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

Annotation of mandoc/mdoc.c, Revision 1.172

1.172   ! kristaps    1: /*     $Id: mdoc.c,v 1.171 2010/12/22 11:15:16 kristaps Exp $ */
1.1       kristaps    2: /*
1.157     schwarze    3:  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
                      4:  * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
1.1       kristaps    5:  *
                      6:  * Permission to use, copy, modify, and distribute this software for any
1.75      kristaps    7:  * purpose with or without fee is hereby granted, provided that the above
                      8:  * copyright notice and this permission notice appear in all copies.
1.1       kristaps    9:  *
1.75      kristaps   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
                     11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
                     12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
                     13:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     14:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     15:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     16:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
1.1       kristaps   17:  */
1.114     kristaps   18: #ifdef HAVE_CONFIG_H
                     19: #include "config.h"
                     20: #endif
                     21:
1.106     kristaps   22: #include <sys/types.h>
                     23:
1.1       kristaps   24: #include <assert.h>
                     25: #include <stdarg.h>
1.73      kristaps   26: #include <stdio.h>
1.1       kristaps   27: #include <stdlib.h>
                     28: #include <string.h>
1.120     kristaps   29: #include <time.h>
1.1       kristaps   30:
1.136     kristaps   31: #include "mandoc.h"
1.70      kristaps   32: #include "libmdoc.h"
1.113     kristaps   33: #include "libmandoc.h"
1.1       kristaps   34:
                     35: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.82      kristaps   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",
                     46:        /* LINTED */
1.114     kristaps   47:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   48:        /* LINTED */
1.114     kristaps   49:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps   50:        /* LINTED */
1.114     kristaps   51:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   52:        "Ac",           "Ao",           "Aq",           "At",
                     53:        "Bc",           "Bf",           "Bo",           "Bq",
                     54:        "Bsx",          "Bx",           "Db",           "Dc",
                     55:        "Do",           "Dq",           "Ec",           "Ef",
                     56:        "Em",           "Eo",           "Fx",           "Ms",
                     57:        "No",           "Ns",           "Nx",           "Ox",
                     58:        "Pc",           "Pf",           "Po",           "Pq",
                     59:        "Qc",           "Ql",           "Qo",           "Qq",
                     60:        "Re",           "Rs",           "Sc",           "So",
                     61:        "Sq",           "Sm",           "Sx",           "Sy",
                     62:        "Tn",           "Ux",           "Xc",           "Xo",
                     63:        "Fo",           "Fc",           "Oo",           "Oc",
                     64:        "Bk",           "Ek",           "Bt",           "Hf",
1.82      kristaps   65:        "Fr",           "Ud",           "Lb",           "Lp",
                     66:        "Lk",           "Mt",           "Brq",          "Bro",
1.64      kristaps   67:        /* LINTED */
1.114     kristaps   68:        "Brc",          "%C",           "Es",           "En",
1.69      kristaps   69:        /* LINTED */
1.114     kristaps   70:        "Dx",           "%Q",           "br",           "sp",
1.110     kristaps   71:        /* LINTED */
1.141     kristaps   72:        "%U",           "Ta"
1.1       kristaps   73:        };
                     74:
                     75: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                     76:        "split",                "nosplit",              "ragged",
                     77:        "unfilled",             "literal",              "file",
                     78:        "offset",               "bullet",               "dash",
                     79:        "hyphen",               "item",                 "enum",
                     80:        "tag",                  "diag",                 "hang",
                     81:        "ohang",                "inset",                "column",
                     82:        "width",                "compact",              "std",
1.52      kristaps   83:        "filled",               "words",                "emphasis",
1.108     kristaps   84:        "symbolic",             "nested",               "centered"
1.1       kristaps   85:        };
