BUG: `line terminator` and '\n to \r\n' problem in Windows(Issue #20353) #21406

deflatSOCO · 2018-06-10T07:01:37Z

closes DataFrame.to_csv not using correct line terminator value #20353
tests added / passed
passes git diff upstream/master -u -- "*.py" | flake8 --diff
whatsnew entry

gfyoung · 2018-06-10T07:11:36Z

doc/source/whatsnew/v0.24.0.txt

@@ -147,7 +147,7 @@ MultiIndex
 I/O
 ^^^

-
+- (Only for Windows) Bug in :meth:`DataFrame.to_csv`, in which all `\n`s are converted to `\r\n` (:issue:`20353`)


I would move the Windows portion as follows:

Bug in :meth:`DataFrame.to_csv`, in which all `\n`s are converted to `\r\n` on Windows (:issue:`20353`)

gfyoung · 2018-06-10T07:12:47Z

pandas/tests/io/formats/test_to_csv.py

+                b'int,str_crlf\n' \
+                b'1,abc\n' \
+                b'2,"d\r\nef"\n' \
+                b'3,"g\r\nh\r\n\r\ni"\n'


Use parentheses instead of slashes for multi-line strings.

gfyoung · 2018-06-10T07:15:02Z

pandas/io/common.py

@@ -394,13 +395,13 @@ def _get_handle(path_or_buf, mode, encoding=None, compression=None,
    elif is_path:
        if compat.PY2:
            # Python 2
-            f = open(path_or_buf, mode)
+            f = io.open(path_or_buf, mode, newline='\n')


Why do we need to use io.open for this one?

I referenced this stackoverflow article, but it was not appropreate as all CI tests failed in py2 environment.
I will fix it.

pep8speaks · 2018-06-10T09:48:43Z

Hello @deflatSOCO! Thanks for updating the PR.

There are no PEP8 issues in the file pandas/core/generic.py !
There are no PEP8 issues in the file pandas/io/common.py !
There are no PEP8 issues in the file pandas/io/formats/csvs.py !
There are no PEP8 issues in the file pandas/tests/frame/test_to_csv.py !
There are no PEP8 issues in the file pandas/tests/io/formats/test_to_csv.py !
There are no PEP8 issues in the file pandas/tests/io/test_html.py !
There are no PEP8 issues in the file pandas/tests/util/test_testing.py !
There are no PEP8 issues in the file pandas/util/testing.py !
Complete extra results for this file :

file_to_check.py:2521:-258: W605 invalid escape sequence '('
file_to_check.py:2521:-255: W605 invalid escape sequence ')'

Comment last updated on October 19, 2018 at 05:43 Hours UTC

codecov · 2018-06-10T09:48:54Z

Codecov Report

Merging #21406 into master will decrease coverage by <.01%.
The diff coverage is 88.88%.

@@            Coverage Diff             @@
##           master   #21406      +/-   ##
==========================================
- Coverage   92.19%   92.19%   -0.01%     
==========================================
  Files         169      169              
  Lines       50950    50956       +6     
==========================================
+ Hits        46975    46980       +5     
- Misses       3975     3976       +1

Flag	Coverage Δ
#multiple	`90.61% <88.88%> (-0.01%)`	⬇️
#single	`42.27% <33.33%> (-0.01%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/generic.py	`96.65% <ø> (ø)`	⬆️
pandas/util/testing.py	`86.5% <100%> (+0.05%)`	⬆️
pandas/io/formats/csvs.py	`98.22% <100%> (+0.01%)`	⬆️
pandas/io/common.py	`70.76% <66.66%> (-0.28%)`	⬇️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ecc5cbc...23ce5c6. Read the comment docs.

bbibentyo · 2018-06-10T11:41:41Z

pandas/io/common.py

            f = open(path_or_buf, mode)
        elif encoding:
            # Python 3 and encoding
-            f = open(path_or_buf, mode, encoding=encoding)
+            f = open(path_or_buf, mode, encoding=encoding, newline='\n')


@deflatSOCO Is there a specific reason for not making newline a function parameter? I would suggest that you make newline a _get_handle parameter and default it to None then let open handle it from there.
Then update formats.csvs.CSVFormatter.save to pass self.line_terminator to the function

@bibenb1 Thanks for feedback.

