Issue1896
Created on 2012-05-25.06:32:49 by alex.gronholm, last changed 2014-05-22.00:13:57 by zyasoft.
msg7146 (view) |
Author: Alex Grönholm (alex.gronholm) |
Date: 2012-05-25.06:32:49 |
|
This is required by the compileall module, which I need to make binary jars so I can test my ERP client on 2.7.
|
msg7147 (view) |
Author: Alex Grönholm (alex.gronholm) |
Date: 2012-05-25.12:24:50 |
|
Renamed the issue.
I also committed the missing imp.get_magic() function, but I think I was too hasty. For one, the value should be unique to Jython. Second, there may be a good reason why the function is missing from Jython 2.5. It's highly likely that my commit needs to be reverted. However, I have no clue how to fix the compileall module then.
|
msg7148 (view) |
Author: Alex Grönholm (alex.gronholm) |
Date: 2012-05-25.12:29:20 |
|
On further investigation I discovered that in 2.7, the compileall module reads the first 8 bytes of an existing target file and checks the magic number and the modification time. Python 2.5 does not do this. I don't know enough about class files to know if or how this applies to Jython.
|
msg7203 (view) |
Author: Dietrich Schulten (dschulten) |
Date: 2012-06-12.05:24:54 |
|
compileall works when run like this within the jython27a2 REPL:
import compileall
compileall.compile_dir("src", force=1)
but it fails when run as a subprocess:
subprocess.call(['java', '-jar', jython_jar, '-m', 'compileall', src])
|
msg7211 (view) |
Author: Dietrich Schulten (dschulten) |
Date: 2012-06-13.05:53:39 |
|
The reason is of course the force=1, it prevents compileall from trying to read in the magic number.
Therefore, -f will work, too:
subprocess.call(['java', '-jar', jython_jar, '-m', 'compileall', '-f', tmpdir])
|
msg8516 (view) |
Author: Jim Baker (zyasoft) |
Date: 2014-05-22.00:13:55 |
|
Works fine now, in the ways I would assume compileall is actually used - by setuptools, pip, etc, etc - basically all the stuff we have been working on for a while. Or jython27 -m compileall
I assume this bug has just been sitting here unnoticed until now, certainly by me. Please respond if this is not the case.
|
|
Date |
User |
Action |
Args |
2014-05-22 00:13:57 | zyasoft | set | status: open -> closed resolution: fixed messages:
+ msg8516 |
2013-05-21 21:15:26 | zyasoft | set | assignee: fwierzbicki -> zyasoft nosy:
+ zyasoft |
2013-02-20 00:31:37 | fwierzbicki | set | priority: normal -> urgent |
2013-02-20 00:28:57 | fwierzbicki | set | versions:
+ Jython 2.7, - 2.7a2 |
2012-06-13 05:53:40 | dschulten | set | messages:
+ msg7211 |
2012-06-12 16:29:16 | Arfrever | set | nosy:
+ Arfrever |
2012-06-12 05:24:55 | dschulten | set | messages:
+ msg7203 |
2012-06-12 05:21:13 | dschulten | set | versions:
+ 2.7a2, - 2.7a1 |
2012-06-12 04:38:35 | dschulten | set | nosy:
+ dschulten |
2012-05-25 12:29:20 | alex.gronholm | set | messages:
+ msg7148 |
2012-05-25 12:24:50 | alex.gronholm | set | messages:
+ msg7147 title: get_magic() missing from the "imp" module -> compileall does not work |
2012-05-25 06:32:49 | alex.gronholm | create | |
|