aboutsummaryrefslogtreecommitdiff
path: root/xmlwf
diff options
context:
space:
mode:
authorXin LI <delphij@FreeBSD.org>2024-02-17 21:54:42 +0000
committerXin LI <delphij@FreeBSD.org>2024-02-17 21:54:42 +0000
commit345cf360f8b824329558275653518e5f514627ac (patch)
tree9963c00cb06c950848b3d9f900e29dec0b2976e5 /xmlwf
parent9fc8132ecaf2ab1867df588de1907dda6875a3e3 (diff)
Diffstat (limited to 'xmlwf')
-rw-r--r--xmlwf/Makefile.in10
-rw-r--r--xmlwf/readfilemap.c21
-rw-r--r--xmlwf/xmlfile.c16
-rw-r--r--xmlwf/xmlfile.h4
-rw-r--r--xmlwf/xmlwf.c156
-rwxr-xr-xxmlwf/xmlwf_helpgen.py31
6 files changed, 152 insertions, 86 deletions
diff --git a/xmlwf/Makefile.in b/xmlwf/Makefile.in
index 9179e1cb1119..a8a84dc4e6de 100644
--- a/xmlwf/Makefile.in
+++ b/xmlwf/Makefile.in
@@ -134,6 +134,8 @@ am__aclocal_m4_deps = $(top_srcdir)/m4/libtool.m4 \
$(top_srcdir)/conftools/ax-append-compile-flags.m4 \
$(top_srcdir)/conftools/ax-append-link-flags.m4 \
$(top_srcdir)/conftools/expatcfg-compiler-supports-visibility.m4 \
+ $(top_srcdir)/conftools/ax-cxx-compile-stdcxx.m4 \
+ $(top_srcdir)/conftools/ax-cxx-compile-stdcxx-11.m4 \
$(top_srcdir)/configure.ac
am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \
$(ACLOCAL_M4)
@@ -267,6 +269,7 @@ FGREP = @FGREP@
FILECMD = @FILECMD@
FILEMAP = @FILEMAP@
GREP = @GREP@
+HAVE_CXX11 = @HAVE_CXX11@
INSTALL = @INSTALL@
INSTALL_DATA = @INSTALL_DATA@
INSTALL_PROGRAM = @INSTALL_PROGRAM@
@@ -286,6 +289,7 @@ LIPO = @LIPO@
LN_S = @LN_S@
LTLIBOBJS = @LTLIBOBJS@
LT_SYS_LIBRARY_PATH = @LT_SYS_LIBRARY_PATH@
+MAINT = @MAINT@
MAKEINFO = @MAKEINFO@
MANIFEST_TOOL = @MANIFEST_TOOL@
MKDIR_P = @MKDIR_P@
@@ -392,7 +396,7 @@ all: all-am
.SUFFIXES:
.SUFFIXES: .c .lo .o .obj
-$(srcdir)/Makefile.in: $(srcdir)/Makefile.am $(am__configure_deps)
+$(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps)
@for dep in $?; do \
case '$(am__configure_deps)' in \
*$$dep*) \
@@ -416,9 +420,9 @@ Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status
$(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES)
cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh
-$(top_srcdir)/configure: $(am__configure_deps)
+$(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps)
cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh
-$(ACLOCAL_M4): $(am__aclocal_m4_deps)
+$(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps)
cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh
$(am__aclocal_m4_deps):
install-binPROGRAMS: $(bin_PROGRAMS)
diff --git a/xmlwf/readfilemap.c b/xmlwf/readfilemap.c
index 59121f5b4dfe..2cb53feef8d5 100644
--- a/xmlwf/readfilemap.c
+++ b/xmlwf/readfilemap.c
@@ -13,6 +13,7 @@
