Default to utf-8 for csv opening. by julienmalard · Pull Request #5712 · pypa/pip

julienmalard · 2018-08-17T16:54:06Z

This solves problems when installing packages with unicode module or package names on Windows and resolves this issue originally reported to setuptools:
pypa/setuptools#1399

benoit-pierre · 2018-08-17T17:11:53Z

encoding is not used... You really need to add a test for this.

Thanks! I had copied over the changes from my local copy of pip.
Would you have any guidance as to where/how to add a test for this case? The error occurs when installing wheels which have unicode module names.

Create a minimal project that triggers the bug, add its sources to tests/data/src/ and a resulting wheel to tests/data/packages/, then add a test similar to test_basic_install_from_wheel in tests/functional/test_install_wheel.py. Do that without your changes, run your new test with something like tox -- -k test_install_wheel_with_unicode_filenames, and make sure you can reproduce the failure, then fix the test by adding your changes. Check Python 2 still work, and push.

You also need to add a news entry: https://pip.pypa.io/en/stable/development/#adding-a-news-entry

And encoding is not a valid open keyword argument in Python 2.

Thanks for the guidance! I will work on this and get back to you.

Hello,
I just made some changes for open and added a test. Will see how the pull request works now (I am not working from a Windows machine at the moment).
Could you help with the wheel test please? I could not figure out how to point a new wheel test to the new unicode_file example in the data folder.
Also, mypy seems to be failing but I am not sure why.

Thanks!

Hello,
Any help please?
Many thanks for your support!

benoit-pierre · 2018-08-17T17:16:58Z

Again, you really need to add a test: what about Python 2 support (the file is opened in binary mode)?

pradyunsg

As noted by @benoit-pierre

BrownTruck · 2018-12-16T19:30:01Z

Hello!

I am an automated bot and I have noticed that this pull request is not currently able to be merged. If you are able to either merge the master branch into this pull request or rebase this pull request against master then it will be eligible for code review and hopefully merging!

pradyunsg · 2019-01-02T10:20:15Z

I think that merge went wrong @julienmalard. :/

julienmalard · 2019-01-02T20:25:55Z

Oops...
I guess could simply clone a new branch from the main repo but then would have to reopen a new pull request and close this one.
Would you happen to know how I could fix it instead? (I suppose from the terminal?)

Thanks and sorry,
Julien

pradyunsg · 2019-01-03T07:27:21Z

This should be a simple rebase. You'll need to fix a merge conflict, for updating the type signature for open_for_csv to include the encoding parameter type.

There's instructions for this at https://pip.pypa.io/en/latest/development/contributing/#updating-your-branch, but if they're not super helpful, please feel free to ping me and I'll be happy to provide step-by-step guidance for the rebase to you. :)

julienmalard · 2019-01-08T22:08:55Z

Hello, Would it be possible to provide step-by-step guidance given where I am right now (i.e., with a merge already gone wrong?) Thanks a lot!

…

________________________________ દ્વારા: Pradyun Gedam <notifications@github.com> મોકલ્યું: 03 જાન્યુઆરી 2019 02:27:28 પ્રતિ: pypa/pip Cc: Julien Malard; Mention વિષય: Re: [pypa/pip] Default to utf-8 for csv opening. (#5712) This should be a simple rebase. You'll need to fix a merge conflict, for updating the type signature for open_for_csv to include the encoding parameter type. There's instructions for this at https://pip.pypa.io/en/latest/development/contributing/#updating-your-branch, but if they're not super helpful, please feel free to ping me and I'll be happy to provide step-by-step guidance for the rebase to you. :) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#5712 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AKMb68pD906SgRS6BR61ZH4TDBcFnM0Fks5u_bDggaJpZM4WB4XV>.

pradyunsg · 2019-01-09T08:29:44Z

Sure!

This is a simple rebase, you have to fix one merge conflict and ignore one patch manually.

To start the rebase, run:
```
$ git rebase master
```
The rebase will stop due to a merge conflict, around line 83 in pip._internal.wheel, change that section of code to be:
```
def open_for_csv(name, mode, encoding='utf8'):
    # type: (str, Text, str) -> IO
```
Then, run:
```
$ git add src/pip/_internal/wheel.py
```
This will add this resolved merge conflict to git's "staging area", so we can proceed now.
To continue the rebase, you will run:
```
$ git rebase --continue
```
After that, it will complain about another commit. This time, you want to ignore it since it's applied. You should just run:
```
$ git rebase --skip
```
That will proceed and complete the rebase.
Now you can do a git push --force (or however you usually push, with --force flag added) to push the change to your repository's branch, which will update this PR. :)

Feel free to ping me in case something is not clear.

