Author Topic: Clipboard Help+Spell Extended Clipboard Format Data Discussion Thread (Read 166,659 times)

mouser · « **Reply #50 on:** February 20, 2020, 12:53 PM »

Next version will include a little clipboard viewer tool you can bring up:

mouser · « **Reply #51 on:** February 20, 2020, 12:54 PM »

Quote from: tomos on February 20, 2020, 11:05 AM

Can we actually *stop* image thumbnails showing in the quick-paste menu /sub-menus?

Try these options:
[attach]

mouser · « **Reply #52 on:** February 20, 2020, 12:55 PM »

Quote

nice would be the ability to name image (and vector) clips. Could e.g. then be saved in favourites with clarity as to what exactly it is (often not clearly visible in thumbnail).

You can customize clip titles.

tomos · « **Reply #53 on:** February 20, 2020, 01:06 PM »

^ thanks mouser. I obviously need to get to know CHS better

mouser · « **Reply #54 on:** February 21, 2020, 11:02 AM »

New version uploaded, with clipboard blob viewer. You can access it through the View menu.. Let me know what you think and of course report any bugs.

djMot · « **Reply #55 on:** February 21, 2020, 02:23 PM »

Could we get some scroll bars on the HTML Source preview tab? (and any other tab that doesn't have them) Thanks.

I am also seeing duplicate clips, although I'm not sure it's all that big of a deal. Would be nice to get it collapsed down into one, though. OR, does the grid control you're using support multi-level (aka grouped) entries? If so, you could put clips that have multiple captures in an expandable row.

Example duplication:

Awesome updates so far! THANK YOU for the effort and all the new features!

mouser · « **Reply #56 on:** February 21, 2020, 02:44 PM »

~~Can you see if the duplicate clips have identical clob hash values?~~
I see the hashes are different. But now with the new Clip Blob interrogator you should be able to view both blobs with hashes for each format and see WHERE they differ. If they only differ in a format chunk that is irrelevant, I will soon have options that let you ignore those formats.

djMot · « **Reply #57 on:** February 21, 2020, 05:04 PM »

Yes, the blob hash values are all different.

Here's an interesting twist. After I posted the last screenshot, two more of the exact same clip showed up out of nowhere spaced a couple of minutes from the original two.

djMot · « **Reply #58 on:** February 21, 2020, 05:11 PM »

Quote from: mouser on February 21, 2020, 02:44 PM

~~Can you see if the duplicate clips have identical clob hash values?~~
I see the hashes are different. But now with the new Clip Blob interrogator you should be able to view both blobs with hashes for each format and see WHERE they differ. If they only differ in a format chunk that is irrelevant, I will soon have options that let you ignore those formats.

Ah okay. Is the blob interrogator in there yet, or is that yet to come?

mouser · « **Reply #59 on:** February 22, 2020, 02:12 AM »

Quote

Could we get some scroll bars on the HTML Source preview tab?

yep, done.

Quote

Is the blob interrogator in there yet, or is that yet to come?

it's there, in view menu.

Quote

Here's an interesting twist. After I posted the last screenshot, two more of the exact same clip showed up out of nowhere spaced a couple of minutes from the original two.

Ok I see these have different hashes, so the contents of the seemingly duplicate clips are different. See my reply above about this, and go into the blobs with the blob interrogator and let us see which formats are different...

I have noticed that MS excel this can happen also -- its happening maybe because these MS Office apps are doing something strange.. perhaps copying to the clipboard in two separate events in quick succession, for some reason I don't understand. Or perhaps CHS is trying to read the clipboard before the app finishes with it..

Here's something that may fix it. I've added a new option to increase the delay between the triggering of a clipboard event and CHS acting on that trigger, so that 2 quick success triggers will only result in the final capture:
[attach]

mouser · « **Reply #60 on:** February 22, 2020, 02:55 AM »

Or the problem could be my hash calculation on blobs has some bugs.

mouser · « **Reply #61 on:** February 22, 2020, 06:43 AM »

Updated with some miscellaneous bug fixes to handling of duplicates and other things.

Most notable fix was that the new multi-format code was not deleting external image files associated with clips when the clips were deleted and removed from the recycle bin, so if you have image clips from a previous alpha version, you'll need to delete those image files manually when the clips are deleted.

tomos · « **Reply #62 on:** February 22, 2020, 05:51 PM »

Mouser, this all just reporting CHS treatment of vector clips -- summary: it appears to work correctly but still saves duplicates.

Testing with Inkscape 0.92.4 (5da689c313, 2019-01-14)

With Clipboard Trigger Copy Delay set to 400ms, two clips get copied (1 x Ctrl+v).
With the same vector drawing and delay set to 500ms -- one clip gets copied.

Also note: Clips remain duplicated (setting is to update clip)
All clips appear identical when pasted:

The only difference in blob info is time taken for the queries. One clip is 0.01kb smaller, other two same size. All have different hashes.
Mouser ~~I cant find where to compare blob info -- you say under the view menu~~ never mind it's under Tools menu. Can I actually compare two clips in this 'Interrogator' -- IIUC I can compare the clip that is selected in the main CHS window and the current clipboard clip. Here the difference between two of those duplicated clips shown above (they both paste correctly):

One seems to have much less formats (?)
Both appear to have a PNG -- this is not shown in the CHS 'Memo' pane
EDIT// note the blob info of both of these clips is the same in the CHS memo pane -- apart from the milliseconds taken for each step

mouser · « **Reply #63 on:** February 23, 2020, 12:24 AM »

Quote

Can I actually compare two clips in this 'Interrogator'

There is no explicit compare function -- it's just the list of formats and hash values you would compare manually.
So what I'd like you to do is look at the two *seemingly* duplicate entries within CHS and view the database views of the blobs from the interrogator window, and see what's different in the interrogator list of formats for those two clips.

