diff --git a/app/pepowo.hs b/app/pepowo.hs new file mode 100644 index 0000000..75c9bf3 --- /dev/null +++ b/app/pepowo.hs @@ -0,0 +1,49 @@ + +{-# 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 "//h5/a[contains(@href,'-r')]" -- pary adres-tytuł + >>> first (extractLinksWithText "//li/a[contains(@href,'.pdf')]") -- pobieramy stronę z adresu URL i wyciągamy linki z tej strony pasujące do wyrażenia XPathowego + >>> second (arr $ replace "\r\n " " ") + >>> second (arr $ replace "\347" "s") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków + >>> second (arr $ replace "\281" "e") -- czyścimy drugi element pary, czyli tytuł z niepotrzebnych białych znaków + -- 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, articleNumber), mainTitle) = + (defaultShadowItem url title) { + originalDate = date, + itype = "periodical", + format = Just "pdf", + finalUrl = url + } + where title = "Wiesci Pepowa" ++ (replace ".pdf" "" (replace "WP" "numer" articleNumber)) + date = getDate mainTitle + +getDate :: String -> Maybe String +getDate title = + case title =~~ "(19[0-9][0-9]|20[0-9][0-9])" :: Maybe [[String]] of + Just [[_, year]] -> (Just year) + otherwise -> Nothing + +main = do + let start = "http://archiwum2019.pepowo.pl/news/c-10/gazeta" + let start2 = "http://archiwum2019.pepowo.pl/news/c-10/gazeta/page/2" + let shadowLibrary = ShadowLibrary {logoUrl=Nothing, + lname="Wieści Pępowa", + abbrev="WiesPep", + lLevel=0, + webpage="http://archiwum2019.pepowo.pl"} + extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem) + extractItemsStartingFromUrl shadowLibrary start2 (extractRecords >>> arr toShadowItem) diff --git a/shadow-library.cabal b/shadow-library.cabal index cd77195..276a0ac 100644 --- a/shadow-library.cabal +++ b/shadow-library.cabal @@ -59,6 +59,18 @@ executable almanachmuszyny , shadow-library default-language: Haskell2010 +executable pepowo + hs-source-dirs: app + main-is: pepowo.hs + ghc-options: -threaded -rtsopts -with-rtsopts=-N + build-depends: base + , hxt + , hxt-xpath + , MissingH + , regex-posix + , shadow-library + default-language: Haskell2010 + source-repository head type: git diff --git a/stack.yaml.lock b/stack.yaml.lock new file mode 100644 index 0000000..cc80ef3 --- /dev/null +++ b/stack.yaml.lock @@ -0,0 +1,19 @@ +# This file was autogenerated by Stack. +# You should not edit this file by hand. +# For more information, please see the documentation at: +# https://docs.haskellstack.org/en/stable/lock_files + +packages: +- completed: + hackage: hxt-xpath-9.1.2.2@sha256:9cd590ae93a04573db8f90fa4094625ebd97dded45da7667c577ce6b38a42900,1999 + pantry-tree: + size: 2225 + sha256: aee2f75974e868ff429b8ff349a29667536c60397098f5dfedc968d1951511bb + original: + hackage: hxt-xpath-9.1.2.2 +snapshots: +- completed: + size: 507596 + url: https://raw.githubusercontent.com/commercialhaskell/stackage-snapshots/master/lts/11/9.yaml + sha256: 42f472dbf06482da1b3319241f3e3b3593a45bd7d4f537d2789f21386b9b2ad3 + original: lts-11.9