version 1.3, 2011/03/20 16:02:05 |
version 1.32, 2012/11/16 22:21:05 |
|
|
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF |
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF |
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. |
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. |
*/ |
*/ |
#include <sys/stat.h> |
#ifdef HAVE_CONFIG_H |
#include <sys/mman.h> |
#include "config.h" |
|
#endif |
|
|
|
#ifdef HAVE_MMAP |
|
# include <sys/stat.h> |
|
# include <sys/mman.h> |
|
#endif |
|
|
#include <assert.h> |
#include <assert.h> |
#include <ctype.h> |
#include <ctype.h> |
#include <fcntl.h> |
#include <fcntl.h> |
#include <stdarg.h> |
#include <stdarg.h> |
|
#include <stdint.h> |
#include <stdio.h> |
#include <stdio.h> |
#include <stdlib.h> |
#include <stdlib.h> |
#include <string.h> |
#include <string.h> |
|
|
#include "libmandoc.h" |
#include "libmandoc.h" |
#include "mdoc.h" |
#include "mdoc.h" |
#include "man.h" |
#include "man.h" |
#include "roff.h" |
#include "main.h" |
|
|
#ifndef MAP_FILE |
#ifndef MAP_FILE |
#define MAP_FILE 0 |
#define MAP_FILE 0 |
|
|
struct man *man; /* man parser */ |
struct man *man; /* man parser */ |
struct mdoc *mdoc; /* mdoc parser */ |
struct mdoc *mdoc; /* mdoc parser */ |
struct roff *roff; /* roff parser (!NULL) */ |
struct roff *roff; /* roff parser (!NULL) */ |
struct regset regs; /* roff registers */ |
|
int reparse_count; /* finite interp. stack */ |
int reparse_count; /* finite interp. stack */ |
mandocmsg mmsg; /* warning/error message handler */ |
mandocmsg mmsg; /* warning/error message handler */ |
void *arg; /* argument to mmsg */ |
void *arg; /* argument to mmsg */ |
mevt_open evt_open; /* file-open event */ |
|
mevt_close evt_close; /* file-close event */ |
|
const char *file; |
const char *file; |
|
struct buf *secondary; |
|
char *defos; /* default operating system */ |
}; |
}; |
|
|
static void resize_buf(struct buf *, size_t); |
static void resize_buf(struct buf *, size_t); |
static void mparse_buf_r(struct mparse *, struct buf, int); |
static void mparse_buf_r(struct mparse *, struct buf, int); |
static void mparse_readfd_r(struct mparse *, int, const char *, int); |
static void mparse_readfd_r(struct mparse *, int, const char *, int); |
static void pset(const char *, int, struct mparse *); |
static void pset(const char *, int, struct mparse *); |
static void pdesc(struct mparse *, const char *, int); |
|
static int read_whole_file(const char *, int, struct buf *, int *); |
static int read_whole_file(const char *, int, struct buf *, int *); |
static void mparse_end(struct mparse *); |
static void mparse_end(struct mparse *); |
|
|
Line 81 static const enum mandocerr mandoclimits[MANDOCLEVEL_M |
|
Line 86 static const enum mandocerr mandoclimits[MANDOCLEVEL_M |
|
MANDOCERR_MAX |
MANDOCERR_MAX |
}; |
}; |
|
|
|
static const char * const mandocerrs[MANDOCERR_MAX] = { |
|
"ok", |
|
|
|
"generic warning", |
|
|
|
/* related to the prologue */ |
|
"no title in document", |
|
"document title should be all caps", |
|
"unknown manual section", |
|
"unknown manual volume or arch", |
|
"date missing, using today's date", |
|
"cannot parse date, using it verbatim", |
|
"prologue macros out of order", |
|
