Fixed all flake8 errors and added flake8 in the build script by hackaugusto · Pull Request #169 · raiden-network/raiden

hackaugusto · 2016-10-19T04:01:09Z

Note that I did not try to fix any flake8 erros, I assumed that the code is correct and only silenced it

hackaugusto · 2016-10-19T05:28:41Z

I did not choose the max-line-length, I think pep8 running on the server inherited from the flake8 configuration.

My system was using 160, and I known some prefer 79/80. I would ask to merge this as is, this way we can get the linter running, because there are 700 violations of the max-line-length as 79 and 645 as 80.

LefterisJP · 2016-10-19T07:56:56Z

Max line length is setup here.

I would like us to try and stick to a soft-limit of 80 lines and a hard limit of 99 or 100.
This certainly does not need to be part of this PR since line length violations are most probably everywhere and a whole category on their own.

LefterisJP · 2016-10-19T07:59:42Z

@hackaugusto you are riddling the codebase with a lot of
#pylint: XXX type of comments. From what I can see so far we are just using flake8 and not pylint. If you are the only one using pylint it would be nice to not fill the codebase with comments on the tool you use alone.

Do more people use it?

LefterisJP · 2016-10-19T08:02:22Z

raiden/encoding/format.py

            length = len(value)
            if length > field.size_bytes:
-                raise ValueError('value with length {length} for {attr} is to big'.format(length=length, attr=name))
+                msg = 'value with length {length} for {attr} is to big'.format(


Since we are fixing style here, let's also fix this typo.
is to big -> is too big

Yep, nice catch :)

LefterisJP · 2016-10-19T08:05:33Z

raiden/network/discovery.py

    your ethereum-/raiden-address and looking up endpoints for other ethereum-/raiden-addressess.
    """
-    def __init__(self, blockchainservice, discovery_contract_address, poll_timeout=DEFAULT_POLL_TIMEOUT):
+    def __init__(self, blockchainservice, discovery_contract_address,


This is a matter of style that we have discussed before. If the line breaks on a function's arguments then please format it like:

def __init__( self, blockchainservice, discovery_contract_address, poll_timeout=DEFAULT_POLL_TIMEOUT ):

Why? It's actually quite simple. First it looks symmetric, simple and organized. But for a more practical reason, when refactoring the code later and adding/removing arguments you just need to change one entry and don't have to mess with rearranging the entire arguments line.

Humm, the suggestion doesnt fully comply with the Python's coding standard PEP8:

When using a hanging indent the following should be considered; there should be no arguments on the first line and further indentation should be used to clearly distinguish itself as a continuation line.

# More indentation included to distinguish this from the rest. def long_function_name( var_one, var_two, var_three, var_four): print(var_one)

Pylint is the tool erroring in this (the styling tools are not), so I will have a look on how to configure this (I have no preference among the hanging indent options).

Oh I am fine with placing the closing parentheses after the last argument. What I really want to emphasize is each argument on its own line if you are to break any arguments.

But note that flake8 accepts both closing parentheses styles. I don't have any strong opinion on that. Whatever you prefer.

LefterisJP · 2016-10-19T08:06:54Z

raiden/network/rpc/client.py

    # pylint: disable=too-many-instance-attributes,unused-argument

-    def __init__(self, privatekey_bin, registry_address, host, port, poll_timeout=DEFAULT_POLL_TIMEOUT, **kwargs):
+    def __init__(self, privatekey_bin, registry_address, host, port,


Same comment as above. Break all arguments each in its own line.

LefterisJP · 2016-10-19T08:07:21Z

raiden/network/rpc/client.py


 class Asset(object):
-    def __init__(self, jsonrpc_client, asset_address, startgas=GAS_LIMIT,  # pylint: disable=too-many-arguments
+    def __init__(self, jsonrpc_client, asset_address, startgas=GAS_LIMIT,


Same comment as above. Break all arguments each in its own line.

LefterisJP · 2016-10-19T08:07:34Z

raiden/network/rpc/client.py


 class ChannelManager(object):
-    def __init__(self, jsonrpc_client, manager_address, startgas=GAS_LIMIT,  # pylint: disable=too-many-arguments
+    def __init__(self, jsonrpc_client, manager_address, startgas=GAS_LIMIT,


Same comment as above. Break all arguments each in its own line.

LefterisJP · 2016-10-19T08:14:09Z

tox.ini

 deps =
    -rrequirements.txt
    pdbpp
+    pep8


Why pep8? I thought we had agreed on using flake8 for this ...

flake8 is a wrapper around pep8/pyflakes/mccabe, I decided to start with the simpler option, in other words, these changes are required but not sufficient to use flake8 and I did not look at flake8 at this moment.

Go with flake8 then, it's quite a complete tool and will help us a lot in keeping our styling checked. Moreover we need consistency, can't have each dev trying out a different tool.

LefterisJP · 2016-10-19T08:14:23Z

tox.ini

    pdbpp
+    pep8
 commands =
+    pep8 --exclude .eggs,.tox,devenv,tools/ansible .


Please use flake8

LefterisJP · 2016-10-19T08:14:33Z

tox.ini

 deps =
    -rrequirements.txt
    coverage==4.0
+    pep8


Please use flake8

LefterisJP · 2016-10-19T08:14:39Z

tox.ini

    coverage==4.0
+    pep8
 commands =
+    pep8 --exclude .eggs,.tox,devenv,tools/ansible .


