Hi guys,
Got a weird one for you today. When entering non-ASCII characters in the
Post Body section of the Http Sampler, the structure of the test plan is
corrupted. The GUI fails to update and I've found out the hard way: when
quick saving (I've looked in the saved jmx, I think, the structure is not
proper). Used the following test char-group to reproduce this one:
!#$%_?§ÄäÖöÜüßÁÀÂÃÇÉÈÊÍÌÎÏÑÓÒÔÕÚÙÛÝáàâãçéèêíìîïñóòôõúùûýÅÆÐËØßþåæðëøµÞ£ƒ®©¥¿
--------------------------------------
My plan looks like:
+ thread group
++ http request [gui configs: Method = POST / Content encoding = utf-8
/ use "Post Body" to add raw post data of multi-part type]
+++ Pre-processors / Counters / Http Header Manager (defines custom
multi-part content type & boundary) / Assertions / Timer / Regex / Xpath
checker
* the last hierarchic level doesn't affect reproduction of issue
This is needed because some basics tweaks of the post data must be made
before making the requests.
After saving the plan, the jmx seems to be messed up, I think the nodes are
broken or not properly formatted.
Other details: ubuntu,
----------------------------------------
Two problems:
1. The raw field should work with any characters from the encoding
specified. Shouldn't it?
- I thought of a quick fix here, to paste the URI-encoded version of
the string
2. This issue messed up some changes that are rather important. Do you
have any suggestion on what tools to use to repair an XML? :)
I've tried removing the content of the Raw Post Body or the entire sampler
that caused the problem, but the first doesn't solve the problem and the
second only messes the test plan more.
Regards,
Adrian
