We're updating the issue view to help you get more done.Ā 

Fix test that locale's emoji name and keywords ā‰  root's

Description

When I looked at the vxml, I found a problem:

Example: vxml/common/annotations/ha.xml

26: <annotation cp="šŸ˜€">ā†‘ā†‘ā†‘</annotation>
27: <annotation cp="šŸ˜€" type="tts">ā†‘ā†‘ā†‘</annotation>

And when I look at the ST, I see 2 vetters voting for that.

https://st.unicode.org/cldr-apps/v#/ha/Smileys/257d9e3f66adac9

However, a vote for "E10-836" should have triggered an hard error; it is a complete mistake. Those IDs were chosen so that we could raise errors if someone picked them.

The problem lies in CheckNew.java

1 2 3 String rootValue = annotationsRoot.getStringValue(path); if (value.equals(rootValue)) { result.add(new CheckStatus().setCause(this) ...

The first line should have been:

1 String rootValue = annotationsRoot.getStringValueWithBailey(path);

With that change, the error is caught.

Now, what will happen for this release is that all such values will be nuked: the only place we should see ā†‘ā†‘ā†‘ in /annotations/ could be in sublocales, such as en_GB.

xpath

None

locale

None

Status

Priority

critical

Assignee

Mark Davis

Reporter

Mark Davis

tracReporter

None

Reviewer

None

Labels

None

Components

Fix versions

phase

None