Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Ansi c frontend feqpkqp #1224

Merged
merged 3 commits into from
Feb 7, 2025
Merged

Ansi c frontend feqpkqp #1224

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b6c501b
floating
andrew-johnson-4 Feb 7, 2025
2306cc8
tokens carry smart strings now
andrew-johnson-4 Feb 7, 2025
3326180
c-parse
andrew-johnson-4 Feb 7, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
48,192 changes: 24,701 additions & 23,491 deletions BOOTSTRAP/cli.c
View file
Open in desktop

Large diffs are not rendered by default.

173 changes: 88 additions & 85 deletions PLUGINS/FRONTEND/C/c-parse.lsts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,10 @@
# However, there are also a large number of compiler-specific extensions that are parsed but mostly ignored
# Example: __extension__ ( f, g )

type CConstant = CConstantInteger{value:CString}
| CConstantCharacter{value:CString}
| CConstantFloating{value:CString}
| CConstantEnumeration{value:CString};
type CConstant = CConstantInteger{value:String}
| CConstantCharacter{value:String}
| CConstantFloating{value:String}
| CConstantEnumeration{value:String};

let cmp(l: CConstant, r: CConstant): Ord = (
if $".0"(l) != $".0"(r) then cmp($".0"(l), $".0"(r))
Expand All @@ -25,78 +25,81 @@ let std-c-parse(tokens: List<Token>): Nil = (
# while non-zero(tokens) { tokens = std-c-parse-external-declaration(tokens); }
);

let std-c-is-reserved-word(tk: CString): U64 = (
let std-c-is-reserved-word(tk: String): U64 = (
let reserved = false;
if tk == c"auto" then (reserved = true);
if tk == c"double" then (reserved = true);
if tk == c"int" then (reserved = true);
if tk == c"struct" then (reserved = true);
if tk == c"break" then (reserved = true);
if tk == c"else" then (reserved = true);
if tk == c"long" then (reserved = true);
if tk == c"switch" then (reserved = true);
if tk == c"case" then (reserved = true);
if tk == c"enum" then (reserved = true);
if tk == c"register" then (reserved = true);
if tk == c"typedef" then (reserved = true);
if tk == c"char" then (reserved = true);
if tk == c"extern" then (reserved = true);
if tk == c"return" then (reserved = true);
if tk == c"union" then (reserved = true);
if tk == c"const" then (reserved = true);
if tk == c"float" then (reserved = true);
if tk == c"short" then (reserved = true);
if tk == c"unsigned" then (reserved = true);
if tk == c"continue" then (reserved = true);
if tk == c"for" then (reserved = true);
if tk == c"signed" then (reserved = true);
if tk == c"void" then (reserved = true);
if tk == c"default" then (reserved = true);
if tk == c"goto" then (reserved = true);
if tk == c"sizeof" then (reserved = true);
if tk == c"volatile" then (reserved = true);
if tk == c"do" then (reserved = true);
if tk == c"if" then (reserved = true);
if tk == c"static" then (reserved = true);
if tk == c"while" then (reserved = true);
if tk == c"_Bool" then (reserved = true);
if tk == c"_Imaginary" then (reserved = true);
if tk == c"restrict" then (reserved = true);
if tk == c"_Complex" then (reserved = true);
if tk == c"inline" then (reserved = true);
if tk == c"_Alignas" then (reserved = true);
if tk == c"_Generic" then (reserved = true);
if tk == c"_Thread_local" then (reserved = true);
if tk == c"_Alignof" then (reserved = true);
if tk == c"_Noreturn" then (reserved = true);
if tk == c"_Atomic" then (reserved = true);
if tk == c"_Static_assert" then (reserved = true);
if tk == "auto" then (reserved = true);
if tk == "double" then (reserved = true);
if tk == "int" then (reserved = true);
if tk == "struct" then (reserved = true);
if tk == "break" then (reserved = true);
if tk == "else" then (reserved = true);
if tk == "long" then (reserved = true);
if tk == "switch" then (reserved = true);
if tk == "case" then (reserved = true);
if tk == "enum" then (reserved = true);
if tk == "register" then (reserved = true);
if tk == "typedef" then (reserved = true);
if tk == "char" then (reserved = true);
if tk == "extern" then (reserved = true);
if tk == "return" then (reserved = true);
if tk == "union" then (reserved = true);
if tk == "const" then (reserved = true);
if tk == "float" then (reserved = true);
if tk == "short" then (reserved = true);
if tk == "unsigned" then (reserved = true);
if tk == "continue" then (reserved = true);
if tk == "for" then (reserved = true);
if tk == "signed" then (reserved = true);
if tk == "void" then (reserved = true);
if tk == "default" then (reserved = true);
if tk == "goto" then (reserved = true);
if tk == "sizeof" then (reserved = true);
if tk == "volatile" then (reserved = true);
if tk == "do" then (reserved = true);
if tk == "if" then (reserved = true);
if tk == "static" then (reserved = true);
if tk == "while" then (reserved = true);
if tk == "_Bool" then (reserved = true);
if tk == "_Imaginary" then (reserved = true);
if tk == "restrict" then (reserved = true);
if tk == "_Complex" then (reserved = true);
if tk == "inline" then (reserved = true);
if tk == "_Alignas" then (reserved = true);
if tk == "_Generic" then (reserved = true);
if tk == "_Thread_local" then (reserved = true);
if tk == "_Alignof" then (reserved = true);
if tk == "_Noreturn" then (reserved = true);
if tk == "_Atomic" then (reserved = true);
if tk == "_Static_assert" then (reserved = true);
reserved
);