Python document says,

When writing output to the stream, if newline is None, any '\n' characters written are translated to the system default line separator, os.linesep. If newline is '' or '\n', no translation takes place. If newline is any of the other legal values, any '\n' characters written are translated to the given string.

formats.csvs.CSVFormatter.save applies self.line_terminator to csv.writer(in line157) before calling get_handle(in line163), so it should not be passed in open() again.
From this thing, I thought newline value should ALWAYS be '' or '\n', to avoid different behavior between Windows and others, so I did not make it a function parameter.

chris-b1 · 2018-06-11T14:54:38Z

Willing to be convinced otherwise but I don't think the default should change to \n on Windows. It's an annoying convention, but convention all the same.

gfyoung · 2018-06-11T15:28:56Z

@chris-b1 : The issue more relates to an inconsistency, as here.

chris-b1 · 2018-06-11T15:35:17Z

Right, and I'm good with fixing the original bug (\r\n terminator when \n explicitly passed), but this PR would, if I'm understanding correctly, also change the default to \n when no line terminator is passed.

gfyoung · 2018-06-11T16:57:16Z

but this PR would, if I'm understanding correctly, also change the default to \n when no line terminator is passed.

@chris-b1 : Ah, I see what you mean. That is indeed correct based on the tests.

@deflatSOCO : Could you change it so that the default remains \r\n unless otherwise specified?

deflatSOCO · 2018-06-11T21:57:10Z

@chris-b1 @gfyoung Please let me convince.
The current document says the default value of line_terminator in to_csv is '\n' no matter what OS is, so I just followed that in this PR.
I suppose that you mean that should be changed to '\r\n' only in Windows.

If so, the expected output undet the Windows environment should be like this. Is that right?

code

data = pd.DataFrame({
    "integer":[1,2,3],
    "string_lf":["abc","d\nef","gh\r\ni"]
})
data.to_csv('test.txt', index=False)
print(pd.read_csv('test.csv'))
data.to_csv('test.csv', index=False)
print(pd.read_csv('test.csv'))
with open("test.csv", mode='rb') as f:
    print(f.read())

current output(v0.23)

   integer  string_lf
0        1        abc
1        2    d\r\nef
2        3  gh\r\r\ni
b'integer,string_lf\r\n1,abc\r\n2,"d\r\nef"\r\n3,"gh\r\r\ni"\r\n'

current output(this PR)

   integer  string_lf
0        1        abc
1        2      d\nef
2        3    gh\r\ni
b'integer,string_lf\n1,abc\n2,"d\nef"\n3,"gh\ni"\n'

expected output

   integer  string_lf
0        1        abc
1        2      d\nef
2        3    gh\r\ni
b'integer,string_lf\r\n1,abc\r\n2,"d\nef"\r\n3,"gh\ni"\r\n'

gfyoung · 2018-06-11T22:41:16Z

@chris-b1 @gfyoung Please let me convince.
The current document says the default value of line_terminator in to_csv is '\n' no matter what OS is, so I just followed that in this PR.

@chris-b1 : Thoughts? IMO @deflatSOCO has a point here.

gfyoung · 2018-06-19T16:29:51Z

@jreback @chris-b1 : Could you take a look at this?

chris-b1 · 2018-06-23T15:43:28Z

I know the current docs say contrary, but my vote is that we continue defaulting to the platform default line terminator (\r\n on Windows) and updating the docs as such.

It preserves backwards compat - so if someone was relying on the \r being there, their code won't break, and is consistent with other programs on Windows that save csvs, including Excel, and python's csv writer.

In [1]: import csv

In [3]: with open('tmp.csv', 'wt', newline='') as f:
   ...:     writer = csv.writer(f)
   ...:     writer.writerow(['a', 'b'])
   ...:     writer.writerow([1, 2])
   ...:     writer.writerow([3, 4])

In [4]: open('tmp.csv', 'rb').read()
Out[4]: b'a,b\r\n1,2\r\n3,4\r\n'

(again fully in support of merging this with bugfix for case with explicit ending)

jreback · 2018-06-23T15:54:20Z

can u rebase

gfyoung · 2018-06-24T00:24:25Z

I know the current docs say contrary, but my vote is that we continue defaulting to the platform default line terminator (\r\n on Windows) and updating the docs as such.

