diff options
author | Joris | 2018-06-17 23:24:47 +0200 |
---|---|---|
committer | Joris | 2018-06-18 11:13:55 +0200 |
commit | 0a4d3c8f12dc5797a919a00b6bcaf759947687cc (patch) | |
tree | bcb89781e22c2314bf0c064ebb37cb7f8a362f5c /src/parser/haskell/Parser/LeboncoinParser.hs | |
parent | e2a5c7c5c596d057b6fa9c08a8204ce1429cfdc4 (diff) | |
download | ad-listener-0a4d3c8f12dc5797a919a00b6bcaf759947687cc.tar.gz ad-listener-0a4d3c8f12dc5797a919a00b6bcaf759947687cc.tar.bz2 ad-listener-0a4d3c8f12dc5797a919a00b6bcaf759947687cc.zip |
Add ouest france parser
Diffstat (limited to 'src/parser/haskell/Parser/LeboncoinParser.hs')
-rw-r--r-- | src/parser/haskell/Parser/LeboncoinParser.hs | 24 |
1 files changed, 24 insertions, 0 deletions
diff --git a/src/parser/haskell/Parser/LeboncoinParser.hs b/src/parser/haskell/Parser/LeboncoinParser.hs new file mode 100644 index 0000000..48eb80f --- /dev/null +++ b/src/parser/haskell/Parser/LeboncoinParser.hs @@ -0,0 +1,24 @@ +module Parser.LeboncoinParser + ( 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 (~== "<a>") tags + where tags = getTagsBetween "<li itemtype=http://schema.org/Offer>" "<div class=information-immo_content>" (parseTags page) + +parseAd :: [Tag Text] -> Maybe Ad +parseAd tags = do + name <- getTagTextAfter "<h2 class=item_title>" tags + location <- getTagAttribute "<meta itemprop=address>" (T.pack "content") tags + let price = getTagTextAfter "<h3 class=item_price>" tags + url <- getTagAttribute "<a>" (T.pack "href") tags + return (Ad name location price (T.concat [T.pack "https:", url])) |