Recent changes to bugs

easy_install broken

a zwinkau — Sat, 02 Sep 2006 11:00:53 -0000

While easy_installing Pylons:

Searching for Myghty>=1.0.2
Reading http://www.python.org/pypi/Myghty/
Reading http://www.myghty.org
Reading http://www.myghty.org/links.myt?linkid=download
Reading http://www.python.org/pypi/Myghty/1.0.2
Best match: Myghty 1.0.2
Downloading
http://69.9.164.2/sourceforge/myghty/Myghty-1.0.2.tar.gz
error: Can't download
http://69.9.164.2/sourceforge/myghty/Myghty-1.0.2.tar.gz:
504 Gateway Time-out

test/Container.py: unittest fails

Geoffrey T. Dairiki — Thu, 27 Oct 2005 17:59:30 -0000

In test/Container.py, ContainerTest.testDbmContainer2
fails for me.

Details of failure are:

testDbmContainer2 (__main__.ContainerTest) ...
testing myghty.container.DBMContainer for 10 secs
with expiretime 2
delay 0
create thread 1 starting
create thread 2 starting
[...]
create thread 3 exiting
total object creates 2
total object gets 6788
FAIL

(Translation: The test script expects (approximately)
five object creates, but it only sees two.)

This is in the latest CVS version of test/Container.py
(file revision 1.13). The version from release 0.99
also fails, but in that case it is
ContainerTest.testcontainer which fails.

Too much white-space in subcomponent/method output.

Geoffrey T. Dairiki — Tue, 13 Sep 2005 17:47:21 -0000

This is more of an anoying behavior than a bug.

The compiler inserts (probably) unintended whitespace
in subcomponents and methods.

Here's an example:
<%flags>
inherit = None
</%flags>
<%def mycomp>
% m.write("Binary data")
</%def>
<& mycomp &>

which produces "\nBinary data" (note the leading "\n").

This behavior is particularly troublesome in
subcomenents/methods which are producing binary data
(say a GIF image).

It can be worked around using m.clear_buffer() (which,
btw, does not seem to be mentioned in the
documentation.) (Using the trim attribute is not a
safe workaround for subcomponents which produce binary
data, since the binary data may include white-space
characters.)

My guess as to the best fix is to make the lexer ignore
all white-space which is not contiguous with either
some non-white-space text, or a substitution or
component call.

r is not available within <%requestlocal> blocks

Geoffrey T. Dairiki — Fri, 09 Sep 2005 02:50:47 -0000

The global http request object, r, does not seem to be
available within <%requestlocal> blocks.

This is in constrast to the documentation, which
states: "The good news is, the regular Myghty variables
m, ARGS, r etc. are all available within a
request-scoped block."
(http://www.myghty.org/docs/scopedpython.myt#scopedpython_request)

Component.attributes is broken

Geoffrey T. Dairiki — Thu, 28 Jul 2005 03:13:59 -0000

The Component.attributes accessor as in (release 0.98a)
will only look up a maximum of one level in the
inheritance graph to find an attribute.

I've attached a patch.

For specificity, here's untested example code which
will exercise the bug:

===in /a.myt===
<%attr>
foo = 'fooval'
</%attr>

===in /b.myt===
<%flags>
inherit = '/a.myt'
</%flags>
<%attr>
bar = 'barval'
</%attr>

===in /c.myt===
<%flags>
inherit = '/b.myt'
</%flags>

# This is okay; it finds the 'bar' attribute from /b.myt:
<% m.base_component.attributes['bar'] %>

# This throws an AttributeError, doesn't find /a.myt's,
'foo':
<% m.base_component.attributes['foo'] %>

=============

session object does not resemble python mapping type

Anonymous — Sat, 23 Jul 2005 13:29:07 -0000

The session object as returned by m.get_session() does
not support basic python mapping type methods as
specified in http://www.python.org/doc/2.4.1/lib/
typesmapping.html .
Especially the methods get and setdefault would be
useful. Shouldn't be too hard to fix this, thanks.

Creating/Executing components from strings, via Interpreter

Philip Jenvey — Sat, 19 Mar 2005 01:37:11 -0000

I am attempting to compile a myghty template from a string. I'm
using the interpreter to generate the component from the string data
(via make_component), and then attemping to execute that
component via the interpreter (execute). If there is another way of
doing this, let me know (I noticed this is how Mason describes how
to do it ;) )

This is under 0.96f

The interpreter seems to lose my passed in parameter, possibly
during the the clone of the interp's request (the first thing execute
does)??

I've included a brief snippet of code that will reproduce this. It
prints out:

id: 137136108MissingArgument: required argument username not
found

file: test.py line 12

message = interp.execute(comp, username = 'pjenvey')

test.py:12

---------------------------------------------
Original Stack Trace:
MissingArgument: required argument username not found

file: /usr/local/lib/python2.3/site-packages/myghty/component.py line
110

raise exception.MissingArgument("required argument %s not found"
% arg)

/usr/local/lib/python2.3/site-packages/myghty/component.py:110
/usr/local/lib/python2.3/site-packages/myghty/request.py:389
/usr/local/lib/python2.3/site-packages/myghty/request.py:253
/usr/local/lib/python2.3/site-packages/myghty/exception.py:70
/usr/local/lib/python2.3/site-packages/myghty/request.py:285
/usr/local/lib/python2.3/site-packages/myghty/interp.py:129
test.py:12

Mod_python error: "PythonHandler myghty.ApacheHandler::handl

Anonymous — Thu, 30 Dec 2004 04:36:01 -0000

I have this error below with 0.96a,b but works fine
with 0.95x

Mod_python error: "PythonHandler
myghty.ApacheHandler::handle"

Traceback (most recent call last):

File
"D:\Python23\lib\site-packages\mod_python\apache.py",
line 287, in HandlerDispatch
log=debug)

File
"D:\Python23\lib\site-packages\mod_python\apache.py",
line 454, in import_module
f, p, d = imp.find_module(parts[i], path)

ImportError: No module named ApacheHandler

Any idea why?
Thanks!

one-line else:/except:/finally: statements glitchy

Mike Bayer — Wed, 01 Dec 2004 17:50:36 -0000

if you write code like this, using ':' lines with the
statement immediately following it:

% if foo: m.write('foo')
% else: m.write('no foo')

the object generator does the "if:" line correctly, but
mistakes the "else:" line for an "else:" all by itself,
and incorrectly decreases the indentation count.

dhandler decline() broken

Mike Bayer — Wed, 10 Nov 2004 05:12:10 -0000

release .95b has a class instantiation error in
exceptions.Declined, the dhandler_arg is not set up
correctly for declined components, calling decline() in
a root dhandler can create endless loops.