profile
viewpoint
Gram orsinium @eclecticiq Amsterdam https://orsinium.dev/ Python, Django, web, security, DevOps, golang. All my projects: orsinium.dev

delete branch life4/deal

delete branch : better-hypothesis

delete time in 15 hours

delete branch life4/deal

delete branch : deeper-raises

delete time in 15 hours

push eventlife4/deal

Gram

commit sha ad6491a1f10c0300afe537a1388e1b4a535438f3

decouple decorators from func

view details

Gram

commit sha 3cc0e88b422aa2e0a6a16b689cdddf38f543e66d

adopt raises check for new format

view details

Gram

commit sha 3974941381ed190ca5009a66c4354bedb49aaddd

adopt all rules

view details

Gram

commit sha 470c018ef46a3205d13cc776507bf1b75d89bc13

use run in contract

view details

Gram

commit sha f86fbfac6ea854f40cac3d2e86cc2f9502e034eb

drop categories proxy from func

view details

Gram

commit sha 4b8c4f969c9b112e9f5cea7098927098e81e0f9b

put all contracts in one func

view details

Gram

commit sha a3859b0bf281a5c768e4e2e7a57071ff958c4273

do a simple inference for raises

view details

Gram

commit sha f0526c4518dfcde2e9d7b9533d7a213b6024c1a4

simple exceptions inference

view details

Gram

commit sha 2c030bfcdaab3a3f304951991db7a03f2b18f817

infer exceptions in assignment call.

view details

Gram

commit sha 4492f896a5aaffd2c5a873015afeb55f5124723e

infer exceptions in subcalls

view details

Gram

commit sha 2cd7407d3044e181cd241c59e2c310f6c9f71668

fix for relative modules

view details

Gram

commit sha f31083556955db01a3e3df0a0e37fe1c46330daf

Merge pull request #34 from life4/deeper-raises Find raised exceptions in called functions

view details

push time in 15 hours

PR merged life4/deal

Find raised exceptions in called functions

If the analyzed function calls another one, analyze it too for possible exceptions. Do not go deeper, only on one level.

+381 -163

0 comment

11 changed files

orsinium

pr closed time in 15 hours

push eventlife4/deal

Gram

commit sha 2cd7407d3044e181cd241c59e2c310f6c9f71668

fix for relative modules

view details

push time in a day

PR opened life4/deal

Find raised exceptions in called functions

If the analyzed function calls another one, analyze it too for possible exceptions. Do not go deeper, only on level.

+379 -163

0 comment

11 changed files

pr created time in a day

push eventlife4/deal

Gram

commit sha 4492f896a5aaffd2c5a873015afeb55f5124723e

infer exceptions in subcalls

view details

push time in a day

issue openedlife4/deal

Generators support

created time in a day

push eventlife4/deal

Gram

commit sha f0526c4518dfcde2e9d7b9533d7a213b6024c1a4

simple exceptions inference

view details

Gram

commit sha 2c030bfcdaab3a3f304951991db7a03f2b18f817

infer exceptions in assignment call.

view details

push time in a day

push eventlife4/deal

Gram

commit sha 4b8c4f969c9b112e9f5cea7098927098e81e0f9b

put all contracts in one func

view details

Gram

commit sha a3859b0bf281a5c768e4e2e7a57071ff958c4273

do a simple inference for raises

view details

push time in a day

issue commentwemake-services/wemake-python-styleguide

Forbid inline if-else

What about if inside comprehensions?

If you about the condition at the end of the comprehension, it's ok. It's a nice way to do filtering. However, when it's as a value, things become even more complicated.

Good:

[a for a in b if a]

bad:

[t if a else f for a in b]

I'm also not sure about the rule in the issue description, but this example definitely hurts me :eyes:

orsinium

comment created time in a day

issue commentlife4/deal

Multi-run properties.

The idea is we want to be able to implement any property as a contract. All existing contract runs when the function is called, and just check input and output values, and side-effects in the process. However, it's not enough when you must do one more function call to check the property. If we'll do it implicitly, it can be unintuitive and raise long debugging why the function was called twice. As an option, we can run these properties only in special tests, but then it can raise a question why the contract wasn't tested in runtime.

And also one more difference for such contracts from other: we can't check it statically at all, it's too complicated. However, it's not a big deal.

orsinium

comment created time in a day

issue commentlife4/deal

Multi-run properties.

@sobolevn I need your advice.

orsinium

comment created time in a day

issue openedlife4/deal

Multi-run properties.

