From 3e9ad1fa0bb3daca60bc9dfc369659ef5948ff7a Mon Sep 17 00:00:00 2001
From: Anna Nowak <annadnowak98@gmail.com>
Date: Tue, 6 Apr 2021 23:25:55 +0200
Subject: [PATCH] Funkcja do wydobycia roku z url, funkcja do wydobycia url i
 tekstu

---
 app/ptd.hs | 32 +++++++++++++++++++++-----------
 1 file changed, 21 insertions(+), 11 deletions(-)

diff --git a/app/ptd.hs b/app/ptd.hs
index 2adf3f6..6c11946 100644
--- a/app/ptd.hs
+++ b/app/ptd.hs
@@ -1,5 +1,7 @@
 
+{-# LANGUAGE UTF #-}
 {-# LANGUAGE Arrows, NoMonomorphismRestriction #-}
+
 import ShadowLibrary.Core
 
 import Text.XML.HXT.Core
@@ -10,27 +12,35 @@ import Data.List.Utils (replace)
 import Text.Regex.Posix
 import Text.Printf
 
+extractNestedLinksWithText xpathCondition = (downloadDocumentWithEncoding "UTF-8" &&& this)
+                                      >>> first (getXPathTrees xpathCondition
+                                                 >>> ((getXPathTrees "//a" >>> getAttrValue "href")
+                                                      &&& (listA (deep isText >>> getText)
+                                                           >>> arr (intercalate " "))))
+                                      >>> arr rotateSecTh
+                                      >>> first expandURIFixed
 
-extractRecords = extractLinksWithText "//div[@class='entry-content']/p[count(a)=1]/a"
-                 >>> second (arr $ replace "\r\n " " ")
-                 >>> first (extractLinksWithText "//li/a[contains(@href,'.pdf')]")
+extractRecords = extractLinksWithText "//div[@class='entry-content']/p/a[contains(@href, 'id')]"
+                >>> first (extractNestedLinksWithText "//div[@class='entry-content']/p[strong[a]] | //div[@class='entry-content']/p[a]")
 
--- ... a tutaj te trójki przerabiamy do docelowej struktury ShadowItem
 toShadowItem :: ((String, String), String) -> ShadowItem
-toShadowItem ((url, articleTitle), yearlyTitle) =
+toShadowItem ((url, articleTitle), magazineTitle) =
   (defaultShadowItem url title) {
     originalDate = Just date,
     itype = "periodical",
     format = Just "pdf",
     finalUrl = url
     }
-  where title = "Almanach Muszyny " ++ yearlyTitle ++ " " ++ (replace "\r\n" "" (replace "\r\n          " "" articleTitle))
-        date = getDate url
+  where title = magazineTitle ++ " - " ++ (replace " \8211 pdf" "" articleTitle)
+        date = getYear url
 
-getDate url =
-  case url =~~ "/(19[0-9][0-9]|20[0-9][0-9])/" :: Maybe [[String]] of
+getYear :: String -> String
+getYear url =
+  case url =~~ "/rocznik[0-9]{2}/" :: Maybe [[String]] of
     Just [[_, year]] -> year
-    otherwise -> error $ "unexpected url: " ++ url
+    otherwise ->  case url =~~ "/(19[0-9][0-9]|20[0-9][0-9])/" :: Maybe [[String]] of
+                  Just [[_, year]] ->  "19" ++ (replace "rocznik" "" year)
+                  otherwise -> ""
 
 
 main = do
@@ -40,4 +50,4 @@ main = do
                                     abbrev="ptd",
                                     lLevel=0,
                                     webpage=baseUrl}
-    extractItemsStartingFromUrl shadowLibrary baseUrl (extractRecords >>> arr toShadowItem)
+    extractItemsStartingFromUrl shadowLibrary baseUrl (extractRecords >>> arr toShadowItem)
\ No newline at end of file