I just about have a complete account file from MS Money to line up with a KMM export. Here is the QIF export settings ..

Date format: %m/%d/%yyyy
Apostrophe handling: 2000-2009

However, dates from MS Money 'loose' QIF export

D2/24'2005
D3/3'2005

and the dates from a KMM export, using the export settings above.

D02/24/2005
D03/03/2005

so MS Money is stripping leading zeros and I have no control over that. I need to somehow tweak the KMM QIF export settings to also strip leading zeros. Also I need to find out how to 'force' the apostrophe from a KMM export.

The QIF export is a plugin, so maybe I need to find out how to modify that. I'm already running 5.1.2, so the file at https://altlinux.pkgs.org/p10/classic-x ... 4.rpm.html may not be an update ??

So it would appear that the KMyMoney QIF export has a bug, as the apostrophe was not included in the date format.

I do not fully understand the post, but perhaps this will be useful. I think these two lines are not formatted correctly:
D2/24'2005
D3/3'2005

These two lines are formatted correctly:
D02/24/2005
D03/03/2005

Using regular expressions in a text editor would convert the incorrect lines to the correct format:

For Month
Find what:      D([1-9])/
Replace with: D0\1/

For Day
Find what:      /([1-9])'
Replace with: /0\1'

Find what:       /([0-9]{2,2})'
Replace with: /\1/

jehoshua wrote:So it would appear that the KMyMoney QIF export has a bug, as the apostrophe was not included in the date format.

No, it's not a bug because the apostrophe itself is only a workaround to differentiate the century when only two-digit year information is used. KMyMoney provides 4 digits for the year and therefore doesn't need the apostrophe in this context.

Stephen Leibowitz wrote:I do not fully understand the post, but perhaps this will be useful. I think these two lines are not formatted correctly:
D2/24'2005
D3/3'2005

These two lines are formatted correctly:
D02/24/2005
D03/03/2005

Yes, correct.

Stephen Leibowitz wrote:Using regular expressions in a text editor would convert the incorrect lines to the correct format:

For Month
Find what:      D([1-9])/
Replace with: D0\1/

For Day
Find what:      /([1-9])'
Replace with: /0\1'

Find what:       /([0-9]{2,2})'
Replace with: /\1/

I tried all of those in Kate, a search/replace, but the 'find' didn't find anything. Some test data at https://pastebin.com/raw/FnueXARj

There are a number of Python solutions at https://python-forum.io/thread-38549.html

ipwizard wrote:No, it's not a bug because the apostrophe itself is only a workaround to differentiate the century when only two-digit year information is used.

Possibly "Bug" was too strong a word , and a perceived "limitation" of the QIF export plugin would be more appropriate ? Meaning the plugin has no provision to span a century. That said, I just exported from an account that spans a century and the QIF data looks fine. So, the "Apostrophe Handling" did nothing in that export, because AFAIK the KMM (XML) data stores 4 digit year.

Hmm, ..so the "Apostrophe Handling" is only for imports I guess.

ipwizard wrote:KMyMoney provides 4 digits for the year and therefore doesn't need the apostrophe in this context.

Oh yes, .. I should have kept reading...

You have to “Enable regular expressions” in Kate by clicking the {} icon. Also, use Replace Checked. The three operations should be done in the order shown in my post: Month, Day, and “apostrophe” (third search/replace).

The Day regular expression did not find any matches in the pastebin test data, because that data already had two digits for the day field in each of the entries. But in the first post of this topic, the second example had only one digit: “D3/3'2005”. My Day regular expression corrects that, by inserting a leading zero.

Stephen Leibowitz wrote:You have to “Enable regular expressions” in Kate by clicking the {} icon. Also, use Replace Checked. The three operations should be done in the order shown in my post: Month, Day, and “apostrophe” (third search/replace).

I didn't know Kate can do that. All the examples you provided worked on another file, thanks. No doubt there is a method to use those 3 regex's in Python; I see

Code: Select all

re.sub(pattern, repl, string, count=0, flags=0)

for example. And only do the regex on the "D" records.