Output Binary data to browser as file

Question

I know this has been covered a thousand time, but every
example I am given does not work.
I am accessing a Java WebService to get a document (in this
case PDF). This WS has no WDSL, so I am hitting it through CFHTTP.
The WS returns me a hashmap with error codes etc. One element is a
Binary string.
I have done everything imaginable and the most I can do is
get the raw text to display. Unfortunately, saving a temp file then
using CFFILE is not an option for security reasons; however, if I
try this the file is corrupted and will not open. When I use
CFHEADER I get a complaint about java.lang.String. However,
ToBinary gives me a CF error with garbage text and no real error,
and BinaryDecode always whines about input and output encodings not
matching.
This is what I use the call the service:
<CFHTTP URL="HIDDEN" METHOD="GET" RESULT="testFN"
CHARSET="ISO-8859-1"></CFHTTP>
And these are somethings I have tried:
<cfheader name="Content-Type" value="pdf">
<CFHEADER NAME="Content-Disposition" VALUE="inline;
filename=my.pdf">
<CFCONTENT TYPE="application/pdf"
VARIABLE="#BinaryDecode(testFN.Filecontent, "Base64")#">
<cfscript>
context = getPageContext();
context.setFlushOutput(false);
response = context.getResponse().getResponse();
out = response.getOutputStream();
response.setContentType("application/pdf");
response.setContentLength(arrayLen(testFN.Filecontent));
out.write(testFN.Filecontent);
out.flush();
out.close();
</cfscript>
<cfcontent type="application/pdf; charset=ISO-8859-1">
<CFSCRIPT>
writeOutput(toString(testFN.Filecontent));
</cfscript>
None work. I either get a blank page, a CF error, or plain
text. Other Java application use this same WS without issue, so any
complaints about the WS have been ignored.
I need help badly. I am not sure what else to try.

Answer 1

Thanks for the reply.
I have had problems using BinaryDecode. It states the input
and output are not the same not matter what I use as the
binaryencoding. I can get a pdf binary object sent to the browser
by using this:
<CFHEADER NAME="Content-Disposition" VALUE="attachment;
filename=#fileNAME#.#fileEXT#">
<CFCONTENT TYPE="*/*"
VARIABLE="#ToBinary(ToBase64(docBINARY))#">
ToBinary nor ToBase64 work on their own (I get an error about
ByteArrays), but they seem to work together. However, now I get a
12 page PDF with 1 blank page and all the other pages generate
errors in the Adobe Reader. I have confirmed that the document
comes up fine in the Java applications that do basically the same
thing, but it is like ColdFusion does not understand the data type.
Unfortunately, our DEV and UAT systems are not exposed for me
to give you access. The Web Service I am consuming will never be
available externally, which is why I am building a CF passthrough
to display the documents to public users. The internal system will
output private contracts, so there are privacy issues related to
opening it up. Our prod deployment is very strict, so I cannot even
put up public pages for testing until all involved sign off on
functionality.
Obviously, the CFDUMP exceeds the text limit, so I cannot
post it either. :(
Edit: Also, I found out that the first byte of object is a
status code. I have parsed that off (why you now see docBINARY),
but I only see this as a square and have not been able to convert
this to the number code I should be getting.