<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">On 9/15/25 12:36 PM, Brendan Coupe via
KMyMoney-devel wrote:<br>
</div>
<blockquote type="cite"
cite="mid:0074c156-06e7-4a57-bb83-a5e4feefaa3a@Coupe7.com">On
2025-09-15 12:31 AM, Thomas Baumgart via KMyMoney-devel wrote: <br>
<blockquote type="cite">On Mittwoch, 10. September 2025 00:03:29
CEST Brendan Coupe via KMyMoney-devel wrote: <br>
<br>
<blockquote type="cite">I have no idea how to fix this. <br>
<br>
grep A000432 MyMoney.kmy | grep '<ACCOUNT>' <br>
<br>
returns nothing. <br>
<br>
grep A000432 MyMoney.kmy | wc <br>
<br>
returns 9 lines <br>
</blockquote>
<br>
I read this before and wondered what you were doing to see <br>
nothing but finding 9 matches and just spotted the problem
today: <br>
<br>
grep A000432 MyMoney.kmy | grep '<ACCOUNT>' <br>
<br>
needs to be changed to <br>
<br>
grep A000432 MyMoney.kmy | grep '<SPLIT' <br>
<br>
because the account is missing and there won't be an ACCOUNT
item <br>
(unless there is and it is not loaded for other reasons, but I
doubt <br>
that). <br>
<br>
That should show those 9 matches as well (maybe less, but that
is <br>
not so important right now). Using <br>
<br>
grep -n A000432 MyMoney.kmy | grep '<SPLIT' <br>
<br>
provides you the line numbers of those splits and you can try to
<br>
analyze of which type this account needs to be by looking at the
<br>
transaction. <br>
<br>
</blockquote>
grep -n A000432 MyMoney.kmy | grep '<SPLIT' <br>
<br>
returns nothing <br>
<br>
grep -n A000432 MyMoney.kmy | grep SPLIT <br>
<br>
returns 9 lines. <br>
<br>
Each of the 9 lines has 1,500 plus characters. The length is
pretty unique in my KMM file although I have found a few that are
not part of the lines with A000432. My guess is A000432 is not the
only problem. <br>
<br>
grep -n '.\{1500\}' MyMoney.kmy | wc <br>
<br>
returns 77 lines that are over 1500 characters long so hopefully
they all don't need to be fixed. <br>
</blockquote>
<p>In my personal xml file, all lines over 1200 characters are <span
style="font-family:monospace"><span
style="background-color: rgb(255, 255, 255);"><PAIR
key="reconciliationHistory" value="very long list of dates and
amounts"/>. If that is true for you, those lines are not
relevant to your problem. However, none of those lines have
any account ID in them, so this finding may end up not
relevant to your data.</span></span></p>
<p><span style="font-family:monospace"><span
style="background-color: rgb(255, 255, 255);">My understanding
is you have a number of bad TRANSACTIONS where one (or more)
of the SPLITs within each one has an account="Annnnnn" where
there is no <ACCOUNT entity with that id in the data file.
For the very long lines, do they look like reasonable lines
which just happen to have long strings as the value of
attributes, or do they look like there is actually something
wrong with them? For the lines with SPLIT but not "<SPLIT"
what is before SPLIT?</span></span></p>
</body>
</html>