This commit is contained in:
MikolajPaterka 2022-04-12 23:00:34 +02:00
parent 8883a924b4
commit b3032cbd44
6 changed files with 110 additions and 0 deletions

7
README.md Normal file
View File

@ -0,0 +1,7 @@
# Strona
Strona: https://bip.um.gdynia.pl/wyszukiwarka-uchwal-rady-miasta,7485/wyszukiwarka-uchwal,504923
# Wyniki
Lata 1990 - 1999 -> oldFiles.json
Lata 2000 - 2022 -> newFiles.json

71
app/test.hs Normal file
View File

@ -0,0 +1,71 @@
{-# 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
import System.IO
extractYears = extractLinksWithText "//div[@class='inner clearfix']//div[@class='row']//div[@class='col-sm-6']//ul//li//a" -- pary adres-tytuł
runYearsExtractor url = runX $ (arr (const url) >>> setTraceLevel 1 >>> extractYears)
extractNewFiles = extractLinksWithText "//a[@title='Pobierz plik']" -- pary adres-tytuł
runNewFilesExtractor url = runX $ (arr (const url) >>> setTraceLevel 1 >>> extractNewFiles)
extractOldFiles = extractLinksWithText "//div[@class='content']//div[@class='inner clearfix']//ul//li//a" -- pary adres-tytuł
runOldFilesExtractor url = runX $ (arr (const url) >>> setTraceLevel 1 >>> extractOldFiles)
--367-376
extractOldYears = extractLinksWithText "//nav[@class='nav-left']//ul//li[@class='single-article']/a" -- pary adres-tytuł
runOldYearsExtractor url = runX $ (arr (const url) >>> setTraceLevel 1 >>> extractOldYears)
mapToUrl :: ([Char], [Char]) -> [Char]
mapToUrl (url, title) = url
merge [] ys = ys
merge (x:xs) ys = x:merge ys xs
repeatNTimes 0 = return []
repeatNTimes n = do
let link = "https://bip.um.gdynia.pl/module/Contents/controller/Default/action/loadSubcategory/subcategory_id/" ++ show (377-n)
items <- runOldYearsExtractor link
let urls = map mapToUrl items
recursive <- repeatNTimes (n-1)
let results = merge urls recursive
return results
flatten :: [[a]] -> [a]
flatten xs = (\z n -> foldr (\x y -> foldr z y x) n xs) (:) []
getAllPages url = do
items <- runYearsExtractor url
let urls = map mapToUrl items
newFiles <- mapM runNewFilesExtractor urls
let fileUrls = map mapToUrl (flatten newFiles)
return fileUrls
getOldFiles = do
oldPages <- repeatNTimes 10
--let sub = take 2 oldPages
oldFiles <- mapM runOldFilesExtractor oldPages
let fileUrls = map mapToUrl (flatten oldFiles)
return fileUrls
main = do
let start = "https://bip.um.gdynia.pl/wyszukiwarka-uchwal-rady-miasta,7485/wyszukiwarka-uchwal,504923"
oldResults <- getOldFiles
oldF <- openFile "oldFiles.json" WriteMode
hPrint oldF oldResults
hClose oldF
--print (length results)
newResults <- getAllPages start
newF <- openFile "newFiles.json" WriteMode
hPrint newF newResults
hClose newF

1
newFiles.json Normal file
View File

@ -0,0 +1 @@
[]

1
oldFiles.json Normal file
View File

@ -0,0 +1 @@
[]

View File

@ -59,6 +59,17 @@ executable almanachmuszyny
, shadow-library
default-language: Haskell2010
executable test
hs-source-dirs: app
main-is: test.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

19
stack.yaml.lock Normal file
View File

@ -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