2011-07-01 49 views
5

我想從使用HXT的6行XML輸入中提取一些數據。我也想保留HXT,因爲Curl整合,因爲我有其他XML文件,後面有成千上萬行。從簡單的XML獲取數據

我的XML看起來是這樣的:

<?xml version = "1.0" encoding = "UTF-8"?> 
<find> 
    <set_number>228461</set_number> 
    <no_records>000000008</no_records> 
    <no_entries>000000008</no_entries> 
</find> 

而且我一直試圖扎堆如何解析。不幸的是,HXT的Wiki頁面並沒有很大的幫助(或者我只是忽略了一些東西)。

data FindResult = FindResult { 
     resultSetNumber :: String, 
     resultNoRecords :: Int, 
     resultNoEntries :: Int 
    } deriving (Eq, Show) 

resultParser :: ArrowXml a => a XmlTree FindResult 
resultParser = hasName "find" >>> getChildren >>> proc x -> do 
    setNumber <- isElem >>> hasName "set_number" >>> getChildren >>> getText -< x 
    noRecords <- isElem >>> hasName "no_records" >>> getChildren >>> getText -< x 
    noEntries <- isElem >>> hasName "no_entries" >>> getChildren >>> getText -< x 
    returnA -< FindResult setNumber (read noRecords) (read noEntries) 

find str = return . head =<< (runX $ readDocument [withValidate no, withCurl []] query >>> resultParser) 
    where query = "http://" ++ server ++ "/find?request=" ++ str 

我總是得到的是

*** Exception: Prelude.head: empty list 

所以,我想,解析必須去可怕的錯誤,因爲我檢查,並正確地獲得從查詢XML。

回答

6

我下面的作品(this example模仿):

{-# LANGUAGE Arrows #-} 

module Main 
     where 

import Text.XML.HXT.Core 
import System.Environment 

data FindResult = FindResult { 
     resultSetNumber :: String, 
     resultNoRecords :: Int, 
     resultNoEntries :: Int 
    } deriving (Eq, Show) 

resultParser :: ArrowXml a => a XmlTree FindResult 
resultParser = 
    deep (isElem >>> hasName "find") >>> proc x -> do 
    setNumber <- getText <<< getChildren <<< deep (hasName "set_number") -< x 
    noRecords <- getText <<< getChildren <<< deep (hasName "no_records") -< x 
    noEntries <- getText <<< getChildren <<< deep (hasName "no_entries") -< x 
    returnA -< FindResult setNumber (read noRecords) (read noEntries) 

main :: IO() 
main = do [src] <- getArgs 
      res <- runX $ (readDocument [withValidate no] src >>> resultParser) 
      print . head $ res 

測試:

$ dist/build/test/test INPUT 
FindResult {resultSetNumber = "228461", resultNoRecords = 8, resultNoEntries = 8} 
+0

非常感謝您! – Lanbo