forked from osmocom/wireshark
1233 lines
33 KiB
C
1233 lines
33 KiB
C
/* strutil.c
|
|
* String utility routines
|
|
*
|
|
* Wireshark - Network traffic analyzer
|
|
* By Gerald Combs <gerald@wireshark.org>
|
|
* Copyright 1998 Gerald Combs
|
|
*
|
|
* SPDX-License-Identifier: GPL-2.0-or-later
|
|
*/
|
|
|
|
#include "config.h"
|
|
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
#include <glib.h>
|
|
#include "strutil.h"
|
|
|
|
#include <wsutil/str_util.h>
|
|
#include <epan/proto.h>
|
|
|
|
#ifdef _WIN32
|
|
#include <windows.h>
|
|
#include <tchar.h>
|
|
#include <wchar.h>
|
|
#endif
|
|
|
|
static const char hex[16] = { '0', '1', '2', '3', '4', '5', '6', '7',
|
|
'8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
|
|
|
|
/*
|
|
* Given a pointer into a data buffer, and to the end of the buffer,
|
|
* find the end of the (putative) line at that position in the data
|
|
* buffer.
|
|
* Return a pointer to the EOL character(s) in "*eol".
|
|
*/
|
|
const guchar *
|
|
find_line_end(const guchar *data, const guchar *dataend, const guchar **eol)
|
|
{
|
|
const guchar *lineend;
|
|
|
|
lineend = (guchar *)memchr(data, '\n', dataend - data);
|
|
if (lineend == NULL) {
|
|
/*
|
|
* No LF - line is probably continued in next TCP segment.
|
|
*/
|
|
lineend = dataend;
|
|
*eol = dataend;
|
|
} else {
|
|
/*
|
|
* Is the LF at the beginning of the line?
|
|
*/
|
|
if (lineend > data) {
|
|
/*
|
|
* No - is it preceded by a carriage return?
|
|
* (Perhaps it's supposed to be, but that's not guaranteed....)
|
|
*/
|
|
if (*(lineend - 1) == '\r') {
|
|
/*
|
|
* Yes. The EOL starts with the CR.
|
|
*/
|
|
*eol = lineend - 1;
|
|
} else {
|
|
/*
|
|
* No. The EOL starts with the LF.
|
|
*/
|
|
*eol = lineend;
|
|
|
|
/*
|
|
* I seem to remember that we once saw lines ending with LF-CR
|
|
* in an HTTP request or response, so check if it's *followed*
|
|
* by a carriage return.
|
|
*/
|
|
if (lineend < (dataend - 1) && *(lineend + 1) == '\r') {
|
|
/*
|
|
* It's <non-LF><LF><CR>; say it ends with the CR.
|
|
*/
|
|
lineend++;
|
|
}
|
|
}
|
|
} else {
|
|
/*
|
|
* Yes - the EOL starts with the LF.
|
|
*/
|
|
*eol = lineend;
|
|
}
|
|
|
|
/*
|
|
* Point to the character after the last character.
|
|
*/
|
|
lineend++;
|
|
}
|
|
return lineend;
|
|
}
|
|
|
|
/*
|
|
* Get the length of the next token in a line, and the beginning of the
|
|
* next token after that (if any).
|
|
* Return 0 if there is no next token.
|
|
*/
|
|
int
|
|
get_token_len(const guchar *linep, const guchar *lineend,
|
|
const guchar **next_token)
|
|
{
|
|
const guchar *tokenp;
|
|
int token_len;
|
|
|
|
tokenp = linep;
|
|
|
|
/*
|
|
* Search for a blank, a CR or an LF, or the end of the buffer.
|
|
*/
|
|
while (linep < lineend && *linep != ' ' && *linep != '\r' && *linep != '\n')
|
|
linep++;
|
|
token_len = (int) (linep - tokenp);
|
|
|
|
/*
|
|
* Skip trailing blanks.
|
|
*/
|
|
while (linep < lineend && *linep == ' ')
|
|
linep++;
|
|
|
|
*next_token = linep;
|
|
|
|
return token_len;
|
|
}
|
|
|
|
|
|
#define INITIAL_FMTBUF_SIZE 128
|
|
|
|
/*
|
|
* Given a string, generate a string from it that shows non-printable
|
|
* characters as C-style escapes, and return a pointer to it.
|
|
*/
|
|
gchar *
|
|
format_text(wmem_allocator_t* allocator, const guchar *string, size_t len)
|
|
{
|
|
gchar *fmtbuf = (gchar*)wmem_alloc(allocator, INITIAL_FMTBUF_SIZE);
|
|
int fmtbuf_len = INITIAL_FMTBUF_SIZE;
|
|
int column;
|
|
const guchar *stringend = string + len;
|
|
guchar c;
|
|
int i;
|
|
|
|
column = 0;
|
|
while (string < stringend) {
|
|
/*
|
|
* Is there enough room for this character, if it expands to
|
|
* a backslash plus 3 octal digits (which is the most it can
|
|
* expand to), and also enough room for a terminating '\0'?
|
|
*/
|
|
if (column+3+1 >= fmtbuf_len) {
|
|
/*
|
|
* Double the buffer's size if it's not big enough.
