Make the generated code a little more pep8 compliant about xcffib

[tych0]

Interesting idea, thanks!

[spinus]

For some of my code (4 cores):

$time (for file in `find -iname '*py'`; do autopep8 --in-place --aggressive --aggressive $file& done; wait)
real    0m32.994s
user    1m56.732s
sys 0m3.244s

$ time autopep8 --in-place --aggressive --aggressive `find -iname '*py'`
real    0m55.356s
user    0m55.152s
sys 0m0.176s

[tych0]

On Wed, Aug 13, 2014 at 04:04:12PM -0700, Tomasz Czyż wrote:

For some of my code (4 cores):

$time (for file in `find -iname '*py'`; do autopep8 --in-place --aggressive --aggressive $file& done; wait)
real  0m32.994s
user  1m56.732s
sys   0m3.244s

$ time autopep8 --in-place --aggressive --aggressive `find -iname '*py'`
real  0m55.356s
user  0m55.152s
sys   0m0.176s

Wow, that is wild. Thanks for the tip! Any idea why they're different?

---

Reply to this email directly or view it on GitHub:
#24 (comment)

[spinus]

autopep8 cannot use multiple cores (yet?).
In this example I spawn one autopep8 process (autopep8 ... &) per python file (which in my case is a little crazy, ~240 files, probably some queueing or gnu parallel should be used here). And then wait is waiting for all those forks.

[tych0]

On Wed, Aug 13, 2014 at 7:33 PM, Tomasz Czyż [email protected]
wrote:

autopep8 cannot use multiple cores (yet?).
In this example I spawn one autopep8 process (autopep8 ... &) per python
file (which in my case is a little crazy, ~240 files, probably some
queueing or gnu parallel should be used here). And then wait is waiting
for all those forks.

Ah, of course, I totally missed that. In that case, you might be able to
save yourself some more time by not fork bombing the machine:

~/packages/xcffib master 2 time $(find ./xcffib/*.py | parallel -j $(grep
-c processor /proc/cpuinfo) autopep8 --aggressive --aggressive --in-place
'{}')

real 0m10.553s
user 0m24.760s
sys 0m0.968s
~/packages/xcffib master time (for file in find -iname '*py'; do autopep8
--in-place --aggressive --aggressive $file& done; wait)

real 0m12.749s
user 0m32.038s
sys 0m1.480s

—

Reply to this email directly or view it on GitHub
#24 (comment).

[spinus]

Very nice.