diff --git a/app/almanachmuszyny.hs b/app/almanachmuszyny.hs index ebc4f5c..775ff27 100644 --- a/app/almanachmuszyny.hs +++ b/app/almanachmuszyny.hs @@ -13,9 +13,9 @@ 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 + -- >>> 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) -- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem @@ -43,4 +43,4 @@ main = do abbrev="AlmMusz", lLevel=0, webpage=start} - extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem) + extractItemsStartingFromUrl shadowLibrary start (extractRecords)-- >>> arr toShadowItem) diff --git a/app/robot.hs b/app/robot.hs new file mode 100644 index 0000000..f48644e --- /dev/null +++ b/app/robot.hs @@ -0,0 +1,37 @@ +{-# LANGUAGE Arrows, NoMonomorphismRestriction #-} +import ShadowLibrary.Core +import Text.XML.HXT.Core +import Text.XML.HXT.XPath +import Data.List +import Data.List.Utils (replace) +import Text.Regex.Posix +import Text.Printf + +extractRecords = extractLinksWithText "//a" + >>> first (extractLinksWithText "//a[contains(@href,'.pdf')]") + + +toShadowItem :: ((String, String), String) -> ShadowItem +toShadowItem ((url, articleTitle), yearlyTitle) = + (defaultShadowItem url title) { + originalDate = extractMonthAndYear title, + itype = "periodical", + format = Just "pdf", + finalUrl = url + } + where title = "Kosmos " ++ yearlyTitle ++ (replace ".pdf" "" articleTitle) + +extractMonthAndYear :: String -> Maybe String +extractMonthAndYear n = + case n =~~ ("(1[6789]|20)[0-9][0-9]" :: String) of + Just year -> Just year + otherwise -> Nothing + +main = do + let start = "http://kosmos.icm.edu.pl/PDF/" + let shadowLibrary = ShadowLibrary {logoUrl=Nothing, + lname="kosmos.icm.edu", + abbrev="Kos", + lLevel=0, + webpage=start} + extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem) diff --git a/shadow-library.cabal b/shadow-library.cabal index cd77195..18ef722 100644 --- a/shadow-library.cabal +++ b/shadow-library.cabal @@ -58,8 +58,21 @@ executable almanachmuszyny , regex-posix , shadow-library default-language: Haskell2010 + +executable robot + hs-source-dirs: app + main-is: robot.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 + location: https://git.wmi.amu.edu.pl/s426229/twilight-library.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