47 lines
1.9 KiB
Haskell
47 lines
1.9 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 "//div[@class='sidebar']/ul/li/a" -- pary adres-tytuł
|
||
|
>>> second (arr $ replace "\r\n " " ") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków
|
||
|
-- >>> first (arr ((++"tr") . init)) -- modyfikujemy pierwszy element pary, czyli adres URL
|
||
|
>>> first (extractLinksWithText "//a[contains(@href,'post.php?wpfb')]") -- 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 = "Historia Informatyki " ++ yearlyTitle ++ " " ++ (replace "\r\n" "" (replace "\r\n " "" articleTitle))
|
||
|
date = "date"
|
||
|
|
||
|
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 = "http://klio.spit.iq.pl/"
|
||
|
let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
|
||
|
lname="Historia polskiej informatyki",
|
||
|
abbrev="HistPolInf",
|
||
|
lLevel=0,
|
||
|
webpage=start}
|
||
|
extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem)
|