Consistent history and tabs

[milen-prg]

Is your feature request related to a problem? Please describe.

1. In the history are visible files, that already were deleted from the disk. Also in the history can be repeating items, must not be duplicates (the old one must be deleted, when reopened recently).
2. There can be opened several tabs with the same file.

Describe the solution you'd like

1. Let have button to automatically clear the history of non existing files. (If be automatically, as I use history of hundreds files (IMPORTANT EXTREMELY USEFUL FEATURE (!), please don't change), I suppose it will take too much time to wait at each start...
2. If I open file it must check if it is already on some tab and switch to it, instead to reopen in new tab.

Describe alternatives you've considered
Now I can clear with context menu an item in history.

Additional context

[GitHubRulesOK]

there is no way an app can know if a file in history is going to come online in a minute or next year.

This was well known in the past.
Your device may be using virtual or network or usb drives and the file only be dead for a few seconds but rise again later. the logical solution is to parse the file with an app that lists current and non current and then ask one by one if non current were deleted.

Another option is to save files from history that exist in each drive but again that's a niche requirement, more easily solved by delete ALL history and add the ones you still have. you can scan the file and build a list of entries that you can use to build a startlist

So this shows those that are in my short history but some may only be found in this user session

type sumatrapdf-settings.txt |findstr "FilePath"
                FilePath = C:\Users\WDAGUtilityAccount\Downloads\HAUTES_FAGNES_Eupen_Paturage_FR_DE_NL_page_1.jpg
                FilePath = C:\Users\WDAGUtilityAccount\Downloads\HAUTES_FAGNES_Eupen_Paturage_FR_DE_NL.pdf
                FilePath = L:\msys-64\home\K\signed.pdf
                FilePath = C:\Users\WDAGUtilityAccount\Downloads\mupdf-sign-out.pdf
                FilePath = C:\Users\WDAGUtilityAccount\Downloads\Bluey.pdf
                FilePath = C:\Users\WDAGUtilityAccount\Downloads\Bluey-hl.txt
                FilePath = C:\Users\WDAGUtilityAccount\AppData\Local\Temp\work.pdf
                FilePath = C:\Users\WDAGUtilityAccount\AppData\Local\Temp\totals.txt
                                FilePath = L:\msys-64\home\K\signed.pdf
                                FilePath = C:\Users\WDAGUtilityAccount\Downloads\HAUTES_FAGNES_Eupen_Paturage_FR_DE_NL.pdf
                                FilePath = C:\Users\WDAGUtilityAccount\Downloads\HAUTES_FAGNES_Eupen_Paturage_FR_DE_NL_page_1.jpg

[milen-prg]

I didn't understood you. I have documents on my disks, local discs. I opened them, read them, they are in history. Good. BUT, in one moment, I deleted some of these files from the disk with file explorer. So there can be button that starts this function:

1. For each history item check if the file exists on the path (this path is in history, from where it tries to reopen it).
2. If the file does not exists - delete the item from the history.
   That's all.

May be I'm wrong in my thoughts, so close this request. Only please do not remove the infinite history from the home tab and the list, it's absolutely necessary for my work and is one of the best features of the SumatraPDF for me!

[GitHubRulesOK]

@milen-prg
to show a simple purge tool I have done something you can TRY put a copy of the big settings file in a folder with this script https://github.com/GitHubRulesOK/SumatraPDF-Plus/blob/master/purgeHistory.vbs (download button is top right) and doubleclick the script, it should backup the settings file then test and remove dead entries so the final settings is a CLEANER history of current files !!

[milen-prg]

@GitHubRulesOK
Seems the script works (it deleted about half of my history, it's very hard to find if it removed some existing file - I have documents with complex names - cyrillic or with icons and hieroglyphs, I hope that it checks properly when the name is complex).
The script, but keeps the duplicates, it must not add file which was already added.

[GitHubRulesOK]

@milen-prg I tried to make it accept global file names but have no large mix of chinese asian and cyrillic to use for test it is supposed to ONLY remove a block where a file cannot be found like DIR blah.ext it could ONLY be that simple as you may have a laptop at home and a file in the office and no way the script could know your at home.
This is why there is no built in manager for the task as every user wants different.
I tried to be close to your description of keep files that exist now.

once proven reliable for your needs it could potentially be added as a shortcut to run at any time but that is then more risky if it fails to deliver during an update

[milen-prg]

@GitHubRulesOK I see this function very useful, my history from over 300kb now us about 80kb. There was so much distraction broken records. Also I checked several files with cyrillic and hieroglyphic names, and they are properly processed. If the script deletes also the duplicate records, it will be perfect.
I think that this feature is useful, because, usually the SumatraPDF and the documents are locally on the local disks on the local PC of anyone and in this way the history (and if implemented - the tabs) will be kept consistent. There is also some security aspect - if someone deleted the file, it must not be kept in the history as track!

[GitHubRulesOK]

@milen-prg it does not check duplicates as such ( but in theory there should not be 2 "master" entries for one file unless in 2 locations, if so they count as 2 separate files) and you may find during a session it does not remove SESSION files that have just been deleted only their "master" entry However that does mean next start there should not be a "master" entry and the session entry will be automatically removed by SumatraPDF itself.