Add NYTimes recipe extraction from __NEXT_DATA__ JSON tag

- Add Roux.NYTimes module with:
  - extractNextData: finds <script id="__NEXT_DATA__"> tag in HTML
    and parses its JSON content via aeson
  - parseNYTRecipe: maps NYTimes recipe JSON structure into a
    typed intermediate representation (NYTRecipe, NYTIngredientGroup,
    NYTIngredientItem, NYTStepGroup, NYTStepItem)
- Supports all NYTimes Cooking data: title, URL, total time, yield,
  topnote, ingredient groups (with section names), step groups
- Add aeson to dependencies
- Add 4 tests: extraction + parsing for both fried-rice and
  carrot-risotto example HTML files
- Wire NYTimesSpec into test runner
This commit is contained in:
2026-05-19 15:51:37 -04:00
parent 262faca90e
commit 21a4cea083
7 changed files with 443 additions and 4 deletions
+76
View File
@@ -0,0 +1,76 @@
module Roux.NYTimesSpec (spec) where
import qualified Data.Text.IO as TIO
import Test.Hspec (Spec, describe, it, shouldBe, shouldSatisfy)
import Roux.NYTimes
spec :: Spec
spec = do
describe "extractNextData" $ do
it "extracts JSON from fried-rice.html" $ do
html <- TIO.readFile "test-data/fried-rice.html"
let result = extractNextData html
result `shouldSatisfy` isJust
it "extracts JSON from carrot-risotto.html" $ do
html <- TIO.readFile "test-data/carrot-risotto.html"
let result = extractNextData html
result `shouldSatisfy` isJust
describe "parseNYTRecipe" $ do
it "parses fried-rice.html recipe correctly" $ do
html <- TIO.readFile "test-data/fried-rice.html"
let result = do
val <- extractNextData html
parseNYTRecipe val
case result of
Nothing -> error "failed to parse fried-rice recipe"
Just recipe -> do
nytTitle recipe `shouldBe` "Fried Rice"
nytUrl recipe `shouldBe` Just "https://cooking.nytimes.com/recipes/12177-fried-rice"
nytTotalTime recipe `shouldBe` Just "20 minutes"
nytRecipeYield recipe `shouldBe` Just "4 to 6 servings"
nytTopnote recipe `shouldSatisfy` isJust
length (nytIngredientGroups recipe) `shouldBe` 1
case nytIngredientGroups recipe of
[group] -> do
length (nytIngredients group) `shouldBe` 14
case nytIngredients group of
(firstIng : _) -> do
nytIngredientText firstIng `shouldBe` "tablespoons neutral oil, like canola or grapeseed"
nytIngredientQuantity firstIng `shouldBe` "3"
[] -> error "empty ingredient group"
_ -> error "expected 1 ingredient group"
case nytStepGroups recipe of
[stepGroup] -> do
length (nytSteps stepGroup) `shouldBe` 4
case nytSteps stepGroup of
(firstStep : _) -> nytStepNumber firstStep `shouldBe` 1
[] -> error "empty step group"
_ -> error "expected 1 step group"
it "parses carrot-risotto.html recipe correctly" $ do
html <- TIO.readFile "test-data/carrot-risotto.html"
let result = do
val <- extractNextData html
parseNYTRecipe val
case result of
Nothing -> error "failed to parse carrot-risotto recipe"
Just recipe -> do
nytTitle recipe `shouldBe` "Carrot Risotto With Chile Crisp"
nytUrl recipe `shouldBe` Just "https://cooking.nytimes.com/recipes/1024086-carrot-risotto-with-chile-crisp"
nytTotalTime recipe `shouldBe` Just "30 minutes"
nytRecipeYield recipe `shouldBe` Just "4 servings"
length (nytIngredientGroups recipe) `shouldBe` 1
case nytIngredientGroups recipe of
[group] -> length (nytIngredients group) `shouldBe` 11
_ -> error "expected 1 ingredient group"
case nytStepGroups recipe of
[stepGroup] -> length (nytSteps stepGroup) `shouldBe` 7
_ -> error "expected 1 step group"
-- | Check if a Maybe value is Just.
isJust :: Maybe a -> Bool
isJust (Just _) = True
isJust Nothing = False
+3
View File
@@ -3,13 +3,16 @@ module Main (main) where
import Test.Tasty (defaultMain, testGroup)
import Test.Tasty.Hspec (testSpec)
import qualified Roux.NYTimesSpec
import qualified Roux.ParserSpec
main :: IO ()
main = do
parserTests <- testSpec "Parser" Roux.ParserSpec.spec
nytTests <- testSpec "NYTimes" Roux.NYTimesSpec.spec
defaultMain $
testGroup
"roux-server"
[ parserTests
, nytTests
]