Jedi - an awesome Python auto-completion
If you have any comments or feature requests, please tell me! I really want to know, what you think about Jedi.
Jedi is an autocompletion tool for Python. It works. With and without syntax errors. Sometimes it sucks, but that's normal in dynamic languages. But it sucks less than other tools. It understands almost all of the basic Python syntax elements including many builtins.
Jedi suports two different goto functions and has support for renaming. Probably it will also have some support for refactoring in the future.
Jedi uses a very simple interface to connect with IDE's. As an reference, there is a VIM implementation, which uses Jedi's autocompletion. However, I encourage you to use Jedi in your IDEs. Start writing plugins! If there are problems with licensing, just contact me.
At the moment Jedi can be used as a VIM-Plugin. So, if you want to test Jedi for now, you'll have to use VIM. But there are new plugins emerging:
Here are some pictures:
Completion for almost anything (Ctrl+Space).
Display of function/class bodies, docstrings.
Pydoc support (with highlighting, Shift+k).
There is also support for goto and renaming.
Get the latest from github.
You can either include Jedi as a submodule in your text editor plugin (like jedi-vim does it by default), or you can install Jedi systemwide.
The preferred way to install the Jedi library into your system is by using pip:
sudo pip install jedi
If you want to install the current development version:
sudo pip install -e git://github.com/davidhalter/jedi.git#egg=jedi
Note: This just installs the Jedi library, not the editor plugins. For information about how to make it work with your editor, refer to the corresponding documentation.
Jedi supports Python 2.5 up to 3.x. There is just one code base, for both Python 2 and 3. Jedi supports many of the widely used Python features:
- builtin functions/classes support
- complex module / function / class structures
- ignores syntax and indentation errors
- multiple returns / yields
- tuple assignments / array indexing / dictionary indexing
- exceptions / with-statement
- *args / **kwargs
- descriptors -> property / staticmethod / classmethod
- generators (yield statement) / iterators
- support for some magic methods:
- support for list.append, set.add, list.extend, etc.
- (nested) list comprehensions / ternary expressions
- relative imports
- function annotations (py3k feature, are ignored right now, but being parsed. I don't know what to do with them.)
- class decorators (py3k feature, are being ignored too, until I find a use case, that doesn't work with Jedi)
isinstancechecks for if/while/assert
However, it does not yet support (and probably will in future versions, because they are on my todo list):
- manipulations of instances outside the instance variables, without using functions
It does not support (and most probably will not in future versions):
- metaclasses (how could an auto-completion ever support this)
This framework should work for both Python 2/3. However, some things were just not as pythonic in Python 2 as things should be. To keep things simple, some things have been held back:
- Classes: Always Python 3 like, therefore all classes inherit from
- Generators: No
__next__method is used instead.
- Exceptions are only looked at in the form of
Exception as e, no comma!
Syntax errors and other strange stuff, that is defined differently in the Python language, may lead to undefined behaviour of the completion. Jedi is NOT a Python compiler, that tries to correct you. It is a tool that wants to help you. But YOU have to know Python, not Jedi.
numpy can be quite slow sometimes, as well as loading the builtins
the first time. If you want to speed it up, you could write import hooks in
jedi, which preloads this stuff. However, once loaded, this is not a problem
anymore. The same is true for huge modules like
Security is an important issue for Jedi. Therefore no Python code is executed. As long as you write pure python, everything is evaluated statically. But: If you use builtin modules (c_builtin) there is no other option than to execute those modules. However: Execute isn't that critical (as e.g. in pythoncomplete, which used to execute every import!), because it means one import and no more. So basically the only dangerous thing is using the import itself. If your c_builtin uses some strange initializations, it might be dangerous. But if it does you're screwed anyways, because eventualy you're going to execute your code, which executes the import.
A little history
The Star Wars Jedi are awesome. My Jedi software tries to imitate a little bit of the precognition the Jedi have. There is even an awesome scene of Monty Python Jedi's :-).
But actually the name hasn't so much to do with Star Wars. It's part of my second name.
After I explained Guido van Rossum, how some parts of my auto-completion work, he said (we drank a beer or two):
Oh, that worries me
When it's finished, I hope he'll like it :-)
I actually started Jedi, because there were no good solutions available for VIM. Most auto-completions just didn't work well. The only good solution was PyCharm. I just like my good old VIM. Rope was never really intended to be an auto-completion (and also I really hate project folders for my Python scripts). It's more of a refactoring suite. So I decided to do my own version of a completion, which would execute non-dangerous code. But I soon realized, that this wouldn't work. So I built an extremely recursive thing which understands many of Python's key features.
By the way, I really tried to program it as understandable as possible. But I think understanding it might need quite some time, because of its recursive nature.
API-Design for IDEs
If you want to set up an IDE with Jedi, you need to
import jedi. You should
have the following objects available:
Script(source, line, column, source_path)
source would be the source of your python file/script, separated by new
line is the current line you want to perform actions on (starting
with line #1 as the first line).
column represents the current
column/indent of the cursor (starting with zero).
source_path should be the
path of your file in the file system.
It returns a script object that contains the relevant information for the other functions to work without params.
api.Completion objects. Those objects have got
informations about the completions. More than just names.
Similar to complete. The returned
api.Definition objects contain
information about the definitions found.
Mostly used for tests. Like goto, but follows statements and imports and doesn't break there. You probably don't want to use this function. It's mostly for testing.
Returns all names that point to the definition of the name under the cursor. This is also very useful for refactoring (renaming).
Function object of the call you're currently in, e.g.:
with the cursor at the end would return the builtin
If you use the goto function and no valid identifier (name) is at the place of the cursor (position). It will raise this exception.
Sets a callback function for
debug.py. This function is called with
multiple text objects, in python 3 you could insert
Access to the
settings.py module. The settings are described there.