Page 1 of 1

something is broken in search

Posted: 24 Oct 2017 10:29
by yusef88
in XYplorer 18.50.0011 this search doesn't seem to work:

Code: Select all

!>[أ-ي]
latest 18.50.0011.jpg
but it works in old version, XYplorer 15.80.0200
old 15.80.0200.jpg
another question: why this pattern doesn't work?

Code: Select all

 !>[a-z]

Re: something is broken in search

Posted: 24 Oct 2017 10:54
by highend
There isn't a simple NOT for complex regex expressions so your pattern !>[a-z] is
interpreted in a different way. Use the help switch to find out how...

Re: something is broken in search

Posted: 26 Oct 2017 14:37
by yusef88
highend wrote:There isn't a simple NOT for complex regex expressions so your pattern !>[a-z] is
interpreted in a different way.
you mean in general or XYplorer in particular because Search Everything can ignore !>[a-z]

Code: Select all

regex:![a-z]

Re: something is broken in search

Posted: 26 Oct 2017 14:44
by highend
Different regex languages have different options / capabilities. As I said, use the help
switch to find out how XY interprets it...

Re: something is broken in search

Posted: 26 Oct 2017 20:09
by yusef88
sorry but The help doesn't explain why :
>[^\u0000-\u007F]+ can ignore Basic Latin characters but >[\u0000-\u007F]+ can't match them!
>[\u0980-\u09FF]+ can match Bengali characters but >[^\u0980-\u09FF]+ can't ignore them!

Re: something is broken in search

Posted: 26 Oct 2017 20:29
by highend
I didn't say anything about unicode, I was only referring to your NOT expression

But:
>[\u0000-\u007F]+ can't match them!
Sorry, but I can't reproduce this...
match.png
Can't say anything about Bengali stuff, not my language...

Re: something is broken in search

Posted: 26 Oct 2017 21:09
by yusef88
does your folder contain only English names? because it matches everything here
en.jpg
only matches Bengali
ban.jpg

Re: something is broken in search

Posted: 27 Oct 2017 13:07
by admin
Help Switch means you append /? to the search pattern. You will get something like this then:

Code: Select all

Query:  >[^\u0000-\u007F]+ /?
Parsed: RegEx[[^\u0000-\u007F]+]

Search: Quick Search
Mode: RegExp
FindBoolSmartDetect: Yes
ExtendedPatternMatching: No
MasterInvert: No
Tags DB Only: No