VimPlug/jedi-fork
1
0
Fork 0
forked from VimPlug/jedi
Code Pull Requests Activity
4,825 Commits 3 Branches 31 Tags
a83b43ccfdd07da91e06dfee87cc8f6d7d1ca5c8
Commit Graph

3198 Commits

Author SHA1 Message Date
Dave Halter
a83b43ccfd Merge branch 'dev' of github.com:davidhalter/jedi into dev 2017-03-15 19:12:51 +01:00
Dave Halter
93f14157a6 Cleanup the ParseError stuff. 2017-03-15 18:41:58 +01:00
Dave Halter
0effd348e8 Add a note about the grammar creation. 2017-03-15 18:18:06 +01:00
Dave Halter
c332fba488 Fix a namespace packages related issue. 2017-03-15 08:59:24 +01:00
Dave Halter
375749c5c3 Small restructuring. 2017-03-15 08:56:49 +01:00
Dave Halter
55c9fd3227 Fix an issue in the fake parser 2017-03-15 08:44:49 +01:00
Matthias Bussonnier
9a851165ad Lookup optimisation 
avoid a couple of dynamic lookup in the core of the parsing loop.
Performance improvement will be minimal, but adding a little many time
can be consequent.
 2017-03-14 23:55:03 +01:00
Matthias Bussonnier
bb8fe0b24c A missing docstring. 2017-03-14 23:54:38 +01:00
Dave Halter
68a7365a0a A few docstring improvements. 2017-03-14 21:03:03 +01:00
Dave Halter
9efb3f0af2 More direct parser usage removals. 2017-03-14 19:31:54 +01:00
Dave Halter
717bfeb574 Remove an occurance of the complicated parser creation. 2017-03-14 19:27:03 +01:00
Dave Halter
97fc3bc23c Refactored the parser calls. Now it's possible to use jedi.parser.python.parse to quickly parse something. 2017-03-14 00:38:58 +01:00
Dave Halter
9b5e6d16da Improved grammar loading API. 2017-03-13 20:33:29 +01:00
Dave Halter
595ffc24d4 Move some more stuff to a python directory in the parser. 2017-03-13 00:54:39 +01:00
Dave Halter
922c480e2e Moved the parser to a new file. 2017-03-12 21:33:41 +01:00
Dave Halter
a635b6839a Remove unused code. 2017-03-12 21:28:32 +01:00
Dave Halter
af9b0ba8d6 Merge branch 'master' into dev 2017-03-12 20:51:17 +01:00
Alex Wiltschko
82d165a723 Missing paren 2017-03-12 20:41:17 +01:00
Dave Halter
a7b1e3fe70 Fixed another diff parser error. 2017-03-12 15:58:14 +01:00
Dave Halter
6e3b00802c Another endless while loop issue, add an assert. 2017-03-11 14:54:44 +01:00
Dave Halter
818fb4f60c Fix a bug that might have caused an endless while loop a while ago. Fixes #878. 2017-03-09 21:47:16 +01:00
Dave Halter
ccef008376 Python 2 compatibility. 2017-03-09 21:06:20 +01:00
Dave Halter
c7a74e6d1c Make the tokenizer a generator. 2017-03-09 18:53:09 +01:00
Dave Halter
989e4bac89 Speed up splitlines. 
We use the python function again with the modifications we need.
I ran it with:

    python3 -m timeit  -n 10000 -s 'from jedi.common import splitlines; x = open("test_regression.py").read()'

The speed differences are quite remarkable, it's ~3 times faster:

    10000 loops, best of 3: 52.1 usec per loop

vs. the old:

