936439 - EncodingUtils should have a method for asking if a string is an encoding that FindEncodingForLabel could return

Status: RESOLVED WORKSFORME

(Core :: Internationalization, defect)

Description

[Henri Sivonen (:hsivonen)]

I need this for implementing the replacement encoding.

Comment 1

[Henri Sivonen (:hsivonen)]

Attachment: Add IsEncoding

Despite the comment, this isn't #ifdef DEBUG, because I have one non-DEBUG use case.

Comment 2

[Masatoshi Kimura [:emk]]

Comment on attachment 829217 [details] [diff] [review]
Add IsEncoding

>+  // Using labelsEncodings isn't as efficient as possible, but this is mostly
>+  // used for debug build asserts anyway.

What is the non-DEBUG use case do you have? Is this acceptable to it?
How about
  nsCString label(aEncoding);
  ToLowerCase(label);
  return NS_SUCCEEDED(nsUConvPropertySearch::SearchPropertyValue(
      labelsEncodings, ArrayLength(labelsEncodings), label, enc)) &&
      aEncoding.Equals(enc);
?

Comment 3

[Henri Sivonen (:hsivonen)]

(In reply to Masatoshi Kimura [:emk] from comment #2)
> Comment on attachment 829217 [details] [diff] [review]
> Add IsEncoding
> 
> >+  // Using labelsEncodings isn't as efficient as possible, but this is mostly
> >+  // used for debug build asserts anyway.
> 
> What is the non-DEBUG use case do you have? Is this acceptable to it?

Rejecting non-Encoding Standard chardet guesses before we figure out if mailnews needs charder to be ably to guess ISO-2022-KR/CN and some Cyrillic encoding whose name escapes me now.

> How about
>   nsCString label(aEncoding);
>   ToLowerCase(label);
>   return NS_SUCCEEDED(nsUConvPropertySearch::SearchPropertyValue(
>       labelsEncodings, ArrayLength(labelsEncodings), label, enc)) &&
>       aEncoding.Equals(enc);
> ?

OK. I didn't do that, because "replacement" needs special casing, but then I special-cased it anyway for another reason. I'll change the patch.

Comment 4

[Masatoshi Kimura [:emk]]

Comment on attachment 829217 [details] [diff] [review]
Add IsEncoding

Assuming the new patch will be created.

Comment 5

[Henri Sivonen (:hsivonen)]

It turns out that I didn't need this after all.