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

Annotation of mandoc/mdoc.c, Revision 1.178

1.178   ! kristaps    1: /*     $Id: mdoc.c,v 1.177 2011/01/03 11:27:33 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);
1.175     kristaps  101: static int               mdoc_span_alloc(struct mdoc *,
                    102:                                const struct tbl_span *);
1.124     kristaps  103:
1.88      kristaps  104:
1.1       kristaps  105: const struct mdoc_node *
1.71      kristaps  106: mdoc_node(const struct mdoc *m)
1.1       kristaps  107: {
                    108:
1.177     kristaps  109:        assert( ! (MDOC_HALT & m->flags));
                    110:        return(m->first);
1.1       kristaps  111: }
                    112:
                    113:
1.37      kristaps  114: const struct mdoc_meta *
1.71      kristaps  115: mdoc_meta(const struct mdoc *m)
1.37      kristaps  116: {
                    117:
1.177     kristaps  118:        assert( ! (MDOC_HALT & m->flags));
                    119:        return(&m->meta);
1.37      kristaps  120: }
                    121:
                    122:
1.85      kristaps  123: /*
                    124:  * Frees volatile resources (parse tree, meta-data, fields).
                    125:  */
1.73      kristaps  126: static void
                    127: mdoc_free1(struct mdoc *mdoc)
1.67      kristaps  128: {
                    129:
                    130:        if (mdoc->first)
1.121     kristaps  131:                mdoc_node_delete(mdoc, mdoc->first);
1.67      kristaps  132:        if (mdoc->meta.title)
                    133:                free(mdoc->meta.title);
                    134:        if (mdoc->meta.os)
                    135:                free(mdoc->meta.os);
                    136:        if (mdoc->meta.name)
                    137:                free(mdoc->meta.name);
                    138:        if (mdoc->meta.arch)
                    139:                free(mdoc->meta.arch);
                    140:        if (mdoc->meta.vol)
                    141:                free(mdoc->meta.vol);
1.133     kristaps  142:        if (mdoc->meta.msec)
                    143:                free(mdoc->meta.msec);
1.73      kristaps  144: }
                    145:
                    146:
1.85      kristaps  147: /*
                    148:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    149:  */
1.113     kristaps  150: static void
1.73      kristaps  151: mdoc_alloc1(struct mdoc *mdoc)
                    152: {
1.67      kristaps  153:
1.112     kristaps  154:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.67      kristaps  155:        mdoc->flags = 0;
1.85      kristaps  156:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.113     kristaps  157:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.70      kristaps  158:        mdoc->first = mdoc->last;
1.67      kristaps  159:        mdoc->last->type = MDOC_ROOT;
                    160:        mdoc->next = MDOC_NEXT_CHILD;
1.73      kristaps  161: }
                    162:
                    163:
                    164: /*
1.85      kristaps  165:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    166:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    167:  * and the parser is ready for re-invocation on a new tree; however,
                    168:  * cross-parse non-volatile data is kept intact.
1.73      kristaps  169:  */
1.113     kristaps  170: void
1.73      kristaps  171: mdoc_reset(struct mdoc *mdoc)
                    172: {
                    173:
                    174:        mdoc_free1(mdoc);
1.113     kristaps  175:        mdoc_alloc1(mdoc);
1.67      kristaps  176: }
                    177:
                    178:
1.68      kristaps  179: /*
1.85      kristaps  180:  * Completely free up all volatile and non-volatile parse resources.
                    181:  * After invocation, the pointer is no longer usable.
1.68      kristaps  182:  */
1.67      kristaps  183: void
1.38      kristaps  184: mdoc_free(struct mdoc *mdoc)
1.34      kristaps  185: {
                    186:
1.73      kristaps  187:        mdoc_free1(mdoc);
1.1       kristaps  188:        free(mdoc);
                    189: }
                    190:
                    191:
1.85      kristaps  192: /*
                    193:  * Allocate volatile and non-volatile parse resources.
