Hi,

I encountered strange behavior in the GUI (version from September 30 of 7.5 beta):

In my database of collected node-ids ('collect'), I had to remove some thousands, because the documents the nodes where referring to in the main database ('DB') had to be removed as well.

So I had a database ('remove') with the documents like this:

<list> <item>abc.xml</item> <item>def.xml</item> </list>

And then I used this for the 'collect' database:

for $i in //entry let $n := $i/node let $d := replace(db:open-id('DB', $n)/base-uri(), 'DB/', '') where db:open('remove')//item[text() = $d] return delete node $i

I tested it (it removed the correct nodes) and then applied it to the whole thing and everything worked fine.

In a second step, I deleted the respective texts from the original DB (to be precise, I created a new one without some texts and with whitespace chopping OFF, because the whitespace problem was another issue).

And then I have to change the node-ids in my 'collect' database, so I tried this:

for $node in //entry/node let $id := data($node) let $path := replace(db:open-id('DB', $id)/path(), 'Q{.*?}', '*:') let $base := replace(base-uri(db:open-id('DB', $id)), 'DB', 'newDB') let $newid := xquery:eval(concat('db:node-id(doc("', $base, '")', $path, ')')) return replace value of node $node with $newid

It worked OKish on my test data: I removed irrelevant nodes first, made a copy and than applied the node-changing. But I got an error message, that some resources would not exist -- the ones I wanted to remove and thus the ones for which no node should be in 'collect' anymore. I created new test data, removed the nodes and immediately applied the changing, and then it worked.

For the whole data, I successfully removed the nodes, I thought. And then regardless of using this database or making a copy and using this, I get an error message that a resource would not exist.

How can I find out (in the GUI) where exactly the error occurs, i.e., for which $node? In the query info I only see "Stopped at line 1, column 14 in ...xq"

Somehow, nodes that should have been deleted seem to still be there. The names of the files I created 'DB' from in the first place included special characters and spaces. The file I list the documents to be removed is a UTF-8 file where the file names appear also with special characters and spaces. Maybe the error is in there:

One file name is "Brontë,-Charlotte.xml". Apparently, matching doesn't work. When I list the base-uri the node-ids belong to, I get Brontë,-Charlotte.xml, but the node will not be deleted, so it doesn't match the respective entry in the corresponding list. So how do I deal with this?

Best regards

Cerstin