Newest at the top
| 2025-02-10 17:41:29 +0100 | <bwe> | dminuoso: yeah, that's actually what lookahead is doing :). Thanks for pointing me at that. |
| 2025-02-10 17:38:32 +0100 | jespada | (~jespada@2800:a4:224d:2b00:41f1:b3c7:471b:62da) jespada |
| 2025-02-10 17:38:15 +0100 | <haskellbridge> | <Bowuigi> tomsmeding re:VariantF oh, is it useful for anything else? I can cover that usecase with other features. I think it was used in an effect system at some point |
| 2025-02-10 17:37:24 +0100 | <dminuoso> | bwe: If all you want is just parse without consuming, use lookahead. |
| 2025-02-10 17:37:04 +0100 | <bwe> | dminuoso: so, swapping out the underlying buffer to the beginning, that means saving the buffer before parsing, then accessing the saved value and reset it to that when I want it, right? |
| 2025-02-10 17:35:52 +0100 | <dminuoso> | do { a <- lookahead parser1; b <- parser2 ... } |
| 2025-02-10 17:35:27 +0100 | <dminuoso> | Yeah that would work for you. |
| 2025-02-10 17:35:21 +0100 | <dminuoso> | Or I guess you could use `lookahead` |
| 2025-02-10 17:35:14 +0100 | <dminuoso> | bwe: Like I said, you need something like `refocus` that I mentioned above. |
| 2025-02-10 17:34:35 +0100 | <bwe> | dminuoso: https://play.haskell.org/saved/HGeKdnSW |
| 2025-02-10 17:34:07 +0100 | pavonia | (~user@user/siracusa) (Quit: Bye!) |
| 2025-02-10 17:33:55 +0100 | jespada | (~jespada@2800:a4:2243:2100:5cc9:2329:b53c:b25f) (Ping timeout: 252 seconds) |
| 2025-02-10 17:31:43 +0100 | <dminuoso> | What do you mean with "again"? |
| 2025-02-10 17:31:26 +0100 | <bwe> | dminuoso: I basically want to parse `s` again with another parser in a bigger parser function. |
| 2025-02-10 17:30:47 +0100 | <dminuoso> | Not sure what the question means |
| 2025-02-10 17:30:42 +0100 | <dminuoso> | bwe: Well you dont if it succeeds.. |
| 2025-02-10 17:30:26 +0100 | <bwe> | dminuoso: How do I backtrack with Megaparsec if one parser just succeeded? |
| 2025-02-10 17:25:05 +0100 | <dminuoso> | euouae: Start by writing an instance Functor for that. Once done, give me a ping. |
| 2025-02-10 17:24:27 +0100 | <dminuoso> | euouae: So imagine you have a simple 3-dimensional vector type: https://paste.tomsmeding.com/RshSBpbw |
| 2025-02-10 17:23:16 +0100 | <bwe> | Actually you are right: I am solving the wrong problem. I should better introduce a dto intermediary and from that construct a unified result based on the results. |
| 2025-02-10 17:23:08 +0100 | <dminuoso> | But even without ambient parser state, you could do this. |
| 2025-02-10 17:22:22 +0100 | <dminuoso> | But it feels awkward and might not make sense if you some outer ambient parser state |
| 2025-02-10 17:21:53 +0100 | <dminuoso> | Now you could write this for ParsecT too, that way you could do something like `let (b1, b2) = tup in try (refocus b1 $ ...) <|> try (refocus b2 $ ...) |
| 2025-02-10 17:20:59 +0100 | <dminuoso> | This just swaps out the underlying buffer, runs the parser and then stitches the continuations together |
| 2025-02-10 17:20:30 +0100 | <dminuoso> | So what I have is something like `refocus :: ByteString -> Get a -> Get a` where `Get` is just some type alias over a fancy parser type. |
| 2025-02-10 17:19:45 +0100 | <dminuoso> | Lets ignore the fact that this is a very XY problem. |
| 2025-02-10 17:18:57 +0100 | <bwe> | What I've got is that `Parser a` abstracts away the input that is `Text`. Therefore I conclude that I ought define another `ParserTuple a b` that abstracts away a tuple? |
| 2025-02-10 17:17:32 +0100 | <dminuoso> | Absolutely, you can do what I do with megaparsec too. |
| 2025-02-10 17:17:19 +0100 | <bwe> | …is this generally possible with Megaparsec? |
| 2025-02-10 17:15:15 +0100 | <dminuoso> | There's this combinator I mentioned earlier that I have.. |
| 2025-02-10 17:15:06 +0100 | <dminuoso> | bwe: So I know how this could be done with flatparse... |
| 2025-02-10 17:14:54 +0100 | jespada | (~jespada@2800:a4:2243:2100:5cc9:2329:b53c:b25f) jespada |
| 2025-02-10 17:13:27 +0100 | <bwe> | Tabular data. And at times stuff belonging into the second cell is being found in the first. |
| 2025-02-10 17:12:37 +0100 | <dminuoso> | bwe: What does this tuple encode, exactly? |
| 2025-02-10 17:12:13 +0100 | <bwe> | …and of course, I am referring to Megaparsec |
| 2025-02-10 17:11:58 +0100 | <bwe> | for `Text` I define `Parser MyDataType`. But how do I parse for example a tuple `(Text, Text)`? I feel defining a function `:: (Text, Text) -> Maybe MyDataType` is not right because I can't combine it with other parsers easily. What's your take on it? |
| 2025-02-10 17:11:40 +0100 | acidjnk_new3 | (~acidjnk@p200300d6e7283f99c4dbaee3a15423f1.dip0.t-ipconnect.de) (Ping timeout: 272 seconds) |
| 2025-02-10 17:11:34 +0100 | alfiee | (~alfiee@user/alfiee) (Ping timeout: 252 seconds) |
| 2025-02-10 17:11:33 +0100 | lortabac | (~lortabac@2a01:e0a:541:b8f0:55ab:e185:7f81:54a4) (Quit: WeeChat 4.4.2) |
| 2025-02-10 17:09:23 +0100 | <dminuoso> | Otherwise you deprive yourself of the wonders of coming up with this yhourself. |
| 2025-02-10 17:09:16 +0100 | <euouae> | learn fromsome real code |
| 2025-02-10 17:09:13 +0100 | <dminuoso> | Perhaps try and derive this yohurself first. |
| 2025-02-10 17:09:03 +0100 | <euouae> | i thought maybe to look into the microlens implementation too |
| 2025-02-10 17:07:21 +0100 | alfiee | (~alfiee@user/alfiee) alfiee |
| 2025-02-10 17:07:02 +0100 | <dminuoso> | taught me lenses with. |
| 2025-02-10 17:06:53 +0100 | <dminuoso> | (In fact, these are the very excercises that someone tought me with) |
| 2025-02-10 17:06:45 +0100 | <euouae> | sure, i will look together with the book |
| 2025-02-10 17:06:27 +0100 | <dminuoso> | If you are interested. |
| 2025-02-10 17:06:06 +0100 | <dminuoso> | euouae: You can derive this yourself. I have a set of simple excercises where you could come up with how they work on your own. |
| 2025-02-10 17:05:22 +0100 | <dminuoso> | That is, in optics you dont ever see any profunctors or some strange missing constraint when misusing the library |