                    194:  */
1.1       kristaps  195: struct mdoc *
1.163     schwarze  196: mdoc_alloc(struct regset *regs, void *data, mandocmsg msg)
1.1       kristaps  197: {
                    198:        struct mdoc     *p;
                    199:
1.113     kristaps  200:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    201:
1.136     kristaps  202:        p->msg = msg;
1.1       kristaps  203:        p->data = data;
1.149     kristaps  204:        p->regs = regs;
1.73      kristaps  205:
1.113     kristaps  206:        mdoc_hash_init();
                    207:        mdoc_alloc1(p);
                    208:        return(p);
1.1       kristaps  209: }
                    210:
                    211:
1.68      kristaps  212: /*
                    213:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.85      kristaps  214:  * through to macro_end() in macro.c.
1.68      kristaps  215:  */
1.1       kristaps  216: int
1.72      kristaps  217: mdoc_endparse(struct mdoc *m)
1.20      kristaps  218: {
                    219:
1.177     kristaps  220:        assert( ! (MDOC_HALT & m->flags));
                    221:        if (mdoc_macroend(m))
1.20      kristaps  222:                return(1);
1.72      kristaps  223:        m->flags |= MDOC_HALT;
                    224:        return(0);
1.20      kristaps  225: }
                    226:
1.175     kristaps  227: int
                    228: mdoc_addspan(struct mdoc *m, const struct tbl_span *sp)
                    229: {
                    230:
1.177     kristaps  231:        assert( ! (MDOC_HALT & m->flags));
1.175     kristaps  232:
                    233:        /* No text before an initial macro. */
                    234:
                    235:        if (SEC_NONE == m->lastnamed) {
                    236:                /* FIXME: grab from span. */
                    237:                mdoc_pmsg(m, 0, 0, MANDOCERR_NOTEXT);
                    238:                return(1);
                    239:        }
                    240:
                    241:        return(mdoc_span_alloc(m, sp));
                    242: }
                    243:
1.20      kristaps  244:
1.50      kristaps  245: /*
1.53      kristaps  246:  * Main parse routine.  Parses a single line -- really just hands off to
1.123     kristaps  247:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.50      kristaps  248:  */
1.20      kristaps  249: int
1.149     kristaps  250: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  251: {
                    252:
1.177     kristaps  253:        assert( ! (MDOC_HALT & m->flags));
1.50      kristaps  254:
1.130     kristaps  255:        m->flags |= MDOC_NEWLINE;
1.153     schwarze  256:
                    257:        /*
                    258:         * Let the roff nS register switch SYNOPSIS mode early,
                    259:         * such that the parser knows at all times
                    260:         * whether this mode is on or off.
                    261:         * Note that this mode is also switched by the Sh macro.
                    262:         */
                    263:        if (m->regs->regs[(int)REG_nS].set) {
                    264:                if (m->regs->regs[(int)REG_nS].v.u)
                    265:                        m->flags |= MDOC_SYNOPSIS;
                    266:                else
                    267:                        m->flags &= ~MDOC_SYNOPSIS;
                    268:        }
                    269:
1.135     kristaps  270:        return(('.' == buf[offs] || '\'' == buf[offs]) ?
1.149     kristaps  271:                        mdoc_pmacro(m, ln, buf, offs) :
                    272:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  273: }
                    274:
                    275:
                    276: int
1.136     kristaps  277: mdoc_vmsg(struct mdoc *mdoc, enum mandocerr t,
                    278:                int ln, int pos, const char *fmt, ...)
