From: Oren Ben-K. <or...@ri...> - 2001-06-14 07:45:33
|
Jason Diamond [mailto:ja...@in...] wrote: > I've been playing with my parser and have noticed that it's > possible to fold scalars in a way that makes it unintuitive > for human readers but still pares able (at least according > to the way I've implemented my parser). Yes, that's the current state of affairs. It reflects a compromise between making cut&paste easier and making the text more readable. This compromise is still being worked out. > [ Examples ] > Are these correct? Yes, they are (at least that was my intention in the new draft). > If they are, could examples like these be > included in the > draft? (I think more examples are needed overall, by the way.) Good idea. > Or, should it be required that all the successive lines of a > folded scalar > be at the exact same indentation so that we never have to worry about > encountering this kind of YAML? I'd go with that - I think cut&paste is overrated. But the nature of a compromise is that nobody is completely happy... Also note that the second line of a multi-line simple scalar in a list *must* be more indented then the first line: foo: @ baz quux bar I proposed we allow a leading ':' for the value in such a case, to make it: foo: @ : baz quux bar : foo xuuq : etc. etc. Have fun, Oren Ben-Kiki |