Copyright (c) 2016-2017 Sebastian Pipping <sebastian@pipping.org>
Copyright (c) 2017 Rhodri James <rhodri@wildebeest.org.uk>
Copyright (c) 2017 Franek Korta <fkorta@gmail.com>
+ Copyright (c) 2022 Sean McBride <sean@rogue-research.com>
Licensed under the MIT license:
Permission is hereby granted, free of charge, to any person obtaining
@@ -50,14 +51,14 @@
#if defined(_MSC_VER)
# include <io.h>
/* https://msdn.microsoft.com/en-us/library/wyssk1bs(v=vs.100).aspx */
-# define _EXPAT_read _read
-# define _EXPAT_read_count_t int
-# define _EXPAT_read_req_t unsigned int
+# define EXPAT_read _read
+# define EXPAT_read_count_t int
+# define EXPAT_read_req_t unsigned int
#else /* POSIX */
/* http://pubs.opengroup.org/onlinepubs/009695399/functions/read.html */
-# define _EXPAT_read read
-# define _EXPAT_read_count_t ssize_t
-# define _EXPAT_read_req_t size_t
+# define EXPAT_read read
+# define EXPAT_read_count_t ssize_t
+# define EXPAT_read_req_t size_t
#endif
#ifndef S_ISREG
@@ -67,7 +68,7 @@
# ifndef S_IFMT
# define S_IFMT _S_IFMT
# endif
-# define S_ISREG(m) (((m)&S_IFMT) == S_IFREG)
+# define S_ISREG(m) (((m) & S_IFMT) == S_IFREG)
#endif /* not S_ISREG */
#ifndef O_BINARY
@@ -87,7 +88,7 @@ filemap(const tchar *name,
void *arg) {
size_t nbytes;
int fd;
- _EXPAT_read_count_t n;
+ EXPAT_read_count_t n;
struct stat sb;
void *p;
@@ -125,14 +126,14 @@ filemap(const tchar *name,
close(fd);
return 0;
}
- n = _EXPAT_read(fd, p, (_EXPAT_read_req_t)nbytes);
+ n = EXPAT_read(fd, p, (EXPAT_read_req_t)nbytes);
if (n < 0) {
tperror(name);
free(p);
close(fd);
return 0;
}
- if (n != (_EXPAT_read_count_t)nbytes) {
+ if (n != (EXPAT_read_count_t)nbytes) {
ftprintf(stderr, T("%s: read unexpected number of bytes\n"), name);
free(p);
close(fd);
diff --git a/xmlwf/xmlfile.c b/xmlwf/xmlfile.c
index e3521676c5b9..0598b86b5fb7 100644
--- a/xmlwf/xmlfile.c
+++ b/xmlwf/xmlfile.c
@@ -11,10 +11,10 @@
Copyright (c) 2002-2003 Fred L. Drake, Jr. <fdrake@users.sourceforge.net>
Copyright (c) 2004-2006 Karl Waclawek <karl@waclawek.net>
Copyright (c) 2005-2007 Steven Solie <steven@solie.ca>
- Copyright (c) 2016-2021 Sebastian Pipping <sebastian@pipping.org>
+ Copyright (c) 2016-2023 Sebastian Pipping <sebastian@pipping.org>
Copyright (c) 2017 Rhodri James <rhodri@wildebeest.org.uk>
Copyright (c) 2019 David Loffredo <loffredo@steptools.com>
- Copyright (c) 2021 Dong-hee Na <donghee.na@python.org>
+ Copyright (c) 2021 Donghee Na <donghee.na@python.org>
Licensed under the MIT license:
Permission is hereby granted, free of charge, to any person obtaining
@@ -37,7 +37,7 @@
USE OR OTHER DEALINGS IN THE SOFTWARE.
