CodeDB 0.3.2b

[Skipper]

Can't open the 0.2.2b ZIP file, Windows 11 reports it as being invalid...

Cheers
Skipper

[idle]

Can't open the 0.2.2b ZIP file, Windows 11 reports it as being invalid...

Cheers
Skipper

Thanks don't know what happened there. redone

[Skipper]

Hi Idle,

this time the ZIP archive is OK. Thanks for that!
The CodeDB version contained within didn't seem to work for me right away.
I dropped the database from the previous version and had the app scan 'myPC' anew.
Only after I then killed CodeDB and restarted it, the database became searchable.
Later today, I'm going to play with the new features you added.

Cheers
Skipper

[idle]

thanks let me know how it goes.

[idle]

updated 0.3.0a
Added option to save DB where you want
Added Prefs file
Changed indexing

it should prompt for a rescan on the major or minor version change

[Skipper]

thanks let me know how it goes.

Hi Idle,

CodeDB 0.3.0a (29508) works fine, including the new options you added.

One suggestion: if I search with a boolean AND, only the first of the two arguments get highlighted. It would be nice if the second one gets the same treatment. The same goes for the boolean OR operator, as far as indeed both arguments would be found in the same snippet.

Another suggestion: when starting CodeDB, it would be nice to be able to somehow have a list of the top-level paths presented that have already been scanned, i.e. not the entire scanning tree, but only the top level paths where the scanning started.

Cheers
Skipper

[idle]

thanks let me know how it goes.

One suggestion: if I search with a boolean AND, only the first of the two arguments get highlighted. It would be nice if the second one gets the same treatment. The same goes for the boolean OR operator, as far as indeed both arguments would be found in the same snippet.

Another suggestion: when starting CodeDB, it would be nice to be able to somehow have a list of the top-level paths presented that have already been scanned, i.e. not the entire scanning tree, but only the top level paths where the scanning started.

I should probable add an option to switch search modes as it's searches the main index which is the procedure and macro names, a *search is a secondary index and the Full texts search of the procedure body is triggered by and , or onto another secondary index.
As for the precedence AND comes before OR

I haven't worked out how to highlight the instances either, I will end up switching to scintilla so it can run on linux and osx eventually.

cheers

[Skipper]

Cool, would appreciate those changes & additions. I really like how this tool is developing, even if - for you - it is only a demo of your DB developments.

cheers
Skipper

[idle]

Cool, would appreciate those changes & additions. I really like how this tool is developing, even if - for you - it is only a demo of your DB developments.

cheers
Skipper

It's helping me figure out what changes I need to make and if the test is useful, why not!
I'd be curious to see Sqlite do full text search, it does have an extension to do it so it could with the dll.
currently the Full text search uses a lot of memory is slow to build and from 3m tokens its using 100mb to index the code and that's probably 2 to 2.5 times the input, I haven't actually checked but that's what it looks like and considering my db on disk is 45mb, 100mb does appear to be a bit large.

When you say show the top level file do you mean while it's scanning?

[Skipper]

When you say show the top level file do you mean while it's scanning?

No, I mean that if I do not scan my entire PC (in which case I know what I did) but individual directories only, it would be nice to be able to see which directories were used to start a scan. This would prevent scanning them twice and would allow me to see if another directory tree would need to be scanned sometime later...

cheers
Skipper

[idle]

When you say show the top level file do you mean while it's scanning?

No, I mean that if I do not scan my entire PC (in which case I know what I did) but individual directories only, it would be nice to be able to see which directories were used to start a scan. This would prevent scanning them twice and would allow me to see if another directory tree would need to be scanned sometime later...

cheers
Skipper

Yes that would be better populate a list with the scan dirs and you can check them or uncheck them, the DB won't double up on the items only when ones changed. I will look at doing it soon.

[idle]

V 0.3.2b
Added support for c backend.