version 1.139, 2015/04/19 14:25:41 |
version 1.148, 2016/01/08 02:53:13 |
|
|
/* $Id$ */ |
/* $Id$ */ |
/* |
/* |
* Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv> |
* Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv> |
* Copyright (c) 2010-2015 Ingo Schwarze <schwarze@openbsd.org> |
* Copyright (c) 2010-2016 Ingo Schwarze <schwarze@openbsd.org> |
* Copyright (c) 2010, 2012 Joerg Sonnenberger <joerg@netbsd.org> |
* Copyright (c) 2010, 2012 Joerg Sonnenberger <joerg@netbsd.org> |
* |
* |
* Permission to use, copy, modify, and distribute this software for any |
* Permission to use, copy, modify, and distribute this software for any |
|
|
#include <sys/mman.h> |
#include <sys/mman.h> |
#include <sys/stat.h> |
#include <sys/stat.h> |
#endif |
#endif |
#include <sys/wait.h> |
|
|
|
#include <assert.h> |
#include <assert.h> |
#include <ctype.h> |
#include <ctype.h> |
|
#if HAVE_ERR |
|
#include <err.h> |
|
#endif |
#include <errno.h> |
#include <errno.h> |
#include <fcntl.h> |
#include <fcntl.h> |
#include <signal.h> |
|
#include <stdarg.h> |
#include <stdarg.h> |
#include <stdint.h> |
#include <stdint.h> |
#include <stdio.h> |
#include <stdio.h> |
#include <stdlib.h> |
#include <stdlib.h> |
#include <string.h> |
#include <string.h> |
#include <unistd.h> |
#include <unistd.h> |
|
#include <zlib.h> |
|
|
#include "mandoc_aux.h" |
#include "mandoc_aux.h" |
#include "mandoc.h" |
#include "mandoc.h" |
|
|
struct mparse { |
struct mparse { |
struct roff_man *man; /* man parser */ |
struct roff_man *man; /* man parser */ |
struct roff *roff; /* roff parser (!NULL) */ |
struct roff *roff; /* roff parser (!NULL) */ |
const struct mchars *mchars; /* character table */ |
|
char *sodest; /* filename pointed to by .so */ |
char *sodest; /* filename pointed to by .so */ |
const char *file; /* filename of current input file */ |
const char *file; /* filename of current input file */ |
struct buf *primary; /* buffer currently being parsed */ |
struct buf *primary; /* buffer currently being parsed */ |
|
|
enum mandoclevel file_status; /* status of current parse */ |
enum mandoclevel file_status; /* status of current parse */ |
enum mandoclevel wlevel; /* ignore messages below this */ |
enum mandoclevel wlevel; /* ignore messages below this */ |
int options; /* parser options */ |
int options; /* parser options */ |
|
int gzip; /* current input file is gzipped */ |
int filenc; /* encoding of the current file */ |
int filenc; /* encoding of the current file */ |
int reparse_count; /* finite interp. stack */ |
int reparse_count; /* finite interp. stack */ |
int line; /* line number in the file */ |
int line; /* line number in the file */ |
pid_t child; /* the gunzip(1) process */ |
|
}; |
}; |
|
|
static void choose_parser(struct mparse *); |
static void choose_parser(struct mparse *); |
Line 130 static const char * const mandocerrs[MANDOCERR_MAX] = |
|
Line 131 static const char * const mandocerrs[MANDOCERR_MAX] = |
|
"blocks badly nested", |
"blocks badly nested", |
"nested displays are not portable", |
"nested displays are not portable", |
"moving content out of list", |
"moving content out of list", |
".Vt block has child macro", |
|
"fill mode already enabled, skipping", |
"fill mode already enabled, skipping", |
"fill mode already disabled, skipping", |
"fill mode already disabled, skipping", |