*/
-#include <expat_config.h>
+#include "expat_config.h"
#include <stdio.h>
#include <stdlib.h>
@@ -71,11 +71,7 @@
# endif
#endif
-#ifdef _DEBUG
-# define READ_SIZE 16
-#else
-# define READ_SIZE (1024 * 8)
-#endif
+int g_read_size_bytes = 1024 * 8;
typedef struct {
XML_Parser parser;
@@ -195,7 +191,7 @@ processStream(const XML_Char *filename, XML_Parser parser) {
}
for (;;) {
int nread;
- char *buf = (char *)XML_GetBuffer(parser, READ_SIZE);
+ char *buf = (char *)XML_GetBuffer(parser, g_read_size_bytes);
if (! buf) {
if (filename != NULL)
close(fd);
@@ -203,7 +199,7 @@ processStream(const XML_Char *filename, XML_Parser parser) {
filename != NULL ? filename : T("xmlwf"));
return 0;
}
- nread = read(fd, buf, READ_SIZE);
+ nread = read(fd, buf, g_read_size_bytes);
if (nread < 0) {
tperror(filename != NULL ? filename : T("STDIN"));
if (filename != NULL)
diff --git a/xmlwf/xmlfile.h b/xmlwf/xmlfile.h
index d75dda222792..579201ccf498 100644
--- a/xmlwf/xmlfile.h
+++ b/xmlwf/xmlfile.h
@@ -10,7 +10,7 @@
Copyright (c) 2000 Clark Cooper <coopercc@users.sourceforge.net>
Copyright (c) 2002 Fred L. Drake, Jr. <fdrake@users.sourceforge.net>
Copyright (c) 2005 Karl Waclawek <karl@waclawek.net>
- Copyright (c) 2016-2019 Sebastian Pipping <sebastian@pipping.org>
+ Copyright (c) 2016-2023 Sebastian Pipping <sebastian@pipping.org>
Licensed under the MIT license:
Permission is hereby granted, free of charge, to any person obtaining
@@ -42,5 +42,7 @@
# define XML_FMT_INT_MOD "l"
#endif
+extern int g_read_size_bytes;
+
extern int XML_ProcessFile(XML_Parser parser, const XML_Char *filename,
unsigned flags);
diff --git a/xmlwf/xmlwf.c b/xmlwf/xmlwf.c
index 471f2a20f57c..7c0a8cd4d6a4 100644
--- a/xmlwf/xmlwf.c
+++ b/xmlwf/xmlwf.c
@@ -11,13 +11,14 @@
Copyright (c) 2001-2003 Fred L. Drake, Jr. <fdrake@users.sourceforge.net>
Copyright (c) 2004-2009 Karl Waclawek <karl@waclawek.net>
Copyright (c) 2005-2007 Steven Solie <steven@solie.ca>
- Copyright (c) 2016-2022 Sebastian Pipping <sebastian@pipping.org>
+ Copyright (c) 2016-2023 Sebastian Pipping <sebastian@pipping.org>
Copyright (c) 2017 Rhodri James <rhodri@wildebeest.org.uk>
Copyright (c) 2019 David Loffredo <loffredo@steptools.com>
Copyright (c) 2020 Joe Orton <jorton@redhat.com>
Copyright (c) 2020 Kleber TarcĂ­sio <klebertarcisio@yahoo.com.br>
Copyright (c) 2021 Tim Bray <tbray@textuality.com>
Copyright (c) 2022 Martin Ettl <ettl.martin78@googlemail.com>
+ Copyright (c) 2022 Sean McBride <sean@rogue-research.com>
Licensed under the MIT license:
Permission is hereby granted, free of charge, to any person obtaining
@@ -40,7 +41,7 @@
USE OR OTHER DEALINGS IN THE SOFTWARE.