Some really useful properties can be described only by a few calls of the function. A few examples:

reverse(reverse(x)) == x
sum(a, b + 1) == sum(a, b) + 1

In other words, we need to do an additional function call to check the property. While it's ok for pure and fast functions, in other cases (and we have them a lot in python) such behavior can be really painful. So, how to check contract with such properties?

created time in a day

push eventlife4/deal

Gram

commit sha f86fbfac6ea854f40cac3d2e86cc2f9502e034eb

drop categories proxy from func

view details

push time in a day

push eventlife4/deal

Gram

commit sha 3974941381ed190ca5009a66c4354bedb49aaddd

adopt all rules

view details

Gram

commit sha 470c018ef46a3205d13cc776507bf1b75d89bc13

use run in contract

view details

push time in a day

create barnchlife4/deal

branch : deeper-raises

created branch time in a day

issue openedwemake-services/wemake-python-styleguide

Forbid inline if-else

Rule request

<!-- Thanks for coming up with a proposal! -->

Thesis

https://pypi.org/project/flake8-if-expr/

Bad:

r = t if c else f

Good:

if c:
  r = t
else:
  r = f

Reasoning

It's not so easy to read. Most of the languages have a condition first (= cond ? true : false), but Python is different (true if cond else false).

Solution

It's implemented in the plugin above, but I'm quite skeptical about one-rule plugins. I propose to re-implement this simple check instead of adding one more dependency.

<!--

Love wemake-python-styleguide? ❤️ Do you want your features to be implemented faster?

Please consider supporting our collective: 👉 https://opencollective.com/wemake-python-styleguide/donate

-->

created time in a day

issue commentwemake-services/wemake-python-styleguide

Integration problems with VSCode

Do you have a flake8 config? Could you show it, please?

AlexisNava

comment created time in 2 days

issue commentwemake-services/wemake-python-styleguide

Integration problems with VSCode

If you just run flake8 . does it work? Does it show the error messages?

AlexisNava

comment created time in 2 days

issue openedwemake-services/wemake-python-styleguide

Warn about possibly infinite `while`

Rule request

Check that while True block has raise or return.

<!-- Thanks for coming up with a proposal! -->

Thesis

It's OK to not have them for CLI entry point, but in other cases, it can be an error. So, let's ask users explicitly noqa such cases to be sure that they really know what they're doing.

<!-- What exactly do you propose? Please, provide code examples. -->

Reasoning

<!-- Why do you think this is a good idea?

Please, include some strong points and considerations. -->

<!--

Love wemake-python-styleguide? ❤️ Do you want your features to be implemented faster?

Please consider supporting our collective: 👉 https://opencollective.com/wemake-python-styleguide/donate

-->

created time in 2 days

issue openedwemake-services/wemake-python-styleguide

Forbid constants in `while`

Rule request

<!-- Thanks for coming up with a proposal! -->

Thesis

The only constant in while condition that makes sense is True for the "infinite" loop.

Bad:

while 1: ...
while False: ...
while "infinity and beyond": ...

Good:

while x: ...
while True: ...

<!-- What exactly do you propose? Please, provide code examples. -->

Reasoning

Consistency and fighting typos.

Interesting fact: for Python <=2.7 while 1 was faster than while True. Luckily, it's not the case for Python 3.

<!-- Why do you think this is a good idea?

Please, include some strong points and considerations. -->

<!--

Love wemake-python-styleguide? ❤️ Do you want your features to be implemented faster?

Please consider supporting our collective: 👉 https://opencollective.com/wemake-python-styleguide/donate

-->

created time in 2 days

issue openedwemake-services/wemake-python-styleguide

Forbid constant as condition in `if`

Rule request

<!-- Thanks for coming up with a proposal! -->

Thesis

Bad:

if 1: ...
if True: ...
if False: ...
if None: ...
if '': ...

We already have a rule forbidding magic numbers. Also, Pylint checks the unreachable code. It should cover a few cases above, but only a few.

Good:

TYPING = False

if TYPING:
   ...

while True: ...

<!-- What exactly do you propose? Please, provide code examples. -->

Reasoning

I should say something here, but I'm not sure about the possible reasoning behind it. Is it typo? Fast commenting out the code for debugging?

<!-- Why do you think this is a good idea?

Please, include some strong points and considerations. -->

<!--

Love wemake-python-styleguide? ❤️ Do you want your features to be implemented faster?

Please consider supporting our collective: 👉 https://opencollective.com/wemake-python-styleguide/donate

-->

created time in 2 days

issue openedwemake-services/wemake-python-styleguide

Forbid repeating conditions in the one condition block

Rule request

<!-- Thanks for coming up with a proposal! -->

Thesis

In examples below, x, y and z are any Python expressions. For example, x and x can be func(a, b, c=3) and func(a, b, c=3).

Bad:

if x and x: ...
if x or x: ...
if x and not x: ...
if x or not x: ...