|
|
* The size of the buffer starts at 128, so doubling its size
|
|
* adds at least another 128 bytes, which is more than enough
|
|
* for one more character plus a terminating '\0'.
|
|
*/
|
|
fmtbuf_len *= 2;
|
|
fmtbuf = (gchar *)wmem_realloc(allocator, fmtbuf, fmtbuf_len);
|
|
}
|
|
c = *string++;
|
|
|
|
if (g_ascii_isprint(c)) {
|
|
fmtbuf[column] = c;
|
|
column++;
|
|
} else {
|
|
fmtbuf[column] = '\\';
|
|
column++;
|
|
switch (c) {
|
|
|
|
case '\a':
|
|
fmtbuf[column] = 'a';
|
|
column++;
|
|
break;
|
|
|
|
case '\b':
|
|
fmtbuf[column] = 'b'; /* BS */
|
|
column++;
|
|
break;
|
|
|
|
case '\f':
|
|
fmtbuf[column] = 'f'; /* FF */
|
|
column++;
|
|
break;
|
|
|
|
case '\n':
|
|
fmtbuf[column] = 'n'; /* NL */
|
|
column++;
|
|
break;
|
|
|
|
case '\r':
|
|
fmtbuf[column] = 'r'; /* CR */
|
|
column++;
|
|
break;
|
|
|
|
case '\t':
|
|
fmtbuf[column] = 't'; /* tab */
|
|
column++;
|
|
break;
|
|
|
|
case '\v':
|
|
fmtbuf[column] = 'v';
|
|
column++;
|
|
break;
|
|
|
|
default:
|
|
i = (c>>6)&03;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
i = (c>>3)&07;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
i = (c>>0)&07;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
fmtbuf[column] = '\0';
|
|
return fmtbuf;
|
|
}
|
|
|
|
|
|
/*
|
|
* Given a string, generate a string from it that shows non-printable
|
|
* characters as C-style escapes except a whitespace character
|
|
* (space, tab, carriage return, new line, vertical tab, or formfeed)
|
|
* which will be replaced by a space, and return a pointer to it.
|
|
*/
|
|
gchar *
|
|
format_text_wsp(wmem_allocator_t* allocator, const guchar *string, size_t len)
|
|
{
|
|
gchar *fmtbuf = (gchar*)wmem_alloc(allocator, INITIAL_FMTBUF_SIZE);
|
|
int fmtbuf_len = INITIAL_FMTBUF_SIZE;
|
|
int column;
|
|
const guchar *stringend = string + len;
|
|
guchar c;
|
|
int i;
|
|
|
|
column = 0;
|
|
while (string < stringend) {
|
|
/*
|
|
* Is there enough room for this character, if it expands to
|
|
* a backslash plus 3 octal digits (which is the most it can
|
|
* expand to), and also enough room for a terminating '\0'?
|
|
*/
|
|
if (column+3+1 >= fmtbuf_len) {
|
|
/*
|
|
* Double the buffer's size if it's not big enough.
|
|
* The size of the buffer starts at 128, so doubling its size
|
|
* adds at least another 128 bytes, which is more than enough
|
|
* for one more character plus a terminating '\0'.
|
|
*/
|
|
fmtbuf_len *= 2;
|
|
fmtbuf = (gchar *)wmem_realloc(allocator, fmtbuf, fmtbuf_len);
|
|
}
|
|
c = *string++;
|
|
|
|
if (g_ascii_isprint(c)) {
|
|
fmtbuf[column] = c;
|
|
column++;
|
|
} else if (g_ascii_isspace(c)) {
|
|
fmtbuf[column] = ' ';
|
|
column++;
|
|
} else {
|
|
fmtbuf[column] = '\\';
|
|
column++;
|
|
switch (c) {
|
|
|
|
case '\a':
|
|
fmtbuf[column] = 'a';
|
|
column++;
|
|
break;
|
|
|
|
case '\b':
|
|
fmtbuf[column] = 'b'; /* BS */
|
|
column++;
|
|
break;
|
|
|
|
case '\f':
|
|
fmtbuf[column] = 'f'; /* FF */
|
|
column++;
|
|
break;
|
|
|
|
case '\n':
|
|
fmtbuf[column] = 'n'; /* NL */
|
|
column++;
|
|
break;
|
|
|
|
case '\r':
|
|
fmtbuf[column] = 'r'; /* CR */
|
|
column++;
|
|
break;
|
|
|
|
case '\t':
|
|
fmtbuf[column] = 't'; /* tab */
|
|
column++;
|
|
break;
|
|
|
|
case '\v':
|
|
fmtbuf[column] = 'v';
|
|
column++;
|
|
break;
|
|
|
|
default:
|
|
i = (c>>6)&03;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
i = (c>>3)&07;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
i = (c>>0)&07;
|
|
fmtbuf[column] = i + '0';
|
|
column++;
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
fmtbuf[column] = '\0';
|
|
return fmtbuf;
|
|
}
|
|
|
|
/*
|
|
* Given a string, generate a string from it that shows non-printable
|
|
* characters as the chr parameter passed, except a whitespace character
|
|
* (space, tab, carriage return, new line, vertical tab, or formfeed)
|
|
* which will be replaced by a space, and return a pointer to it.
