Closed xenore78 closed 2 years ago
Which language server are you using?
It is python-lsp-server 1.2.4.
(base) root@fdf3cce460e1:/# conda list |grep lsp
jupyter-lsp 1.5.1 pyhd8ed1ab_0 conda-forge
jupyter-lsp-python 1.5.1 pyhd8ed1ab_0 conda-forge
jupyterlab-lsp 3.9.3 pyhd8ed1ab_0 conda-forge
python-lsp-black 1.0.0 pyhd3eb1b0_0
python-lsp-jsonrpc 1.0.0 pyhd3eb1b0_0
python-lsp-server 1.2.4 pyhd3eb1b0_0
I edited the original post so now you can see the error messages from the command line.
When I had installed jupyter-lsp before, I did:
conda install -c conda-forge jupyter-lsp-python
But the extension manager of Jupyter lab shows that no jupyter-lsp is installed, so I installed jupyter-lsp again using the manager.
I wondered that this 'unclear' installation made some inconsistency, so I removed all packages related to jupyter-lsp and reinstalled them by following the instruction given in the git page exactly:
conda install -c conda-forge 'jupyterlab>=3.0.0,<4.0.0a0' jupyterlab-lsp
Note: the language server was not removed or reinstalled.
During the removal, the following packages are deleted:
jupyter-lsp-1.5.1-pyhd8ed1ab_0
jupyter-lsp-python-1.5.1-pyhd8ed1ab_0
jupyterlab-lsp-3.9.3-pyhd8ed1ab_0
libgcc-7.2.0-h69d50b8_2
python_abi-3.9-2_cp39
and these are updated
conda-forge::smmap-3.0.5 - 4.0.0
During the reinstallation, the followings are installed
jupyter-lsp-1.5.1
jupyterlab-lsp-3.9.3
python_abi-3.9-2_cp39
After the reinstallation, everything seems working except it is too slow and sometimes stucks. So I installed jedi-language-server using pip. Now jupyter-lsp is running on jedi-language server and no problem until now.
Thank you for getting back to me. I cannot reproduce the issue with the latest python-lsp-server
; it could have been a bug in an old version. In any case it would not be a bug in jupyterlab-lsp
, so I will close this issue now.
Again, thank you for the report and following up.
Description
Auto completion feature stops working if I import two different python modules as the same name. For example, when I run the following codes in a Jupyter notebook,
If I try to invoke the auto completion feature by press 'tab' key repeatedly, some error messages saying 'broken pipe' appear in the log of the Jupyter lab.
Restarting the python kernel does not resolve the problem. It never works again unless I refresh the webpage displaying the Jupyter lab. After the refreshing, auto completion works but when I try to access the duplicated name 'n', it stops again. In this case, error messages in the log are just for jedi when I hit the tab key.
Edited: LSP status indicator on the Jupyter Lab screen always remains green(initialized).
Reproduce
Expected behavior
Context
Edited : language server
Required: installed server extensions
Required: installed lab extensions
Troubleshoot Output
Command Line Output
## output from command line, when the error occurs and I hit 'ctrl-c' to try to stop the server: ``` Traceback (most recent call last): File "/root/anaconda3/lib/python3.9/site-packages/jedi/inference/syntax_tree.py", line 727, in tree_name_to_values types = context.predefined_names[node][tree_name.value] KeyError:Browser Output (recommended for all interface issues)