HDDS-5564. feature/PrefixFSO.md translation

[iamabug]

What changes were proposed in this pull request?

doc translation

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-5564

How was this patch tested?

hugo server under docs diectory.

[iamabug]

cc @sky76093016

[sky76093016]

支持 and support appear repeatedly, it would be better to change to the above.

Suggested change

	summary: 支持Supports atomic rename and delete operation.
	summary: 支持 atomic 重命名和删除操作

[cku328]

I think it translates to 原子的.

[sky76093016]

1.Extra space
2.要么 is too colloquial.Would it be better to use 可能?

Suggested change

	在简单布局格式中，OM 采用全路径名来存储每个键的元数据。在基于前缀的优化布局格式中，OM 将中间目录的元数据存储在 `DirectoryTable` 中，将文件的元数据存储在 `FileTable` 中，如下图所示。两张表中的主键要么是目录名，要么是以父目录唯一标识符为前缀的文件名， 形如 `<parent
	在简单布局格式中，OM 采用全路径名来存储每个键的元数据。在基于前缀的优化布局格式中，OM 将中间目录的元数据存储在 `DirectoryTable` 中，将文件的元数据存储在 `FileTable` 中，如下图所示。两张表中的主键可能是目录名，也可能是以父目录唯一标识符为前缀的文件名，形如 `<parent

[yuangu002]

Thanks @iamabug and @sky76093016 for adding zh doc. I think you misunderstood this sentence:

The key to the table is the name of a directory or a file prefixed by the unique identifier of its parent directory, <parent unique-id>/<filename>.

This means that the key is in the format of <parent-id>/<name>, where the name is either a filename or a directory name (depends on DirTable or FileTable). So a correct translation should be something like 主键是以父目录唯一标识符为前缀的文件名或目录名.

[sky76093016]

I think so.

[iamabug]

thanks @yuangu002 I think you are right.

[yuangu002]

Also, @iamabug could you please add zh explanations on the picture? since the picture comes with English.

[cku328]

Thanks @iamabug for the translation.
Overall good, I only have some minor comments.

[cku328]

Suggested change

	该特性默认未启用，可以通过在 `ozone-site.xml` 进行如下配置启用：
	该特性默认未启用，可以通过在 `ozone-site.xml` 进行如下配置来启用：

I think it will read more smoothly.

[cku328]

Would it be better to replace 全路径 with 完整路径?
I think the existing in the original text should be mentioned.

[iamabug]

thanks @cku328 for the suggestion.

1. I think there is no harm using 完整路径
2. I am not sure about here. One reason is that the word existing seems to be indicating that this is a design doc before PrefixFSO is implemented. Now that Simple and PrefixFSO both exist, it seems unnecessary to mention Simple as existing.

[cku328]

I think it translates to 原子的.

[iamabug]

Also, @iamabug could you please add zh explanations on the picture? since the picture comes with English.

Thanks for the suggestion. I agree it would be better to use a picture with Chinese. I think we need the original to do that since we only have .png file here. cc @elek @rakeshadr

[yuangu002]

Also, @iamabug could you please add zh explanations on the picture? since the picture comes with English.

Thanks for the suggestion. I agree it would be better to use a picture with Chinese. I think we need the original to do that since we only have .png file here. cc @elek @rakeshadr

I don't think you need separate pictures. We can use the original pictures, but add an explanation in Chinese which explains what the picture is saying. For example, for the first picture (Legacy layout), you can add a short description (alt text) in Chinese like: 在传统的OM元数据格式中，我们采用完整路径，并且用尾部斜杠来区分目录和键。

[adoroszlai]

Thanks @iamabug @sky76093016 @cku328 @yuangu002 for the patch and reviews. What's the next step here?

[smengcl]

I've amended the usage part of the doc as the original doc is updated.

[adoroszlai]

Thanks @smengcl for updating the translation. It seems the configuration excerpt does not match in the original and translated pages.

[smengcl]

Thanks @adoroszlai for pointing that out. Looks like the configuration section has been updated in HDDS-6351. I have pushed updates accordingly.

[adoroszlai]

Thanks @iamabug and @smengcl for the translation, @cku328 and @yuangu002 for the review. Please feel free to open a new PR for further improvements if needed.