C Implementation of 'Writing an interpreter in GO - Thorsten Ball'
parse identifier -> let statement -> sla identifier op in OBJ_VAL(copyString(token->start,token ->lenght)).lees chpater 17 en 18 van crafting interpreters en lees ook over type punning in c https://en.wikipedia.org/wiki/Type_punning https://www.youtube.com/watch?v=TsUOhPsZk6k
- read about type pruning:
- do like c and create a Value struct
- create also an Obj struct
Store identifiers and strings in Obj see how to do typreuning also the numbers and operators we'll store here
source :: let 2 = 4; parser :: Token: Type = TOKEN_LET Token: Type = TOKEN_INT Token: Type = TOKEN_ASSIGN Token: Type = TOKEN_INT Correct that 2 is parsed as an INT when it is in the spot of an IDENTIEFER?
Token value names get put put in the top namespace in clang. This caused an issue where TokenType EOF conflicted with EOF in <stdio.h>. The solution is prepending all the enum values, so they don't conflict with existing keywords in CLANG.
Go works with string slices to get a shallow copy to the source String(points to the original string). In go strings are immutable so the source String will never change. In Chimp, we point to the original source stream, this means that we must make sure that source string outlives the token literal and that the source string doesn't get changed as long as the token hasn't been processed!
The way we look for Keywords is different from the go version. In go we use a map and in ChimpanC we use a simple switch statement (minimalistic TRIE) TRIE wiki. I got the idea from Clox. Ch 16.4 Identifiers and keywords.
Struggeling a bit with implementing the pasrse. I think the issue is that I don't know in which direction the code is going and I'm staying to close to the Golang gode in the book. After chapter 2.6 - the first part of the PRATT parser, I'll do some heavy refactoring.