                     86:
                     87: const  char * const *mdoc_macronames = __mdoc_macronames;
                     88: const  char * const *mdoc_argnames = __mdoc_argnames;
                     89:
1.121     kristaps   90: static void              mdoc_node_free(struct mdoc_node *);
                     91: static void              mdoc_node_unlink(struct mdoc *,
                     92:                                struct mdoc_node *);
1.73      kristaps   93: static void              mdoc_free1(struct mdoc *);
1.113     kristaps   94: static void              mdoc_alloc1(struct mdoc *);
1.73      kristaps   95: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.117     kristaps   96:                                enum mdoct, enum mdoc_type);
1.73      kristaps   97: static int               node_append(struct mdoc *,
1.71      kristaps   98:                                struct mdoc_node *);
1.149     kristaps   99: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                    100: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
1.124     kristaps  101:
1.88      kristaps  102:
1.1       kristaps  103: const struct mdoc_node *
1.71      kristaps  104: mdoc_node(const struct mdoc *m)
1.1       kristaps  105: {
                    106:
1.71      kristaps  107:        return(MDOC_HALT & m->flags ? NULL : m->first);
1.1       kristaps  108: }
                    109:
                    110:
1.37      kristaps  111: const struct mdoc_meta *
1.71      kristaps  112: mdoc_meta(const struct mdoc *m)
1.37      kristaps  113: {
                    114:
1.71      kristaps  115:        return(MDOC_HALT & m->flags ? NULL : &m->meta);
1.37      kristaps  116: }
                    117:
                    118:
1.85      kristaps  119: /*
                    120:  * Frees volatile resources (parse tree, meta-data, fields).
                    121:  */
1.73      kristaps  122: static void
                    123: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  124: {
                    125:
                    126:        if (mdoc->first)
1.121     kristaps  127:                mdoc_node_delete(mdoc, mdoc->first);
1.67      kristaps  128:        if (mdoc->meta.title)
                    129:                free(mdoc->meta.title);
                    130:        if (mdoc->meta.os)
                    131:                free(mdoc->meta.os);
                    132:        if (mdoc->meta.name)
                    133:                free(mdoc->meta.name);
                    134:        if (mdoc->meta.arch)
                    135:                free(mdoc->meta.arch);
                    136:        if (mdoc->meta.vol)
                    137:                free(mdoc->meta.vol);
1.133     kristaps  138:        if (mdoc->meta.msec)
                    139:                free(mdoc->meta.msec);
1.73      kristaps  140: }
                    141:
                    142:
1.85      kristaps  143: /*
                    144:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    145:  */
1.113     kristaps  146: static void
1.73      kristaps  147: mdoc_alloc1(struct mdoc *mdoc)
                    148: {
1.67      kristaps  149:
1.112     kristaps  150:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  151:        mdoc->flags = 0;
1.85      kristaps  152:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  153:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  154:        mdoc->first = mdoc->last;
1.67      kristaps  155:        mdoc->last->type = MDOC_ROOT;
                    156:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  157: }
                    158:
                    159:
                    160: /*
1.85      kristaps  161:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    162:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    163:  * and the parser is ready for re-invocation on a new tree; however,
                    164:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  165:  */
1.113     kristaps  166: void
1.73      kristaps  167: mdoc_reset(struct mdoc *mdoc)
                    168: {
                    169:
                    170:        mdoc_free1(mdoc);
1.113     kristaps  171:        mdoc_alloc1(mdoc);
1.67      kristaps  172: }
                    173:
                    174:
1.68      kristaps  175: /*
1.85      kristaps  176:  * Completely free up all volatile and non-volatile parse resources.
                    177:  * After invocation, the pointer is no longer usable.
1.68      kristaps  178:  */
1.67      kristaps  179: void
1.38      kristaps  180: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  181: {
                    182:
1.73      kristaps  183:        mdoc_free1(mdoc);
1.1       kristaps  184:        free(mdoc);
                    185: }
                    186:
                    187:
1.85      kristaps  188: /*
                    189:  * Allocate volatile and non-volatile parse resources.