Fair enough. @deflatSOCO : Can you update?

(If you can't, I can pick this up)

gfyoung · 2018-06-24T00:25:22Z

Marking for 0.24.0, as I can bring this to the finish line if need be.

* re-defined testcases that suits conversations in PR pandas-dev#21406 * changed default value of line_terminator to os.linesep * changed API document of DataFrame.to_csv * changed "newline" value of "open()" from '\n' to '' * Updated whatsnew document related pages: * Issue pandas-dev#20353 * PR pandas-dev#21406

* Updates: * Updated expected values for some tests about 'to_csv()' method, to deal with new default value of 'line_terminator' arg. * Related Issue: * Issue pandas-dev#20353 * PR pandas-dev#21406

gfyoung · 2018-06-27T05:06:21Z

pandas/tests/frame/test_to_csv.py

+        expected = ('"A","B"' + os.linesep +
+                    '1,"foo"' + os.linesep +
+                    '2,"bar"' + os.linesep +
+                    '3,"baz"' + os.linesep)


I wonder if we can make a testing utility (i.e. function) for this os.linesep "magic"

jreback · 2018-07-06T16:22:35Z

pandas/tests/frame/test_to_csv.py

+            with open(path, mode='rb') as f:
+                assert f.read() == expected
+
+            df.to_csv(path, line_terminator='\n')


can you do separate ensure_cleans() for each case & put a comment for each case

jreback · 2018-07-06T16:26:03Z

pandas/tests/io/formats/test_to_csv.py

+                b'2,"d\nef"' + os_linesep +
+                b'3,"g\nh\n\ni"' + os_linesep
+            )
+


can you make separate ensure_cleans here

jreback · 2018-07-06T16:26:51Z

pandas/tests/io/formats/test_to_csv.py

        df.to_csv(sys.stdout, encoding='ascii')
        output = sys.stdout.getvalue()
        assert output == expected_ascii
        assert not sys.stdout.closed

+    @pytest.mark.xfail(
+        os.name == 'nt',


use compat.is_windows_platform()

jreback · 2018-07-06T16:27:22Z

pandas/util/testing.py

+
+
+def convert_rows_list_to_csv_str(rows_list):
+    csv_str = ''


can you add a mini-doc-string

jreback · 2018-07-06T16:28:38Z

pandas/util/testing.py

+    csv_str = ''
+    for row in rows_list:
+        csv_str += row + os.linesep
+    return csv_str


can you use

return os.linesep.join(rows_list) ?

* Requested in PR pandas-dev#21406

TomAugspurger · 2018-07-24T16:23:48Z

doc/source/whatsnew/v0.24.0.txt

@@ -89,6 +89,8 @@ Other Enhancements

 Backwards incompatible API changes
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+- When file path is given by string format in :func:`DataFrame.to_csv`, it disables universal newline support.


It's not clear form this whether disabling universal newline support is the previous or new behavior. I think I would phrase it as

:func:`DataFrame.to_csv` now uses :func:`os.linesep` rather than ``'\n'`` for the default line terminator.

TomAugspurger · 2018-07-24T16:24:35Z

pandas/io/common.py

@@ -409,13 +409,15 @@ def _get_handle(path_or_buf, mode, encoding=None, compression=None,
    elif is_path:
        if compat.PY2:
            # Python 2
+            if mode == 'w':


Why was this change necessary?

Sorry for late reply.

Document of csv.writer() in python2 says:

If csvfile is a file object, it must be opened with the ‘b’ flag on platforms where that makes a difference.

Document of open() in python2 says:

Python on Windows makes a distinction between text and binary files; the end-of-line characters in text files are automatically altered slightly when data is read or written.

From these two, I thought binary mode should always be activated in python2 when writing.

pandas/util/testing.py

gfyoung · 2018-08-25T07:27:11Z

@deflatSOCO : Could you address the merge conflicts here?

