If you add an EFU into an index, does it have to remain in the same location forever?
I had 2 EFUs in my Monolith instance and neither has presented their entries for a few years. It's like the second time I rebuilt with Monolith and the efus weren't there, their contents were erased but still kept in the filelist section.
I added several new efus and they're in a folder I can keep them in for years but it would be nice if I could add EFUs and then remove the original files.
Is that how it's meant to work? If so, maybe I added the 2 EFUs to Monolith and deleted them while it was rebuilding so it never added their contents..
Also how do I modify an EFU with a custom non-existent sub-folder at the second-top level?
Say I have real F:\, F:\A1 and F:\A1\A1 folders in an EFU.
For it to correctly put everything under F:\ in a sub-folder that never existed, do I just duplicate the F:\ line and change the folder?
I never made any tests before adding new EFUs to Monolith so it spent 6h45min rebuilding into broken filelists. I did also add 4 new NTFS monitors so it wasn't a total waste..And it still seems to have most of the file entries under F:\ itself and then duplicated F:\Fake versions that only has folders in them.
Anyway, instead of
F:\Fake
F:\Fake\A1\A1
F:\Fake\A1\A1
I make it
F:\
F:\Fake\
F:\Fake\A1\A1
F:\Fake\A1\A1
?
Cause I did the first one and as I was writing a previous block of text I realised maybe that's why it broke.
I thought it was because I was using an illegal character, the sub-folder has # in it.
When I opened one that has plain text with the same first setup above it didn't load as broken so that's why I thought maybe it's the symbol.
The one with # in it even loaded the file path to the EFU itself, while the second EFU didn't seem to. But both have been modified without having a line for the fake folder itself.
[Solved] 2 curiosities about adding filelists to indexes
-
Herkules97
- Posts: 321
- Joined: Tue Oct 08, 2019 6:42 am
[Solved] 2 curiosities about adding filelists to indexes
Last edited by Herkules97 on Thu Nov 12, 2026 8:29 am, edited 1 time in total.
-
santilli
- Posts: 35
- Joined: Tue Apr 30, 2024 7:05 am
Re: 2 curiosities about adding filelists to indexes
When you add a new entry to the File Lists index, the path to the EFU file is stored in a .ini file. If you move, rename or delete the EFU file, that alone will not update/remove it from that ini file, so it will still show up in the list despite pointing to an nonexistent file.
When you first load an EFU file, its content is indexed into the database, so they remain indexed if the EFU file is moved, renamed or deleted. So answering your question, this is how it is meant to work.
As for the second question, you can create fake entries and they can even have special characters, except for ones Windows forbids.
Just make sure to escape each path with quotation marks, and if you want it to point to a folder, it must end with a slash. For example:
Any item that doesn't end with a slash will be treated as a file.
When you first load an EFU file, its content is indexed into the database, so they remain indexed if the EFU file is moved, renamed or deleted. So answering your question, this is how it is meant to work.
As for the second question, you can create fake entries and they can even have special characters, except for ones Windows forbids.
Just make sure to escape each path with quotation marks, and if you want it to point to a folder, it must end with a slash. For example:
Code: Select all
"F:\#A1\"
"F:\Fake\#A1\"-
Herkules97
- Posts: 321
- Joined: Tue Oct 08, 2019 6:42 am
Re: 2 curiosities about adding filelists to indexes
Yah it should be how it works cause I have an entire efu instance, but it has several thousand EFUs under it so if a few vanish on a rebuild, I might not notice.
I don't have to do any sort of modifying of file paths for the second thing, as I'm only adding a fake sub-folder to the root of everything.
But I think I messed up and only realised afterward that I didn't do the sub-foldering correctly.
I probably answered my own question in the same message as I asked it..I just haven't tested it.
Maybe my entire post was pointless and I should've just done the tests first. I do for a bunch of would-be-posts but I decided for some reason to send this one. Laziness and stupidity on my part.
So I think I found the issue for the second thing..
Maybe it was to make the root device letter exist as well as the fake folder's path, I cba to check.
But it turns out when I modified a 200MB+ file with Notepad++ 32-bit it simply failed most of the lines.
It's also much slower, it took a few seconds to modify 1.3mil lines with 64-bit..
32-bit took more than 25min and I just force-exited it cause it was still stuck.
Now that I have a 64-bit modified version, it finished all the lines successfully and at least for one of the EFUs it loaded with Open File List.. properly.
This wasn't the EFU with # in it, but I will modify that one later. Presumably it will work now that I realise this.
I keep trying to use the 32-bit version for larger files as that's the main Notepad++ I use.
Regardless, Monolith still has all the files, so even if it failed it still failed "successfully". In a few weeks maybe I will rebuild Monolith with the fixed EFUs. Unless there are more issues I will find out later, this marks the end of that issue for me.
For the first issue, I will try to do some tests later but it is very likely that when Monolith was rebuilding, I deleted the EFUs I added forgetting that it was meant to index them first. Or Monolith is too corrupt and removes EFU contents on rebuilds if they don't exist in the specified paths.
Monolith IS corrupt, just not in a terribly big way, some file/folder paths that don't exist remain and typically also with random symbols that weren't there. I guess it's similar to the symbols one might see when opening certain incompatible files with Windows' Notepad.
I imagine the only way to fix that is either /command 40405 or re-adding the devices or even starting over from a fresh db. Hell no with how much work it has processed over time..
I don't have to do any sort of modifying of file paths for the second thing, as I'm only adding a fake sub-folder to the root of everything.
But I think I messed up and only realised afterward that I didn't do the sub-foldering correctly.
I probably answered my own question in the same message as I asked it..I just haven't tested it.
Maybe my entire post was pointless and I should've just done the tests first. I do for a bunch of would-be-posts but I decided for some reason to send this one. Laziness and stupidity on my part.
So I think I found the issue for the second thing..
Maybe it was to make the root device letter exist as well as the fake folder's path, I cba to check.
But it turns out when I modified a 200MB+ file with Notepad++ 32-bit it simply failed most of the lines.
It's also much slower, it took a few seconds to modify 1.3mil lines with 64-bit..
32-bit took more than 25min and I just force-exited it cause it was still stuck.
Now that I have a 64-bit modified version, it finished all the lines successfully and at least for one of the EFUs it loaded with Open File List.. properly.
This wasn't the EFU with # in it, but I will modify that one later. Presumably it will work now that I realise this.
I keep trying to use the 32-bit version for larger files as that's the main Notepad++ I use.
Regardless, Monolith still has all the files, so even if it failed it still failed "successfully". In a few weeks maybe I will rebuild Monolith with the fixed EFUs. Unless there are more issues I will find out later, this marks the end of that issue for me.
For the first issue, I will try to do some tests later but it is very likely that when Monolith was rebuilding, I deleted the EFUs I added forgetting that it was meant to index them first. Or Monolith is too corrupt and removes EFU contents on rebuilds if they don't exist in the specified paths.
Monolith IS corrupt, just not in a terribly big way, some file/folder paths that don't exist remain and typically also with random symbols that weren't there. I guess it's similar to the symbols one might see when opening certain incompatible files with Windows' Notepad.
I imagine the only way to fix that is either /command 40405 or re-adding the devices or even starting over from a fresh db. Hell no with how much work it has processed over time..