Can you show me screenshots of the interrogator for each of the two clips?

It's EITHER going to be the case that the clipboard blobs are different, or that CHS has a bug in dupe detection.

Quote

With the same vector drawing and delay set to 500ms -- one clip gets copied.

So I think that probably suggests that your having that same problem with apps that put a lot of info on the clipboard where CHS first grabs the clipboard and then the app adds more info and chs has to go again. As before, there are a couple of options here -- either forcing a longer delay for CHS from some apps, or some process where i auto-delete the prior copy when there are 2 in close succession.

mouser · « **Reply #64 on:** February 23, 2020, 01:37 AM »

I'm going to go ahead and move the clipboard capture delay option up to a more prominent location in the "More Capture Options" tab, and add an explanation for it; that will make it easier for people to notice.

mouser · « **Reply #65 on:** February 23, 2020, 04:04 AM »

Slight tweek in new update; *might* (but probably won't) improve the issue of double clips.

tomos · « **Reply #66 on:** February 23, 2020, 06:27 AM »

Quote from: mouser on February 23, 2020, 12:24 AM

So what I'd like you to do is look at the two *seemingly* duplicate entries within CHS and view the database views of the blobs from the interrogator window, and see what's different in the interrogator list of formats for those two clips.

Can you show me screenshots of the interrogator for each of the two clips?

It's EITHER going to be the case that the clipboard blobs are different, or that CHS has a bug in dupe detection.

only difference I can see there is the milliseconds taken. There could be a difference in the png -- I havent looked for them (from Inkscape they are not shown in CHS).
Will try new version later

mouser · « **Reply #67 on:** February 23, 2020, 06:51 AM »

Quote

only difference I can see there is the milliseconds taken.

Don't trust that view, it doesnt show HASH values -- I will add that now.

Ok I've uploaded a new version that shows blob format hashes now in main Blob tab; should make it easier to see if there are differences in clips.

mouser · « **Reply #68 on:** February 23, 2020, 11:44 AM »

Updated with improved ability to set limits on different format sizes.

mouser · « **Reply #69 on:** February 23, 2020, 01:14 PM »

I found an issue that *might* be responsible for the double duplicate captures. Can you download and try the new version and see if it still occurs even if you lower the delay to 200 or so.

tomos · « **Reply #70 on:** February 23, 2020, 03:26 PM »

Clipboard Help+Spell
v2.47.15 by DonationCoder

Quote from: mouser on February 23, 2020, 01:14 PM

I found an issue that *might* be responsible for the double duplicate captures. Can you download and try the new version and see if it still occurs even if you lower the delay to 200 or so.

with delay set to 200 only one clip gets copied.

I still get duplicates if I copy the same vector clip twice:

Code: (Generic)

39.12kb (5 formats)

5 format chunks.

"image/x-inkscape-svg" [C507]: Hash 41AFB00E | 2.88kb (query 31ms)
"PNG" [C49B]: Hash F408A94C | 36.13kb (query 31ms)
"CF_DIB" [0008]: Hash 00000000 | 0b (query 31ms)
"CF_ENHMETAFILE" [000E]: Hash 00000000 | 0b (query 31ms)
"CF_METAFILEPICT" [0003]: Hash 49E26D2A | 16b (query 0ms)
-----

Extended (non-redundant) data save size: 39.12kb
Extended clipboard scan time: 124ms
Hash: 30403722

Code: (Generic)

39.12kb (5 formats)

5 format chunks.

"image/x-inkscape-svg" [C507]: Hash 89894DD6 | 2.88kb (query 31ms)
"PNG" [C49B]: Hash F408A94C | 36.13kb (query 31ms)
"CF_DIB" [0008]: Hash 00000000 | 0b (query 31ms)
"CF_ENHMETAFILE" [000E]: Hash 00000000 | 0b (query 31ms)
"CF_METAFILEPICT" [0003]: Hash 16F78EE7 | 16b (query 0ms)
-----

Extended (non-redundant) data save size: 39.12kb
Extended clipboard scan time: 124ms
Hash: CA51C702

Some hashes (lines 5 and 9) are different there, not sure what that means... or why they should be different, but I presume that's to do with the editor involved.

mouser · « **Reply #71 on:** February 24, 2020, 03:45 AM »

I've added a new tab to the main view so you can see more easily the values of all the fields for the clip, whether they are shown in the grid or not:

[attachthumb]

tomos · « **Reply #72 on:** February 24, 2020, 07:13 PM »

Sorry, not sure if that was directed to me, in case, here again two duplicate clips:

mouser · « **Reply #73 on:** February 25, 2020, 04:48 AM »

Since it says the hashes of the CF_METAFILEPICT are different and are only 16 bytes, I'm thinking that actually what happens is that for CF_METAFILEPICT, the data is NOT actually the 16 bytes, but rather that it's a handle to data, and so CHS needs to do some special processing to grab that metafilepict data to store and restore it properly. I'll work on it.

wraith808 · « **Reply #74 on:** February 25, 2020, 05:30 AM »

What hasn't been looked it is do both clips restore correctly when pasted somewhere that is supposed to be able to interpret the data?