Automatically download labvariables CSS instead of shipping them

[martinRenou]

Remove labvariables.css and materialcolors.css files from the repo. Instead we download them automatically, similar to what nbconvert does: https://github.com/jupyter/nbconvert/blob/main/setup.py

Code changes

Automatically download CSS files on install

User-facing changes

None

Backwards-incompatible changes

None

[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.

Results table

Test file	basics.ipynb	bqplot.ipynb	dashboard.ipynb	gridspecLayout.ipynb	interactive.ipynb	ipympl.ipynb	ipyvolume.ipynb	multiple_widgets.ipynb	query-strings.ipynb	reveal.ipynb
Render
chromium
actual	3526 <- [3671 - 3921 - 4190] -> 5384	2889 <- [2948 - 2961 - 2976] -> 3418	3155 <- [3211 - 3213 - 3256] -> 3528	4161 <- [4190 - 4201 - 4381] -> 4541	2361 <- [2376 - 2382 - 2480] -> 2723	4782 <- [4838 - 4958 - 5173] -> 5553	8240 <- [11823 - 12077 - 13010] -> 13386	12915 <- [13096 - 13202 - 13248] -> 13316	1598 <- [1606 - 1681 - 1772] -> 2004	2547 <- [2553 - 2561 - 2568] -> 2798
expected	3379 <- [3442 - 3517 - 3701] -> 3876	2976 <- [3227 - 3321 - 3421] -> 3604	3608 <- [3623 - 3709 - 3793] -> 3825	4453 <- [4453 - 4523 - 4661] -> 4748	2559 <- [2655 - 2656 - 2660] -> 2674	3982 <- [4079 - 4213 - 4356] -> 4743	12183 <- [18509 - 19553 - 20811] -> 21515	15319 <- [15660 - 15796 - 15912] -> 16056	1517 <- [1920 - 1997 - 2103] -> 2113

❗ Test metadata have changed

--- /dev/fd/63	2022-01-18 14:41:00.313419431 +0000
+++ /dev/fd/62	2022-01-18 14:41:00.313419431 +0000
@@ -4,37 +4,37 @@
     "BENCHMARK_REFERENCE": "actual"
   },
   "browsers": {
-    "chromium": "97.0.4666.0"
+    "chromium": "94.0.4595.0"
   },
   "systemInformation": {
     "cpu": {
-      "brand": "Xeon® E5-2673 v4",
+      "brand": "Xeon® E5-2673 v3",
       "cache": {
         "l1d": 65536,
         "l1i": 65536,
         "l2": 524288,
-        "l3": 52428800
+        "l3": 31457280
       },
       "cores": 2,
       "family": "6",
-      "flags": "fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt md_clear",
+      "flags": "fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm invpcid_single pti fsgsbase bmi1 avx2 smep bmi2 erms invpcid xsaveopt md_clear",
       "governor": "",
       "manufacturer": "Intel®",
-      "model": "79",
+      "model": "63",
       "physicalCores": 2,
       "processors": 1,
       "revision": "",
       "socket": "",
-      "speed": 2.3,
+      "speed": 2.4,
       "speedMax": null,
       "speedMin": null,
-      "stepping": "1",
+      "stepping": "2",
       "vendor": "GenuineIntel",
       "virtualization": false,
       "voltage": ""
     },
     "mem": {
-      "total": 7289610240
+      "total": 7291699200
     },
     "osInfo": {
       "arch": "x64",
@@ -42,11 +42,11 @@
       "codename": "Focal Fossa",
       "codepage": "UTF-8",
       "distro": "Ubuntu",
-      "kernel": "5.11.0-1025-azure",
+      "kernel": "5.8.0-1040-azure",
       "logofile": "ubuntu",
       "platform": "linux",
       "release": "20.04.3 LTS",
-      "serial": "bfb5cc16a13d479a868c817b847bcf8f",
+      "serial": "cfc067bfcb844f35865e279a1b0e66c5",
       "servicepack": "",
       "uefi": false
     }

[martinRenou]

I don't get the UI test failure. It looks like a font-size difference but the font-size has not been changed using this PR.

This CSS variables are still properly discovered by the reveal template.

I will update the reference screenshots.

[trungleduc]

Hi @martinRenou, by fetching CSS files from the internet on installation, users who install Voila in an offline environment (e.g directly via tar.gz file or via local pypi repository) will not be able to do it anymore.
Should we instead download the CSS files at build time via CI script to keep them up-to-date with upstream?
My bad, this PR actually downloads CSS files at build time, which is great!

[martinRenou]

users who install Voila in an offline environment (e.g directly via tar.gz file or via local pypi repository) will not be able to do it anymore.

I am not sure this is a use-case we actually support right now? Those users would also need to install nbconvert in some way, and nbconvert does the same as this PR proposes. (EDIT: this is not true, see #1062 (comment))

Should we instead download the CSS files at build time via CI script to keep them up-to-date with upstream?

Another idea could be to include those CSS files in the sdist when publishing on pypi?

[martinRenou]

My bad, this PR actually downloads CSS files at build time, which is great!

No I think you had a point. This CSS, not being in the sdist, will not be part of the tar.gz published on Pypi.

[trungleduc]

My bad, this PR actually downloads CSS files at build time, which is great!

No I think you had a point. This CSS, not being in the sdist, will not be part of the tar.gz published on Pypi.

I looked at nbconvert, both the tag.gz and .whl have the downloaded CSS files.

[martinRenou]

I looked at nbconvert, both the tag.gz and .whl have the downloaded CSS files.

With this PR in voila, it's not included in the tar.gz, it is included in the whl though.

It's not included in the tar.gz because I added an exclude in the MANIFEST.IN (which nbconvert does not do). We could remove this exclude and ship the CSS with the tar.gz (I am personally totally ok with this, it would fix the "non internet connection" setup), but the check-manifest CI check is complaining if there are files included in the sdist that are not tracked by git.

[trungleduc]

Can we update tool.check-manifest ignore list?

[martinRenou]

Yes, we can do that. I wonder if that could be considered bad practice?

[trungleduc]

Yes, we can do that. I wonder if that could be considered bad practice?

I don't know, but we had the same pattern for ignoring js files in check-manifest.

[martinRenou]

Ok, thanks :) I updated my PR

[martinRenou]

I accidentally canceled the CI, retriggering it

[trungleduc]

Thanks @martinRenou !