config: Allow equal signs (=) between variable names and value

[Moguri]

This allows for more INI-style syntax as proposed by #523. Sections are
still not supported.

[codecov]

Codecov Report

Merging #926 into master will increase coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #926      +/-   ##
==========================================
+ Coverage   14.99%   15.01%   +0.01%     
==========================================
  Files        3701     3701              
  Lines      355662   355668       +6     
==========================================
+ Hits        53339    53390      +51     
+ Misses     302323   302278      -45     

Impacted Files	Coverage Δ
dtool/src/prc/configPage.cxx	62.43% <100.00%> (+3.64%)	⬆️
panda/src/cocoadisplay/cocoaPandaApp.mm	0.00% <0.00%> (-29.17%)	⬇️
panda/src/cocoadisplay/cocoaGraphicsWindow.mm	6.74% <0.00%> (-0.37%)	⬇️
panda/src/putil/bamCache.cxx	34.45% <0.00%> (+0.17%)	⬆️
dtool/src/prc/configDeclaration.cxx	52.77% <0.00%> (+0.92%)	⬆️
dtool/src/prc/notifyCategory.I	47.82% <0.00%> (+4.34%)	⬆️
dtool/src/prc/configDeclaration.I	59.75% <0.00%> (+6.09%)	⬆️
dtool/src/prc/configVariableCore.cxx	58.20% <0.00%> (+14.81%)	⬆️
dtool/src/prc/configFlags.cxx	60.00% <0.00%> (+60.00%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9f1289b...9cfc090. Read the comment docs.

[rdb]

Interesting. I think this PR would allow var = = value, though?

I think the is_split_character, which is not declared externally, should be marked static (since it's not declared elsewhere, no point in giving it external linkage), and hoisted to the top of the .cxx file by Panda convention.

Do we want to enable this optionally, as part of a special mode for now (possibly toggled by extension), or do we want to allow this syntax for all .prc files?

[Moguri]

Yes, it allows any combination of spaces and equal signs. I figured this would be the simplest approach and PRC isn't very strict as is (e.g., anything start with "t" being "true").

I'll fix up the scope of is_split_character() and move it.

[Moguri]

I forced pushed a new version that only allows one equal sign.

[Moguri]

@rdb Is this PR something we want? What is missing to move this along? More discussion?

[rdb]

One question I raised for discussion remains unaddressed, and I'd like to carefully consider that before we make this otherwise irreversible change.

[Moguri]

I am going to abandon this PR for now and we can continue discussion in #523. I started this PR in the hopes of quickly getting an issue closed, but I am otherwise not strongly invested in the feature. The code will still be around for someone to pick up.