*/
-#include <expat_config.h>
+#include "expat_config.h"
#include <assert.h>
#include <stdio.h>
@@ -178,7 +179,7 @@ is equivalent to lexicographically comparing based on the character number. */
static int
attcmp(const void *att1, const void *att2) {
- return tcscmp(*(const XML_Char **)att1, *(const XML_Char **)att2);
+ return tcscmp(*(const XML_Char *const *)att1, *(const XML_Char *const *)att2);
}
static void XMLCALL
@@ -215,8 +216,8 @@ endElement(void *userData, const XML_Char *name) {
static int
nsattcmp(const void *p1, const void *p2) {
- const XML_Char *att1 = *(const XML_Char **)p1;
- const XML_Char *att2 = *(const XML_Char **)p2;
+ const XML_Char *att1 = *(const XML_Char *const *)p1;
+ const XML_Char *att2 = *(const XML_Char *const *)p2;
int sep1 = (tcsrchr(att1, NSSEP) != 0);
int sep2 = (tcsrchr(att2, NSSEP) != 0);
if (sep1 != sep2)
@@ -370,8 +371,8 @@ xcscmp(const XML_Char *xs, const XML_Char *xt) {
static int
notationCmp(const void *a, const void *b) {
- const NotationList *const n1 = *(NotationList **)a;
- const NotationList *const n2 = *(NotationList **)b;
+ const NotationList *const n1 = *(const NotationList *const *)a;
+ const NotationList *const n2 = *(const NotationList *const *)b;
return xcscmp(n1->notationName, n2->notationName);
}
@@ -871,6 +872,9 @@ showVersion(XML_Char *prog) {
}
}
+#if defined(__GNUC__)
+__attribute__((noreturn))
+#endif
static void
usage(const XML_Char *prog, int rc) {
ftprintf(
@@ -883,50 +887,54 @@ usage(const XML_Char *prog, int rc) {
/* clang-format off */
T("usage:\n")
T(" %s [OPTIONS] [FILE ...]\n")
- T(" %s -h\n")
- T(" %s -v\n")
+ T(" %s -h|--help\n")
+ T(" %s -v|--version\n")
T("\n")
T("xmlwf - Determines if an XML document is well-formed\n")
T("\n")
T("positional arguments:\n")
- T(" FILE file to process (default: STDIN)\n")
+ T(" FILE file to process (default: STDIN)\n")
T("\n")
T("input control arguments:\n")
- T(" -s print an error if the document is not [s]tandalone\n")
- T(" -n enable [n]amespace processing\n")
- T(" -p enable processing external DTDs and [p]arameter entities\n")
- T(" -x enable processing of e[x]ternal entities\n")
- T(" -e ENCODING override any in-document [e]ncoding declaration\n")
- T(" -w enable support for [W]indows code pages\n")
- T(" -r disable memory-mapping and use normal file [r]ead IO calls instead\n")
- T(" -k when processing multiple files, [k]eep processing after first file with error\n")
+ T(" -s print an error if the document is not [s]tandalone\n")
+ T(" -n enable [n]amespace processing\n")
+ T(" -p enable processing of external DTDs and [p]arameter entities\n")
+ T(" -x enable processing of e[x]ternal entities\n")
+ T(" -e ENCODING override any in-document [e]ncoding declaration\n")
+ T(" -w enable support for [W]indows code pages\n")
+ T(" -r disable memory-mapping and use [r]ead calls instead\n")
+ T(" -g BYTES buffer size to request per call pair to XML_[G]etBuffer and read (default: 8 KiB)\n")
+ T(" -k when processing multiple files, [k]eep processing after first file with error\n")
T("\n")
T("output control arguments:\n")
- T(" -d DIRECTORY output [d]estination directory\n")
- T(" -c write a [c]opy of input XML, not canonical XML\n")
- T(" -m write [m]eta XML, not canonical XML\n")
- T(" -t write no XML output for [t]iming of plain parsing\n")
- T(" -N enable adding doctype and [n]otation declarations\n")
+ T(" -d DIRECTORY output [d]estination directory\n")
+ T(" -c write a [c]opy of input XML, not canonical XML\n")
+ T(" -m write [m]eta XML, not canonical XML\n")
+ T(" -t write no XML output for [t]iming of plain parsing\n")
+ T(" -N enable adding doctype and [n]otation declarations\n")
T("\n")
T("billion laughs attack protection:\n")
T(" NOTE: If you ever need to increase these values for non-attack payload, please file a bug report.\n")
T("\n")
