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

Annotation of src/usr.bin/mandoc/mdoc.c, Revision 1.64

1.64    ! schwarze    1: /*     $Id: mdoc.c,v 1.63 2010/08/07 18:06:45 schwarze Exp $ */
1.1       kristaps    2: /*
1.61      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
1.3       schwarze    7:  * purpose with or without fee is hereby granted, provided that the above
                      8:  * copyright notice and this permission notice appear in all copies.
1.1       kristaps    9:  *
1.3       schwarze   10:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
                     11:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
                     12:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
                     13:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     14:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     15:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     16:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
1.1       kristaps   17:  */
1.28      schwarze   18: #include <sys/types.h>
                     19:
1.1       kristaps   20: #include <assert.h>
                     21: #include <stdarg.h>
                     22: #include <stdio.h>
                     23: #include <stdlib.h>
                     24: #include <string.h>
1.39      schwarze   25: #include <time.h>
1.1       kristaps   26:
1.54      schwarze   27: #include "mandoc.h"
1.1       kristaps   28: #include "libmdoc.h"
1.32      schwarze   29: #include "libmandoc.h"
1.1       kristaps   30:
                     31: const  char *const __mdoc_macronames[MDOC_MAX] = {
1.7       schwarze   32:        "Ap",           "Dd",           "Dt",           "Os",
1.1       kristaps   33:        "Sh",           "Ss",           "Pp",           "D1",
                     34:        "Dl",           "Bd",           "Ed",           "Bl",
                     35:        "El",           "It",           "Ad",           "An",
                     36:        "Ar",           "Cd",           "Cm",           "Dv",
                     37:        "Er",           "Ev",           "Ex",           "Fa",
                     38:        "Fd",           "Fl",           "Fn",           "Ft",
                     39:        "Ic",           "In",           "Li",           "Nd",
                     40:        "Nm",           "Op",           "Ot",           "Pa",
                     41:        "Rv",           "St",           "Va",           "Vt",
                     42:        /* LINTED */
1.33      schwarze   43:        "Xr",           "%A",           "%B",           "%D",
1.1       kristaps   44:        /* LINTED */
1.33      schwarze   45:        "%I",           "%J",           "%N",           "%O",
1.1       kristaps   46:        /* LINTED */
1.33      schwarze   47:        "%P",           "%R",           "%T",           "%V",
1.1       kristaps   48:        "Ac",           "Ao",           "Aq",           "At",
                     49:        "Bc",           "Bf",           "Bo",           "Bq",
                     50:        "Bsx",          "Bx",           "Db",           "Dc",
                     51:        "Do",           "Dq",           "Ec",           "Ef",
                     52:        "Em",           "Eo",           "Fx",           "Ms",
                     53:        "No",           "Ns",           "Nx",           "Ox",
                     54:        "Pc",           "Pf",           "Po",           "Pq",
                     55:        "Qc",           "Ql",           "Qo",           "Qq",
                     56:        "Re",           "Rs",           "Sc",           "So",
                     57:        "Sq",           "Sm",           "Sx",           "Sy",
                     58:        "Tn",           "Ux",           "Xc",           "Xo",
                     59:        "Fo",           "Fc",           "Oo",           "Oc",
                     60:        "Bk",           "Ek",           "Bt",           "Hf",
1.7       schwarze   61:        "Fr",           "Ud",           "Lb",           "Lp",
                     62:        "Lk",           "Mt",           "Brq",          "Bro",
1.1       kristaps   63:        /* LINTED */
1.33      schwarze   64:        "Brc",          "%C",           "Es",           "En",
1.1       kristaps   65:        /* LINTED */
1.33      schwarze   66:        "Dx",           "%Q",           "br",           "sp",
1.31      schwarze   67:        /* LINTED */
1.56      schwarze   68:        "%U",           "Ta"
1.1       kristaps   69:        };
                     70:
                     71: const  char *const __mdoc_argnames[MDOC_ARG_MAX] = {
                     72:        "split",                "nosplit",              "ragged",
                     73:        "unfilled",             "literal",              "file",
                     74:        "offset",               "bullet",               "dash",
                     75:        "hyphen",               "item",                 "enum",
                     76:        "tag",                  "diag",                 "hang",
                     77:        "ohang",                "inset",                "column",
                     78:        "width",                "compact",              "std",
                     79:        "filled",               "words",                "emphasis",
1.30      schwarze   80:        "symbolic",             "nested",               "centered"
1.1       kristaps   81:        };
                     82:
                     83: const  char * const *mdoc_macronames = __mdoc_macronames;
                     84: const  char * const *mdoc_argnames = __mdoc_argnames;
                     85:
1.40      schwarze   86: static void              mdoc_node_free(struct mdoc_node *);
                     87: static void              mdoc_node_unlink(struct mdoc *,
                     88:                                struct mdoc_node *);
1.1       kristaps   89: static void              mdoc_free1(struct mdoc *);
1.32      schwarze   90: static void              mdoc_alloc1(struct mdoc *);
1.1       kristaps   91: static struct mdoc_node *node_alloc(struct mdoc *, int, int,
1.37      schwarze   92:                                enum mdoct, enum mdoc_type);
1.1       kristaps   93: static int               node_append(struct mdoc *,
                     94:                                struct mdoc_node *);
1.53      schwarze   95: static int               mdoc_ptext(struct mdoc *, int, char *, int);
                     96: static int               mdoc_pmacro(struct mdoc *, int, char *, int);
                     97: static int               macrowarn(struct mdoc *, int,
                     98:                                const char *, int);
1.46      schwarze   99:
1.1       kristaps  100:
                    101: const struct mdoc_node *
                    102: mdoc_node(const struct mdoc *m)
                    103: {
                    104:
                    105:        return(MDOC_HALT & m->flags ? NULL : m->first);
                    106: }
                    107:
                    108:
                    109: const struct mdoc_meta *
                    110: mdoc_meta(const struct mdoc *m)
                    111: {
                    112:
                    113:        return(MDOC_HALT & m->flags ? NULL : &m->meta);
                    114: }
                    115:
                    116:
1.8       schwarze  117: /*
                    118:  * Frees volatile resources (parse tree, meta-data, fields).
                    119:  */
1.1       kristaps  120: static void
                    121: mdoc_free1(struct mdoc *mdoc)
                    122: {
                    123:
                    124:        if (mdoc->first)
1.40      schwarze  125:                mdoc_node_delete(mdoc, mdoc->first);
1.1       kristaps  126:        if (mdoc->meta.title)
                    127:                free(mdoc->meta.title);
                    128:        if (mdoc->meta.os)
                    129:                free(mdoc->meta.os);
                    130:        if (mdoc->meta.name)
                    131:                free(mdoc->meta.name);
                    132:        if (mdoc->meta.arch)
                    133:                free(mdoc->meta.arch);
                    134:        if (mdoc->meta.vol)
                    135:                free(mdoc->meta.vol);
1.50      schwarze  136:        if (mdoc->meta.msec)
                    137:                free(mdoc->meta.msec);
1.1       kristaps  138: }
                    139:
                    140:
1.8       schwarze  141: /*
                    142:  * Allocate all volatile resources (parse tree, meta-data, fields).
                    143:  */
1.32      schwarze  144: static void
1.1       kristaps  145: mdoc_alloc1(struct mdoc *mdoc)
                    146: {
                    147:
1.32      schwarze  148:        memset(&mdoc->meta, 0, sizeof(struct mdoc_meta));
1.1       kristaps  149:        mdoc->flags = 0;
1.9       schwarze  150:        mdoc->lastnamed = mdoc->lastsec = SEC_NONE;
1.32      schwarze  151:        mdoc->last = mandoc_calloc(1, sizeof(struct mdoc_node));
1.1       kristaps  152:        mdoc->first = mdoc->last;
                    153:        mdoc->last->type = MDOC_ROOT;
                    154:        mdoc->next = MDOC_NEXT_CHILD;
                    155: }
                    156:
                    157:
                    158: /*
1.8       schwarze  159:  * Free up volatile resources (see mdoc_free1()) then re-initialises the
                    160:  * data with mdoc_alloc1().  After invocation, parse data has been reset
                    161:  * and the parser is ready for re-invocation on a new tree; however,
                    162:  * cross-parse non-volatile data is kept intact.
1.1       kristaps  163:  */
1.32      schwarze  164: void
1.1       kristaps  165: mdoc_reset(struct mdoc *mdoc)
                    166: {
                    167:
                    168:        mdoc_free1(mdoc);
1.32      schwarze  169:        mdoc_alloc1(mdoc);
1.1       kristaps  170: }
                    171:
                    172:
                    173: /*
1.8       schwarze  174:  * Completely free up all volatile and non-volatile parse resources.
                    175:  * After invocation, the pointer is no longer usable.
1.1       kristaps  176:  */
                    177: void
                    178: mdoc_free(struct mdoc *mdoc)
                    179: {
                    180:
                    181:        mdoc_free1(mdoc);
                    182:        free(mdoc);
                    183: }
                    184:
                    185:
1.8       schwarze  186: /*
                    187:  * Allocate volatile and non-volatile parse resources.