"duplicate prologue macro", |
|
"macro not allowed in prologue", |
|
"macro not allowed in body", |
|
|
|
/* related to document structure */ |
|
".so is fragile, better use ln(1)", |
|
"NAME section must come first", |
|
"bad NAME section contents", |
|
"sections out of conventional order", |
|
"duplicate section name", |
|
"section not in conventional manual section", |
|
|
|
/* related to macros and nesting */ |
|
"skipping obsolete macro", |
|
"skipping paragraph macro", |
|
"moving paragraph macro out of list", |
|
"skipping no-space macro", |
|
"blocks badly nested", |
|
"child violates parent syntax", |
|
"nested displays are not portable", |
|
"already in literal mode", |
|
"line scope broken", |
|
|
|
/* related to missing macro arguments */ |
|
"skipping empty macro", |
|
"argument count wrong", |
|
"missing display type", |
|
"list type must come first", |
|
"tag lists require a width argument", |
|
"missing font type", |
|
"skipping end of block that is not open", |
|
|
|
/* related to bad macro arguments */ |
|
"skipping argument", |
|
"duplicate argument", |
|
"duplicate display type", |
|
"duplicate list type", |
|
"unknown AT&T UNIX version", |
|
"bad Boolean value", |
|
"unknown font", |
|
"unknown standard specifier", |
|
"bad width argument", |
|
|
|
/* related to plain text */ |
|
"blank line in non-literal context", |
|
"tab in non-literal context", |
|
"end of line whitespace", |
|
"bad comment style", |
|
"bad escape sequence", |
|
"unterminated quoted string", |
|
|
|
/* related to equations */ |
|
"unexpected literal in equation", |
|
|
|
"generic error", |
|
|
|
/* related to equations */ |
|
"unexpected equation scope closure", |
|
"equation scope open on exit", |
|
"overlapping equation scopes", |
|
"unexpected end of equation", |
|
"equation syntax error", |
|
|
|
/* related to tables */ |
|
"bad table syntax", |
|
"bad table option", |
|
"bad table layout", |
|
"no table layout cells specified", |
|
"no table data cells specified", |
|
"ignore data in cell", |
|
"data block still open", |
|
"ignoring extra data cells", |
|
|
|
"input stack limit exceeded, infinite loop?", |
|
"skipping bad character", |
|
"escaped character not allowed in a name", |
|
"manual name not yet set", |
|
"skipping text before the first section header", |
|
"skipping unknown macro", |
|
"NOT IMPLEMENTED, please use groff: skipping request", |
|
"argument count wrong", |
|
"skipping end of block that is not open", |
|
"missing end of block", |
|
"scope open on exit", |
|
"uname(3) system call failed", |
|
"macro requires line argument(s)", |
|
"macro requires body argument(s)", |
|
"macro requires argument(s)", |
|
"missing list type", |
|
"line argument(s) will be lost", |
|
"body argument(s) will be lost", |
|
|
|
"generic fatal error", |
|
|
|
"not a manual", |
|
"column syntax is inconsistent", |
|
"NOT IMPLEMENTED: .Bd -file", |
|
"argument count wrong, violates syntax", |
|
"child violates parent syntax", |
|
"argument count wrong, violates syntax", |
|
"NOT IMPLEMENTED: .so with absolute path or \"..\"", |
|
"no document body", |
|
"no document prologue", |
|
"static buffer exhausted", |
|
}; |
|
|
|
static const char * const mandoclevels[MANDOCLEVEL_MAX] = { |