|
|
*/
|
|
gchar *
|
|
format_text_chr(wmem_allocator_t* allocator, const guchar *string, const size_t len, const guchar chr)
|
|
{
|
|
gchar *fmtbuf = (gchar*)wmem_alloc(allocator, INITIAL_FMTBUF_SIZE);
|
|
int fmtbuf_len = INITIAL_FMTBUF_SIZE;
|
|
int column;
|
|
const guchar *stringend = string + len;
|
|
guchar c;
|
|
|
|
column = 0;
|
|
while (string < stringend)
|
|
{
|
|
/*
|
|
* Is there enough room for this character,
|
|
* and also enough room for a terminating '\0'?
|
|
*/
|
|
if (column+1 >= fmtbuf_len)
|
|
{
|
|
/*
|
|
* Double the buffer's size if it's not big enough.
|
|
* The size of the buffer starts at 128, so doubling its size
|
|
* adds at least another 128 bytes, which is more than enough
|
|
* for one more character plus a terminating '\0'.
|
|
*/
|
|
fmtbuf_len *= 2;
|
|
fmtbuf = (gchar *)wmem_realloc(allocator, fmtbuf, fmtbuf_len);
|
|
}
|
|
c = *string++;
|
|
|
|
if (g_ascii_isprint(c))
|
|
{
|
|
fmtbuf[column] = c;
|
|
column++;
|
|
}
|
|
else if (g_ascii_isspace(c))
|
|
{
|
|
fmtbuf[column] = ' ';
|
|
column++;
|
|
}
|
|
else
|
|
{
|
|
fmtbuf[column] = chr;
|
|
column++;
|
|
}
|
|
}
|
|
fmtbuf[column] = '\0';
|
|
return fmtbuf;
|
|
}
|
|
|
|
static gboolean
|
|
is_byte_sep(guint8 c)
|
|
{
|
|
return (c == '-' || c == ':' || c == '.');
|
|
}
|
|
|
|
/* Turn a string of hex digits with optional separators (defined by
|
|
* is_byte_sep() into a byte array.
|
|
*/
|
|
gboolean
|
|
hex_str_to_bytes(const char *hex_str, GByteArray *bytes, gboolean force_separators)
|
|
{
|
|
guint8 val;
|
|
const gchar *p, *q, *r, *s, *punct;
|
|
char four_digits_first_half[3];
|
|
char four_digits_second_half[3];
|
|
char two_digits[3];
|
|
char one_digit[2];
|
|
|
|
if (! hex_str || ! bytes) {
|
|
return FALSE;
|
|
}
|
|
g_byte_array_set_size(bytes, 0);
|
|
p = hex_str;
|
|
while (*p) {
|
|
q = p+1;
|
|
r = p+2;
|
|
s = p+3;
|
|
|
|
if (*q && *r && *s
|
|
&& g_ascii_isxdigit(*p) && g_ascii_isxdigit(*q) &&
|
|
g_ascii_isxdigit(*r) && g_ascii_isxdigit(*s)) {
|
|
four_digits_first_half[0] = *p;
|
|
four_digits_first_half[1] = *q;
|
|
four_digits_first_half[2] = '\0';
|
|
four_digits_second_half[0] = *r;
|
|
four_digits_second_half[1] = *s;
|
|
four_digits_second_half[2] = '\0';
|
|
|
|
/*
|
|
* Four or more hex digits in a row.
|
|
*/
|
|
val = (guint8) strtoul(four_digits_first_half, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
val = (guint8) strtoul(four_digits_second_half, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
|
|
punct = s + 1;
|
|
if (*punct) {
|
|
/*
|
|
* Make sure the character after
|
|
* the forth hex digit is a byte
|
|
* separator, i.e. that we don't have
|
|
* more than four hex digits, or a
|
|
* bogus character.
|
|
*/
|
|
if (is_byte_sep(*punct)) {
|
|
p = punct + 1;
|
|
continue;
|
|
}
|
|
else if (force_separators) {
|
|
return FALSE;
|
|
}
|
|
}
|
|
p = punct;
|
|
continue;
|
|
}
|
|
else if (*q && g_ascii_isxdigit(*p) && g_ascii_isxdigit(*q)) {
|
|
two_digits[0] = *p;
|
|
two_digits[1] = *q;
|
|
two_digits[2] = '\0';
|
|
|
|
/*
|
|
* Two hex digits in a row.
|
|
*/
|
|
val = (guint8) strtoul(two_digits, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
punct = q + 1;
|
|
if (*punct) {
|
|
/*
|
|
* Make sure the character after
|
|
* the second hex digit is a byte
|
|
* separator, i.e. that we don't have
|
|
* more than two hex digits, or a
|
|
* bogus character.