                    188:  */
1.1       kristaps  189: struct mdoc *
1.58      schwarze  190: mdoc_alloc(struct regset *regs, void *data,
                    191:                int pflags, mandocmsg msg)
1.1       kristaps  192: {
                    193:        struct mdoc     *p;
                    194:
1.32      schwarze  195:        p = mandoc_calloc(1, sizeof(struct mdoc));
                    196:
1.54      schwarze  197:        p->msg = msg;
1.1       kristaps  198:        p->data = data;
                    199:        p->pflags = pflags;
1.58      schwarze  200:        p->regs = regs;
1.1       kristaps  201:
1.32      schwarze  202:        mdoc_hash_init();
                    203:        mdoc_alloc1(p);
                    204:        return(p);
1.1       kristaps  205: }
                    206:
                    207:
                    208: /*
                    209:  * Climb back up the parse tree, validating open scopes.  Mostly calls
1.8       schwarze  210:  * through to macro_end() in macro.c.
1.1       kristaps  211:  */
                    212: int
                    213: mdoc_endparse(struct mdoc *m)
                    214: {
                    215:
                    216:        if (MDOC_HALT & m->flags)
                    217:                return(0);
                    218:        else if (mdoc_macroend(m))
                    219:                return(1);
                    220:        m->flags |= MDOC_HALT;
                    221:        return(0);
                    222: }
                    223:
                    224:
                    225: /*
                    226:  * Main parse routine.  Parses a single line -- really just hands off to
1.45      schwarze  227:  * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
1.1       kristaps  228:  */
                    229: int
1.53      schwarze  230: mdoc_parseln(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  231: {
                    232:
                    233:        if (MDOC_HALT & m->flags)
                    234:                return(0);
                    235:
1.47      schwarze  236:        m->flags |= MDOC_NEWLINE;
1.60      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:
1.53      schwarze  251:        return(('.' == buf[offs] || '\'' == buf[offs]) ?
                    252:                        mdoc_pmacro(m, ln, buf, offs) :
                    253:                        mdoc_ptext(m, ln, buf, offs));
1.1       kristaps  254: }
                    255:
                    256:
                    257: int
1.54      schwarze  258: mdoc_vmsg(struct mdoc *mdoc, enum mandocerr t,
                    259:                int ln, int pos, const char *fmt, ...)
1.1       kristaps  260: {
                    261:        char             buf[256];
                    262:        va_list          ap;
                    263:
                    264:        va_start(ap, fmt);
1.54      schwarze  265:        vsnprintf(buf, sizeof(buf) - 1, fmt, ap);
1.1       kristaps  266:        va_end(ap);
1.17      schwarze  267:
1.54      schwarze  268:        return((*mdoc->msg)(t, mdoc->data, ln, pos, buf));
1.5       schwarze  269: }
                    270:
                    271:
                    272: int
1.58      schwarze  273: mdoc_macro(MACRO_PROT_ARGS)
1.1       kristaps  274: {
1.46      schwarze  275:        assert(tok < MDOC_MAX);
                    276:
                    277:        /* If we're in the body, deny prologue calls. */
1.37      schwarze  278:
1.1       kristaps  279:        if (MDOC_PROLOGUE & mdoc_macros[tok].flags &&
1.9       schwarze  280:                        MDOC_PBODY & m->flags)
1.58      schwarze  281:                return(mdoc_pmsg(m, line, ppos, MANDOCERR_BADBODY));
1.46      schwarze  282:
                    283:        /* If we're in the prologue, deny "body" macros.  */
                    284:
1.1       kristaps  285:        if ( ! (MDOC_PROLOGUE & mdoc_macros[tok].flags) &&
1.39      schwarze  286:                        ! (MDOC_PBODY & m->flags)) {
1.58      schwarze  287:                if ( ! mdoc_pmsg(m, line, ppos, MANDOCERR_BADPROLOG))
1.39      schwarze  288:                        return(0);
                    289:                if (NULL == m->meta.title)
1.56      schwarze  290:                        m->meta.title = mandoc_strdup("UNKNOWN");
1.39      schwarze  291:                if (NULL == m->meta.vol)
1.56      schwarze  292:                        m->meta.vol = mandoc_strdup("LOCAL");
1.39      schwarze  293:                if (NULL == m->meta.os)
1.56      schwarze  294:                        m->meta.os = mandoc_strdup("LOCAL");
1.39      schwarze  295:                if (0 == m->meta.date)
                    296:                        m->meta.date = time(NULL);
                    297:                m->flags |= MDOC_PBODY;
                    298:        }
1.1       kristaps  299:
1.58      schwarze  300:        return((*mdoc_macros[tok].fp)(m, tok, line, ppos, pos, buf));
1.1       kristaps  301: }
                    302:
                    303:
                    304: static int
                    305: node_append(struct mdoc *mdoc, struct mdoc_node *p)
                    306: {
                    307:
                    308:        assert(mdoc->last);
                    309:        assert(mdoc->first);
                    310:        assert(MDOC_ROOT != p->type);
                    311:
                    312:        switch (mdoc->next) {
                    313:        case (MDOC_NEXT_SIBLING):
                    314:                mdoc->last->next = p;
                    315:                p->prev = mdoc->last;
                    316:                p->parent = mdoc->last->parent;
                    317:                break;
                    318:        case (MDOC_NEXT_CHILD):
                    319:                mdoc->last->child = p;
                    320:                p->parent = mdoc->last;
                    321:                break;
                    322:        default:
                    323:                abort();
                    324:                /* NOTREACHED */
                    325:        }
                    326:
1.10      schwarze  327:        p->parent->nchild++;
                    328:
1.1       kristaps  329:        if ( ! mdoc_valid_pre(mdoc, p))
                    330:                return(0);
                    331:        if ( ! mdoc_action_pre(mdoc, p))
                    332:                return(0);
                    333:
                    334:        switch (p->type) {
                    335:        case (MDOC_HEAD):
                    336:                assert(MDOC_BLOCK == p->parent->type);
                    337:                p->parent->head = p;
                    338:                break;
                    339:        case (MDOC_TAIL):
                    340:                assert(MDOC_BLOCK == p->parent->type);
                    341:                p->parent->tail = p;
                    342:                break;
                    343:        case (MDOC_BODY):
1.59      schwarze  344:                if (p->end)
                    345:                        break;
1.1       kristaps  346:                assert(MDOC_BLOCK == p->parent->type);
                    347:                p->parent->body = p;
                    348:                break;
                    349:        default:
                    350:                break;
                    351:        }
                    352:
                    353:        mdoc->last = p;
                    354:
                    355:        switch (p->type) {
                    356:        case (MDOC_TEXT):
                    357:                if ( ! mdoc_valid_post(mdoc))
                    358:                        return(0);
                    359:                if ( ! mdoc_action_post(mdoc))
                    360:                        return(0);
                    361:                break;
                    362:        default:
                    363:                break;
                    364:        }
                    365:
                    366:        return(1);
                    367: }
                    368:
                    369:
                    370: static struct mdoc_node *
1.37      schwarze  371: node_alloc(struct mdoc *m, int line, int pos,
                    372:                enum mdoct tok, enum mdoc_type type)
1.1       kristaps  373: {
                    374:        struct mdoc_node *p;
                    375:
1.32      schwarze  376:        p = mandoc_calloc(1, sizeof(struct mdoc_node));
1.19      schwarze  377:        p->sec = m->lastsec;
1.1       kristaps  378:        p->line = line;
                    379:        p->pos = pos;
                    380:        p->tok = tok;
1.37      schwarze  381:        p->type = type;
1.58      schwarze  382:
                    383:        /* Flag analysis. */
                    384:
1.60      schwarze  385:        if (MDOC_SYNOPSIS & m->flags)
                    386:                p->flags |= MDOC_SYNPRETTY;
                    387:        else
                    388:                p->flags &= ~MDOC_SYNPRETTY;
1.47      schwarze  389:        if (MDOC_NEWLINE & m->flags)
                    390:                p->flags |= MDOC_LINE;
                    391:        m->flags &= ~MDOC_NEWLINE;
1.58      schwarze  392:
1.1       kristaps  393:        return(p);
                    394: }
                    395:
                    396:
                    397: int
1.37      schwarze  398: mdoc_tail_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  399: {
                    400:        struct mdoc_node *p;
                    401:
1.19      schwarze  402:        p = node_alloc(m, line, pos, tok, MDOC_TAIL);
1.25      schwarze  403:        if ( ! node_append(m, p))
                    404:                return(0);
                    405:        m->next = MDOC_NEXT_CHILD;
                    406:        return(1);
1.1       kristaps  407: }
                    408:
                    409:
                    410: int