1.1       kristaps  279: {
1.31      kristaps  280:        char             buf[256];
                    281:        va_list          ap;
1.1       kristaps  282:
1.31      kristaps  283:        va_start(ap, fmt);
1.136     kristaps  284:        vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
1.31      kristaps  285:        va_end(ap);
1.88      kristaps  286:
1.136     kristaps  287:        return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
1.88      kristaps  288: }
                    289:
                    290:
                    291: int
1.148     kristaps  292: mdoc_macro(MACRO_PROT_ARGS)
1.88      kristaps  293: {
1.122     kristaps  294:        assert(tok < MDOC_MAX);
                    295:
                    296:        /* If we're in the body, deny prologue calls. */
1.117     kristaps  297:
1.88      kristaps  298:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.174     kristaps  299:                        MDOC_PBODY & m->flags) {
                    300:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY);
                    301:                return(1);
                    302:        }
1.122     kristaps  303:
                    304:        /* If we're in the prologue, deny "body" macros.  */
                    305:
                    306:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
                    307:                        ! (MDOC_PBODY & m->flags)) {
1.174     kristaps  308:                mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG);
1.167     kristaps  309:                if (NULL == m->meta.msec)
                    310:                        m->meta.msec = mandoc_strdup("1");
1.120     kristaps  311:                if (NULL == m->meta.title)
1.140     kristaps  312:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.120     kristaps  313:                if (NULL == m->meta.vol)
1.140     kristaps  314:                        m->meta.vol = mandoc_strdup("LOCAL");
1.120     kristaps  315:                if (NULL == m->meta.os)
1.140     kristaps  316:                        m->meta.os = mandoc_strdup("LOCAL");
1.120     kristaps  317:                if (0 == m->meta.date)
                    318:                        m->meta.date = time(NULL);
                    319:                m->flags |= MDOC_PBODY;
                    320:        }
1.88      kristaps  321:
1.149     kristaps  322:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.73      kristaps  323: }
                    324:
                    325:
                    326: static int
                    327: node_append(struct mdoc *mdoc, struct mdoc_node *p)
1.1       kristaps  328: {
                    329:
1.25      kristaps  330:        assert(mdoc->last);
                    331:        assert(mdoc->first);
                    332:        assert(MDOC_ROOT != p->type);
1.1       kristaps  333:
1.13      kristaps  334:        switch (mdoc->next) {
                    335:        case (MDOC_NEXT_SIBLING):
1.6       kristaps  336:                mdoc->last->next = p;
                    337:                p->prev = mdoc->last;
1.13      kristaps  338:                p->parent = mdoc->last->parent;
1.1       kristaps  339:                break;
1.13      kristaps  340:        case (MDOC_NEXT_CHILD):
                    341:                mdoc->last->child = p;
1.1       kristaps  342:                p->parent = mdoc->last;
                    343:                break;
                    344:        default:
1.13      kristaps  345:                abort();
                    346:                /* NOTREACHED */
1.1       kristaps  347:        }
                    348:
1.86      kristaps  349:        p->parent->nchild++;
                    350:
1.172     kristaps  351:        /*
                    352:         * Copy over the normalised-data pointer of our parent.  Not
                    353:         * everybody has one, but copying a null pointer is fine.