                    190:  */
1.1       kristaps  191: struct mdoc *
1.163     schwarze  192: mdoc_alloc(struct regset *regs, void *data, mandocmsg msg)
1.1       kristaps  193: {
                    194:        struct mdoc     *p;
                    195:
1.113     kristaps  196:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    197:
1.136     kristaps  198:        p->msg = msg;
1.1       kristaps  199:        p->data = data;
1.149     kristaps  200:        p->regs = regs;
1.73      kristaps  201:
1.113     kristaps  202:        mdoc_hash_init();
                    203:        mdoc_alloc1(p);
                    204:        return(p);
1.1       kristaps  205: }
                    206:
                    207:
1.68      kristaps  208: /*
                    209:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  210:  * through to macro_end() in macro.c.
1.68      kristaps  211:  */
1.1       kristaps  212: int
1.72      kristaps  213: mdoc_endparse(struct mdoc *m)
1.20      kristaps  214: {
                    215:
1.72      kristaps  216:        if (MDOC_HALT & m->flags)
1.20      kristaps  217:                return(0);
1.72      kristaps  218:        else if (mdoc_macroend(m))
1.20      kristaps  219:                return(1);
1.72      kristaps  220:        m->flags |= MDOC_HALT;
                    221:        return(0);
1.20      kristaps  222: }
                    223:
                    224:
1.50      kristaps  225: /*
1.53      kristaps  226:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  227:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  228:  */
1.20      kristaps  229: int
1.149     kristaps  230: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  231: {
                    232:
1.53      kristaps  233:        if (MDOC_HALT & m->flags)
1.20      kristaps  234:                return(0);
1.50      kristaps  235:
1.130     kristaps  236:        m->flags |= MDOC_NEWLINE;
1.153     schwarze  237:
                    238:        /*
                    239:         * Let the roff nS register switch SYNOPSIS mode early,
                    240:         * such that the parser knows at all times
                    241:         * whether this mode is on or off.
                    242:         * Note that this mode is also switched by the Sh macro.
                    243:         */
                    244:        if (m->regs->regs[(int)REG_nS].set) {
                    245:                if (m->regs->regs[(int)REG_nS].v.u)
                    246:                        m->flags |= MDOC_SYNOPSIS;
                    247:                else
                    248:                        m->flags &= ~MDOC_SYNOPSIS;
                    249:        }
                    250:
1.135     kristaps  251:        return(('.' == buf[offs] || '\'' == buf[offs]) ?
1.149     kristaps  252:                        mdoc_pmacro(m, ln, buf, offs) :
                    253:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  254: }
                    255:
                    256:
                    257: int
1.136     kristaps  258: mdoc_vmsg(struct mdoc *mdoc, enum mandocerr t,
                    259:                int ln, int pos, const char *fmt, ...)
1.1       kristaps  260: {
1.31      kristaps  261:        char             buf[256];
                    262:        va_list          ap;
1.1       kristaps  263:
1.31      kristaps  264:        va_start(ap, fmt);
1.136     kristaps  265:        vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
1.31      kristaps  266:        va_end(ap);
1.88      kristaps  267:
1.136     kristaps  268:        return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
1.88      kristaps  269: }
                    270:
                    271:
                    272: int
1.148     kristaps  273: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  274: {
1.122     kristaps  275:        assert(tok < MDOC_MAX);
                    276:
                    277:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  278:
1.88      kristaps  279:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.122     kristaps  280:                        MDOC_PBODY & m->flags)
1.148     kristaps  281:                return(mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY));
1.122     kristaps  282:
                    283:        /* If we're in the prologue, deny "body" macros.  */
                    284:
                    285:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    286:                        ! (MDOC_PBODY & m->flags)) {
1.148     kristaps  287:                if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG))
1.120     kristaps  288:                        return(0);
1.167     kristaps  289:                if (NULL == m->meta.msec)
                    290:                        m->meta.msec = mandoc_strdup("1");
1.120     kristaps  291:                if (NULL == m->meta.title)
1.140     kristaps  292:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.120     kristaps  293:                if (NULL == m->meta.vol)
1.140     kristaps  294:                        m->meta.vol = mandoc_strdup("LOCAL");
1.120     kristaps  295:                if (NULL == m->meta.os)
1.140     kristaps  296:                        m->meta.os = mandoc_strdup("LOCAL");
1.120     kristaps  297:                if (0 == m->meta.date)
                    298:                        m->meta.date = time(NULL);
                    299:                m->flags |= MDOC_PBODY;
                    300:        }
1.88      kristaps  301:
1.149     kristaps  302:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.73      kristaps  303: }