|
"SUCCESS", |
|
"RESERVED", |
|
"WARNING", |
|
"ERROR", |
|
"FATAL", |
|
"BADARG", |
|
"SYSERR" |
|
}; |
|
|
static void |
static void |
resize_buf(struct buf *buf, size_t initial) |
resize_buf(struct buf *buf, size_t initial) |
{ |
{ |
Line 116 pset(const char *buf, int pos, struct mparse *curp) |
|
Line 250 pset(const char *buf, int pos, struct mparse *curp) |
|
switch (curp->inttype) { |
switch (curp->inttype) { |
case (MPARSE_MDOC): |
case (MPARSE_MDOC): |
if (NULL == curp->pmdoc) |
if (NULL == curp->pmdoc) |
curp->pmdoc = mdoc_alloc(&curp->regs, curp); |
curp->pmdoc = mdoc_alloc(curp->roff, curp, |
|
curp->defos); |
assert(curp->pmdoc); |
assert(curp->pmdoc); |
curp->mdoc = curp->pmdoc; |
curp->mdoc = curp->pmdoc; |
return; |
return; |
case (MPARSE_MAN): |
case (MPARSE_MAN): |
if (NULL == curp->pman) |
if (NULL == curp->pman) |
curp->pman = man_alloc(&curp->regs, curp); |
curp->pman = man_alloc(curp->roff, curp); |
assert(curp->pman); |
assert(curp->pman); |
curp->man = curp->pman; |
curp->man = curp->pman; |
return; |
return; |
Line 132 pset(const char *buf, int pos, struct mparse *curp) |
|
Line 267 pset(const char *buf, int pos, struct mparse *curp) |
|
|
|
if (pos >= 3 && 0 == memcmp(buf, ".Dd", 3)) { |
if (pos >= 3 && 0 == memcmp(buf, ".Dd", 3)) { |
if (NULL == curp->pmdoc) |
if (NULL == curp->pmdoc) |
curp->pmdoc = mdoc_alloc(&curp->regs, curp); |
curp->pmdoc = mdoc_alloc(curp->roff, curp, |
|
curp->defos); |
assert(curp->pmdoc); |
assert(curp->pmdoc); |
curp->mdoc = curp->pmdoc; |
curp->mdoc = curp->pmdoc; |
return; |
return; |
} |
} |
|
|
if (NULL == curp->pman) |
if (NULL == curp->pman) |
curp->pman = man_alloc(&curp->regs, curp); |
curp->pman = man_alloc(curp->roff, curp); |
assert(curp->pman); |
assert(curp->pman); |
curp->man = curp->pman; |
curp->man = curp->pman; |
} |
} |
Line 195 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
Line 331 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
* Warn about bogus characters. If you're using |
* Warn about bogus characters. If you're using |
* non-ASCII encoding, you're screwing your |
* non-ASCII encoding, you're screwing your |
* readers. Since I'd rather this not happen, |
* readers. Since I'd rather this not happen, |
* I'll be helpful and drop these characters so |
* I'll be helpful and replace these characters |
* we don't display gibberish. Note to manual |
* with "?", so we don't display gibberish. |
* writers: use special characters. |
* Note to manual writers: use special characters. |
*/ |
*/ |
|
|
c = (unsigned char) blk.buf[i]; |
c = (unsigned char) blk.buf[i]; |
Line 205 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
Line 341 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
if ( ! (isascii(c) && |
if ( ! (isascii(c) && |
(isgraph(c) || isblank(c)))) { |
(isgraph(c) || isblank(c)))) { |
mandoc_msg(MANDOCERR_BADCHAR, curp, |
mandoc_msg(MANDOCERR_BADCHAR, curp, |
curp->line, pos, "ignoring byte"); |
curp->line, pos, NULL); |
i++; |
i++; |
|
if (pos >= (int)ln.sz) |
|
resize_buf(&ln, 256); |
|
ln.buf[pos++] = '?'; |
continue; |
continue; |
} |
} |
|
|
Line 235 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
Line 374 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