                    354:         */
                    355:
                    356:        switch (p->type) {
                    357:        case (MDOC_BODY):
                    358:                /* FALLTHROUGH */
                    359:        case (MDOC_TAIL):
                    360:                /* FALLTHROUGH */
                    361:        case (MDOC_HEAD):
                    362:                p->norm = p->parent->norm;
                    363:                break;
                    364:        default:
                    365:                break;
                    366:        }
                    367:
1.23      kristaps  368:        if ( ! mdoc_valid_pre(mdoc, p))
                    369:                return(0);
1.27      kristaps  370:
                    371:        switch (p->type) {
                    372:        case (MDOC_HEAD):
                    373:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  374:                p->parent->head = p;
1.27      kristaps  375:                break;
                    376:        case (MDOC_TAIL):
                    377:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  378:                p->parent->tail = p;
1.27      kristaps  379:                break;
                    380:        case (MDOC_BODY):
1.152     schwarze  381:                if (p->end)
                    382:                        break;
1.27      kristaps  383:                assert(MDOC_BLOCK == p->parent->type);
1.53      kristaps  384:                p->parent->body = p;
1.27      kristaps  385:                break;
                    386:        default:
                    387:                break;
                    388:        }
                    389:
1.1       kristaps  390:        mdoc->last = p;
1.71      kristaps  391:
                    392:        switch (p->type) {
1.176     kristaps  393:        case (MDOC_TBL):
                    394:                /* FALLTHROUGH */
1.71      kristaps  395:        case (MDOC_TEXT):
                    396:                if ( ! mdoc_valid_post(mdoc))
                    397:                        return(0);
                    398:                break;
                    399:        default:
                    400:                break;
                    401:        }
                    402:
1.23      kristaps  403:        return(1);
1.1       kristaps  404: }
                    405:
                    406:
1.45      kristaps  407: static struct mdoc_node *
1.117     kristaps  408: node_alloc(struct mdoc *m, int line, int pos,
                    409:                enum mdoct tok, enum mdoc_type type)
1.45      kristaps  410: {
1.46      kristaps  411:        struct mdoc_node *p;
                    412:
1.113     kristaps  413:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.91      kristaps  414:        p->sec = m->lastsec;
1.73      kristaps  415:        p->line = line;
                    416:        p->pos = pos;
                    417:        p->tok = tok;
1.118     kristaps  418:        p->type = type;
1.150     kristaps  419:
                    420:        /* Flag analysis. */
                    421:
1.153     schwarze  422:        if (MDOC_SYNOPSIS & m->flags)
                    423:                p->flags |= MDOC_SYNPRETTY;
                    424:        else
                    425:                p->flags &= ~MDOC_SYNPRETTY;
1.130     kristaps  426:        if (MDOC_NEWLINE & m->flags)
                    427:                p->flags |= MDOC_LINE;
                    428:        m->flags &= ~MDOC_NEWLINE;
1.150     kristaps  429:
1.46      kristaps  430:        return(p);
1.45      kristaps  431: }
                    432:
                    433:
1.23      kristaps  434: int
1.117     kristaps  435: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.17      kristaps  436: {
                    437:        struct mdoc_node *p;
                    438:
1.91      kristaps  439:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.102     kristaps  440:        if ( ! node_append(m, p))
                    441:                return(0);
                    442:        m->next = MDOC_NEXT_CHILD;
                    443:        return(1);
1.17      kristaps  444: }
                    445:
                    446:
1.23      kristaps  447: int
1.117     kristaps  448: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  449: {
                    450:        struct mdoc_node *p;
                    451:
1.91      kristaps  452:        assert(m->first);
                    453:        assert(m->last);
1.1       kristaps  454:
1.91      kristaps  455:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.102     kristaps  456:        if ( ! node_append(m, p))
                    457:                return(0);
                    458:        m->next = MDOC_NEXT_CHILD;
                    459:        return(1);
1.1       kristaps  460: }
                    461:
                    462:
1.23      kristaps  463: int
1.117     kristaps  464: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  465: {
                    466:        struct mdoc_node *p;
                    467:
1.91      kristaps  468:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.102     kristaps  469:        if ( ! node_append(m, p))
                    470:                return(0);
                    471:        m->next = MDOC_NEXT_CHILD;
1.152     schwarze  472:        return(1);
                    473: }
                    474:
                    475:
                    476: int
