possible gsoap bug/corruption?

16 views
Skip to first unread message

hyuen

unread,
Jun 18, 2010, 5:51:27 PM6/18/10
to WS4D-gSOAP Development
Hi all,

I am trying to debug a problem, and on the way I found the following:
This is an excerpt from soap->buf

(gdb) p soap->buf
$13 = "HTTP/1.1 200 OK\r\nServer: gSOAP/2.7\r\nContent-Type: text/xml;
charset=utf-8\r\nContent-Length: 445\r\nConnection: close\r\n\r\n<?xml
version=\"1.0\" encoding=\"UTF-8\"?>\n<SOAP-ENV:Envelope xmlns:SOAP-
ENV=\"http://schemas.xmlsoap.org/soap/envelope/\" xmlns:SOAP-ENC=
\"http://schemas.xmlsoap.org/soap/encoding/\" xmlns:xsi=\"http://
www.w3.org/2001/XMLSchema-instance\" xmlns:xsd=\"http://www.w3.org/
2001/XMLSchema\" xmlns:pmd=\"urn:foo\"><SOAP-
ENV:Body><method:method1><response>6</response></
method:method1Response></SOAP-ENV:Body></SOAP-ENV:Envelope>nvelope>",
'\0'


Please note the ending, which has an extra 'nvelope>', is this normal?
Could this be a bug? Btw, I have changed the original request names to
method:method1 and so on, so maybe things like content-length will not
match

This is gsoap-2.7.10 running on linux


Thanks
Hector

S. Poehlsen

unread,
Jun 21, 2010, 4:39:47 AM6/21/10
to ws4d-gs...@googlegroups.com
Hi,

> Please note the ending, which has an extra 'nvelope>', is this normal?

I don't know, but why must the buffer be null terminated? In case of a
known buffer length it is not necessary to null terminate the buffer.

> Could this be a bug? Btw, I have changed the original request names to
> method:method1 and so on, so maybe things like content-length will not
> match

Maybe it is better to to track the problem without additionally changing
the method names. I don't know where you changed them, so it is quite
difficult to help you. Lock at the SENT.log and RECV.log to see what the
exchanged messages looks like.

Steph

Reply all
Reply to author
Forward
0 new messages