let std-c-has-class(tk: CString, cls: String): U64 = (
let std-c-has-class(tks: String, cls: String): U64 = (
let tk = untern(tks);
match cls {
"identifier" => tk == r/^[a-zA-Z_][a-zA-Z0-9_]*/ && not(std-c-is-reserved-word(tk));
"identifier" => tk == r/^[a-zA-Z_][a-zA-Z0-9_]*/ && not(std-c-is-reserved-word(tks));
"integer" => tk == r/^[0-9]+([uU]|[lL]|wb|WB)*/ # decimal constant
|| tk == r/^[0][0-7]+([uU]|[lL]|wb|WB)*/ # octal constant
|| tk == r/^[0][x][0-9a-fA-F]+([uU]|[lL]|wb|WB)*/ # hexadecimal constant
|| tk == r/^[0][bB][01]+([uU]|[lL]|wb|WB)*/; # binary constant
"character" => tk == r/^(u8|u|U|L)?[']([^']|([\\][']))+[']/; # character constant
"floating" => tk == r/^[0-9]+([.][0-9]+)?([eE][0-9]+)?[fF]?/ # decimal constant
|| tk == r/^[0][x][0-9a-fA-F]+([.][0-9a-fA-F]+)?([eE][0-9a-fA-F]+)?([pP][0-9]+)[fF]?/; # hexadecimal constant
_ => tk == cls;
}
);

let std-c-can-take(tokens: List<Token>, cls: String): U64 = (
non-zero(tokens) && std-c-has-class(head(tokens).key, cls)
non-zero(tokens) && std-c-has-class(head(tokens).skey, cls)
);

let std-c-take-expect(tokens: List<Token>, cls: String): List<Token> = (
if non-zero(tokens) && std-c-has-class(head(tokens).key, cls) then tail(tokens)
if non-zero(tokens) && std-c-has-class(head(tokens).skey, cls) then tail(tokens)
else (print("Expected token [\{cls}] at \{tokens.formatted-location}\n"); exit(1); tokens);
);

let std-c-take-maybe(tokens: List<Token>, cls: String): List<Token> = (
if non-zero(tokens) && std-c-has-class(head(tokens).key, cls) then tail(tokens)
if non-zero(tokens) && std-c-has-class(head(tokens).skey, cls) then tail(tokens)
else tokens;
);

Expand Down Expand Up @@ -435,54 +438,54 @@ let std-c-take-maybe(tokens: List<Token>, cls: String): List<Token> = (
#struct-declarator = ':', constant-expression
# | declarator, [':', constant-expression];

let std-c-parse-assignment-operator(tokens: List<Token>): Tuple<Maybe<CString>,List<Token>> = (
let no = None :: Maybe<CString>;
let std-c-parse-assignment-operator(tokens: List<Token>): Tuple<Maybe<String>,List<Token>> = (
let no = None :: Maybe<String>;
if not(non-zero(tokens)) then Tuple{ no, tokens }
else if head(tokens).key == c"=" then Tuple{ Some{c"="}, tail(tokens) }
else if head(tokens).key == c"*=" then Tuple{ Some{c"*="}, tail(tokens) }
else if head(tokens).key == c"/=" then Tuple{ Some{c"/="}, tail(tokens) }
else if head(tokens).key == c"%=" then Tuple{ Some{c"%="}, tail(tokens) }
else if head(tokens).key == c"+=" then Tuple{ Some{c"+="}, tail(tokens) }
else if head(tokens).key == c"-=" then Tuple{ Some{c"-="}, tail(tokens) }
else if head(tokens).key == c"<<=" then Tuple{ Some{c"<<="}, tail(tokens) }
else if head(tokens).key == c">>=" then Tuple{ Some{c">>="}, tail(tokens) }
else if head(tokens).key == c"&=" then Tuple{ Some{c"&="}, tail(tokens) }
else if head(tokens).key == c"^=" then Tuple{ Some{c"^="}, tail(tokens) }
else if head(tokens).key == c"|=" then Tuple{ Some{c"|="}, tail(tokens) }
else if head(tokens).skey == "=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "*=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "/=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "%=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "+=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "-=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "<<=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == ">>=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "&=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "^=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "|=" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else Tuple{ no, tokens }
);

