diff --git a/app/archiwapilsudski.hs b/app/archiwapilsudski.hs new file mode 100644 index 0000000..cf45714 --- /dev/null +++ b/app/archiwapilsudski.hs @@ -0,0 +1,45 @@ + +{-# 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 "//a[@class='roczniki']" -- 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 "//li/a[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, title) = + (defaultShadowItem url title) { + originalDate = Just date, + itype = "periodical", + format = Just "pdf", + finalUrl = url + } + where title = title + date = '12-34-5678' + +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://archiwa.pilsudski.org/" + let shadowLibrary = ShadowLibrary {logoUrl=Nothing, + lname="Archiwa Pilsudski", + abbrev="ArchPil", + lLevel=0, + webpage=start} + extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem) diff --git a/shadow-library.cabal b/shadow-library.cabal index cd77195..0792e6e 100644 --- a/shadow-library.cabal +++ b/shadow-library.cabal @@ -60,6 +60,19 @@ executable almanachmuszyny default-language: Haskell2010 +executable archiwapilsudski + hs-source-dirs: app + main-is: archiwapilsudski.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 location: https://github.com/name/project 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