CLN: Fix CRLFs in repo + add .gitattributes #3915

jtratner · 2013-06-15T15:28:08Z

View this pull request with ?w=t added to the end of the url to see that there are not differences [aside from the .gitattributes files] - this should prevent more commits with CRLF entering the repo.

I think it'd be good to merge this (especially because we can avoid this happening in the future.), though (I think) it'll mess up the history for 10 minutes to pandas and pandas/core/expressions.py.

After this, will never need to deal with CRLF again! Yay!

jtratner · 2013-06-15T15:33:23Z

Looks like it's not a huge deal in re core/expressions - it's all @jreback and looks like only 3 commits total https://github.com/jtratner/pandas/blame/8bde194aef964ff4294271bbf9da587f144ab9bd/pandas/core/expressions.py . More of a loss in 10 minutes to pandas section - https://github.com/pydata/pandas/blame/d1556237eee69aaeee47a4d1014f782413ad2d7f/doc/source/10min.rst

jreback · 2013-06-15T15:52:22Z

yep that was before changed my git settings to not update line endings
as I edit on windows

jtratner · 2013-06-15T17:47:26Z

@jreback Well, this would prevent this from occurring with anyone else's contributions.

cpcloud · 2013-06-15T18:51:56Z

👍 on this because this gets annoying sometimes when a script barfs the oh-so-cryptic file not found error and of course there's no way to realize this until git or [text-editor-here] shows u ^M everywher

cpcloud · 2013-06-16T23:36:24Z

did u do a doc build?

jtratner · 2013-06-16T23:56:03Z

@cpcloud ? do you mean check if this builds docs okay?

cpcloud · 2013-06-16T23:56:35Z

yep sorry didn't mean to be so terse

cpcloud · 2013-06-16T23:56:44Z

i can do it if u want

jtratner · 2013-06-17T00:01:49Z

it's no problem...made me install pytables and hdf5 anyways :) took me a
while to realize that pytables is pip install tables. Anyways it'll be a
while. Def should revisit the travis doc build.

On Sun, Jun 16, 2013 at 7:56 PM, Phillip Cloud [email protected]:

i can do it if u want

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19521496
.

cpcloud · 2013-06-17T00:04:36Z

there's a json error but i'm not sure if that's from here it might be fixed after u rebase. try it and c

cpcloud · 2013-06-17T00:04:48Z

well not error but "invalid" sphinx

cpcloud · 2013-06-17T00:07:51Z

oh what is the ?w=t that is kind of cool. where did u find out about that?

jreback · 2013-06-17T00:08:07Z

FYI I think the doc string is screwed up somehow in read_json which is causing that sphix error
never could format them correctly

jtratner · 2013-06-17T00:09:33Z

@cpcloud I was trying to figure out why I was getting CRLF errors all over the place, so I explored and found that somewhere.

Okay, I need to redo this off of master (going to keep it on the same branch name though). Give me a minute.

jtratner · 2013-06-17T00:33:15Z

Which file is that? I can probably fix it.

On Sun, Jun 16, 2013 at 8:08 PM, jreback [email protected] wrote:

FYI I think the doc string is screwed up somehow in read_json which is
causing that sphix error
never could format them correctly

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19521636
.

jreback · 2013-06-17T00:44:39Z

iirc pandas.io.json

ghost · 2013-06-17T00:50:21Z

word of caution, the 11.0 release raised issues with pickle files being detected as text
by git autocrlf, should probably flag .pickle as binary in .gitattributes as they are binary
files with large chunks of embedded text.

xref #3360

cpcloud · 2013-06-17T01:05:50Z

problem in json docstring is nested lists, not sure exactly where but if u nest u need to separate each node from its children by a newline

jtratner · 2013-06-17T01:15:53Z

@y-p added that and rebased.

On Sun, Jun 16, 2013 at 9:06 PM, Phillip Cloud [email protected]:

problem in json is nested lists, not sure exactly where but if u nest u
need to separate each node from its children by a newline

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19522456
.

jtratner · 2013-06-17T01:47:33Z

@cpcloud not going to try to fix this for this commit, given that the issue
isn't caused by this change (right?) Plus, I'm having a difficult time
trying to build the docs and then find that error...

On Sun, Jun 16, 2013 at 9:15 PM, Jeffrey Tratner
[email protected]:

@y-p added that and rebased.

