You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am using transformations for nlog.config and I have some conditions that check the length of the thread name using the ">" symbol. slow-cheetah changes these even with no transformation at all. Here is an example:
> is the escaped version of > and should be used anywhere you need a > symbol in xml. You can use the unescaped version in some circumstances - such as in an attribute value, but generally it should be avoided.
It is not 'damaged' as an xml parser will correctly unescape the attribute value when it is read.
I'm reasonably certain that the way transforms are done, the whole document is loaded by a parser, then transformed then written back out, so while the specific node won't be touched if no transform is applied, the whole file is written back out with full escaping applied.
@CZEMacLeod Thanks for the comment, makes sense. Seems to be working as well. All nlog samples seem to use > without escapting it which lead to this confusion but it's a problem of these samples.
I am using transformations for nlog.config and I have some conditions that check the length of the thread name using the ">" symbol. slow-cheetah changes these even with no transformation at all. Here is an example:
nlog.config:
nlog.Debug.config (no transformations at all):
Result:
Notice the > in the layout property.
The simplest solution would be to not touch any values that have no transformation applied.
The text was updated successfully, but these errors were encountered: