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

Annotation of src/usr.bin/mandoc/man.c, Revision 1.108

1.108   ! schwarze    1: /*     $OpenBSD: man.c,v 1.107 2015/04/19 13:59:37 schwarze Exp $ */
1.1       kristaps    2: /*
1.52      schwarze    3:  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.97      schwarze    4:  * Copyright (c) 2013, 2014, 2015 Ingo Schwarze <schwarze@openbsd.org>
1.74      schwarze    5:  * Copyright (c) 2011 Joerg Sonnenberger <joerg@netbsd.org>
1.1       kristaps    6:  *
                      7:  * Permission to use, copy, modify, and distribute this software for any
1.2       schwarze    8:  * purpose with or without fee is hereby granted, provided that the above
                      9:  * copyright notice and this permission notice appear in all copies.
1.1       kristaps   10:  *
1.98      schwarze   11:  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.2       schwarze   12:  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.98      schwarze   13:  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.2       schwarze   14:  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
                     15:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
                     16:  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
                     17:  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
1.1       kristaps   18:  */
1.14      schwarze   19: #include <sys/types.h>
                     20:
1.1       kristaps   21: #include <assert.h>
1.76      schwarze   22: #include <ctype.h>
1.1       kristaps   23: #include <stdarg.h>
                     24: #include <stdlib.h>
                     25: #include <stdio.h>
                     26: #include <string.h>
                     27:
1.98      schwarze   28: #include "mandoc_aux.h"
                     29: #include "mandoc.h"
                     30: #include "roff.h"
1.59      schwarze   31: #include "man.h"
1.98      schwarze   32: #include "libmandoc.h"
1.106     schwarze   33: #include "roff_int.h"
1.1       kristaps   34: #include "libman.h"
                     35:
1.79      schwarze   36: const  char *const __man_macronames[MAN_MAX] = {
1.3       schwarze   37:        "br",           "TH",           "SH",           "SS",
1.79      schwarze   38:        "TP",           "LP",           "PP",           "P",
1.1       kristaps   39:        "IP",           "HP",           "SM",           "SB",
                     40:        "BI",           "IB",           "BR",           "RB",
                     41:        "R",            "B",            "I",            "IR",
1.96      schwarze   42:        "RI",           "sp",           "nf",
1.50      schwarze   43:        "fi",           "RE",           "RS",           "DT",
                     44:        "UC",           "PD",           "AT",           "in",
1.70      schwarze   45:        "ft",           "OP",           "EX",           "EE",
1.78      schwarze   46:        "UR",           "UE",           "ll"
1.1       kristaps   47:        };
                     48:
                     49: const  char * const *man_macronames = __man_macronames;
                     50:
1.101     schwarze   51: static void             man_descope(struct roff_man *, int, int);
                     52: static int              man_ptext(struct roff_man *, int, char *, int);
                     53: static int              man_pmacro(struct roff_man *, int, char *, int);
1.1       kristaps   54:
                     55:
1.16      schwarze   56: void
1.101     schwarze   57: man_endparse(struct roff_man *man)
1.1       kristaps   58: {
                     59:
1.92      schwarze   60:        man_macroend(man);
1.1       kristaps   61: }
                     62:
                     63: int
1.101     schwarze   64: man_parseln(struct roff_man *man, int ln, char *buf, int offs)
1.1       kristaps   65: {
1.25      schwarze   66:
1.98      schwarze   67:        if (man->last->type != ROFFT_EQN || ln > man->last->line)
1.89      schwarze   68:                man->flags |= MAN_NEWLINE;
1.54      schwarze   69:
1.69      schwarze   70:        return (roff_getcontrol(man->roff, buf, &offs) ?
1.79      schwarze   71:            man_pmacro(man, ln, buf, offs) :
                     72:            man_ptext(man, ln, buf, offs));
1.1       kristaps   73: }
                     74:
1.92      schwarze   75: void
1.101     schwarze   76: man_elem_alloc(struct roff_man *man, int line, int pos, int tok)
1.1       kristaps   77: {
1.99      schwarze   78:        struct roff_node *p;
1.1       kristaps   79:
1.106     schwarze   80:        p = roff_node_alloc(man, line, pos, ROFFT_ELEM, tok);
                     81:        roff_node_append(man, p);
1.101     schwarze   82:        man->next = ROFF_NEXT_CHILD;
1.9       schwarze   83: }
                     84:
1.92      schwarze   85: void
1.101     schwarze   86: man_block_alloc(struct roff_man *man, int line, int pos, int tok)
1.9       schwarze   87: {
1.99      schwarze   88:        struct roff_node *p;
1.9       schwarze   89:
1.106     schwarze   90:        p = roff_node_alloc(man, line, pos, ROFFT_BLOCK, tok);
                     91:        roff_node_append(man, p);
1.101     schwarze   92:        man->next = ROFF_NEXT_CHILD;
1.9       schwarze   93: }
                     94:
1.92      schwarze   95: static void
1.101     schwarze   96: man_descope(struct roff_man *man, int line, int offs)
1.52      schwarze   97: {
                     98:        /*
                     99:         * Co-ordinate what happens with having a next-line scope open:
                    100:         * first close out the element scope (if applicable), then close
                    101:         * out the block scope (also if applicable).
