From 149a0470b73781022e584aaeaa7ce871d6f4173b Mon Sep 17 00:00:00 2001 From: Joris Date: Tue, 19 Jun 2018 22:49:16 +0200 Subject: Add automatic tests on remote pages --- src/lib/haskell/Parser/OuestFranceParser.hs | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) create mode 100644 src/lib/haskell/Parser/OuestFranceParser.hs (limited to 'src/lib/haskell/Parser/OuestFranceParser.hs') diff --git a/src/lib/haskell/Parser/OuestFranceParser.hs b/src/lib/haskell/Parser/OuestFranceParser.hs new file mode 100644 index 0000000..f46ed03 --- /dev/null +++ b/src/lib/haskell/Parser/OuestFranceParser.hs @@ -0,0 +1,25 @@ +module Parser.OuestFranceParser + ( parse + ) where + +import Data.Maybe (catMaybes) +import Data.Text (Text) +import qualified Data.Text as T +import Text.HTML.TagSoup + +import Model.Ad (Ad (Ad)) +import Parser.Utils + +parse :: Text -> [Ad] +parse page = + catMaybes . fmap parseAd $ partitions (~== (T.unpack "")) tags + where tags = getTagsBetween "
" "
" (parseTags page) + +parseAd :: [Tag Text] -> Maybe Ad +parseAd tags = do + name <- getTagTextAfter "" tags + location <- getTagTextAfter "" tags + let price = getTagTextAfter "" tags + let startUrl = "https://www.ouestfrance-immo.com/" + url <- getTagAttribute "" "href" tags + return (Ad name location price (T.concat [startUrl, url])) -- cgit v1.2.3