Hi, I have taken a run at adding some more functionality to `evancz/url-parser` <https://github.com/evancz/url-parser>. The repo is *here <https://github.com/enetsee/elm-url-parser>.*
(Note: I actually wrote this in F# originally and have only tested it in that environment. I will look to add tests and a working example in due course.) Whilst the approach works for the most part, I have a few problems with the design. From the README: # elm-url-parser This library is an attempt to add more functionality to `evancz/url-parser`. Specifically, the library has combinators to deal with query parameters and hash fragments: - (</>) : composes two parsers with the second being applied to `PathPart`s only (see below) - (<#>) : composes two parsers with the second being applied to `Hash`s only (see below) - (<?>) : composes two parsers with the second being applied to `Query`s only (see below) In addition, there are a couple other combinators: - `many` : matches zero or more accepts URL chunks which match the supplied parser - `param` : matches a query parameter with the given key and a value matching the supplied parser - `optParam`: matches a query parameter with the given key and a value matching the supplied parser which may or may not be present in the `location` `search` (I'm using a `Dict String String` to hold the key-value pairs in the `location` `search` so you don't need to worry about the ordering of the query parameters when using the `param` and `optParam` combinators.) ## Approach I have introduced an ADT (`UrlPart`) to differentiate between different parts of a URL. There are four data constructors: - PathPart : an element of the `location` `pathname` - Query : a dictionary containing the key-value pairs in the `location` `search` - Hash : the `location` `hash` - Param : an individual key-value pair in the `location` `search` Only the first three of these are used for scanning a URL; `Param` is pushed into the `seen` list when it is parsed from the `Query` URL part. ## Problems The types are *very* confusing. I have introduced a type alias `UrlPartExtractor a` (for `UrlPart -> Result String a`) which is used to determine which part of URL a parser should be applied to. It feels like this could be moved into the definition of the `Parser` data type but I cannot figure out how. Secondly, I am totally lost with the function `formatPart` and `formatPath`. In @evancz's library there is a single `format` function but I had to twiddle the types until I got the result I wanted. If the types can be lined up, I suspect these functions are unnecessary. I would appreciate any thoughts or pointers on the additions. I would appreciate thoughts and suggestions on the design (especially if I am way off track with the approach!). Thanks, Michael -- You received this message because you are subscribed to the Google Groups "Elm Discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to elm-discuss+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