                    102:         */
                    103:
1.92      schwarze  104:        if (man->flags & MAN_ELINE) {
1.69      schwarze  105:                man->flags &= ~MAN_ELINE;
1.92      schwarze  106:                man_unscope(man, man->last->parent);
1.52      schwarze  107:        }
1.92      schwarze  108:        if ( ! (man->flags & MAN_BLINE))
                    109:                return;
1.69      schwarze  110:        man->flags &= ~MAN_BLINE;
1.92      schwarze  111:        man_unscope(man, man->last->parent);
1.106     schwarze  112:        roff_body_alloc(man, line, offs, man->last->tok);
1.52      schwarze  113: }
                    114:
1.1       kristaps  115: static int
1.101     schwarze  116: man_ptext(struct roff_man *man, int line, char *buf, int offs)
1.1       kristaps  117: {
1.27      schwarze  118:        int              i;
1.26      schwarze  119:
1.10      schwarze  120:        /* Literal free-form text whitespace is preserved. */
                    121:
1.92      schwarze  122:        if (man->flags & MAN_LITERAL) {
1.108   ! schwarze  123:                roff_word_alloc(man, line, offs, buf + offs);
1.92      schwarze  124:                man_descope(man, line, offs);
                    125:                return(1);
1.10      schwarze  126:        }
                    127:
1.92      schwarze  128:        for (i = offs; buf[i] == ' '; i++)
1.10      schwarze  129:                /* Skip leading whitespace. */ ;
1.18      schwarze  130:
1.71      schwarze  131:        /*
                    132:         * Blank lines are ignored right after headings
                    133:         * but add a single vertical space elsewhere.
                    134:         */
                    135:
1.92      schwarze  136:        if (buf[i] == '\0') {
1.27      schwarze  137:                /* Allocate a blank entry. */
1.92      schwarze  138:                if (man->last->tok != MAN_SH &&
                    139:                    man->last->tok != MAN_SS) {
                    140:                        man_elem_alloc(man, line, offs, MAN_sp);
1.101     schwarze  141:                        man->next = ROFF_NEXT_SIBLING;
1.71      schwarze  142:                }
1.68      schwarze  143:                return(1);
1.10      schwarze  144:        }
1.1       kristaps  145:
1.79      schwarze  146:        /*
1.27      schwarze  147:         * Warn if the last un-escaped character is whitespace. Then
1.79      schwarze  148:         * strip away the remaining spaces (tabs stay!).
1.27      schwarze  149:         */
1.18      schwarze  150:
1.27      schwarze  151:        i = (int)strlen(buf);
                    152:        assert(i);
1.18      schwarze  153:
1.27      schwarze  154:        if (' ' == buf[i - 1] || '\t' == buf[i - 1]) {
                    155:                if (i > 1 && '\\' != buf[i - 2])
1.81      schwarze  156:                        mandoc_msg(MANDOCERR_SPACE_EOL, man->parse,
                    157:                            line, i - 1, NULL);
1.18      schwarze  158:
1.27      schwarze  159:                for (--i; i && ' ' == buf[i]; i--)
                    160:                        /* Spin back to non-space. */ ;
1.10      schwarze  161:
1.27      schwarze  162:                /* Jump ahead of escaped whitespace. */
                    163:                i += '\\' == buf[i] ? 2 : 1;
1.18      schwarze  164:
1.27      schwarze  165:                buf[i] = '\0';
1.10      schwarze  166:        }
1.108   ! schwarze  167:        roff_word_alloc(man, line, offs, buf + offs);
1.28      schwarze  168:
                    169:        /*
                    170:         * End-of-sentence check.  If the last character is an unescaped
                    171:         * EOS character, then flag the node as being the end of a
                    172:         * sentence.  The front-end will know how to interpret this.
                    173:         */
                    174:
                    175:        assert(i);
1.72      schwarze  176:        if (mandoc_eos(buf, (size_t)i))
1.69      schwarze  177:                man->last->flags |= MAN_EOS;
1.10      schwarze  178:
1.92      schwarze  179:        man_descope(man, line, offs);
                    180:        return(1);
1.1       kristaps  181: }
                    182:
1.53      schwarze  183: static int
1.101     schwarze  184: man_pmacro(struct roff_man *man, int ln, char *buf, int offs)
1.1       kristaps  185: {
1.99      schwarze  186:        struct roff_node *n;
1.91      schwarze  187:        const char      *cp;
1.99      schwarze  188:        int              tok;
1.84      schwarze  189:        int              i, ppos;
                    190:        int              bline;
1.91      schwarze  191:        char             mac[5];
1.1       kristaps  192:
1.59      schwarze  193:        ppos = offs;
1.1       kristaps  194:
1.23      schwarze  195:        /*
1.59      schwarze  196:         * Copy the first word into a nil-terminated buffer.
