Instead of a type that has to be tested, store the actual parser against each XPath mapping. This means fewer conditionals and (hopefully) clearer code.
Also in the case of certain decimals, we avoid mutating the DOM itself for clarity (and maybe micro performance improvement)
Instead of a type that has to be tested, store the actual parser against each XPath mapping. This means fewer conditionals and (hopefully) clearer code.
Also in the case of certain decimals, we avoid mutating the DOM itself for clarity (and maybe micro performance improvement)