0

我在使用 alpakka 的 XmlParsing Flow 时遇到问题

val files: List[String] = ... // file paths locally on disk

// simple source emitting the contents of 2 XML files
val documentSource = FileIO.fromPath(Paths.get(files.head))
  .concat(FileIO.fromPath(Paths.get(files(1))))

val contentFlow: Flow[ParseEvent, CustomContent, Notused] =
  Flow.fromGraph(new ContentProcessorFlow)

documentSource
  .via(XmlParsing.parser)
  .via(contentFlow)
  .to(Sink.foreach(println))
  .run

运行此程序时,图表会打印由 发出的元素contentFlow,这些元素是正确的并且与第一个文件的预期值匹配。之后,抛出此异常:

[ERROR] [12/20/2018 16:32:23.648] [Sync-akka.actor.default-dispatcher-2] [akka://Sync/system/StreamSupervisor-0/flow-0-0-ignoreSink] Error in stage [akka.stream.alpakka.xml.impl.StreamingXmlParser@36b80955]: Illegal processing instruction target: 'xml' (case insensitive) is reserved by the xml specification
 at [row,col {unknown-source}]: [44,17]
com.fasterxml.aalto.WFCException: Illegal processing instruction target: 'xml' (case insensitive) is reserved by the xml specification
 at [row,col {unknown-source}]: [44,17]
at com.fasterxml.aalto.in.XmlScanner.reportInputProblem(XmlScanner.java:1333)
at com.fasterxml.aalto.async.AsyncByteScanner.checkPITargetName(AsyncByteScanner.java:665)
at com.fasterxml.aalto.async.AsyncByteArrayScanner.handlePI(AsyncByteArrayScanner.java:2091)
at com.fasterxml.aalto.async.AsyncByteArrayScanner.nextFromProlog(AsyncByteArrayScanner.java:1064)
at com.fasterxml.aalto.stax.StreamReaderImpl.next(StreamReaderImpl.java:802)
at akka.stream.alpakka.xml.impl.StreamingXmlParser$$anon$1.advanceParser(StreamingXmlParser.scala:55)

我了解这里发生的基本情况 - 解析器抱怨<?xml version="1.0" encoding="UTF-8"?>第二个文件顶部的指令的 ByteString,但我对流不够熟悉,不知道该怎么做。如果我删除该指令,我会得到一个稍微不同的例外,即拥有 2 个根元素。

我的目标是构建一个从某个位置读取文件并发出CustomContent以供进一步处理的图。我怎样才能修改它以将每个文件视为一个不同的输入单元?

4

1 回答 1

1

将文件视为不同Source的 s,然后将它们合并为一个Source

val files: List[String] = ???

val sources: List[Source[CustomContent, Future[IOResult]]] =
  files
    .map { f =>
      FileIO.fromPath(Paths.get(f))
        .via(XMLParsing.parser)
        .via(contentFlow)
    }

val mergedSource: Source[CustomContent, NotUsed] =
  Source(sources).flatMapConcat(identity)

mergedSource.runForeach(println)
于 2018-12-20T22:31:24.313 回答