Jonathan Dowland

Hi @snowleopard findVertices p = foldg [] [ x | p x ] (++) (++) I think there must be a typo in the list comprehension here, my GHCI doesn't...

NP. I've just pushed that, let's see what the performance is like. I'd quite like to reproduce the performance measurement stuff locally (to avoid round-tripping github PRs to test things)....

As an aside, would you like PRs to add things like `{-# LANGUAGE TupleSections #-}` where necessary — I see this is already specified in the cabal file, but is...

[travis](https://travis-ci.org/snowleopard/alga/jobs/449782302) and [appveyor](https://ci.appveyor.com/project/snowleopard/alga/builds/19999443) results for the `findVertices` change only (`ab78e059`) show a regression > hasVertex: 1.55 (bad) I'm going to push an initial `findEdges` now to see how that performs....

Results for `findEdges` as of `5ed33b63284afb275d7584d0203d0b4327ecc02e` > hasEdge: 1.35 (bad) > hasVertex: 2.03 (bad)

@snowleopard wrote: > Also, I'd like to understand your use case a bit better: how do you plan to use these functions? Merely checking if a vertex with the right...

Thanks @snowleopard. I'm going to see what PoC I can build on top of these primitives to further assess how this all might work for me.

Things *I* know need doing (at least) - [ ] tests - [x] GHC warns about redundant constraints for some functions, to investigate - [ ] docs - [x] do...

Simple test program ``` import Algebra.Graph import Algebra.Graph.Export.Dot main = do putStrLn test putStrLn $ show $ test == correct putStrLn (export defaultStyleViaShow (path [1, 2]) :: String) test =...

> As for generic export, I don't know if we want to give the user full control and escaping themselves, or shall we provide built-in escaping? I suppose the question...