gbdev/rgbds
1
0
Fork 0
mirror of https://github.com/gbdev/rgbds.git synced 2025-11-20 18:22:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

Reduce nesting depth in lexer.cpp

Browse Source
This commit is contained in:
Rangi42
2025-07-12 07:54:16 -04:00
parent d32b1912ed
commit da133baf17
1 changed files with 47 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

95
src/asm/lexer.cpp
View File

@@ -1953,62 +1953,61 @@ static Token yylex_NORMAL() {
bool raw = c == '#'; bool raw = c == '#';
if (raw && startsIdentifier(peek())) { if (raw && startsIdentifier(peek())) {
c = nextChar(); c = nextChar();
} else if (!startsIdentifier(c)) {
// Do not report weird characters when capturing, it'll be done later
if (!lexerState->capturing) {
reportGarbageCharacters(c);
}
break;
} }
if (startsIdentifier(c)) { Token token = readIdentifier(c, raw);
Token token = readIdentifier(c, raw);
// An ELIF after a taken IF needs to not evaluate its condition // An ELIF after a taken IF needs to not evaluate its condition
if (token.type == T_(POP_ELIF) && lexerState->lastToken == T_(NEWLINE) if (token.type == T_(POP_ELIF) && lexerState->lastToken == T_(NEWLINE)
&& lexer_GetIFDepth() > 0 && lexer_RanIFBlock() && !lexer_ReachedELSEBlock()) { && lexer_GetIFDepth() > 0 && lexer_RanIFBlock() && !lexer_ReachedELSEBlock()) {
return yylex_SKIP_TO_ENDC(); return yylex_SKIP_TO_ENDC();
} }
// If a keyword, don't try to expand
if (token.type != T_(SYMBOL) && token.type != T_(LOCAL)) {
return token;
}
// `token` is either a `SYMBOL` or a `LOCAL`, and both have a `std::string` value.
assume(std::holds_alternative<std::string>(token.value));
// Raw symbols and local symbols cannot be string expansions
if (!raw && token.type == T_(SYMBOL) && lexerState->expandStrings) {
// Attempt string expansion
Symbol const *sym = sym_FindExactSymbol(std::get<std::string>(token.value));
if (sym && sym->type == SYM_EQUS) {
std::shared_ptr<std::string> str = sym->getEqus();
assume(str);
beginExpansion(str, sym->name);
continue; // Restart, reading from the new buffer
}
}
// This is a "lexer hack"! We need it to distinguish between label definitions
// (which start with `LABEL`) and macro invocations (which start with `SYMBOL`).
//
// If we had one `IDENTIFIER` token, the parser would need to perform "lookahead"
// to determine which rule applies. But since macros need to enter "raw" mode to
// parse their arguments, which may not even be valid tokens in "normal" mode, we
// cannot use lookahead to check for the presence of a `COLON`.
//
// Instead, we have separate `SYMBOL` and `LABEL` tokens, lexing as a `LABEL` if a
// ':' character *immediately* follows the identifier. Thus, at the beginning of a
// line, "Label:" and "mac:" are treated as label definitions, but "Label :" and
// "mac :" are treated as macro invocations.
if (token.type == T_(SYMBOL) && peek() == ':') {
token.type = T_(LABEL);
}
// If a keyword, don't try to expand
if (token.type != T_(SYMBOL) && token.type != T_(LOCAL)) {
return token; return token;
} }
// Do not report weird characters when capturing, it'll be done later // `token` is either a `SYMBOL` or a `LOCAL`, and both have a `std::string` value.
if (!lexerState->capturing) { assume(std::holds_alternative<std::string>(token.value));
reportGarbageCharacters(c);
// Raw symbols and local symbols cannot be string expansions
if (!raw && token.type == T_(SYMBOL) && lexerState->expandStrings) {
// Attempt string expansion
Symbol const *sym = sym_FindExactSymbol(std::get<std::string>(token.value));
if (sym && sym->type == SYM_EQUS) {
std::shared_ptr<std::string> str = sym->getEqus();
assume(str);
beginExpansion(str, sym->name);
continue; // Restart, reading from the new buffer
}
} }
// This is a "lexer hack"! We need it to distinguish between label definitions
// (which start with `LABEL`) and macro invocations (which start with `SYMBOL`).
//
// If we had one `IDENTIFIER` token, the parser would need to perform "lookahead"
// to determine which rule applies. But since macros need to enter "raw" mode to
// parse their arguments, which may not even be valid tokens in "normal" mode, we
// cannot use lookahead to check for the presence of a `COLON`.
//
// Instead, we have separate `SYMBOL` and `LABEL` tokens, lexing as a `LABEL` if a
// ':' character *immediately* follows the identifier. Thus, at the beginning of a
// line, "Label:" and "mac:" are treated as label definitions, but "Label :" and
// "mac :" are treated as macro invocations.
if (token.type == T_(SYMBOL) && peek() == ':') {
token.type = T_(LABEL);
}
return token;
} }
lexerState->atLineStart = false; lexerState->atLineStart = false;
} }