utils: not to convert codepoint string (#3280) (#3284)

* utils: not to convert codepoint string (#3280)
* tests: internal: utils: compare utf8 encoding data

Signed-off-by: Takahiro Yamashita <[email protected]>

Author: nokute78

src/flb_utils.c

@@ -716,29 +716,11 @@ int flb_utils_write_str(char *buf, int *off, size_t size,
             if (i + hex_bytes > str_len) {
                 break; /* skip truncated UTF-8 */
             }
-
-            state = FLB_UTF8_ACCEPT;
-            codepoint = 0;
             for (b = 0; b < hex_bytes; b++) {
-                s = (unsigned char *) str + i + b;
-                ret = flb_utf8_decode(&state, &codepoint, *s);
-                if (ret == 0) {
-                    break;
-                }
-            }
-
-            if (state != FLB_UTF8_ACCEPT) {
-                /* Invalid UTF-8 hex, just skip utf-8 bytes */
-                flb_warn("[pack] invalid UTF-8 bytes found, skipping bytes");
-            }
-            else {
-                len = snprintf(tmp, sizeof(tmp) - 1, "\\u%04x", codepoint);
-                if ((available - written) < len) {
-                    return FLB_FALSE;
-                }
-                encoded_to_buf(p, tmp, len);
-                p += len;
+                tmp[b] = str[i+b];
             }
+            encoded_to_buf(p, tmp, hex_bytes);
+            p += hex_bytes;
             i += (hex_bytes - 1);
         }
         else {

tests/internal/data/pack/mixed_002.json

@@ -1 +1 @@
-"mixed_002 =>\n\n  \u00e1\u00e9\u00ed\u00f3\u00fa\n\n\n'\n\\t\n"
+"mixed_002 =>\n\n  áéíóú\n\n\n'\n\\t\n"

tests/internal/data/pack/mixed_003.json

@@ -1 +1 @@
-"\u00e1\n"
+"á\n"

tests/internal/data/pack/utf8_bell.json

@@ -1 +1 @@
-"\u1f514"
+"🔔"

tests/internal/data/pack/utf8_copyright.json

@@ -1 +1 @@
-"\u00a9"
+"©"

tests/internal/data/pack/utf8_hokke.json

@@ -1 +1 @@
-"\u29e3d"
+"𩸽"

tests/internal/data/pack/utf8_relaxed.json

@@ -1 +1 @@
-"\u263a"
+"☺"

tests/internal/utils.c

@@ -115,6 +115,7 @@ void test_url_split()
 void test_write_str()
 {
     char buf[10];
+    char japanese_a[4] = {0xe3, 0x81, 0x82};
     int size = sizeof(buf);
     int off;
     int ret;
@@ -132,13 +133,13 @@ void test_write_str()
     off = 0;
     ret = flb_utils_write_str(buf, &off, size, "\xe3\x81\x82", 3);
     TEST_CHECK(ret == FLB_TRUE);
-    TEST_CHECK(memcmp(buf, "\\u3042", off) == 0);
+    TEST_CHECK(memcmp(buf, japanese_a, off) == 0);
 
     // Truncated bytes
     off = 0;
     ret = flb_utils_write_str(buf, &off, size, "\xe3\x81\x82\xe3", 1);
     TEST_CHECK(ret == FLB_TRUE);
-    TEST_CHECK(memcmp(buf, "\\u3042", off) == 0);
+    TEST_CHECK(memcmp(buf, japanese_a, off) == 0);
 
     // Error: buffer too small
     off = 0;