Good:

if x: ...
if not x: ...

If we're brave enough, we can after this rule care a little bit about conditions simplification:

x or x and y == x or y
x and y or y and z == y and (x or z)

Reasoning

In most of cases it's a logical mistake. Every condition above can be reduced to True or False, and usually, it's not what you expect.

<!-- Why do you think this is a good idea?

Please, include some strong points and considerations. -->

<!--

Love wemake-python-styleguide? ❤️ Do you want your features to be implemented faster?

Please consider supporting our collective: 👉 https://opencollective.com/wemake-python-styleguide/donate

-->

created time in 2 days

pull request commentdephell/dephell_setuptools

Add appveyor

Ugh, it can't be easy

orsinium

comment created time in 2 days

push eventdephell/dephell_setuptools

Gram

commit sha 441e18c1b93594f18a19d8af98f7579eaf92532f

add comment

view details

push time in 2 days

push eventdephell/dephell_setuptools

Gram

commit sha a703c0871c876e0eb1468c839fea1e300eaac1a7

fix env vars

view details

push time in 2 days

push eventdephell/dephell_setuptools

Gram

commit sha 8aa23f55fd5ebf3a7a1653863090d74f64c4c06f

fix dependencies

view details

push time in 2 days

PR opened dephell/dephell_setuptools

Add appveyor
+37 -0

0 comment

1 changed file

pr created time in 2 days

create barnchdephell/dephell_setuptools

branch : appveyor

created branch time in 2 days

pull request commentdephell/dephell

setuppy: Allow setup.py to import local modules

That's interesting, thank you, I will try it on my work :) However, in this case we should fix the cause, not tests. It happens because of network requests, and it can happen for users as well. So, we should do retry to be more stable.

jayvdb

comment created time in 2 days

push eventdephell/dephell

Gram

commit sha 34093fa808395fe02170657009fdc7ba38fa5bb0

sort imports

view details

push time in 2 days

PR opened dephell/dephell

Integrate dephell_setuptools
+53 -134

0 comment

5 changed files

pr created time in 2 days

push eventdephell/dephell

Gram

commit sha bcbe9e31e76f0e143f417d2363cc9d106f9bd7b8

update deps

view details

push time in 2 days

push eventdephell/dephell

Gram

commit sha bc82f7400f6f38f0c3b89c22f88dc1d6e16f5ec7

uncomment test

view details

push time in 2 days

push eventdephell/dephell_setuptools

Gram

commit sha 11f4904429f0804181f447b7cc6bfc0a93a412db

fix a few details and release

view details

push time in 2 days

create barnchdephell/dephell

branch : setuppy-again

created branch time in 2 days

delete branch dephell/dephell_setuptools

delete branch : typing

delete time in 2 days

push eventdephell/dephell_setuptools

Gram

commit sha da12f8be49f2dbcc4e8f8a986057db575aff0f3e

check type annotations

view details

Gram

commit sha ee34eca82501f5db72c0c765ac934dd1728d007a

Merge pull request #2 from dephell/typing check type annotations

view details

push time in 2 days

PR merged dephell/dephell_setuptools

check type annotations

Breaking change: on error readers raise exception instead of returning None.

+42 -22

0 comment

9 changed files

orsinium

pr closed time in 2 days

PR opened dephell/dephell_setuptools

check type annotations
+42 -22

0 comment

9 changed files

pr created time in 2 days

create barnchdephell/dephell_setuptools

branch : typing

created branch time in 2 days

push eventlife4/textdistance

Gram

commit sha 404d1e7c020c67bc750b7a3f2663d14fdafcb64f

Add articles section

view details

Gram

commit sha bbba2eb9660aa0b360ad54ef7aa5a8348d9a9924

Merge pull request #49 from life4/articles Add articles section

view details

push time in 2 days

PR merged life4/textdistance

Add articles section
+10 -1

0 comment

1 changed file

orsinium

pr closed time in 2 days

PR opened life4/textdistance

Add articles section
+10 -1

0 comment

1 changed file

pr created time in 2 days

create barnchlife4/textdistance

branch : articles

created branch time in 2 days

pull request commentdephell/dephell

setuppy: Add traceback in _execute

Thank you :)

jayvdb

comment created time in 2 days

push eventdephell/dephell

John Vandenberg

commit sha 626c6714e8d5e2dfd2a8f6d09f4bf5946393f60e

setuppy: Add traceback in _execute When an exception occurs while running the fake setup.py in _execute, the traceback is useful for determining how to improve the logic.

view details

Gram

commit sha 5ffab4fa71ea56b111650ec4264930f2853f321d

Merge pull request #301 from jayvdb/setuppy-add-traceback setuppy: Add traceback in _execute

view details

push time in 2 days

PR merged dephell/dephell

setuppy: Add traceback in _execute

When an exception occurs while running the fake setup.py in _execute, the traceback is useful for determining how to improve the logic.

