If I may add here, having gotten a much deeper understanding of how the matching software works tonight from the dev himself,
I am using the example given. I’ve not tested in it in this form. I am writing here only to explain how it works and this was the first-available example.
During automatic matching,
- Everything found is used.
Dark.Waters.2019.720p.AMZN.WEB-DL.DDP5.1.H.264-NTG.mkv)
is broken into words Dark+Water+2019+720p+AMZN+WEB-DL+1+H+264+NTG and used as the search criteria.
- The number of items (tokens) in that list is 10.
- The best match it can get would
Dark+Waters (year-hint=2019) = 2 items.
- 2 of 10 = 20% match score.
- 20% does not meet the minimum match score required for automatic.
When the name is freed of the extra information, either by enclosing in [ & ] (anything between brackets is ignored), or removing from the name entirely, the resultant searcn becomes Dark+Waters+2019 (now only 3 items).
Based on heuristic match data (everyone else who has used fix match) the match engine knows 2019 is the year and not part of the name.
It now can search using Dark+Waters+(year-hint=2019) (2 items)
The simple percentage / probability of match is 2 versus 2 yields a match score of 100.
The software sees this value is greater than the required minimum and returned “Matched” status.
Per the references to FileBot,
What it does is
- Take the file name as given
- Start the same brute force search using everything (after having parsed it)
- Iterates down , removing words until it finds the best match
- It displays that match candidate in the right-hand pane.
- That match candidate is the pristine name inclusive of other directory & year (if included) in the final file name.
- When you click “Rename”, it movies & renames (copying then deleting after if necessary) to put the file in pristine form.
The naming template I use for movies is very simple:
/nas/movies/{n} ({y})/{n} ({y})
{n} = name
{y} = year
I can choose to add extra information if I wish. Here is how I would add resolution.
If I were to do that, I would likely have something of the form below which conveys the info and is compliant with Plex naming standards.
/nas/movies/{n} ({y}) [{resolution}]/{n} ({y}) [{resolution}]
This would produce:
/nas/movies/Star Trek Nemesis (2002) [2160p]/Star Trek Nemesis (2002) [2160p].mkv
caveat: I haven’t used resolution in a while so the output might now be different than what I show here.
I hope this helps take some of the mystery out of all this. I will also add; using a regex, in isolation, can only address some of the elements found. . They either include or exclude characters found. There will be failures on the pro and con side. It will include what it should have skipped and the reverse, all because of the serial LR-parse nature of the expression. Anything out of sequence from what the regex is expecting will result in a fail.
My best advice is:
- Get FileBot. It’s more than worth its cost many times over
- Take the time to workout the perfect renaming expression for your needs.
- When you have that, drop the existing files in and let FileBot rename everything.
a. It will rename my entire library in less than 5 minutes.