continue; |
continue; |
} |
} |
|
|
if ('"' == blk.buf[i + 1]) { |
if ('"' == blk.buf[i + 1] || '#' == blk.buf[i + 1]) { |
i += 2; |
i += 2; |
/* Comment, skip to end of line */ |
/* Comment, skip to end of line */ |
for (; i < (int)blk.sz; ++i) { |
for (; i < (int)blk.sz; ++i) { |
Line 281 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
Line 420 mparse_buf_r(struct mparse *curp, struct buf blk, int |
|
|
|
of = 0; |
of = 0; |
|
|
|
/* |
|
* Maintain a lookaside buffer of all parsed lines. We |
|
* only do this if mparse_keep() has been invoked (the |
|
* buffer may be accessed with mparse_getkeep()). |
|
*/ |
|
|
|
if (curp->secondary) { |
|
curp->secondary->buf = |
|
mandoc_realloc |
|
(curp->secondary->buf, |
|
curp->secondary->sz + pos + 2); |
|
memcpy(curp->secondary->buf + |
|
curp->secondary->sz, |
|
ln.buf, pos); |
|
curp->secondary->sz += pos; |
|
curp->secondary->buf |
|
[curp->secondary->sz] = '\n'; |
|
curp->secondary->sz++; |
|
curp->secondary->buf |
|
[curp->secondary->sz] = '\0'; |
|
} |
rerun: |
rerun: |
rr = roff_parseln |
rr = roff_parseln |
(curp->roff, curp->line, |
(curp->roff, curp->line, |
|
|
assert(MANDOCLEVEL_FATAL <= curp->file_status); |
assert(MANDOCLEVEL_FATAL <= curp->file_status); |
break; |
break; |
case (ROFF_SO): |
case (ROFF_SO): |
|
/* |
|
* We remove `so' clauses from our lookaside |
|
* buffer because we're going to descend into |
|
* the file recursively. |
|
*/ |
|
if (curp->secondary) |
|
curp->secondary->sz -= pos + 1; |
mparse_readfd_r(curp, -1, ln.buf + of, 1); |
mparse_readfd_r(curp, -1, ln.buf + of, 1); |
if (MANDOCLEVEL_FATAL <= curp->file_status) |
if (MANDOCLEVEL_FATAL <= curp->file_status) |
break; |
break; |
|
|
|
|
/* |
/* |
* If input parsers have not been allocated, do so now. |
* If input parsers have not been allocated, do so now. |
* We keep these instanced betwen parsers, but set them |
* We keep these instanced between parsers, but set them |
* locally per parse routine since we can use different |
* locally per parse routine since we can use different |
* parsers with each one. |
* parsers with each one. |
*/ |
*/ |
|
|
free(ln.buf); |
free(ln.buf); |
} |
} |
|
|
static void |
|
pdesc(struct mparse *curp, const char *file, int fd) |
|
{ |
|
struct buf blk; |
|
int with_mmap; |
|
|
|
/* |
|
* Run for each opened file; may be called more than once for |
|
* each full parse sequence if the opened file is nested (i.e., |
|
* from `so'). Simply sucks in the whole file and moves into |
|
* the parse phase for the file. |
|
*/ |
|
|
|
if ( ! read_whole_file(file, fd, &blk, &with_mmap)) { |
|
curp->file_status = MANDOCLEVEL_SYSERR; |
|
return; |
|
} |
|
|
|
/* Line number is per-file. */ |
|
|
|
curp->line = 1; |
|
|
|
mparse_buf_r(curp, blk, 1); |
|
|
|
if (with_mmap) |
|
munmap(blk.buf, blk.sz); |
|
else |
|
free(blk.buf); |
|
} |
|
|
|
static int |
static int |
read_whole_file(const char *file, int fd, struct buf *fb, int *with_mmap) |
read_whole_file(const char *file, int fd, struct buf *fb, int *with_mmap) |
{ |
{ |
struct stat st; |
|
size_t off; |
size_t off; |
ssize_t ssz; |
ssize_t ssz; |
|
|
|
#ifdef HAVE_MMAP |
|
struct stat st; |
if (-1 == fstat(fd, &st)) { |
if (-1 == fstat(fd, &st)) { |
