Issue1634405
Created on 2007-01-12.22:17:05 by leouserz, last changed 2007-03-09.17:26:39 by cgroves.
|
| msg2595 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:17:05 |
|
See:
http://sourceforge.net/tracker/index.php?func=detail&aid=1605847&group_id=12867&atid=112867
Ok here comes some patches.
|
| msg2596 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:17:58 |
|
File Added: jythonDiff.txt
|
| msg2597 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:18:32 |
|
File Added: BytecodeLoaderDiff.txt
|
| msg2598 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:19:02 |
|
File Added: PyDiff.txt
|
| msg2599 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:19:52 |
|
File Added: PyFunctionTableDiff.txt
|
| msg2600 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:20:31 |
|
File Added: impDiff.txt
|
| msg2601 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:21:43 |
|
org.python.compiler.PyFunctionTable patch
should be
org.python.code.PyFunctionTable patch
ok the patch changes the static initialiser to a instance method. This allows the system to provide a String file name to the module.
|
| msg2602 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:29:56 |
|
another possibility is to move all the initialise code into the constructor for the module. It would reduce the code down to one method and make the BytecodeLoader patch simpler.
|
| msg2603 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:49:45 |
|
yes, moving the code into the constructor works. Attached is the Module diff and following will be the ByteCodeloader diff 2.
File Added: ModuleDiff2.txt
|
| msg2604 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-12.22:50:45 |
|
File Added: BytecodeLoaderDiff2.txt
|
| msg2605 (view) |
Author: Nobody/Anonymous (nobody) |
Date: 2007-01-15.18:31:39 |
|
Logged In: NO
I assume this should have been submitted into http://jython.org/patches/. Same is true also for http://jython.org/bugs/1636030. But obviously a patch in a bit wrong place is much better than no patch. =)
|
| msg2606 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-15.18:37:14 |
|
well, I asked the best way to submit it and they told me to do it this way.
|
| msg2607 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-15.18:40:12 |
|
hmm, maybe that is the best home for it...
|
| msg2608 (view) |
Author: Charlie Groves (cgroves) |
Date: 2007-01-15.18:40:45 |
|
If you do a 'svn diff . > co_filename.patch' in the root of your jython checkout, you can get all of this in a single patch file instead of several. If you have changes in your checkout you don't want in the patch, just list the files you do want in the patch rather in place of the '.' in the command. It'll make this much more likely to be evaluated and applied if it's in a single patch file.
|
| msg2609 (view) |
Author: Deleted User leouserz (leouserz) |
Date: 2007-01-16.00:15:15 |
|
I believe the PyFunctionTable patch is not needed if the constructor version is used.
Charles,
Ill look into it. I hope this doesn't veto the patches for a possible evaluation. They really aren't that involved...
|
| msg2610 (view) |
Author: Charlie Groves (cgroves) |
Date: 2007-01-20.17:23:26 |
|
It doesn't veto it, it just means that it's much less likely that I'll look at it. The more you do to make your changes easily comprehensible and evaluated the better. A single patch file instead of 7 is in that direction. A single, concise message describing what you've done instead of a stream of consciousness narrative as you're working on something is in that direction.
Shrink this down to one patch and I'll happily evaluate it. If not, maybe someone else will.
|
| msg2611 (view) |
Author: Charlie Groves (cgroves) |
Date: 2007-03-09.17:26:39 |
|
Committed in r3138.
|
|
| Date |
User |
Action |
Args |
| 2007-01-12 22:17:05 | leouserz | create | |
|
