From a399c7e42a8e780dc47edbac45d64db75e031fac Mon Sep 17 00:00:00 2001 From: Aleksander Morgado Date: Tue, 26 May 2020 11:45:44 +0200 Subject: charsets: take_and_convert() methods should support GSM encoding The iconv() operation would fail because we wouldn't give any proper charset string to convert to/from. Use our custom GSM encoding support instead. --- src/mm-charsets.c | 22 ++++++++++++++++++---- src/tests/test-charsets.c | 14 ++++++++++++++ 2 files changed, 32 insertions(+), 4 deletions(-) (limited to 'src') diff --git a/src/mm-charsets.c b/src/mm-charsets.c index 4b08200d..e48cec3e 100644 --- a/src/mm-charsets.c +++ b/src/mm-charsets.c @@ -458,7 +458,8 @@ mm_charset_gsm_unpacked_to_utf8 (const guint8 *gsm, guint32 len) g_byte_array_append (utf8, (guint8 *) "?", 1); } - g_byte_array_append (utf8, (guint8 *) "\0", 1); /* NULL terminator */ + /* Always make sure returned string is NUL terminated */ + g_byte_array_append (utf8, (guint8 *) "\0", 1); return g_byte_array_free (utf8, FALSE); } @@ -471,7 +472,6 @@ mm_charset_utf8_to_unpacked_gsm (const char *utf8, guint32 *out_len) int i = 0; g_return_val_if_fail (utf8 != NULL, NULL); - g_return_val_if_fail (out_len != NULL, NULL); g_return_val_if_fail (g_utf8_validate (utf8, -1, NULL), NULL); /* worst case initial length */ @@ -480,7 +480,8 @@ mm_charset_utf8_to_unpacked_gsm (const char *utf8, guint32 *out_len) if (*utf8 == 0x00) { /* Zero-length string */ g_byte_array_append (gsm, (guint8 *) "\0", 1); - *out_len = 0; + if (out_len) + *out_len = 0; return g_byte_array_free (gsm, FALSE); } @@ -501,7 +502,12 @@ mm_charset_utf8_to_unpacked_gsm (const char *utf8, guint32 *out_len) i++; } - *out_len = gsm->len; + /* Output length doesn't consider terminating NUL byte */ + if (out_len) + *out_len = gsm->len; + + /* Always make sure returned string is NUL terminated */ + g_byte_array_append (gsm, (guint8 *) "\0", 1); return g_byte_array_free (gsm, FALSE); } @@ -757,6 +763,10 @@ mm_charset_take_and_convert_to_utf8 (gchar *str, MMModemCharset charset) break; case MM_MODEM_CHARSET_GSM: + utf8 = (gchar *) mm_charset_gsm_unpacked_to_utf8 ((const guint8 *) str, strlen (str)); + g_free (str); + break; + case MM_MODEM_CHARSET_8859_1: case MM_MODEM_CHARSET_PCCP437: case MM_MODEM_CHARSET_PCDN: { @@ -881,6 +891,10 @@ mm_utf8_take_and_convert_to_charset (gchar *str, break; case MM_MODEM_CHARSET_GSM: + encoded = (gchar *) mm_charset_utf8_to_unpacked_gsm (str, NULL); + g_free (str); + break; + case MM_MODEM_CHARSET_8859_1: case MM_MODEM_CHARSET_PCCP437: case MM_MODEM_CHARSET_PCDN: { diff --git a/src/tests/test-charsets.c b/src/tests/test-charsets.c index d223d706..0931d7e8 100644 --- a/src/tests/test-charsets.c +++ b/src/tests/test-charsets.c @@ -350,6 +350,19 @@ test_take_convert_ucs2_bad_ascii2 (void) g_assert (converted == NULL); } +static void +test_take_convert_gsm_utf8 (void) +{ + gchar *src, *converted, *utf8; + + src = g_strdup ("T-Mobile"); + converted = mm_charset_take_and_convert_to_utf8 (src, MM_MODEM_CHARSET_GSM); + g_assert_cmpstr (converted, ==, "T-Mobile"); + utf8 = mm_utf8_take_and_convert_to_charset (converted, MM_MODEM_CHARSET_GSM); + g_assert_cmpstr (utf8, ==, "T-Mobile"); + g_free (utf8); +} + struct charset_can_convert_to_test_s { const char *utf8; gboolean to_gsm; @@ -432,6 +445,7 @@ int main (int argc, char **argv) g_test_add_func ("/MM/charsets/take-convert/ucs2/hex", test_take_convert_ucs2_hex_utf8); g_test_add_func ("/MM/charsets/take-convert/ucs2/bad-ascii", test_take_convert_ucs2_bad_ascii); g_test_add_func ("/MM/charsets/take-convert/ucs2/bad-ascii-2", test_take_convert_ucs2_bad_ascii2); + g_test_add_func ("/MM/charsets/take-convert/gsm", test_take_convert_gsm_utf8); g_test_add_func ("/MM/charsets/can-convert-to", test_charset_can_covert_to); -- cgit v1.2.3-70-g09d2