Fixing parsing errors, and use Text from now

3 files changed, 36 insertions, 21 deletions

diff --git a/src/Parser/Detail.hs b/src/Parser/Detail.hs
index ab1b0ca..b787772 100644
--- a/src/Parser/Detail.hs
+++ b/src/Parser/Detail.hs
@@ -2,14 +2,23 @@ module Parser.Detail
   ( parseDetail
   ) where
 
+import qualified Data.Text as T
+
 import Text.HTML.TagSoup
 
 import Model.Detail
 
 import Parser.Utils
 
-parseDetail :: String -> Detail
+parseDetail :: T.Text -> Detail
 parseDetail page =
   let tags = parseTags page
-      description = getTagText "<div class=content>" tags
+      descriptionTags = getTagsBetween "<div itemprop=description>" "</div>" tags
+      description =
+        if null descriptionTags
+          then
+            Nothing
+          else
+            let replaceBr = map (\tag -> if tag ~== "<br>" then TagText (T.pack "\n") else tag)
+            in  Just . T.strip . renderTags . replaceBr $ descriptionTags
   in  Detail { description = description }
diff --git a/src/Parser/Resume.hs b/src/Parser/Resume.hs
index bd73912..6cd4415 100644
--- a/src/Parser/Resume.hs
+++ b/src/Parser/Resume.hs
@@ -3,6 +3,7 @@ module Parser.Resume
   ) where
 
 import Data.Maybe (catMaybes)
+import qualified Data.Text as T
 
 import Text.HTML.TagSoup
 
@@ -10,7 +11,7 @@ import Model.Resume
 
 import Parser.Utils
 
-parseResumes :: String -> [Resume]
+parseResumes :: T.Text -> [Resume]
 parseResumes page =
   case sections (~== "<div class=list-lbc>") (parseTags page) of
     [] ->
@@ -19,9 +20,9 @@ parseResumes page =
       let lbcTags = takeWhile (~/= "<div id=alertesCartouche>") sectionTags
       in  catMaybes . fmap parseResume $ partitions (~== "<a>") lbcTags
 
-parseResume :: [Tag String] -> Maybe Resume
+parseResume :: [Tag T.Text] -> Maybe Resume
 parseResume item = do
-  name <- getTagText "<h2 class=title>" item
-  let price = getTagText "<div class=price>" item
-  url <- getTagAttribute "<a>" "href" item
-  return Resume { name = name, price = price, url = url }
+  name <- getTagTextAfter "<h2 class=title>" item
+  let price = getTagTextAfter "<div class=price>" item
+  url <- getTagAttribute "<a>" (T.pack "href") item
+  return Resume { name = name, price = price, url = T.unpack url }
diff --git a/src/Parser/Utils.hs b/src/Parser/Utils.hs
index 4864e00..8527777 100644
--- a/src/Parser/Utils.hs
+++ b/src/Parser/Utils.hs
@@ -1,30 +1,35 @@
 module Parser.Utils
-  ( getTagAttribute
-  , getTagText
+  ( getTagsBetween
+  , getTagAttribute
+  , getTagTextAfter
   ) where
 
 import Data.List (find, findIndex)
 import Data.Maybe (listToMaybe)
+import qualified Data.Text as T
 
 import Text.HTML.TagSoup
 
-getTagAttribute :: String -> String -> [Tag String] -> Maybe String
-getTagAttribute selector attribute item =
-  find (~== selector) item >>= maybeTagAttribute attribute
+getTagsBetween :: String -> String -> [Tag T.Text] -> [Tag T.Text]
+getTagsBetween beginSelector endSelector =
+  takeWhile (~/= endSelector)
+  . drop 1
+  . dropWhile (~/= beginSelector)
 
-getTagText :: String -> [Tag String] -> Maybe String
-getTagText selector item =
-  case findIndex (~== selector) item of
-    Just index -> fmap trim $ safeGetAt (index + 1) item >>= maybeTagText
+getTagAttribute :: String -> T.Text -> [Tag T.Text] -> Maybe T.Text
+getTagAttribute selector attribute tags =
+  find (~== selector) tags >>= maybeTagAttribute attribute
+
+getTagTextAfter :: String -> [Tag T.Text] -> Maybe T.Text
+getTagTextAfter selector tags =
+  case findIndex (~== selector) tags of
+    Just index -> fmap T.strip $ safeGetAt (index + 1) tags >>= maybeTagText
     Nothing -> Nothing
 
-maybeTagAttribute :: String -> Tag String -> Maybe String
+maybeTagAttribute :: T.Text -> Tag T.Text -> Maybe T.Text
 maybeTagAttribute name (TagOpen _ xs) =
   fmap snd . find (\(x, _) -> x == name) $ xs
 maybeTagAttribute attribute _ = Nothing
 
-trim :: String -> String
-trim = unwords . words
-
 safeGetAt :: Int -> [a] -> Maybe a
 safeGetAt index = listToMaybe . drop index