diff options
Diffstat (limited to 'protocols/Sametime/src/glib/gbase64.c')
-rw-r--r-- | protocols/Sametime/src/glib/gbase64.c | 440 |
1 files changed, 440 insertions, 0 deletions
diff --git a/protocols/Sametime/src/glib/gbase64.c b/protocols/Sametime/src/glib/gbase64.c new file mode 100644 index 0000000000..8a492de363 --- /dev/null +++ b/protocols/Sametime/src/glib/gbase64.c @@ -0,0 +1,440 @@ +/* gbase64.c - Base64 encoding/decoding + * + * Copyright (C) 2006 Alexander Larsson <alexl@redhat.com> + * Copyright (C) 2000-2003 Ximian Inc. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Library General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Library General Public License for more details. + * + * You should have received a copy of the GNU Library General Public + * License along with this library; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 02111-1307, USA. + * + * This is based on code in camel, written by: + * Michael Zucchi <notzed@ximian.com> + * Jeffrey Stedfast <fejj@ximian.com> + */ + +#include "config.h" + +#include <string.h> + +#include "gbase64.h" +#include "gtestutils.h" +#include "glibintl.h" + + +/** + * SECTION:base64 + * @title: Base64 Encoding + * @short_description: encodes and decodes data in Base64 format + * + * Base64 is an encoding that allows a sequence of arbitrary bytes to be + * encoded as a sequence of printable ASCII characters. For the definition + * of Base64, see <ulink url="http://www.ietf.org/rfc/rfc1421.txt">RFC + * 1421</ulink> or <ulink url="http://www.ietf.org/rfc/rfc2045.txt">RFC + * 2045</ulink>. Base64 is most commonly used as a MIME transfer encoding + * for email. + * + * GLib supports incremental encoding using g_base64_encode_step() and + * g_base64_encode_close(). Incremental decoding can be done with + * g_base64_decode_step(). To encode or decode data in one go, use + * g_base64_encode() or g_base64_decode(). To avoid memory allocation when + * decoding, you can use g_base64_decode_inplace(). + * + * Support for Base64 encoding has been added in GLib 2.12. + */ + +static const char base64_alphabet[] = + "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; + +/** + * g_base64_encode_step: + * @in: the binary data to encode + * @len: the length of @in + * @break_lines: whether to break long lines + * @out: pointer to destination buffer + * @state: Saved state between steps, initialize to 0 + * @save: Saved state between steps, initialize to 0 + * + * Incrementally encode a sequence of binary data into its Base-64 stringified + * representation. By calling this function multiple times you can convert + * data in chunks to avoid having to have the full encoded data in memory. + * + * When all of the data has been converted you must call + * g_base64_encode_close() to flush the saved state. + * + * The output buffer must be large enough to fit all the data that will + * be written to it. Due to the way base64 encodes you will need + * at least: (@len / 3 + 1) * 4 + 4 bytes (+ 4 may be needed in case of + * non-zero state). If you enable line-breaking you will need at least: + * ((@len / 3 + 1) * 4 + 4) / 72 + 1 bytes of extra space. + * + * @break_lines is typically used when putting base64-encoded data in emails. + * It breaks the lines at 72 columns instead of putting all of the text on + * the same line. This avoids problems with long lines in the email system. + * + * Return value: The number of bytes of output that was written + * + * Since: 2.12 + */ +gsize +g_base64_encode_step (const guchar *in, + gsize len, + gboolean break_lines, + gchar *out, + gint *state, + gint *save) +{ + char *outptr; + const guchar *inptr; + + g_return_val_if_fail (in != NULL, 0); + g_return_val_if_fail (out != NULL, 0); + g_return_val_if_fail (state != NULL, 0); + g_return_val_if_fail (save != NULL, 0); + + if (len <= 0) + return 0; + + inptr = in; + outptr = out; + + if (len + ((char *) save) [0] > 2) + { + const guchar *inend = in+len-2; + int c1, c2, c3; + int already; + + already = *state; + + switch (((char *) save) [0]) + { + case 1: + c1 = ((unsigned char *) save) [1]; + goto skip1; + case 2: + c1 = ((unsigned char *) save) [1]; + c2 = ((unsigned char *) save) [2]; + goto skip2; + } + + /* + * yes, we jump into the loop, no i'm not going to change it, + * it's beautiful! + */ + while (inptr < inend) + { + c1 = *inptr++; + skip1: + c2 = *inptr++; + skip2: + c3 = *inptr++; + *outptr++ = base64_alphabet [ c1 >> 2 ]; + *outptr++ = base64_alphabet [ c2 >> 4 | + ((c1&0x3) << 4) ]; + *outptr++ = base64_alphabet [ ((c2 &0x0f) << 2) | + (c3 >> 6) ]; + *outptr++ = base64_alphabet [ c3 & 0x3f ]; + /* this is a bit ugly ... */ + if (break_lines && (++already) >= 19) + { + *outptr++ = '\n'; + already = 0; + } + } + + ((char *)save)[0] = 0; + len = 2 - (inptr - inend); + *state = already; + } + + if (len>0) + { + char *saveout; + + /* points to the slot for the next char to save */ + saveout = & (((char *)save)[1]) + ((char *)save)[0]; + + /* len can only be 0 1 or 2 */ + switch(len) + { + case 2: *saveout++ = *inptr++; + case 1: *saveout++ = *inptr++; + } + ((char *)save)[0] += len; + } + + return outptr - out; +} + +/** + * g_base64_encode_close: + * @break_lines: whether to break long lines + * @out: pointer to destination buffer + * @state: Saved state from g_base64_encode_step() + * @save: Saved state from g_base64_encode_step() + * + * Flush the status from a sequence of calls to g_base64_encode_step(). + * + * The output buffer must be large enough to fit all the data that will + * be written to it. It will need up to 4 bytes, or up to 5 bytes if + * line-breaking is enabled. + * + * Return value: The number of bytes of output that was written + * + * Since: 2.12 + */ +gsize +g_base64_encode_close (gboolean break_lines, + gchar *out, + gint *state, + gint *save) +{ + int c1, c2; + char *outptr = out; + + g_return_val_if_fail (out != NULL, 0); + g_return_val_if_fail (state != NULL, 0); + g_return_val_if_fail (save != NULL, 0); + + c1 = ((unsigned char *) save) [1]; + c2 = ((unsigned char *) save) [2]; + + switch (((char *) save) [0]) + { + case 2: + outptr [2] = base64_alphabet[ ( (c2 &0x0f) << 2 ) ]; + g_assert (outptr [2] != 0); + goto skip; + case 1: + outptr[2] = '='; + skip: + outptr [0] = base64_alphabet [ c1 >> 2 ]; + outptr [1] = base64_alphabet [ c2 >> 4 | ( (c1&0x3) << 4 )]; + outptr [3] = '='; + outptr += 4; + break; + } + if (break_lines) + *outptr++ = '\n'; + + *save = 0; + *state = 0; + + return outptr - out; +} + +/** + * g_base64_encode: + * @data: the binary data to encode + * @len: the length of @data + * + * Encode a sequence of binary data into its Base-64 stringified + * representation. + * + * Return value: a newly allocated, zero-terminated Base-64 encoded + * string representing @data. The returned string must + * be freed with g_free(). + * + * Since: 2.12 + */ +gchar * +g_base64_encode (const guchar *data, + gsize len) +{ + gchar *out; + gint state = 0, outlen; + gint save = 0; + + g_return_val_if_fail (data != NULL || len == 0, NULL); + + /* We can use a smaller limit here, since we know the saved state is 0, + +1 is needed for trailing \0, also check for unlikely integer overflow */ + if (len >= ((G_MAXSIZE - 1) / 4 - 1) * 3) + g_error("%s: input too large for Base64 encoding (%"G_GSIZE_FORMAT" chars)", + G_STRLOC, len); + + out = g_malloc ((len / 3 + 1) * 4 + 1); + + outlen = g_base64_encode_step (data, len, FALSE, out, &state, &save); + outlen += g_base64_encode_close (FALSE, out + outlen, &state, &save); + out[outlen] = '\0'; + + return (gchar *) out; +} + +static const unsigned char mime_base64_rank[256] = { + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255, 62,255,255,255, 63, + 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,255,255,255, 