module Parser.Detail
( parse
) where
import Data.Text (Text)
import qualified Data.Text as T
import Data.Map (Map)
import qualified Data.Map as M
import Data.Maybe (catMaybes, fromMaybe)
import Text.HTML.TagSoup
import Model.Detail
import Parser.Utils
parse :: Text -> Detail
parse page =
let tags = parseTags page
in Detail
{ description = parseDescription tags
, images = map (\url -> T.concat [T.pack "https:", url]) $ getTagAttributes "" (T.pack "content") tags
, properties = parseProperties tags
}
parseDescription :: [Tag Text] -> Maybe Text
parseDescription tags =
let descriptionTags = getTagsBetween "" "
" tags
in if null descriptionTags
then
Nothing
else
let replaceBr = map (\tag -> if tag ~== "
" then TagText (T.pack "\n") else tag)
in Just . T.strip . renderTags . replaceBr $ descriptionTags
parseProperties :: [Tag Text] -> Map Text Text
parseProperties tags =
let mbUtagData = getTagTextAfter "" . getTagsAfter "" $ tags
in fromMaybe M.empty (fmap parseUtagData mbUtagData)
parseUtagData :: Text -> Map Text Text
parseUtagData =
M.fromList
. catMaybes
. fmap parseUtag
. T.splitOn (T.pack ",")
. T.takeWhile (/= '}')
. T.drop 1
. T.dropWhile (/= '{')
parseUtag :: Text -> Maybe (Text, Text)
parseUtag utag =
case T.splitOn (T.pack ":") utag of
[x, y] -> Just (T.strip x, removeQuotes y)
_ -> Nothing
removeQuotes :: Text -> Text
removeQuotes =
T.takeWhile (/= '\"')
. T.dropWhile (== '\"')
. T.strip