_deroff(char **dest, const struct mdoc_node *n)
                    859: {
                    860:        char    *cp;
                    861:        size_t   sz;
                    862:
                    863:        if (MDOC_TEXT != n->type) {
                    864:                for (n = n->child; n; n = n->next)
                    865:                        mdoc_deroff(dest, n);
                    866:                return;
                    867:        }
                    868:
                    869:        /* Skip leading whitespace. */
                    870:
                    871:        for (cp = n->string; '\0' != *cp; cp++)
                    872:                if (0 == isspace((unsigned char)*cp))
                    873:                        break;
                    874:
                    875:        /* Skip trailing whitespace. */
                    876:
                    877:        for (sz = strlen(cp); sz; sz--)
                    878:                if (0 == isspace((unsigned char)cp[sz-1]))
                    879:                        break;
                    880:
                    881:        /* Skip empty strings. */
                    882:
                    883:        if (0 == sz)
                    884:                return;
                    885:
                    886:        if (NULL == *dest) {
                    887:                *dest = mandoc_strndup(cp, sz);
                    888:                return;
                    889:        }
                    890:
                    891:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    892:        free(*dest);
                    893:        *dest = cp;
1.83      schwarze  894: }