Not yet. Very close though.

Have a nice day everyone!

I took a day off today and we wont have the kids its awesome.

so what you're saying is that you're finally going to sleep a useful amount?

@IvenBach I look forward to it!

[Zomis/Games/master] build 180 Build success.

[Zomis/Games] build for commit bef4ee92 on master: This commit looks good

@Marc-Andre Sounds relaxing :) What have you planned for the day?

@Vogel612 Maybe ! The morning will be quieter that's for sure. But the weird things is since it's so unusual normally I have more troubled sleeping

@SimonForsberg My wife have a dentist appointment, we have to go to costco then do the groceries and latter she's getting a tatoo. After well be really free

But right bow im playing the switch in the car waiting

@Marc-Andre Cool! What will be the tattoo?

@M.Doerner I've seen the various *Context objects within RD. For what reason do you include the Indices?

What is the difference between TokenStream and GetText? My naive understanding thinks they would be the same.

^ Better if I rubberduck it out since I know they shouldn't be.

( 5 + 2 * 7) can be a `ParenthesizedExpressionContext. The text is what is shown. The tokens should be the contexts they represent?

2 * 7 converted to a MultiplicationExpressionContext would have Left = 2, Right - 7.

Rewriting with the new token ( 5 + MultiplicationToken ) which can be rewritten.

Nope... being too hasty.

2 * 7 needs to be composed first.

IntegerExpressionContext.Value = 2 and IntegerExpressionContext.Value = 7 for the respective left and right operands.

From those, create the MultiplicationExpressionContext with Left/Right expression containing the IntegerExpressionContext.

Refactor attempt on my code to reflect this.

@IvenBach For example to show in the editor where the code is, so that you can for example show a warning, hint, or error

Working towards the ability to understand that.

@SimonForsberg Splunk Alert: MFE Long SQL Query - [max(runtime):162801, avg(runtime):162801, min(runtime):162801]

The TokenStream and the indices are not necessary for the problem. They are just there to be able to reproduce the matched text and in case you want to do some operations based on the context.

Let me explain what the TokenStream is. A typical way to parse code is to first turn it in a lexer into a stream of tokens, i.e. atomic items for parsing.

hey.late

E.g. if you have the formula (111 + 13) * 22 would be turned into tokens LPAREN, 111, PLUS, 13, RPAREN, MULT, 22. That simplifies the parsing since unnecessary things like whitespace are already gone and things like number literals area already condensed into a single token.

Your explanation is good. The idea I can understand. Translating from idea into actual code is my failing right now.

My updated result is a step towards progress.

It feels more robotically done than with comprehension.

[IvenBach/AdventOfCode2020] IvenBach pushed commit 8bfbe0e7 to main

Maybe I can give some inspiration too.. if I can find the project

TokenStream is a contiguous sequence of tokens? Each token is an expression represented as code?

github.com/skiwi2/BrainfuckCompiler

The BrainfuckCompiler project actually looks pretty damn complex..

A lexer is used to define the smallest (atomic) parts that make up a token. From these lexical lego-like tokens more complex constructs can be made.

That. E.g. in Rubberduck we identify all kinds of literals and identifiers already in the lexer in order not to deal with those in the actual parser.

VBALexer.g4 defines the smallest atomic parts of the language.

@skiwi It's your code, what did you expect?

VBAParser.g4 takes those atomic pieces and builds larger constructs with them.

@IvenBach Most stuff makes sense.. then there's DEFLNGLNG

Well, those are the definitions for the generator.

Antlr uses them to generate the actual lexer and parser.

@SimonForsberg Well.. I don't know

@IvenBach That file is pretty complex

@skiwi Deftype statement for a longlong: docs.microsoft.com/en-us/office/vba/Language/Concepts/…

@M.Doerner Is the lexer and parser different from the files within which the rules are defined?

Btw, if you want to look at the parser I used for AoC day 18, github.com/MDoerner/AdventOfCode2020/blob/main/TypeScript/src/….

The generated files are the one with the extension.cs.

Since they are generated, they do not look too straightforward.

BTW, we exclude the generated lexer and parser from git because they are just a derivative. So, you have to look after a build in order to see the generated lexer and parser.