jython-bugs — Private bug notification list. Do not post to this list.

[Jython-bugs] [Bug #122800] __getslice__ bug

[<no...@so...>]

Bug #122800, was updated on 2000-Nov-18 10:54
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Wont Fix
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : nobody
Summary: __getslice__ bug

Details: __getitem__ has a bug at boundaries.

Example:
>>> "012345"[-100:-100:-1]
'0' # should be ''

The problem appears to be in getStop and(?) getStart in PySequence. These
functions 
should return values in [0,length] for positive steps and [-1,length-1] for
negative
steps. Currently, getStart always returns values in [0,length] and getStop 
returns values between [0, length] and [-1, length] for positive and negative
values of step respectively. I believe the above problem is due to getStart, and
I 
haven't found a problem related to getStop, but it probably wouldn't hurt to
fix
it too...

Follow-Ups:

Date: 2000-Dec-12 01:00
By: bckfnn

Comment:
I'm not sure what this example should return. When looking at NumPy, I get

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
>>> from Numeric import array
>>> a = array([0, 1,2,3,4,5,6,7,8])
>>> print a[::-1]
[8 7 6 5 4 3 2 1 0]
>>> print a[-100:-100:-1]
[0]

It seems like a huge negative start value is interpreted as starting at 0. Based on this, I'm going to close this bug. 

-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122800&group_id=12867

[Jython-bugs] [Bug #125440] Minor nit: "as" not recognised as valid function argument

[<no...@so...>]

Bug #125440, was updated on 2000-Dec-11 21:45
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: nickm
Assigned to : bckfnn
Summary: Minor nit: "as" not recognised as valid function argument

Details: Although python2.0 introduces "as" as a keyword, it allows it to appear where it would ordinarily be acceptable, as in:

def fn(as): as = as + 1; return as
fn(as=3)
class as: pass

This feature allows for backward compatibility with code written for earlier versions of python.  I actually have some such code

Jython (as in CVS as of 12 Dec 2000) does not allow this.  I believe that adding it would be as simple as replacing the declaration for Name
in python.jjt to allow <AS> as an alternative for <NAME>, but I haven't tried this out myself, due to unfamiliarity with the codebase.


Follow-Ups:

Date: 2000-Dec-12 00:36
By: bckfnn

Comment:
The suggested solution should work. I'll check in the patch.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=125440&group_id=12867

[Jython-bugs] [Bug #125440] Minor nit: "as" not recognised as valid function argument

[<no...@so...>]

Bug #125440, was updated on 2000-Dec-11 21:45
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: nickm
Assigned to : nobody
Summary: Minor nit: "as" not recognised as valid function argument

Details: Although python2.0 introduces "as" as a keyword, it allows it to appear where it would ordinarily be acceptable, as in:

def fn(as): as = as + 1; return as
fn(as=3)
class as: pass

This feature allows for backward compatibility with code written for earlier versions of python.  I actually have some such code

Jython (as in CVS as of 12 Dec 2000) does not allow this.  I believe that adding it would be as simple as replacing the declaration for Name
in python.jjt to allow <AS> as an alternative for <NAME>, but I haven't tried this out myself, due to unfamiliarity with the codebase.


For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=125440&group_id=12867

[Jython-bugs] [Bug #122852] re.RegexObject.groupindex

[<no...@so...>]

Bug #122852, was updated on 2000-Nov-18 11:33
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : nobody
Summary: re.RegexObject.groupindex

Details: Erronous handling of '(' for groupindex count:

Here the correct output of (C-)Python:
>>> re.compile(r'\\()(?P<x>)').groupindex
{'x': 2}
>>> re.compile(r'[()](?P<x>)').groupindex
{'x': 1}
>>> re.compile(r'[()](?P<x>)').groupindex
{'x': 1}
>>>

And here the buggy output of JPython:
>>> re.compile(r'\\()(?P<x>)').groupindex
{'x': 1}
>>> re.compile(r'[()](?P<x>)').groupindex
{'x': 2}
>>> re.compile(r'\\(?P<x>)').groupindex
Traceback (innermost last):
  File "<console>", line 1, in ?
re.error: Sequence (?P...) not recognized
>>>


Follow-Ups:

Date: 2000-Dec-11 10:50
By: bckfnn

Comment:
Closed when we switched to sre where the bug is solved.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122852&group_id=12867

[Jython-bugs] [Bug #122815] Bug in MatchObject.group(PyString)

[<no...@so...>]

Bug #122815, was updated on 2000-Nov-18 11:13
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : nobody
Summary: Bug in MatchObject.group(PyString)

Details: There appears to be a bug either in MatchObject.java or in the regular
expression package (I've only tested using a distribution WITH
OROMatcher so far).  For more complex regular expressions, it appears
that referencing a group in a match object by its name can cause an
IndexError.  Following is a code snippet that demonstrates the
problem.  I haven't been able to determine why exactly Case 3 works
whereas Case 1 fails inappropriately.  Case 2 seems to fail
appropriately.

