diff options
| -rw-r--r-- | CHANGES | 2 | ||||
| -rw-r--r-- | Makefile.in | 14 | ||||
| -rw-r--r-- | encoding/apr_escape.c | 1156 | ||||
| -rw-r--r-- | include/apr_escape.h | 375 | ||||
| -rw-r--r-- | test/Makefile.in | 2 | ||||
| -rw-r--r-- | test/abts_tests.h | 1 | ||||
| -rw-r--r-- | test/testescape.c | 255 | ||||
| -rw-r--r-- | test/testutil.h | 1 | ||||
| -rw-r--r-- | tools/gen_test_char.c | 134 |
9 files changed, 1937 insertions, 3 deletions
@@ -1,6 +1,8 @@ -*- coding: utf-8 -*- Changes for APR 2.0.0 + *) Add the apr_escape interface. [Graham Leggett] + *) Add support to apr_memcache for unix domain sockets. PR 54573 [Remi Gacogne <rgacogne+asf aquaray.com>] diff --git a/Makefile.in b/Makefile.in index 343920f35..87cdc5f84 100644 --- a/Makefile.in +++ b/Makefile.in @@ -43,7 +43,7 @@ APRUTIL_EXPORT_LIBS = @APRUTIL_EXPORT_LIBS@ # building the entire package. # TARGETS = $(TARGET_LIB) $(APR_DSO_MODULES) \ - apr.exp apr-config.out build/apr_rules.out + include/private/apr_escape_test_char.h apr.exp apr-config.out build/apr_rules.out LT_VERSION = @LT_VERSION@ @@ -52,7 +52,7 @@ LT_VERSION = @LT_VERSION@ @INCLUDE_OUTPUTS@ CLEAN_TARGETS = apr-config.out apr.exp exports.c export_vars.c .make.dirs \ - build/apr_rules.out + build/apr_rules.out tools/gen_test_char DISTCLEAN_TARGETS = config.cache config.log config.status \ include/apr.h include/arch/unix/apr_private.h \ libtool $(APR_CONFIG) build/apr_rules.mk apr.pc \ @@ -157,5 +157,15 @@ check: $(TARGET_LIB) etags: etags `find . -name '*.[ch]'` +OBJECTS_gen_test_char = tools/gen_test_char.lo $(LOCAL_LIBS) +tools/gen_test_char@EXEEXT@: $(OBJECTS_gen_test_char) + $(LINK_PROG) $(OBJECTS_gen_test_char) $(ALL_LIBS) + +include/private/apr_escape_test_char.h: tools/gen_test_char + tools/gen_test_char > include/private/apr_escape_test_char.h + +LINK_PROG = $(LIBTOOL) $(LTFLAGS) --mode=link $(COMPILE) $(LT_LDFLAGS) \ + @LT_NO_INSTALL@ $(ALL_LDFLAGS) -o $@ + # DO NOT REMOVE docs: $(INCDIR)/*.h diff --git a/encoding/apr_escape.c b/encoding/apr_escape.c new file mode 100644 index 000000000..0516e276c --- /dev/null +++ b/encoding/apr_escape.c @@ -0,0 +1,1156 @@ +/* Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +/* escape/unescape functions. + * + * These functions perform various escaping operations, and are provided in + * pairs, a function to query the length of and escape existing buffers, as + * well as companion functions to perform the same process to memory + * allocated from a pool. + * + * The API is designed to have the smallest possible RAM footprint, and so + * will only allocate the exact amount of RAM needed for each conversion. + */ + +#include "apr_escape.h" +#include "apr_escape_test_char.h" +#include "apr_lib.h" +#include "apr_strings.h" + +/* helper for Latin1 <-> entity encoding */ +#if APR_CHARSET_EBCDIC +#include "apr_xlate.h" +#define RAW_ASCII_CHAR(ch) apr_xlate_conv_byte(ap_hdrs_from_ascii, \ + (unsigned char)ch) +#else /* APR_CHARSET_EBCDIC */ +#define RAW_ASCII_CHAR(ch) (ch) +#endif /* !APR_CHARSET_EBCDIC */ + +/* we assume the folks using this ensure 0 <= c < 256... which means + * you need a cast to (unsigned char) first, you can't just plug a + * char in here and get it to work, because if char is signed then it + * will first be sign extended. + */ +#define TEST_CHAR(c, f) (test_char_table[(unsigned)(c)] & (f)) + +APR_DECLARE(apr_status_t) apr_escape_shell(char *escaped, const char *str, + apr_ssize_t slen, apr_size_t *len) +{ + unsigned char *d; + const unsigned char *s; + apr_size_t size = 1; + int found = 0; + + d = (unsigned char *) escaped; + s = (const unsigned char *) str; + + if (s) { + if (d) { + for (; *s && slen; ++s, slen--) { +#if defined(OS2) || defined(WIN32) + /* + * Newlines to Win32/OS2 CreateProcess() are ill advised. + * Convert them to spaces since they are effectively white + * space to most applications + */ + if (*s == '\r' || *s == '\n') { + if (d) { + *d++ = ' '; + found = 1; + } + continue; + } +#endif + if (TEST_CHAR(*s, T_ESCAPE_SHELL_CMD)) { + *d++ = '\\'; + size++; + found = 1; + } + *d++ = *s; + size++; + } + *d = '\0'; + } + else { + for (; *s && slen; ++s, slen--) { + if (TEST_CHAR(*s, T_ESCAPE_SHELL_CMD)) { + size++; + found = 1; + } + size++; + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_shell(apr_pool_t *p, const char *str) +{ + apr_size_t len; + + switch (apr_escape_shell(NULL, str, APR_ESCAPE_STRING, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_escape_shell(cmd, str, APR_ESCAPE_STRING, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +static char x2c(const char *what) +{ + register char digit; + +#if !APR_CHARSET_EBCDIC + digit = + ((what[0] >= 'A') ? ((what[0] & 0xdf) - 'A') + 10 : (what[0] - '0')); + digit *= 16; + digit += (what[1] >= 'A' ? ((what[1] & 0xdf) - 'A') + 10 : (what[1] - '0')); +#else /*APR_CHARSET_EBCDIC*/ + char xstr[5]; + xstr[0]='0'; + xstr[1]='x'; + xstr[2]=what[0]; + xstr[3]=what[1]; + xstr[4]='\0'; + digit = apr_xlate_conv_byte(ap_hdrs_from_ascii, + 0xFF & strtol(xstr, NULL, 16)); +#endif /*APR_CHARSET_EBCDIC*/ + return (digit); +} + +APR_DECLARE(apr_status_t) apr_unescape_url(char *escaped, const char *url, + apr_ssize_t slen, const char *forbid, const char *reserved, int plus, + apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const char *s = (const char *) url; + char *d = (char *) escaped; + register int badesc, badpath; + + if (!url) { + return APR_NOTFOUND; + } + + badesc = 0; + badpath = 0; + if (s) { + if (d) { + for (; *s && slen; ++s, d++, slen--) { + if (plus && *s == '+') { + *d = ' '; + found = 1; + } + else if (*s != '%') { + *d = *s; + } + else { + if (!apr_isxdigit(*(s + 1)) || !apr_isxdigit(*(s + 2))) { + badesc = 