demangler: Support the future final mangling prefix $S #10588
Conversation
As it’s not clear if we will need underscores, the demangler accepts $S and _$S. Note that a double underscore is handled by the demangler client. rdar://problem/32251811
|
@swift-ci Please smoke test and merge |
|
@swift-ci Please smoke test macos |
| /// Returns the length of the swift mangling prefix of the \p SymbolName. | ||
| /// | ||
| /// Returns 0 if \p SymbolName is not a mangled swift (>= swift 4.x) name. | ||
| int getManglingPrefixLength(const char *mangledName); |
There was a problem hiding this comment.
Could this be Optional<StringRef> getManglingPrefix(StringRef), where None represents no prefix and a StringRef is the actual data of the prefix? It seems like a lot of the users are thinking in those terms anyway e.g. the use in swift-demangler below could be:
auto prefix = swift::Demangle::getManglingPrefix(remangled);
if (!name.startswith(*prefix)) {
auto namePrefix = swift::Demangle::getManglingPrefix(name);
remangled = namePrefix + remangled.substr(prefix->size());
}There was a problem hiding this comment.
It could, but I don't see the benefit of it.
I don't see the need of Optional because the None case could simply be an empty StringRef. But then, why returning the (char_ptr, size) (= StringRef) and not just the size if the char-ptr is not used? We would introduce the implicit assumption that the returned StringRef is always the prefix of the argument. Well, I prefer to just return the size.
As it’s not clear if we will need underscores, the demangler accepts $S and _$S.
Note that a double underscore is handled by the demangler client.
rdar://problem/32251811