Precedence isn't hard -- but it's often not done with recursive descent. In other words, it's not just "typing out the code".
I think you're misreading my reply. A hand-written recursive descent works a lot of the time, and is almost certainly the right "default". But my point is that it doesn't work all the time. If it works for you, great! That doesn't mean all parsing is easy. Try adding an autoformatter like gofmt or IDE completion for Jai, and see how your parser changes (or if you have to write an entirely new parser).
In particular, I'm NOT advocating that all parsers should written with parsing tools. I think you must have read that somewhere else.
In fact, I used EXACTLY the same strategy as you mention on my recent shell parser. I machine-checked the grammar with ANTLR, and wrote some test cases for it. Then I transcribed that grammar to recursive descent by hand, along with several other ad hoc strategies that you need to parse an ad hoc language like shell.
(ANTLR is a great front end, but not necessarily a good back end. The generated Java code is probably OK, but the C++ code it generates is preposterous. Yacc is not that much better, i.e. with globals all over the place.)
"Just type out the code" is a horrible strategy because you will end up not knowing what language you've designed. If your post advocated "prototype your grammar with a tool, and then manually transcribe the grammar to code", then I'd be inclined to agree.
But that's not what you said. It seems like you are more interested in telling everyone what a great programmer you are, and how things are so easy for you, rather than spreading any interesting knowledge about parsing.
"Try adding an autoformatter like gofmt or IDE completion for Jai, and see how your parser changes (or if you have to write an entirely new parser)."
It would not change at all, and I have no idea why you think it would, except to guess that the model you have in your head of a hand-written parser kind of sucks. They don't have to suck.
"...not knowing what language you've designed." I have no idea what you're on about here either.
Look, I think you are making things a lot harder than they are. I am not bragging ... I used to build lexers and parsers by hand 23+ years ago when I was a student in college and had almost no programming experience compared to what I have now. It is not hard. If you think it's hard, something is missing in your knowledge set.
(I also built stuff using parser tools 23+ years ago, and was able to very clearly contrast the two methods. Parser tools have gotten slightly better since then, but not much.)
I doubt your parser is even close to doing what he is talking about. I think you're an advocate of Microsoft's debugger over GDB, so hopefully what they are saying carries some weight (i.e. a little computer science doesn't hurt). They aren't just a bunch of eggheads; they are actually building language tools that people use.
And again, I'm not claiming that ALL parsing is hard. Sometimes you can just write a hand written lexer and recursive descent parser and be done with it. But that's not true in all cases, particularly for "real world" languages, or when you are designing a brand new language.
Saying parsing is easy is like saying "I wrote a game in BASIC in 8th grade, so writing games is easy". (And yes I've actually heard that statements like that ... )
I was researching Jai awhile back, and was definitely curious about it. I don't think you have released the code, so I can't tell which of these statements is true:
1) Parsing is easy for you
2) You think parsing is easy, but your language and your parser are full of bugs
I'm perfectly willing to believe #1 (honestly). But that doesn't mean all parsing is easy, for say a competent programmer in some other domain. It doesn't help anyone to say "just type out the code". Your comment about how you do precedence is a little more helpful.
I watched that talk when it appeared on the HN front page, and I actually think the whole methodology he is talking about is misguided. I don't find any of the "incremental program understanding" stuff in Visual Studio to be useful at all. I wish it were not there because it only causes problems and distractions.
It's a case where some people are choosing to do something that is a lot harder than a straightforward parse ... but as a user, a straightforward parse is actually what I want.
That said, even if you thought this was the right way to go,
I am not sure that the internals of their code would look anything like the kinds of parsing tools you are talking about, so I am not sure it supports your point in any way.
> And again, I'm not claiming that ALL parsing is hard.
Parsing is easy. The video you link above is harder, but that's not really parsing any more, it's more like "make sense of this text that is sort of like a working program", which is more like an AI problem.
But anyway. It's pretty clear you haven't written many parsers (or any) so I am going to stop arguing. If I were to "win" this argument I wouldn't get anything out of it. I am trying to help by disavowing people of the notion that certain things are harder than they've been indoctrinated to think. If you don't want that help, fine ... just keep doing what you do and the world will keep going onward.
> I doubt your parser is even close to doing what he is talking about.
You act like if someone hasn't architecture-astronauted all their software to meet any hypothetical future requirement, you've won. The truth is, making a parser that can restart incrementally and reuse above-and-below data structures is just another requirement. It shapes the software design and would probably result in a more iterator-oriented approach -- see, we can already imagine how it might be built. It's like you're saying, "Writing a piddly game in BASIC isn't all that, let me show you how hard it is to write this Tetris program." Maybe a better analogy is those people that thought OOP was some deep advanced stuff.
And note again parsing is the easiest part of that software, a neat and well-contained problem.
I think you're misreading my reply. A hand-written recursive descent works a lot of the time, and is almost certainly the right "default". But my point is that it doesn't work all the time. If it works for you, great! That doesn't mean all parsing is easy. Try adding an autoformatter like gofmt or IDE completion for Jai, and see how your parser changes (or if you have to write an entirely new parser).
In particular, I'm NOT advocating that all parsers should written with parsing tools. I think you must have read that somewhere else.
In fact, I used EXACTLY the same strategy as you mention on my recent shell parser. I machine-checked the grammar with ANTLR, and wrote some test cases for it. Then I transcribed that grammar to recursive descent by hand, along with several other ad hoc strategies that you need to parse an ad hoc language like shell.
(ANTLR is a great front end, but not necessarily a good back end. The generated Java code is probably OK, but the C++ code it generates is preposterous. Yacc is not that much better, i.e. with globals all over the place.)
"Just type out the code" is a horrible strategy because you will end up not knowing what language you've designed. If your post advocated "prototype your grammar with a tool, and then manually transcribe the grammar to code", then I'd be inclined to agree.
But that's not what you said. It seems like you are more interested in telling everyone what a great programmer you are, and how things are so easy for you, rather than spreading any interesting knowledge about parsing.