1; + *d = '%'; + } + else { + char decoded; + decoded = x2c(s + 1); + if ((decoded == '\0') + || (forbid && strchr(forbid, decoded))) { + badpath = 1; + *d = decoded; + s += 2; + slen -= 2; + } + else if (reserved && strchr(reserved, decoded)) { + *d++ = *s++; + *d++ = *s++; + *d = *s; + size += 2; + } + else { + *d = decoded; + s += 2; + slen -= 2; + found = 1; + } + } + } + size++; + } + *d = '\0'; + } + else { + for (; *s && slen; ++s, slen--) { + if (plus && *s == '+') { + found = 1; + } + else if (*s != '%') { + /* character unchanged */ + } + else { + if (!apr_isxdigit(*(s + 1)) || !apr_isxdigit(*(s + 2))) { + badesc = 1; + } + else { + char decoded; + decoded = x2c(s + 1); + if ((decoded == '\0') + || (forbid && strchr(forbid, decoded))) { + badpath = 1; + s += 2; + slen -= 2; + } + else if (reserved && strchr(reserved, decoded)) { + s += 2; + slen -= 2; + size += 2; + } + else { + s += 2; + slen -= 2; + found = 1; + } + } + } + size++; + } + } + } + + if (len) { + *len = size; + } + if (badesc) { + return APR_EINVAL; + } + else if (badpath) { + return APR_BADCH; + } + else if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_punescape_url(apr_pool_t *p, const char *url, + const char *forbid, const char *reserved, int plus) +{ + apr_size_t len; + + switch (apr_unescape_url(NULL, url, APR_ESCAPE_STRING, forbid, reserved, + plus, &len)) { + case APR_SUCCESS: { + char *buf = apr_palloc(p, len); + apr_unescape_url(buf, url, APR_ESCAPE_STRING, forbid, reserved, plus, + NULL); + return buf; + } + case APR_EINVAL: + case APR_BADCH: { + return NULL; + } + case APR_NOTFOUND: { + break; + } + } + + return url; +} + +/* c2x takes an unsigned, and expects the caller has guaranteed that + * 0 <= what < 256... which usually means that you have to cast to + * unsigned char first, because (unsigned)(char)(x) first goes through + * signed extension to an int before the unsigned cast. + * + * The reason for this assumption is to assist gcc code generation -- + * the unsigned char -> unsigned extension is already done earlier in + * both uses of this code, so there's no need to waste time doing it + * again. + */ +static const char c2x_table[] = "0123456789abcdef"; + +static APR_INLINE unsigned char *c2x(unsigned what, unsigned char prefix, + unsigned char *where) +{ +#if APR_CHARSET_EBCDIC + what = apr_xlate_conv_byte(ap_hdrs_to_ascii, (unsigned char)what); +#endif /*APR_CHARSET_EBCDIC*/ + *where++ = prefix; + *where++ = c2x_table[what >> 4]; + *where++ = c2x_table[what & 0xf]; + return where; +} + +APR_DECLARE(apr_status_t) apr_escape_path_segment(char *escaped, + const char *str, apr_ssize_t slen, apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const unsigned char *s = (const unsigned char *) str; + unsigned char *d = (unsigned char *) escaped; + unsigned c; + + if (s) { + if (d) { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_PATH_SEGMENT)) { + d = c2x(c, '%', d); + size += 2; + found = 1; + } + else { + *d++ = c; + } + ++s; + size++; + slen--; + } + *d = '\0'; + } + else { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_PATH_SEGMENT)) { + size += 2; + found = 1; + } + ++s; + size++; + slen--; + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_path_segment(apr_pool_t *p, + const char *str) +{ + apr_size_t len; + + switch (apr_escape_path_segment(NULL, str, APR_ESCAPE_STRING, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_escape_path_segment(cmd, str, APR_ESCAPE_STRING, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +APR_DECLARE(apr_status_t) apr_escape_path(char *escaped, const char *path, + apr_ssize_t slen, int partial, apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const unsigned char *s = (const unsigned char *) path; + unsigned char *d = (unsigned char *) escaped; + unsigned c; + + if (!path) { + return APR_NOTFOUND; + } + + if (!partial) { + const char *colon = strchr(path, ':'); + const char *slash = strchr(path, '/'); + + if (colon && (!slash || colon < slash)) { + if (d) { + *d++ = '.'; + *d++ = '/'; + } + size += 2; + found = 1; + } + } + if (d) { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_OS_ESCAPE_PATH)) { + d = c2x(c, '%', d); + } + else { + *d++ = c; + } + ++s; + size++; + slen--; + } + *d = '\0'; + } + else { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_OS_ESCAPE_PATH)) { + size += 2; + found = 1; + } + ++s; + size++; + slen--; + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_path(apr_pool_t *p, const char *str, + int partial) +{ + apr_size_t len; + + switch (apr_escape_path(NULL, str, APR_ESCAPE_STRING, partial, &len)) { + case APR_SUCCESS: { + char *path = apr_palloc(p, len); + apr_escape_path(path, str, APR_ESCAPE_STRING, partial, NULL); + return path; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +APR_DECLARE(apr_status_t) apr_escape_urlencoded(char *escaped, const char *str, + apr_ssize_t slen, apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const unsigned char *s = (const unsigned char *) str; + unsigned char *d = (unsigned char *) escaped; + unsigned c; + + if (s) { + if (d) { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_URLENCODED)) { + d = c2x(c, '%', d); + size += 2; + found = 1; + } + else if (c == ' ') { + *d++ = '+'; + found = 1; + } + else { + *d++ = c; + } + ++s; + size++; + slen--; + } + *d = '\0'; + } + else { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_URLENCODED)) { + size += 2; + found = 1; + } + else if (c == ' ') { + found = 1; + } + ++s; + size++; + slen--; + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_urlencoded(apr_pool_t *p, const char *str) +{ + apr_size_t len; + + switch (apr_escape_urlencoded(NULL, str, APR_ESCAPE_STRING, &len)) { + case APR_SUCCESS: { + char *encoded = apr_palloc(p, len); + apr_escape_urlencoded(encoded, str, APR_ESCAPE_STRING, NULL); + return encoded; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +APR_DECLARE(apr_status_t) apr_escape_entity(char *escaped, const char *str, + apr_ssize_t slen, int toasc, apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const unsigned char *s = (const unsigned char *) str; + unsigned char *d = (unsigned char *) escaped; + unsigned c; + + if (s) { + if (d) { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_XML)) { + switch (c) { + case '>': { + memcpy(d, ">", 4); + size += 4; + d += 4; + break; + } + case '<': { + memcpy(d, "<", 4); + size += 4; + d += 4; + break; + } + case '&': { + memcpy(d, "&", 5); + size += 5; + d += 5; + break; + } + case '\"': { + memcpy(d, """, 6); + size += 6; + d += 6; + break; + } + case '\'': { + memcpy(d, "'", 6); + size += 6; + d += 6; + break; + } + } + found = 1; + } + else if (toasc && !apr_isascii(c)) { + int offset = apr_snprintf((char *) d, 6, "&#%3.3d;", c); + size += offset; + d += offset; + found = 1; + } + else { + *d++ = c; + size++; + } + ++s; + slen--; + } + *d = '\0'; + } + else { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_XML)) { + switch (c) { + case '>': { + size += 4; + break; + } + case '<': { + size += 4; + break; + } + case '&': { + size += 5; + break; + } + case '\"': { + size += 6; + break; + } + case '\'': { + size += 6; + break; + } + } + found = 1; + } + else if (toasc && !apr_isascii(c)) { + char buf[8]; + size += apr_snprintf(buf, 6, "&#%3.3d;", c); + found = 1; + } + else { + size++; + } + ++s; + slen--; + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_entity(apr_pool_t *p, const char *str, + int toasc) +{ + apr_size_t len; + + switch (apr_escape_entity(NULL, str, APR_ESCAPE_STRING, toasc, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_escape_entity(cmd, str, APR_ESCAPE_STRING, toasc, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +/* maximum length of any ISO-LATIN-1 HTML entity name. */ +#define MAXENTLEN (6) + +APR_DECLARE(apr_status_t) apr_unescape_entity(char *unescaped, const char *str, + apr_ssize_t slen, apr_size_t *len) +{ + int found = 0; + apr_size_t size = 1; + int val, i, j; + char *d = unescaped; + const char *s = str; + const char *ents; + static const char * const entlist[MAXENTLEN + 1] = + { + NULL, /* 0 */ + NULL, /* 1 */ + "lt\074gt\076", /* 2 */ + "amp\046ETH\320eth\360", /* 3 */ + "quot\042Auml\304Euml\313Iuml\317Ouml\326Uuml\334auml\344euml" + "\353iuml\357ouml\366uuml\374yuml\377", /* 4 */ + "Acirc\302Aring\305AElig\306Ecirc\312Icirc\316Ocirc\324Ucirc" + "\333THORN\336szlig\337acirc\342aring\345aelig\346ecirc\352" + "icirc\356ocirc\364ucirc\373thorn\376", /* 5 */ + "Agrave\300Aacute\301Atilde\303Ccedil\307Egrave\310Eacute\311" + "Igrave\314Iacute\315Ntilde\321Ograve\322Oacute\323Otilde" + "\325Oslash\330Ugrave\331Uacute\332Yacute\335agrave\340" + "aacute\341atilde\343ccedil\347egrave\350eacute\351igrave" + "\354iacute\355ntilde\361ograve\362oacute\363otilde\365" + "oslash\370ugrave\371uacute\372yacute\375" /* 6 */ + }; + + if (s) { + if (d) { + for (; *s != '\0' && slen; s++, d++, size++, slen--) { + if (*s != '&') { + *d = *s; + continue; + } + /* find end of entity */ + for (i = 1; s[i] != ';' && s[i] != '\0' && (slen - i) != 0; + i++) { + continue; + } + + if (s[i] == '\0' || (slen - i) == 0) { /* treat as normal data */ + *d = *s; + continue; + } + + /* is it numeric ? */ + if (s[1] == '#') { + for (j = 2, val = 0; j < i && apr_isdigit(s[j]); j++) { + val = val * 10 + s[j] - '0'; + } + s += i; + if (j < i || val <= 8 || (val >= 11 && val <= 31) + || (val >= 127 && val <= 160) || val >= 256) { + d--; /* no data to output */ + size--; + } + else { + *d = RAW_ASCII_CHAR(val); + found = 1; + } + } + else { + j = i - 1; + if (j > MAXENTLEN || entlist[j] == NULL) { + /* wrong length */ + *d = '&'; + continue; /* skip it */ + } + for (ents = entlist[j]; *ents != '\0'; ents += i) { + if (strncmp(s + 1, ents, j) == 0) { + break; + } + } + + if (*ents == '\0') { + *d = '&'; /* unknown */ + } + else { + *d = RAW_ASCII_CHAR(((const unsigned char *) ents)[j]); + s += i; + slen -= i; + found = 1; + } + } + } + *d = '\0'; + } + else { + for (; *s != '\0' && slen; s++, size++, slen--) { + if (*s != '&') { + continue; + } + /* find end of entity */ + for (i = 1; s[i] != ';' && s[i] != '\0' && (slen - i) != 0; + i++) { + continue; + } + + if (s[i] == '\0' || (slen - i) == 0) { /* treat as normal data */ + continue; + } + + /* is it numeric ? */ + if (s[1] == '#') { + for (j = 2, val = 0; j < i && apr_isdigit(s[j]); j++) { + val = val * 10 + s[j] - '0'; + } + s += i; + if (j < i || val <= 8 || (val >= 11 && val <= 31) + || (val >= 127 && val <= 160) || val >= 256) { + /* no data to output */ + size--; + } + else { + found = 1; + } + } + else { + j = i - 1; + if (j > MAXENTLEN || entlist[j] == NULL) { + /* wrong length */ + continue; /* skip it */ + } + for (ents = entlist[j]; *ents != '\0'; ents += i) { + if (strncmp(s + 1, ents, j) == 0) { + break; + } + } + + if (*ents == '\0') { + /* unknown */ + } + else { + s += i; + slen -= i; + found = 1; + } + } + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_punescape_entity(apr_pool_t *p, const char *str) +{ + apr_size_t len; + + switch (apr_unescape_entity(NULL, str, APR_ESCAPE_STRING, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_unescape_entity(cmd, str, APR_ESCAPE_STRING, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +APR_DECLARE(apr_status_t) apr_escape_echo(char *escaped, const char *str, + apr_ssize_t slen, int quote, apr_size_t *len) +{ + apr_size_t size = 1; + int found = 0; + const unsigned char *s = (const unsigned char *) str; + unsigned char *d = (unsigned char *) escaped; + unsigned c; + + if (s) { + if (d) { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_ECHO)) { + *d++ = '\\'; + size++; + switch (c) { + case '\a': + *d++ = 'a'; + size++; + found = 1; + break; + case '\b': + *d++ = 'b'; + size++; + found = 1; + break; + case '\e': + *d++ = 'e'; + size++; + found = 1; + break; + case '\f': + *d++ = 'f'; + size++; + found = 1; + break; + case '\n': + *d++ = 'n'; + size++; + found = 1; + break; + case '\r': + *d++ = 'r'; + size++; + found = 1; + break; + case '\t': + *d++ = 't'; + size++; + found = 1; + break; + case '\v': + *d++ = 'v'; + size++; + found = 1; + break; + case '\\': + *d++ = '\\'; + size++; + found = 1; + break; + case '"': + if (quote) { + *d++ = c; + size++; + found = 1; + } + else { + d[-1] = c; + } + break; + default: + c2x(c, 'x', d); + d += 3; + size += 3; + found = 1; + break; + } + } + else { + *d++ = c; + size++; + } + ++s; + slen--; + } + *d = '\0'; + } + else { + while ((c = *s) && slen) { + if (TEST_CHAR(c, T_ESCAPE_ECHO)) { + size++; + switch (c) { + case '\a': + case '\b': + case '\e': + case '\f': + case '\n': + case '\r': + case '\t': + case '\v': + case '\\': + size++; + found = 1; + break; + case '"': + if (quote) { + size++; + found = 1; + } + break; + default: + size += 3; + found = 1; + break; + } + } + else { + size++; + } + ++s; + slen--; + } + } + } + + if (len) { + *len = size; + } + if (!found) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_echo(apr_pool_t *p, const char *str, + int quote) +{ + apr_size_t len; + + switch (apr_escape_echo(NULL, str, APR_ESCAPE_STRING, quote, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_escape_echo(cmd, str, APR_ESCAPE_STRING, quote, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return str; +} + +APR_DECLARE(apr_status_t) apr_escape_hex(char *dest, const void *src, + apr_size_t srclen, int colon, apr_size_t *len) +{ + const unsigned char *in = src; + apr_size_t size; + + if (!src) { + return APR_NOTFOUND; + } + + if (dest) { + for (size = 0; size < srclen; size++) { + if (colon && size) { + *dest++ = ':'; + } + *dest++ = c2x_table[in[size] >> 4]; + *dest++ = c2x_table[in[size] & 0xf]; + } + *dest = '\0'; + } + + if (len) { + if (colon && srclen) { + *len = srclen * 3; + } + else { + *len = srclen * 2 + 1; + } + } + + return APR_SUCCESS; +} + +APR_DECLARE(const char *) apr_pescape_hex(apr_pool_t *p, const void *src, + apr_size_t srclen, int colon) +{ + apr_size_t len; + + switch (apr_escape_hex(NULL, src, srclen, colon, &len)) { + case APR_SUCCESS: { + char *cmd = apr_palloc(p, len); + apr_escape_hex(cmd, src, srclen, colon, NULL); + return cmd; + } + case APR_NOTFOUND: { + break; + } + } + + return src; +} + +APR_DECLARE(apr_status_t) apr_unescape_hex(void *dest, const char *str, + apr_ssize_t slen, int colon, apr_size_t *len) +{ + apr_size_t size = 0; + int flip = 0; + const unsigned char *s = (const unsigned char *) str; + unsigned char *d = (unsigned char *) dest; + unsigned c; + unsigned char u = 0; + + if (s) { + if (d) { + while ((c = *s) && slen) { + + if (!flip) { + u = 0; + } + + if (colon && c == ':' && !flip) { + ++s; + slen--; + continue; + } + else if (apr_isdigit(c)) { + u |= c - '0'; + } + else if (apr_isupper(c) && c <= 'F') { + u |= c - ('A' - 10); + } + else if (apr_islower(c) && c <= 'f') { + u |= c - ('a' - 10); + } + else { + return APR_BADCH; + } + + if (flip) { + *d++ = u; + size++; + } + else { + u <<= 4; + *d = u; + } + flip = !flip; + + ++s; + slen--; + } + } + else { + while ((c = *s) && slen) { + + if (colon && c == ':' && !flip) { + ++s; + slen--; + continue; + } + else if (apr_isdigit(c)) { + /* valid */ + } + else if (apr_isupper(c) && c <= 'F') { + /* valid */ + } + else if (apr_islower(c) && c <= 'f') { + /* valid */ + } + else { + return APR_BADCH; + } + + if (flip) { + size++; + } + flip = !flip; + + ++s; + slen--; + } + } + } + + if (len) { + *len = size; + } + if (!s) { + return APR_NOTFOUND; + } + + return APR_SUCCESS; +} + +APR_DECLARE(const void *) apr_punescape_hex(apr_pool_t *p, const char *str, + int colon, apr_size_t *len) +{ + apr_size_t size; + + switch (apr_unescape_hex(NULL, str, APR_ESCAPE_STRING, colon, &size)) { + case APR_SUCCESS: { + void *cmd = apr_palloc(p, size); + apr_unescape_hex(cmd, str, APR_ESCAPE_STRING, colon, len); + return cmd; + } + case APR_BADCH: + case APR_NOTFOUND: { + break; + } + } + + return NULL; +} diff --git a/include/apr_escape.h b/include/apr_escape.h new file mode 100644 index 000000000..bc77c3c9d --- /dev/null +++ b/include/apr_escape.h @@ -0,0 +1,375 @@ +/* Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +/** + * @file apr_escape.h + * @brief APR-UTIL Escaping + */ +#ifndef APR_ESCAPE_H +#define APR_ESCAPE_H +#include "apu.h" +#include "apr_general.h" +#ifdef __cplusplus +extern "C" { +#endif + +/** + * @defgroup APR_Util_Escaping Escape functions + * @ingroup APR + * @{ + */ + +/* Simple escape/unescape functions. + * + */ + +/** + * When passing a string to one of the escape functions, this value can be + * passed to indicate a string-valued key, and have the length computed + * automatically. + */ +#define APR_ESCAPE_STRING (-1) + +/** + * Perform shell escaping on the provided string. + * + * Shell escaping causes characters to be prefixed with a '\' character. