diff --git a/README.md b/README.md new file mode 100644 index 0000000..8c6fa5a --- /dev/null +++ b/README.md @@ -0,0 +1,7 @@ +# Build and run project + +`stack install` + +`stack build` + +`stack exec pti` \ No newline at end of file diff --git a/app/pti.hs b/app/pti.hs new file mode 100644 index 0000000..e402e5b --- /dev/null +++ b/app/pti.hs @@ -0,0 +1,46 @@ + +{-# 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 "//div[@class='sidebar']/ul/li/a" -- 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 "//a[contains(@href,'post.php?wpfb')]") -- 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 +toShadowItem :: ((String, String), String) -> ShadowItem +toShadowItem ((url, articleTitle), yearlyTitle) = + (defaultShadowItem url title) { + originalDate = Just date, + itype = "periodical", + format = Just "pdf", + finalUrl = url + } + where title = "Historia Informatyki " ++ yearlyTitle ++ " " ++ (replace "\r\n" "" (replace "\r\n " "" articleTitle)) + date = "date" + +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 = "http://klio.spit.iq.pl/" + let shadowLibrary = ShadowLibrary {logoUrl=Nothing, + lname="Historia polskiej informatyki", + abbrev="HistPolInf", + lLevel=0, + webpage=start} + extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem) diff --git a/shadow-library.cabal b/shadow-library.cabal index cd77195..1aa312e 100644 --- a/shadow-library.cabal +++ b/shadow-library.cabal @@ -5,8 +5,8 @@ description: Please see README.md homepage: http://github.com/name/project license: Proprietary license-file: LICENSE -author: Your name here -maintainer: your.address@example.com +author: Adam Wojdyła +maintainer: adawoj5@st.amu.edu.pl -- copyright: category: Web build-type: Simple @@ -34,18 +34,18 @@ library , tz default-language: Haskell2010 --- executable maly-modelarz-exe --- hs-source-dirs: app --- main-is: malymodelarz.hs --- ghc-options: -threaded -rtsopts -with-rtsopts=-N --- build-depends: base --- , hxt --- , hxt-curl --- , hxt-xpath --- , MissingH --- , regex-posix --- , shadow-library --- default-language: Haskell2010 +executable pti + hs-source-dirs: app + main-is: pti.hs + ghc-options: -threaded -rtsopts -with-rtsopts=-N + build-depends: base + , hxt + , hxt-curl + , hxt-xpath + , MissingH + , regex-posix + , shadow-library + default-language: Haskell2010 executable almanachmuszyny hs-source-dirs: app @@ -62,4 +62,4 @@ executable almanachmuszyny source-repository head type: git - location: https://github.com/name/project + location: https://git.wmi.amu.edu.pl/s444507/twilight-library 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