                    304:
                    305:
                    306: static int
                    307: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  308: {
                    309:
1.25      kristaps  310:        assert(mdoc->last);
                    311:        assert(mdoc->first);
                    312:        assert(MDOC_ROOT != p->type);
1.1       kristaps  313:
1.13      kristaps  314:        switch (mdoc->next) {
                    315:        case (MDOC_NEXT_SIBLING):
1.6       kristaps  316:                mdoc->last->next = p;
                    317:                p->prev = mdoc->last;
1.13      kristaps  318:                p->parent = mdoc->last->parent;
1.1       kristaps  319:                break;
1.13      kristaps  320:        case (MDOC_NEXT_CHILD):
                    321:                mdoc->last->child = p;
1.1       kristaps  322:                p->parent = mdoc->last;
                    323:                break;
                    324:        default:
1.13      kristaps  325:                abort();
                    326:                /* NOTREACHED */
1.1       kristaps  327:        }
                    328:
1.86      kristaps  329:        p->parent->nchild++;
                    330:
1.172   ! kristaps  331:        /*
        !           332:         * Copy over the normalised-data pointer of our parent.  Not
        !           333:         * everybody has one, but copying a null pointer is fine.
        !           334:         */
        !           335:
        !           336:        switch (p->type) {
        !           337:        case (MDOC_BODY):
        !           338:                /* FALLTHROUGH */
        !           339:        case (MDOC_TAIL):
        !           340:                /* FALLTHROUGH */
        !           341:        case (MDOC_HEAD):
        !           342:                p->norm = p->parent->norm;
        !           343:                break;
        !           344:        default:
        !           345:                break;
        !           346:        }
        !           347:
1.23      kristaps  348:        if ( ! mdoc_valid_pre(mdoc, p))
                    349:                return(0);
1.27      kristaps  350:
                    351:        switch (p->type) {
                    352:        case (MDOC_HEAD):
                    353:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  354:                p->parent->head = p;
1.27      kristaps  355:                break;
                    356:        case (MDOC_TAIL):
                    357:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  358:                p->parent->tail = p;
1.27      kristaps  359:                break;
                    360:        case (MDOC_BODY):
1.152     schwarze  361:                if (p->end)
                    362:                        break;
1.27      kristaps  363:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  364:                p->parent->body = p;
1.27      kristaps  365:                break;
                    366:        default:
                    367:                break;
                    368:        }
                    369:
1.1       kristaps  370:        mdoc->last = p;
1.71      kristaps  371:
                    372:        switch (p->type) {
                    373:        case (MDOC_TEXT):
                    374:                if ( ! mdoc_valid_post(mdoc))
                    375:                        return(0);
                    376:                break;
                    377:        default:
                    378:                break;
                    379:        }
                    380:
1.23      kristaps  381:        return(1);
1.1       kristaps  382: }
                    383:
                    384:
1.45      kristaps  385: static struct mdoc_node *
1.117     kristaps  386: node_alloc(struct mdoc *m, int line, int pos,
                    387:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  388: {
1.46      kristaps  389:        struct mdoc_node *p;
                    390:
1.113     kristaps  391:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.91      kristaps  392:        p->sec = m->lastsec;
1.73      kristaps  393:        p->line = line;
                    394:        p->pos = pos;
                    395:        p->tok = tok;
1.118     kristaps  396:        p->type = type;
1.150     kristaps  397:
                    398:        /* Flag analysis. */
                    399:
1.153     schwarze  400:        if (MDOC_SYNOPSIS & m->flags)
                    401:                p->flags |= MDOC_SYNPRETTY;
                    402:        else
                    403:                p->flags &= ~MDOC_SYNPRETTY;
1.130     kristaps  404:        if (MDOC_NEWLINE & m->flags)
                    405:                p->flags |= MDOC_LINE;
                    406:        m->flags &= ~MDOC_NEWLINE;
1.150     kristaps  407:
1.46      kristaps  408:        return(p);
1.45      kristaps  409: }
                    410:
                    411:
1.23      kristaps  412: int
1.117     kristaps  413: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.17      kristaps  414: {
                    415:        struct mdoc_node *p;
                    416:
1.91      kristaps  417:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.102     kristaps  418:        if ( ! node_append(m, p))
                    419:                return(0);
                    420:        m->next = MDOC_NEXT_CHILD;
                    421:        return(1);
1.17      kristaps  422: }
                    423:
                    424:
