PythonScript detect saved/unsaved file state at startup

[Alan Kilborn]

I’m having trouble detecting the saved/unsaved state of the filetabs at Notepad++ 8.4.7 startup using a script. Specifically, I’m being told that I have no unsaved tabs, when I do have some:

Here’s the code as text:

if 1:
    import os
    for (pathname, buffer_id, index, view) in notepad.getFiles():
        notepad.activateFile(pathname)
        state = 'unsaved' if editor.getModify() else 'saved'
        print('on startup, activating a {st} "{fn}"'.format(st=state, fn=pathname.rsplit(os.sep, 1)[-1]))

Can anyone that’s inclined to, try and duplicate this?

Also, if there are any ideas for successful retrieval of the saved/unsaved state of the tabs at startup, I’d like to hear them. I have some “wild” ideas, but I lack simple solutions (the kind I like).

[PeterJones]

@Alan-Kilborn said in PythonScript detect saved/unsaved file state at startup:

Can anyone that’s inclined to, try and duplicate this?

It gives the values I expect: new1 was unsaved but no changes (no content yet), LICENSE was changed and unsaved, new2 was unsaved with changes. And the script output agreed.

Actually, I ran that from the console instead of a script. Run it again from a saved script:

Still works for me.

Maybe it’s a timing issue for you. Try putting a quarter or half-second sleep between the activateFile and the getModify – I know that some of my scripts, I have to give Notepad++ enough time to switch active files and potentially re-lex the document (especially if it’s big)

[Alan Kilborn]

@PeterJones said in PythonScript detect saved/unsaved file state at startup:

Maybe it’s a timing issue for you. Try putting a quarter or half-second sleep

I have never before found this kind of “sleep” to be necessary, in PythonScripting.
But…I tried it, using a full 2 seconds of sleep.
No change for me. :-(

Then I thought maybe I should try it with PS 2.0, because, from my earlier output one can see that I was using PS 3. No change for me from that test either (file still all present “saved”).

I’m stumped. :-(

new1 was unsaved but no changes (no content yet)

This tab would be empty of content, correct?
When I try having such a tab, Notepad++ won’t even “bring it back” after a restart – the tab is eliminated. I suppose that makes sense: if you’ve created a “scratch” tab, but haven’t put anything in it, why should N++ keep it around? But…this is not your experience, so…(I don’t know what to make of that).

Thanks for doing some experimentation on my behalf, Peter Jones.

[PeterJones]

@Alan-Kilborn ,

Sorry, that wasn’t at startup; that was just “live”.

When I did enable periodic-backup and made sure that I had toggled through all the files so it made the 7sec backup for the edited ones, then restarted Notepad++, you are right: it didn’t bring back new 1, and all the files claimed to be saved, even though they are properly shown as edited/unsaved in the tabbar list.

on startup, activating a saved "Peter's Scratchpad.md"
on startup, activating a saved "LICENSE"
on startup, activating a saved "change.log"
on startup, activating a saved "new 2"
on startup, activating a saved "23908-saved-unsaved-list.py"
on startup, activating a saved "startup.py"

So at startup, the answer is wrong. And if I rerun the script manually without making new changes, it still shows them as all being saved.

So yes, I concur that getModify() apparently returns whether it’s been changed since the load, rather than whether it’s been changed from the “saved” state.

[Ekopalypse]

@Alan-Kilborn

I think the safest way is something like this…

# python3
editor.getText() == open(WHAT_EVER_FILE, 'rb').read().decode()

[Alan Kilborn]

@Ekopalypse

Yes. Before when I said I have some “wild” ideas…, yours was one of them. I didn’t want to go that far to solve it. :-)

Maybe I’m attempting to solve something that isn’t really a problem for me, anyway, because I don’t operate with a setup that permits unsaved files when exiting Notepad++ (and thus there is no possibility to have unsaved files when N++ starts up).

However, I was working on a script for posting that I was trying to make work in all modes…

[Alan Kilborn]

Another “wild” idea is to read the N++ “backup” folder for files present there.

Example: If these files are there:

change.log@2022-12-28_064818
new 2@2022-12-26_075922

One can infer that the change.log and new 2 filetabs have unsaved changes.

[Alan Kilborn]

After long delay, I made an official Notepad++ issue out of this:

https://github.com/notepad-plus-plus/notepad-plus-plus/issues/16514

In the issue, I switched to using NppExec to reproduce, rather than PythonScript, lest it be interpreted as a PS bug and not a N++ bug.