|
|
*/
|
|
if (is_byte_sep(*punct)) {
|
|
p = punct + 1;
|
|
continue;
|
|
}
|
|
else if (force_separators) {
|
|
return FALSE;
|
|
}
|
|
}
|
|
p = punct;
|
|
continue;
|
|
}
|
|
else if (*q && g_ascii_isxdigit(*p) && is_byte_sep(*q)) {
|
|
one_digit[0] = *p;
|
|
one_digit[1] = '\0';
|
|
|
|
/*
|
|
* Only one hex digit (not at the end of the string)
|
|
*/
|
|
val = (guint8) strtoul(one_digit, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
p = q + 1;
|
|
continue;
|
|
}
|
|
else if (!*q && g_ascii_isxdigit(*p)) {
|
|
one_digit[0] = *p;
|
|
one_digit[1] = '\0';
|
|
|
|
/*
|
|
* Only one hex digit (at the end of the string)
|
|
*/
|
|
val = (guint8) strtoul(one_digit, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
p = q;
|
|
continue;
|
|
}
|
|
else {
|
|
return FALSE;
|
|
}
|
|
}
|
|
return TRUE;
|
|
}
|
|
|
|
static inline gchar
|
|
get_valid_byte_sep(gchar c, const guint encoding)
|
|
{
|
|
gchar retval = -1; /* -1 means failure */
|
|
|
|
switch (c) {
|
|
case ':':
|
|
if (encoding & ENC_SEP_COLON)
|
|
retval = c;
|
|
break;
|
|
case '-':
|
|
if (encoding & ENC_SEP_DASH)
|
|
retval = c;
|
|
break;
|
|
case '.':
|
|
if (encoding & ENC_SEP_DOT)
|
|
retval = c;
|
|
break;
|
|
case ' ':
|
|
if (encoding & ENC_SEP_SPACE)
|
|
retval = c;
|
|
break;
|
|
case '\0':
|
|
/* we were given the end of the string, so it's fine */
|
|
retval = 0;
|
|
break;
|
|
default:
|
|
if (g_ascii_isxdigit(c) && (encoding & ENC_SEP_NONE))
|
|
retval = 0;
|
|
/* anything else means we've got a failure */
|
|
break;
|
|
}
|
|
|
|
return retval;
|
|
}
|
|
|
|
/* Turn a string of hex digits with optional separators (defined by is_byte_sep())
|
|
* into a byte array. Unlike hex_str_to_bytes(), this will read as many hex-char
|
|
* pairs as possible and not error if it hits a non-hex-char; instead it just ends
|
|
* there. (i.e., like strtol()/atoi()/etc.) Unless fail_if_partial is TRUE.
|
|
*
|
|
* The **endptr, if not NULL, is set to the char after the last hex character.
|
|
*/
|
|
gboolean
|
|
hex_str_to_bytes_encoding(const gchar *hex_str, GByteArray *bytes, const gchar **endptr,
|
|
const guint encoding, const gboolean fail_if_partial)
|
|
{
|
|
gint8 c, d;
|
|
guint8 val;
|
|
const gchar *end = hex_str;
|
|
gboolean retval = FALSE;
|
|
gchar sep = -1;
|
|
|
|
/* a map from ASCII hex chars to their value */
|
|
static const gint8 str_to_nibble[256] = {
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1,
|
|
-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
|
|
-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
};
|
|
|
|
/* we must see two hex chars at the beginning, or fail */
|
|
if (bytes && *end && g_ascii_isxdigit(*end) && g_ascii_isxdigit(*(end+1))) {
|
|
retval = TRUE;
|
|
|
|
/* set the separator character we'll allow; if this returns a -1, it means something's
|
|
* invalid after the hex, but we'll let the while-loop grab the first hex-pair anyway
|
|
*/
|
|
sep = get_valid_byte_sep(*(end+2), encoding);
|
|
|
|
while (*end) {
|
|
c = str_to_nibble[(guchar)*end];
|
|
if (c < 0) {
|
|
if (fail_if_partial) retval = FALSE;
|
|
break;
|
|
}
|
|
++end;
|
|
|
|
d = str_to_nibble[(guchar)*end];
|
|
if (d < 0) {
|
|
if (fail_if_partial) retval = FALSE;
|
|
break;
|
|
}
|
|
val = ((guint8)c * 16) + d;
|
|
g_byte_array_append(bytes, &val, 1);
|
|
++end;
|
|
|
|
/* check for separator and peek at next char to make sure we should keep going */
|
|
if (sep > 0 && *end == sep && str_to_nibble[(guchar)*(end+1)] > -1) {
|
|
/* yes, it's the right sep and followed by more hex, so skip the sep */
|
|
++end;
|
|
} else if (sep != 0 && *end) {
|
|
/* we either need a separator, but we don't see one; or the get_valid_byte_sep()
|
|
earlier didn't find a valid one to begin with */
|
|
if (fail_if_partial) retval = FALSE;
|
|
break;
|
|
}
|
|
/* otherwise, either no separator allowed, or *end is null, or *end is an invalid
|
|
* sep, or *end is a valid sep but after it is not a hex char - in all those
|
|
* cases, just loop back up and let it fail later naturally.
|
|
*/
|
|
}
|
|
}
|
|
|
|
if (!retval) {
|
|
if (bytes) g_byte_array_set_size(bytes, 0);
|
|
end = hex_str;
|
|
}
|
|
|
|
if (endptr) *endptr = end;
|
|
|
|
return retval;
|
|
}
|
|
|
|
/*
|
|
* Turn an RFC 3986 percent-encoded string into a byte array.
