Add `jupyter-server-proxy`, `dask-labextension` and enable
`panel.io.jupyter_server_extension`.

The `jupyter-server-proxy` allows to proxy URL like
"http://127.0.0.1:8787/status" (dask dashboard URL sample) to
"https://pavics.ouranos.ca/jupyter/user-redirect/proxy/8787/status"
which the `dask-labextension` will accept.

http://127.0.0.1:8787/status
is mapped to
https://pavics.ouranos.ca/jupyter/user-redirect/proxy/8787/status

For "panel serve", we can do it this way:

`! panel serve --allow-websocket-origin=pavics.ouranos.ca --port 5007 /notebook_dir/writable-workspace/tmp-test/test-panel.ipynb`

2022-04-08 01:33:50,268 Starting Bokeh server version 2.4.2 (running on Tornado 6.1)
2022-04-08 01:33:50,270 User authentication hooks NOT provided (default user enabled)
2022-04-08 01:33:50,272 Bokeh app running at: http://localhost:5007/test-panel
2022-04-08 01:33:50,272 Starting Bokeh server with process id: 2155
2022-04-08 01:34:06,179 WebSocket connection opened
2022-04-08 01:34:06,179 ServerConnection created

Then open a new browser tab to https://pavics.ouranos.ca/jupyter, login
again if it asks, and to go this URL:
https://pavics.ouranos.ca/jupyter/user-redirect/proxy/5007/test-panel

http://localhost:5007/test-panel
is mapped to
https://pavics.ouranos.ca/jupyter/user-redirect/proxy/5007/test-panel

…serve display url

Found solution here
https://discourse.holoviz.org/t/is-it-possible-to-deploy-and-run-panel-apps-via-jupyter-hub/189

Not sure if it works.

… testing

…anty reproducibility

… launch dir

… launch dir

```
! conda install -c pyviz/label/dev jupyter-panel-proxy

Collecting package metadata (current_repodata.json): failed

NotWritableError: The current user does not have write permissions to a
required path.
  path: /opt/conda/pkgs/cache/37adb69c.json
  uid: 1000
  gid: 1000

If you feel that permissions on this path are set incorrectly, you can
manually
change them by executing

  $ sudo chown 1000:1000 /opt/conda/pkgs/cache/37adb69c.json

In general, it's not advisable to use 'sudo conda'.
```

Fix for this error:
```
! conda install -c pyviz/label/dev jupyter-panel-proxy

Collecting package metadata (current_repodata.json): failed

NotWritableError: The current user does not have write permissions to a
required path.
  path: /opt/conda/pkgs/cache/37adb69c.json
  uid: 1000
  gid: 1000

If you feel that permissions on this path are set incorrectly, you can
manually
change them by executing

  $ sudo chown 1000:1000 /opt/conda/pkgs/cache/37adb69c.json

In general, it's not advisable to use 'sudo conda'.
```

…almost 2 years old

…isting executable only

The conda install step did not work, error below:

```
Step 10/13 : RUN conda install -c pyviz/label/dev jupyter-panel-proxy
 ---> Running in 3bf123abf1a8
Collecting package metadata (current_repodata.json): ...working... done
Solving environment: ...working... failed with initial frozen solve.
Retrying with flexible solve.
Solving environment: ...working... failed with repodata from
current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): ...working... done
Solving environment: ...working... failed with initial frozen solve.
Retrying with flexible solve.
Examining @/linux-64::__linux==5.4.170=0:  17%|█▋        | 1/6
[00:00<00:00, 3591.01it/Examining @/linux-64::__glibc==2.31=0:  33%|███▎
| 2/6 [00:00<00:00, 4232.40it/s] Examining
@/linux-64::__archspec==1=x86_64:  83%|████████▎ | 5/6 [00:00<00:00,
38.59it/
Examining conflict for jupyter-panel-proxy python:  17%|█▋        | 1/6
[00:00<00:00, 9
Found conflicts! Looking for incompatible packages.
```

…blem

See comment holoviz/panel#3440 (comment)

To hopefully fix this error:
```
ServerApp - ERROR - Uncaught exception GET
/panel-preview/render/workdir/PAVICS-e2e-workflow-tests/binder/test-notebooks/TEST-panel-servable.ipynb
(10.10.10.7)
HTTPServerRequest(protocol='http', host='lvu.ouranos.ca:8890',
method='GET',
uri='/panel-preview/render/workdir/PAVICS-e2e-workflow-tests/binder/test-notebooks/TEST-panel-servable.ipynb',
version='HTTP/1.1', remote_ip='10.10.10.7')
Traceback (most recent call last):
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/tornado/web.py", line
1704, in _execute
    result = await result
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/panel/io/jupyter_server_extension.py",
line 116, in get
    page = server_html_page_for_session(
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/bokeh/embed/server.py",
line 255, in server_html_page_for_session
    bundle = bundle_for_objs_and_resources(None, resources)
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/bokeh/embed/bundle.py",
line 211, in bundle_for_objs_and_resources
    js_files.extend(js_resources.js_files)
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/panel/io/resources.py",
line 299, in js_files
    dist_dir = self.dist_dir
  File
"/opt/conda/envs/birdy/lib/python3.8/site-packages/panel/io/resources.py",
line 259, in dist_dir
    if self.absolute:
AttributeError: 'Resources' object has no attribute 'absolute'
```

Previous docker build error:
```
Encountered problems while solving:
- package clisops-0.9.0-pyh6c4a22f_0 requires cf_xarray >=0.7.0, but none of the providers can be installed
```

cf_xarray was previously pinned because of climex.ipynb (not
WPS_example.ipynb, my mistake reading the Jenkins output) performance
but it was already happening before the pin so cf_xarray might not be
the real culprit.

See issue about cf_xarray pinning here
#100

…utton

See comment holoviz/panel#3440 (comment)

Remove this patch once a Panel release contain this PR
holoviz/panel#3469

…s it

…on fix