-/* $Id: main.c,v 1.112 2010/12/01 16:28:23 kristaps Exp $ */
+/* $Id: main.c,v 1.139 2011/01/22 13:16:02 schwarze Exp $ */
/*
- * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
+ * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
* Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
*
* Permission to use, copy, modify, and distribute this software for any
#define MAP_FILE 0
#endif
+#define REPARSE_LIMIT 1000
#define UNCONST(a) ((void *)(uintptr_t)(const void *)(a))
/* FIXME: Intel's compiler? LLVM? pcc? */
struct mdoc *mdoc; /* mdoc parser */
struct roff *roff; /* roff parser (!NULL) */
struct regset regs; /* roff registers */
+ int reparse_count; /* finite interpolation stack */
enum outt outtype; /* which output to use */
out_mdoc outmdoc; /* mdoc output ptr */
out_man outman; /* man output ptr */
"generic warning",
- "text should be uppercase",
- "sections out of conventional order",
- "section name repeats",
- "out of order prologue",
- "repeated prologue entry",
- "list type must come first",
- "tab in non-literal context",
- "bad escape sequence",
- "unterminated quoted string",
- "argument requires the width argument",
- "superfluous width argument",
- "bad date argument",
- "bad width argument",
+ /* related to the prologue */
+ "no title in document",
+ "document title should be all caps",
"unknown manual section",
+ "cannot parse date argument",
+ "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",
+ "manual name not yet set",
+ "sections out of conventional order",
+ "duplicate section name",
"section not in conventional manual section",
- "end of line whitespace",
- "blocks badly nested",
- "generic error",
+ /* related to macros and nesting */
+ "skipping obsolete macro",
+ "skipping paragraph macro",
+ "blocks badly nested",
+ "child violates parent syntax",
+ "nested displays are not portable",
+ "already in literal mode",
- "NAME section must come first",
+ /* 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",
- "child violates parent syntax",
- "bad AT&T symbol",
- "bad standard",
- "list type repeated",
- "display type repeated",
- "argument repeated",
- "ignoring argument",
- "manual name not yet set",
- "obsolete macro ignored",
- "empty macro ignored",
- "macro not allowed in body",
- "macro not allowed in prologue",
- "bad character",
- "bad NAME section contents",
- "no blank lines",
- "no text in this context",
+ "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",
- "unknown macro will be lost",
- "NOT IMPLEMENTED: skipping request",
+ "unknown escape sequence",
+ "unterminated quoted string",
+
+ "generic 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",
+ "skipping text before the first section header",
+ "skipping unknown macro",
+ "NOT IMPLEMENTED, please use groff: skipping request",
"line scope broken",
"argument count wrong",
- "request scope close w/none open",
- "scope already open",
+ "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)",
- "no title in document",
"missing list type",
- "missing display type",
- "missing font type",
"line argument(s) will be lost",
"body argument(s) will be lost",
- "paragraph macro ignored",
"generic fatal error",
"column syntax is inconsistent",
- "displays may not be nested",
- "unsupported display type",
- "blocks badly nested",
- "no such block is open",
+ "NOT IMPLEMENTED: .Bd -file",
"line scope broken, syntax violated",
"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",
- "utsname system call failed",
"static buffer exhausted",
};
static int woptions(struct curparse *, char *);
static const char *progname;
+static enum mandoclevel file_status = MANDOCLEVEL_OK;
static enum mandoclevel exit_status = MANDOCLEVEL_OK;
int
if (-1 == (fd = open(file, O_RDONLY, 0))) {
perror(file);
- exit_status = MANDOCLEVEL_SYSERR;
+ file_status = MANDOCLEVEL_SYSERR;
return(0);
}
if (-1 == close(fd))
perror(file);
- return(MANDOCLEVEL_FATAL > exit_status ? 1 : 0);
+ return(MANDOCLEVEL_FATAL > file_status ? 1 : 0);
}
resize_buf(struct buf *buf, size_t initial)
{
- buf->sz = buf->sz ? 2 * buf->sz : initial;
+ buf->sz = buf->sz > initial/2 ? 2 * buf->sz : initial;
buf->buf = realloc(buf->buf, buf->sz);
if (NULL == buf->buf) {
perror(NULL);
curp->mdoc = NULL;
curp->man = NULL;
+ file_status = MANDOCLEVEL_OK;
/* Make sure the mandotory roff parser is initialised. */
pdesc(curp);
- if (MANDOCLEVEL_FATAL <= exit_status)
+ if (MANDOCLEVEL_FATAL <= file_status)
goto cleanup;
/* NOTE a parser may not have been assigned, yet. */
if ( ! (curp->man || curp->mdoc)) {
fprintf(stderr, "%s: Not a manual\n", curp->file);
- exit_status = MANDOCLEVEL_FATAL;
+ file_status = MANDOCLEVEL_FATAL;
goto cleanup;
}
/* Clean up the parse routine ASTs. */
if (curp->mdoc && ! mdoc_endparse(curp->mdoc)) {
- assert(MANDOCLEVEL_FATAL <= exit_status);
+ assert(MANDOCLEVEL_FATAL <= file_status);
goto cleanup;
}
if (curp->man && ! man_endparse(curp->man)) {
- assert(MANDOCLEVEL_FATAL <= exit_status);
+ assert(MANDOCLEVEL_FATAL <= file_status);
goto cleanup;
}
assert(curp->roff);
- if ( ! roff_endparse(curp->roff)) {
- assert(MANDOCLEVEL_FATAL <= exit_status);
- goto cleanup;
- }
+ roff_endparse(curp->roff);
/*
* With -Wstop and warnings or errors of at least
* the requested level, do not produce output.