|
|
* XXX - We don't check for reserved characters.
|
|
*/
|
|
#define HEX_DIGIT_BUF_LEN 3
|
|
gboolean
|
|
uri_str_to_bytes(const char *uri_str, GByteArray *bytes)
|
|
{
|
|
guint8 val;
|
|
const gchar *p;
|
|
gchar hex_digit[HEX_DIGIT_BUF_LEN];
|
|
|
|
g_byte_array_set_size(bytes, 0);
|
|
if (! uri_str) {
|
|
return FALSE;
|
|
}
|
|
|
|
p = uri_str;
|
|
|
|
while (*p) {
|
|
if (!g_ascii_isprint(*p))
|
|
return FALSE;
|
|
if (*p == '%') {
|
|
p++;
|
|
if (*p == '\0') return FALSE;
|
|
hex_digit[0] = *p;
|
|
p++;
|
|
if (*p == '\0') return FALSE;
|
|
hex_digit[1] = *p;
|
|
hex_digit[2] = '\0';
|
|
if (! g_ascii_isxdigit(hex_digit[0]) || ! g_ascii_isxdigit(hex_digit[1]))
|
|
return FALSE;
|
|
val = (guint8) strtoul(hex_digit, NULL, 16);
|
|
g_byte_array_append(bytes, &val, 1);
|
|
} else {
|
|
g_byte_array_append(bytes, (const guint8 *) p, 1);
|
|
}
|
|
p++;
|
|
|
|
}
|
|
return TRUE;
|
|
}
|
|
|
|
/*
|
|
* Given a GByteArray, generate a string from it that shows non-printable
|
|
* characters as percent-style escapes, and return a pointer to it.
|
|
*/
|
|
gchar *
|
|
format_uri(wmem_allocator_t* allocator, const GByteArray *bytes, const gchar *reserved_chars)
|
|
{
|
|
gchar *fmtbuf = (gchar*)wmem_alloc(allocator, INITIAL_FMTBUF_SIZE);
|
|
guint fmtbuf_len = INITIAL_FMTBUF_SIZE;
|
|
static const guchar *reserved_def = ":/?#[]@!$&'()*+,;= ";
|
|
const guchar *reserved = reserved_def;
|
|
guint8 c;
|
|
guint column, i;
|
|
gboolean is_reserved = FALSE;
|
|
|
|
if (! bytes)
|
|
return "";
|
|
|
|
if (reserved_chars)
|
|
reserved = reserved_chars;
|
|
|
|
for (column = 0; column < bytes->len; column++) {
|
|
/*
|
|
* Is there enough room for this character, if it expands to
|
|
* a percent plus 2 hex digits (which is the most it can
|
|
* expand to), and also enough room for a terminating '\0'?
|
|
*/
|
|
if (column+2+1 >= fmtbuf_len) {
|
|
/*
|
|
* Double the buffer's size if it's not big enough.
|
|
* The size of the buffer starts at 128, so doubling its size
|
|
* adds at least another 128 bytes, which is more than enough
|
|
* for one more character plus a terminating '\0'.
|
|
*/
|
|
fmtbuf_len *= 2;
|
|
fmtbuf = (gchar *)wmem_realloc(allocator, fmtbuf, fmtbuf_len);
|
|
}
|
|
c = bytes->data[column];
|
|
|
|
if (!g_ascii_isprint(c) || c == '%') {
|
|
is_reserved = TRUE;
|
|
}
|
|
|
|
for (i = 0; reserved[i]; i++) {
|
|
if (c == reserved[i])
|
|
is_reserved = TRUE;
|
|
}
|
|
|
|
if (!is_reserved) {
|
|
fmtbuf[column] = c;
|
|
} else {
|
|
fmtbuf[column] = '%';
|
|
column++;
|
|
fmtbuf[column] = hex[c >> 4];
|
|
column++;
|
|
fmtbuf[column] = hex[c & 0xF];
|
|
}
|
|
}
|
|
fmtbuf[column] = '\0';
|
|
return fmtbuf;
|
|
}
|
|
|
|
/**
|
|
* Create a copy of a GByteArray
|
|
*
|
|
* @param ba The byte array to be copied.
|
|
* @return If ba exists, a freshly allocated copy. NULL otherwise.