1.23      kristaps  425: int
1.117     kristaps  426: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  427: {
                    428:        struct mdoc_node *p;
                    429:
1.91      kristaps  430:        assert(m->first);
                    431:        assert(m->last);
1.1       kristaps  432:
1.91      kristaps  433:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.102     kristaps  434:        if ( ! node_append(m, p))
                    435:                return(0);
                    436:        m->next = MDOC_NEXT_CHILD;
                    437:        return(1);
1.1       kristaps  438: }
                    439:
                    440:
1.23      kristaps  441: int
1.117     kristaps  442: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  443: {
                    444:        struct mdoc_node *p;
                    445:
1.91      kristaps  446:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.102     kristaps  447:        if ( ! node_append(m, p))
                    448:                return(0);
                    449:        m->next = MDOC_NEXT_CHILD;
1.152     schwarze  450:        return(1);
                    451: }
                    452:
                    453:
                    454: int
                    455: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    456:                struct mdoc_node *body, enum mdoc_endbody end)
                    457: {
                    458:        struct mdoc_node *p;
                    459:
                    460:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    461:        p->pending = body;
                    462:        p->end = end;
                    463:        if ( ! node_append(m, p))
                    464:                return(0);
                    465:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  466:        return(1);
1.1       kristaps  467: }
                    468:
                    469:
1.23      kristaps  470: int
1.91      kristaps  471: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  472:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  473: {
                    474:        struct mdoc_node *p;
                    475:
1.91      kristaps  476:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  477:        p->args = args;
                    478:        if (p->args)
1.53      kristaps  479:                (args->refcnt)++;
1.172   ! kristaps  480:
        !           481:        switch (tok) {
        !           482:        case (MDOC_Bd):
        !           483:                /* FALLTHROUGH */
        !           484:        case (MDOC_Bf):
        !           485:                /* FALLTHROUGH */
        !           486:        case (MDOC_Bl):
        !           487:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
        !           488:                break;
        !           489:        default:
        !           490:                break;
        !           491:        }
        !           492:
1.102     kristaps  493:        if ( ! node_append(m, p))
                    494:                return(0);
                    495:        m->next = MDOC_NEXT_CHILD;
                    496:        return(1);
1.1       kristaps  497: }
                    498:
                    499:
1.23      kristaps  500: int
1.91      kristaps  501: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  502:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  503: {
                    504:        struct mdoc_node *p;
                    505:
1.91      kristaps  506:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.77      kristaps  507:        p->args = args;
                    508:        if (p->args)
1.53      kristaps  509:                (args->refcnt)++;
1.172   ! kristaps  510:
        !           511:        switch (tok) {
        !           512:        case (MDOC_An):
        !           513:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
        !           514:                break;
        !           515:        default:
        !           516:                break;
        !           517:        }
        !           518:
1.102     kristaps  519:        if ( ! node_append(m, p))
                    520:                return(0);
                    521:        m->next = MDOC_NEXT_CHILD;
                    522:        return(1);
1.1       kristaps  523: }
                    524:
                    525:
1.124     kristaps  526: int
                    527: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  528: {
1.91      kristaps  529:        struct mdoc_node *n;
1.124     kristaps  530:        size_t            sv, len;
                    531:
                    532:        len = strlen(p);
1.1       kristaps  533:
1.125     kristaps  534:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.113     kristaps  535:        n->string = mandoc_malloc(len + 1);
1.91      kristaps  536:        sv = strlcpy(n->string, p, len + 1);
                    537:
                    538:        /* Prohibit truncation. */
                    539:        assert(sv < len + 1);
                    540:
1.101     kristaps  541:        if ( ! node_append(m, n))
                    542:                return(0);
1.124     kristaps  543:
1.101     kristaps  544:        m->next = MDOC_NEXT_SIBLING;
                    545:        return(1);
1.91      kristaps  546: }
                    547:
                    548:
1.155     kristaps  549: static void
1.53      kristaps  550: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  551: {
                    552:
