Keywords are different on the same files across 2 different IMatch databases

[Damit]

I was trying to work on moving some keywords around and preparing the thesaurus.  I had a bunch of artwork photos I had downloaded from museums which had all sorts of keywords attached to them. I wanted to move all these keywords into a parent keyword named "Outsourced." I was originally doing this with the clipboard tool in the Category workspace using the @keywords category, but then decided it was best to just create a new database with just the files with the external keywords so I could just cut and past in one gross sweep.  I had some difficulties as I had some keywords that did not want to move and re-propagated themselves outside of the "Outsource" parent directory, no matter how many times I moved them, but I was able to resolve this with the keyword panel and deleting said keyword rather than moving them.

I wrote all the metadata changes and in the new database, all the files had all their keywords under "Outsourced." I then made a thesaurus from the database, which gave me the outsourced child directories under the Outsourced Parent, along with all the keywords included by default.

I then opened up my original database, which included the files I just changed. I re-scanned the folders.  When I checked on the same exact files on the original database, the keywords were not the same as those that showed on the new database. Some of the original keywords I had moved remained and were listed under keywords in the metadata panel and as categories under @keyword. I was confused. I closed the database and opened up the new one and the keywords were as I had left them, under outsourced.

I do not understand how this could be.  I thought since everything was embedded in the files that the would show up with the same keywords in any database, after I updated the metadata. There must be a simple explanation as to why this is happening but I can't figure it out. Any ideas on why this is happening?

[Mario]

Why do you work with multiple databases?
IMatch databases can easily manage 500,000 to one million assets.
Managing the same files in multiple databases is a recipe for trouble and requires a lot of discipline from your side to always keep the database-specific settings etc. in-sync. I would not waste time on this.

1. Use the ExifTool Command Processor with the "All Keywords" preset to see the keywords in your files.
This is what IMatch has written in database A. Seeing any issues?

2. Check your Thesaurus and Edit & Preferences > Metadata and Edit > Preferences > Metadata 2 settings. All settings which deal with how flat keywords are imported.

If IMatch imports files, it maps flat keywords (IPTC keywords and XMP:dc-Subject) into hierarchical keywords using the thesaurus.  If the keywords end up differently in your second database, most likely the thesaurus does not match or you have used different settings for mapping flat to hierarchical files.

Again, managing the same files in different IMatch databases is not something I would ever recommend or do. So many settings to keep in sync, the thesaurus and everything. A lot of extra work. Many ways to shot yourself in your own foot and cause support tickets.

[Damit]

I explained why I was doing this. It was simply to clean up the external keywords in a separate database. I would rather work with one database.

I checked the ExifTool and on both databases, it shows the same keyword, which show hierarchically as they should, with all keywords under the parent keyword "Outsourced." However, the keyword panel, The lightroom hierarchical, the xmp dc and the IPTC (and perhaps others) show all these keywords and then a list of some of those which are included in "[] Outsourced," but are not listed, hierarchically as part of "[] Outsourced."

I do think it may have something to do with the Thesaurus, as they were different, but even after I imported the new thesaurus and re-scanned, the old keywords remained. So, I have to delete this folders from the database and then re-import them now that the thesauri are identical. 

Man, photo metadata is a pain! And I thought music metadata was bad...

[Mario]

Man, photo metadata is a pain! And I thought music metadata was bad...

I call this Metadata Mess.

And it's even worse with video metadata. The most common used is Apple's QuickTime, all 80 different versions and revisions of it. XMP is trying something and the IPTC is also trying to do something. Should only take a decade or two until some common standard emerges...

[Damit]

I have held out for so long hoping this mess would be sorted out.   

I have removed the folders from the database, which I assumed I had to do for a re-import.  However now I have all these keyword categories that I don't want.  I thought they would disappear when I took out the files, as the @ keyword is supposed to represent your keywords and these keywords no longer exist, but they remain as categories with 0 files in them. I don't know of a quick or easy way to delete them. This whole thing has become a mess, but there has to be an easy way to remove empty categories.

[Damit]

Ah, the "Keep Empty Categories" property! I guess I had to uncheck that.  I am surprised it is unchecked by default.

