WIP: Use Jedi as the underlying parser #240
Conversation
…import problems) # Conflicts: # src/pydocstyle/checker.py # src/pydocstyle/parser.py # src/tests/test_definitions.py
…official API so that's my fault).
|
This is still in the works, but I would appreciate any thoughts or comments from anyone who has a bit of time (esp. |
|
I haven't looked at the code at all. My first question, however, is how does a Jedi AST differ from a Python stdlib AST? I ask specifically because Flake8 defaults to using a stdlib AST. It provides that parsed AST (doing it only once) to each plugin that traverses an AST. So if pydocstyle develops its own Flake8 plugin, it'll need to know how to take a stdlib AST and use that instead of a Jedi AST (or we'll have to work on converting Flake8 to using Jedi ASTs). |
sigmavirus24
left a comment
sigmavirus24
left a comment
There was a problem hiding this comment.
I'll probably have to do a few more passes to give you good feedback on the parser module rewrite.
setup.py
Outdated
| packages=('pydocstyle',), | ||
| package_dir={'': 'src'}, | ||
| install_requires=[ | ||
| 'jedi==0.9.0', |
There was a problem hiding this comment.
Please, let's not pin this. This causes unending headaches for users and downstream redistributors.
There was a problem hiding this comment.
Eventually this will not be pinned. I temporarily pinned this because a new jedi was released while I was working on this, which broke some of my changes.
src/__init__.py
Outdated
| @@ -1,0 +1 @@ | |||
| """Pydocstyle - a docstring style checker.""" | |||
There was a problem hiding this comment.
Is src/ meant to be a package? If yes, why? To be clear, I don't think src/ should be a package.
| source = file.read() | ||
| try: | ||
| source = source.decode('utf-8') | ||
| except: |
There was a problem hiding this comment.
I think this except should be more narrowly scoped. There are a surprising number of exceptions that can happen here and they depend on platform, how the python interpreter is compiled, etc.
src/tests/__init__.py
Outdated
| @@ -1,0 +1 @@ | |||
| """Test for pydocstyle.""" | |||
There was a problem hiding this comment.
py.test does not require an __init__.py and often recommends you don't make your tests module importable.
| return {True: 'public', False: 'private'}[self.is_public] | ||
|
|
||
| @property | ||
| def source_lines(self): |
There was a problem hiding this comment.
Will _source ever be in _fields now? Can't you just make this property a proxy for self.parent.source_lines now?
|
|
||
|
|
||
| class DunderAll(Value): | ||
| """A list of exported names in a module.""" |
There was a problem hiding this comment.
>>> 'names'.split()
['names']Why not simply do?
_fields = ['names']|
|
||
|
|
||
| class FutureImport(Value): | ||
| """A list of __future__ imports in a module.""" |
src/pydocstyle/parser.py
Outdated
| skipped_error_codes = ''.join( | ||
| self.current.value.split('noqa: ')[1:]) | ||
| elif self.current.value.startswith('# noqa'): | ||
| skipped_error_codes = 'all' |
There was a problem hiding this comment.
If I stripped the noqa logic into a separate library, would you use that instead?
| source = file.read() | ||
| try: | ||
| source = source.decode('utf-8') | ||
| except: |
There was a problem hiding this comment.
To catch any exception use the except Exception, instead of except BaseException or just except.
Indicate the most specific exception type in except block.
For example:
# Bad
try:
mapping[key]
except Exception:
...
# Better
try:
mapping[key]
except KeyError:
...…e pinning before merge).
|
Closing this as it has been inactive for a while. Please feel free to re-open if this is still relevant. |
sambhav
added
the
STALE-PR
4 participants
WIP, description to come.