let std-c-parse-unary-operator(tokens: List<Token>): Tuple<Maybe<CString>,List<Token>> = (
let no = None :: Maybe<CString>;
let std-c-parse-unary-operator(tokens: List<Token>): Tuple<Maybe<String>,List<Token>> = (
let no = None :: Maybe<String>;
if not(non-zero(tokens)) then Tuple{ no, tokens }
else if head(tokens).key == c"&" then Tuple{ Some{c"&"}, tail(tokens) }
else if head(tokens).key == c"*" then Tuple{ Some{c"*"}, tail(tokens) }
else if head(tokens).key == c"+" then Tuple{ Some{c"+"}, tail(tokens) }
else if head(tokens).key == c"-" then Tuple{ Some{c"-"}, tail(tokens) }
else if head(tokens).key == c"~" then Tuple{ Some{c"~"}, tail(tokens) }
else if head(tokens).key == c"!" then Tuple{ Some{c"!"}, tail(tokens) }
else if head(tokens).skey == "&" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "*" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "+" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "-" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "~" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "!" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else Tuple{ no, tokens }
);

let std-c-parse-struct-or-union(tokens: List<Token>): Tuple<Maybe<CString>,List<Token>> = (
let no = None :: Maybe<CString>;
let std-c-parse-struct-or-union(tokens: List<Token>): Tuple<Maybe<String>,List<Token>> = (
let no = None :: Maybe<String>;
if not(non-zero(tokens)) then Tuple{ no, tokens }
else if head(tokens).key == c"struct" then Tuple{ Some{c"struct"}, tail(tokens) }
else if head(tokens).key == c"union" then Tuple{ Some{c"union"}, tail(tokens) }
else if head(tokens).skey == "struct" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else if head(tokens).skey == "union" then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else Tuple{ no, tokens }
);

let std-c-parse-identifier(tokens: List<Token>): Tuple<Maybe<CString>,List<Token>> = (
let no = None :: Maybe<CString>;
if std-c-can-take(tokens, "identifier") then Tuple{ Some{head(tokens).key}, tail(tokens) }
let std-c-parse-identifier(tokens: List<Token>): Tuple<Maybe<String>,List<Token>> = (
let no = None :: Maybe<String>;
if std-c-can-take(tokens, "identifier") then Tuple{ Some{head(tokens).skey}, tail(tokens) }
else Tuple{ no, tokens }
);

let std-c-parse-constant(tokens: List<Token>): Tuple<Maybe<CConstant>,List<Token>> = (
let no = None :: Maybe<CConstant>;
if std-c-can-take(tokens, "integer") then Tuple{ Some{CConstantInteger{head(tokens).key}}, tail(tokens) }
else if std-c-can-take(tokens, "character") then Tuple{ Some{CConstantCharacter{head(tokens).key}}, tail(tokens) }
#else if std-c-can-take(tokens, "floating") then Tuple{ Some{CConstantFloating{head(tokens).key}}, tail(tokens) }
if std-c-can-take(tokens, "integer") then Tuple{ Some{CConstantInteger{head(tokens).skey}}, tail(tokens) }
else if std-c-can-take(tokens, "character") then Tuple{ Some{CConstantCharacter{head(tokens).skey}}, tail(tokens) }
else if std-c-can-take(tokens, "floating") then Tuple{ Some{CConstantFloating{head(tokens).skey}}, tail(tokens) }
#else if std-c-can-take(tokens, "enumeration") then Tuple{ Some{CConstantFloating{head(tokens).key}}, tail(tokens) }
else Tuple{ no, tokens }
);
Expand Down
17 changes: 15 additions & 2 deletions PLUGINS/FRONTEND/C/c-smart-tokenize.lsts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,8 +56,17 @@ let std-c-tokenize-string(file-path: String, text: String): List<Token> = (
".".. rest => (tokens = cons(text[:".".length], tokens); text = rest;);
"-".. rest => (tokens = cons(text[:"-".length], tokens); text = rest;);

(lit=r/^["]([^"\\]|([\\].))*["]/).. rest => (
tokens = cons(text[:lit.length], tokens); text = rest;
(cl=r/^[0][x][0-9a-fA-F]+([.][0-9a-fA-F]+)?([eE][0-9a-fA-F]+)?([pP][0-9]+)(fF)?/).. rest => (
tokens = cons(text[:cl.length], tokens); text = rest;
);
(cl=r/^[0][x][0-9a-fA-F]+([uU]|[lL]|wb|WB)*/).. rest => (
tokens = cons(text[:cl.length], tokens); text = rest;
);
(cl=r/^[0][bB][01]+([uU]|[lL]|wb|WB)*/).. rest => (
tokens = cons(text[:cl.length], tokens); text = rest;
);
(cl=r/^[0-9]+([uU]|[lL]|wb|WB)*([.][0-9]+)?([eE][0-9]+)?[fF]?/).. rest => (
tokens = cons(text[:cl.length], tokens); text = rest;
);
(cl=r/^(u8|u|U|L)?[']([^']|([\\][']))+[']/).. rest => (
tokens = cons(text[:cl.length], tokens); text = rest;
Expand All @@ -67,6 +76,10 @@ let std-c-tokenize-string(file-path: String, text: String): List<Token> = (
tokens = cons(text[:id.length], tokens); text = rest;
);

(lit=r/^["]([^"\\]|([\\].))*["]/).. rest => (
tokens = cons(text[:lit.length], tokens); text = rest;
);

(comment=r/^#[^\n]*[\n]/).. rest => (
text = rest;
);
Expand Down
Loading