"line scope broken", |
"line scope broken", |
Line 217 static const char * const mandocerrs[MANDOCERR_MAX] = |
|
Line 217 static const char * const mandocerrs[MANDOCERR_MAX] = |
|
/* related to request and macro arguments */ |
/* related to request and macro arguments */ |
"escaped character not allowed in a name", |
"escaped character not allowed in a name", |
"NOT IMPLEMENTED: Bd -file", |
"NOT IMPLEMENTED: Bd -file", |
|
"skipping display without arguments", |
"missing list type, using -item", |
"missing list type, using -item", |
"missing manual name, using \"\"", |
"missing manual name, using \"\"", |
"uname(3) system call failed, using UNKNOWN", |
"uname(3) system call failed, using UNKNOWN", |
Line 327 mparse_buf_r(struct mparse *curp, struct buf blk, size |
|
Line 328 mparse_buf_r(struct mparse *curp, struct buf blk, size |
|
int of; |
int of; |
int lnn; /* line number in the real file */ |
int lnn; /* line number in the real file */ |
int fd; |
int fd; |
pid_t save_child; |
|
unsigned char c; |
unsigned char c; |
|
|
memset(&ln, 0, sizeof(ln)); |
memset(&ln, 0, sizeof(ln)); |
|
|
if (curp->secondary) |
if (curp->secondary) |
curp->secondary->sz -= pos + 1; |
curp->secondary->sz -= pos + 1; |
save_file = curp->file; |
save_file = curp->file; |
save_child = curp->child; |
if ((fd = mparse_open(curp, ln.buf + of)) != -1) { |
if (mparse_open(curp, &fd, ln.buf + of) == |
|
MANDOCLEVEL_OK) { |
|
mparse_readfd(curp, fd, ln.buf + of); |
mparse_readfd(curp, fd, ln.buf + of); |
|
close(fd); |
curp->file = save_file; |
curp->file = save_file; |
} else { |
} else { |
curp->file = save_file; |
curp->file = save_file; |
|
|
of = 0; |
of = 0; |
mparse_buf_r(curp, ln, of, 0); |
mparse_buf_r(curp, ln, of, 0); |
} |
} |
curp->child = save_child; |
|
pos = 0; |
pos = 0; |
continue; |
continue; |
default: |
default: |
|
|
read_whole_file(struct mparse *curp, const char *file, int fd, |
read_whole_file(struct mparse *curp, const char *file, int fd, |
struct buf *fb, int *with_mmap) |
struct buf *fb, int *with_mmap) |
{ |
{ |
|
gzFile gz; |
size_t off; |
size_t off; |
ssize_t ssz; |
ssize_t ssz; |
|
|
#if HAVE_MMAP |
#if HAVE_MMAP |
struct stat st; |
struct stat st; |
if (-1 == fstat(fd, &st)) { |
|
perror(file); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
|
|
|
if (fstat(fd, &st) == -1) |
|
err((int)MANDOCLEVEL_SYSERR, "%s", file); |
|
|
/* |
/* |
* If we're a regular file, try just reading in the whole entry |
* If we're a regular file, try just reading in the whole entry |
* via mmap(). This is faster than reading it into blocks, and |
* via mmap(). This is faster than reading it into blocks, and |
Line 628 read_whole_file(struct mparse *curp, const char *file, |
|
Line 626 read_whole_file(struct mparse *curp, const char *file, |
|
* concerned that this is going to tank any machines. |
* concerned that this is going to tank any machines. |
*/ |
*/ |
|
|
if (S_ISREG(st.st_mode)) { |
if (curp->gzip == 0 && S_ISREG(st.st_mode)) { |
if (st.st_size > 0x7fffffff) { |
if (st.st_size > 0x7fffffff) { |
mandoc_msg(MANDOCERR_TOOLARGE, curp, 0, 0, NULL); |
mandoc_msg(MANDOCERR_TOOLARGE, curp, 0, 0, NULL); |
return(0); |
return 0; |
} |
} |
*with_mmap = 1; |
*with_mmap = 1; |
fb->sz = (size_t)st.st_size; |
fb->sz = (size_t)st.st_size; |
fb->buf = mmap(NULL, fb->sz, PROT_READ, MAP_SHARED, fd, 0); |