                    477: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    478:                struct mdoc_node *body, enum mdoc_endbody end)
                    479: {
                    480:        struct mdoc_node *p;
                    481:
                    482:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    483:        p->pending = body;
                    484:        p->end = end;
                    485:        if ( ! node_append(m, p))
                    486:                return(0);
                    487:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  488:        return(1);
1.1       kristaps  489: }
                    490:
                    491:
1.23      kristaps  492: int
1.91      kristaps  493: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  494:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  495: {
                    496:        struct mdoc_node *p;
                    497:
1.91      kristaps  498:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.77      kristaps  499:        p->args = args;
                    500:        if (p->args)
1.53      kristaps  501:                (args->refcnt)++;
1.172     kristaps  502:
                    503:        switch (tok) {
                    504:        case (MDOC_Bd):
                    505:                /* FALLTHROUGH */
                    506:        case (MDOC_Bf):
                    507:                /* FALLTHROUGH */
                    508:        case (MDOC_Bl):
1.173     kristaps  509:                /* FALLTHROUGH */
                    510:        case (MDOC_Rs):
1.172     kristaps  511:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    512:                break;
                    513:        default:
                    514:                break;
                    515:        }
                    516:
1.102     kristaps  517:        if ( ! node_append(m, p))
                    518:                return(0);
                    519:        m->next = MDOC_NEXT_CHILD;
                    520:        return(1);
1.1       kristaps  521: }
                    522:
                    523:
1.23      kristaps  524: int
1.91      kristaps  525: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.117     kristaps  526:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  527: {
                    528:        struct mdoc_node *p;
                    529:
1.91      kristaps  530:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.77      kristaps  531:        p->args = args;
                    532:        if (p->args)
1.53      kristaps  533:                (args->refcnt)++;
1.172     kristaps  534:
                    535:        switch (tok) {
                    536:        case (MDOC_An):
                    537:                p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
                    538:                break;
                    539:        default:
                    540:                break;
                    541:        }
                    542:
1.102     kristaps  543:        if ( ! node_append(m, p))
                    544:                return(0);
                    545:        m->next = MDOC_NEXT_CHILD;
1.175     kristaps  546:        return(1);
                    547: }
                    548:
                    549: static int
                    550: mdoc_span_alloc(struct mdoc *m, const struct tbl_span *sp)
                    551: {
                    552:        struct mdoc_node *n;
                    553:
1.178   ! kristaps  554:        n = node_alloc(m, sp->line, 0, MDOC_MAX, MDOC_TBL);
1.175     kristaps  555:        n->span = sp;
                    556:
                    557:        if ( ! node_append(m, n))
                    558:                return(0);
                    559:
                    560:        m->next = MDOC_NEXT_SIBLING;
1.102     kristaps  561:        return(1);
1.1       kristaps  562: }
                    563:
                    564:
1.124     kristaps  565: int
                    566: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  567: {
1.91      kristaps  568:        struct mdoc_node *n;
1.124     kristaps  569:        size_t            sv, len;
                    570:
                    571:        len = strlen(p);
1.1       kristaps  572:
1.125     kristaps  573:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.113     kristaps  574:        n->string = mandoc_malloc(len + 1);
1.91      kristaps  575:        sv = strlcpy(n->string, p, len + 1);
                    576:
                    577:        /* Prohibit truncation. */
                    578:        assert(sv < len + 1);
                    579:
1.101     kristaps  580:        if ( ! node_append(m, n))
                    581:                return(0);
1.124     kristaps  582:
1.101     kristaps  583:        m->next = MDOC_NEXT_SIBLING;
                    584:        return(1);
1.91      kristaps  585: }
                    586:
                    587:
1.155     kristaps  588: static void
1.53      kristaps  589: mdoc_node_free(struct mdoc_node *p)
1.1       kristaps  590: {
                    591:
1.172     kristaps  592:        if (MDOC_BLOCK == p->type || MDOC_ELEM == p->type)
1.171     kristaps  593:                free(p->norm);
1.53      kristaps  594:        if (p->string)
                    595:                free(p->string);
                    596:        if (p->args)
                    597:                mdoc_argv_free(p->args);
1.1       kristaps  598:        free(p);
