What existing works are there

[anthonykulis]

Just curious to see what open source plugins exist related to this topic. This includes variants that do not solve our problem but may provide insights to solving it, e.g., a handlebars plugin. Compile a list with pros and cons about plugins. Remember, we can leverage other open source plugins to minimize our future work, so make sure to read their source. Is it set up such a way that if we were to leverage their plugin? If so, how, if not, and it is a good plugin, what work would it take to port it over? These are things I seek.

[ghost]

Open Source Plugins Mentioned in #9

AAAPackageDev

AAAPackageDev is a Sublime Text 2 and 3 plugin that helps create and edit syntax definitions, snippets, completions files, build systems and other Sublime Text extensions.

Pros

• Easily installed through Package Control.
• Abstracts out the minutiae of creating syntax definitions - no need to reinvent the wheel.
• Easy tutorial specifically for AAAPackageDev here.
• Even supports completion.

Cons

• Abstraction does not lend well to our understanding of our own project.
• If used, our project is extremely dependent on this package.

Other Thoughts

By default, AAAPackageManager transcompiles YAML to Plist. This is because Sublime Text is dependent on Plist files for its syntax definitions, and the developers didn't like working with Plist. So, if we want to use YAML, this is a pro. If we want to use something else, it's a con and may result in some "hackery," as they say. I will look into reconciling another transcompilation pathway if there is strong opposition to YAML.

Aroma

Command line utility to compile CoffeeScript objects into property list files.

Pros

• By itself, extremely easy to install and use.

Cons

• Ease of reconciliation with the use of AAAPackageDev is questionable.

Other Thoughts

It's not really a plugin so much as it is a helper command line tool. It's dependent on Node.js.

ScopeHunter

This is a simple plugin get [sic] the scope under the cursor(s) in Sublime Text. This is useful for plugin development.

Pros

• Just a helper plugin to aid in our package development.
• Should be extremely useful when troubleshooting scopes and their inheritance.

Cons

• None - if it doesn't work, or if it breaks, we just won't use it.

BracketHighlighter

Bracket Highlighter matches a variety of brackets such as: [], (), {}, "", '', , and even custom brackets.

I'll leave this one up to @abandimh01 since he's been assigned #11. I believe this package has major implications for Handlebars in Sublime Text, but I haven't researched it in breadth or depth.

Final Notes

I'm curious to see what the others will bring to this discussion. If any of the above mentioned plugins (or tools) sound at all viable, I'll get my hands dirty with them, and we'll find out just how viable they are in practice.

[anthonykulis]

By pros/cons, if the similar work deals say in hamlbars, what works, what doesn't, can ours be a better version. Means you will have to use these in the evaluation.

[ghost]

[DOCUMENTATION]

Similar Work to Hamlbars

Handlebars as Introduced in #11

Pros

• Considering this plugin has been around for two years, it's far more well-rounded and tested than anything we'll come up with by alpha.
• Consistent and varied support - spans commits from multiple developers.
• Could be easily forked into Hamlbars, and then we could programmatically add the existing support for Haml in Sublime Text 3. (Haml has a .tmLanguage file associated with it that we can use.)
• Properly formatted for Package Control - meaning, if forked, our plugin would be easily formatted for Package Control.
• Even if we didn't fork this project, it's so similar to what we would do that I consider it an invaluable resource - it implements everything Handlebars-related we were going to try to do ourselves.

Cons

• Predominately supported by one developer.
• Dependent on AAAPackageDev for converting from json to plist, but we were probably going to use that anyway.
• May have hiccups with corner cases specific to Greplytix, and therefore we need to emphasize on-site testing.

Conclusion

Ultimately, if we can leverage this project like I mentioned in the pros, then we could get away with writing exceptionally little code, and just glue code at that, for Hamlbars.