fixes directory not found error when preferred_dir is set

[andrewfulton9]

References

#12111

Code changes

Made change to filebrowser FileBrowserModel.restore so that if it is restoring a path from a previous session, it cds to the root directory before restoring the path. This fixes the directory not found error thrown when preferred_dir was set and the user was restoring a previous session with a sub-folder of the preferred_dir.

User-facing changes

User will no longer get a directory not found error when restoring a previous session in a sub-folder of their set preferred_dir.

Backwards-incompatible changes

None that I am aware.

[jupyterlab-probot]

Thanks for making a pull request to jupyterlab!
To try out this branch on binder, follow this link:

[echarles]

Tested with the identified failing scenario (both root_dir and preferred_dir set), and now no more error message is shown and the opened notebook is restored. LGTM. Thx @andrewfulton9

[echarles]

@meeseeksdev please backport to 3.3.x

[github-actions]

Benchmark report

The execution time (in milliseconds) are grouped by test file, test type and browser.
For each case, the following values are computed: min <- [1st quartile - median - 3rd quartile] -> max.

The mean relative comparison is computed with 95% confidence.

Results table

Test file	large_code_notebook	large_md_notebook
open
chromium
actual	4454 <- [4624 - 4678 - 4736] -> 5111	2940 <- [3035 - 3076 - 3144] -> 3355
expected	4374 <- [4559 - 4620 - 4681] -> 5268	2878 <- [3011 - 3054 - 3102] -> 3285
Mean relative change	1.0% ± 0.7%	1.4% ± 0.8%
switch-from
chromium
actual	694 <- [755 - 783 - 818] -> 1068	526 <- [560 - 572 - 584] -> 1216
expected	692 <- [737 - 763 - 825] -> 1235	515 <- [549 - 565 - 584] -> 1270
Mean relative change	0.6% ± 3.5%	-0.4% ± 4.5%
switch-to
chromium
actual	364 <- [417 - 436 - 455] -> 609	293 <- [339 - 349 - 359] -> 407
expected	357 <- [402 - 425 - 441] -> 613	301 <- [325 - 341 - 354] -> 384
Mean relative change	2.8% ± 2.8%	2.9% ± 1.7%
close
chromium
actual	684 <- [759 - 1085 - 1114] -> 1237	546 <- [577 - 592 - 607] -> 639
expected	669 <- [722 - 1048 - 1088] -> 1209	556 <- [572 - 588 - 601] -> 629
Mean relative change	7.5% ± 5.8%	0.9% ± 0.9%

Changes are computed with expected as reference.

[echarles]

@jtpio @fcollonval @meeseeksdev please backport to 3.3.x is completely silent and does not generate the expected PR. Is it a temporary issue?

[jtpio]

Maybe Meeseeks is down at the moment. We can give it another try and otherwise report on the repo.

Looks like this issue might be related: scientific-python/MeeseeksDev#59

[jtpio]

@meeseeksdev please backport to 3.3.x

[echarles]

@messeeksdev is back up and running. thx @jtpio