                    599: }
                    600:
                    601:
1.121     kristaps  602: static void
                    603: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    604: {
                    605:
                    606:        /* Adjust siblings. */
                    607:
                    608:        if (n->prev)
                    609:                n->prev->next = n->next;
                    610:        if (n->next)
                    611:                n->next->prev = n->prev;
                    612:
                    613:        /* Adjust parent. */
                    614:
                    615:        if (n->parent) {
                    616:                n->parent->nchild--;
                    617:                if (n->parent->child == n)
                    618:                        n->parent->child = n->prev ? n->prev : n->next;
1.169     kristaps  619:                if (n->parent->last == n)
                    620:                        n->parent->last = n->prev ? n->prev : NULL;
1.121     kristaps  621:        }
                    622:
                    623:        /* Adjust parse point, if applicable. */
                    624:
                    625:        if (m && m->last == n) {
                    626:                if (n->prev) {
                    627:                        m->last = n->prev;
                    628:                        m->next = MDOC_NEXT_SIBLING;
                    629:                } else {
                    630:                        m->last = n->parent;
                    631:                        m->next = MDOC_NEXT_CHILD;
                    632:                }
                    633:        }
                    634:
                    635:        if (m && m->first == n)
                    636:                m->first = NULL;
                    637: }
                    638:
                    639:
1.53      kristaps  640: void
1.121     kristaps  641: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  642: {
                    643:
1.121     kristaps  644:        while (p->child) {
                    645:                assert(p->nchild);
                    646:                mdoc_node_delete(m, p->child);
                    647:        }
                    648:        assert(0 == p->nchild);
1.1       kristaps  649:
1.121     kristaps  650:        mdoc_node_unlink(m, p);
1.53      kristaps  651:        mdoc_node_free(p);
1.1       kristaps  652: }
                    653:
                    654:
1.53      kristaps  655: /*
                    656:  * Parse free-form text, that is, a line that does not begin with the
                    657:  * control character.
                    658:  */
                    659: static int
1.149     kristaps  660: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  661: {
1.142     kristaps  662:        char             *c, *ws, *end;
                    663:        struct mdoc_node *n;
1.1       kristaps  664:
1.123     kristaps  665:        /* Ignore bogus comments. */
                    666:
1.135     kristaps  667:        if ('\\' == buf[offs] &&
                    668:                        '.' == buf[offs + 1] &&
1.174     kristaps  669:                        '"' == buf[offs + 2]) {
                    670:                mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT);
                    671:                return(1);
                    672:        }
1.123     kristaps  673:
1.124     kristaps  674:        /* No text before an initial macro. */
                    675:
1.174     kristaps  676:        if (SEC_NONE == m->lastnamed) {
                    677:                mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT);
                    678:                return(1);
                    679:        }
1.142     kristaps  680:
                    681:        assert(m->last);
                    682:        n = m->last;
                    683:
                    684:        /*
1.144     kristaps  685:         * Divert directly to list processing if we're encountering a
1.142     kristaps  686:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
1.144     kristaps  687:         * (a MDOC_BODY means it's already open, in which case we should
                    688:         * process within its context in the normal way).
1.142     kristaps  689:         */
                    690:
1.143     kristaps  691:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  692:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  693:                /* `Bl' is open without any children. */
1.142     kristaps  694:                m->flags |= MDOC_FREECOL;
1.149     kristaps  695:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  696:        }
                    697:
                    698:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    699:                        NULL != n->parent &&
                    700:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  701:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  702:                /* `Bl' has block-level `It' children. */
1.142     kristaps  703:                m->flags |= MDOC_FREECOL;
1.149     kristaps  704:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
1.142     kristaps  705:        }
1.124     kristaps  706:
1.137     schwarze  707:        /*
                    708:         * Search for the beginning of unescaped trailing whitespace (ws)
                    709:         * and for the first character not to be output (end).