perror(file); |
perror(file); |
return(0); |
return(0); |
Line 447 read_whole_file(const char *file, int fd, struct buf * |
|
Line 585 read_whole_file(const char *file, int fd, struct buf * |
|
if (fb->buf != MAP_FAILED) |
if (fb->buf != MAP_FAILED) |
return(1); |
return(1); |
} |
} |
|
#endif |
|
|
/* |
/* |
* If this isn't a regular file (like, say, stdin), then we must |
* If this isn't a regular file (like, say, stdin), then we must |
Line 499 mparse_end(struct mparse *curp) |
|
Line 638 mparse_end(struct mparse *curp) |
|
return; |
return; |
} |
} |
|
|
#if 0 |
|
/* FIXME: NOTE a parser may not have been assigned, yet. */ |
|
|
|
if ( ! (curp->man || curp->mdoc)) { |
if ( ! (curp->man || curp->mdoc)) { |
/* FIXME: make into an mandoc.h error. */ |
mandoc_msg(MANDOCERR_NOTMANUAL, curp, 1, 0, NULL); |
fprintf(stderr, "%s: Not a manual\n", curp->file); |
|
curp->file_status = MANDOCLEVEL_FATAL; |
curp->file_status = MANDOCLEVEL_FATAL; |
goto cleanup; |
return; |
} |
} |
#endif |
|
|
|
roff_endparse(curp->roff); |
roff_endparse(curp->roff); |
} |
} |
|
|
static void |
static void |
mparse_readfd_r(struct mparse *curp, int fd, const char *file, int re) |
mparse_parse_buffer(struct mparse *curp, struct buf blk, const char *file, |
|
int re) |
{ |
{ |
const char *svfile; |
const char *svfile; |
|
|
if ( ! (*curp->evt_open)(curp->arg, file)) { |
/* Line number is per-file. */ |
curp->file_status = MANDOCLEVEL_SYSERR; |
svfile = curp->file; |
return; |
curp->file = file; |
} |
curp->line = 1; |
|
|
|
mparse_buf_r(curp, blk, 1); |
|
|
|
if (0 == re && MANDOCLEVEL_FATAL > curp->file_status) |
|
mparse_end(curp); |
|
|
|
curp->file = svfile; |
|
} |
|
|
|
enum mandoclevel |
|
mparse_readmem(struct mparse *curp, const void *buf, size_t len, |
|
const char *file) |
|
{ |
|
struct buf blk; |
|
|
|
blk.buf = UNCONST(buf); |
|
blk.sz = len; |
|
|
|
mparse_parse_buffer(curp, blk, file, 0); |
|
return(curp->file_status); |
|
} |
|
|
|
static void |
|
mparse_readfd_r(struct mparse *curp, int fd, const char *file, int re) |
|
{ |
|
struct buf blk; |
|
int with_mmap; |
|
|
if (-1 == fd) |
if (-1 == fd) |
if (-1 == (fd = open(file, O_RDONLY, 0))) { |
if (-1 == (fd = open(file, O_RDONLY, 0))) { |
perror(file); |
perror(file); |
curp->file_status = MANDOCLEVEL_SYSERR; |
curp->file_status = MANDOCLEVEL_SYSERR; |
return; |
return; |
} |
} |
|
/* |
|
* Run for each opened file; may be called more than once for |
|
* each full parse sequence if the opened file is nested (i.e., |
|
* from `so'). Simply sucks in the whole file and moves into |
|
* the parse phase for the file. |
|
*/ |
|
|
svfile = curp->file; |
if ( ! read_whole_file(file, fd, &blk, &with_mmap)) { |
curp->file = file; |
curp->file_status = MANDOCLEVEL_SYSERR; |
|
return; |
|
} |
|
|
pdesc(curp, file, fd); |
mparse_parse_buffer(curp, blk, file, re); |
|
|
if (0 == re && MANDOCLEVEL_FATAL > curp->file_status) |
#ifdef HAVE_MMAP |
mparse_end(curp); |
if (with_mmap) |
|
munmap(blk.buf, blk.sz); |
|
else |
|
#endif |
|
free(blk.buf); |
|
|
if (STDIN_FILENO != fd && -1 == close(fd)) |
if (STDIN_FILENO != fd && -1 == close(fd)) |
perror(file); |
perror(file); |
|
|
(*curp->evt_close)(curp->arg, file); |
|
curp->file = svfile; |