0,255,255, + 255, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, + 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,255,255,255,255,255, + 255, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, + 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, + 255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255, +}; + +/** + * g_base64_decode_step: + * @in: binary input data + * @len: max length of @in data to decode + * @out: output buffer + * @state: Saved state between steps, initialize to 0 + * @save: Saved state between steps, initialize to 0 + * + * Incrementally decode a sequence of binary data from its Base-64 stringified + * representation. By calling this function multiple times you can convert + * data in chunks to avoid having to have the full encoded data in memory. + * + * The output buffer must be large enough to fit all the data that will + * be written to it. Since base64 encodes 3 bytes in 4 chars you need + * at least: (@len / 4) * 3 + 3 bytes (+ 3 may be needed in case of non-zero + * state). + * + * Return value: The number of bytes of output that was written + * + * Since: 2.12 + **/ +gsize +g_base64_decode_step (const gchar *in, + gsize len, + guchar *out, + gint *state, + guint *save) +{ + const guchar *inptr; + guchar *outptr; + const guchar *inend; + guchar c, rank; + guchar last[2]; + unsigned int v; + int i; + + g_return_val_if_fail (in != NULL, 0); + g_return_val_if_fail (out != NULL, 0); + g_return_val_if_fail (state != NULL, 0); + g_return_val_if_fail (save != NULL, 0); + + if (len <= 0) + return 0; + + inend = (const guchar *)in+len; + outptr = out; + + /* convert 4 base64 bytes to 3 normal bytes */ + v=*save; + i=*state; + inptr = (const guchar *)in; + last[0] = last[1] = 0; + while (inptr < inend) + { + c = *inptr++; + rank = mime_base64_rank [c]; + if (rank != 0xff) + { + last[1] = last[0]; + last[0] = c; + v = (v<<6) | rank; + i++; + if (i==4) + { + *outptr++ = v>>16; + if (last[1] != '=') + *outptr++ = v>>8; + if (last[0] != '=') + *outptr++ = v; + i=0; + } + } + } + + *save = v; + *state = i; + + return outptr - out; +} + +/** + * g_base64_decode: + * @text: zero-terminated string with base64 text to decode + * @out_len: The length of the decoded data is written here + * + * Decode a sequence of Base-64 encoded text into binary data + * + * Return value: a newly allocated buffer containing the binary data + * that @text represents. The returned buffer must + * be freed with g_free(). + * + * Since: 2.12 + */ +guchar * +g_base64_decode (const gchar *text, + gsize *out_len) +{ + guchar *ret; + gsize input_length; + gint state = 0; + guint save = 0; + + g_return_val_if_fail (text != NULL, NULL); + g_return_val_if_fail (out_len != NULL, NULL); + + input_length = strlen (text); + + /* We can use a smaller limit here, since we know the saved state is 0, + +1 used to avoid calling g_malloc0(0), and hence retruning NULL */ + ret = g_malloc0 ((input_length / 4) * 3 + 1); + + *out_len = g_base64_decode_step (text, input_length, ret, &state, &save); + + return ret; +} + +/** + * g_base64_decode_inplace: + * @text: zero-terminated string with base64 text to decode + * @out_len: The length of the decoded data is written here + * + * Decode a sequence of Base-64 encoded text into binary data + * by overwriting the input data. + * + * Return value: The binary data that @text responds. This pointer + * is the same as the input @text. + * + * Since: 2.20 + */ +guchar * +g_base64_decode_inplace (gchar *text, + gsize *out_len) +{ + gint input_length, state = 0; + guint save = 0; + + g_return_val_if_fail (text != NULL, NULL); + g_return_val_if_fail (out_len != NULL, NULL); + + input_length = strlen (text); + + g_return_val_if_fail (input_length > 1, NULL); + + *out_len = g_base64_decode_step (text, input_length, (guchar *) text, &state, &save); + + return (guchar *) text; +} |