test: utf8 - fix utf16 tests on BE machines

[elogind.git] / src / test / test-utf8.c

diff --git a/src/test/test-utf8.c b/src/test/test-utf8.c
index 26cc37bd6ea5642d7405531c7aefb20aa1476b58..346f8524c620c4fbdda4337b8db4489cbab15d00 100644 (file)
--- a/src/test/test-utf8.c
+++ b/src/test/test-utf8.c
@@ -19,50 +19,100 @@
   along with systemd; If not, see <http://www.gnu.org/licenses/>.
 ***/
 
-
 #include "utf8.h"
 #include "util.h"
 
-/* helpers for test_udev_encode_string */
-static char *do_encode_string(const char *in) {
-        size_t out_len = strlen(in) * 4;
-        char *out = malloc(out_len);
+static void test_utf8_is_printable(void) {
+        assert_se(utf8_is_printable("ascii is valid\tunicode", 22));
+        assert_se(utf8_is_printable("\342\204\242", 3));
+        assert_se(!utf8_is_printable("\341\204", 2));
+        assert_se(utf8_is_printable("ąę", 4));
+}
 
-        assert_se(out);
-        assert_se(udev_encode_string(in, out, out_len) >= 0);
-        puts(out);
+static void test_utf8_is_valid(void) {
+        assert_se(utf8_is_valid("ascii is valid unicode"));
+        assert_se(utf8_is_valid("\342\204\242"));
+        assert_se(!utf8_is_valid("\341\204"));
+}
 
-        return out;
+static void test_ascii_is_valid(void) {
+        assert_se(ascii_is_valid("alsdjf\t\vbarr\nba z"));
+        assert_se(!ascii_is_valid("\342\204\242"));
+        assert_se(!ascii_is_valid("\341\204"));
 }
 
-static bool expect_encoded_as(const char *in, const char *expected) {
-        _cleanup_free_ char *encoded = do_encode_string(in);
-        return streq(encoded, expected);
+static void test_utf8_encoded_valid_unichar(void) {
+        assert_se(utf8_encoded_valid_unichar("\342\204\242") == 3);
+        assert_se(utf8_encoded_valid_unichar("\302\256") == 2);
+        assert_se(utf8_encoded_valid_unichar("a") == 1);
+        assert_se(utf8_encoded_valid_unichar("\341\204") < 0);
+        assert_se(utf8_encoded_valid_unichar("\341\204\341\204") < 0);
 }
 
-static void test_udev_encode_string(void) {
-        assert_se(expect_encoded_as("systemd sucks", "systemd\\x20sucks"));
-        assert_se(expect_encoded_as("pinkiepie", "pinkiepie"));
-        assert_se(expect_encoded_as("valíd\\ųtf8", "valíd\\x5cųtf8"));
-        assert_se(expect_encoded_as("s/ash/ng", "s\\x2fash\\x2fng"));
+static void test_utf8_escaping(void) {
+        _cleanup_free_ char *p1, *p2, *p3;
+
+        p1 = utf8_escape_invalid("goo goo goo");
+        puts(p1);
+        assert_se(utf8_is_valid(p1));
+
+        p2 = utf8_escape_invalid("\341\204\341\204");
+        puts(p2);
+        assert_se(utf8_is_valid(p2));
+
+        p3 = utf8_escape_invalid("\341\204");
+        puts(p3);
+        assert_se(utf8_is_valid(p3));
 }
 
-static void test_utf8_is_printable(void) {
-        assert_se(utf8_is_printable("ascii is valid\tunicode", 22));
-        assert_se(utf8_is_printable("\342\204\242", 3));
-        assert_se(!utf8_is_printable("\341\204", 2));
+static void test_utf8_escaping_printable(void) {
+        _cleanup_free_ char *p1, *p2, *p3, *p4, *p5, *p6;
+
+        p1 = utf8_escape_non_printable("goo goo goo");
+        puts(p1);
+        assert_se(utf8_is_valid(p1));
+
+        p2 = utf8_escape_non_printable("\341\204\341\204");
+        puts(p2);
+        assert_se(utf8_is_valid(p2));
+
+        p3 = utf8_escape_non_printable("\341\204");
+        puts(p3);
+        assert_se(utf8_is_valid(p3));
+
+        p4 = utf8_escape_non_printable("ąę\n가너도루\n1234\n\341\204\341\204\n\001 \019\20\a");
+        puts(p4);
+        assert_se(utf8_is_valid(p4));
+
+        p5 = utf8_escape_non_printable("\001 \019\20\a");
+        puts(p5);
+        assert_se(utf8_is_valid(p5));
+
+        p6 = utf8_escape_non_printable("\xef\xbf\x30\x13");
+        puts(p6);
+        assert_se(utf8_is_valid(p6));
 }
 
-static void test_utf8_is_valid(void) {
-        assert_se(utf8_is_valid("ascii is valid unicode"));
-        assert_se(utf8_is_valid("\341\204\242"));
-        assert_se(!utf8_is_valid("\341\204"));
+static void test_utf16_to_utf8(void) {
+        char *a = NULL;
+        const uint16_t utf16[] = { htole16('a'), htole16(0xd800), htole16('b'), htole16(0xdc00), htole16('c'), htole16(0xd801), htole16(0xdc37) };
+        const char utf8[] = { 'a', 'b', 'c', 0xf0, 0x90, 0x90, 0xb7, 0 };
+
+        a = utf16_to_utf8(utf16, 14);
+        assert_se(a);
+        assert_se(streq(a, utf8));
+
+        free(a);
 }
 
 int main(int argc, char *argv[]) {
         test_utf8_is_valid();
         test_utf8_is_printable();
-        test_udev_encode_string();
+        test_ascii_is_valid();
+        test_utf8_encoded_valid_unichar();
+        test_utf8_escaping();
+        test_utf8_escaping_printable();
+        test_utf16_to_utf8();
 
         return 0;
 }