1.37      schwarze  411: mdoc_head_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  412: {
                    413:        struct mdoc_node *p;
                    414:
1.19      schwarze  415:        assert(m->first);
                    416:        assert(m->last);
1.1       kristaps  417:
1.19      schwarze  418:        p = node_alloc(m, line, pos, tok, MDOC_HEAD);
1.25      schwarze  419:        if ( ! node_append(m, p))
                    420:                return(0);
                    421:        m->next = MDOC_NEXT_CHILD;
                    422:        return(1);
1.1       kristaps  423: }
                    424:
                    425:
                    426: int
1.37      schwarze  427: mdoc_body_alloc(struct mdoc *m, int line, int pos, enum mdoct tok)
1.1       kristaps  428: {
                    429:        struct mdoc_node *p;
                    430:
1.19      schwarze  431:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
1.25      schwarze  432:        if ( ! node_append(m, p))
                    433:                return(0);
                    434:        m->next = MDOC_NEXT_CHILD;
1.59      schwarze  435:        return(1);
                    436: }
                    437:
                    438:
                    439: int
                    440: mdoc_endbody_alloc(struct mdoc *m, int line, int pos, enum mdoct tok,
                    441:                struct mdoc_node *body, enum mdoc_endbody end)
                    442: {
                    443:        struct mdoc_node *p;
                    444:
                    445:        p = node_alloc(m, line, pos, tok, MDOC_BODY);
                    446:        p->pending = body;
                    447:        p->end = end;
                    448:        if ( ! node_append(m, p))
                    449:                return(0);
                    450:        m->next = MDOC_NEXT_SIBLING;
1.25      schwarze  451:        return(1);
1.1       kristaps  452: }
                    453:
                    454:
                    455: int
1.19      schwarze  456: mdoc_block_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  457:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  458: {
                    459:        struct mdoc_node *p;
                    460:
1.19      schwarze  461:        p = node_alloc(m, line, pos, tok, MDOC_BLOCK);
1.4       schwarze  462:        p->args = args;
                    463:        if (p->args)
1.1       kristaps  464:                (args->refcnt)++;
1.25      schwarze  465:        if ( ! node_append(m, p))
                    466:                return(0);
                    467:        m->next = MDOC_NEXT_CHILD;
                    468:        return(1);
1.1       kristaps  469: }
                    470:
                    471:
                    472: int
1.19      schwarze  473: mdoc_elem_alloc(struct mdoc *m, int line, int pos,
1.37      schwarze  474:                enum mdoct tok, struct mdoc_arg *args)
1.1       kristaps  475: {
                    476:        struct mdoc_node *p;
                    477:
1.19      schwarze  478:        p = node_alloc(m, line, pos, tok, MDOC_ELEM);
1.4       schwarze  479:        p->args = args;
                    480:        if (p->args)
1.1       kristaps  481:                (args->refcnt)++;
1.25      schwarze  482:        if ( ! node_append(m, p))
                    483:                return(0);
                    484:        m->next = MDOC_NEXT_CHILD;
                    485:        return(1);
1.1       kristaps  486: }
                    487:
                    488:
1.46      schwarze  489: int
                    490: mdoc_word_alloc(struct mdoc *m, int line, int pos, const char *p)
1.1       kristaps  491: {
1.19      schwarze  492:        struct mdoc_node *n;
1.46      schwarze  493:        size_t            sv, len;
                    494:
                    495:        len = strlen(p);
1.1       kristaps  496:
1.46      schwarze  497:        n = node_alloc(m, line, pos, MDOC_MAX, MDOC_TEXT);
1.32      schwarze  498:        n->string = mandoc_malloc(len + 1);
1.19      schwarze  499:        sv = strlcpy(n->string, p, len + 1);
                    500:
                    501:        /* Prohibit truncation. */
                    502:        assert(sv < len + 1);
                    503:
1.25      schwarze  504:        if ( ! node_append(m, n))
                    505:                return(0);
1.46      schwarze  506:
1.25      schwarze  507:        m->next = MDOC_NEXT_SIBLING;
                    508:        return(1);
1.19      schwarze  509: }
                    510:
                    511:
1.61      schwarze  512: static void
1.1       kristaps  513: mdoc_node_free(struct mdoc_node *p)
                    514: {
                    515:
1.61      schwarze  516:        /*
                    517:         * XXX: if these end up being problematic in terms of memory
                    518:         * management and dereferencing freed blocks, then make them
                    519:         * into reference-counted double-pointers.