This solves problems when installing packages with unicode module or package names.

julienmalard · 2019-07-14T11:02:14Z

Hello,
Any news? (See my comment above asking for clarification on how to implement some of the changes you had requested.)
Many thanks for your help!

xavfernandez · 2019-07-18T08:54:11Z

Many thanks for your help!

From this PR, it is unclear what you are trying to fix.

As asked in pypa/setuptools#1399 it would be helpful to produce a minimal example failing in your case.

Your tests/data/src/unicode_files/ doesn't seem to be it since it fails with:

$ pip install  tests/data/src/unicode_files/
Looking in indexes: https://pypi.org/simple
Processing ./tests/data/src/unicode_files
    Complete output from command python setup.py egg_info:
    running egg_info
    error: Invalid distribution name or version syntax: --0.0.1
    
    ----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /tmp/pip-req-build-a7fdv7vd/

which does not seem to fail on a csv.

julienmalard · 2019-07-18T09:01:18Z

Thanks for the quick response! Will investigate and get back to you soon.

pradyunsg · 2019-07-15T05:19:05Z


-def open_for_csv(name, mode):
+def open_for_csv(name, mode, encoding='utf8'):
    # type: (str, Text) -> IO


# type: (str, Text, str)

chrahunt · 2019-09-17T23:20:32Z

Since it has been some time without progress here I will close this, but anyone should feel free to re-raise this as a PR or issue if the information above can be provided.

Thanks!

julienmalard · 2019-10-02T13:29:09Z

Hello,

Any chance this could be reopened? I am trying to fix this PR and it would be helpful to see if the tests now pass.

Thanks,
Julien

chrahunt · 2019-10-02T23:13:36Z

GitHub won't allow reopening at this time. The hover message on the disabled "Reopen and comment" button states: "The master branch was force-pushed or recreated."

I would try creating a new branch off of your current commit, fetch and overwrite the master from the main pypa/pip repository, and then rebase your new branch on your local master. It would look something like (assuming you have the main pypa/pip repo set as a remote called upstream):

git checkout -b my-fixes
git fetch-f upstream master:master
git rebase master

Then you can push my-fixes back to your GitHub repository and submit a new PR, or git branch --move master my-fixes then git push -f origin master to update this PR.

julienmalard · 2019-10-04T19:07:16Z

Done, thank you! I continued it in a new pull request (#7138).

benoit-pierre reviewed Aug 17, 2018

View reviewed changes

pradyunsg requested changes Aug 26, 2018

View reviewed changes

BrownTruck added the needs rebase or merge PR has conflicts with current master label Dec 16, 2018

pypa-bot removed the needs rebase or merge PR has conflicts with current master label Jan 1, 2019

julienmalard added 3 commits January 10, 2019 13:17

Default to utf-8 for csv opening.

cfca7a8

This solves problems when installing packages with unicode module or package names.

Added tests

40ee62e

News item

2f660e2

pradyunsg reviewed Jul 18, 2019

View reviewed changes

chrahunt added the S: awaiting response Waiting for a response/more information label Aug 16, 2019

chrahunt closed this Sep 17, 2019

julienmalard mentioned this pull request Oct 4, 2019

Fix unicode isses in pip install #7138

Closed

lock Bot added the auto-locked Outdated issues that have been locked by automation label Nov 13, 2019

lock Bot locked as resolved and limited conversation to collaborators Nov 13, 2019

Conversation

julienmalard commented Aug 17, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pradyunsg left a comment

Choose a reason for hiding this comment

Uh oh!

BrownTruck commented Dec 16, 2018

Uh oh!

pradyunsg commented Jan 2, 2019

Uh oh!

julienmalard commented Jan 2, 2019

Uh oh!

pradyunsg commented Jan 3, 2019

Uh oh!

julienmalard commented Jan 8, 2019 via email

Uh oh!

pradyunsg commented Jan 9, 2019

Uh oh!

julienmalard commented Jul 14, 2019

Uh oh!

xavfernandez commented Jul 18, 2019

Uh oh!

julienmalard commented Jul 18, 2019

Uh oh!

Choose a reason for hiding this comment

Uh oh!

chrahunt commented Sep 17, 2019

Uh oh!

julienmalard commented Oct 2, 2019

Uh oh!

chrahunt commented Oct 2, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

julienmalard commented Oct 4, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

chrahunt commented Oct 2, 2019 •

edited

Loading

julienmalard commented Oct 4, 2019 •

edited

Loading