[clangd] Mark colon as a safe character when percent-encoding.

[kadircet]

Summary

Also change output of percent-encoding to use upper-case letters.

[sammccall]

clangd/URI.cpp

94 ↗

(On Diff #168758)

add a comment: ':' is only reserved in relative URI paths, which we never produce

[kadircet]

• Add comments.

clang-tools-extra/trunk/clangd/URI.cpp

if ((C >= 'a' && C <= 'z') || (C >= 'A' && C <= 'Z') ||
       (C >= '0' && C <= '9'))
     return false;
   switch (C) {
   case '-':
   case '_':
   case '.':
   case '~':
   case '/': // '/' is only reserved when parsing.
+  // ':' is only reserved for relative URI paths, which clangd doesn't produce.
+  case ':':
     return false;
   }
   return true;
 }
 
 /// Encodes a string according to percent-encoding.
 /// - Unreserved characters are not escaped.
 /// - Reserved characters always escaped with exceptions like '/'.
 /// - All other characters are escaped.
 std::string percentEncode(llvm::StringRef Content) {
   std::string Result;
   llvm::raw_string_ostream OS(Result);
   for (unsigned char C : Content)
     if (shouldEscape(C))
-      OS << '%' << llvm::format_hex_no_prefix(C, 2);
+      OS << '%' << llvm::format_hex_no_prefix(C, 2, /*Upper = */true);
     else
       OS << C;
 
   OS.flush();
   return Result;
 }
 
 /// Decodes a string according to percent-encoding.

clang-tools-extra/trunk/unittests/clangd/URITests.cpp

URI parseOrDie(llvm::StringRef Uri) {
   auto U = URI::parse(Uri);
   if (!U)
     llvm_unreachable(llvm::toString(U.takeError()).c_str());
   return *U;
 }
 
 TEST(PercentEncodingTest, Encode) {
   EXPECT_EQ(URI("x", /*Authority=*/"", "a/b/c").toString(), "x:a/b/c");
-  EXPECT_EQ(URI("x", /*Authority=*/"", "a!b;c~").toString(), "x:a%21b%3bc~");
+  EXPECT_EQ(URI("x", /*Authority=*/"", "a!b;c~").toString(), "x:a%21b%3Bc~");
   EXPECT_EQ(URI("x", /*Authority=*/"", "a123b").toString(), "x:a123b");
+  EXPECT_EQ(URI("x", /*Authority=*/"", "a:b;c").toString(), "x:a:b%3Bc");
 }
 
 TEST(PercentEncodingTest, Decode) {
   EXPECT_EQ(parseOrDie("x:a/b/c").body(), "a/b/c");
 
   EXPECT_EQ(parseOrDie("s%2b://%3a/%3").scheme(), "s+");
   EXPECT_EQ(parseOrDie("s%2b://%3a/%3").authority(), ":");
   EXPECT_EQ(parseOrDie("s%2b://%3a/%3").body(), "/%3");
 
   EXPECT_EQ(parseOrDie("x:a%21b%3ac~").body(), "a!b:c~");
+  EXPECT_EQ(parseOrDie("x:a:b%3bc").body(), "a:b;c");
 }
 
 std::string resolveOrDie(const URI &U, llvm::StringRef HintPath = "") {
   auto Path = URI::resolve(U, HintPath);
   if (!Path)
     llvm_unreachable(llvm::toString(Path.takeError()).c_str());
   return *Path;
 }
 
 TEST(URITest, Create) {
 #ifdef _WIN32
-  EXPECT_THAT(createOrDie("c:\\x\\y\\z"), "file:///c%3a/x/y/z");
+  EXPECT_THAT(createOrDie("c:\\x\\y\\z"), "file:///c:/x/y/z");
 #else
   EXPECT_THAT(createOrDie("/x/y/z"), "file:///x/y/z");
-  EXPECT_THAT(createOrDie("/(x)/y/\\ z"), "file:///%28x%29/y/%5c%20z");
+  EXPECT_THAT(createOrDie("/(x)/y/\\ z"), "file:///%28x%29/y/%5C%20z");
 #endif
 }
 
 TEST(URITest, FailedCreate) {
   auto Fail = [](llvm::Expected<URI> U) {
     if (!U) {
       llvm::consumeError(U.takeError());
       return true;
TEST(URITest, ParseFailed) {
   EXPECT_TRUE(FailedParse(""));
   EXPECT_TRUE(FailedParse(":/a/b/c"));
   EXPECT_TRUE(FailedParse("\"/a/b/c\" IWYU pragma: abc"));
 }
 
 TEST(URITest, Resolve) {
 #ifdef _WIN32
   EXPECT_THAT(resolveOrDie(parseOrDie("file:///c%3a/x/y/z")), "c:\\x\\y\\z");
+  EXPECT_THAT(resolveOrDie(parseOrDie("file:///c:/x/y/z")), "c:\\x\\y\\z");
 #else
   EXPECT_EQ(resolveOrDie(parseOrDie("file:/a/b/c")), "/a/b/c");
   EXPECT_EQ(resolveOrDie(parseOrDie("file://auth/a/b/c")), "/a/b/c");
   EXPECT_THAT(resolveOrDie(parseOrDie("file://au%3dth/%28x%29/y/%20z")),
               "/(x)/y/ z");
   EXPECT_THAT(resolveOrDie(parseOrDie("file:///c:/x/y/z")), "c:/x/y/z");
 #endif
   EXPECT_EQ(resolveOrDie(parseOrDie("unittest:///a"), testPath("x")),