{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreibet6zhu7ada52xekzraihg6rsajicvvibo7anu7m5dnofu4pgkbm",
    "uri": "at://did:plc:pi6woz4d47bkuws673w2il2r/app.bsky.feed.post/3mgm3mio7wuo2"
  },
  "path": "/t/announcing-scrod-like-haddock-but-faster/13761#post_18",
  "publishedAt": "2026-03-08T23:15:14.000Z",
  "site": "https://discourse.haskell.org",
  "tags": [
    "graphex"
  ],
  "textContent": "What Haddock needs is more involved, but - anecdotally - I first tried to write graphex’s parser using the GHC parser. But that was a pain in the butt, so I just wrote something for what graphex needs (imports) using megaparsec in a quick sitting.\n\nI definitely am glad I went that way. I didn’t consider it, but not failing when irrelevant stuff is borked is a nice side effect.",
  "title": "Announcing Scrod, like Haddock but faster"
}