forked from filipg/twilight-library
45 lines
1.6 KiB
Haskell
45 lines
1.6 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 "" -- pary adres-tytuł
|
||
|
>>> first (extractLinksWithText "//a[@target='_blank'][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)
|
||
|
|
||
|
|
||
|
toShadowItem :: ((String, String), String) -> ShadowItem
|
||
|
toShadowItem ((url, articleTitle), yearlyTitle) =
|
||
|
(defaultShadowItem url title) {
|
||
|
originalDate = Just date,
|
||
|
itype = "periodical",
|
||
|
format = Just "pdf",
|
||
|
finalUrl = url
|
||
|
}
|
||
|
where title = replace "\"" "'" (replace "\t\t\t\t\t" "" (replace "\n" "" yearlyTitle)) ++ "articleTitle: " ++ 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 start = "https://www.czasopisma.centralnabibliotekapttk.pl"
|
||
|
let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
|
||
|
lname="Chelmek",
|
||
|
abbrev="Chelmek",
|
||
|
lLevel=0,
|
||
|
webpage=start}
|
||
|
extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem)
|