Re: [code] v strange find/replace behaviour

From: Mitchell <>
Date: Mon, 27 Aug 2018 08:35:50 -0400 (EDT)

Hi Thomas,

On Mon, 27 Aug 2018, Thomas Lauer wrote:

> I am pretty new to TA and so far things have been looking good. However,
> I've just run into something that's utterly puzzling. This happens with
> a fresh install of TA x86_64 (ie no changes in the textadept directory)
> on a Debian-based system.
> I have a few largish files with filenames for music titles (m3u8). The
> names all look like this:
> /dir/dir2/Music/MusicM4A/Artist/Album/Title.m4a or
> /dir/dir2/Music/MusicMP3/Artist/Album/Title.mp3
> ie they all start with either /dir/dir2/Music/MusicM4A/ or
> /dir/dir2/Music/MusicMP3/.
> I load one of those files with say 500 titles (ie 500 lines). I go to
> the first line and select the part "/dir/dir2/Music/MusicM4A/" with the
> mouse and copy it to the clipboard (Ctrl-C). I press Ctrl-F, enter the
> copied string into the Find field (Ctrl-V)and press Find Next
> repeatedly. This does absolutely nothing.
> However, if I clear the Find field and enter the *exact same string*
> manually, ie character for character, Find Next works. It also
> sometimes, not always, works if I remove the first / character after
> copying from the clipboard.
> This is so strange that I'm inclined to think it's a problem with my
> installation.

I doubt it's an installation issue.

> Can anyone reproduce that and if so, any ideas as to what might be the
> problem?

I can think of two things: encoding issue, or copy-pasted text is not what you expect.

What encoding is your file in? You can use the Buffer -> Encoding menu to try temporarily changing to ASCII or ISO-8859-1 and then copy-pasting your text to search for (just don't save, which will commit any encoding changes). I'm wondering if your file has a UTF-8 BOM and that's accidentally being copied. Does this happen for other lines? (Not just the first.)

Open a new buffer and paste what you just copied and inspect it character by character. Tools -> Show Style also shows character bytes, which can help identify multi-byte UTF-8 characters you did not expect to see.


You are subscribed to
To change subscription settings, send an e-mail to
To unsubscribe, send an e-mail to
Received on Mon 27 Aug 2018 - 08:35:50 EDT

This archive was generated by hypermail 2.2.0 : Tue 28 Aug 2018 - 06:27:31 EDT