@TomAugspurger @jreback : Can you have a look again at this?

deflatSOCO · 2018-09-22T12:01:49Z

@TomAugspurger @jreback : Can you have a look again at this?

jreback

looks ok to me. i would just request an exapanded whatsnew note. maybe a small sub-section on when a user should care and what they need to do about it. e.g. i write csv's on windows, do I need to do anything?what if I explicity pass line_terminator

jreback · 2018-09-23T13:02:30Z

doc/source/whatsnew/v0.24.0.txt

@@ -189,6 +189,7 @@ Other Enhancements

 Backwards incompatible API changes
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+- :func:`DataFrame.to_csv` now uses :func:`os.linesep` rather than ``'\n'`` for the default line terminator.(:issue:`20353`)


can you elabortate on what if anything a user needs to do (e.g. when should they care about this)

jreback · 2018-09-23T13:02:45Z

doc/source/whatsnew/v0.24.0.txt

@@ -731,6 +732,7 @@ MultiIndex
 I/O
 ^^^

+- Bug in :meth:`DataFrame.to_csv`, in which all `\n`s are converted to `\r\n` on Windows (:issue:`20353`)


is this a repeat?

PR pandas-dev#21406 * Expanded whatsnew documents about the change of to_csv * Resolved duplication

jreback · 2018-09-24T13:42:05Z

doc/source/whatsnew/v0.24.0.txt

+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+:func:`DataFrame.to_csv` now uses :func:`os.linesep` rather than ``'\n'``
+ for the default line terminator(:issue:`20353`).


space after terminator

doc/source/whatsnew/v0.24.0.txt

jreback · 2018-09-24T13:44:35Z

doc/source/whatsnew/v0.24.0.txt

+
+.. code-block:: ipython
+
+  In [1]: import pandas as pd


you can skip the pd import

jreback · 2018-09-24T13:46:27Z

doc/source/whatsnew/v0.24.0.txt

+
+  In [5]: with open("test.csv", mode='rb') as f:
+     ...:     print(f.read())
+  b'string_with_lf,string_with_crlf\r\nabc,abc\r\n"d\r\nef","d\r\r\nef"\r\n"g\r\nh


can make this simpler, just show [3] , [4]

Can I make data smaller, and keep [5] instead of [4]?
With [4], I can't show the terminator of each line in CSV, which is the main point of this change. That's why I want to keep [5].

yes, the point is we want a small example that shows the most common case where a user would want to know what is changing. Too long and you lose the reader, too short and you don't know if the change applies to youl.

jreback · 2018-09-24T13:46:36Z

doc/source/whatsnew/v0.24.0.txt

+
+  In [1]: import pandas as pd
+
+  In [2]: data = pd.DataFrame({


same as above

jreback · 2018-09-24T13:47:09Z

doc/source/whatsnew/v0.24.0.txt

+     ...:     data.to_csv(f,index=False,line_terminator='\n')
+
+  In [7]: pd.read_csv("test2.csv")
+  Out[7]:


thanks for writing all this, but just show a single, most important change

PR pandas-dev#21406 - Applied jreback's feedback in PR pandas-dev#21406

doc/source/whatsnew/v0.24.0.txt

jreback · 2018-10-07T23:23:33Z

@deflatSOCO can you update the whatsnew as discussed above, and rebase on master.

deflatSOCO · 2018-10-08T06:51:52Z

@jreback I applied in commit e4badc4, and rebased. Please check them.

doc/source/whatsnew/v0.24.0.txt

gfyoung · 2018-10-09T18:18:03Z

doc/source/whatsnew/v0.24.0.txt

+b'string_with_lf,string_with_crlf\n"a\nbc","a\r\nbc"\n'
+
+
+- On windows, the value of ``os.linesep`` is ``'\r\n'``,


windows --> Windows

doc/source/whatsnew/v0.24.0.txt

gfyoung · 2018-10-09T18:20:30Z

doc/source/whatsnew/v0.24.0.txt

+
+
+- As default value of ``line_terminator`` changes, just passing file object with ``newline='\n'`` does not set ``'\n'`` to line terminator.
+  Pass ``line_terminator='\n'`` explicitly.


