40461 - implement NS_ConvertUCS2toUTF8()

Status: VERIFIED FIXED

(Core :: XPCOM, defect, P1)

Description

[Chris Waterson]

Provide UCS2-to-UTF8 conversion (other than the always-allocating 
ToNewUTF8String).

Comment 1

[Chris Waterson]

Attachment: proposed implementation

Comment 2

[Chris Waterson]

Attachment: proposed implementation, try #2

Comment 3

[Chris Waterson]

Please ignore patch #1. I had the "hand off ownership" done wrong in 
ToNewUTF8String(), and had also carried over a bug with first-pass length 
calculation from the original source.

Comment 4

[Chris Waterson]

ftang: could you also take a look at these changes? I'm re-implementing 
ToNewUTF8String() in terms of a global helper class, NS_ConvertUCS2toUTF8(). 
N.B. that I think I fixed a bug with UTF-8 first-pass length calculation as 
well (you had the masks inverted bitwise).

Comment 5

[Chris Waterson]

I need this for 10373, and probably 28787, too. Can I get transitive closure of 
[nsbeta+] from these bugs, please?

Comment 6

[Scott Collins]

These changes look reasonable, but be aware that I will probably later refactor 
to move the work of the |Init| method to a function |NS_CopyUCS2toUTF8( const 
nsAReadableString& aUCS2source, nsAWritableCString& aUTF8destination )|, which I 
will then call from |Init|.  This work will actually be done by creating a 
special char-sink so that all the work is done in |copy_string|.

If you have some reasonable tests cases, and this code passes, check in at will.

Comment 7

[Chris Waterson]

Attachment: after review from the Style Council

Comment 8

[Brendan Eich [:brendan]]

Didn't I convince you guys to follow the nsFoo class-naming convention?

/be

Comment 9

[Chris Waterson]

fix checked in, r=scc,brendan

Comment 10

[Tracy Walker [:tracy]]

- Per last comments, age of bug, and no reopen - Marking Verified/Fixed.  Please 
reopen if still a problem.