+1 -1

2 comments

1 changed file

jayvdb

pr closed time in 2 days

pull request commentdephell/dephell

setuppy: Allow setup.py to import local modules

flaky plugin?

Sorry, I didn't catch what do you mean. These tests fail because there are smoke tests and do requests to pypi,org. Adding backoff logic can only help us to get them failed less often.

jayvdb

comment created time in 2 days

push eventdephell/dephell

John Vandenberg

commit sha 850536144ee84affde15d9a5f54e13f9457c654b

setuppy: Allow setup.py to import local modules Add the directory containing setup.py to sys.path during _execute and run_setup, so that import of local modules are successful. These imports may be the package to be installed, or other modules beside setup.py that contain extra logic. Related to https://github.com/dephell/dephell/issues/284

view details

Gram

commit sha d64f552794935d68cfad3713fbc753911be01e48

Merge pull request #302 from jayvdb/setuppy-add-sys-path setuppy: Allow setup.py to import local modules

view details

push time in 2 days

PR merged dephell/dephell

setuppy: Allow setup.py to import local modules

Add the directory containing setup.py to sys.path during _execute and run_setup, so that import of local modules are successful. These imports may be the package to be installed, or other modules beside setup.py that contain extra logic.

Related to https://github.com/dephell/dephell/issues/284

+25 -0

5 comments

2 changed files

jayvdb

pr closed time in 2 days

pull request commentdephell/dephell

setuppy: Allow setup.py to import local modules

Failed smoke tests make me think about adding backoff and retry logic :thinking:

jayvdb

comment created time in 2 days

pull request commentdephell/dephell

setuppy: Add traceback in _execute

Yes, right. I've tried a few times, though. TravisCI feels bad :/

jayvdb

comment created time in 2 days

created taglife4/deal

tagv.3.4.0

Design by contract for Python with many validators support.

created time in 2 days

release life4/deal

v.3.4.0

released time in 2 days

push eventlife4/deal

Gram

commit sha 629873986f0cad09ab5ace1199e589c8b5bc2227

release

view details

push time in 2 days

delete branch life4/deal

delete branch : async

delete time in 2 days

push eventlife4/deal

Gram

commit sha 440d133a1813057fb51cd59f9f609bc4716065b0

Support async for decorators

view details

Gram

commit sha 2d6e8843a7c6b0fceff0c9bc2852fc440bdf93c9

mv decorators tests

view details

Gram

commit sha 3e2eab53a316c1f635486c193a1d1a1274639bd6

test async pre

view details

Gram

commit sha e50899884b3ed61ad66cefee95d50d76afdc0917

cover more async decos

view details

Gram

commit sha 047007be86ab3bf6ce0752649cf0eca72bec57e5

cover everything

view details

Gram

commit sha beac85da72689fd588a44f5fe48098614b392c53

update readme

view details

Gram

commit sha 8c125896bd24239fd9db4a4e319e1f1037ea2d9d

Merge pull request #31 from life4/async Support async for decorators

view details

push time in 2 days

PR merged life4/deal

Support async for decorators

Close #21

+309 -82

1 comment

24 changed files

orsinium

pr closed time in 2 days

issue closedlife4/deal

Asyncio support

Is it possible to decorate async def functions? Should it be possible to do it? In case it already works, I guess we can add a docs about it.

closed time in 2 days

sobolevn

push eventlife4/deal

Gram

commit sha 2d6e8843a7c6b0fceff0c9bc2852fc440bdf93c9

mv decorators tests

view details

Gram

commit sha 3e2eab53a316c1f635486c193a1d1a1274639bd6

test async pre

view details

Gram

commit sha e50899884b3ed61ad66cefee95d50d76afdc0917

cover more async decos

view details

Gram

commit sha 047007be86ab3bf6ce0752649cf0eca72bec57e5

cover everything

view details

Gram

commit sha beac85da72689fd588a44f5fe48098614b392c53

update readme

view details

push time in 3 days

pull request commentdephell/dephell

setuppy: Remove _patched_open

Yeah, finally merged. TravisCI is quite nightly for the last few days.

