twilight-library/app/zadanie.hs

65 lines
2.7 KiB
Haskell
Raw Normal View History

2022-03-30 21:18:39 +02:00
{-# 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
extractNestedRecords = extractLinksWithText "//a[@class='archiveVolume' and not(contains(@href, '.PDF')) and not(contains(@href, '.pdf'))]" -- pary adres-tytuł
2022-05-05 21:24:26 +02:00
>>> second (arr $ replace "%09" "") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków
2022-03-30 21:18:39 +02:00
>>> first (extractLinksWithText "//a[contains(@href,'.pdf')]") -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego
-- ostatecznie wyjdą trójki ((adres URL, tytuł artykułu), tytuł rocznika)
extractRecords = extractLinksWithText "//a[@class='archiveVolume' and (contains(@href,'.pdf') or contains(@href,'.PDF'))]"
>>> second (arr $ replace "\t" " ")
-- ... 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
}
2022-05-05 21:24:26 +02:00
where title = "Zagadnienia ekonomiki rolnej " ++ (replace "\t" "" (replace "\r\n" "" yearlyTitle)) ++ " " ++ (replace "\t" "" (replace "\r\n" "" (replace " " "" articleTitle)))
2022-03-30 21:18:39 +02:00
date = getDate yearlyTitle
toShadowItemTop :: (String, String) -> ShadowItem
toShadowItemTop (url, articleTitle) =
(defaultShadowItem url title) {
originalDate = Just date,
itype = "periodical",
format = Just "pdf",
finalUrl = url
}
2022-05-05 21:24:26 +02:00
where title = "Zagadnienia ekonomiki rolnej " ++ " " ++ (replace "\t" "" (replace "\r\n " "" articleTitle))
2022-03-30 21:18:39 +02:00
date = getDateTop url
getDate url =
case url =~~ "[0-9]?[0-9]/(19[0-9][0-9]|20[0-9][0-9])" :: Maybe [[String]] of
Just [[month, year]] -> month
otherwise -> error $ "unexpected url: " ++ url
getDateTop url =
case url =~~ "[0-9]*(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 = "http://www.zer.waw.pl/Archive"
let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
lname="Ekonomika Rolna",
abbrev="EkonRol",
lLevel=0,
webpage=start}
extractItemsStartingFromUrl shadowLibrary start (extractNestedRecords >>> arr toShadowItem)
extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItemTop)