<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">I finally managed to clean my tags.<br>
<br>
Here is how I proceded: first, as Maik suggested, ensure there are
no duplicates except at root level. Once this is done, simply
delete the wrong tags from the root. If dK recreates a wrong root
tag immediately after you delete it, it means that you missed a
duplicate somewhere. I had found another way by dragging the wrong
root tag to where it should be, thus triggering a merge, but I
deduce from Maik's explanation this dK behaviour only triggers
duplicates, so deleting a duplicate will always leave the original
untouched, and simply deleting the wrong root tag is simpler than
merging it.<br>
<br>
Ensuring that there are no duplicates may not be easy, because
there are situations where using the exact same name in different
places is makes a lot of sense, but there is currently no way to
get around this limitation in dK (at least not that I know of).
For example, I had 2 "piscine", one for our swimming pool, one as
a general swimming pool tag. I had to rename one of those in order
to ensure unicity.<br>
<br>
I don't understand why dK does this. When I check A/A1/X in the
UI, why can't dK simply do the same thing in the selected photos?
I tell dK where tag X is, so why explore the full tags list? Why
does it matter if there exists for example B/B1/X too? I guess
such a complicated mechanism has been included deliberately, but I
fail to see to what purpose. Note that dK does what I ask him to
if I ask him to add a tag which is used in more than one place. In
my example above, dK adds the tag A/A1/X to the selected photo.
But it adds a tag X too which I did not as for. What for?<br>
<br>
<pre class="moz-signature" cols="72"></pre>
Here are some remarks:<br>
<ol>
<li>dK sometimes created duplicates not only at root level, but
also in intermediate levels. Those "intermediate" duplicates
were always somewhere in the path to a correctly placed tag. </li>
<li>the tags created by dK by duplication have a character
coding issue. I found that many (all?) the duplicated tags
which contained accentuated characters had those replaced with
an inverted question mark.</li>
<li>when selecting a group of photos and changing tags in a way
which affects only a subset of the selected photos, dK updates
*all* the selected photos, not only the subset; I feel this is
both unnecessary and wrong (nothing was changed in the
selected photos which weren't part of the subset). Example:
photo1 has tags A and B and photo2 has only tag A, I select
both photo1 and photo2 and I check tag B, dK updates both
photo1 and photo2, when actually only photo2 was really
changed.<br>
</li>
</ol>
<br>
<pre class="moz-signature" cols="72">--
Frédéric Da Vitoria</pre>
<br>
On 07/12/2024 22:05, Frédéric Da Vitoria wrote:<br>
</div>
<blockquote type="cite"
cite="mid:0c3ef6aa-34cb-4795-952b-3f96a93f8317@free.fr">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<div class="moz-cite-prefix">Hello Maik,<br>
<br>
my replies/questions below:<br>
<br>
On 07/12/2024 19:21, Maik Qualmann wrote:<br>
</div>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">Hi Frédéric,
I can't find any errors with these images.
The tag "piscine" is created by the image "piscine
20200311_092408_vHDR_Auto.jpg" in the root as a single tag.</pre>
</blockquote>
Let me explain again, and expand at the same time: <br>
- I don't want the "piscine" at the root level, only as a
hierarchical tag<br>
- the tag piscine has 2 "legit" positions:
in Sujets/technique/piscine and in Lieux|France|Charente
Maritime|Le Mung|piscine. <br>
- I did not create the tag "piscine" at root level. dK decided to
create it by itself (but you have explained below how and why this
happens). <br>
- if I select "<span style="white-space: pre-wrap">piscine </span><span
style="white-space: pre-wrap">20200311_092408_vHDR_Auto.jpg</span>
" in the thumbnails view and uncheck the root piscine, dK
recreates it as soon as I click on Apply. Again, I know now why,
but I don't know how to fix it.<br>
<br>
Hmm, BTW, I just copied those 2 paths above from "<span
style="white-space: pre-wrap">piscine 2</span><span
style="white-space: pre-wrap">0200311_092408_vHDR_Auto.jpg", and I notice the difference in separators. I guess this can't be good, but do </span>you
know any method to fix this? I guess that many photos are affected
by this...<br>
<br>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">This is actually how it is shown in the metadata and can be easily corrected.</pre>
</blockquote>
Well, the method may be easy, but I am unable to find it. Could
you explain?<br>
<br>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">I don't have the tag "mésange bleue" in the root with this set of images.</pre>
</blockquote>
It is there. Search for "sange" and you will find it :-) Obviously
a character encoding issue, but here again I don't know what to do
about it. This could be inherited from the photo tagging software
I used before, but I don't have it any more.<br>
<br>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">Please note the following: you should always read an entry for tags in the
advanced metadata that supports hierarchical tags. Either LR hierarchical tags
or digikam TagsList.</pre>
</blockquote>
For several years now, dK has been my only way of editing tags.
The only other graphical tools I use are Gimp and Hugin (and
sometimes RawTherapee or darktable) and AFAIK none of those know
what a tag is. I never used hierarchical tags before, hierarchical
tags were one of the reasons I chose dK. Currently, dK is set to
Xmp.digiKam.TagsList. Here are my tags settings:<br>
<img src="cid:part1.0b2RfXQk.R0Q8l85l@free.fr" alt="" class=""><br>
Does Xmp.lr.hierarchicalSubject mean LR hierarchical tags? If so,
IIUC, I should uncheck this, correct?<br>
<br>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">Please note the following: if an image has a single tag entry, digiKam tries
to find the name in the digiKam tag tree. However, if this exists multiple
times in different hierarchies, the tag logically cannot be assigned to one. In
this case, the single tag is created in the root.</pre>
</blockquote>
I guess this is the cause of the issue. <br>
<br>
But how should I handle this? Should I ensure there aren't 2
different tags at different places? How can I clean my tags even
assuming I <span style="white-space: pre-wrap">am willing to replace </span>Lieux|France|Charente
Maritime|Le Mung|piscine with Lieux|France|Charente Maritime|Le
Mung|piscine_dv?<br>
<br>
Is there a tool showing which tags are "duplicated"?<br>
<br>
Shouldn't this dK behaviour be enhanced in the future? Imagine a
tag named "garden", there are bound to be many places in a
hierarchy where such a tag would be appropriate. And one would not
necessarily be willing to give a unique name to each "garden",
because being able to pull all the "garden" from all over the
hierarchy could be meaningful.<br>
<br>
<pre class="moz-signature" cols="72">--
Frédéric Da Vitoria
</pre>
<br>
<span style="white-space: pre-wrap">
</span><span style="white-space: pre-wrap">
</span>
<blockquote type="cite"
cite="mid:2526214.XAFRqVoOGU@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">Maik
Am Samstag, 7. Dezember 2024, 10:13:51 Mitteleuropäische Normalzeit schrieb
Frédéric Da Vitoria:
</pre>
<blockquote type="cite">
<pre wrap="" class="moz-quote-pre">Hello Maik,
Thank you for looking into this.
Here are 8 examples. I did not touch anything in them since the problem
occured except changing their file names, and I did so after copying
them into a directory outside of dK's scope. So these should exactly
what dK sees.
I included:
- 4 photos with the "piscine" tag, taken with 4 defferent cameras, each
time the "piscine" is duplicated
- 2 photos with a "mésange bleue" tag, one where this itag is duplicated
at the root level, one where there is only one copy of the tag; I did
not do anything recently with those old photos so I don't have any idea
why dK decided to duplicate those tags
- 2 photos with a "macaque à longue queue" tag, one where this itag is
duplicated, one where there is only one copy of the tag; I did not do
anything recently with those old photos so I don't have any idea why dK
decided to duplicate those tags either; here, dK duplicated tag under
"animal" instead of putting the duplicate in the root.
I noticed that the duplicated tags had character set issues (when there
were non-standard characters, of course).
Note that the problem occurs when I select a group of photos and edit
the tags by checking / unchecking the tags in the right pane. AFAICT,
this does not happen when I merge tags using "Tags" in the left pane. I
guess I should refrain from using the right pane to edit tags until this
issue is solved.
<a class="moz-txt-link-freetext"
href="https://drive.google.com/drive/folders/1K2Q9Nf0F-D5UEuVK8CqSX5SG_9P3eCsc?usp"
moz-do-not-send="true">https://drive.google.com/drive/folders/1K2Q9Nf0F-D5UEuVK8CqSX5SG_9P3eCsc?usp</a>=sharing
</pre>
</blockquote>
</blockquote>
</blockquote>
<br>
</body>
</html>