database design and location questions

[Ty Mayn]

 using a small number of test images I have examined XMP sidecar files as they have been added in JPT.

I have seen the changes of markup text  by added keywords and description labels

I also want to have a similar tangible  understanding of the database file which I have heard mentioned

Where is the database file located?

Is the database file text readable?

What are the database elements (fields) of a basic entry for an image that is know to exist in the folder navigator/selector?

What triggers the addition of one or more new entries into that database?

Programmatically: does each image first join the database before a thumbnail is generated or are thumbnails generated/presented first so that click activation of a thumbnail becomes the gateway for database update and image entry?

 

  Is the database always indexed in such a way that duplicate files are discovered at time of intake?

Or are duplicate files discovered only in a temporally separate program process?

Thanks for JPT . So much going on  generates a hunger for info

Ty

[Elmar Baumann]

Am 31.12.2018 um 07:57 schrieb Ty Mayn:
> Where is the database file located?

You can see the database location in: "File > Database > Info". It's in
the user's home folder below ".de.elmar_baumann/ImageMetaDataViewer" and
is a set of files with the name "database.*".

> Is the database file text readable?

No, it's a binary format (HSQLDB 1.8.0.10). Tools such as SQuirreL with
the same HSQLDB library (using the one, which is distributed with
JPhotoTagger) can open and manipulate the database. But this is not
recommended if you are not firm with SQL databases.

> What are the database elements (fields) of a basic entry for an image
> that is know to exist in the folder navigator/selector?

It's a relational database containing multiple tables with 1:n or m:n
relationships .

> What triggers the addition of one or more new entries into that database?

If a folder is selected, JPhotoTagger checks whether it contains new
images or known images with modified date/time/XMP (comparsion of
modification date/time and size) and if that is the case, the database
will be updated.

> Programmatically: does each image first join the database before a
> thumbnail is generated or are thumbnails generated/presented first so
> that click activation of a thumbnail becomes the gateway for database
> update and image entry?

Database update is done by the presense of image files in a folder.
Either via selecting the folder in the "Folder" or "Favorites" tab or
through automatic scanning via "Edit > Settings > Automize", which scans
predefined folders at a scheduled time whe JPhotoTagger runs. An other
scanning trigger is "File > Scan folders".

>  
>   Is the database always indexed in such a way that duplicate files are
> discovered at time of intake?
> Or are duplicate files discovered only in a temporally separate program
> process?

There is no such concept as "duplicate files". All file systems forbid
files of the same name in the same folder and because a file is
identified by it's path (folder hierarchy and file name), every file is
treated as unique. You can scan for duplicate images (same binary
content in different folders) via "Tools > Find duplicate images".