Navigation in generated docs.

Navigation in generated docs.

Committed: https://code.google.com/p/dart/source/detail?r=2102

[pquitslund, 2011-12-05 18:14:58]

LGTM.

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/files.dart
File utils/dartdoc/files.dart (right):

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/files.dart#newcode6
utils/dartdoc/files.dart:6: 
Aside: it'd be great to start collecting handy utilities like this somewhere
shareable/discoverable.  Ideally, we could see some bits promoted to the
standard IO lib...  but maybe in the interm we could cook up something like a
"commons" a la apache?  All of that being said, clearly the right and most
expedient thing in the interest of hammering out docs is not sweat this at the
moment.  Just food for thought!

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart
File utils/dartdoc/utils.dart (right):

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode8
utils/dartdoc/utils.dart:8: // Unlike world.withTiming, returns the elapsed
time.
Maybe this comment would be more useful as a proper function dartdoc?

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode28
utils/dartdoc/utils.dart:28: // Offsetting by needle length means overlapping
needles are not counted.
needle->text...

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode37
utils/dartdoc/utils.dart:37: // TODO(rnystrom): Should be in corelib.
Aha!  You read my mind. ;)

Along those lines, if we were to do that I'm guessing we'd want to harden the
impl and be wary of null args (or explicitly disallow them in the docs).  Or
maybe defensive programming is not the vogue in dynamic languages...  (Honestly
I don't know the answer here; edify me please! ;))

[Bob Nystrom, 2011-12-05 23:14:28]

Thanks!

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart
File utils/dartdoc/utils.dart (right):

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode8
utils/dartdoc/utils.dart:8: // Unlike world.withTiming, returns the elapsed
time.
On 2011/12/05 18:14:58, pquitslund wrote:
> Maybe this comment would be more useful as a proper function dartdoc?

Done.

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode28
utils/dartdoc/utils.dart:28: // Offsetting by needle length means overlapping
needles are not counted.
On 2011/12/05 18:14:58, pquitslund wrote:
> needle->text...

Done.

http://codereview.chromium.org/8772067/diff/1/utils/dartdoc/utils.dart#newcode37
utils/dartdoc/utils.dart:37: // TODO(rnystrom): Should be in corelib.
On 2011/12/05 18:14:58, pquitslund wrote:
> Aha!  You read my mind. ;)
> 
> Along those lines, if we were to do that I'm guessing we'd want to harden the
> impl and be wary of null args (or explicitly disallow them in the docs).  Or
> maybe defensive programming is not the vogue in dynamic languages... 
(Honestly
> I don't know the answer here; edify me please! ;)) 

Well, if you asked *me*, I'd say the type system should explicitly support
non-nullable types and then the VM would do those checks for you automatically
in checked mode.

[Siggi Cherem (dart-lang), 2011-12-06 17:31:45]

lgtm