twilight-library/app/elektronikapraktyczna.hs

91 lines
3.6 KiB
Haskell
Raw Normal View History

2021-04-06 17:51:18 +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
import Data.Char (toLower)
2021-04-06 17:51:18 +02:00
class Nothingish a where
nada :: a
instance Nothingish [a] where
nada = []
eliminate :: (Nothingish a) => Maybe a -> a
eliminate (Just a) = a
eliminate Nothing = nada
toLowerString :: [Char] -> [Char]
toLowerString str = [ toLower x | x <- str]
mToString :: Maybe String -> String
mToString n
| n == Nothing = ""
| otherwise = eliminate n
extractMonth :: String -> String
extractMonth n =
case n =~~ ("[A-za-z]+" :: String) of
2021-04-17 18:13:05 +02:00
Just month -> eliminate (baseMonthNameToNumber (toLowerString month))
otherwise -> ""
mExtractYear :: String -> String
mExtractYear n =
case n =~~ ("(1[6789]|20)[0-9][0-9]" :: String) of
Just year -> year
otherwise -> ""
changeDate :: String -> String
2021-04-17 18:13:05 +02:00
changeDate a = eliminate (extractYear a) ++ "-" ++ extractMonth a
extractRecords = extractLinksWithText "(//a[@class='magazine-list__year-item'])[last()]" -- pary adres-tytuł
--extractRecords = extractLinksWithText "//a[@class='magazine-list__year-item']" -- pary adres-tytuł
>>> second (arr $ replace "\r\n " "") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków
>>> second (arr $ replace " " "") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków
2021-04-06 17:51:18 +02:00
-- >>> first (arr ((++"tr") . init)) -- modyfikujemy pierwszy element pary, czyli adres URL
>>> first (extractLinksWithText "//div[@class='magazine-list__item']/a") -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego
>>> first (second (arr $ replace "\r\n " ""))
>>> first (first (
extractLinksWithText "//div[@class='files__item']/a[contains(@href,'.pdf')]"
>>> second (arr $ replace "\r\n " "")
-- >>> first (arr $ replace "//" "/")
)
>>> second (arr $ changeDate)
) -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego
2021-04-06 17:51:18 +02:00
-- ostatecznie wyjdą trójki ((adres URL, tytuł artykułu), tytuł rocznika)
-- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem
toShadowItem :: (((String, String), String), String) -> ShadowItem
toShadowItem (((url, chapterTitle), articleTitle), yearlyTitle) =
2021-04-06 17:51:18 +02:00
(defaultShadowItem url title) {
originalDate = Just date,
itype = "periodical",
format = Just "pdf",
finalUrl = url
}
where title = "Elektronika Praktyczna " ++ (replace " " "" articleTitle)
date = yearlyTitle
2021-04-06 17:51:18 +02:00
getDate yearlyTitle =
case yearlyTitle =~~ "/(19[0-9][0-9]|20[0-9][0-9])/" :: Maybe [[String]] of
2021-04-06 17:51:18 +02:00
Just [[_, year]] -> year
otherwise -> error $ "unexpected yearlyTitle: " ++ yearlyTitle
2021-04-06 17:51:18 +02:00
main = do
let start = "https://ep.com.pl/archiwum/"
let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
lname="Elektronika praktyczna",
abbrev="EP",
lLevel=0,
webpage=start}
extractItemsStartingFromUrl shadowLibrary start (extractRecords)-- >>> arr toShadowItem)