- T(" -a FACTOR set maximum tolerated [a]mplification factor (default: 100.0)\n")
- T(" -b BYTES set number of output [b]ytes needed to activate (default: 8 MiB)\n")
+ T(" -a FACTOR set maximum tolerated [a]mplification factor (default: 100.0)\n")
+ T(" -b BYTES set number of output [b]ytes needed to activate (default: 8 MiB)\n")
+ T("\n")
+ T("reparse deferral:\n")
+ T(" -q disable reparse deferral, and allow [q]uadratic parse runtime with large tokens\n")
T("\n")
T("info arguments:\n")
- T(" -h show this [h]elp message and exit\n")
- T(" -v show program's [v]ersion number and exit\n")
+ T(" -h, --help show this [h]elp message and exit\n")
+ T(" -v, --version show program's [v]ersion number and exit\n")
T("\n")
T("exit status:\n")
- T(" 0 the input files are well-formed and the output (if requested) was written successfully\n")
- T(" 1 could not allocate data structures, signals a serious problem with execution environment\n")
- T(" 2 one or more input files were not well-formed\n")
- T(" 3 could not create an output file\n")
- T(" 4 command-line argument error\n")
+ T(" 0 the input files are well-formed and the output (if requested) was written successfully\n")
+ T(" 1 could not allocate data structures, signals a serious problem with execution environment\n")
+ T(" 2 one or more input files were not well-formed\n")
+ T(" 3 could not create an output file\n")
+ T(" 4 command-line argument error\n")
T("\n")
T("xmlwf of libexpat is software libre, licensed under the MIT license.\n")
- T("Please report bugs at https://github.com/libexpat/libexpat/issues. Thank you!\n")
+ T("Please report bugs at https://github.com/libexpat/libexpat/issues -- thank you!\n")
, /* clang-format on */
prog, prog, prog);
exit(rc);
@@ -940,8 +948,10 @@ int wmain(int argc, XML_Char **argv);
#define XMLWF_SHIFT_ARG_INTO(constCharStarTarget, argc, argv, i, j) \
{ \
if (argv[i][j + 1] == T('\0')) { \
- if (++i == argc) \
+ if (++i == argc) { \
usage(argv[0], XMLWF_EXIT_USAGE_ERROR); \
+ /* usage called exit(..), never gets here */ \
+ } \
constCharStarTarget = argv[i]; \
} else { \
constCharStarTarget = argv[i] + j + 1; \
@@ -964,9 +974,11 @@ tmain(int argc, XML_Char **argv) {
int continueOnError = 0;
float attackMaximumAmplification = -1.0f; /* signaling "not set" */
- unsigned long long attackThresholdBytes;
+ unsigned long long attackThresholdBytes = 0;
XML_Bool attackThresholdGiven = XML_FALSE;
+ XML_Bool disableDeferral = XML_FALSE;
+
int exitCode = XMLWF_EXIT_SUCCESS;
enum XML_ParamEntityParsing paramEntityParsing
= XML_PARAM_ENTITY_PARSING_NEVER;
@@ -983,9 +995,17 @@ tmain(int argc, XML_Char **argv) {
if (j == 0) {
if (argv[i][0] != T('-'))
break;
- if (argv[i][1] == T('-') && argv[i][2] == T('\0')) {
- i++;
- break;
+ if (argv[i][1] == T('-')) {
+ if (argv[i][2] == T('\0')) {
+ i++;
+ break;
+ } else if (tcscmp(argv[i] + 2, T("help")) == 0) {
+ usage(argv[0], XMLWF_EXIT_SUCCESS);
+ // usage called exit(..), never gets here
+ } else if (tcscmp(argv[i] + 2, T("version")) == 0) {
+ showVersion(argv[0]);
+ return XMLWF_EXIT_SUCCESS;
+ }
}
j++;
}
@@ -1038,10 +1058,30 @@ tmain(int argc, XML_Char **argv) {
break;
case T('h'):
usage(argv[0], XMLWF_EXIT_SUCCESS);
- return 0;
+ // usage called exit(..), never gets here
case T('v'):
showVersion(argv[0]);
- return 0;
+ return XMLWF_EXIT_SUCCESS;
+ case T('g'): {
+ const XML_Char *valueText = NULL;
+ XMLWF_SHIFT_ARG_INTO(valueText, argc, argv, i, j);
+
+ errno = 0;
+ XML_Char *afterValueText = (XML_Char *)valueText;
+ const long long read_size_bytes_candidate
+ = tcstoull(valueText, &afterValueText, 10);
+ if ((errno != 0) || (afterValueText[0] != T('\0'))
+ || (read_size_bytes_candidate < 1)
+ || (read_size_bytes_candidate > (INT_MAX / 2 + 1))) {
+ // This prevents tperror(..) from reporting misleading "[..]: Success"
+ errno = ERANGE;
+ tperror(T("invalid buffer size") T(
