This PR replaces all instance of xml-rs and xmltree with (lower level) quick-xml.
I couldn't find any benchmark but it is supposed to be faster / use less memory.
On the other hand, it might be more complicated to read (in particular the xmltree replacement).
No offense taken if you prefer NOT to merge it. This is just a proposal.
If someone has a realistic scenario to test, I would be happy to add a benchmark.
This PR replaces all instance of xml-rs and xmltree with (lower level) quick-xml.
I couldn't find any benchmark but it is supposed to be faster / use less memory. On the other hand, it might be more complicated to read (in particular the xmltree replacement).
No offense taken if you prefer NOT to merge it. This is just a proposal.
If someone has a realistic scenario to test, I would be happy to add a benchmark.