One of the options I have when submiting the post to request the file is
to have it returned as a file or on screen. When I select to have it on
screen it all works correctly, when i select return a file curl never
gives me the file. I have check live httpheaders and --trace and they
match perfectly. I believe there is something wrong with how curls
dealing with

0000: Content-disposition: inline; filename=CaseFiledRPT.txt
0000: Transfer-Encoding: chunked
0000: Content-Type: application/binary; name=CaseFiledRPT.txt

I have now also tried with the most recent version of curl with the same
result.

bart wrote:

> When using a browser i am prompted to save a file, i say yes save the
> file then it displays the html below, i go look at the file i saved
> and its a tab delimited text file that i expected. The html that is
> returned by curl is not the file i receive when using a browser.
>
> Daniel Stenberg wrote:
>
>> On Wed, 18 May 2005, bart wrote:
>>
>>> So i use a procedure in the database to build the curl line it takes
>>> 4 posts to get logged in and to the point where I am requesting the
>>> file I want and requires probably way to much effort
>>
>>
>>
>>> here is the output of --trace-ascii
>>>
>>> => Send header, 576 bytes (0x240)
>>> 0000: POST /cgi-bin/CaseFiled-Rpt.pl?101551106276801-L_56_1-1 HTTP/1.1
>>
>>
>>
>>> 0000: Content-disposition: inline; filename=CaseFiledRPT.txt
>>> 0000: Transfer-Encoding: chunked
>>> 0000: Content-Type: application/binary; name=CaseFiledRPT.txt
>>
>>
>>
>> ...
>>
>>> 0000: 3ff
>>> 0005: <HR><CENTER><TABLE BORDER=1 BGCOLOR=WHITE WIDTH="400"><TR><TH CO
>>> 0045: LSPAN=4><FONT SIZE=+1 COLOR=DARKRED>blah blah
>>>
>>> As you can see I get the html, it mentions the file but I dont get
>>> the file. Any help or ideas would be greatly appreciated.
>>
>>
>>
>> That's the file you get. It even claims application/binary and a set
>> name.
>>