TC7rc4 - no longer finds text in UTF8 files

[marekjed]

(Thank you so much for the quick response to my previous report!)

config: TC7 rc 4, WinXP SP2, Polish locale

Steps to reproduce:
Use "Find files" and search for text in known UTF8-encoded files. (UTF8 checkbox should be checked). Search expression must contain "extended" characters (i.e. those which take up more than 1 byte in UTF8 encoding).

Result:
TC7 does not find any matches if the search expression contains "extended" characters.

By comparison, TC 6.56 finds the same text in the same set of files. I am only working with UTF8 files, so I have not been able to see if the bug affects also other Unicode encodings ("Unicode" checkbox).

Notes:
1. The bug does not affect lister.
2. The files I am working with do not use BOM signature.

[Flint]

Hm... I confirm this. No matter whether BOM signature present or not.

I checked now, this problem takes place from the version 7.0 private beta 2.5. I suppose, it has to do something with this change:

13.12.06 Added: Case-insensitive search for UTF8 text in the search function and in lister (currently only the forward direction works with non-English characters)

Also, I tested the same search with the option "Case-sensitive" turned on: in beta 2.5 and later it worked fine, till private beta 4.5, where it stopped working completely, for both case-sensitive and case-insensitive searches. The change here seems to be:

27.02.07 Fixed: Case-insensitive search in UTF-8 not always working properly

[ghisler(Author)]

I could reproduce it and fix it, thanks!

[Lefteous]

UTF-8 encoded words containing characters higher than 0x7F can be found using TC 7 RC 5. I have tested only files containing the UTF-8 BOM though.

[Flint]

Yes, works fine in TC7rc5 without BOM as well.

[ghisler(Author)]

Great, thanks!