Proper Asyncify list name handling (#2275)

The lists are comma separated, but the names can have internal commas since they are human-readable. This adds awareness of bracketing things, so void foo(int, double) is parsed as a single function name, properly.

Helps emscripten-core/emscripten#9128

Author: kripken

src/passes/Asyncify.cpp

@@ -1043,6 +1043,9 @@ struct Asyncify : public Pass {
     String::Split whitelist(
       runner->options.getArgumentOrDefault("asyncify-whitelist", ""), ",");
 
+    blacklist = handleBracketingOperators(blacklist);
+    whitelist = handleBracketingOperators(whitelist);
+
     // The lists contain human-readable strings. Turn them into the internal
     // escaped names for later comparisons
     auto processList = [module](String::Split& list, const std::string& which) {

src/support/string.h

@@ -31,6 +31,8 @@ namespace String {
 // Creates a vector of the split parts of a string, by a delimiter.
 class Split : public std::vector<std::string> {
 public:
+  Split() = default;
+
   Split(const std::string& input, const std::string& delim) {
     size_t lastEnd = 0;
     while (lastEnd < input.size()) {
@@ -44,6 +46,46 @@ class Split : public std::vector<std::string> {
   }
 };
 
+// Handles bracketing in a list initially split by ",", but the list may
+// contain nested ","s. For example,
+//   void foo(int, double)
+// must be kept together because of the "(". Likewise, "{", "<", "[" are
+// handled.
+inline String::Split handleBracketingOperators(String::Split split) {
+  String::Split ret;
+  std::string last;
+  int nesting = 0;
+  auto handlePart = [&](std::string part) {
+    if (part.empty()) {
+      return;
+    }
+    for (const char c : part) {
+      if (c == '(' || c == '<' || c == '[' || c == '{') {
+        nesting++;
+      } else if (c == ')' || c == '>' || c == ']' || c == '}') {
+        nesting--;
+      }
+    }
+    if (last.empty()) {
+      last = part;
+    } else {
+      last += ',' + part;
+    }
+    if (nesting == 0) {
+      ret.push_back(last);
+      last.clear();
+    }
+  };
+  for (auto& part : split) {
+    handlePart(part);
+  }
+  handlePart("");
+  if (nesting != 0) {
+    Fatal() << "Asyncify: failed to parse lists";
+  }
+  return ret;
+}
+
 // Does a simple wildcard match between a pattern and a value. Currently
 // supports a '*' at the end of the pattern.
 inline bool wildcardMatch(const std::string& pattern,

test/unit/input/asyncify-pure.wast

@@ -55,5 +55,8 @@
     (call $main)
     (call $print (i32.const 500))
   )
+  ;; interesting escaped name
+  (func $DOS_ReadFile\28unsigned\20short\2c\20unsigned\20char*\2c\20unsigned\20short*\2c\20bool\29 (param $0 i32) (param $1 i32) (param $2 i32) (param $3 i32)
+  )
 )
 

test/unit/test_asyncify.py

@@ -35,6 +35,8 @@ def test_asyncify_list_bad(self):
       ('--pass-arg=asyncify-whitelist@nonexistent', 'nonexistent'),
       ('--pass-arg=asyncify-blacklist@main', None),
       ('--pass-arg=asyncify-whitelist@main', None),
+      ('--pass-arg=asyncify-whitelist@main', None),
+      ('--pass-arg=asyncify-whitelist@DOS_ReadFile(unsigned short, unsigned char*, unsigned short*, bool)', None),
     ]:
       print(arg, warning)
       err = run_process(WASM_OPT + [self.input_path('asyncify-pure.wast'), '--asyncify', arg], stdout=subprocess.PIPE, stderr=subprocess.PIPE).stderr.strip()