I think we can reword this a little. Say something like:

For files objects, specifying newline is not sufficient to set the line terminator. You must pass in the line terminator explicitly, even in this case.

PR pandas-dev#21406

doc/source/whatsnew/v0.24.0.txt

pandas/core/generic.py

jreback · 2018-10-11T01:23:28Z

just a couple of minor comments / doc-string things. @gfyoung ping when happy.

jreback · 2018-10-18T16:14:41Z

@gfyoung you ok with this, can you rebase this? or @deflatSOCO

gfyoung · 2018-10-18T16:16:42Z

@jreback : Can do, given that no reply how has come since the latest comment in > 1 week.

gfyoung · 2018-10-18T20:57:25Z

Alas, there were too many merge conflicts with master that I had to nuke the old commits and replace them with a squashed version.

@deflatSOCO

* Use OS line terminator if none is provided * Enforce line terminator selection if one is Originally authored by @deflatSOCO, but reapplied by @gfyoung due to enormous merge conflicts. Closes pandas-devgh-20353.

gfyoung · 2018-10-19T08:47:50Z

@jreback : Addressed all comments, rebased, and all is green! PTAL.

jreback · 2018-10-19T12:55:35Z

thanks @deflatSOCO and @gfyoung

@deflatSOCO

* Use OS line terminator if none is provided * Enforce line terminator selection if one is Originally authored by @deflatSOCO, but reapplied by @gfyoung due to enormous merge conflicts. Closes pandas-devgh-20353.

gfyoung added IO CSV read_csv, to_csv Windows Windows OS labels Jun 10, 2018

gfyoung reviewed Jun 10, 2018

View reviewed changes

bbibentyo reviewed Jun 10, 2018

View reviewed changes

gfyoung added this to the 0.24.0 milestone Jun 24, 2018

deflatSOCO force-pushed the windows_crlf branch from 293dcd7 to 48f5685 Compare June 24, 2018 01:13

gfyoung reviewed Jun 27, 2018

View reviewed changes

jreback requested changes Jul 6, 2018

View reviewed changes

jreback requested a review from TomAugspurger July 6, 2018 16:28

deflatSOCO added a commit to deflatSOCO/pandas that referenced this pull request Jul 7, 2018

CLN: applied jreback's request

33c20e4

* Requested in PR pandas-dev#21406

TomAugspurger reviewed Jul 24, 2018

View reviewed changes

WillAyd mentioned this pull request Aug 24, 2018

Saving and reloading a csv on Windows with a \n-character in a field replaces \n with \r\n. #22497

Open

jreback requested changes Sep 23, 2018

View reviewed changes

deflatSOCO added a commit to deflatSOCO/pandas that referenced this pull request Sep 24, 2018

DOC: expanded whatsnew docs

8bde6c6

PR pandas-dev#21406 * Expanded whatsnew documents about the change of to_csv * Resolved duplication

jreback requested changes Sep 24, 2018

View reviewed changes

deflatSOCO added a commit to deflatSOCO/pandas that referenced this pull request Sep 25, 2018

DOC: applied jreback's feedback

74a71fe

PR pandas-dev#21406 - Applied jreback's feedback in PR pandas-dev#21406

jreback requested changes Sep 26, 2018

View reviewed changes

doc/source/whatsnew/v0.24.0.txt Show resolved Hide resolved

doc/source/whatsnew/v0.24.0.txt Show resolved Hide resolved

jreback requested changes Oct 9, 2018

View reviewed changes

doc/source/whatsnew/v0.24.0.txt Show resolved Hide resolved

gfyoung reviewed Oct 9, 2018

View reviewed changes

doc/source/whatsnew/v0.24.0.txt Show resolved Hide resolved

gfyoung reviewed Oct 9, 2018

View reviewed changes

deflatSOCO added a commit to deflatSOCO/pandas that referenced this pull request Oct 10, 2018

DOC: applied gfyoung's feedback

eeec91c

PR pandas-dev#21406

jreback reviewed Oct 11, 2018

View reviewed changes

doc/source/whatsnew/v0.24.0.txt Outdated Show resolved Hide resolved

jreback reviewed Oct 11, 2018

View reviewed changes

pandas/core/generic.py Show resolved Hide resolved

jreback approved these changes Oct 11, 2018

View reviewed changes

gfyoung force-pushed the windows_crlf branch from eeec91c to 48dd218 Compare October 18, 2018 20:56