1.172   ! kristaps  553:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  554:                free(p->norm);
1.53      kristaps  555:        if (p->string)
                    556:                free(p->string);
                    557:        if (p->args)
                    558:                mdoc_argv_free(p->args);
1.1       kristaps  559:        free(p);
                    560: }
                    561:
                    562:
1.121     kristaps  563: static void
                    564: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    565: {
                    566:
                    567:        /* Adjust siblings. */
                    568:
                    569:        if (n->prev)
                    570:                n->prev->next = n->next;
                    571:        if (n->next)
                    572:                n->next->prev = n->prev;
                    573:
                    574:        /* Adjust parent. */
                    575:
                    576:        if (n->parent) {
                    577:                n->parent->nchild--;
                    578:                if (n->parent->child == n)
                    579:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  580:                if (n->parent->last == n)
                    581:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  582:        }
                    583:
                    584:        /* Adjust parse point, if applicable. */
                    585:
                    586:        if (m && m->last == n) {
                    587:                if (n->prev) {
                    588:                        m->last = n->prev;
                    589:                        m->next = MDOC_NEXT_SIBLING;
                    590:                } else {
                    591:                        m->last = n->parent;
                    592:                        m->next = MDOC_NEXT_CHILD;
                    593:                }
                    594:        }
                    595:
                    596:        if (m && m->first == n)
                    597:                m->first = NULL;
                    598: }
                    599:
                    600:
1.53      kristaps  601: void
1.121     kristaps  602: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  603: {
                    604:
1.121     kristaps  605:        while (p->child) {
                    606:                assert(p->nchild);
                    607:                mdoc_node_delete(m, p->child);
                    608:        }
                    609:        assert(0 == p->nchild);
1.1       kristaps  610:
1.121     kristaps  611:        mdoc_node_unlink(m, p);
1.53      kristaps  612:        mdoc_node_free(p);
1.1       kristaps  613: }
                    614:
                    615:
1.53      kristaps  616: /*
                    617:  * Parse free-form text, that is, a line that does not begin with the
                    618:  * control character.
                    619:  */
                    620: static int
1.149     kristaps  621: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  622: {
1.142     kristaps  623:        char             *c, *ws, *end;
                    624:        struct mdoc_node *n;
1.1       kristaps  625:
1.123     kristaps  626:        /* Ignore bogus comments. */
                    627:
1.135     kristaps  628:        if ('\\' == buf[offs] &&
                    629:                        '.' == buf[offs + 1] &&
                    630:                        '"' == buf[offs + 2])
1.136     kristaps  631:                return(mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT));
1.123     kristaps  632:
1.124     kristaps  633:        /* No text before an initial macro. */
                    634:
1.85      kristaps  635:        if (SEC_NONE == m->lastnamed)
1.136     kristaps  636:                return(mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT));
1.142     kristaps  637:
                    638:        assert(m->last);
                    639:        n = m->last;
                    640:
                    641:        /*
1.144     kristaps  642:         * Divert directly to list processing if we're encountering a
1.142     kristaps  643:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  644:         * (a MDOC_BODY means it's already open, in which case we should
                    645:         * process within its context in the normal way).
1.142     kristaps  646:         */
                    647:
1.143     kristaps  648:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172   ! kristaps  649:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  650:                /* `Bl' is open without any children. */
1.142     kristaps  651:                m->flags |= MDOC_FREECOL;
1.149     kristaps  652:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  653:        }
                    654:
                    655:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    656:                        NULL != n->parent &&
                    657:                        MDOC_Bl == n->parent->tok &&
1.172   ! kristaps  658:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  659:                /* `Bl' has block-level `It' children. */
1.142     kristaps  660:                m->flags |= MDOC_FREECOL;
1.149     kristaps  661:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  662:        }
1.124     kristaps  663:
1.137     schwarze  664:        /*
                    665:         * Search for the beginning of unescaped trailing whitespace (ws)
                    666:         * and for the first character not to be output (end).
