Quoting mediadogg:

Quote:
One other thing you mentioned. To uninstall any of my plugins, you can use the Windows Control Panel, or you can, for example, navigate to the DVDP \Plugins\BulkEdit folder and execute the unins000.exe program.

Great!

Thanks

Quoting mediadogg:

Quote:
Quoting ljckbailey:

Quote:

Maybe you can tell me, is it possible to cause a secondary sort by click order on columns?

Perhaps sort is via columns left to right? I noticed the "Disc" column was causing issues so I moved it to after the Slot column and that seemed to fix that issue.

Next question: Is there a way to paste into a selected block in the BulkEdit sheet? I noticed ctrl-v does not work. That would make things much easier once I get the sort order dealt with ...

Do you have a Regex primer? I can't see a way to proceed replacing the Location or Slot data with the pattern I need via search & replace. You said that you thought Regex could possibly handle the replacement pattern I need ...

I don't want to impose too much on your valuable time! Again, thanks so much for all your help.

First of all, I chose to spend my time this way. Don't give it another thought. After all, if BulkEdit can't be made to help you, then all my prior efforts are for nothing. So I want to find a way.

Unfortunately the block copy/paste, as I alluded to before, I haven't been able to implement, and the sorting is one column at a time. Sorry.

I'm not sure about the sorting issue. Did you do it manually or via the context menu? I believe my code just scans the column and figures out how many "0"s are needed, and then does the padding. Perhaps you should just reload the flagged profiles and start over. As long as you haven't saved, all your original data is still there.

There are a bunch of Regex tutorials out there. I will Google and update this post with the ones I recognize having been helpful to me in the past.

Code Project Tutorial
MSDN Reference
Regex Tester
Tester with Match / Replace

Using the tester, I created a match pattern in a couple of minutes:  D\d*\s*-\s*\d*

This matches "D" followed by one or more digits, followed by 0 or more spaces, followed by "-", followed by 0 or more spaces, followed by one or more digits. That is the easy one. The harder one is the expression for the replacement. Working ...

Thanks for the link. I'm busy educating myself now!
I have induced a few screen shots of BulkEdit across a sample of my DB sorted by location and then by slot below (A online free file drop service - you have to download the jpg there. Sorry there wasn't a more elegant solution).

Location Sort Screen Shot

Slot Sort Screen Shot

Hopefully you can see the padded "Slot" rows stay highlighted and do not sort as selected.

Thanks again.

Quoting ljckbailey:

Quote:

Hopefully you can see the padded "Slot" rows stay highlighted and do not sort as selected.

Thanks again.

Given that BulkEdit cannot sort on multile columns, your pictures look correct to me. The Slot sort is correct (but screws up Location), and the Location sort is correct (but screws up Slot).

Did I miss something? If so, I'll look more closely ...

Quoting ljckbailey:

Quote:

I guess that my question is why do the "0 padded" cells remain highlighted and sort out of order? If you notice all the non-highlighted cells are sorted correctly. The slot cell are highlighted because of the previous padding operation, or so it seems. It seems like if the padding operation left the altered cell/rows "un-highlighted", the sort may turn out correct.

I don't mind waiting at all if that is where this takes us. Also, I am flexible. I wanted to replace D1 with the string "Page: XX" and 1 - 400 in the slot field with the string "Disk: X" as I previously indicated, but I'm not married to the nomenclature.  I could also concatenate the two strings and just use the location field. Whatever makes the whole thing possible.

Thanks again.

Regex can easily create any text-based nomenclature. It just can't do the modulo arithmetic. To update the two columns, you would run the match code on Location with a specific replace string, then run it again on the Slot column with a different string. If nor for the modulo arithmetic required, the whle thing would be done in a matter of minutes.

I still don't get the sorting issue. Remember, you cannot expect that two columns stay correctly sorted except by accident, because my code does not do it explicitly. It sorts each column independently - I have not looked at the Microsoft code beneath the sort, in order to affect the other columns. I could have, but I haven't so far. The padded cells are highlighted because you have changed them, indicating that a save is required before your actual profiles are updated. If you do a "Save Validated Changes", your actual database will be updated, and the highlighted cells will turn back white, indicating that they have been successfully saved and then retrieved from the database.

Anyways, I'll give the bulk load a go over the weekend, while you continue to look for a constructive way to use what is already there.

We are going a bit in circles. You need to save your changes to get rid of the highlights. Then see what happens. To me, the Slot column is correctly sorted, even with the highlights. One more time, you cannot expect the Location column to remain sorted with the correctly sorted Slot column and vice-versa. It is not a bug - two column sort is just not supported.

Quoting mediadogg:

Quote:
Using the tester, I created a match pattern in a couple of minutes:  D\d*\s*-\s*\d*

This matches "D" followed by one or more digits, followed by 0 or more spaces, followed by "-", followed by 0 or more spaces, followed by one or more digits. That is the easy one. The harder one is the expression for the replacement. Working ...

That's not completely true. This matches also "D-". If you want to get sure that at least one digit is given in front and back of the "-" you should use this: "D\d+\s*-s*\d+"

An asterisk means zero or more, a plus means one or more.

Quoting ljckbailey:

Quote:
Quoting mediadogg:

Quote:
We are going a bit in circles. You need to save your changes to get rid of the highlights. Then see what happens. To me, the Slot column is correctly sorted, even with the highlights. One more time, you cannot expect the Location column to remain sorted with the correctly sorted Slot column and vice-versa. It is not a bug - two column sort is just not supported.

Ok. I save the changes and things look better. It is sorting as expected. Looks like I' may have to just use BulkEdit as a convenient way to hand edit all the fields, definitely better than drilling down into every profile!

Thanks

Yup, it is certainly good for that. Perhaps you can use the global find / replace and Regex to make mass partial changes, and then do the rest by hand ...

For example, you could change Location "D#" to "Page: #", and then manually fix the page #'s. Similarly, you can change the Slot #'s to "Disk: #", and then manually fix the #'s by applying the modulo rule.

Maybe you can play with this, say 50 profiles at a time. Then, if my bulk load comes along before you finish, you can use that to complete the job. It's up to you.

Quoting Mithi:

Quote:
Quoting ljckbailey:

Quote:
Looks like I' may have to just use BulkEdit as a convenient way to hand edit all the fields, definitely better than drilling down into every profile!

1) Have a look at TinyTask a small program that can record keystrokes and repeat them, could do wonders on the slots as it is basically "1 ENTER 2 ENTER 3 ENTER 4 ENTER" again and again, when sorted correct.

2) If you can hold out one or two days longer I'll try if there is something to be done via the XML-Export and how well it can be loaded back into DVD Profiler with BulkEdit.

cya, Mithi

Cool!    Might also be useful in entering a set of DVDs by Title?

In terms of re-importing using BE, that is available now just by editing the standard collection.xml output of DVDP. What some people get messed up on, is that if you don't have an XML editor and you use Excel instead, the ProfileID column will get converted into a floating point number if you don't first reformat the column as text before saving.

In addition to that, I'm reasonably well along on the new BulkEdit menu "Bulk Load Personal Data."

Initial supported fields will be: location, slot, notes, collection#, collectiontype. Input file will be in standard CSV format.

Any others I should add?

(Edit: Ok, it's done. Just waiting a bit to see if any more fields requested.)