                    520:         */
                    521:
                    522:        if (MDOC_Bd == p->tok && MDOC_BLOCK == p->type)
                    523:                if (p->data.Bd)
                    524:                        free(p->data.Bd);
                    525:        if (MDOC_Bl == p->tok && MDOC_BLOCK == p->type)
                    526:                if (p->data.Bl)
                    527:                        free(p->data.Bl);
                    528:        if (MDOC_Bf == p->tok && MDOC_HEAD == p->type)
                    529:                if (p->data.Bf)
                    530:                        free(p->data.Bf);
                    531:
1.1       kristaps  532:        if (p->string)
                    533:                free(p->string);
                    534:        if (p->args)
                    535:                mdoc_argv_free(p->args);
                    536:        free(p);
                    537: }
                    538:
                    539:
1.40      schwarze  540: static void
                    541: mdoc_node_unlink(struct mdoc *m, struct mdoc_node *n)
                    542: {
                    543:
                    544:        /* Adjust siblings. */
                    545:
                    546:        if (n->prev)
                    547:                n->prev->next = n->next;
                    548:        if (n->next)
                    549:                n->next->prev = n->prev;
                    550:
                    551:        /* Adjust parent. */
                    552:
                    553:        if (n->parent) {
                    554:                n->parent->nchild--;
                    555:                if (n->parent->child == n)
                    556:                        n->parent->child = n->prev ? n->prev : n->next;
                    557:        }
                    558:
                    559:        /* Adjust parse point, if applicable. */
                    560:
                    561:        if (m && m->last == n) {
                    562:                if (n->prev) {
                    563:                        m->last = n->prev;
                    564:                        m->next = MDOC_NEXT_SIBLING;
                    565:                } else {
                    566:                        m->last = n->parent;
                    567:                        m->next = MDOC_NEXT_CHILD;
                    568:                }
                    569:        }
                    570:
                    571:        if (m && m->first == n)
                    572:                m->first = NULL;
                    573: }
                    574:
                    575:
1.1       kristaps  576: void
1.40      schwarze  577: mdoc_node_delete(struct mdoc *m, struct mdoc_node *p)
1.1       kristaps  578: {
                    579:
1.40      schwarze  580:        while (p->child) {
                    581:                assert(p->nchild);
                    582:                mdoc_node_delete(m, p->child);
                    583:        }
                    584:        assert(0 == p->nchild);
1.1       kristaps  585:
1.40      schwarze  586:        mdoc_node_unlink(m, p);
1.1       kristaps  587:        mdoc_node_free(p);
                    588: }
                    589:
                    590:
                    591: /*
                    592:  * Parse free-form text, that is, a line that does not begin with the
                    593:  * control character.
                    594:  */
                    595: static int
1.53      schwarze  596: mdoc_ptext(struct mdoc *m, int line, char *buf, int offs)
1.1       kristaps  597: {
1.56      schwarze  598:        char             *c, *ws, *end;
                    599:        struct mdoc_node *n;
1.44      schwarze  600:
                    601:        /* Ignore bogus comments. */
                    602:
1.53      schwarze  603:        if ('\\' == buf[offs] &&
                    604:                        '.' == buf[offs + 1] &&
                    605:                        '"' == buf[offs + 2])
1.54      schwarze  606:                return(mdoc_pmsg(m, line, offs, MANDOCERR_BADCOMMENT));
1.1       kristaps  607:
1.46      schwarze  608:        /* No text before an initial macro. */
                    609:
1.9       schwarze  610:        if (SEC_NONE == m->lastnamed)
1.54      schwarze  611:                return(mdoc_pmsg(m, line, offs, MANDOCERR_NOTEXT));
1.46      schwarze  612:
1.56      schwarze  613:        assert(m->last);
                    614:        n = m->last;
                    615:
                    616:        /*
                    617:         * Divert directly to list processing if we're encountering a
                    618:         * columnar MDOC_BLOCK with or without a prior MDOC_BLOCK entry
                    619:         * (a MDOC_BODY means it's already open, in which case we should
                    620:         * process within its context in the normal way).
                    621:         */
                    622:
                    623:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.61      schwarze  624:                        LIST_column == n->data.Bl->type) {
1.56      schwarze  625:                /* `Bl' is open without any children. */
                    626:                m->flags |= MDOC_FREECOL;
                    627:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    628:        }
                    629:
                    630:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    631:                        NULL != n->parent &&
                    632:                        MDOC_Bl == n->parent->tok &&
1.61      schwarze  633:                        LIST_column == n->parent->data.Bl->type) {
1.56      schwarze  634:                /* `Bl' has block-level `It' children. */
                    635:                m->flags |= MDOC_FREECOL;
                    636:                return(mdoc_macro(m, MDOC_It, line, offs, &offs, buf));
                    637:        }
                    638:
1.52      schwarze  639:        /*
                    640:         * Search for the beginning of unescaped trailing whitespace (ws)
                    641:         * and for the first character not to be output (end).
                    642:         */
1.56      schwarze  643:
                    644:        /* FIXME: replace with strcspn(). */
1.52      schwarze  645:        ws = NULL;
1.53      schwarze  646:        for (c = end = buf + offs; *c; c++) {
1.52      schwarze  647:                switch (*c) {
1.55      schwarze  648:                case '-':
                    649:                        if (mandoc_hyph(buf + offs, c))
                    650:                                *c = ASCII_HYPH;
1.56      schwarze  651:                        ws = NULL;
1.55      schwarze  652:                        break;
1.52      schwarze  653:                case ' ':
                    654:                        if (NULL == ws)
                    655:                                ws = c;
                    656:                        continue;
                    657:                case '\t':
                    658:                        /*
                    659:                         * Always warn about trailing tabs,
                    660:                         * even outside literal context,
                    661:                         * where they should be put on the next line.