jayvdb

comment created time in 3 days

push eventdephell/dephell

John Vandenberg

commit sha b82fba0093e38768cbfa02feff8a63337fb07afb

setuppy: Remove _patched_open _patched_open used the filename as the file contents to avoid loading the file. This fails when used like exec(open('..')) as in the case of idna 2.8 Related to https://github.com/dephell/dephell/issues/284

view details

Gram

commit sha 29e5e1ab60fe7fc4e2b409406402c84ef3f0a1c0

Merge pull request #299 from jayvdb/setuppy-fix-patched-open setuppy: Remove _patched_open

view details

push time in 3 days

PR merged dephell/dephell

setuppy: Remove _patched_open

_patched_open used the filename as the file contents to avoid loading the file. This fails when used like exec(open('..')) as in the case of idna 2.8

Related to https://github.com/dephell/dephell/issues/284

+0 -8

5 comments

1 changed file

jayvdb

pr closed time in 3 days

pull request commentdephell/dephell

setuppy: Improve _execute setup.py rewriter

Great, thank you :)

jayvdb

comment created time in 3 days

push eventdephell/dephell

John Vandenberg

commit sha b6d65ec6277a9c0d7c569ca03f445cb10d52e37d

setuppy: Improve _execute setup.py rewriter - Allow invocation like setuptools.setup() - Allow setup() to have been invoked within a function, by making _dist a global, and avoiding `return setup(..)`. - Allow funcions called foosetup to not be seen as setup() calls. Related to https://github.com/dephell/dephell/issues/284

view details

Gram

commit sha adf4f4bd54c72144a3264be28b08041421fc4cc1

Merge pull request #303 from jayvdb/setuppy-improve-parse setuppy: Improve _execute setup.py rewriter

view details

push time in 3 days

PR merged dephell/dephell

setuppy: Improve _execute setup.py rewriter
  • Allow invocation like setuptools.setup()
  • Allow setup() to have been invoked within a function, by making _dist a global, and avoiding return setup(..).
  • Allow funcions called foosetup to not be seen as setup() calls.

Related to https://github.com/dephell/dephell/issues/284

+57 -1

0 comment

2 changed files

jayvdb

pr closed time in 3 days

pull request commentdephell/dephell

setuppy: Allow setup.py to import local modules

The testy fails on CI

jayvdb

comment created time in 3 days

pull request commentlife4/deal

WIP Support async for decorators

What color is your code? I have the same code in both colors.

orsinium

comment created time in 3 days

PR opened life4/deal

Support async for decorators

Close #21

+83 -0

0 comment

8 changed files

pr created time in 3 days

create barnchlife4/deal

branch : async

created branch time in 3 days

issue commentdephell/dephell

Forbid to specify `--from` and `--from-format` at the same time.

Sure, keep going. Read CONTRIBUTING.md for a guide how to run the code, put debugger into the function, see what you have. Do PR when you have PoC. Good luck!

orsinium

comment created time in 3 days

pull request commentdephell/dephell

DNM: setuppy: Remove _patched_open

I'm going to merge it when CI passes. I want to start integrating a new project after all your changes are merged to have them in git log and be able to get good ideas from it later. You've done a lot of cool research, I don't want to lost it.

jayvdb

comment created time in 3 days

push eventjaibhageria/dephell

orsinium

commit sha ece6712595eecdfd4b0adac6e3c1f092e5e9477f

sort imports

view details

push time in 3 days

issue commentdephell/dephell

TypeError: attrs() got an unexpected keyword argument 'eq'

That's strange, pip should install the latest version of attrs by default. Anyway, we'll release it very soon, with dephell_setuptools integration. Thank you for your feedback :)

jayvdb

comment created time in 3 days

issue commentdephell/dephell

Forbid to specify `--from` and `--from-format` at the same time.

Too late for Hacktoberfest, but how about DepHellofest? :upside_down_face:

orsinium

comment created time in 3 days

issue closeddephell/dephell

convert setuppy output filename ignored

python3 -m dephell deps convert --traceback --from setup.py --to setup.py.new --to-format setuppy should create a file setup.py.new which is a parse and dump of the existing setup.py

Instead it writes to setup.py (and for dephell repo it changes stuff :P)

closed time in 3 days

jayvdb

issue commentdephell/dephell

convert setuppy output filename ignored

You specified --to instead of --to-path :) I've created #307 to be explicit about it.

jayvdb

comment created time in 3 days

issue openeddephell/dephell

Forbid to specify `--from` and `--from-format` at the same time.

--from redefines --from-format and --from-path and vice versa.

  1. Forbid specify --from and --from-format at the same time.
  2. Forbid specify --to and --to-format at the same time.
  3. Forbid specify --from and --from-path at the same time.
  4. Forbid specify --to and --to-path at the same time.

