More changes to extension omnibox API for styles.

More changes to extension omnibox API for styles.
- Overlapping styles are now bitwise-ORed together (instead of replacing
previous ones).
- The "length" parameter for style ranges is optional and can be 0.

BUG=63077, 63078
TEST=covered by tests

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=66327

[asargent_no_longer_on_chrome, 2010-11-16 19:05:02]

lgtm

http://codereview.chromium.org/5064001/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_omnibox_api.cc (right):

http://codereview.chromium.org/5064001/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_omnibox_api.cc:146: if
(!style->GetInteger(kDescriptionStylesLength, &length) || length < 0)
It seems like passing a negative length value is likely to indicate a bug in the
calling extension's arithmetic. Would it make more sense to either return an
error or just ignore this style (ie treat it the same as length 0)? Your call.

On a related note, it might actually be useful to be able to send negative
values for the *offset* - in python I find it's often convenient to use negative
indices on strings and arrays to do things like "give me the last element in
this array" or "give me the last 5 characters from this string".

http://codereview.chromium.org/5064001/diff/1/chrome/common/extensions/api/ex...
File chrome/common/extensions/api/extension_api.json (right):

http://codereview.chromium.org/5064001/diff/1/chrome/common/extensions/api/ex...
chrome/common/extensions/api/extension_api.json:3931: "description": "An array
of style objects, created using styleUrl, styleMatch, or styleDim. A style
applies to the region of text specified by the style's starting offset and
length. If there are any overlapping regions of text covered by multiple styles,
they will be added together (e.g. 'match' + 'dim' will display a dimmed match).
Not all style combinations will be visually distinct.",
nit: I'm not quite sure what you mean by "Not all style combinations will be
visually distinct" - perhaps an example would illuminate?

[Matt Perry, 2010-11-16 22:26:44]

http://codereview.chromium.org/5064001/diff/1/chrome/browser/extensions/exten...
File chrome/browser/extensions/extension_omnibox_api.cc (right):

http://codereview.chromium.org/5064001/diff/1/chrome/browser/extensions/exten...
chrome/browser/extensions/extension_omnibox_api.cc:146: if
(!style->GetInteger(kDescriptionStylesLength, &length) || length < 0)
On 2010/11/16 19:05:02, Antony Sargent wrote:
> It seems like passing a negative length value is likely to indicate a bug in
the
> calling extension's arithmetic. Would it make more sense to either return an
> error or just ignore this style (ie treat it the same as length 0)? Your call.

The negative checks here are just me being paranoid - the JSON schema should
prohibit negative values in the extension process.

If GetInteger fails, that means that length was omitted, which means "to the end
of the range".

> On a related note, it might actually be useful to be able to send negative
> values for the *offset* - in python I find it's often convenient to use
negative
> indices on strings and arrays to do things like "give me the last element in
> this array" or "give me the last 5 characters from this string". 

That's a good idea. Using python-style ranges would probably be fairly intuitive
for developers.

http://codereview.chromium.org/5064001/diff/1/chrome/common/extensions/api/ex...
File chrome/common/extensions/api/extension_api.json (right):

http://codereview.chromium.org/5064001/diff/1/chrome/common/extensions/api/ex...
chrome/common/extensions/api/extension_api.json:3931: "description": "An array
of style objects, created using styleUrl, styleMatch, or styleDim. A style
applies to the region of text specified by the style's starting offset and
length. If there are any overlapping regions of text covered by multiple styles,
they will be added together (e.g. 'match' + 'dim' will display a dimmed match).
Not all style combinations will be visually distinct.",
On 2010/11/16 19:05:02, Antony Sargent wrote:
> nit: I'm not quite sure what you mean by "Not all style combinations will be
> visually distinct" - perhaps an example would illuminate?

Done.