                    662:                         */
                    663:                        if (NULL == ws)
                    664:                                ws = c;
                    665:                        /*
                    666:                         * Strip trailing tabs in literal context only;
                    667:                         * outside, they affect the next line.
                    668:                         */
                    669:                        if (MDOC_LITERAL & m->flags)
                    670:                                continue;
                    671:                        break;
                    672:                case '\\':
                    673:                        /* Skip the escaped character, too, if any. */
                    674:                        if (c[1])
                    675:                                c++;
                    676:                        /* FALLTHROUGH */
                    677:                default:
                    678:                        ws = NULL;
                    679:                        break;
                    680:                }
                    681:                end = c + 1;
                    682:        }
                    683:        *end = '\0';
1.19      schwarze  684:
1.52      schwarze  685:        if (ws)
1.54      schwarze  686:                if ( ! mdoc_pmsg(m, line, (int)(ws-buf), MANDOCERR_EOLNSPACE))
1.52      schwarze  687:                        return(0);
1.34      schwarze  688:
1.53      schwarze  689:        if ('\0' == buf[offs] && ! (MDOC_LITERAL & m->flags)) {
1.54      schwarze  690:                if ( ! mdoc_pmsg(m, line, (int)(c-buf), MANDOCERR_NOBLANKLN))
1.38      schwarze  691:                        return(0);
1.46      schwarze  692:
1.40      schwarze  693:                /*
1.46      schwarze  694:                 * Insert a `Pp' in the case of a blank line.  Technically,
                    695:                 * blank lines aren't allowed, but enough manuals assume this
                    696:                 * behaviour that we want to work around it.
1.40      schwarze  697:                 */
1.53      schwarze  698:                if ( ! mdoc_elem_alloc(m, line, offs, MDOC_Pp, NULL))
1.42      schwarze  699:                        return(0);
1.46      schwarze  700:
1.42      schwarze  701:                m->next = MDOC_NEXT_SIBLING;
                    702:                return(1);
1.38      schwarze  703:        }
1.1       kristaps  704:
1.53      schwarze  705:        if ( ! mdoc_word_alloc(m, line, offs, buf+offs))
1.1       kristaps  706:                return(0);
                    707:
1.52      schwarze  708:        if (MDOC_LITERAL & m->flags)
                    709:                return(1);
                    710:
1.35      schwarze  711:        /*
1.48      schwarze  712:         * End-of-sentence check.  If the last character is an unescaped
                    713:         * EOS character, then flag the node as being the end of a
                    714:         * sentence.  The front-end will know how to interpret this.
1.35      schwarze  715:         */
                    716:
1.52      schwarze  717:        assert(buf < end);
1.48      schwarze  718:
1.62      schwarze  719:        if (mandoc_eos(buf+offs, (size_t)(end-buf-offs), 0))
1.48      schwarze  720:                m->last->flags |= MDOC_EOS;
                    721:
1.1       kristaps  722:        return(1);
                    723: }
1.19      schwarze  724:
1.1       kristaps  725:
                    726: static int
1.53      schwarze  727: macrowarn(struct mdoc *m, int ln, const char *buf, int offs)
1.1       kristaps  728: {
1.54      schwarze  729:        int              rc;
                    730:
                    731:        rc = mdoc_vmsg(m, MANDOCERR_MACRO, ln, offs,
                    732:                        "unknown macro: %s%s",
                    733:                        buf, strlen(buf) > 3 ? "..." : "");
                    734:
                    735:        /* FIXME: logic should be in driver. */
1.56      schwarze  736:        /* FIXME: broken, will error out and not omit a message. */
1.54      schwarze  737:        return(MDOC_IGN_MACRO & m->pflags ? rc : 0);
1.1       kristaps  738: }
                    739:
                    740:
                    741: /*
                    742:  * Parse a macro line, that is, a line beginning with the control
                    743:  * character.
                    744:  */
1.61      schwarze  745: static int
1.53      schwarze  746: mdoc_pmacro(struct mdoc *m, int ln, char *buf, int offs)
1.1       kristaps  747: {
1.56      schwarze  748:        enum mdoct        tok;
                    749:        int               i, j, sv;
                    750:        char              mac[5];
                    751:        struct mdoc_node *n;
1.1       kristaps  752:
1.7       schwarze  753:        /* Empty lines are ignored. */
1.1       kristaps  754:
1.53      schwarze  755:        offs++;
                    756:
                    757:        if ('\0' == buf[offs])
1.1       kristaps  758:                return(1);
                    759:
1.53      schwarze  760:        i = offs;
1.24      schwarze  761:
1.63      schwarze  762:        /* Accept tabs/whitespace after the initial control char. */
1.24      schwarze  763:
1.63      schwarze  764:        if (' ' == buf[i] || '\t' == buf[i]) {
1.24      schwarze  765:                i++;
1.63      schwarze  766:                while (buf[i] && (' ' == buf[i] || '\t' == buf[i]))
1.1       kristaps  767:                        i++;
1.34      schwarze  768:                if ('\0' == buf[i])
1.1       kristaps  769:                        return(1);
                    770:        }
                    771:
1.47      schwarze  772:        sv = i;
                    773:
1.63      schwarze  774:        /*
1.64    ! schwarze  775:         * Copy the first word into a nil-terminated buffer.