BUG: Use correct line terminator on Windows

23ce5c6

* Use OS line terminator if none is provided * Enforce line terminator selection if one is Originally authored by @deflatSOCO, but reapplied by @gfyoung due to enormous merge conflicts. Closes pandas-devgh-20353.

gfyoung force-pushed the windows_crlf branch from 48dd218 to 23ce5c6 Compare October 19, 2018 05:43

jreback merged commit 31f86d6 into pandas-dev:master Oct 19, 2018

BenRussert mentioned this pull request Dec 14, 2018

BUG: Fix line_terminator option #23608

Closed

2 tasks

BenRussert added a commit to BenRussert/pandas that referenced this pull request Dec 15, 2018

DOC: Update io docs for pandas-dev#21406

f562d82

TomAugspurger mentioned this pull request Sep 17, 2019

Google cloud storage csv file write with pandas empty lines fsspec/gcsfs#176

Closed

		b'string_with_lf,string_with_crlf\n"a\nbc","a\r\nbc"\n'


		- On windows, the value of ``os.linesep`` is ``'\r\n'``,



		- As default value of ``line_terminator`` changes, just passing file object with ``newline='\n'`` does not set ``'\n'`` to line terminator.
		Pass ``line_terminator='\n'`` explicitly.

BUG: line terminator and '\n to \r\n' problem in Windows(Issue #20353) #21406

BUG: line terminator and '\n to \r\n' problem in Windows(Issue #20353) #21406

Conversation

deflatSOCO commented Jun 10, 2018

gfyoung Jun 10, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pep8speaks commented Jun 10, 2018 • edited Loading

Comment last updated on October 19, 2018 at 05:43 Hours UTC

codecov bot commented Jun 10, 2018 • edited Loading

Codecov Report

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chris-b1 commented Jun 11, 2018

gfyoung commented Jun 11, 2018

chris-b1 commented Jun 11, 2018

gfyoung commented Jun 11, 2018

deflatSOCO commented Jun 11, 2018 • edited Loading

gfyoung commented Jun 11, 2018 • edited Loading

gfyoung commented Jun 19, 2018

chris-b1 commented Jun 23, 2018

jreback commented Jun 23, 2018

gfyoung commented Jun 24, 2018 • edited Loading

gfyoung commented Jun 24, 2018

gfyoung Jun 27, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

deflatSOCO Aug 4, 2018 • edited Loading

Choose a reason for hiding this comment

gfyoung commented Aug 25, 2018

deflatSOCO commented Sep 22, 2018

jreback left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jreback commented Oct 7, 2018

deflatSOCO commented Oct 8, 2018

Choose a reason for hiding this comment

gfyoung Oct 9, 2018 • edited Loading

Choose a reason for hiding this comment

jreback commented Oct 11, 2018

jreback commented Oct 18, 2018

gfyoung commented Oct 18, 2018

gfyoung commented Oct 18, 2018 • edited Loading

gfyoung commented Oct 19, 2018 • edited Loading

jreback commented Oct 19, 2018

BUG: `line terminator` and '\n to \r\n' problem in Windows(Issue #20353) #21406

BUG: `line terminator` and '\n to \r\n' problem in Windows(Issue #20353) #21406

gfyoung Jun 10, 2018 •

edited

Loading

pep8speaks commented Jun 10, 2018 •

edited

Loading

codecov bot commented Jun 10, 2018 •

edited

Loading

deflatSOCO commented Jun 11, 2018 •

edited

Loading

gfyoung commented Jun 11, 2018 •

edited

Loading

gfyoung commented Jun 24, 2018 •

edited

Loading

gfyoung Jun 27, 2018 •

edited

Loading

deflatSOCO Aug 4, 2018 •

edited

Loading

gfyoung Oct 9, 2018 •

edited

Loading

gfyoung commented Oct 18, 2018 •

edited

Loading

gfyoung commented Oct 19, 2018 •

edited

Loading