{-# LANGUAGE Arrows, NoMonomorphismRestriction #-}
import ShadowLibrary.Core

import Text.XML.HXT.Core
import Text.XML.HXT.XPath
import Data.List
import Data.List.Utils (replace)

import Text.Regex.Posix
import Text.Printf


extractRecords = extractLinksWithText "//div[@class='entry-content']/p[count(a)=1]/a"
                 >>> second (arr $ replace "\r\n " " ")
                 >>> first (extractLinksWithText "//li/a[contains(@href,'.pdf')]")

-- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem
toShadowItem :: ((String, String), String) -> ShadowItem
toShadowItem ((url, articleTitle), yearlyTitle) =
  (defaultShadowItem url title) {
    originalDate = Just date,
    itype = "periodical",
    format = Just "pdf",
    finalUrl = url
    }
  where title = "Almanach Muszyny " ++ yearlyTitle ++ " " ++ (replace "\r\n" "" (replace "\r\n          " "" articleTitle))
        date = getDate url

getDate url =
  case url =~~ "/(19[0-9][0-9]|20[0-9][0-9])/" :: Maybe [[String]] of
    Just [[_, year]] -> year
    otherwise -> error $ "unexpected url: " ++ url


main = do
    let baseUrl = "https://www.ptd.pl/?page_id=7"
    let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
                                    lname="Polskie Towarzystwo Dendrologiczne",
                                    abbrev="ptd",
                                    lLevel=0,
                                    webpage=baseUrl}
    extractItemsStartingFromUrl shadowLibrary baseUrl (extractRecords >>> arr toShadowItem)