Pismo sulimczyk - bot

app/sulimczykpismo.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[contains(@href,'.pdf')]"
+                 >>> second (arr $ replace "\r\n            " " ")
+
+toShadowItem :: ((String, String)) -> ShadowItem
+toShadowItem ((url, periodicalTitle)) =
+  (defaultShadowItem url title) {
+    originalDate = Just date,
+    itype = "periodical",
+    format = Just "pdf",
+    finalUrl = url
+    }
+  where title = "Sulimczyk - pismo harcerskie" ++ " " ++ (replace "\r\n" "" (replace "\r\n          " "" periodicalTitle))
+        date = getDate $ show $ periodicalTitle
+        
+getDate title = 
+    case title =~~ "(([0-9][0-9]{0,1}) ([I|V|X]{1,3}) (19[0-9][0-9]|20[0-9][0-9]))" :: Maybe String of
+        Just date -> date
+        otherwise ->
+          case title =~~ "(0[1-9]|[12][0-9]|3[01]).(0[1-9]|1[0-2]).((1[789]|20)[0-9][0-9])" :: Maybe String of
+            Just date -> date
+            otherwise ->
+              case title =~~ "(19[0-9][0-9]|20[0-9][0-9])" :: Maybe String of
+                Just year -> year
+                otherwise -> error $ "unexpected title: " ++ title
+
+main = do
+    let start = "http://www.16wdh.pl/s28,Sulimczyk"
+    let shadowLibrary = ShadowLibrary {logoUrl=Nothing,
+                                       lname="Sulimczyk - pismo harcerskie",
+                                       abbrev="SulPis",
+                                       lLevel=0,
+                                       webpage=start}
+    extractItemsStartingFromUrl shadowLibrary start (extractRecords >>> arr toShadowItem)

shadow-library.cabal

@@ -59,6 +59,18 @@ executable almanachmuszyny
                       , shadow-library
    default-language:    Haskell2010
 
+executable sulimczykpismo
+   hs-source-dirs:      app
+   main-is:             sulimczykpismo.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

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