On Sun, Jun 16, 2013 at 9:06 PM, Phillip Cloud [email protected]:

problem in json is nested lists, not sure exactly where but if u nest u
need to separate each node from its children by a newline

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19522456
.

cpcloud · 2013-06-17T02:38:40Z

i fixed it already. was just merged #3930

cpcloud · 2013-06-17T02:42:23Z

@jtratner can u rebase? (just did a doc build and it was fine). will merge after the rebase...

cpcloud · 2013-06-17T02:43:45Z

btw i just saw the blames at the top i had no idea about this feature of gh. kind of cool

jtratner · 2013-06-17T03:01:49Z

@cpcloud that's a git feature too (github makes it look nicer though),
checkout git blame. I rebased it, let's try merging it :)

On Sun, Jun 16, 2013 at 10:43 PM, Phillip Cloud [email protected]:

btw i just saw the blames at the top i had no idea about this feature of
gh. kind of cool

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19524151
.

cpcloud · 2013-06-17T03:02:36Z

i like git blame but it's a bit ugly. i'm sure there are ways to make it pretty but i don't use it that often...

cpcloud · 2013-06-17T03:02:59Z

any objections to merging this?

cpcloud · 2013-06-17T03:11:51Z

@jtratner sorry 2 be a pain but what about adding lines in .gitattributes for stata, h5, xls(x) (all binary)

cpcloud · 2013-06-17T03:13:56Z

so that would be

+*.dta binary
+*.xls binary
+*.xlsx binary # this is xml  but god help u if ur looking at the guts of xlsx
+*.h5 binary

i think

jtratner · 2013-06-17T03:14:24Z

stata is .dta?

jtratner · 2013-06-17T03:16:53Z

Okay, that's done. Not going to squash b/c it's time consuming.

cpcloud · 2013-06-17T03:21:55Z

should be ok not to rebase, just 3 commits here.

cpcloud · 2013-06-17T13:35:00Z

@jtratner can u rebase? this is good 2 go after that. @jreback @y-p anything add before merging after rebase?

jtratner · 2013-06-17T14:26:06Z

I'll do it after work tonight.
On Jun 17, 2013 9:35 AM, "Phillip Cloud" [email protected] wrote:

@jtratner https://github.com/jtratner can u rebase? this is good 2 go
after that. @jreback https://github.com/jreback @y-phttps://github.com/y-panything add before merging after rebase?

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3915#issuecomment-19544972
.

jtratner · 2013-06-17T22:55:50Z

@cpcloud @jreback ready to be merged.

cpcloud · 2013-06-17T23:44:21Z

looks good 2 me.

jreback · 2013-06-17T23:45:55Z

ok here too....

CLN: Fix CRLFs in repo + add .gitattributes

jreback · 2013-06-17T23:46:26Z

thank you @jtratner

jtratner · 2013-06-18T02:20:10Z

@jreback @cpcloud @y-p et al - if you have issues rebasing on to master after this because of the line endings (should really only come into play if you were changing core/expressions.py or one of the other files changed here, you can accomplish the merge/rebase automatically by using

git rebase <branch> -s recursive -X renormalize

and this will make git ignore the differences in line endings. Thanks to this SO answer for the solution!

cpcloud · 2013-06-18T02:21:53Z

that is a nice solution, maybe i will webscrape pandas github repo using pandas' own read_json when i finally write up the upcoming dev workflows section of the pandas website...whoohoo! so much raw power in git

jtratner closed this Jun 17, 2013

jtratner reopened this Jun 17, 2013

jtratner added 3 commits June 17, 2013 18:54

CLN: Add .gitattribtes to fix crlf issues.

ffebe97

CLN: Change CRLF --> LF

51c6799

BLD: Add binary files to gitattributes too

9bb9f57

jreback added a commit that referenced this pull request Jun 17, 2013

Merge pull request #3915 from jtratner/fix-crlf-in-repo

7ea6bdc

CLN: Fix CRLFs in repo + add .gitattributes

jreback merged commit 7ea6bdc into pandas-dev:master Jun 17, 2013

jtratner deleted the fix-crlf-in-repo branch June 18, 2013 02:06

CLN: Fix CRLFs in repo + add .gitattributes #3915

CLN: Fix CRLFs in repo + add .gitattributes #3915

Conversation

jtratner commented Jun 15, 2013

jtratner commented Jun 15, 2013

jreback commented Jun 15, 2013

jtratner commented Jun 15, 2013

cpcloud commented Jun 15, 2013

cpcloud commented Jun 16, 2013

jtratner commented Jun 16, 2013

cpcloud commented Jun 16, 2013

cpcloud commented Jun 16, 2013

jtratner commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jreback commented Jun 17, 2013

jtratner commented Jun 17, 2013

jtratner commented Jun 17, 2013

jreback commented Jun 17, 2013

ghost commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jtratner commented Jun 17, 2013

jtratner commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jtratner commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jtratner commented Jun 17, 2013

jtratner commented Jun 17, 2013

cpcloud commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jtratner commented Jun 17, 2013

jtratner commented Jun 17, 2013

cpcloud commented Jun 17, 2013

jreback commented Jun 17, 2013

jreback commented Jun 17, 2013

jtratner commented Jun 18, 2013

cpcloud commented Jun 18, 2013