Do a check in the beginning of Config.attach_cli, and raise ValueError('cannot specify --from and --from-path at the same time. Probably, you mean --from-format'). Keys in the error message should be the real keys that cause the problem, of course.

created time in 3 days

issue commentdephell/dephell

dephell deps --help should give refined help result

@sadielbartholomew thank you for taking care of it! Just do as you see it most helpful :) How is it going? Do you need my help?

jayvdb

comment created time in 3 days

push eventjayvdb/dephell

John Vandenberg

commit sha cfc3f66785c2f9c2a6d73605771856e78cd5e469

setuppy: Emit sorted dictionaries Closes https://github.com/dephell/dephell/issues/289

view details

515hikaru

commit sha 7bbdaab3470a2417f98790eee348b2b4a9e7798f

ignore __pycache__ directory when dephell creates tarball

view details

515hikaru

commit sha bef52ae7f851a7d68d7f719deb1e371b9686e8d2

preserve the directory structure

view details

515hikaru

commit sha ef82b1f8c684a8e610f1fe4a5497b69ad1904da2

add trailing comma

view details

515hikaru

commit sha 14ed9a1dfbea3a94f201835317ff82f7e648055a

fix flake8 error: prefer to using generator rather than list

view details

515hikaru

commit sha 331d203706bce627b1dc19b021664e94f40cc689

review: Use filter function instead of adding each file manually

view details

515hikaru

commit sha 55803983c9a0bcc74599bdd16ce91a91b79cd5e1

review: rename to correct

view details

Gram

commit sha 58337a15b8bb44a94d567197410043ca01fbdc73

Merge pull request #298 from 515hikaru/feature/flake8-error Fix flake8 error

view details

Gram

commit sha 904b15e4e8531c0a4aee97933c6de040a5126cb1

Merge pull request #297 from 515hikaru/master Exclude `__pycache__` from tests in sdist

view details

Gram

commit sha da99b3936e8af9310ca55f0724320e3c6141ad2d

Merge pull request #292 from jayvdb/setuppy-sort-extras setuppy: Emit sorted dictionaries

view details

orsinium

commit sha aa617452f01cfd02b88394cb835eb1660efee46a

Merge remote-tracking branch 'origin/master' into setuppy-add-sys-path

view details

push time in 3 days

push eventjayvdb/dephell

John Vandenberg

commit sha cfc3f66785c2f9c2a6d73605771856e78cd5e469

setuppy: Emit sorted dictionaries Closes https://github.com/dephell/dephell/issues/289

view details

515hikaru

commit sha 7bbdaab3470a2417f98790eee348b2b4a9e7798f

ignore __pycache__ directory when dephell creates tarball

view details

515hikaru

commit sha bef52ae7f851a7d68d7f719deb1e371b9686e8d2

preserve the directory structure

view details

515hikaru

commit sha ef82b1f8c684a8e610f1fe4a5497b69ad1904da2

add trailing comma

view details

515hikaru

commit sha 14ed9a1dfbea3a94f201835317ff82f7e648055a

fix flake8 error: prefer to using generator rather than list

view details

515hikaru

commit sha 331d203706bce627b1dc19b021664e94f40cc689

review: Use filter function instead of adding each file manually

view details

515hikaru

commit sha 55803983c9a0bcc74599bdd16ce91a91b79cd5e1

review: rename to correct

view details

Gram

commit sha 58337a15b8bb44a94d567197410043ca01fbdc73

Merge pull request #298 from 515hikaru/feature/flake8-error Fix flake8 error

view details

Gram

commit sha 904b15e4e8531c0a4aee97933c6de040a5126cb1

Merge pull request #297 from 515hikaru/master Exclude `__pycache__` from tests in sdist

view details

Gram

commit sha da99b3936e8af9310ca55f0724320e3c6141ad2d

Merge pull request #292 from jayvdb/setuppy-sort-extras setuppy: Emit sorted dictionaries

view details

orsinium

commit sha c19ff9f9f40c195090a5fd73fd145c1a7b6572f8

Merge remote-tracking branch 'origin/master' into setuppy-improve-parse

view details

push time in 3 days

push eventjayvdb/dephell

John Vandenberg

commit sha cfc3f66785c2f9c2a6d73605771856e78cd5e469

setuppy: Emit sorted dictionaries Closes https://github.com/dephell/dephell/issues/289

view details

515hikaru

commit sha 7bbdaab3470a2417f98790eee348b2b4a9e7798f

ignore __pycache__ directory when dephell creates tarball

view details

515hikaru

commit sha bef52ae7f851a7d68d7f719deb1e371b9686e8d2

preserve the directory structure

view details

515hikaru

commit sha ef82b1f8c684a8e610f1fe4a5497b69ad1904da2

add trailing comma

view details

515hikaru

commit sha 14ed9a1dfbea3a94f201835317ff82f7e648055a

fix flake8 error: prefer to using generator rather than list

view details

515hikaru

commit sha 331d203706bce627b1dc19b021664e94f40cc689

review: Use filter function instead of adding each file manually

view details

515hikaru

commit sha 55803983c9a0bcc74599bdd16ce91a91b79cd5e1

review: rename to correct

view details

Gram

commit sha 58337a15b8bb44a94d567197410043ca01fbdc73

Merge pull request #298 from 515hikaru/feature/flake8-error Fix flake8 error

view details

Gram

commit sha 904b15e4e8531c0a4aee97933c6de040a5126cb1

Merge pull request #297 from 515hikaru/master Exclude `__pycache__` from tests in sdist

