Compatability issue with advagg module [#1459502]

The Smarter Styles plugin appears to be conflicting with the Advanced CSS/JS Aggregation (advagg) module. It looks to be a similar issue as with the LESS module: #1231462: Please make Hexagon/Canvas compatible with the LESS module.

I've tried to do a workaround similar to the one in #1231462, but with no luck yet. Any ideas?

Comments

Comment #1

mikeytown2 CreditAttribution: mikeytown2 commented 28 February 2012 at 03:31

This is the file that is causing issues with AdvAgg
http://drupalcode.org/project/hexagon.git/blob/52a67271c650f40560baf8c88...

What you can do is load up a page that is having issues and place a ?advagg-debug=1. This will output a large debug string to watchdog. Paste that below in a code block. From here I can guess at some solutions for this.

Comment #2

dvessel CreditAttribution: dvessel commented 28 February 2012 at 04:11

Thanks for digging through this issue. I’m currently swamped with work so I can’t look at it until the weekend.

Comment #3

mikeytown2 CreditAttribution: mikeytown2 commented 28 February 2012 at 19:48

@dvessel
Long story short, AdvAgg has a bunch of hooks that do similar things that the smarter_styles_re_process_page() function does. Here are the list of hooks http://drupalcode.org/project/advagg.git/blob/refs/heads/6.x-1.x:/README...

Looking at the code seems like you would want to use these 2 hooks
- hook_advagg_css_pre_alter. Modify the raw $variables['css'] before it gets processed. Useful for file replacement.
- hook_advagg_css_extra_alter. Allows one to set the a prefix and suffix to be added into the HTML DOM. Useful for CSS conditionals.

Let me know if you need any changes to AdvAgg or if you need help implementing these hooks :)

Comment #4

rwohleb

he/him

English

CreditAttribution: rwohleb commented 29 February 2012 at 09:32

The workaround hexagon plugin I started writing is already using the approach mikeytown2 mentions in #3, so it's good to know I'm on the right path. I just need to finish debugging it. I'll try and post the debug output requested in #1 as soon as I can.