|
|
*
|
|
*/
|
|
GByteArray *
|
|
byte_array_dup(const GByteArray *ba)
|
|
{
|
|
GByteArray *new_ba;
|
|
|
|
if (!ba)
|
|
return NULL;
|
|
|
|
new_ba = g_byte_array_new();
|
|
g_byte_array_append(new_ba, ba->data, ba->len);
|
|
return new_ba;
|
|
}
|
|
|
|
#define SUBID_BUF_LEN 5
|
|
gboolean
|
|
oid_str_to_bytes(const char *oid_str, GByteArray *bytes)
|
|
{
|
|
return rel_oid_str_to_bytes(oid_str, bytes, TRUE);
|
|
}
|
|
gboolean
|
|
rel_oid_str_to_bytes(const char *oid_str, GByteArray *bytes, gboolean is_absolute)
|
|
{
|
|
guint32 subid0, subid, sicnt, i;
|
|
const char *p, *dot;
|
|
guint8 buf[SUBID_BUF_LEN];
|
|
|
|
g_byte_array_set_size(bytes, 0);
|
|
|
|
/* check syntax */
|
|
p = oid_str;
|
|
dot = NULL;
|
|
while (*p) {
|
|
if (!g_ascii_isdigit(*p) && (*p != '.')) return FALSE;
|
|
if (*p == '.') {
|
|
if (p == oid_str && is_absolute) return FALSE;
|
|
if (!*(p+1)) return FALSE;
|
|
if ((p-1) == dot) return FALSE;
|
|
dot = p;
|
|
}
|
|
p++;
|
|
}
|
|
if (!dot) return FALSE;
|
|
|
|
p = oid_str;
|
|
sicnt = is_absolute ? 0 : 2;
|
|
if (!is_absolute) p++;
|
|
subid0 = 0; /* squelch GCC complaints */
|
|
while (*p) {
|
|
subid = 0;
|
|
while (g_ascii_isdigit(*p)) {
|
|
subid *= 10;
|
|
subid += *p - '0';
|
|
p++;
|
|
}
|
|
if (sicnt == 0) {
|
|
subid0 = subid;
|
|
if (subid0 > 2) return FALSE;
|
|
} else if (sicnt == 1) {
|
|
if ((subid0 < 2) && (subid > 39)) return FALSE;
|
|
subid += 40 * subid0;
|
|
}
|
|
if (sicnt) {
|
|
i = SUBID_BUF_LEN;
|
|
do {
|
|
i--;
|
|
buf[i] = 0x80 | (subid % 0x80);
|
|
subid >>= 7;
|
|
} while (subid && i);
|
|
buf[SUBID_BUF_LEN-1] &= 0x7F;
|
|
g_byte_array_append(bytes, buf + i, SUBID_BUF_LEN - i);
|
|
}
|
|
sicnt++;
|
|
if (*p) p++;
|
|
}
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
/**
|
|
* Compare the contents of two GByteArrays
|
|
*
|
|
* @param ba1 A byte array
|
|
* @param ba2 A byte array
|
|
* @return If both arrays are non-NULL and their lengths are equal and
|
|
* their contents are equal, returns TRUE. Otherwise, returns
|
|
* FALSE.
|
|
*
|
|
* XXX - Should this be in strutil.c?
|
|
*/
|
|
gboolean
|
|
byte_array_equal(GByteArray *ba1, GByteArray *ba2)
|
|
{
|
|
if (!ba1 || !ba2)
|
|
return FALSE;
|
|
|
|
if (ba1->len != ba2->len)
|
|
return FALSE;
|
|
|
|
if (memcmp(ba1->data, ba2->data, ba1->len) != 0)
|
|
return FALSE;
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/* Return a XML escaped representation of the unescaped string.
|
|
* The returned string must be freed when no longer in use. */
|
|
gchar *
|
|
xml_escape(const gchar *unescaped)
|
|
{
|
|
GString *buffer = g_string_sized_new(128);
|
|
const gchar *p;
|
|
gchar c;
|
|
|
|
p = unescaped;
|
|
while ( (c = *p++) ) {
|
|
switch (c) {
|
|
case '<':
|
|
g_string_append(buffer, "<");
|
|
break;
|
|
case '>':
|
|
g_string_append(buffer, ">");
|
|
break;
|
|
case '&':
|
|
g_string_append(buffer, "&");
|
|
break;
|
|
case '\'':
|
|
g_string_append(buffer, "'");
|
|
break;
|
|
case '"':
|
|
g_string_append(buffer, """);
|
|
break;
|
|
default:
|
|
g_string_append_c(buffer, c);
|
|
break;
|
|
}
|
|
}
|
|
/* Return the string value contained within the GString
|
|
* after getting rid of the GString structure.
|
|
* This is the way to do this, see the GLib reference. */
|
|
return g_string_free(buffer, FALSE);
|
|
}
|
|
|
|
|
|
/* Return the first occurrence of needle in haystack.
|
|
* If not found, return NULL.
|
|
* If either haystack or needle has 0 length, return NULL.
|
|
* Algorithm copied from GNU's glibc 2.3.2 memmem() under LGPL 2.1+ */
|
|
const guint8 *
|
|
epan_memmem(const guint8 *haystack, guint haystack_len,
|
|
const guint8 *needle, guint needle_len)
|
|
{
|
|
const guint8 *begin;
|
|
const guint8 *const last_possible = haystack + haystack_len - needle_len;
|
|
|
|
if (needle_len == 0) {
|
|
return NULL;
|
|
}
|
|
|
|
if (needle_len > haystack_len) {
|
|
return NULL;
|
|
}
|
|
|
|
for (begin = haystack ; begin <= last_possible; ++begin) {
|
|
if (begin[0] == needle[0] &&
|
|
!memcmp(&begin[1], needle + 1,
|
|
needle_len - 1)) {
|
|
return begin;
|
|
}
|
|
}
|
|
|
|
return NULL;
|
|
}
|
|
|
|
/*
|
|
* Scan the search string to make sure it's valid hex. Return the
|
|
* number of bytes in nbytes.