view details

Gram

commit sha da99b3936e8af9310ca55f0724320e3c6141ad2d

Merge pull request #292 from jayvdb/setuppy-sort-extras setuppy: Emit sorted dictionaries

view details

orsinium

commit sha b7e25e692e8483663b7650592022cd444bc84768

Merge remote-tracking branch 'origin/master' into setuppy-fix-patched-open

view details

push time in 3 days

push eventjayvdb/dephell

John Vandenberg

commit sha cfc3f66785c2f9c2a6d73605771856e78cd5e469

setuppy: Emit sorted dictionaries Closes https://github.com/dephell/dephell/issues/289

view details

515hikaru

commit sha 7bbdaab3470a2417f98790eee348b2b4a9e7798f

ignore __pycache__ directory when dephell creates tarball

view details

515hikaru

commit sha bef52ae7f851a7d68d7f719deb1e371b9686e8d2

preserve the directory structure

view details

515hikaru

commit sha ef82b1f8c684a8e610f1fe4a5497b69ad1904da2

add trailing comma

view details

515hikaru

commit sha 14ed9a1dfbea3a94f201835317ff82f7e648055a

fix flake8 error: prefer to using generator rather than list

view details

515hikaru

commit sha 331d203706bce627b1dc19b021664e94f40cc689

review: Use filter function instead of adding each file manually

view details

515hikaru

commit sha 55803983c9a0bcc74599bdd16ce91a91b79cd5e1

review: rename to correct

view details

Gram

commit sha 58337a15b8bb44a94d567197410043ca01fbdc73

Merge pull request #298 from 515hikaru/feature/flake8-error Fix flake8 error

view details

Gram

commit sha 904b15e4e8531c0a4aee97933c6de040a5126cb1

Merge pull request #297 from 515hikaru/master Exclude `__pycache__` from tests in sdist

view details

Gram

commit sha da99b3936e8af9310ca55f0724320e3c6141ad2d

Merge pull request #292 from jayvdb/setuppy-sort-extras setuppy: Emit sorted dictionaries

view details

orsinium

commit sha 4fe04dff4cc8ea0b728564ffe519e508bba779b8

Merge remote-tracking branch 'origin/master' into setuppy-add-traceback

view details

push time in 3 days

issue closeddephell/dephell

setup.py package_data should be sorted for smaller diffs

Whenever I run dephell deps convert --from=pyproject.toml --to=setup.py, the ordering of the package_data={...} dict is unstable and changes between every run. The actual contents don't change, just their ordering.

We're in the long process of upgrading a particular project from setup.py to Poetry, and as part of our new build process we run dephell deps convert quite frequently. This makes for some very noisy diffs as the resulting setup.py is constantly changing, even though it's not really.

Would it be possible to sort the keys in package_data before writing them out so that the output only changes whenever something is legitimately different?

closed time in 3 days

kstrauser

issue commentdephell/dephell

setup.py package_data should be sorted for smaller diffs

Fixed in #292

kstrauser

comment created time in 3 days

pull request commentdephell/dephell

Sort 'package_data' keys in setup.py (#265)

I think it was fixed in #292. Thank you for your help :)