+ " (needs an integer from 1 to INT_MAX/2+1 i.e. 1,073,741,824 on most platforms)"));
+ exit(XMLWF_EXIT_USAGE_ERROR);
+ }
+ g_read_size_bytes = (int)read_size_bytes_candidate;
+ break;
+ }
case T('k'):
continueOnError = 1;
j++;
@@ -1051,7 +1091,7 @@ tmain(int argc, XML_Char **argv) {
XMLWF_SHIFT_ARG_INTO(valueText, argc, argv, i, j);
errno = 0;
- XML_Char *afterValueText = (XML_Char *)valueText;
+ XML_Char *afterValueText = NULL;
attackMaximumAmplification = tcstof(valueText, &afterValueText);
if ((errno != 0) || (afterValueText[0] != T('\0'))
|| isnan(attackMaximumAmplification)
@@ -1062,9 +1102,10 @@ tmain(int argc, XML_Char **argv) {
" (needs a floating point number greater or equal than 1.0)"));
exit(XMLWF_EXIT_USAGE_ERROR);
}
-#ifndef XML_DTD
- ftprintf(stderr, T("Warning: Given amplification limit ignored") T(
- ", xmlwf has been compiled without DTD support.\n"));
+#if XML_GE == 0
+ ftprintf(stderr,
+ T("Warning: Given amplification limit ignored")
+ T(", xmlwf has been compiled without DTD/GE support.\n"));
#endif
break;
}
@@ -1083,12 +1124,18 @@ tmain(int argc, XML_Char **argv) {
exit(XMLWF_EXIT_USAGE_ERROR);
}
attackThresholdGiven = XML_TRUE;
-#ifndef XML_DTD
- ftprintf(stderr, T("Warning: Given attack threshold ignored") T(
- ", xmlwf has been compiled without DTD support.\n"));
+#if XML_GE == 0
+ ftprintf(stderr,
+ T("Warning: Given attack threshold ignored")
+ T(", xmlwf has been compiled without DTD/GE support.\n"));
#endif
break;
}
+ case T('q'): {
+ disableDeferral = XML_TRUE;
+ j++;
+ break;
+ }
case T('\0'):
if (j > 1) {
i++;
@@ -1098,6 +1145,7 @@ tmain(int argc, XML_Char **argv) {
/* fall through */
default:
usage(argv[0], XMLWF_EXIT_USAGE_ERROR);
+ // usage called exit(..), never gets here
}
}
if (i == argc) {
@@ -1120,13 +1168,13 @@ tmain(int argc, XML_Char **argv) {
}
if (attackMaximumAmplification != -1.0f) {
-#ifdef XML_DTD
+#if XML_GE == 1
XML_SetBillionLaughsAttackProtectionMaximumAmplification(
parser, attackMaximumAmplification);
#endif
}
if (attackThresholdGiven) {
-#ifdef XML_DTD
+#if XML_GE == 1
XML_SetBillionLaughsAttackProtectionActivationThreshold(
parser, attackThresholdBytes);
#else
@@ -1134,6 +1182,16 @@ tmain(int argc, XML_Char **argv) {
#endif
}
+ if (disableDeferral) {
+ const XML_Bool success = XML_SetReparseDeferralEnabled(parser, XML_FALSE);
+ if (! success) {
+ // This prevents tperror(..) from reporting misleading "[..]: Success"
+ errno = EINVAL;
+ tperror(T("Failed to disable reparse deferral"));
+ exit(XMLWF_EXIT_INTERNAL_ERROR);
+ }
+ }
+
if (requireStandalone)
XML_SetNotStandaloneHandler(parser, notStandalone);
XML_SetParamEntityParsing(parser, paramEntityParsing);
diff --git a/xmlwf/xmlwf_helpgen.py b/xmlwf/xmlwf_helpgen.py
index c2a527fde27d..3d32f5d148b7 100755
--- a/xmlwf/xmlwf_helpgen.py
+++ b/xmlwf/xmlwf_helpgen.py
@@ -6,7 +6,7 @@
# \___/_/\_\ .__/ \__,_|\__|
# |_| XML parser
#
-# Copyright (c) 2019-2021 Sebastian Pipping <sebastian@pipping.org>
+# Copyright (c) 2019-2023 Sebastian Pipping <sebastian@pipping.org>
# Copyright (c) 2021 Tim Bray <tbray@textuality.com>
# Licensed under the MIT license:
#
@@ -33,20 +33,20 @@ import argparse
epilog = """
exit status:
- 0 the input files are well-formed and the output (if requested) was written successfully
- 1 could not allocate data structures, signals a serious problem with execution environment
- 2 one or more input files were not well-formed
- 3 could not create an output file
- 4 command-line argument error
+ 0 the input files are well-formed and the output (if requested) was written successfully
+ 1 could not allocate data structures, signals a serious problem with execution environment
+ 2 one or more input files were not well-formed
+ 3 could not create an output file
+ 4 command-line argument error
xmlwf of libexpat is software libre, licensed under the MIT license.