1.91      schwarze  197:         * Stop when a space, tab, escape, or eoln is encountered.
1.23      schwarze  198:         */
1.27      schwarze  199:
1.59      schwarze  200:        i = 0;
1.91      schwarze  201:        while (i < 4 && strchr(" \t\\", buf[offs]) == NULL)
1.59      schwarze  202:                mac[i++] = buf[offs++];
1.1       kristaps  203:
1.59      schwarze  204:        mac[i] = '\0';
1.1       kristaps  205:
1.107     schwarze  206:        tok = (i > 0 && i < 4) ? man_hash_find(mac) : TOKEN_NONE;
1.1       kristaps  207:
1.107     schwarze  208:        if (tok == TOKEN_NONE) {
1.85      schwarze  209:                mandoc_msg(MANDOCERR_MACRO, man->parse,
                    210:                    ln, ppos, buf + ppos - 1);
1.1       kristaps  211:                return(1);
                    212:        }
                    213:
1.91      schwarze  214:        /* Skip a leading escape sequence or tab. */
                    215:
                    216:        switch (buf[offs]) {
                    217:        case '\\':
                    218:                cp = buf + offs + 1;
                    219:                mandoc_escape(&cp, NULL, NULL);
                    220:                offs = cp - buf;
                    221:                break;
                    222:        case '\t':
                    223:                offs++;
                    224:                break;
                    225:        default:
                    226:                break;
                    227:        }
                    228:
                    229:        /* Jump to the next non-whitespace word. */
1.1       kristaps  230:
1.92      schwarze  231:        while (buf[offs] && buf[offs] == ' ')
1.59      schwarze  232:                offs++;
1.18      schwarze  233:
1.79      schwarze  234:        /*
1.27      schwarze  235:         * Trailing whitespace.  Note that tabs are allowed to be passed
                    236:         * into the parser as "text", so we only warn about spaces here.
                    237:         */
1.18      schwarze  238:
1.92      schwarze  239:        if (buf[offs] == '\0' && buf[offs - 1] == ' ')
1.81      schwarze  240:                mandoc_msg(MANDOCERR_SPACE_EOL, man->parse,
                    241:                    ln, offs - 1, NULL);
1.1       kristaps  242:
1.79      schwarze  243:        /*
1.97      schwarze  244:         * Some macros break next-line scopes; otherwise, remember
                    245:         * whether we are in next-line scope for a block head.
1.21      schwarze  246:         */
                    247:
1.97      schwarze  248:        man_breakscope(man, tok);
1.84      schwarze  249:        bline = man->flags & MAN_BLINE;
1.24      schwarze  250:
                    251:        /* Call to handler... */
1.1       kristaps  252:
1.22      schwarze  253:        assert(man_macros[tok].fp);
1.92      schwarze  254:        (*man_macros[tok].fp)(man, tok, ln, ppos, &offs, buf);
1.73      schwarze  255:
                    256:        /* In quick mode (for mandocdb), abort after the NAME section. */
                    257:
1.92      schwarze  258:        if (man->quick && tok == MAN_SH) {
1.80      schwarze  259:                n = man->last;
1.98      schwarze  260:                if (n->type == ROFFT_BODY &&
1.80      schwarze  261:                    strcmp(n->prev->child->string, "NAME"))
                    262:                        return(2);
                    263:        }
1.1       kristaps  264:
1.79      schwarze  265:        /*
1.84      schwarze  266:         * If we are in a next-line scope for a block head,
                    267:         * close it out now and switch to the body,
                    268:         * unless the next-line scope is allowed to continue.
1.21      schwarze  269:         */
                    270:
1.84      schwarze  271:        if ( ! bline || man->flags & MAN_ELINE ||
                    272:            man_macros[tok].flags & MAN_NSCOPED)
1.21      schwarze  273:                return(1);
1.1       kristaps  274:
1.92      schwarze  275:        assert(man->flags & MAN_BLINE);
1.69      schwarze  276:        man->flags &= ~MAN_BLINE;
1.1       kristaps  277:
1.92      schwarze  278:        man_unscope(man, man->last->parent);
1.106     schwarze  279:        roff_body_alloc(man, ln, ppos, man->last->tok);
1.92      schwarze  280:        return(1);
1.97      schwarze  281: }
                    282:
                    283: void
1.101     schwarze  284: man_breakscope(struct roff_man *man, int tok)
1.97      schwarze  285: {
1.99      schwarze  286:        struct roff_node *n;
1.97      schwarze  287:
                    288:        /*
                    289:         * An element next line scope is open,
                    290:         * and the new macro is not allowed inside elements.
