gbase64: Fix base-64 stepped encoding with small chunks

If encoding with small chunks such that the call to
g_base64_encode_close() has to deal with 0 < x < 24 bits of remaining
state, the encoding code would not correctly use zeroes to pad out the
state — it would use left-over state from an earlier iteration in the
encoding process.

This resulted in invalid base-64 encodings.

Add a unit test for incremental encoding using different sized chunks
too.

Thanks to Rainier Perske for reporting and analysing the bug.

https://bugzilla.gnome.org/show_bug.cgi?id=780066

Signed-off-by: Philip Withnall <withnall@endlessm.com>
This commit is contained in:
Philip Withnall 2017-03-17 11:30:47 +00:00
parent 32aae79db5
commit 35c0dd2755
2 changed files with 59 additions and 0 deletions

View File

@ -222,6 +222,7 @@ g_base64_encode_close (gboolean break_lines,
goto skip;
case 1:
outptr[2] = '=';
c2 = 0; /* saved state here is not relevant */
skip:
outptr [0] = base64_alphabet [ c1 >> 2 ];
outptr [1] = base64_alphabet [ c2 >> 4 | ( (c1&0x3) << 4 )];

View File

@ -236,6 +236,58 @@ test_base64_encode (void)
}
}
/* Test that incremental and all-in-one encoding of strings of a length which
* is not a multiple of 3 bytes behave the same, as the state carried over
* between g_base64_encode_step() calls varies depending on how the input is
* split up. This is like the test_base64_decode_smallblock() test, but for
* encoding. */
static void
test_base64_encode_incremental_small_block (gconstpointer block_size_p)
{
gsize i;
struct MyRawData myraw;
g_test_bug ("780066");
generate_databuffer_for_base64 (&myraw);
for (i = 0; ok_100_encode_strs[i] != NULL; i++)
{
const guint block_size = GPOINTER_TO_UINT (block_size_p);
gchar *encoded_complete = NULL;
gchar encoded_stepped[1024];
gint state = 0, save = 0;
gsize len_written, len_read, len_to_read, input_length;
input_length = i + 1;
/* Do it all at once. */
encoded_complete = g_base64_encode (myraw.data, input_length);
/* Split the data up so some number of bits remain after each step. */
for (len_written = 0, len_read = 0; len_read < input_length; len_read += len_to_read)
{
len_to_read = MIN (block_size, input_length - len_read);
len_written += g_base64_encode_step (myraw.data + len_read, len_to_read,
FALSE,
encoded_stepped + len_written,
&state, &save);
}
len_written += g_base64_encode_close (FALSE, encoded_stepped + len_written,
&state, &save);
g_assert_cmpuint (len_written, <, G_N_ELEMENTS (encoded_stepped));
/* Nul-terminate to make string comparison easier. */
encoded_stepped[len_written] = '\0';
/* Compare results. They should be the same. */
g_assert_cmpstr (encoded_complete, ==, ok_100_encode_strs[i]);
g_assert_cmpstr (encoded_stepped, ==, encoded_complete);
g_free (encoded_complete);
}
}
static void
decode_and_compare (const gchar *datap,
@ -361,6 +413,7 @@ main (int argc, char *argv[])
gint i;
g_test_init (&argc, &argv, NULL);
g_test_bug_base ("https://bugzilla.gnome.org/browse.cgi?product=");
for (i = 0; i < DATA_SIZE; i++)
data[i] = (guchar)i;
@ -370,6 +423,11 @@ main (int argc, char *argv[])
g_test_add_data_func ("/base64/full/3", GINT_TO_POINTER (2), test_full);
g_test_add_data_func ("/base64/full/4", GINT_TO_POINTER (3), test_full);
g_test_add_data_func ("/base64/encode/incremental/small-block/1", GINT_TO_POINTER (1), test_base64_encode_incremental_small_block);
g_test_add_data_func ("/base64/encode/incremental/small-block/2", GINT_TO_POINTER (2), test_base64_encode_incremental_small_block);
g_test_add_data_func ("/base64/encode/incremental/small-block/3", GINT_TO_POINTER (3), test_base64_encode_incremental_small_block);
g_test_add_data_func ("/base64/encode/incremental/small-block/4", GINT_TO_POINTER (4), test_base64_encode_incremental_small_block);
g_test_add_data_func ("/base64/incremental/nobreak/1", GINT_TO_POINTER (DATA_SIZE), test_incremental_nobreak);
g_test_add_data_func ("/base64/incremental/break/1", GINT_TO_POINTER (DATA_SIZE), test_incremental_break);