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_shell(char *escaped, const char *str, + apr_ssize_t slen, apr_size_t *len); + +/** + * Perform shell escaping on the provided string, returning the result + * from the pool. + * + * Shell escaping causes characters to be prefixed with a '\' character. + * + * If no characters were escaped, the original string is returned. + * @param p Pool to allocate from + * @param str The original string + * @return the encoded string, allocated from the pool, or the original + * string if no escaping took place or the string was NULL. + */ +APR_DECLARE(const char *) apr_pescape_shell(apr_pool_t *p, const char *str) + __attribute__((nonnull(1))); + +/* + * Unescapes a URL, leaving reserved characters intact. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param url String to be unescaped + * @param slen The length of the original url, or APR_ESCAPE_STRING + * @param forbid Optional list of forbidden characters, in addition to + * 0x00 + * @param reserved Optional list of reserved characters that will be + * left unescaped + * @param plus If non zero, '+' is converted to ' ' as per + * application/x-www-form-urlencoded encoding + * @param len If set, the length of the escaped string will be returned + * @return APR_SUCCESS on success, APR_NOTFOUND if no characters are + * decoded or the string is NULL, APR_EINVAL if a bad escape sequence is + * found, APR_BADCH if a character on the forbid list is found. + */ +APR_DECLARE(apr_status_t) apr_unescape_url(char *escaped, const char *url, + apr_ssize_t slen, const char *forbid, const char *reserved, int plus, + apr_size_t *len); + +/* + * Unescapes a URL, leaving reserved characters intact, returning the + * result from a pool. + * @param p Pool to allocate from + * @param url String to be unescaped in place + * @param forbid Optional list of forbidden characters, in addition to + * 0x00 + * @param reserved Optional list of reserved characters that will be + * left unescaped + * @param plus If non zero, '+' is converted to ' ' as per + * application/x-www-form-urlencoded encoding + * @return A string allocated from the pool on success, the original string + * if no characters are decoded, or NULL if a bad escape sequence is found + * or if a character on the forbid list is found, or if the original string + * was NULL. + */ +APR_DECLARE(const char *) apr_punescape_url(apr_pool_t *p, const char *url, + const char *forbid, const char *reserved, int plus) + __attribute__((nonnull(1))); + +/** + * Escape a path segment, as defined in RFC1808. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_path_segment(char *escaped, + const char *str, apr_ssize_t slen, apr_size_t *len); + +/** + * Escape a path segment, as defined in RFC1808, returning the result from a + * pool. + * @param p Pool to allocate from + * @param url String to be escaped + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or the string is NULL. + */ +APR_DECLARE(const char *) apr_pescape_path_segment(apr_pool_t *p, + const char *str) __attribute__((nonnull(1))); + +/** + * Converts an OS path to a URL, in an OS dependent way, as defined in RFC1808. + * In all cases if a ':' occurs before the first '/' in the URL, the URL should + * be prefixed with "./" (or the ':' escaped). In the case of Unix, this means + * leaving '/' alone, but otherwise doing what escape_path_segment() does. For + * efficiency reasons, we don't use escape_path_segment(), which is provided for + * reference. Again, RFC 1808 is where this stuff is defined. + * + * If partial is set, os_escape_path() assumes that the path will be appended to + * something with a '/' in it (and thus does not prefix "./"). + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param partial If non zero, suppresses the prepending of "./" + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or if the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_path(char *escaped, const char *path, + apr_ssize_t slen, int partial, apr_size_t *len); + +/** + * Converts an OS path to a URL, in an OS dependent way, as defined in RFC1808, + * returning the result from a pool. + * + * In all cases if a ':' occurs before the first '/' in the URL, the URL should + * be prefixed with "./" (or the ':' escaped). In the case of Unix, this means + * leaving '/' alone, but otherwise doing what escape_path_segment() does. For + * efficiency reasons, we don't use escape_path_segment(), which is provided for + * reference. Again, RFC 1808 is where this stuff is defined. + * + * If partial is set, os_escape_path() assumes that the path will be appended to + * something with a '/' in it (and thus does not prefix "./"). + * @param p Pool to allocate from + * @param str The original string + * @param partial If non zero, suppresses the prepending of "./" + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or if the string was NULL. + */ +APR_DECLARE(const char *) apr_pescape_path(apr_pool_t *p, const char *str, + int partial) __attribute__((nonnull(1))); + +/** + * Urlencode a string, as defined in + * http://www.w3.org/TR/html401/interact/forms.html#h-17.13.4.1. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or if the stirng was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_urlencoded(char *escaped, const char *str, + apr_ssize_t slen, apr_size_t *len); + +/** + * Urlencode a string, as defined in + * http://www.w3.org/TR/html401/interact/forms.html#h-17.13.4.1, returning + * the result from a pool. + * @param p Pool to allocate from + * @param url String to be escaped + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or if the string was NULL. + */ +APR_DECLARE(const char *) apr_pescape_urlencoded(apr_pool_t *p, + const char *str) __attribute__((nonnull(1))); + +/** + * Apply entity encoding to a string. Characters are replaced as follows: + * '<' becomes '<', '>' becomes '>', '&' becomes '&', the + * double quote becomes '"" and the single quote becomes '''. + * + * If toasc is not zero, any non ascii character will be encoded as + * '%#ddd;', where ddd is the decimal code of the character. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param toasc If non zero, encode non ascii characters + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_entity(char *escaped, const char *str, + apr_ssize_t slen, int toasc, apr_size_t *len); + +/** + * Apply entity encoding to a string, returning the result from a pool. + * Characters are replaced as follows: '<' becomes '<', '>' becomes + * '>', '&' becomes '&', the double quote becomes '"" and the + * single quote becomes '''. + * @param p Pool to allocate from + * @param str The original string + * @param toasc If non zero, encode non ascii characters + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or the string is NULL. + */ +APR_DECLARE(const char *) apr_pescape_entity(apr_pool_t *p, const char *str, + int toasc) __attribute__((nonnull(1))); + +/* + * Decodes html entities or numeric character references in a string. If + * the string to be unescaped is syntactically incorrect, then the + * following fixups will be made: + * unknown entities will be left undecoded; + * references to unused numeric characters will be deleted. + * In particular, � will not be decoded, but will be deleted. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or the string was NULL + */ +APR_DECLARE(apr_status_t) apr_unescape_entity(char *unescaped, const char *str, + apr_ssize_t slen, apr_size_t *len); + +/* + * Decodes html entities or numeric character references in a string. If + * the string to be unescaped is syntactically incorrect, then the + * following fixups will be made: + * unknown entities will be left undecoded; + * references to unused numeric characters will be deleted. + * In particular, � will not be decoded, but will be deleted. + * @param p Pool to allocate from + * @param str The original string + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or the string is NULL. + */ +APR_DECLARE(const char *) apr_punescape_entity(apr_pool_t *p, const char *str) + __attribute__((nonnull(1))); + +/** + * Escape control characters in a string, as performed by the shell's + * 'echo' command. Characters are replaced as follows: + * \a alert (bell), \b backspace, \e an escape character, \f form feed, + * \n new line, \r carriage return, \t horizontal tab, \v vertical tab, + * \\ backslash. + * + * Any non ascii character will be encoded as '\xHH', where HH is the hex + * code of the character. + * + * If quote is not zero, the double quote character will also be escaped. + * @param escaped Optional buffer to write the encoded string, can be + * NULL + * @param str The original string + * @param slen The length of the original string, or APR_ESCAPE_STRING + * @param quote If non zero, encode double quotes + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if no changes to the string were + * detected or the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_echo(char *escaped, const char *str, + apr_ssize_t slen, int quote, apr_size_t *len); + +/** + * Escape control characters in a string, as performed by the shell's + * 'echo' command, and return the results from a pool. Characters are + * replaced as follows: \a alert (bell), \b backspace, \e an escape + * character, \f form feed, \n new line, \r carriage return, \t + * horizontal tab, \v vertical tab, \\ backslash. + * + * Any non ascii character will be encoded as '\xHH', where HH is the hex + * code of the character. + * + * If quote is not zero, the double quote character will also be escaped. + * @param p Pool to allocate from + * @param str The original string + * @param quote If non zero, encode double quotes + * @return A string allocated from the pool on success, the original string + * if no characters are encoded or the string is NULL. + */ +APR_DECLARE(const char *) apr_pescape_echo(apr_pool_t *p, const char *str, + int quote); + +/** + * Convert binary data to a hex encoding. + * @param dest The destination buffer, can be NULL + * @param src The original buffer + * @param slen The length of the original buffer + * @param colon If not zero, insert colon characters between hex digits. + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if the string was NULL + */ +APR_DECLARE(apr_status_t) apr_escape_hex(char *dest, const void *src, + apr_size_t srclen, int colon, apr_size_t *len); + +/** + * Convert binary data to a hex encoding, and return the results from a + * pool. + * @param p Pool to allocate from + * @param src The original buffer + * @param srclen The length of the original buffer + * @param colon If not zero, insert colon characters between hex digits. + * @return A zero padded buffer allocated from the pool on success, or + * NULL if src was NULL. + */ +APR_DECLARE(const char *) apr_pescape_hex(apr_pool_t *p, const void *src, + apr_size_t slen, int colon) __attribute__((nonnull(1))); + +/** + * Convert hex encoded string to binary data. + * @param dest The destination buffer, can be NULL + * @param src The original buffer + * @param slen The length of the original buffer + * @param colon If not zero, ignore colon characters between hex digits. + * @param len If present, returns the length of the string + * @return APR_SUCCESS, or APR_NOTFOUND if the string was NULL, or APR_BADCH + * if a non hex character is present. + */ +APR_DECLARE(apr_status_t) apr_unescape_hex(void *dest, const char *str, + apr_ssize_t slen, int colon, apr_size_t *len); + +/** + * Convert hex encoding to binary data, and return the results from a pool. + * If the colon character appears between pairs of hex digits, it will be + * ignored. + * @param p Pool to allocate from + * @param str The original string + * @param