    10000 loops, best of 3: 148 usec per loop

We might need to speedup splitlines with  as well. It's probably
also a factor 2-3 slower than it should be.
 2017-03-09 08:58:57 +01:00
Dave Halter
b814a91f29 Avoid endless loop with an assert in the diff parser. 2017-03-08 18:33:38 +01:00
Dave Halter
5c9769c5a3 Merge remote-tracking branch 'origin/master' into dev 2017-03-07 19:01:53 +01:00
Dave Halter
ee98eab64c Merge branch 'dev' of github.com:davidhalter/jedi into dev 2017-03-07 19:01:42 +01:00
Dave Halter
05e05252fa Make python -m jedi.parser.tokenize possible for debugging purposes. 2017-03-07 18:31:12 +01:00
micbou
a859add6d7 Only resolve names for actual modules 
A name can be part of an import statement without being a module.
 2017-03-01 21:06:21 +01:00
Matthias Bussonnier
fc27ca1b6a 'fix a couple of error locations' 2017-02-24 13:03:03 +01:00
Matthias Bussonnier
0fb386d7e2 Make sure error token set the new_line flag when necessary 
Should solve #855
 2017-02-24 00:05:38 +01:00
daniel
5513f72987 added support for implicit ns packages and added tests 2017-02-23 23:53:14 +01:00
Matthias Bussonnier
ef1b1f41e4 Use start and end position in repr for simpler debugging. 2017-02-14 23:53:53 +01:00
Mathias Rav
c809aad67f Complete dotted names properly in decorators 
Outside decorators, a dotted name is parsed by the grammar as
stmt -> test -> power -> atom trailer -> (NAME) ('.' NAME)
where the first NAME is an 'atom' and the second NAME is a 'trailer'.

Thus, testing if the most recent variable is a 'trailer' and the most
recent node is a '.' is almost always enough for Jedi to properly
complete dotted names.

However, the grammar for decorators is more restricted and doesn't allow
arbitrary atoms and trailers; instead, a dotted name in a decorator is
decorator -> '@' dotted_name -> '@' (NAME '.' NAME),
meaning the most recent variable will be 'dotted_name', not 'trailer'.

Besides in decorators, the 'dotted_name' variable is only used in import
statements which are handled previously in _get_context_completions,
so checking for 'dotted_name' in this arm of the if only covers decorators
and not inadvertently anything else.
 2017-02-04 18:05:15 +01:00
Dave Halter
1d64a5caa1 Replace first_leaf and last_leaf with get_first_leaf and get_last_leaf. 2017-02-03 17:35:53 +01:00
Dave Halter
90fffd883e Clean up the function docstring. 2017-02-03 17:30:58 +01:00
Dave Halter
647aec11a6 Return None instead of raising an IndexError in get_next_leaf. 2017-02-03 17:26:02 +01:00
Dave Halter
c5071f9f49 Change get_previous_leaf to return None if there is no leaf anymore (at the start of the file). 2017-02-03 17:23:15 +01:00
Dave Halter
445bf6c419 Get rid of get_parent_until. 2017-02-03 09:59:32 +01:00
Dave Halter
b3cb7b5490 Remove the def isinstance from the parser. It was a really bad pattern. 2017-02-03 09:37:59 +01:00
Dave Halter
f2b41b1752 Update already the version number so we don't forget it. 2017-02-03 00:38:08 +01:00
Dave Halter
8027aeebd8 Fix a small bug that was raised two commits ago (in the test suite). 2017-02-02 23:50:29 +01:00
Dave Halter
f627d541b8 Trying to fix the docs. 2017-02-02 23:39:10 +01:00
Dave Halter
a5a54fbc85 Fix a call signature issue. 2017-02-01 19:21:07 +01:00
Dave Halter
68a3a9cf41 Don't do anything with the diff parser if nothing changes. 2017-01-29 22:12:24 +01:00
Dave Halter
e5af996829 Remove old debug code from the diff parser. 2017-01-29 21:57:09 +01:00
Dave Halter
e0c8b3dd3b Fix an context issue in completions. 2017-01-29 19:09:35 +01:00
Dave Halter
3f1e658e1d Set the limit for debugging failed diff parsers a bit higher. 2017-01-29 14:34:20 +01:00
Dave Halter
40b6079ebd If an INDENT is the next supposed token, we should still be able to complete. 2017-01-29 14:06:22 +01:00
Dave Halter
b0f340748c So much work for one simple diff fail. 2017-01-29 00:42:09 +01:00