ko3luhbka

comment created time in 3 days

created tagdephell/dephell_setuptools

tagv.0.1.2

created time in 3 days

push eventdephell/dephell_setuptools

Gram

commit sha 0d1af2183aff4c33df399c680006a56cd021a5a6

release

view details

push time in 3 days

delete branch dephell/dephell_setuptools

delete branch : ci

delete time in 3 days

push eventdephell/dephell_setuptools

Gram

commit sha faf6c7ce68de3b8022e0e4db8207db49e1220182

add metainfo and ci

view details

Gram

commit sha 80096b8921fda7fc94bdf7440aac21e21a515d7b

release

view details

Gram

commit sha ce3713f14de1c6f18b6c5e75e575f63947a1c4b3

build ci only once

view details

Gram

commit sha 0f3d2afe6d67e353f23719007431398dad1e5095

sort imports

view details

Gram

commit sha 5fec4e89aefb50dbc43ecab8fd70adf3f35cc946

fix flake8

view details

Gram

commit sha 7891df69d7cff41dbe83c9a314419c701c3639c9

check more pythons

view details

Gram

commit sha 6c33675758b6c3aec537f7c11dcf63abcb7329ee

trigger ci

view details

Gram

commit sha a905e8b49d0fbef42f42dcfa90cc414707ec0109

py 3.5 compat

view details

Gram

commit sha 7791e4d85ad0ca4c897d6a53d8f6083d36d0e1d1

Merge pull request #1 from dephell/ci CI and metainfo

view details

push time in 3 days

PR merged dephell/dephell_setuptools

CI and metainfo
+215 -7

0 comment

20 changed files

orsinium

pr closed time in 3 days

pull request commentdephell/dephell

added inspect project command

Thank you! Could you add docs for it as well, please?

jaibhageria

comment created time in 3 days

push eventjaibhageria/dephell

RPiAwesomeness

commit sha 38cde34eaac3fcab94765d3454c90c84fe963406

Add issue template

view details

John Vandenberg

commit sha 9e86b8eb06053897b2dd9fa5c2f689e9b6dac44e

setuppy.py: Force version to string When the value is a packaging.specifiers.SpecifierSet, - comparisons to UNKNOWN fail with Invalid specifier: 'UNKNOWN'. - invoking value.strip() fails as it has no attribute 'strip' Related to https://github.com/dephell/dephell/issues/284

view details

Gram

commit sha 15a50e241b4d7e263bd678c00e49178df7df10cb

fix attrs version

view details

Gram

commit sha f8437c8c53d9422181c07eecfc3c4d790cce34d8

Merge pull request #285 from jayvdb/version-object setuppy.py: Force version to string

view details

Gram

commit sha 86c1f3ec758dc34d08c021eaf53f913c11457b47

fix mdlint

view details

Gram

commit sha 34518b8827be99e2c4aac66ae4393620548eaa3c

Merge pull request #287 from dephell/fix-attrs-version fix attrs version

view details

Gram

commit sha 7c86aef0664351a377668672f0ef96027ca7ee15

change config

view details

RPiAwesomeness

commit sha c8f2d5f2bfa40aa166a94c4d1379f15760a7fce2

Fixed CONTRIBUTING.md

view details

Gram

commit sha 092cb4c6b3adcdad3beb8da84b8c34f4c7cea214

Merge pull request #290 from dephell/fix-mdlint fix mdlint

view details

orsinium

commit sha 6618f64b150367bc0b77394c1038869547bdbeff

Merge remote-tracking branch 'origin/master' into issue_template

view details

orsinium

commit sha d9b3c96472ef95df9cbc1959e1a9dd87a0a9a9bc

add package_dir in setup.py on dump

view details

Gram

commit sha 24a822369a54fd45120f691d85f722e8631b2c61

Merge pull request #291 from dephell/fix-package-dir add package_dir in setup.py on dump

view details

Gram

commit sha 97d45928ab8af11e8aaa6778fe744af2c424cd2d

Merge pull request #280 from RPiAwesomeness/issue_template Add issue template

view details

John Vandenberg

commit sha cfc3f66785c2f9c2a6d73605771856e78cd5e469

setuppy: Emit sorted dictionaries Closes https://github.com/dephell/dephell/issues/289

view details

515hikaru

commit sha 7bbdaab3470a2417f98790eee348b2b4a9e7798f

ignore __pycache__ directory when dephell creates tarball

view details

515hikaru

commit sha bef52ae7f851a7d68d7f719deb1e371b9686e8d2

preserve the directory structure

view details

515hikaru

commit sha ef82b1f8c684a8e610f1fe4a5497b69ad1904da2

add trailing comma

view details

515hikaru

commit sha 14ed9a1dfbea3a94f201835317ff82f7e648055a

fix flake8 error: prefer to using generator rather than list

view details

515hikaru

commit sha 331d203706bce627b1dc19b021664e94f40cc689

review: Use filter function instead of adding each file manually

view details

515hikaru

commit sha 55803983c9a0bcc74599bdd16ce91a91b79cd5e1

review: rename to correct

view details

push time in 3 days

PR closed dephell/dephell

DNM: setuppy: Use metaextract

metaextract runs a setuptools command to accurately extract metadata from setup.py

Related to https://github.com/dephell/dephell/issues/284

+13 -2

8 comments

3 changed files

jayvdb

pr closed time in 3 days

pull request commentdephell/dephell

DNM: setuppy: Use metaextract

Thank you for your PoC. We're going to use dephell_setuptools. It has the implementation from metaextract for accuracy and static analysis (based on poetry and improved a lot) for speed.

jayvdb

comment created time in 3 days

more