-Please report bugs at https://github.com/libexpat/libexpat/issues. Thank you!
+Please report bugs at https://github.com/libexpat/libexpat/issues -- thank you!
"""
usage = """
%(prog)s [OPTIONS] [FILE ...]
- %(prog)s -h
- %(prog)s -v
+ %(prog)s -h|--help
+ %(prog)s -v|--version
"""
parser = argparse.ArgumentParser(prog='xmlwf', add_help=False,
@@ -58,11 +58,12 @@ parser = argparse.ArgumentParser(prog='xmlwf', add_help=False,
input_related = parser.add_argument_group('input control arguments')
input_related.add_argument('-s', action='store_true', help='print an error if the document is not [s]tandalone')
input_related.add_argument('-n', action='store_true', help='enable [n]amespace processing')
-input_related.add_argument('-p', action='store_true', help='enable processing external DTDs and [p]arameter entities')
+input_related.add_argument('-p', action='store_true', help='enable processing of external DTDs and [p]arameter entities')
input_related.add_argument('-x', action='store_true', help='enable processing of e[x]ternal entities')
input_related.add_argument('-e', action='store', metavar='ENCODING', help='override any in-document [e]ncoding declaration')
input_related.add_argument('-w', action='store_true', help='enable support for [W]indows code pages')
-input_related.add_argument('-r', action='store_true', help='disable memory-mapping and use normal file [r]ead IO calls instead')
+input_related.add_argument('-r', action='store_true', help='disable memory-mapping and use [r]ead calls instead')
+input_related.add_argument('-g', metavar='BYTES', help='buffer size to request per call pair to XML_[G]etBuffer and read (default: 8 KiB)')
input_related.add_argument('-k', action='store_true', help='when processing multiple files, [k]eep processing after first file with error')
output_related = parser.add_argument_group('output control arguments')
@@ -81,12 +82,16 @@ billion_laughs.add_argument('-a', metavar='FACTOR',
help='set maximum tolerated [a]mplification factor (default: 100.0)')
billion_laughs.add_argument('-b', metavar='BYTES', help='set number of output [b]ytes needed to activate (default: 8 MiB)')
+reparse_deferral = parser.add_argument_group('reparse deferral')
+reparse_deferral.add_argument('-q', metavar='FACTOR',
+ help='disable reparse deferral, and allow [q]uadratic parse runtime with large tokens')
+
parser.add_argument('files', metavar='FILE', nargs='*', help='file to process (default: STDIN)')
info = parser.add_argument_group('info arguments')
info = info.add_mutually_exclusive_group()
-info.add_argument('-h', action='store_true', help='show this [h]elp message and exit')
-info.add_argument('-v', action='store_true', help='show program\'s [v]ersion number and exit')
+info.add_argument('-h', '--help', action='store_true', help='show this [h]elp message and exit')
+info.add_argument('-v', '--version', action='store_true', help='show program\'s [v]ersion number and exit')
if __name__ == '__main__':