java - Spring Integration Aggregator -

i want use aggregator create message out of 2 messages, i'm not sure how this.

at moment i'm reading in 2 files directory , want aggregate messages one.

my whole project looks this:

read in .zip -> pass custom message handler unzips directory -> read files directory -> try aggregate them

it great if send message 2 payloads after unzipping file, aggregating after reading suffice.

my unzipper looks this:

public class ziphandler extends abstractmessagehandler {  file dat; file json;  @override protected void handlemessageinternal(message<?> message) throws exception {     byte[] buffer = new byte[1024];     try {         file file = (file) message.getpayload();         zipfile zip = new zipfile(file);          (enumeration<? extends zipentry> entries = zip.entries(); entries                 .hasmoreelements();) {             zipentry ze = entries.nextelement();             string name = ze.getname();              if (name.endswith(".dat") || name.endswith(".dat")) {                 inputstream input = zip.getinputstream(ze);                 file datfile = new file("d:/lrtrans/zipout"                         + file.separator + name);                 fileoutputstream fos = new fileoutputstream(datfile);                 int len;                 while ((len = input.read(buffer)) > 0) {                     fos.write(buffer, 0, len);                 }                 this.dat = datfile;                 fos.close();             } else if (name.endswith(".json") || name.endswith(".json")) {                 inputstream input = zip.getinputstream(ze);                 file jsonfile = new file("d:/lrtrans/zipout"                         + file.separator + name);                 fileoutputstream fos = new fileoutputstream(jsonfile);                 int len;                 while ((len = input.read(buffer)) > 0) {                     fos.write(buffer, 0, len);                 }                 this.json = jsonfile;                 fos.close();             }         }         zip.close();     } catch (exception e) {         e.printstacktrace();     }  } } 

it takes files , puts them 2 directories, read them in again using filereadingmessagesource. solve using annotation based notation, not xml.

edit:

i want use defaultaggregatingmessagegroupprocssor correlationstrategy based on header called "zip" , releasestrategy based on message since, in case 2 files should combined one.

@aggregator(inputchannel = "toaggregatorchannel", outputchannel = "torouterchannel", discardchannel = "nullchannel") public defaultaggregatingmessagegroupprocessor aggregate(){     defaultaggregatingmessagegroupprocessor aggregator = new defaultaggregatingmessagegroupprocessor();      return aggregator; } @correlationstrategy  public string correlateby(@header("zipfile") string zip){     return zip; } @releasestrategy public boolean isreadytorelease(list<message<?>> messages) {     return messages.size() == 2; } 

i'd go right way. since zip has several files in it, correct requirement unzip , collect files 1 message , send further process.

so, yes, <aggregator> you. there need determine how correlate , group them.

don't know how unzip them, can use zip file name correlationkey , use number of files group size determine signal release group.

feel free ask more questions. first of need see "unzipper".

update

first of all, annotation-based aggregator configuration bit limited , better use @serviceactivator on aggregatingmessagehandler @bean have more control on options.

however, can achieve requirements choice. @aggregator configuration should follow pojo-method invocation principles:

@aggregator(inputchannel = "toaggregatorchannel", outputchannel = "torouterchannel", discardchannel = "nullchannel") public list<file> aggregate(list<file> files){     return files; } 

something that.