Please use flake8

LefterisJP · 2016-10-19T08:14:59Z

tox.ini

 max-complexity = 10
+
+[pep8]
+max-line-length = 99


Not needed, the flake8 configuration is just 3 lines above this.

Hummm, if you don't mind these 3 extra lines I'm going to leave them there, pep8 is a requirement for the flake8 tool, that means both binaries are available if flake8 is installed and tools like syntastic will run both automatically, this makes life easier by having both tools consistent. Note: this happen even if pep8 is not installed directly and we are using just the pycodestyle package.

hackaugusto · 2016-10-19T10:40:56Z

@hackaugusto you are riddling the codebase with a lot of

I wouldn't call it riddling.

From what I can see so far we are just using flake8 and not pylint. If you are the only one using pylint it would be nice to not fill the codebase with comments on the tool you use alone.

It's a nice tool, I recommend you to give it a try. It does more checks than pep8/pyflakes/mccabe, it actually points out bad practices, but saddly has more false errors.

hackaugusto · 2016-10-19T10:42:07Z

Max line length is setup here.

Interesting, I didn't realize we had a second config file, I'm going to remove the configuration section from the tox.ini

LefterisJP · 2016-10-28T09:10:37Z

@hackaugusto There are some minor conflicts. Can you address and rebase?

hackaugusto · 2016-10-31T23:02:13Z

The build will fail until PR 502 is merged.

LefterisJP · 2016-10-31T23:08:20Z

.travis.yml

 install:
 - pip install -r requirements.txt
- pip install coveralls
+- pip install coveralls pep8


@hackaugusto Then maybe remove the pep8 tool invocation from here until PyCQA/pycodestyle#502 is merged?

No reason to leave this PR hanging on our side. It does fix a lot of style issues.

removed from travis.yml, kept in tox.ini.

LefterisJP · 2016-11-02T08:42:55Z

@hackaugusto Looks good to me. Needs a rebase though. We can merge afterwards.

- removed pep8 from the travis.yml until PyCQA/pycodestyle#502 is merged. - fixed one additional style issue

schmir · 2016-11-06T20:04:57Z

I've forced a rebuild on travis. The build is fine now. This can be rebased on master without conflicts.

Simplify contract manager

hackaugusto added the dev: Please Review label Oct 19, 2016

hackaugusto force-pushed the pep8 branch from dcb2355 to 91ad9f6 Compare October 19, 2016 04:01

hackaugusto force-pushed the pep8 branch from d32aef6 to 0551fbb Compare October 19, 2016 05:29

LefterisJP suggested changes Oct 19, 2016

View reviewed changes

hackaugusto changed the title ~~Fixed all pep8 errors and added pep8 in the build script~~ Fixed all flake8 errors and added flake8 in the build script Oct 19, 2016

hackaugusto force-pushed the pep8 branch from a5bfdb4 to bfe1e93 Compare October 26, 2016 03:34

LefterisJP approved these changes Oct 28, 2016

View reviewed changes

hackaugusto force-pushed the pep8 branch from bfe1e93 to 7a9ba98 Compare October 31, 2016 22:38

LefterisJP suggested changes Oct 31, 2016

View reviewed changes

LefterisJP approved these changes Nov 2, 2016

View reviewed changes

hackaugusto force-pushed the pep8 branch from 579ee0b to 631755a Compare November 3, 2016 10:34

hackaugusto added 7 commits November 3, 2016 09:48

all files conforming to pep8 tool

6250f94

build running pep8

2e6774d

pep8 max-line-length 99

7578e5e

style changes

51687d8

flake8 fixes

00e2596

running flake8 instead of pep8

5322829

failling build fix

8e2eb08

- removed pep8 from the travis.yml until PyCQA/pycodestyle#502 is merged. - fixed one additional style issue

hackaugusto force-pushed the pep8 branch from 631755a to 8e2eb08 Compare November 3, 2016 11:51

schmir added the dev: Ready To Merge label Nov 6, 2016

Merge branch 'master' into pep8

34d2150

schmir merged commit 32d14eb into raiden-network:master Nov 6, 2016

hackaugusto deleted the pep8 branch November 6, 2016 23:29

hackaugusto removed dev: Please Review labels Nov 6, 2016

err508 pushed a commit that referenced this pull request Aug 22, 2018

Merge pull request #169 from ulope/refactor_contractmgr

c49924a

Simplify contract manager

Conversation

hackaugusto commented Oct 19, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hackaugusto commented Oct 19, 2016

Uh oh!

LefterisJP commented Oct 19, 2016

Uh oh!

LefterisJP commented Oct 19, 2016

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!

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!

hackaugusto commented Oct 19, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hackaugusto commented Oct 19, 2016

Uh oh!

LefterisJP commented Oct 28, 2016

Uh oh!

hackaugusto commented Oct 31, 2016

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

LefterisJP commented Nov 2, 2016

Uh oh!

schmir commented Nov 6, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

hackaugusto commented Oct 19, 2016 •

edited

Loading

hackaugusto commented Oct 19, 2016 •

edited

Loading