fb->buf = mmap(NULL, fb->sz, PROT_READ, MAP_SHARED, fd, 0); |
if (fb->buf != MAP_FAILED) |
if (fb->buf != MAP_FAILED) |
return(1); |
return 1; |
} |
} |
#endif |
#endif |
|
|
|
if (curp->gzip) { |
|
if ((gz = gzdopen(fd, "rb")) == NULL) |
|
err((int)MANDOCLEVEL_SYSERR, "%s", file); |
|
} else |
|
gz = NULL; |
|
|
/* |
/* |
* 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 |
* go the old way and just read things in bit by bit. |
* go the old way and just read things in bit by bit. |
Line 659 read_whole_file(struct mparse *curp, const char *file, |
|
Line 663 read_whole_file(struct mparse *curp, const char *file, |
|
} |
} |
resize_buf(fb, 65536); |
resize_buf(fb, 65536); |
} |
} |
ssz = read(fd, fb->buf + (int)off, fb->sz - off); |
ssz = curp->gzip ? |
|
gzread(gz, fb->buf + (int)off, fb->sz - off) : |
|
read(fd, fb->buf + (int)off, fb->sz - off); |
if (ssz == 0) { |
if (ssz == 0) { |
fb->sz = off; |
fb->sz = off; |
return(1); |
return 1; |
} |
} |
if (ssz == -1) { |
if (ssz == -1) |
perror(file); |
err((int)MANDOCLEVEL_SYSERR, "%s", file); |
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
off += (size_t)ssz; |
off += (size_t)ssz; |
} |
} |
|
|
free(fb->buf); |
free(fb->buf); |
fb->buf = NULL; |
fb->buf = NULL; |
return(0); |
return 0; |
} |
} |
|
|
static void |
static void |
Line 742 mparse_readmem(struct mparse *curp, void *buf, size_t |
|
Line 746 mparse_readmem(struct mparse *curp, void *buf, size_t |
|
blk.sz = len; |
blk.sz = len; |
|
|
mparse_parse_buffer(curp, blk, file); |
mparse_parse_buffer(curp, blk, file); |
return(curp->file_status); |
return curp->file_status; |
} |
} |
|
|
/* |
/* |
Line 769 mparse_readfd(struct mparse *curp, int fd, const char |
|
Line 773 mparse_readfd(struct mparse *curp, int fd, const char |
|
#endif |
#endif |
free(blk.buf); |
free(blk.buf); |
} |
} |
|
return curp->file_status; |
if (fd != STDIN_FILENO && close(fd) == -1) |
|
perror(file); |
|
|
|
mparse_wait(curp); |
|
return(curp->file_status); |
|
} |
} |
|
|
enum mandoclevel |
int |
mparse_open(struct mparse *curp, int *fd, const char *file) |
mparse_open(struct mparse *curp, const char *file) |
{ |
{ |
int pfd[2]; |
|
int save_errno; |
|
char *cp; |
char *cp; |
|
int fd; |
|
|
curp->file = file; |
curp->file = file; |
|
cp = strrchr(file, '.'); |
|
curp->gzip = (cp != NULL && ! strcmp(cp + 1, "gz")); |
|
|
/* Unless zipped, try to just open the file. */ |
/* First try to use the filename as it is. */ |
|
|
if ((cp = strrchr(file, '.')) == NULL || |
if ((fd = open(file, O_RDONLY)) != -1) |
strcmp(cp + 1, "gz")) { |
return fd; |
curp->child = 0; |
|
if ((*fd = open(file, O_RDONLY)) != -1) |
|
return(MANDOCLEVEL_OK); |
|
|
|
/* Open failed; try to append ".gz". */ |
/* |
|
* If that doesn't work and the filename doesn't |
|
* already end in .gz, try appending .gz. |
|
*/ |
|
|
|
if ( ! curp->gzip) { |
mandoc_asprintf(&cp, "%s.gz", file); |
mandoc_asprintf(&cp, "%s.gz", file); |
file = cp; |
fd = open(file, O_RDONLY); |
} else |
|
cp = NULL; |
|
|
|
/* Before forking, make sure the file can be read. */ |
|
|
|
save_errno = errno; |
|
if (access(file, R_OK) == -1) { |
|
if (cp != NULL) |
|
errno = save_errno; |
|
free(cp); |
free(cp); |
*fd = -1; |
if (fd != -1) { |
curp->child = 0; |
curp->gzip = 1; |
mandoc_msg(MANDOCERR_FILE, curp, 0, 0, strerror(errno)); |
return fd; |
