Factor out a base class from RuntimeFeatureWriter

Factor out a base class from RuntimeFeatureWriter

This base class will be used by other in-file based generators to avoid needing
to copy/paste boilerplate code.

R=eseidel@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=149465

[eseidel, 2013-04-30 21:07:29]

lgtm

I don't think we know enough about what we're going to need to abstract this
yet.  But I'm not going to argue with you.

https://codereview.chromium.org/14668003/diff/1/Source/core/scripts/in_genera...
File Source/core/scripts/in_generator.py (right):

https://codereview.chromium.org/14668003/diff/1/Source/core/scripts/in_genera...
Source/core/scripts/in_generator.py:37: class_name = None
Why is this a class-level property?

https://codereview.chromium.org/14668003/diff/1/Source/core/scripts/in_genera...
Source/core/scripts/in_generator.py:38: defaults = None
This makes more sense as a class-level property.

[abarth-chromium, 2013-04-30 21:09:41]

https://codereview.chromium.org/14668003/diff/1/Source/core/scripts/in_genera...
File Source/core/scripts/in_generator.py (right):

https://codereview.chromium.org/14668003/diff/1/Source/core/scripts/in_genera...
Source/core/scripts/in_generator.py:37: class_name = None
On 2013/04/30 21:07:29, Eric Seidel (Google) wrote:
> Why is this a class-level property?

We might end up making it an instance-level property as we iterate on this code.
 At the moment, RuntimeFeatureWriter always computes this as a constant.  I can
make it a class-level constant in the derived class as well if you like.

[abarth-chromium, 2013-04-30 21:13:29]

Updated to use the class-level constant.

On 2013/04/30 21:07:29, Eric Seidel (Google) wrote:
> I don't think we know enough about what we're going to need to abstract this
> yet.  But I'm not going to argue with you.

Yeah, we'll likely evolve the base class over time.  Having a base class in Perl
was very useful for make_event_factory.pl and make_dom_exceptions.pl. 
Similarly, not having a base class meant that make_names.pl has grown into a
multi-headed monster that generates lots of different kinds of files.

[abarth-chromium, 2013-04-30 21:13:50]

Committed patchset #2 manually as r149465 (presubmit successful).