colon If not zero, ignore colon characters between hex digits. + * @param len If present, returns the length of the final buffer + * @return A buffer allocated from the pool on success, or NULL if src was + * NULL, or a bad character was present. + */ +APR_DECLARE(const void *) apr_punescape_hex(apr_pool_t *p, const char *str, + int colon, apr_size_t *len); + +/** @} */ +#ifdef __cplusplus +} +#endif + +#endif /* !APR_ESCAPE_H */ diff --git a/test/Makefile.in b/test/Makefile.in index 83a88dd8a..d0620934b 100644 --- a/test/Makefile.in +++ b/test/Makefile.in @@ -35,7 +35,7 @@ TESTS = testtime.lo teststr.lo testvsn.lo testipsub.lo testshm.lo \ teststrmatch.lo testpass.lo testcrypto.lo testqueue.lo \ testbuckets.lo testxml.lo testdbm.lo testuuid.lo testmd5.lo \ testreslist.lo testbase64.lo testhooks.lo testlfsabi.lo \ - testlfsabi32.lo testlfsabi64.lo + testlfsabi32.lo testlfsabi64.lo testescape.lo OTHER_PROGRAMS = \ sendfile@EXEEXT@ \ diff --git a/test/abts_tests.h b/test/abts_tests.h index 09713659c..88a4a925c 100644 --- a/test/abts_tests.h +++ b/test/abts_tests.h @@ -28,6 +28,7 @@ const struct testlist { {testdso}, {testdup}, {testenv}, + {testescape}, {testfile}, {testfilecopy}, {testfileinfo}, diff --git a/test/testescape.c b/test/testescape.c new file mode 100644 index 000000000..bbe4ad8a5 --- /dev/null +++ b/test/testescape.c @@ -0,0 +1,255 @@ +/* Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +#include <assert.h> +#include <stdio.h> +#include <stdlib.h> + +#include "apr_escape.h" +#include "apr_strings.h" + +#include "abts.h" +#include "testutil.h" + +static void test_escape(abts_case *tc, void *data) +{ + apr_pool_t *pool; + const char *src, *target; + const char *dest; + const void *vdest; + apr_size_t len, vlen; + + apr_pool_create(&pool, NULL); + + src = "Hello World &;`'\"|*?~<>^()[]{}$\\\n"; + target = "Hello World \\&\\;\\`\\'\\\"\\|\\*\\?\\~\\<\\>\\^\\(\\)\\[\\]\\{\\}\\$\\\\\\\n"; + dest = apr_pescape_shell(pool, src); + ABTS_ASSERT(tc, "shell escaped matches expected output", + (strcmp(dest, target) == 0)); + apr_escape_shell(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_punescape_url(pool, src, NULL, NULL, 0); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "Hello"; + dest = apr_punescape_url(pool, src, NULL, NULL, 1); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "Hello%20"; + dest = apr_punescape_url(pool, src, " ", NULL, 0); + ABTS_PTR_EQUAL(tc, NULL, dest); + + src = "Hello%20World"; + target = "Hello World"; + dest = apr_punescape_url(pool, src, NULL, NULL, 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_url(NULL, src, APR_ESCAPE_STRING, NULL, NULL, 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello+World"; + target = "Hello World"; + dest = apr_punescape_url(pool, src, NULL, NULL, 1); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_url(NULL, src, APR_ESCAPE_STRING, NULL, NULL, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello%20World"; + target = "Hello%20World"; + dest = apr_punescape_url(pool, src, NULL, " ", 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_url(NULL, src, APR_ESCAPE_STRING, NULL, " ", 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_path_segment(pool, src); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "$-_.+!*'(),:@&=/~Hello World"; + target = "$-_.+!*'(),:@&=%2f~Hello%20World"; + dest = apr_pescape_path_segment(pool, src); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_path_segment(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_path(pool, src, 0); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "$-_.+!*'(),:@&=/~Hello World"; + target = "./$-_.+!*'(),:@&=/~Hello%20World"; + dest = apr_pescape_path(pool, src, 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_path(NULL, src, APR_ESCAPE_STRING, 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_path(pool, src, 1); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "$-_.+!*'(),:@&=/~Hello World"; + target = "$-_.+!*'(),:@&=/~Hello%20World"; + dest = apr_pescape_path(pool, src, 1); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_path(NULL, src, APR_ESCAPE_STRING, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_urlencoded(pool, src); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "$-_.+!*'(),:@&=/~Hello World"; + target = "%24-_.%2b%21*%27%28%29%2c%3a%40%26%3d%2f%7eHello+World"; + dest = apr_pescape_urlencoded(pool, src); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_urlencoded(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_entity(pool, src, 0); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "\xFF<>&\'\"Hello World"; + target = "\xFF<>&'"Hello World"; + dest = apr_pescape_entity(pool, src, 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_entity(NULL, src, APR_ESCAPE_STRING, 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_entity(pool, src, 1); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "\xFF<>&\'\"Hello World"; + target = "ÿ<>&'"Hello World"; + dest = apr_pescape_entity(pool, src, 1); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_entity(NULL, src, APR_ESCAPE_STRING, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_punescape_entity(pool, src); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "\xFF<>&'"Hello World"; + target = "\xFF<>&\'\"Hello World"; + dest = apr_punescape_entity(pool, src); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_entity(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "ÿ<>&'"Hello World"; + target = "\xFF<>&\'\"Hello World"; + dest = apr_punescape_entity(pool, src); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_entity(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = " <>&'"Hello World"; + target = " <>&\'\"Hello World"; + dest = apr_punescape_entity(pool, src); + ABTS_STR_EQUAL(tc, target, dest); + apr_unescape_entity(NULL, src, APR_ESCAPE_STRING, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "Hello"; + dest = apr_pescape_echo(pool, src, 0); + ABTS_PTR_EQUAL(tc, src, dest); + + src = "\a\b\e\f\\n\r\t\v\"Hello World\""; + target = "\\a\\b\\e\\f\\\\n\\r\\t\\v\"Hello World\""; + dest = apr_pescape_echo(pool, src, 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_echo(NULL, src, APR_ESCAPE_STRING, 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "\a\b\e\f\\n\r\t\v\"Hello World\""; + target = "\\a\\b\\e\\f\\\\n\\r\\t\\v\\\"Hello World\\\""; + dest = apr_pescape_echo(pool, src, 1); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_echo(NULL, src, APR_ESCAPE_STRING, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "\xFF\x00\xFF\x00"; + target = "ff00ff00"; + dest = apr_pescape_hex(pool, src, 4, 0); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_hex(NULL, src, 4, 0, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "\xFF\x00\xFF\x00"; + target = "ff:00:ff:00"; + dest = apr_pescape_hex(pool, src, 4, 1); + ABTS_STR_EQUAL(tc, target, dest); + apr_escape_hex(NULL, src, 4, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, strlen(dest) + 1), + (len == strlen(dest) + 1)); + + src = "ff:00:ff:00"; + target = "\xFF\x00\xFF\x00"; + vdest = apr_punescape_hex(pool, src, 1, &vlen); + ABTS_ASSERT(tc, "apr_punescape_hex target!=dest", memcmp(target, vdest, 4) == 0); + ABTS_INT_EQUAL(tc, (int)vlen, 4); + apr_unescape_hex(NULL, src, APR_ESCAPE_STRING, 1, &len); + ABTS_ASSERT(tc, + apr_psprintf(pool, "size mismatch (%" APR_SIZE_T_FMT "!=%" APR_SIZE_T_FMT ")", len, (apr_size_t)4), + (len == 4)); + + apr_pool_destroy(pool); +} + +abts_suite *testescape(abts_suite *suite) +{ + suite = ADD_SUITE(suite); + + abts_run_test(suite, test_escape, NULL); + + return suite; +} diff --git a/test/testutil.h b/test/testutil.h index 8bb508d28..286d253bd 100644 --- a/test/testutil.h +++ b/test/testutil.h @@ -110,6 +110,7 @@ abts_suite *testud(abts_suite *suite); abts_suite *testuser(abts_suite *suite); abts_suite *testvsn(abts_suite *suite); +abts_suite *testescape(abts_suite *suite); abts_suite *teststrmatch(abts_suite *suite); abts_suite *testuri(abts_suite *suite); abts_suite *testuuid(abts_suite *suite); diff --git a/tools/gen_test_char.c b/tools/gen_test_char.c new file mode 100644 index 000000000..fca9850d6 --- /dev/null +++ b/tools/gen_test_char.c @@ -0,0 +1,134 @@ +/* Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +#ifdef CROSS_COMPILE + +#define apr_isalnum(c) (isalnum(((unsigned char)(c)))) +#define apr_isalpha(c) (isalpha(((unsigned char)(c)))) +#define apr_iscntrl(c) (iscntrl(((unsigned char)(c)))) +#define apr_isprint(c) (isprint(((unsigned char)(c)))) +#include <ctype.h> +#define APR_HAVE_STDIO_H 1 +#define APR_HAVE_STRING_H 1 + +#else + +#include "apr.h" +#include "apr_lib.h" + +#endif + +#if defined(WIN32) || defined(OS2) +#define NEED_ENHANCED_ESCAPES +#endif + +#if APR_HAVE_STDIO_H +#include <stdio.h> +#endif +#if APR_HAVE_STRING_H +#include <string.h> +#endif + +/* A bunch of functions in util.c scan strings looking for certain characters. + * To make that more efficient we encode a lookup table. + */ +#define T_ESCAPE_SHELL_CMD (0x01) +#define T_ESCAPE_PATH_SEGMENT (0x02) +#define T_OS_ESCAPE_PATH (0x04) +#define T_ESCAPE_ECHO (0x08) +#define T_ESCAPE_URLENCODED (0x10) +#define T_ESCAPE_XML (0x20) + +int main(int argc, char *argv[]) +{ + unsigned c; + unsigned char flags; + + printf("/* this file is automatically generated by gen_test_char, " + "do not edit. \"make include/private/apr_escape_test_char.h\" to regenerate. */\n" + "#define T_ESCAPE_SHELL_CMD (%u)\n" + "#define T_ESCAPE_PATH_SEGMENT (%u)\n" + "#define T_OS_ESCAPE_PATH (%u)\n" + "#define T_ESCAPE_ECHO (%u)\n" + "#define T_ESCAPE_URLENCODED (%u)\n" + "#define T_ESCAPE_XML (%u)\n" + "\n" + "static const unsigned char test_char_table[256] = {", + T_ESCAPE_SHELL_CMD, + T_ESCAPE_PATH_SEGMENT, + T_OS_ESCAPE_PATH, + T_ESCAPE_ECHO, + T_ESCAPE_URLENCODED, + T_ESCAPE_XML); + + for (c = 0; c < 256; ++c) { + flags = 0; + if (c % 20 == 0) + printf("\n "); + + /* escape_shell_cmd */ +#ifdef NEED_ENHANCED_ESCAPES + /* Win32/OS2 have many of the same vulnerable characters + * as Unix sh, plus the carriage return and percent char. + * The proper escaping of these characters varies from unix + * since Win32/OS2 use carets or doubled-double quotes, + * and neither lf nor cr can be escaped. We escape unix + * specific as well, to assure that cross-compiled unix + * applications behave similiarly when invoked on win32/os2. + * + * Rem please keep in-sync with apr's list in win32/filesys.c + */ + if (c && strchr("&;`'\"|*?~<>^()[]{}$\\\n\r%", c)) { + flags |= T_ESCAPE_SHELL_CMD; + } +#else + if (c && strchr("&;`'\"|*?~<>^()[]{}$\\\n", c)) { + flags |= T_ESCAPE_SHELL_CMD; + } +#endif + + if (!apr_isalnum(c) && !strchr("$-_.+!*'(),:@&=~", c)) { + flags |= T_ESCAPE_PATH_SEGMENT; + } + + if (!apr_isalnum(c) && !strchr("$-_.+!*'(),:@&=/~", c)) { + flags |= T_OS_ESCAPE_PATH; + } + + if (!apr_isalnum(c) && !strchr(".-*_ ", c)) { + flags |= T_ESCAPE_URLENCODED; + } + + /* For logging, escape all control characters, + * double quotes (because they delimit the request in the log file) + * backslashes (because we use backslash for escaping) + * and 8-bit chars with the high bit set + */ + if (c && (!apr_isprint(c) || c == '"' || c == '\\' || apr_iscntrl(c))) { + flags |= T_ESCAPE_ECHO; + } + + if (strchr("<>&\"", c)) { + flags |= T_ESCAPE_XML; + } + + printf("%u%c", flags, (c < 255) ? ',' : ' '); + } + + printf("\n};\n"); + + return 0; +} |