                    667:         */
1.139     kristaps  668:
                    669:        /* FIXME: replace with strcspn(). */
1.137     schwarze  670:        ws = NULL;
                    671:        for (c = end = buf + offs; *c; c++) {
                    672:                switch (*c) {
1.138     kristaps  673:                case '-':
                    674:                        if (mandoc_hyph(buf + offs, c))
                    675:                                *c = ASCII_HYPH;
1.145     kristaps  676:                        ws = NULL;
1.138     kristaps  677:                        break;
1.137     schwarze  678:                case ' ':
                    679:                        if (NULL == ws)
                    680:                                ws = c;
                    681:                        continue;
                    682:                case '\t':
                    683:                        /*
                    684:                         * Always warn about trailing tabs,
                    685:                         * even outside literal context,
                    686:                         * where they should be put on the next line.
                    687:                         */
                    688:                        if (NULL == ws)
                    689:                                ws = c;
                    690:                        /*
                    691:                         * Strip trailing tabs in literal context only;
                    692:                         * outside, they affect the next line.
                    693:                         */
                    694:                        if (MDOC_LITERAL & m->flags)
                    695:                                continue;
                    696:                        break;
                    697:                case '\\':
                    698:                        /* Skip the escaped character, too, if any. */
                    699:                        if (c[1])
                    700:                                c++;
                    701:                        /* FALLTHROUGH */
                    702:                default:
                    703:                        ws = NULL;
                    704:                        break;
                    705:                }
                    706:                end = c + 1;
                    707:        }
                    708:        *end = '\0';
1.91      kristaps  709:
1.137     schwarze  710:        if (ws)
                    711:                if ( ! mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE))
                    712:                        return(0);
1.115     kristaps  713:
1.137     schwarze  714:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
                    715:                if ( ! mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN))
1.119     kristaps  716:                        return(0);
1.124     kristaps  717:
1.119     kristaps  718:                /*
1.165     schwarze  719:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  720:                 * blank lines aren't allowed, but enough manuals assume this
                    721:                 * behaviour that we want to work around it.
1.119     kristaps  722:                 */
1.165     schwarze  723:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.119     kristaps  724:                        return(0);
1.124     kristaps  725:
                    726:                m->next = MDOC_NEXT_SIBLING;
                    727:                return(1);
1.119     kristaps  728:        }
1.68      kristaps  729:
1.137     schwarze  730:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
                    731:                return(0);
1.91      kristaps  732:
1.137     schwarze  733:        if (MDOC_LITERAL & m->flags)
                    734:                return(1);
1.128     kristaps  735:
                    736:        /*
                    737:         * End-of-sentence check.  If the last character is an unescaped
                    738:         * EOS character, then flag the node as being the end of a
                    739:         * sentence.  The front-end will know how to interpret this.
                    740:         */
1.132     kristaps  741:
1.137     schwarze  742:        assert(buf < end);
                    743:
1.159     schwarze  744:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.128     kristaps  745:                m->last->flags |= MDOC_EOS;
                    746:
                    747:        return(1);
1.1       kristaps  748: }
                    749:
                    750:
1.53      kristaps  751: /*
                    752:  * Parse a macro line, that is, a line beginning with the control
                    753:  * character.