[Mario]

The optimal setting depends on your workflow. Many users like to keep empty @Keywords so they can use them again to assign files to keywords. If you don't want this, turn this setting off.

See @Keywords Category and there Keep Empty Categories for details.

[Damit]

OK, I understand, but it conflicts with what is written in the help file "By default, empty child categories of a data-driven category are removed when the category is refreshed. By enabling this option you can instruct IMatch to keep empty child categories of @Keywords."This leads one to believe that, by default, the option is set to "No," when it is really set to "Yes." Just letting you know this in case you may want to edit that portion of the help section. This reminds me of my first question on this forum, where I had to uncheck the "shop only tags with data" to find the fields I needed.

Unfortunately, I am still getting empty sub-categories even though I have the Keep empty categories set to "No." I have refreshed the category and even data-driven categories but the empty subcategories remain.

Also, is there a setting to merge keywords with the exact name together when placing them into a parent category?  It seems IMatch will add a " 1" to the end of the keyword when it finds a keyword with the same name already residing in the parent keyword, in order to not create a duplicate.  Is there a way for it to merge the files under the same name?

[Damit]

After about half an hour to an hour the empty sub-directories in @keywords were deleted. I am not surge why it took so long and what caused this action to happen.

Please don't forget to let me know if there is a setting to merge keywords with the same exact name. Thanks!

[sinus]

...
Also, is there a setting to merge keywords with the exact name together when placing them into a parent category?  It seems IMatch will add a " 1" to the end of the keyword when it finds a keyword with the same name already residing in the parent keyword, in order to not create a duplicate.  Is there a way for it to merge the files under the same name?

I do not fully understand your question.
If a file has the same keywords, it will go into the same keyword. It is not possible to have the same keyword twice, as far as I know.
Maybe you do speak from Data driven categories, what is not the same as keywords?

[Mario]

I also don't understand. What should be merged and how.
Keywords are unique. You also mix keywords and categories in your post, which makes things even harder to understand.

The @Keywords category is a data-driven category build from the hierarchical keywords in your files. That's basically it.

If you try to get answers for complex situations, put yourself into the place of a community member.
We don't see what you see. We don't see the keywords you have, what @Keywords contains etc.

It is very easy in Windows to create some screen shots and attach them to your post. It is much more likely that we can understand the problem if you provide more information. We're happy to help, but we need to understand.

Also, it seems that you are mixing different questions in the same thread.
This thread is about "Keywords are different on the same files across 2 different IMatch databases", now it's about merging keywords somehow.

Better to open new threads for new problems. This keeps threads short and to the point. Better answers, better for the search engines and Google.

[Tveloso]

Is the OP saying that given a structure like this in @Keywords:


    @Keywords
    |--Parent1
    |    |--Child1
    |    |--Child2
    |--Parent2
    |    |--Child2
    |    |--Child3



...that if the Category Child2 from Parent1 is dragged and dropped onto Parent2, this does not merge the files into Parent2|Child2, but instead creates  Parent2|Child21?...(so we get this instead?):


    @Keywords
    |--Parent1
    |    |--Child1
    |--Parent2
    |    |--Child2
    |    |--Child21
    |    |--Child3


I did not try it...

[Mario]

In that case (and this is why I asked for a little more detail) you would not drag Child2, but the files in Child2 to the other Child 2. This way you have both options, adding a new child category (unique) and merging the multiple keywords into one.

Or just select the files in a File Window, remove Parent2|Child2 in the Keywords Panel and add Parent1Child2. Should take only two clicks and a <Cltr>+<S> to apply.

Or switch the KW panel to edit mode and replace Parent2 with Parent1 and save.

[Damit]

What happened is that I had two sets of files with the same keywords. I cut & Pasted the keywords into a parent keyword "Outsourced." When I opened up Outsourced I had all the keywords, with the duplicate listing an " 1" at the end.  I don't know if I had a option to merge or not, as I remember seeing a dialog to start a new category when moving another keyword that had nothing to do with this set, but I believe not. Unfortunately its been a crazy day so I have been late to respond and will not be able to get to some of the other topics or explain this with more detail until later tonight or tomorrow. I will see about opening up the panel in edit mode...