I have tested under all combinations of jdk 1.1/jdk 1.2 and jpython
1.1beta2/1.1beta3.  The code snippet seems to behave as expected in
CPython.

-------------------------begin-----------------------------------
import re

# Regular expressions used for parsing

_S = '[ \t\r\n]+'                       # white space
_opS = '[ \t\r\n]*'                     # optional white space
_Name = '[a-zA-Z_:][-a-zA-Z0-9._:]*'    # valid XML name
_QStr = "(?:'[^']*'|\"[^\"]*\")"        # quoted XML string

attrfind = re.compile(
    _S + '(?P<name>' + _Name + ')'
    '(' + _opS + '=' + _opS +
    '(?P<value>'+_QStr+'|[-a-zA-Z0-9.:+*%?!()_#=~]+))?')
starttagend = re.compile(_opS + '(?P<slash>/?)>')


## Case 1
starttagmatch = re.compile('<(?P<tagname>'+_Name+')'
                           '(?P<attrs>(?:'+attrfind.pattern+')*)'+
                           starttagend.pattern)
m = starttagmatch.match('<foo>')
try:
    print '|%s|' % m.group('slash')
except IndexError, e:
    print e


## Case 2
starttagmatch = re.compile('<(?P<tagname>'+_Name+')'
                           '(?P<attrs>(?:'+attrfind.pattern+')*)')
m = starttagmatch.match('<foo>')
try:
    print '|%s|' % m.group('slash')
except IndexError, e:
    print e


## Case 3
r = re.compile('<(?P<tagname>' + _Name + ')'
               + _opS + '(?P<slash>/?)>')
m = r.match('<foo>')
print '|%s|' % m.group('slash')
---------------------------end-----------------------------------

JPython output:
---------------
group 7 is undefined
group 'slash' is undefined
||

CPython output:
---------------
||
group 'slash' is undefined
||


Follow-Ups:

Date: 2000-Dec-11 10:49
By: bckfnn

Comment:
Closed when we switched to sre where the bug is solved.
-------------------------------------------------------

Date: 2000-Nov-19 08:25
By: bckfnn

Comment:
This works when using the "sre" moduled instead of "re".
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122815&group_id=12867

[Jython-bugs] [Bug #122835] Should sys.prefix be added to sys.path?

[<no...@so...>]

Bug #122835, was updated on 2000-Nov-18 11:23
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : Nobody
Summary: Should sys.prefix be added to sys.path?

Details: In CPython sys.prefix (or sys.exec_prefix?) are added to sys.path. 
By default (it seems) 'distutils' will install modules in sys.prefix
which JPython can not find.


Follow-Ups:

Date: 2000-Dec-06 13:29
By: bckfnn

Comment:
The site.py which add sys.prefix to sys.path is not included in the Library installation package. This matches CPython completely.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122835&group_id=12867

[Jython-bugs] [Bug #122835] Should sys.prefix be added to sys.path?

[<no...@so...>]

Bug #122835, was updated on 2000-Nov-18 11:23
Here is a current snapshot of the bug.

Project: Jython
Category: None
Status: Open
Resolution: None
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : Nobody
Summary: Should sys.prefix be added to sys.path?

Details: In CPython sys.prefix (or sys.exec_prefix?) are added to sys.path. 
By default (it seems) 'distutils' will install modules in sys.prefix
which JPython can not find.


Follow-Ups:

Date: 2000-Dec-06 13:29
By: bckfnn

Comment:
The site.py which add sys.prefix to sys.path is not included in the Library installation package. This matches CPython completely.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122835&group_id=12867

[Jython-bugs] [Bug #122834] import * into locals

[<no...@so...>]

Bug #122834, was updated on 2000-Nov-18 11:23
Here is a current snapshot of the bug.

Project: Jython
Category: None
Status: Open
Resolution: None
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : Nobody
Summary: import * into locals

Details: A minor difference between CPython and JPython. Wildcard import into 
a local namespace does not make the imported names available.


def foo():
    from stat import *
    print locals().keys()
    print ST_MTIME

foo()


Follow-Ups:

Date: 2000-Dec-06 12:57
By: bckfnn

Comment:
Fixed in:
SimpleCompiler.py revision: 2.11;
LocalsCompiler.java revision: 2.5;
CodeCompiler.java revision: 2.8;

-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122834&group_id=12867

[Jython-bugs] [Bug #122834] import * into locals

[<no...@so...>]

Bug #122834, was updated on 2000-Nov-18 11:23
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : Nobody
Summary: import * into locals

Details: A minor difference between CPython and JPython. Wildcard import into 
a local namespace does not make the imported names available.


def foo():
    from stat import *
    print locals().keys()
    print ST_MTIME

foo()


Follow-Ups:

Date: 2000-Dec-06 12:57
By: bckfnn

Comment:
Fixed in:
SimpleCompiler.py revision: 2.11;
LocalsCompiler.java revision: 2.5;
CodeCompiler.java revision: 2.8;

-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122834&group_id=12867

[Jython-bugs] [Bug #122816] read () doesn't always read to EOF

[<no...@so...>]

Bug #122816, was updated on 2000-Nov-18 11:09
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 5
Submitted by: bckfnn
Assigned to : Nobody
Summary: read () doesn't always read to EOF

Details: In CPython, the read () method of file objects is stated to return the entire
contents of
the file.