                    754:  */
1.155     kristaps  755: static int
1.149     kristaps  756: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  757: {
1.144     kristaps  758:        enum mdoct        tok;
                    759:        int               i, j, sv;
                    760:        char              mac[5];
                    761:        struct mdoc_node *n;
1.1       kristaps  762:
1.81      kristaps  763:        /* Empty lines are ignored. */
1.63      kristaps  764:
1.135     kristaps  765:        offs++;
                    766:
                    767:        if ('\0' == buf[offs])
1.63      kristaps  768:                return(1);
                    769:
1.135     kristaps  770:        i = offs;
1.100     kristaps  771:
1.160     kristaps  772:        /* Accept tabs/whitespace after the initial control char. */
1.100     kristaps  773:
1.160     kristaps  774:        if (' ' == buf[i] || '\t' == buf[i]) {
1.100     kristaps  775:                i++;
1.160     kristaps  776:                while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
1.63      kristaps  777:                        i++;
1.115     kristaps  778:                if ('\0' == buf[i])
1.63      kristaps  779:                        return(1);
                    780:        }
1.1       kristaps  781:
1.130     kristaps  782:        sv = i;
                    783:
1.160     kristaps  784:        /*
1.162     schwarze  785:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  786:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  787:         */
1.1       kristaps  788:
1.162     schwarze  789:        j = 0;
1.165     schwarze  790:        while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
1.162     schwarze  791:                mac[j++] = buf[i++];
1.135     kristaps  792:        mac[j] = '\0';
1.1       kristaps  793:
1.163     schwarze  794:        tok = (j > 1 || j < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
                    795:        if (MDOC_MAX == tok) {
1.168     kristaps  796:                mdoc_vmsg(m, MANDOCERR_MACRO, ln, sv, "%s", buf + sv - 1);
1.58      kristaps  797:                return(1);
1.53      kristaps  798:        }
1.1       kristaps  799:
1.160     kristaps  800:        /* Disregard the first trailing tab, if applicable. */
                    801:
                    802:        if ('\t' == buf[i])
                    803:                i++;
                    804:
                    805:        /* Jump to the next non-whitespace word. */
1.1       kristaps  806:
1.65      kristaps  807:        while (buf[i] && ' ' == buf[i])
1.53      kristaps  808:                i++;
1.1       kristaps  809:
1.125     kristaps  810:        /*
                    811:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    812:         * into the parser as "text", so we only warn about spaces here.
                    813:         */
1.115     kristaps  814:
                    815:        if ('\0' == buf[i] && ' ' == buf[i - 1])
1.136     kristaps  816:                if ( ! mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE))
1.115     kristaps  817:                        goto err;
                    818:
1.144     kristaps  819:        /*
                    820:         * If an initial macro or a list invocation, divert directly
                    821:         * into macro processing.
                    822:         */
                    823:
                    824:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.149     kristaps  825:                if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.144     kristaps  826:                        goto err;
                    827:                return(1);
                    828:        }
                    829:
                    830:        n = m->last;
                    831:        assert(m->last);
                    832:
                    833:        /*
                    834:         * If the first macro of a `Bl -column', open an `It' block
                    835:         * context around the parsed macro.
                    836:         */
                    837:
                    838:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172   ! kristaps  839:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  840:                m->flags |= MDOC_FREECOL;
1.149     kristaps  841:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  842:                        goto err;
                    843:                return(1);
                    844:        }
                    845:
                    846:        /*
                    847:         * If we're following a block-level `It' within a `Bl -column'
                    848:         * context (perhaps opened in the above block or in ptext()),
                    849:         * then open an `It' block context around the parsed macro.
1.98      kristaps  850:         */
1.144     kristaps  851:
                    852:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    853:                        NULL != n->parent &&
                    854:                        MDOC_Bl == n->parent->tok &&
1.172   ! kristaps  855:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  856:                m->flags |= MDOC_FREECOL;
1.149     kristaps  857:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  858:                        goto err;
                    859:                return(1);
                    860:        }
                    861:
                    862:        /* Normal processing of a macro. */
                    863:
1.149     kristaps  864:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.53      kristaps  865:                goto err;
1.1       kristaps  866:
1.53      kristaps  867:        return(1);
1.1       kristaps  868:
1.53      kristaps  869: err:   /* Error out. */
1.1       kristaps  870:
1.53      kristaps  871:        m->flags |= MDOC_HALT;
                    872:        return(0);
1.1       kristaps  873: }
1.100     kristaps  874:
                    875:

CVSweb