97238 - switch XPIDL from nsAReadableString to const nsAString

Status: RESOLVED FIXED

(Core :: XPCOM, defect)

Description

[Jon Smirl]

> nsAReadbleString.h/nsAwritableString.h. I guess the new plan is to tell
> readable from writable with the const qualifier. Should the XPIDL compiler
> be changed to reflect this?

Funny, i noticed the same thing the other day.

It should be changed from:

const nsAReadableString & foo
to
const nsAString& foo

and

nsAWritableString & foo
to
nsAString& foo

--pete collins

Comment 1

[Doug Turner (:dougt)]

reassign all kandrot xpcom bug.

Comment 2

[Doug Turner (:dougt)]

jaggernaut, please advise. 

Comment 3

[jag (Peter Annema)]

Yes, I think it should. I don't think there should be any problem (since
nsAReadableString is typedef'ed to const nsAString, other than perhaps someone
reading the generated .h and not quite understanding the difference in it (const
nsAString&) and the implementation (const nsAReadableString&). Though at some
point I'll probably sweep the tree and change it all.

Comment 4

[Doug Turner (:dougt)]

assigning to jag.  (assign back if you want me or someone else to try to get to
this problem)

Comment 5

[jag (Peter Annema)]

http://lxr.mozilla.org/seamonkey/source/xpcom/typelib/xpidl/xpidl_header.c#665

bugzilla as a "mental note:"

Comment 6

[jag (Peter Annema)]

Attachment: Emit nsAString instead of nsAReadableString and nsAWritableString

Comment 7

[Mike Shaver (:shaver -- probably not reading bugmail closely)]

Comment on attachment 48741 [details] [diff] [review]
Emit nsAString instead of nsAReadableString and nsAWritableString

sr=shaver, now that jag held my hand through the new AString/readable/writable mappings.

Comment 8

[jag (Peter Annema)]

Here's why this should work flawlessly:

typedef const nsAString   nsAReadableString;
typedef       nsAString   nsAWritableString;

(see nsAReadableString.h and nsAWritableString.h)

Comment 9

[cls]

Comment on attachment 48741 [details] [diff] [review]
Emit nsAString instead of nsAReadableString and nsAWritableString

r=cls

Comment 10

[jag (Peter Annema)]

And checked in.

Comment 11

[jag (Peter Annema)]

Btw, I may end up doing a nsAReadableString/nsAWritableString -> nsAString tree
sweep at some point in the future. Not dealing with that in this bug :-)

Comment 12

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

For some reason this turned speedracer and bismark orange.

Comment 13

[jag (Peter Annema)]

Looking into it.

Comment 14

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

It looks like clobbering bismark fixed the problems.  I have no idea why, though.