return(MANDOCLEVEL_ERROR); |
|
} |
|
|
|
/* Run gunzip(1). */ |
|
|
|
if (pipe(pfd) == -1) { |
|
perror("pipe"); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
|
|
switch (curp->child = fork()) { |
|
case -1: |
|
perror("fork"); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
case 0: |
|
close(pfd[0]); |
|
if (dup2(pfd[1], STDOUT_FILENO) == -1) { |
|
perror("dup"); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
} |
signal(SIGPIPE, SIG_DFL); |
|
execlp("gunzip", "gunzip", "-c", file, NULL); |
|
perror("exec"); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
default: |
|
close(pfd[1]); |
|
*fd = pfd[0]; |
|
return(MANDOCLEVEL_OK); |
|
} |
} |
} |
|
|
|
enum mandoclevel |
/* Neither worked, give up. */ |
mparse_wait(struct mparse *curp) |
|
{ |
|
int status; |
|
|
|
if (curp->child == 0) |
mandoc_msg(MANDOCERR_FILE, curp, 0, 0, strerror(errno)); |
return(MANDOCLEVEL_OK); |
return -1; |
|
|
if (waitpid(curp->child, &status, 0) == -1) { |
|
perror("wait"); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
curp->child = 0; |
|
if (WIFSIGNALED(status)) { |
|
mandoc_vmsg(MANDOCERR_FILE, curp, 0, 0, |
|
"gunzip died from signal %d", WTERMSIG(status)); |
|
return(MANDOCLEVEL_ERROR); |
|
} |
|
if (WEXITSTATUS(status)) { |
|
mandoc_vmsg(MANDOCERR_FILE, curp, 0, 0, |
|
"gunzip failed with code %d", WEXITSTATUS(status)); |
|
return(MANDOCLEVEL_ERROR); |
|
} |
|
return(MANDOCLEVEL_OK); |
|
} |
} |
|
|
struct mparse * |
struct mparse * |
mparse_alloc(int options, enum mandoclevel wlevel, mandocmsg mmsg, |
mparse_alloc(int options, enum mandoclevel wlevel, mandocmsg mmsg, |
const struct mchars *mchars, const char *defos) |
const char *defos) |
{ |
{ |
struct mparse *curp; |
struct mparse *curp; |
|
|
Line 881 mparse_alloc(int options, enum mandoclevel wlevel, man |
|
Line 825 mparse_alloc(int options, enum mandoclevel wlevel, man |
|
curp->mmsg = mmsg; |
curp->mmsg = mmsg; |
curp->defos = defos; |
curp->defos = defos; |
|
|
curp->mchars = mchars; |
curp->roff = roff_alloc(curp, options); |
curp->roff = roff_alloc(curp, curp->mchars, options); |
|
curp->man = roff_man_alloc( curp->roff, curp, curp->defos, |
curp->man = roff_man_alloc( curp->roff, curp, curp->defos, |
curp->options & MPARSE_QUICK ? 1 : 0); |
curp->options & MPARSE_QUICK ? 1 : 0); |
if (curp->options & MPARSE_MDOC) { |
if (curp->options & MPARSE_MDOC) { |
Line 893 mparse_alloc(int options, enum mandoclevel wlevel, man |
|
Line 836 mparse_alloc(int options, enum mandoclevel wlevel, man |
|
curp->man->macroset = MACROSET_MAN; |
curp->man->macroset = MACROSET_MAN; |
} |
} |
curp->man->first->tok = TOKEN_NONE; |
curp->man->first->tok = TOKEN_NONE; |
return(curp); |
return curp; |
} |
} |
|
|
void |
void |
|
|
mparse_strerror(enum mandocerr er) |
mparse_strerror(enum mandocerr er) |
{ |
{ |
|
|
return(mandocerrs[er]); |
return mandocerrs[er]; |
} |
} |
|
|
const char * |
const char * |
mparse_strlevel(enum mandoclevel lvl) |
mparse_strlevel(enum mandoclevel lvl) |
{ |
{ |
return(mandoclevels[lvl]); |
return mandoclevels[lvl]; |
} |
} |
|
|
void |
void |
Line 1001 mparse_getkeep(const struct mparse *p) |
|
Line 944 mparse_getkeep(const struct mparse *p) |
|
{ |
{ |
|
|
assert(p->secondary); |
assert(p->secondary); |
return(p->secondary->sz ? p->secondary->buf : NULL); |
return p->secondary->sz ? p->secondary->buf : NULL; |
} |
} |