FIX: Escape colon in JupyterHub link to repo #556
Conversation
|
Thanks for submitting your first pull request! You are awesome! 🤗 |
I've run into a problem using The Littlest JupyterHub with current `sphinx-book-theme`. At the moment the generated URLs for JupyterHub are of form: ``` <hub-url>/hub/user-redirect/git-pull?repo=https://<repo>&urlpath=tree/<nb_fname> ``` These work fine on a modern Kubernetes JupyterHub, I guess because of a recent version of JupyterHub responding to the query. But, on current The Littlest JupyterHub, still on JupyterHub 1.5.0, this causes Nbgitpuller to stall if the user is not already logged in, with log messages of form: ``` Disallowing redirect outside JupyterHub: '/hub/user-redirect/git-pull?repo=https://github.com/<repo>&urlpath=tree/<nb_fname>'. ``` Once logged in, retrying the same link works correctly. This can be fixed by escaping the `:` in the Nbgitpuller URL with a `%3A`, in this PR.
for more information, see https://pre-commit.ci
|
I've rebased. I tested a modern Kubernetes setup by replacing the I see that a few tests will need fixing if this is the right way to go. |
Replace : with %3A in test files, to match escaping behavior. Note: the file `test_topbar_launchbtns.html` does not appear to have tests that check the URL.
|
It's not realistic for us to add tests for JupyterHubs specifically, so I think it's fine to use this PR as confirmation that "it works" and just get the tests into a working state (my guess is the regression tests will need fixing - to update those we can just replace the relevant launch links w/ the escaped version) |
|
Chris - you probably saw - but I updated the files for the regression tests. Tests now passing. |
| @@ -118,9 +118,10 @@ def add_launch_buttons( | |||
| ) | |||
|
|
|||
| if jupyterhub_url: | |||
| repo_esc = repo_url.replace(":", "%3A") | |||
There was a problem hiding this comment.
A more general replacement is
from urllib.parse import quote
repo_esc = quote(repo_url)which will catch all characters that need encoding, not just :.
This will happen automatically if you use url-building functions to build the query, rather than by hand:
from urllib.parse import urlencode
url_params = urlencode(dict(repo=repo_url, urlpath=f"{ui_pre}/{repo}/{path_rel_repo}", branch=branch), safe="/")
url = f"{jupyterhub_url}/hub/user-redirect/git-pull?{url_params}"There was a problem hiding this comment.
Aha - I was missing the safe="/" kwarg - that makes all the difference to readability.
There was a problem hiding this comment.
it does! I don't know why safe="/" is the default for urllib.parse.quote, but not urllib.parse.urlencode, but it should be fine for our purposes here.
There was a problem hiding this comment.
@matthew-brett wanna give this a shot so that our implementation is a bit less special-casey and more robust to other character wackiness?
Generalize URL encoding using urlencode function.
for more information, see https://pre-commit.ci
choldgraf
changed the title
|
Congrats on your first merged pull request in this project! 🎉 |
I've run into a problem using The Littlest JupyterHub with current
sphinx-book-theme. At the moment the generated URLs for JupyterHub areof form:
These work fine on a modern Kubernetes JupyterHub, I guess because of
a recent version of JupyterHub responding to the query.
But, on current The Littlest JupyterHub, still on JupyterHub 1.5.0, this
causes Nbgitpuller to stall if the user is not already logged in, with
log messages of form:
Once logged in, retrying the same link works correctly.
This can be fixed by escaping the
:in the Nbgitpuller URL with a%3A, in this PR.