2021-04-06 16:38:26 +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
2021-05-01 05:27:55 +02:00
import Data.List.Utils
2021-04-07 12:10:27 +02:00
2021-04-06 16:38:26 +02:00
extractRecords = extractLinksWithText " //a[@class='ramka'] " -- pary adres-tytuł
>>> first ( extractLinksWithText " //div/a[contains(@href, '.zip') and not (contains(@href,'text_offline.zip'))] " ) -- tutaj wybieramy paczki zip z plikami DjVu, jeśli byśmy chcieli bezpośrednio format DjVu wystarczy trochę zmienić zapytanie XPatha.
-- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem
toShadowItem :: ( ( String , String ) , String ) -> ShadowItem
toShadowItem ( ( url , articleTitle ) , yearlyTitle ) =
( defaultShadowItem url title ) {
2021-05-01 05:27:55 +02:00
originalDate = Just newDate ,
2021-04-06 16:38:26 +02:00
itype = " periodical " ,
format = Just " zip " ,
finalUrl = url
}
where title = " Tajemnica Atari " ++ yearlyTitle ++ " " ++ ( replace " \ r \ n " " " ( replace " \ r \ n " " " articleTitle ) )
date = getDate url
2021-04-07 12:10:27 +02:00
splitDate = split " _ " date
2021-05-01 05:27:55 +02:00
month = splitDate !! 0
year = splitDate !! 1
newDate = newFormatDate month year
2021-04-06 16:38:26 +02:00
getDate url =
case url =~~ " /([1-9]_9[0-9]|1[0-2]_9[0-9]|[1-9]-[1-9]_9[0-9]|1[0-2]-1[0-2]_9[0-9])/ " :: Maybe [ [ String ] ] of
Just [ [ _ , year ] ] -> year
otherwise -> error $ " unexpected url: " ++ url
2021-05-01 05:27:55 +02:00
newFormatDate :: String -> String -> String
newFormatDate month year =
if month =~ " ^[1-9]$ " then " 19 " ++ year ++ " -0 " ++ month
else if month =~ " ^1[0-2]$ " then " 19 " ++ year ++ " - " ++ month
else splittedMonth month year
splittedMonth :: String -> String -> String
splittedMonth month year = do
if month =~ " ^1*[0-9]-1*[0-9]$ " then do
let firstMonth = split " - " month !! 0
let secondMonth = split " - " month !! 1
if firstMonth =~ " ^[1-9]$ " && secondMonth =~ " ^[1-9]$ " then " 19 " ++ year ++ " -0 " ++ firstMonth ++ " -0 " ++ secondMonth
else if firstMonth =~ " ^[1-9]$ " && secondMonth =~ " ^1[0-9]$ " then " 19 " ++ year ++ " -0 " ++ firstMonth ++ " - " ++ secondMonth
else if firstMonth =~ " ^1[1-9]$ " && secondMonth =~ " ^1[0-9]$ " then " 19 " ++ year ++ " - " ++ firstMonth ++ " - " ++ secondMonth
else " wrong date "
else " wrong date "
2021-04-06 16:38:26 +02:00
main = do
let start = " http://krap.pl/mirrorz/atari/horror.mirage.com.pl/pixel/ "
let shadowLibrary = ShadowLibrary { logoUrl = Nothing ,
lname = " Tajemnica Atari " ,
abbrev = " TA " ,
lLevel = 0 ,
webpage = start }
extractItemsStartingFromUrl shadowLibrary start ( extractRecords >>> arr toShadowItem )
2021-05-01 05:27:55 +02:00