Re: better flatten()?
- Posted by petelomax Jul 27, 2015
- 1557 views
and (btw) I'm just using the Phix builtin string() function here
What are the differences between this and from the OE std string() function ? http://openeuphoria.org/docs/std_types.html#_2199_string
Aside, of course, from the obvious (speed!)....
The other obvious one is that you don't need to include anything to invoke string() in Phix, which is mainly why I mention it.
Otherwise nothing major, except that in Phix strings can get auto-expanded to dword-sequences, which leads to the subtle difference:
sequence s s = "fred" -- string(s) yields true on Phix and OE s &= -1 -- string(s) yields false on Phix and OE s = s[1..$-1] -- string(s) yields false on Phix but true on OE
Obviously in Phix, if you want something to stay a string you keep it in a string, so that anomaly is very unlikely to happen.
You could also argue that last true on OE is a bit of a "false positive" in the sense that it looks like a string, but it probably ain't really supposed to be one by then.
Oh, I forgot about declaration semantics:
string("fred") -- true on Phix and OE string({'f','r','e','d'}) -- false on Phix, true on OE.
You can of course puts() etc dword_sequences as if they were strings, without any hassle.
Pete
