twilight-library/app/pbsociety.hs

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

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

import Text.Regex.Posix
import Text.Printf


extractRecords = extractLinksWithText "//a[@class='image-link']"  -- pary adres-tytuł
                 >>> second (arr $ replace "\n" " ")
                 >>> first (extractLinksWithText "//div/a[contains(@href,'.pdf') and not(@class)]") -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego
                 -- ostatecznie wyjdą trójki ((adres URL, tytuł artykułu), tytuł rocznika)

-- ... 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 = "Pbsociety " ++ yearlyTitle
        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 start = "https://pbsociety.org.pl/repository/discover?filtertype_1=has_content_in_original_bundle&filter_relational_operator_1=equals&filter_1=true&filtertype_2=title&filter_relational_operator_2=contains&filter_2=&submit_apply_filter=Apply&query=&scope=%2F&rpp=260&sort_by=score&order=desc"
    let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
                                       lname="Polish Botanical Society Repository",
                                       abbrev="Pbsociety",
                                       lLevel=0,
                                       webpage=start}
    extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem)