From 89a4fc4335fe433af6758db84990a08ae5a8e3b2 Mon Sep 17 00:00:00 2001
From: Nora Abi Akar <nora.abiakar@gmail.com>
Date: Thu, 28 Nov 2019 13:17:25 +0100
Subject: [PATCH] Modcc: Add support for `COMMENT` and `ENDCOMMENT` (#906)
Also change the way we handle unitson/unitsoff.
UNITSON, UNITSOFF, COMMENT, ENDCOMMENT are now all handled in Lexer::parse
Fixes #14
Fixes #885
---
modcc/lexer.cpp | 23 ++++++++++++++++-------
modcc/parser.cpp | 19 -------------------
modcc/parser.hpp | 1 -
3 files changed, 16 insertions(+), 27 deletions(-)
diff --git a/modcc/lexer.cpp b/modcc/lexer.cpp
index dc5851ec..086a5a80 100644
--- a/modcc/lexer.cpp
+++ b/modcc/lexer.cpp
@@ -20,7 +20,7 @@ inline bool is_alphanumeric(char c) {
return (is_numeric(c) || is_alpha(c) );
}
inline bool is_whitespace(char c) {
- return (c==' ' || c=='\t' || c=='\v' || c=='\f');
+ return (c==' ' || c=='\t' || c=='\v' || c=='\f' || c=='\n' || c=='\r');
}
inline bool is_eof(char c) {
return (c==0);
@@ -97,14 +97,23 @@ Token Lexer::parse() {
// identifier or keyword
case 'a' ... 'z':
case 'A' ... 'Z':
- case '_':
+ case '_': {
// get std::string of the identifier
- t.spelling = identifier();
- t.type
- = status_==lexerStatus::error
- ? tok::reserved
- : get_identifier_type(t.spelling);
+ auto id = identifier();
+ if (id == "UNITSON" || id == "UNITSOFF") continue;
+ if (id == "COMMENT") {
+ while (!is_eof(*current_)) {
+ while (is_whitespace(*current_) || !is_alpha(*current_)) current_++;
+ if (identifier() == "ENDCOMMENT") break;
+ }
+ continue;
+ }
+ t.spelling = id;
+ t.type = status_ == lexerStatus::error
+ ? tok::reserved
+ : get_identifier_type(t.spelling);
return t;
+ }
case '(':
t.type = tok::lparen;
t.spelling += character();
diff --git a/modcc/parser.cpp b/modcc/parser.cpp
index 3cacb0f1..efeb1cac 100644
--- a/modcc/parser.cpp
+++ b/modcc/parser.cpp
@@ -35,15 +35,6 @@ bool Parser::expect(tok tok, std::string const& str) {
return false;
}
-void Parser::parse_unit() {
- if(token_.type == tok::lparen) {
- while (token_.type != tok::rparen) {
- get_token();
- }
- get_token(); // consume ')'
- }
-}
-
void Parser::error(std::string msg) {
std::string location_info = pprintf(
"%:% ", module_ ? module_->source_name() : "", token_.location);
@@ -138,12 +129,6 @@ bool Parser::parse() {
module_->add_callable(std::move(f));
}
break;
- case tok::unitson :
- get_token();
- break;
- case tok::unitsoff :
- get_token();
- break;
default :
error(pprintf("expected block type, found '%'", token_.spelling));
break;
@@ -835,8 +820,6 @@ expression_ptr Parser::parse_prototype(std::string name=std::string()) {
get_token(); // consume the identifier
- parse_unit(); // consume the unit if provided
-
// look for a comma
if(!(token_.type == tok::comma || token_.type==tok::rparen)) {
error( "expected a comma or closing parenthesis, found '"
@@ -967,8 +950,6 @@ symbol_ptr Parser::parse_function() {
auto p = parse_prototype();
if(p==nullptr) return nullptr;
- parse_unit();
-
// check for opening left brace {
if(!expect(tok::lbrace)) return nullptr;
diff --git a/modcc/parser.hpp b/modcc/parser.hpp
index 09130e9a..49512f7e 100644
--- a/modcc/parser.hpp
+++ b/modcc/parser.hpp
@@ -81,7 +81,6 @@ private:
Parser();
Parser(Parser const &);
- void parse_unit();
bool expect(tok, const char *str="");
bool expect(tok, std::string const& str);
};
--
GitLab