feat: Allow user-specified local dir to be served by Tornado

[wang-boyu]

This is an alternative implementation of #1428 and #1369, and helps implement a feature in Mesa-Geo: projectmesa/mesa-geo#76.

Some related discussions: #1364 (reply in thread) and #1428 (comment).

[rht]

Probably need a comment explaining what this default behavior is. Is it relative to the directory of user launching the Python code or something else? What if local_dir is precomputed on the fly to be the __file__ of the class definition?

[wang-boyu]

Good point! Based on Tornado's source code: [1] and [2], if local_dir is absolute path such as os.path.dirname(__file__) + "/../templates", then this path will be used.

If it's relative path such as the default "" or ../src/templates, I guess the current working directory is used: see os.path.abspath.

[wang-boyu]

Updated docstring.

[codecov]

Codecov Report

Base: 90.85% // Head: 90.89% // Increases project coverage by +0.03% 🎉

Coverage data is based on head (ed00e25) compared to base (4ce9021).
Patch coverage: 81.81% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1435      +/-   ##
==========================================
+ Coverage   90.85%   90.89%   +0.03%     
==========================================
  Files          15       15              
  Lines        1291     1296       +5     
  Branches      216      217       +1     
==========================================
+ Hits         1173     1178       +5     
  Misses         83       83              
  Partials       35       35              

Impacted Files	Coverage Δ
mesa/visualization/ModularVisualization.py	78.75% <81.81%> (+0.68%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[rht]

LGTM

[rht]

We need this feature in Mesa-Geo. And we have had several features merged since 1.0.0. I think we should release 1.1.0 soon? @jackiekazil @tpike3?

[jackiekazil]

@rht sounds like a plan. I don't mile doing a deploy, but could use help this week putting it together. Can you start up a ticket for the release & help tag anything with the Stafford milestone https://github.com/projectmesa/mesa/milestone/31 that should be included in the history file?

[rht]

IMO, the highlighted changes:

• Perf: ContinuousSpace: speed-up add/remove agents #1376 > 6x perf speedup for add/remove agent in ContinuousSpace
• fix: time: Recompute agent_keys between stages #1391 correctness fix for SimultaneousActivation and StagedActivation
• ModularServer: Always set model.running = True on reset #1399 make self.running = True optional. We need to tell existing users that initializing this is no longer necessary, and so, reducing the boilerplate code
• feat: Allow user-specified local dir to be served by Tornado #1435 (this PR)

[rht]

And #1413, which is a huge feature. I somehow missed this one when going through the commits.

[rht]

I have made #1436, but is there a way to automatically add the issues that have been closed since 1.0.0 to the history file? Adding them to the milestone seems like duplication of work.

[rht]

I'm able to get all the newly closed issues with this search query: https://github.com/projectmesa/mesa/issues?q=is%3Aissue+is%3Aclosed+updated%3A%3E%3D2022-07-07. But these are clearly not representative of the commits that were merged without closing a ticket.

[rht]

I used this search query is:closed updated:>=2022-07-07 and had put all the recently closed issues and PRs to the milestone.

[edgeofchaos42]

Yes 100% we need a release, particularly with the tutorial I need to eliminate the requirement for self.running=True.

Thanks @rht for getting the notes together!!!