Add tests for unicode

tjysdsg · May 13, 2020 · e83b10a · e83b10a
1 parent 86fd8a3
commit e83b10a
Show file tree

Hide file tree

Showing 4 changed files with 17 additions and 4 deletions.
diff --git a/.idea/runConfigurations/tanc.xml b/.idea/runConfigurations/tanc.xml
diff --git a/include/reader.h b/include/reader.h
@@ -19,7 +19,6 @@ struct line_info {
   line_info(const size_t lineno, const std::string &code) : lineno(lineno), code(code) {}
 };
 
-// TODO: support unicode
 class Reader final {
 public:
   Reader() = default;

diff --git a/src/lexer/lexer.cpp b/src/lexer/lexer.cpp
@@ -188,7 +188,7 @@ Token *tokenize_char(Reader *reader, cursor &start) {
       }
       value = std::string(1, escape_char(value[1]));
     } else if (value.length() != 1) {
-      report_code_error(reader->get_line(forward.l).code, forward.l, forward.c, "Incomplete character literal");
+      report_code_error(reader->get_line(forward.l).code, forward.l, forward.c, "Invalid character literal");
     }
     t = new Token(TokenType::CHAR, value, start, &reader->get_line(start.l));
     start = (*reader).forward(forward);

diff --git a/src/test/test_src/string.tan b/src/test/test_src/string.tan
@@ -2,11 +2,25 @@ import "runtime/print.tan";
 import "runtime/debug.tan";
 
 pub fn main(argc: int, argv: char**) : int {
-    var s: str = "shit";
+    var s: str = "shit 他妈的";
     assert(s[0] == 's');
     assert(s[1] == 'h');
     assert(s[2] == 'i');
     assert(s[3] == 't');
+    assert(s[4] == ' ');
+    var ta = "他";
+    assert(s[5] == ta[0]);
+    assert(s[6] == ta[1]);
+    assert(s[7] == ta[2]);
+    var ma = "妈";
+    assert(s[8] == ma[0]);
+    assert(s[9] == ma[1]);
+    assert(s[10] == ma[2]);
+    var de = "的";
+    assert(s[11] == de[0]);
+    assert(s[12] == de[1]);
+    assert(s[13] == de[2]);
     print("SUCCESS\n");
+    print(s);
     return 0;
 }