Handling 'chunked' encoded response with HTTPBuilder/Groovy

Question

I'm new to using groovy and have started to use it to test some REST services. I'm having an issue parsing my XML response from our service due to 'Content not allowed in prolog.' After awhile searching I came across a post saying there might be a Byte Order Marker at the beginning. To compensate I followed their approach to trim the characters before the first < and then parse the response. While this works, I was also told the issue is that the response is coming back as 'Transfer-Encoding: chunked'.

Using HTTPBuilder, is there a way to handle chunked responses without trimming characters off? If I try:

def http = new HTTPBuilder('url')
http.request( Method.valueOf("GET"), XML )

I get the 'Content not allowed in prolog message. However:

http.request( Method.valueOf("GET"), TEXT )

Works, but requires trimming the text until the first < before sending the response to XmlParser.

Unfortunately, your issues are neither clearly expressed nor *reproducible* (by me). You'd want to post a [reproducible unit test case](http://svn.jremoting.codehaus.org/browse/~raw,r=831/gmod/httpbuilder/trunk/src/test/groovy/groovyx/net/http/HTTPBuilderTest.groovy). — robbbert, Nov 18 '10 at 16:39
I think I am having a similar problem : http://stackoverflow.com/questions/8760713/groovy-httpbuilder-getting-the-entity-content-from-a-gziped-chunked-response — serty2, Jan 06 '12 at 16:07

score 0 · Accepted Answer · answered Oct 04 '12 at 12:56

0

I had the same issue when I needed to interact with an IIS server. The XML returned had a bogus character in front of the actual XML returned by the web server. I worked around it like this:

StringReader reader = builder.get( path: 'rcserver/systeminfo.xml', contentType: ContentType.TEXT )
def text = reader.getText()
def xml = new XmlSlurper().parseText(text.substring(1));

answered Oct 04 '12 at 12:56

Wim Deblauwe

25,113
20
133
211

This was a long time ago, but that's what I ended up doing as well! – Caleb S Feb 14 '13 at 02:14

score 0 · Answer 2 · answered Mar 21 '13 at 02:07

I was having this problem as well hitting an IIS server over https. Here is a little addition to Wim Deblauwe's answer for a POST request. You have to send a different type in the request than you expect in the response.

Send a POST with XML as the request type and TEXT as the response type. Then, parse the text response into XML. This worked for me.

In Groovy:

def reader = http.request(Method.POST, ContentType.TEXT){
    uri.path = "myPath.api"
    send ContentType.XML, postBodyXml
}
def text = reader.getText()
def resultxml = new XmlSlurper().parseText(text.substring(1));

score 0 · Answer 3 · answered Feb 20 '12 at 17:05

0

The HTTPBuilder class has a setContentEncoding() method that allows you to specify the response's content-type.

Maybe something like:

http.contentEncoding = ContentEncoding.Type.GZIP
http.request( Method.GET, XML)

Hope this helps.

answered Feb 20 '12 at 17:05

Daniel Woods

1,029
7
10

I also have this issue and that does not help. – Wim Deblauwe Oct 04 '12 at 12:54

Handling 'chunked' encoded response with HTTPBuilder/Groovy

3 Answers3