[jira] [Commented] (CMIS-906) doc.getContentStreamLength() returns the length of whole content even it is partial retrieved

["Florian Müller (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CMIS-906?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14486989#comment-14486989 ] 

Florian Müller commented on CMIS-906:
-------------------------------------

{{doc.getContentStreamlength()}} always returns the length of the document. It is independent of the content stream that you request.
I guess you are looking for  {{retrieved.getLength()}}. It returns the length of the (partial) content stream. 

> doc.getContentStreamLength() returns the length of whole content even it is partial retrieved
> ---------------------------------------------------------------------------------------------
>
>                 Key: CMIS-906
>                 URL: https://issues.apache.org/jira/browse/CMIS-906
>             Project: Chemistry
>          Issue Type: Bug
>          Components: opencmis-client-bindings
>    Affects Versions: OpenCMIS 0.13.0
>         Environment: WIN
>            Reporter: James Li
>            Priority: Minor
>
> I have a test to just retrieve partial content:
> {code:java}
> public void getContentRange() throws IOException{
> 	    byte[] content = "Hello World, this is for contentstream range test!".getBytes();  
> 	    InputStream stream = new ByteArrayInputStream(content);
>         ContentStream contentStream = session.getObjectFactory().createContentStream("hello.txt", content.length, "text/plain", stream);
> 	    
>      // (minimal set: name and object type id)
>         Map<String, Object> properties = new HashMap<String, Object>();
>         properties.put(PropertyIds.OBJECT_TYPE_ID, BaseTypeId.CMIS_DOCUMENT.value());
>         properties.put(PropertyIds.NAME, "hello");
>         
>         Document doc = this.scratchFolder.createDocument(properties, contentStream, VersioningState.MAJOR);
>         
>         ContentStream retrieved =  doc.getContentStream("0", BigInteger.valueOf(5), BigInteger.valueOf(5));
>         System.out.println("streamlength: " + doc.getContentStreamLength());
> 	}
> {code}
> The test only retrieve from position 5, length 5 of contentstream, but the {{doc.getContentStreamlength()}} actually return 50 (the length of whole ocntent).
> On server side implmentation, we actually set the length to 5:
> ContentStream [filename=hello.txt, length=5, MIME type=text/plain, has stream=true][extensions=null]
> I traced the client.bindings.spi.atompub.ObjectServiceImple, the response header is:
> {noformat}
> {cache-control=[private, max-age=0], content-type=[text/plain], null=[HTTP/1.1 206 Partial Content], transfer-encoding=[chunked], content-disposition=[attachment; filename=hello.txt], content-language=[en-US], server=[Apache-Chemistry-OpenCMIS/1.0.0-SNAPSHOT], date=[Thu, 09 Apr 2015 06:43:24 GMT], content-range=[bytes 5-9/*]}
> {noformat}
> The code set the length using {{result.setLength(resp.getContentLength());}}
> Is this a bug? 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)