                    291:         * Delete the element that is being broken.
                    292:         */
                    293:
1.107     schwarze  294:        if (man->flags & MAN_ELINE && (tok == TOKEN_NONE ||
1.97      schwarze  295:            ! (man_macros[tok].flags & MAN_NSCOPED))) {
                    296:                n = man->last;
1.98      schwarze  297:                assert(n->type != ROFFT_TEXT);
1.97      schwarze  298:                if (man_macros[n->tok].flags & MAN_NSCOPED)
                    299:                        n = n->parent;
                    300:
                    301:                mandoc_vmsg(MANDOCERR_BLK_LINE, man->parse,
                    302:                    n->line, n->pos, "%s breaks %s",
1.107     schwarze  303:                    tok == TOKEN_NONE ? "TS" : man_macronames[tok],
1.97      schwarze  304:                    man_macronames[n->tok]);
                    305:
1.106     schwarze  306:                roff_node_delete(man, n);
1.97      schwarze  307:                man->flags &= ~MAN_ELINE;
                    308:        }
                    309:
                    310:        /*
                    311:         * A block header next line scope is open,
                    312:         * and the new macro is not allowed inside block headers.
                    313:         * Delete the block that is being broken.
                    314:         */
                    315:
1.107     schwarze  316:        if (man->flags & MAN_BLINE && (tok == TOKEN_NONE ||
1.97      schwarze  317:            man_macros[tok].flags & MAN_BSCOPE)) {
                    318:                n = man->last;
1.98      schwarze  319:                if (n->type == ROFFT_TEXT)
1.97      schwarze  320:                        n = n->parent;
                    321:                if ( ! (man_macros[n->tok].flags & MAN_BSCOPE))
                    322:                        n = n->parent;
                    323:
1.98      schwarze  324:                assert(n->type == ROFFT_HEAD);
1.97      schwarze  325:                n = n->parent;
1.98      schwarze  326:                assert(n->type == ROFFT_BLOCK);
1.97      schwarze  327:                assert(man_macros[n->tok].flags & MAN_SCOPED);
                    328:
                    329:                mandoc_vmsg(MANDOCERR_BLK_LINE, man->parse,
                    330:                    n->line, n->pos, "%s breaks %s",
1.107     schwarze  331:                    tok == TOKEN_NONE ? "TS" : man_macronames[tok],
1.97      schwarze  332:                    man_macronames[n->tok]);
                    333:
1.106     schwarze  334:                roff_node_delete(man, n);
1.97      schwarze  335:                man->flags &= ~MAN_BLINE;
                    336:        }
1.62      schwarze  337: }
                    338:
                    339: const struct mparse *
1.101     schwarze  340: man_mparse(const struct roff_man *man)
1.62      schwarze  341: {
                    342:
1.69      schwarze  343:        assert(man && man->parse);
                    344:        return(man->parse);
1.76      schwarze  345: }
                    346:
                    347: void
1.99      schwarze  348: man_deroff(char **dest, const struct roff_node *n)
1.76      schwarze  349: {
                    350:        char    *cp;
                    351:        size_t   sz;
                    352:
1.98      schwarze  353:        if (n->type != ROFFT_TEXT) {
1.76      schwarze  354:                for (n = n->child; n; n = n->next)
                    355:                        man_deroff(dest, n);
                    356:                return;
                    357:        }
                    358:
1.77      schwarze  359:        /* Skip leading whitespace and escape sequences. */
1.76      schwarze  360:
1.77      schwarze  361:        cp = n->string;
                    362:        while ('\0' != *cp) {
                    363:                if ('\\' == *cp) {
                    364:                        cp++;
                    365:                        mandoc_escape((const char **)&cp, NULL, NULL);
                    366:                } else if (isspace((unsigned char)*cp))
                    367:                        cp++;
                    368:                else
1.76      schwarze  369:                        break;
1.77      schwarze  370:        }
1.76      schwarze  371:
                    372:        /* Skip trailing whitespace. */
                    373:
                    374:        for (sz = strlen(cp); sz; sz--)
                    375:                if (0 == isspace((unsigned char)cp[sz-1]))
                    376:                        break;
                    377:
                    378:        /* Skip empty strings. */
                    379:
                    380:        if (0 == sz)
                    381:                return;
                    382:
                    383:        if (NULL == *dest) {
                    384:                *dest = mandoc_strndup(cp, sz);
                    385:                return;
                    386:        }
                    387:
                    388:        mandoc_asprintf(&cp, "%s %*s", *dest, (int)sz, cp);
                    389:        free(*dest);
                    390:        *dest = cp;
1.4       schwarze  391: }