Traceback (most recent call last):
.....snakian stackfest...
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/adapters.py", line 37, in validate_ticket
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/api.py", line 78, in test
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/filters/regex.py", line 39, in test
TypeError: expected string or buffer
This is apparently some spam containment plug acting up but I can say that having this 7 times _AND_ losing your form's content
_AND_ being kicked back to login page afterwards is _very_ darn aggravating.
Traceback (most recent call last):
.....snakian stackfest...
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/
adapters.py", line 37, in validate_ticket
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/api.py",
line 78, in test
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/filters/
regex.py", line 39, in test
TypeError: expected string or buffer
This is apparently some spam containment plug acting up but I can say
that having this 7 times _AND_ losing your form's content
_AND_ being kicked back to login page afterwards is _very_ darn
aggravating.
Suck! I turned on the spam filter for registered users also to
eliminate future spam floods from registered users.
I tried replying, however, and it worked for me. Can anyone else confirm?
I also retrieved a stack trace when I tried to Submit Changes after
clicking the Reply button:
Traceback (most recent call last):
File "/opt/csw/lib/python2.3/site-packages/trac/web/main.py", line
356, in dispatch_request
dispatcher.dispatch(req)
File "/opt/csw/lib/python2.3/site-packages/trac/web/main.py", line
224, in dispatch
resp = chosen_handler.process_request(req)
File "/opt/csw/lib/python2.3/site-packages/trac/ticket/web_ui.py",
line 279, in process_request
self._do_save(req, db, ticket)
File "/opt/csw/lib/python2.3/site-packages/trac/ticket/web_ui.py",
line 536, in _do_save
self._validate_ticket(req, ticket)
File "/opt/csw/lib/python2.3/site-packages/trac/ticket/web_ui.py",
line 66, in _validate_ticket
for field, message in manipulator.validate_ticket(req, ticket):
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/adapters.py",
line 37, in validate_ticket
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/api.py",
line 78, in test
File "build/bdist.solaris-2.11-i86pc/egg/tracspamfilter/filters/regex.py",
line 39, in test
TypeError: expected string or buffer
I was the original reporter of the bug.
After reviewing the patch and re-reading my original bug report I
understand why it didn't work ...
I stored the information from the request to session['http_lang'] and
in the test I tried to retrieve it with session[:http_lang].
Now it makes me wonder if this was really a bug at all ... while the
patch will make it work (since it converts the argument to and =
to a string thus effectively changing my call to session[:http_lang]
to session['http_lang']) I can't remember if calling session[:symbol]
and session['symbol'] is supposed to be the same ...
Thank you people anyway now I can happily test my accept-language
dependant controllers
Do I have to close the bug myself once I agree with the resolution or
will it be closed by someone else ?
The official Guide To Testing The Rails _specifically_ mentions that of all testing hash containers only
the assigns are sensitive to string/symbol differences.
Therefore I filed it as a bug (especially knowing that the real Session objects are indifferent). You should also consider
that a bug like this might make some _auth tests_ that you have in your app pass while the app actually does the wrong thing
..inside action, up there
unless condition_that_does_not_fire_but_should
session[:user_pwd] = MD5.new(xxx_secret)
...down here in the test
assert_nil session['user_pwd'], 'Authentication should have been cleared from the session'
That's a bummer man (tm)
Maybe you can put the activity meter to some good use and disable this, uhm, measure for some folks
based on the patch count, for example.
