[maniyac]

I had an issue a few days ago in v13.1.12.1 with the "Scrape metadata for selected movies" feature.

I think I know what may have caused it and I have run some tests since and it does not appear to occur again.
I will provide my suspicion of the cause (unverified just speculation) in an attempt to warn others to be careful and run tests before they use the feature on their entire library. 

I used the scrape metadata feature on about 3500 files but I only selected to scrape "ratings", "top 250", "trailers" as I thought this would be the best way to update my IMDB ratings (is there a better way to do this?) and a few other details.

Even though I had only these options selected, it appears TMM scrapped ALL metadata (lucky I had the image scrapers off or I'm guessing it would have wrecked my collection even more).

The only reason I can think this may have occurred is I was using Kodi 16.1 until now (upgraded to 17.6 because TMM v3 does not write compatible .nfo files to v16 anymore).

When on TMM v2 I was scrapping .nfo files in Kodi <16 format. I suspect to update the .nfo data TMM had to scrape all data to make it compatible with the new format?

Most of the changes are manageable as far as I can tell it was not overly detrimental to my collection (although I manually inputted every movie and it's too much work to go over them all again, so I don't know for sure).

It would be great if I could undo this last scrape but I'm almost positive besides going to an outdated manual backup this would not be possible?

One thing I did do manually and could use help with was add titles like "IMAX" or "Directors Cut" etc into the "title" field, to be displayed in Kodi.

Is there a way to automatically write the "edition" tag into the "title" field and leave "original title" untouched?
(this is to do with the data stored in the .nfo file for the title Kodi displays, not the re-namer that would modify the file name itself)

[mlaggner]

IIRC we did not touch the legacy NFO format (for Kodi < 16) for ages - except internal changes.

tmm writes internal backups (just look into the /backups folder in the tmm folder). Backups are being made on a daily interval - you may try to extract the backup file from the day where you scraped or the day before (just back up your actual /data folder ).

with 4.1 there is an enhanced bulk editor built on JMTE where you can change internal fields like title/original title/... the same way as you can do for the renamer!

[maniyac]

IIRC we did not touch the legacy NFO format (for Kodi < 16) for ages - except internal changes.

tmm writes internal backups (just look into the /backups folder in the tmm folder). Backups are being made on a daily interval - you may try to extract the backup file from the day where you scraped or the day before (just back up your actual /data folder ).

with 4.1 there is an enhanced bulk editor built on JMTE where you can change internal fields like title/original title/... the same way as you can do for the renamer!

Thanks for the information, I am just about to donate to get v4 so I will give that a go!

Unfortunately, the back-ups only go back 5 days (literally 1 day after I had the issue), maybe because of the v3 update I did around that time?

I think I will be ok, I have probably lost some manual information but I still have a version of v2 I am going to skim over and manually cross-reference to look for significant differences. 

From what I can tell the scrapper does not overwrite data manually inputted that didn't have a TMDB match or information like sort order, which I have manually done for many movies and special edition tags.

I think I will just have to update the IMAX etc titles and some alternate versions of the same movie I manually had to edit (which I should be able to find with the duplicate detector).

It's a pain in the rear but it could have been worse.

[maniyac]

IIRC we did not touch the legacy NFO format (for Kodi < 16) for ages - except internal changes.

tmm writes internal backups (just look into the /backups folder in the tmm folder). Backups are being made on a daily interval - you may try to extract the backup file from the day where you scraped or the day before (just back up your actual /data folder ).

with 4.1 there is an enhanced bulk editor built on JMTE where you can change internal fields like title/original title/... the same way as you can do for the renamer!

Thanks for the information, I am just about to donate to get v4 so I will give that a go!

Unfortunately, the back-ups only go back 5 days (literally 1 day after I had the issue), maybe because of the v3 update I did around that time?

I think I will be ok, I have probably lost some manual information but I still have a version of v2 I am going to skim over and manually cross-reference to look for significant differences. 

From what I can tell the scrapper does not overwrite data manually inputted that didn't have a TMDB match or information like sort order, which I have manually done for many movies and special edition tags.

I think I will just have to update the IMAX etc titles and some alternate versions of the same movie I manually had to edit (which I should be able to find with the duplicate detector).

It's a pain in the rear but it could have been worse.

Just had a thought.

If I use the "rewrite .nfo" feature in v2 to bulk select say all "IMAX" filter matches, then open v3, update the database and re-scrape just "ratings" on those files with "IMAX" in the title, would that be a fast way to get most of the information back and keep the .nfo files up-to-date?

I still manually need to identify discrepancies but at least this way I can do it in bulk?