Hmm. Well, no objection for mbstring test data, I guess... although the files are provided by the Unicode Consortium and were copied directly into php-src as downloaded from their site.

If ever the Unicode Consortium updates those files and we import the updated files, the same misspellings will come back. So if you really want to fix them for good, it might be good to contribute the fixes upstream.

Anyways, no objection from me.

If ever the Unicode Consortium updates those files and we import the updated files, the same misspellings will come back. So if you really want to fix them for good, it might be good to contribute the fixes upstream.

If some by chance, these data files changes, fixes to each time.
I don't know where to report upstream, but shall I report it.

Is it here? https://corp.unicode.org/reporting/website.html

If ever the Unicode Consortium updates those files and we import the updated files, the same misspellings will come back. So if you really want to fix them for good, it might be good to contribute the fixes upstream.

If some by chance, these data files changes, fixes to each time. I don't know where to report upstream, but shall I report it.

Is it here? https://corp.unicode.org/reporting/website.html

I believe that's for problems with the websites. I believe this should be used for the data files: https://corp.unicode.org/reporting/error.html

@nielsdos Thanks. I tried to send.
スクリーンショット 2023年09月17日 22 08 10

Perhaps a diff would've been easier for them, but I don't think it's a big deal if it's already sent.

Perhaps a diff would've been easier for them, but I don't think it's a big deal if it's already sent.

Ah, I'm lazy😣. I hope member of Unicode Consortium that look this page.

Maybe this one can be added here also:

--- a/.gitattributes
+++ b/.gitattributes
@@ -25,5 +25,5 @@
 /Zend/zend_vm_handlers.h linguist-generated -diff
 /Zend/zend_vm_opcodes.[ch] linguist-generated -diff
 
-# The OSS fuzz files are bunary
+# The OSS fuzz files are binary
 /ext/date/tests/ossfuzz*.txt binary

We did see this, but have decided not to touch the two archival mapping tables. Note those mapping tables are posted in an "OBSOLETE" directory for a reason -- they are basically just sources for historical research, and are not intended to be used for current products. JIS0212.TXT dates from 1994, and KSX1001.TXT from 1999. Later updates have only been to deal with the out-of-date terms of use and contact links, and have left the data (and the rest of the header content) as is, including typos. Please see the warning notice at the top of the ReadMe.txt:
https://www.unicode.org/Public/MAPPINGS/OBSOLETE/EASTASIA/ReadMe.txt

@Ken-Whistler Thank you very much for confirming this. I understand how to handle archived mapping tables. Thanks again.