|
} |
} |
|
|
enum mandoclevel |
enum mandoclevel |
Line 554 mparse_readfd(struct mparse *curp, int fd, const char |
|
Line 725 mparse_readfd(struct mparse *curp, int fd, const char |
|
} |
} |
|
|
struct mparse * |
struct mparse * |
mparse_alloc(enum mparset inttype, mevt_open eopen, |
mparse_alloc(enum mparset inttype, enum mandoclevel wlevel, |
mevt_close eclose, enum mandoclevel wlevel, mandocmsg mmsg, void *arg) |
mandocmsg mmsg, void *arg, char *defos) |
{ |
{ |
struct mparse *curp; |
struct mparse *curp; |
|
|
|
assert(wlevel <= MANDOCLEVEL_FATAL); |
|
|
curp = mandoc_calloc(1, sizeof(struct mparse)); |
curp = mandoc_calloc(1, sizeof(struct mparse)); |
|
|
curp->wlevel = wlevel; |
curp->wlevel = wlevel; |
curp->mmsg = mmsg; |
curp->mmsg = mmsg; |
curp->arg = arg; |
curp->arg = arg; |
curp->inttype = inttype; |
curp->inttype = inttype; |
curp->evt_open = eopen; |
curp->defos = defos; |
curp->evt_close = eclose; |
|
|
|
curp->roff = roff_alloc(&curp->regs, curp); |
curp->roff = roff_alloc(curp); |
return(curp); |
return(curp); |
} |
} |
|
|
|
|
mparse_reset(struct mparse *curp) |
mparse_reset(struct mparse *curp) |
{ |
{ |
|
|
memset(&curp->regs, 0, sizeof(struct regset)); |
|
|
|
roff_reset(curp->roff); |
roff_reset(curp->roff); |
|
|
if (curp->mdoc) |
if (curp->mdoc) |
mdoc_reset(curp->mdoc); |
mdoc_reset(curp->mdoc); |
if (curp->man) |
if (curp->man) |
man_reset(curp->man); |
man_reset(curp->man); |
|
if (curp->secondary) |
|
curp->secondary->sz = 0; |
|
|
curp->file_status = MANDOCLEVEL_OK; |
curp->file_status = MANDOCLEVEL_OK; |
curp->mdoc = NULL; |
curp->mdoc = NULL; |
Line 600 mparse_free(struct mparse *curp) |
|
Line 772 mparse_free(struct mparse *curp) |
|
man_free(curp->pman); |
man_free(curp->pman); |
if (curp->roff) |
if (curp->roff) |
roff_free(curp->roff); |
roff_free(curp->roff); |
|
if (curp->secondary) |
|
free(curp->secondary->buf); |
|
|
|
free(curp->secondary); |
free(curp); |
free(curp); |
} |
} |
|
|
|
|
mparse_result(struct mparse *curp, struct mdoc **mdoc, struct man **man) |
mparse_result(struct mparse *curp, struct mdoc **mdoc, struct man **man) |
{ |
{ |
|
|
*mdoc = curp->mdoc; |
if (mdoc) |
*man = curp->man; |
*mdoc = curp->mdoc; |
|
if (man) |
|
*man = curp->man; |
} |
} |
|
|
void |
void |
Line 639 mandoc_msg(enum mandocerr er, struct mparse *m, |
|
Line 816 mandoc_msg(enum mandocerr er, struct mparse *m, |
|
if (level < m->wlevel) |
if (level < m->wlevel) |
return; |
return; |
|
|
(*m->mmsg)(er, level, m->file, ln, col, msg); |
if (m->mmsg) |
|
(*m->mmsg)(er, level, m->file, ln, col, msg); |
|
|
if (m->file_status < level) |
if (m->file_status < level) |
m->file_status = level; |
m->file_status = level; |
|
} |
|
|
|
const char * |
|
mparse_strerror(enum mandocerr er) |
|
{ |
|
|
|
return(mandocerrs[er]); |
|
} |
|
|
|
const char * |
|
mparse_strlevel(enum mandoclevel lvl) |
|
{ |
|
return(mandoclevels[lvl]); |
|
} |
|
|
|
void |
|
mparse_keep(struct mparse *p) |
|
{ |
|
|
|
assert(NULL == p->secondary); |
|
p->secondary = mandoc_calloc(1, sizeof(struct buf)); |
|
} |
|
|
|
const char * |
|
mparse_getkeep(const struct mparse *p) |
|
{ |
|
|
|
assert(p->secondary); |
|
return(p->secondary->sz ? p->secondary->buf : NULL); |
} |
} |