wgsl-reader: reject identifiers starting with underscrore

Update one test to avoid this error.

Fixed: tint:1179
Change-Id: Id41b0eb0f404648de4e86a835fe43f1729cb4696
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/64464
Auto-Submit: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: David Neto <dneto@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

src/reader/wgsl/lexer.cc

@@ -119,15 +119,15 @@ bool Lexer::is_eof() const {
 }
 
 bool Lexer::is_alpha(char ch) const {
-  return std::isalpha(ch) || ch == '_';
+  return std::isalpha(ch);
 }
 
 bool Lexer::is_digit(char ch) const {
   return std::isdigit(ch);
 }
 
-bool Lexer::is_alphanum(char ch) const {
-  return is_alpha(ch) || is_digit(ch);
+bool Lexer::is_alphanum_underscore(char ch) const {
+  return is_alpha(ch) || is_digit(ch) || ch == '_';
 }
 
 bool Lexer::is_hex(char ch) const {
@@ -660,7 +660,7 @@ Token Lexer::try_integer() {
 }
 
 Token Lexer::try_ident() {
-  // Must begin with an a-zA-Z_
+  // Must begin with an a-zA-Z
   if (!is_alpha(content_->data[pos_])) {
     return {};
   }
@@ -668,7 +668,7 @@ Token Lexer::try_ident() {
   auto source = begin_source();
 
   auto s = pos_;
-  while (!is_eof() && is_alphanum(content_->data[pos_])) {
+  while (!is_eof() && is_alphanum_underscore(content_->data[pos_])) {
     pos_++;
     location_.column++;
   }

src/reader/wgsl/lexer.h

@@ -56,10 +56,19 @@ class Lexer {
   void end_source(Source&) const;
 
   bool is_eof() const;
+  /// @param ch a character
+  /// @returns true if 'ch' is an alphabetic character
   bool is_alpha(char ch) const;
+  /// @param ch a character
+  /// @returns true if 'ch' is a decimal digit
   bool is_digit(char ch) const;
+  /// @param ch a character
+  /// @returns true if 'ch' is a hexadecimal digit
   bool is_hex(char ch) const;
-  bool is_alphanum(char ch) const;
+  /// @param ch a character
+  /// @returns true if 'ch' is a digit, an alphabetic character,
+  /// or an underscore.
+  bool is_alphanum_underscore(char ch) const;
   bool matches(size_t pos, const std::string& substr);
 
   /// The source file path

src/reader/wgsl/lexer_test.cc

@@ -186,10 +186,26 @@ TEST_P(IdentifierTest, Parse) {
   EXPECT_EQ(t.source().range.end.column, 1u + strlen(GetParam()));
   EXPECT_EQ(t.to_str(), GetParam());
 }
-INSTANTIATE_TEST_SUITE_P(
-    LexerTest,
-    IdentifierTest,
-    testing::Values("test01", "_test_", "test_", "_test", "_01", "_test01"));
+INSTANTIATE_TEST_SUITE_P(LexerTest,
+                         IdentifierTest,
+                         testing::Values("a",
+                                         "test",
+                                         "test01",
+                                         "test_",
+                                         "test_01",
+                                         "ALLCAPS",
+                                         "MiXeD_CaSe",
+                                         "abcdefghijklmnopqrstuvwxyz",
+                                         "ABCDEFGHIJKLMNOPQRSTUVWXYZ",
+                                         "alldigits_0123456789"));
+
+TEST_F(LexerTest, IdentifierTest_DoesNotStartWithUnderscore) {
+  Source::FileContent content("_test");
+  Lexer l("test.wgsl", &content);
+
+  auto t = l.next();
+  EXPECT_FALSE(t.IsIdentifier());
+}
 
 TEST_F(LexerTest, IdentifierTest_DoesNotStartWithNumber) {
   Source::FileContent content("01test");