*/
- if (MANDOCLEVEL_OK != exit_status && curp->wstop)
+ if (MANDOCLEVEL_OK != file_status && curp->wstop)
goto cleanup;
/* If unset, allocate output dev now (if applicable). */
assert(curp->roff);
roff_reset(curp->roff);
+ if (exit_status < file_status)
+ exit_status = file_status;
+
return;
}
*/
if ( ! read_whole_file(curp, &blk, &with_mmap)) {
- exit_status = MANDOCLEVEL_SYSERR;
+ file_status = MANDOCLEVEL_SYSERR;
return;
}
parsebuf(struct curparse *curp, struct buf blk, int start)
{
struct buf ln;
+ enum rofferr rr;
int i, of, rc;
int pos; /* byte number in the ln buffer */
int lnn; /* line number in the real file */
if (0 == pos && '\0' == blk.buf[i])
break;
- if (start)
+ if (start) {
curp->line = lnn;
+ curp->reparse_count = 0;
+ }
while (i < (int)blk.sz && (start || '\0' != blk.buf[i])) {
if ('\n' == blk.buf[i]) {
of = 0;
rerun:
- rc = roff_parseln
+ rr = roff_parseln
(curp->roff, curp->line,
&ln.buf, &ln.sz, of, &of);
- switch (rc) {
+ switch (rr) {
case (ROFF_REPARSE):
- parsebuf(curp, ln, 0);
+ if (REPARSE_LIMIT >= ++curp->reparse_count)
+ parsebuf(curp, ln, 0);
+ else
+ mmsg(MANDOCERR_ROFFLOOP, curp,
+ curp->line, pos, NULL);
pos = 0;
continue;
case (ROFF_APPEND):
pos = 0;
continue;
case (ROFF_ERR):
- assert(MANDOCLEVEL_FATAL <= exit_status);
+ assert(MANDOCLEVEL_FATAL <= file_status);
break;
case (ROFF_SO):
if (pfile(ln.buf + of, curp)) {
continue;
} else
break;
- case (ROFF_CONT):
+ default:
break;
}
+ /*
+ * If we encounter errors in the recursive parsebuf()
+ * call, make sure we don't continue parsing.
+ */
+
+ if (MANDOCLEVEL_FATAL <= file_status)
+ break;
+
/*
* If input parsers have not been allocated, do so now.
* We keep these instanced betwen parsers, but set them
* Lastly, push down into the parsers themselves. One
* of these will have already been set in the pset()
* routine.
+ * If libroff returns ROFF_TBL, then add it to the
+ * currently open parse. Since we only get here if
+ * there does exist data (see tbl_data.c), we're
+ * guaranteed that something's been allocated.
*/
- if (curp->man || curp->mdoc) {
+ if (ROFF_TBL == rr) {
+ assert(curp->man || curp->mdoc);
+ if (curp->man)
+ man_addspan(curp->man, roff_span(curp->roff));
+ else
+ mdoc_addspan(curp->mdoc, roff_span(curp->roff));
+
+ } else if (curp->man || curp->mdoc) {
rc = curp->man ?
man_parseln(curp->man,
curp->line, ln.buf, of) :
curp->line, ln.buf, of);
if ( ! rc) {
- assert(MANDOCLEVEL_FATAL <= exit_status);
+ assert(MANDOCLEVEL_FATAL <= file_status);
break;
}
}
fprintf(stderr, ": %s", msg);
fputc('\n', stderr);
- if (exit_status < level)
- exit_status = level;
+ if (file_status < level)
+ file_status = level;
return(level < MANDOCLEVEL_FATAL);
}