                    710:         */
1.139     kristaps  711:
                    712:        /* FIXME: replace with strcspn(). */
1.137     schwarze  713:        ws = NULL;
                    714:        for (c = end = buf + offs; *c; c++) {
                    715:                switch (*c) {
1.138     kristaps  716:                case '-':
                    717:                        if (mandoc_hyph(buf + offs, c))
                    718:                                *c = ASCII_HYPH;
1.145     kristaps  719:                        ws = NULL;
1.138     kristaps  720:                        break;
1.137     schwarze  721:                case ' ':
                    722:                        if (NULL == ws)
                    723:                                ws = c;
                    724:                        continue;
                    725:                case '\t':
                    726:                        /*
                    727:                         * Always warn about trailing tabs,
                    728:                         * even outside literal context,
                    729:                         * where they should be put on the next line.
                    730:                         */
                    731:                        if (NULL == ws)
                    732:                                ws = c;
                    733:                        /*
                    734:                         * Strip trailing tabs in literal context only;
                    735:                         * outside, they affect the next line.
                    736:                         */
                    737:                        if (MDOC_LITERAL & m->flags)
                    738:                                continue;
                    739:                        break;
                    740:                case '\\':
                    741:                        /* Skip the escaped character, too, if any. */
                    742:                        if (c[1])
                    743:                                c++;
                    744:                        /* FALLTHROUGH */
                    745:                default:
                    746:                        ws = NULL;
                    747:                        break;
                    748:                }
                    749:                end = c + 1;
                    750:        }
                    751:        *end = '\0';
1.91      kristaps  752:
1.137     schwarze  753:        if (ws)
1.174     kristaps  754:                mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE);
1.115     kristaps  755:
1.137     schwarze  756:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.174     kristaps  757:                mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN);
1.124     kristaps  758:
1.119     kristaps  759:                /*
1.165     schwarze  760:                 * Insert a `sp' in the case of a blank line.  Technically,
1.124     kristaps  761:                 * blank lines aren't allowed, but enough manuals assume this
                    762:                 * behaviour that we want to work around it.
1.119     kristaps  763:                 */
1.165     schwarze  764:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_sp, NULL))
1.119     kristaps  765:                        return(0);
1.124     kristaps  766:
                    767:                m->next = MDOC_NEXT_SIBLING;
                    768:                return(1);
1.119     kristaps  769:        }
1.68      kristaps  770:
1.137     schwarze  771:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
                    772:                return(0);
1.91      kristaps  773:
1.137     schwarze  774:        if (MDOC_LITERAL & m->flags)
                    775:                return(1);
1.128     kristaps  776:
                    777:        /*
                    778:         * End-of-sentence check.  If the last character is an unescaped
                    779:         * EOS character, then flag the node as being the end of a
                    780:         * sentence.  The front-end will know how to interpret this.
                    781:         */
1.132     kristaps  782:
1.137     schwarze  783:        assert(buf < end);
                    784:
1.159     schwarze  785:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.128     kristaps  786:                m->last->flags |= MDOC_EOS;
                    787:
                    788:        return(1);
1.1       kristaps  789: }
                    790:
                    791:
1.53      kristaps  792: /*
                    793:  * Parse a macro line, that is, a line beginning with the control
                    794:  * character.
                    795:  */
1.155     kristaps  796: static int
1.149     kristaps  797: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  798: {
1.144     kristaps  799:        enum mdoct        tok;
                    800:        int               i, j, sv;
                    801:        char              mac[5];
                    802:        struct mdoc_node *n;
1.1       kristaps  803:
1.81      kristaps  804:        /* Empty lines are ignored. */
1.63      kristaps  805:
1.135     kristaps  806:        offs++;
                    807:
                    808:        if ('\0' == buf[offs])
1.63      kristaps  809:                return(1);
                    810:
1.135     kristaps  811:        i = offs;
1.100     kristaps  812:
1.160     kristaps  813:        /* Accept tabs/whitespace after the initial control char. */
1.100     kristaps  814:
1.160     kristaps  815:        if (' ' == buf[i] || '\t' == buf[i]) {
1.100     kristaps  816:                i++;
1.160     kristaps  817:                while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
1.63      kristaps  818:                        i++;
1.115     kristaps  819:                if ('\0' == buf[i])
1.63      kristaps  820:                        return(1);
                    821:        }
1.1       kristaps  822:
1.130     kristaps  823:        sv = i;
                    824:
1.160     kristaps  825:        /*
1.162     schwarze  826:         * Copy the first word into a nil-terminated buffer.