        !           776:         * Stop copying when a tab, space, or eoln is encountered.
1.63      schwarze  777:         */
1.1       kristaps  778:
1.64    ! schwarze  779:        j = 0;
        !           780:        while (j < 4 && '\0' != buf[i] && ' ' != buf[i] && '\t' != buf[i])
        !           781:                mac[j++] = buf[i++];
1.53      schwarze  782:        mac[j] = '\0';
1.1       kristaps  783:
1.24      schwarze  784:        if (j == 4 || j < 2) {
1.53      schwarze  785:                if ( ! macrowarn(m, ln, mac, sv))
1.1       kristaps  786:                        goto err;
                    787:                return(1);
                    788:        }
                    789:
1.46      schwarze  790:        if (MDOC_MAX == (tok = mdoc_hash_find(mac))) {
1.53      schwarze  791:                if ( ! macrowarn(m, ln, mac, sv))
1.1       kristaps  792:                        goto err;
                    793:                return(1);
                    794:        }
                    795:
1.63      schwarze  796:        /* Disregard the first trailing tab, if applicable. */
                    797:
                    798:        if ('\t' == buf[i])
                    799:                i++;
                    800:
                    801:        /* Jump to the next non-whitespace word. */
1.1       kristaps  802:
                    803:        while (buf[i] && ' ' == buf[i])
                    804:                i++;
1.34      schwarze  805:
1.46      schwarze  806:        /*
                    807:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    808:         * into the parser as "text", so we only warn about spaces here.
                    809:         */
1.34      schwarze  810:
                    811:        if ('\0' == buf[i] && ' ' == buf[i - 1])
1.54      schwarze  812:                if ( ! mdoc_pmsg(m, ln, i - 1, MANDOCERR_EOLNSPACE))
1.34      schwarze  813:                        goto err;
1.1       kristaps  814:
1.56      schwarze  815:        /*
                    816:         * If an initial macro or a list invocation, divert directly
                    817:         * into macro processing.
                    818:         */
                    819:
                    820:        if (NULL == m->last || MDOC_It == tok || MDOC_El == tok) {
                    821:                if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
                    822:                        goto err;
                    823:                return(1);
                    824:        }
                    825:
                    826:        n = m->last;
                    827:        assert(m->last);
                    828:
                    829:        /*
                    830:         * If the first macro of a `Bl -column', open an `It' block
                    831:         * context around the parsed macro.
                    832:         */
                    833:
                    834:        if (MDOC_Bl == n->tok && MDOC_BODY == n->type &&
1.61      schwarze  835:                        LIST_column == n->data.Bl->type) {
1.56      schwarze  836:                m->flags |= MDOC_FREECOL;
1.58      schwarze  837:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
1.56      schwarze  838:                        goto err;
                    839:                return(1);
                    840:        }
                    841:
                    842:        /*
                    843:         * If we're following a block-level `It' within a `Bl -column'
                    844:         * context (perhaps opened in the above block or in ptext()),
                    845:         * then open an `It' block context around the parsed macro.
1.24      schwarze  846:         */
1.56      schwarze  847:
                    848:        if (MDOC_It == n->tok && MDOC_BLOCK == n->type &&
                    849:                        NULL != n->parent &&
                    850:                        MDOC_Bl == n->parent->tok &&
1.61      schwarze  851:                        LIST_column == n->parent->data.Bl->type) {
1.56      schwarze  852:                m->flags |= MDOC_FREECOL;
                    853:                if ( ! mdoc_macro(m, MDOC_It, ln, sv, &sv, buf))
                    854:                        goto err;
                    855:                return(1);
                    856:        }
                    857:
                    858:        /* Normal processing of a macro. */
                    859:
1.47      schwarze  860:        if ( ! mdoc_macro(m, tok, ln, sv, &i, buf))
1.1       kristaps  861:                goto err;
                    862:
                    863:        return(1);
                    864:
                    865: err:   /* Error out. */
                    866:
                    867:        m->flags |= MDOC_HALT;
                    868:        return(0);
                    869: }
1.24      schwarze  870:
                    871: