Why is this on a pull parser? Simple answer is by default it costs little, and by selection the user can optimise whole streams at an early and central place.
Why is this on a pull parser? Simple answer is by default it costs little, and by selection the user can optimise whole streams at an early and central place. Optimising the stream in onQNames or another iteratee or in user code seems to go against the definition of a good design.
(Changed in version 2.8.0) collect has changed. The previous behavior can be reproduced with toSeq.
If the depth is -1 then we haven't hit an element yet, and start should keep pumping until that happens.
If the depth is -1 then we haven't hit an element yet, and start should keep pumping until that happens.
If the depth is -1 after processing the end root (haveRoot), then we keep pumping into the endMisc
Pumps until the first elem, always collecting the prolog
Pumps until the first elem, always collecting the prolog
Basis for xmlpulls, an Iterator[PullType]