1.165     schwarze  827:         * Stop copying when a tab, space, or eoln is encountered.
1.160     kristaps  828:         */
1.1       kristaps  829:
1.162     schwarze  830:        j = 0;
1.165     schwarze  831:        while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
1.162     schwarze  832:                mac[j++] = buf[i++];
1.135     kristaps  833:        mac[j] = '\0';
1.1       kristaps  834:
1.163     schwarze  835:        tok = (j > 1 || j < 4) ? mdoc_hash_find(mac) : MDOC_MAX;
                    836:        if (MDOC_MAX == tok) {
1.168     kristaps  837:                mdoc_vmsg(m, MANDOCERR_MACRO, ln, sv, "%s", buf + sv - 1);
1.58      kristaps  838:                return(1);
1.53      kristaps  839:        }
1.1       kristaps  840:
1.160     kristaps  841:        /* Disregard the first trailing tab, if applicable. */
                    842:
                    843:        if ('\t' == buf[i])
                    844:                i++;
                    845:
                    846:        /* Jump to the next non-whitespace word. */
1.1       kristaps  847:
1.65      kristaps  848:        while (buf[i] && ' ' == buf[i])
1.53      kristaps  849:                i++;
1.1       kristaps  850:
1.125     kristaps  851:        /*
                    852:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    853:         * into the parser as "text", so we only warn about spaces here.
                    854:         */
1.115     kristaps  855:
                    856:        if ('\0' == buf[i] && ' ' == buf[i - 1])
1.174     kristaps  857:                mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE);
1.115     kristaps  858:
1.144     kristaps  859:        /*
                    860:         * If an initial macro or a list invocation, divert directly
                    861:         * into macro processing.
                    862:         */
                    863:
                    864:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
1.149     kristaps  865:                if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.144     kristaps  866:                        goto err;
                    867:                return(1);
                    868:        }
                    869:
                    870:        n = m->last;
                    871:        assert(m->last);
                    872:
                    873:        /*
                    874:         * If the first macro of a `Bl -column', open an `It' block
                    875:         * context around the parsed macro.
                    876:         */
                    877:
                    878:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.172     kristaps  879:                        LIST_column == n->norm->Bl.type) {
1.144     kristaps  880:                m->flags |= MDOC_FREECOL;
1.149     kristaps  881:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  882:                        goto err;
                    883:                return(1);
                    884:        }
                    885:
                    886:        /*
                    887:         * If we're following a block-level `It' within a `Bl -column'
                    888:         * context (perhaps opened in the above block or in ptext()),
                    889:         * then open an `It' block context around the parsed macro.
1.98      kristaps  890:         */
1.144     kristaps  891:
                    892:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    893:                        NULL != n->parent &&
                    894:                        MDOC_Bl == n->parent->tok &&
1.172     kristaps  895:                        LIST_column == n->parent->norm->Bl.type) {
1.144     kristaps  896:                m->flags |= MDOC_FREECOL;
1.149     kristaps  897:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.144     kristaps  898:                        goto err;
                    899:                return(1);
                    900:        }
                    901:
                    902:        /* Normal processing of a macro. */
                    903:
1.149     kristaps  904:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.53      kristaps  905:                goto err;
1.1       kristaps  906:
1.53      kristaps  907:        return(1);
1.1       kristaps  908:
1.53      kristaps  909: err:   /* Error out. */
1.1       kristaps  910:
1.53      kristaps  911:        m->flags |= MDOC_HALT;
                    912:        return(0);
1.1       kristaps  913: }
1.100     kristaps  914:
                    915:

CVSweb