forked from filipg/twilight-library
46 lines
1.8 KiB
Haskell
46 lines
1.8 KiB
Haskell
|
|
{-# 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[contains(@title,'Aneks')]" -- pary tytuł-adres
|
|
>>> first (arr $ replace "\r\n " " ") -- czyścimy pierwszy element pary, czyli tytuł z niepotrzebnych białych znaków
|
|
>>> second (extractLinksWithText "//div/a[contains(@href,'.pdf')]") -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego
|
|
-- ostatecznie wyjdą trójki? (Tytuł, (link do pdf, blank?))
|
|
|
|
-- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem
|
|
toShadowItem :: (String, (String, String)) -> ShadowItem
|
|
toShadowItem (yearlyTitle, (url, blank)) =
|
|
(defaultShadowItem url title) {
|
|
originalDate = Just date,
|
|
itype = "periodical",
|
|
format = Just "pdf",
|
|
finalUrl = url
|
|
}
|
|
where title = yearlyTitle ++ " " ++ (replace "\r\n" "" (replace "\r\n " "" blank))
|
|
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://aneks.kulturaliberalna.pl/archiwum-aneksu/"
|
|
let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
|
|
lname="Archiwum Aneksu",
|
|
abbrev="ArchAnek",
|
|
lLevel=0,
|
|
webpage=start}
|
|
extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem)
|