|
|
*/
|
|
guint8 *
|
|
convert_string_to_hex(const char *string, size_t *nbytes)
|
|
{
|
|
size_t n_bytes;
|
|
const char *p;
|
|
gchar c;
|
|
guint8 *bytes, *q, byte_val;
|
|
|
|
n_bytes = 0;
|
|
p = &string[0];
|
|
for (;;) {
|
|
c = *p++;
|
|
if (c == '\0')
|
|
break;
|
|
if (g_ascii_isspace(c))
|
|
continue; /* allow white space */
|
|
if (c==':' || c=='.' || c=='-')
|
|
continue; /* skip any ':', '.', or '-' between bytes */
|
|
if (!g_ascii_isxdigit(c)) {
|
|
/* Not a valid hex digit - fail */
|
|
return NULL;
|
|
}
|
|
|
|
/*
|
|
* We can only match bytes, not nibbles; we must have a valid
|
|
* hex digit immediately after that hex digit.
|
|
*/
|
|
c = *p++;
|
|
if (!g_ascii_isxdigit(c))
|
|
return NULL;
|
|
|
|
/* 2 hex digits = 1 byte */
|
|
n_bytes++;
|
|
}
|
|
|
|
/*
|
|
* Were we given any hex digits?
|
|
*/
|
|
if (n_bytes == 0) {
|
|
/* No. */
|
|
return NULL;
|
|
}
|
|
|
|
/*
|
|
* OK, it's valid, and it generates "n_bytes" bytes; generate the
|
|
* raw byte array.
|
|
*/
|
|
bytes = (guint8 *)g_malloc(n_bytes);
|
|
p = &string[0];
|
|
q = &bytes[0];
|
|
for (;;) {
|
|
c = *p++;
|
|
if (c == '\0')
|
|
break;
|
|
if (g_ascii_isspace(c))
|
|
continue; /* allow white space */
|
|
if (c==':' || c=='.' || c=='-')
|
|
continue; /* skip any ':', '.', or '-' between bytes */
|
|
/* From the loop above, we know this is a hex digit */
|
|
byte_val = ws_xton(c);
|
|
byte_val <<= 4;
|
|
|
|
/* We also know this is a hex digit */
|
|
c = *p++;
|
|
byte_val |= ws_xton(c);
|
|
|
|
*q++ = byte_val;
|
|
}
|
|
*nbytes = n_bytes;
|
|
return bytes;
|
|
}
|
|
|
|
/*
|
|
* Copy if if it's a case-sensitive search; uppercase it if it's
|
|
* a case-insensitive search.
|
|
*/
|
|
char *
|
|
convert_string_case(const char *string, gboolean case_insensitive)
|
|
{
|
|
|
|
if (case_insensitive) {
|
|
return g_utf8_strup(string, -1);
|
|
} else {
|
|
return g_strdup(string);
|
|
}
|
|
}
|
|
|
|
const char *
|
|
epan_strcasestr(const char *haystack, const char *needle)
|
|
{
|
|
gsize hlen = strlen(haystack);
|
|
gsize nlen = strlen(needle);
|
|
|
|
while (hlen-- >= nlen) {
|
|
if (!g_ascii_strncasecmp(haystack, needle, nlen))
|
|
return haystack;
|
|
haystack++;
|
|
}
|
|
return NULL;
|
|
}
|
|
|
|
const char *
|
|
string_or_null(const char *string)
|
|
{
|
|
if (string)
|
|
return string;
|
|
return "[NULL]";
|
|
}
|
|
|
|
int
|
|
escape_string_len(const char *string)
|
|
{
|
|
const char *p;
|
|
gchar c;
|
|
int repr_len;
|
|
|
|
repr_len = 0;
|
|
for (p = string; (c = *p) != '\0'; p++) {
|
|
/* Backslashes and double-quotes must
|
|
* be escaped */
|
|
if (c == '\\' || c == '"') {
|
|
repr_len += 2;
|
|
}
|
|
/* Values that can't nicely be represented
|
|
* in ASCII need to be escaped. */
|
|
else if (!g_ascii_isprint(c)) {
|
|
/* c --> \xNN */
|
|
repr_len += 4;
|
|
}
|
|
/* Other characters are just passed through. */
|
|
else {
|
|
repr_len++;
|
|
}
|
|
}
|
|
return repr_len + 2; /* string plus leading and trailing quotes */
|
|
}
|
|
|
|
char *
|
|
escape_string(char *buf, const char *string)
|
|
{
|
|
const gchar *p;
|
|
gchar c;
|
|
char *bufp;
|
|
char hexbuf[3];
|
|
|
|
bufp = buf;
|
|
*bufp++ = '"';
|
|
for (p = string; (c = *p) != '\0'; p++) {
|
|
/* Backslashes and double-quotes must
|
|
* be escaped. */
|
|
if (c == '\\' || c == '"') {
|
|
*bufp++ = '\\';
|
|
*bufp++ = c;
|
|
}
|
|
/* Values that can't nicely be represented
|
|
* in ASCII need to be escaped. */
|
|
else if (!g_ascii_isprint(c)) {
|
|
/* c --> \xNN */
|
|
g_snprintf(hexbuf,sizeof(hexbuf), "%02x", (unsigned char) c);
|
|
*bufp++ = '\\';
|
|
*bufp++ = 'x';
|
|
*bufp++ = hexbuf[0];
|
|
*bufp++ = hexbuf[1];
|
|
}
|
|
/* Other characters are just passed through. */
|
|
else {
|
|
*bufp++ = c;
|
|
}
|
|
}
|
|
*bufp++ = '"';
|
|
*bufp = '\0';
|
|
return buf;
|
|
}
|
|
|
|
#define GN_CHAR_ALPHABET_SIZE 128
|
|
|
|
static gunichar IA5_default_alphabet[GN_CHAR_ALPHABET_SIZE] = {
|
|
|
|
/*ITU-T recommendation T.50 specifies International Reference Alphabet 5 (IA5) */
|
|
|
|
'?', '?', '?', '?', '?', '?', '?', '?',
|
|
'?', '?', '?', '?', '?', '?', '?', '?',
|
|
'?', '?', '?', '?', '?', '?', '?', '?',
|
|
'?', '?', '?', '?', '?', '?', '?', '?',
|
|
' ', '!', '\"','#', '$', '%', '&', '\'',
|
|
'(', ')', '*', '+', ',', '-', '.', '/',
|
|
'0', '1', '2', '3', '4', '5', '6', '7',
|
|
'8', '9', ':', ';', '<', '=', '>', '?',
|
|
'@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
|
|
'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
|
|
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
|
|
'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
|
|
'`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
|
|
'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
|
|
'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
|
|
'x', 'y', 'z', '{', '|', '}', '~', '?'
|
|
};
|
|
|
|
static gunichar
|
|
char_def_ia5_alphabet_decode(unsigned char value)
|
|
{
|
|
if (value < GN_CHAR_ALPHABET_SIZE) {
|
|
return IA5_default_alphabet[value];
|
|
}
|
|
else {
|
|
return '?';
|
|
}
|
|
}
|
|
|
|
void
|
|
IA5_7BIT_decode(unsigned char * dest, const unsigned char* src, int len)
|
|
{
|
|
int i, j;
|
|
gunichar buf;
|
|
|
|
for (i = 0, j = 0; j < len; j++) {
|
|
buf = char_def_ia5_alphabet_decode(src[j]);
|
|
i += g_unichar_to_utf8(buf,&(dest[i]));
|
|
}
|
|
dest[i]=0;
|
|
return;
|
|
}
|
|
|
|
/*
|
|
* This function takes a string and copies it, inserting a 'chr' before
|
|
* every 'chr' in it.
|
|
*/
|
|
gchar*
|
|
ws_strdup_escape_char (const gchar *str, const gchar chr)
|
|
{
|
|
const gchar *p;
|
|
gchar *q, *new_str;
|
|
|
|
if(!str)
|
|
return NULL;
|
|
|
|
p = str;
|
|
/* Worst case: A string that is full of 'chr' */
|
|
q = new_str = (gchar *)g_malloc (strlen(str) * 2 + 1);
|
|
|
|
while(*p != 0) {
|
|
if(*p == chr)
|
|
*q++ = chr;
|
|
|
|
*q++ = *p++;
|
|
}
|
|
*q = '\0';
|
|
|
|
return new_str;
|
|
}
|
|
|
|
/*
|
|
* This function takes a string and copies it, removing any occurences of double
|
|
* 'chr' with a single 'chr'.
|
|
*/
|
|
gchar*
|
|
ws_strdup_unescape_char (const gchar *str, const char chr)
|
|
{
|
|
const gchar *p;
|
|
gchar *q, *new_str;
|
|
|
|
if(!str)
|
|
return NULL;
|
|
|
|
p = str;
|
|
/* Worst case: A string that contains no 'chr' */
|
|
q = new_str = (gchar *)g_malloc (strlen(str) + 1);
|
|
|
|
while(*p != 0) {
|
|
*q++ = *p;
|
|
if ((*p == chr) && (*(p+1) == chr))
|
|
p += 2;
|
|
else
|
|
p++;
|
|
}
|
|
*q = '\0';
|
|
|
|
return new_str;
|
|
}
|
|
|
|
/* Create a newly-allocated string with replacement values. */
|
|
gchar *
|
|
string_replace(const gchar* str, const gchar *old_val, const gchar *new_val)
|
|
{
|
|
gchar **str_parts;
|
|
gchar *new_str;
|
|
|
|
if (!str || !old_val) {
|
|
return NULL;
|
|
}
|
|
|
|
str_parts = g_strsplit(str, old_val, 0);
|
|
new_str = g_strjoinv(new_val, str_parts);
|
|
g_strfreev(str_parts);
|
|
|
|
return new_str;
|
|
}
|
|
|
|
/*
|
|
* Editor modelines - http://www.wireshark.org/tools/modelines.html
|
|
*
|
|
* Local variables:
|
|
* c-basic-offset: 4
|
|
* tab-width: 8
|
|
* indent-tabs-mode: nil
|
|
* End:
|
|
*
|
|
* vi: set shiftwidth=4 tabstop=8 expandtab:
|
|
* :indentSize=4:tabSize=8:noTabs=true:
|
|
*/
|