Generate correct C literals for non-ascii characters

spec/compiler/util/string_helpers_spec.cc

@@ -0,0 +1,27 @@
+#include "compiler/compiler_spec_helper.h"
+#include "compiler/util/string_helpers.h"
+
+using namespace rules;
+using util::escape_char;
+
+START_TEST
+
+describe("escape_char", []() {
+  it("returns ascii characters as strings", [&]() {
+    AssertThat(escape_char('x'), Equals("'x'"));
+  });
+
+  it("escapes special characters with backslashes", [&]() {
+    AssertThat(escape_char('\\'), Equals("'\\\\'"));
+    AssertThat(escape_char('\n'), Equals("'\\n'"));
+    AssertThat(escape_char('\t'), Equals("'\\t'"));
+    AssertThat(escape_char('\r'), Equals("'\\r'"));
+    AssertThat(escape_char('\''), Equals("'\\''"));
+  });
+
+  it("prints non-ascii characters as numbers", [&]() {
+    AssertThat(escape_char(256), Equals("256"));
+  });
+});
+
+END_TEST

src/compiler/rules/character_range.cc

@@ -1,4 +1,5 @@
 #include "compiler/rules/character_range.h"
+#include "compiler/util/string_helpers.h"
 #include <algorithm>
 #include <string>
 
@@ -27,29 +28,11 @@ bool CharacterRange::operator<(const CharacterRange &other) const {
   return false;
 }
 
-string escape_character(uint32_t input) {
-  if (input >= 255)
-    return to_string(input);
-
-  switch (input) {
-    case '\0':
-      return "\\0";
-    case '\n':
-      return "\\n";
-    case '\r':
-      return "\\r";
-    case '\t':
-      return "\\t";
-    default:
-      return string() + static_cast<char>(input);
-  }
-}
-
 string CharacterRange::to_string() const {
   if (min == max)
-    return escape_character(min);
+    return util::escape_char(min);
   else
-    return string() + escape_character(min) + "-" + escape_character(max);
+    return string() + util::escape_char(min) + "-" + util::escape_char(max);
 }
 
 ostream &operator<<(ostream &stream, const CharacterRange &range) {

src/compiler/util/string_helpers.cc

@@ -7,6 +7,7 @@ namespace util {
 using std::string;
 using std::vector;
 using std::set;
+using std::to_string;
 
 void str_replace(string *input, const string &search, const string &replace) {
   size_t pos = 0;
@@ -26,7 +27,7 @@ string escape_string(string input) {
   return input;
 }
 
-string escape_char(char character) {
+string escape_char(uint32_t character) {
   switch (character) {
     case '"':
       return "'\\\"'";
@@ -42,11 +43,9 @@ string escape_char(char character) {
       return "'\\\\'";
     default:
       if (character >= ' ' && character <= '~') {
-        return string("'") + character + "'";
+        return string("'") + static_cast<char>(character) + "'";
       } else {
-        char buffer[5];
-        snprintf(buffer, sizeof(buffer), "%d", static_cast<int>(character));
-        return string(buffer);
+        return to_string(character);
       }
   }
 }

src/compiler/util/string_helpers.h

@@ -11,7 +11,7 @@ namespace util {
 void str_replace(std::string *input, const std::string &search,
                  const std::string &replace);
 std::string escape_string(std::string input);
-std::string escape_char(char character);
+std::string escape_char(uint32_t character);
 
 }  // namespace util
 }  // namespace tree_sitter