Found typos in the "zh.xml" CLDR metadata file.

[Ali Sherief]

First of all, I will start this by saying I know absolutely no Chinese - any information I'm posting about it came from Wikipedia. I am in the process of making my C++ date/time library locale-aware and as part of that, I create scripts that parse the CLDR metadata and converts it into code.

There is a small incorrection inside common/main/zh.xml file. In the <exemplarCharacters> node for numbers (XPath: /ldml/characters/exemplarCharacters[@type="numbers"]), the Chinese numbers are completely in the wrong order. It currently reads:

        <exemplarCharacters type="numbers">[\- ‑ , . % ‰ + 0 1 2 3 4 5 6 7 8 9 〇 一 七 三 九 二 五 八 六 四]</exemplarCharacters>

the correct data (in my opinion) should be:

        <exemplarCharacters type="numbers">[\- ‑ , . % ‰ + 0 1 2 3 4 5 6 7 8 9 〇 一 二 三 四 五 六 七  八  九]</exemplarCharacters>

See https://en.wikipedia.org/wiki/Chinese_numerals for reference.

I apologize if I have made any mistake with the description.

- Ali

[Markus Scherer]

On Mon, Jan 30, 2023 at 7:43 AM Ali Sherief <zenula...@gmail.com> wrote:

There is a small incorrection inside common/main/zh.xml file. In the <exemplarCharacters> node for numbers (XPath: /ldml/characters/exemplarCharacters[@type="numbers"]), the Chinese numbers are completely in the wrong order.

There is no order. The exemplarCharacters are defined as *sets*. They are just presented in some sort order for stable and convenient presentation.

In particular, those "Chinese numbers" are ideographs (letters) that *can* be used for writing numbers but are otherwise used to write words, and don't sort in numeric order.

markus