When similar code is run under JPython, however, it frequently only seems to
return part of the
file.  It's as if it relies on the Java "read (buffer)" API which is not
guaranteed to actually read
the entire file, but never checks the results.

The best way to test this appears to be to try loading a reasonably long page
over a network connection,
as this is most likely to cause the JPython code to fail.

For Example:
  slashdotfile = urllib.urlopen ('http://www.slashdot.org/')
  page = slashdotfile.read ()
  print page

This will almost always only get about half the page under JPython.


Follow-Ups:

Date: 2000-Nov-26 07:20
By: bckfnn

Comment:
The original problem is solved, but using the new httplob from CPython2.0 causes a new bug. The socket gets closed right after reading the response with HTTP.getresponse(). The body will then be read through a dup'ed fd with sock.makefile(). 

Jython does not support the dupping. When the socket is closed, the sock.makefile file is close closed.
-------------------------------------------------------

Date: 2000-Dec-05 11:27
By: bckfnn

Comment:
Fixed in Lib/socket.py revision: 1.8.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122816&group_id=12867

[Jython-bugs] [Bug #122816] read () doesn't always read to EOF

[<no...@so...>]

Bug #122816, was updated on 2000-Nov-18 11:09
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: bckfnn
Assigned to : Nobody
Summary: read () doesn't always read to EOF

Details: In CPython, the read () method of file objects is stated to return the entire
contents of
the file.

When similar code is run under JPython, however, it frequently only seems to
return part of the
file.  It's as if it relies on the Java "read (buffer)" API which is not
guaranteed to actually read
the entire file, but never checks the results.

The best way to test this appears to be to try loading a reasonably long page
over a network connection,
as this is most likely to cause the JPython code to fail.

For Example:
  slashdotfile = urllib.urlopen ('http://www.slashdot.org/')
  page = slashdotfile.read ()
  print page

This will almost always only get about half the page under JPython.


Follow-Ups:

Date: 2000-Nov-26 07:20
By: bckfnn

Comment:
The original problem is solved, but using the new httplob from CPython2.0 causes a new bug. The socket gets closed right after reading the response with HTTP.getresponse(). The body will then be read through a dup'ed fd with sock.makefile(). 

Jython does not support the dupping. When the socket is closed, the sock.makefile file is close closed.
-------------------------------------------------------

Date: 2000-Dec-05 11:27
By: bckfnn

Comment:
Fixed in Lib/socket.py revision: 1.8.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=122816&group_id=12867

[Jython-bugs] [Bug #124028] Bad sys.stdin.read(x) in some circumstances

[<no...@so...>]

Bug #124028, was updated on 2000-Dec-01 03:17
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 5
Submitted by: zenzen
Assigned to : Nobody
Summary: Bad sys.stdin.read(x) in some circumstances

Details: Found this when trying to use pickle.

sys.stdin.read(1) sometimes reads in a lot, rather than one byte.

#!/bin/env jython
import sys
print sys.stdin.read(1)

% jython test.py < ~/tmp/fromc
S'Hi Mom'
p0
.

However, the same script works fine if data is entered by typing rather than redirection.


Follow-Ups:

Date: 2000-Dec-04 13:54
By: bckfnn

Comment:
Fixed PyFile.java rev 2.11.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=124028&group_id=12867

[Jython-bugs] [Bug #124028] Bad sys.stdin.read(x) in some circumstances

[<no...@so...>]

Bug #124028, was updated on 2000-Dec-01 03:17
Here is a current snapshot of the bug.

Project: Jython
Category: None
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: zenzen
Assigned to : Nobody
Summary: Bad sys.stdin.read(x) in some circumstances

Details: Found this when trying to use pickle.

sys.stdin.read(1) sometimes reads in a lot, rather than one byte.

#!/bin/env jython
import sys
print sys.stdin.read(1)

% jython test.py < ~/tmp/fromc
S'Hi Mom'
p0
.

However, the same script works fine if data is entered by typing rather than redirection.


Follow-Ups:

Date: 2000-Dec-04 13:54
By: bckfnn

Comment:
Fixed PyFile.java rev 2.11.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=124028&group_id=12867