diff --git a/test/main.cpp b/test/main.cpp
index c0c22a0..52a7145 100644
--- a/test/main.cpp
+++ b/test/main.cpp
@@ -11,33 +11,63 @@ static scan_arc_ctx_t arc_list_ctx;
 static scan_text_ctx_t text_500_ctx;
 
 
-TEST(TextTest, BookCsvContentLen) {
-    const char *filepath = "libscan-test-files/test_files/text/books.csv";
-
+TEST(Text, BookCsvContentLen) {
     vfile_t f;
     document_t doc;
-    doc.meta_head = nullptr;
-    doc.meta_tail = nullptr;
-    load_file(filepath, &f);
+    load_doc_file("libscan-test-files/test_files/text/books.csv", &f, &doc);
+
     parse_text(&text_500_ctx, &f, &doc);
 
     ASSERT_NEAR(strlen(get_meta(&doc, MetaContent)->str_val), 500, 1);
-    CLOSE_FILE(f)
-    destroy_doc(&doc);
+    cleanup(&doc, &f);
 }
 
-TEST(TextTest, MemUtf8_1) {
-    const char *content = "a"; //todo
-
+TEST(Text, MemUtf8_1) {
+    const char *content = "a";
     vfile_t f;
     document_t doc;
-    doc.meta_head = nullptr;
-    doc.meta_tail = nullptr;
-    load_mem((void *) content, strlen(content), &f);
+    load_doc_mem((void *) content, strlen(content), &f, &doc);
+
     parse_text(&text_500_ctx, &f, &doc);
 
     ASSERT_EQ(strlen(get_meta(&doc, MetaContent)->str_val), 1);
-    destroy_doc(&doc);
+    cleanup(&doc, &f);
+}
+
+TEST(Text, MemUtf8_Invalid1) {
+    const char *content = "12345\xE0";
+    vfile_t f;
+    document_t doc;
+    load_doc_mem((void *) content, strlen(content), &f, &doc);
+
+    parse_text(&text_500_ctx, &f, &doc);
+
+    ASSERT_STREQ(get_meta(&doc, MetaContent)->str_val, "12345");
+    cleanup(&doc, &f);
+}
+
+TEST(Text, MemUtf8_2) {
+    const char *content = "最後測試";
+    vfile_t f;
+    document_t doc;
+    load_doc_mem((void *) content, strlen(content), &f, &doc);
+
+    parse_text(&text_500_ctx, &f, &doc);
+
+    ASSERT_STREQ(get_meta(&doc, MetaContent)->str_val, "最後測試");
+    cleanup(&doc, &f);
+}
+
+TEST(Text, MemUtf8_Invalid2) {
+    const char *content = "最後測\xe8\xa9";
+    vfile_t f;
+    document_t doc;
+    load_doc_mem((void *) content, strlen(content), &f, &doc);
+
+    parse_text(&text_500_ctx, &f, &doc);
+
+    ASSERT_STREQ(get_meta(&doc, MetaContent)->str_val, "最後測");
+    cleanup(&doc, &f);
 }
 
 
diff --git a/test/test_util.cpp b/test/test_util.cpp
index 8aef3c1..60bb979 100644
--- a/test/test_util.cpp
+++ b/test/test_util.cpp
@@ -32,6 +32,23 @@ void fs_close(vfile_t *f) {
     }
 }
 
+void load_doc_file(const char *filepath, vfile_t *f, document_t *doc) {
+    doc->meta_head = nullptr;
+    doc->meta_tail = nullptr;
+    load_file(filepath, f);
+}
+
+void load_doc_mem(void *mem, size_t mem_len, vfile_t *f, document_t *doc) {
+    doc->meta_head = nullptr;
+    doc->meta_tail = nullptr;
+    load_mem(mem, mem_len, f);
+}
+
+void cleanup(document_t *doc, vfile_t *f) {
+    destroy_doc(doc);
+    CLOSE_FILE((*f))
+}
+
 void load_file(const char *filepath, vfile_t *f) {
     stat(filepath, &f->info);
     f->fd = open(filepath, O_RDONLY);
diff --git a/test/test_util.h b/test/test_util.h
index 3df9400..abc90b4 100644
--- a/test/test_util.h
+++ b/test/test_util.h
@@ -5,6 +5,9 @@
 
 void load_file(const char *filepath, vfile_t *f);
 void load_mem(void *mem, size_t size, vfile_t *f);
+void load_doc_mem(void *mem, size_t mem_len, vfile_t *f, document_t *doc);
+void load_doc_file(const char *filepath, vfile_t *f, document_t *doc);
+void cleanup(document_t *doc, vfile_t *f);
 
 static void noop_logf(char *filepath, int level, char *format, ...) {
     // noop