psb/tests/Test/Harness.hs
Pagwin 0f356e4d5e
property tests for the djot parser
Expectations come from the djot syntax reference and IR.hs rather than
from what the parser currently produces, so a failure is evidence of a
bug instead of a record of one. 55 properties: 32 inline (emphasis,
verbatim, math, raw, links, images, references, autolinks, spans,
attributes, footnotes, symbols, escapes, smart punctuation) and 23 block
(headings, quotes, breaks, code fences, divs, raw blocks, every ordered
marker style and delimiter, tight vs loose items, nesting, task lists,
definition lists, tables, definitions, block attributes, round trip).

Two rules are deliberately unasserted and noted in the module header:
headings of seven or more hashes, and the general verbatim space
stripping rule, of which only the unambiguous backtick case is tested.

Three deviations the suite found, fixed in Djot.hs:

- an unclosed verbatim span failed the construct back into plain text.
  The reference says it "extends to the end of the text"; manyTill
  cannot say that on its own, it needs the eof alternative.
- a task list checkbox was recognised without the space the reference
  requires after it, so "1. [x](5)" ate a link and turned it into a
  checked item containing a list starting at 5.
- checkboxes were recognised on ordered items, but the reference gives
  them to bullet items only.

Test.Gen.Document is now shared by both suites. Syntax grew
quoteContent and listTypeFor for the two places markdown and djot
disagree about the resulting IR, and two djot facts became generator
invariants rather than assertions: adjacent markup sharing a delimiter
is ambiguous, so runs separate it, and a blank line between two lists
makes one loose list rather than two blocks, so documents never place
lists back to back.

The test suite is no longer markdown only, so it is just "tests" now.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-09-04 17:21:57 -04:00

98 lines
3.2 KiB
Haskell

{-# LANGUAGE OverloadedStrings #-}
-- | Running a document parser inside a property and turning the result into a
-- QuickCheck verdict. Everything here is parser agnostic so the Djot suite can
-- share it with the Markdown one.
module Test.Harness
( DocumentParser,
markdownDocument,
djotDocument,
shouldParseTo,
shouldParse,
succeed,
failWith,
)
where
import Control.Exception (evaluate)
import Data.Functor.Identity (Identity)
import Data.Text (Text)
import Data.Void (Void)
import qualified Djot
import IR (Document)
import qualified Markdown
import System.Timeout (timeout)
import Test.QuickCheck (counterexample)
import Test.QuickCheck.Monadic (PropertyM, run, stop)
import Text.Megaparsec (ParsecT, errorBundlePretty, parse)
-- | A document parser pinned to the concrete stream and monad the tests use.
type DocumentParser = ParsecT Void Text Identity Document
markdownDocument :: DocumentParser
markdownDocument = Markdown.document
djotDocument :: DocumentParser
djotDocument = Djot.document
-- | What running a parser over an input told us. Timeouts are a failure mode in
-- their own right because a parser bug is much more likely to loop than to
-- throw.
data Outcome
= TimedOut
| Failed String
| Passed
-- | 1 second, every test here should finish orders of magnitude below that
parseTimeout :: Int
parseTimeout = 1000000
-- | Both the parse and the check on its result happen inside the timeout, a
-- comparison against a lazily built infinite tree would otherwise hang outside
-- of it.
attempt :: DocumentParser -> Text -> (Document -> Maybe String) -> PropertyM IO Outcome
attempt parser input check = do
result <- run $
timeout parseTimeout $ do
parsed <- evaluate $ parse parser "test_input" input
case parsed of
Left e -> pure . Failed $ errorBundlePretty e
Right actual -> case check actual of
Nothing -> pure Passed
Just message -> do
-- force the message, building it is what forces the tree
_ <- evaluate $ length message
pure $ Failed message
pure $ maybe TimedOut id result
verdict :: Outcome -> PropertyM IO ()
verdict TimedOut = failWith "Hit Timeout"
verdict (Failed message) = failWith message
verdict Passed = succeed
-- | The input must parse to exactly this tree.
shouldParseTo :: DocumentParser -> Text -> Document -> PropertyM IO ()
shouldParseTo parser input expected = verdict =<< attempt parser input check
where
check actual
| actual == expected = Nothing
| otherwise =
Just $
unlines
[ "input: " <> show input,
"expected: " <> show expected,
"actual: " <> show actual
]
-- | The input must parse, with no claim about what it parses to. For the cases
-- where the parser only has to not fall over.
shouldParse :: DocumentParser -> Text -> PropertyM IO ()
shouldParse parser input = verdict =<< attempt parser input (const Nothing)
-- | the equivalent of Hedgehog's `success`
succeed :: PropertyM IO ()
succeed = pure ()
-- | the equivalent of Hedgehog's `fail`
failWith :: String -> PropertyM IO a
failWith message = stop $ counterexample message False