class library: improving Pstep input handling and documentation #2511
Conversation
allow durs to also take raw lists as input
|
PstepDur is an extension from my ddwPatterns quark. Documentation of core classes should not refer to extensions as if every user will have them automatically. Otherwise, at a quick glance, it looks ok. |
|
Ah, didn't realize that. I thought it was odd that a subclass would be better documented than its super. Went back and removed those references. |
| @@ -4,7 +4,10 @@ Pstep : Pattern { | |||
| *new { arg levels, durs = 1, repeats = 1; | |||
| ^super.newCopyArgs(levels, durs, repeats).init | |||
| } | |||
| init { if (list.isKindOf(Collection)) { list = Pseq(list); } } | |||
| init { | |||
| if (list.isKindOf(Collection)) { list = Pseq(list); }; | |||
There was a problem hiding this comment.
instead of list.isKindOf(Collection) it should be list.isSequenceableCollection. A Set, e.g. won't work as an argument to Pseq.
|
changes made, thanks for catching that—I was naively duplicating what was already there. Out of curiosity I searched the repo to see if this was a recurring pattern; luckily it seems to be unique. |
|
it could be that, in the future, we might want to implement multichannel expansion for patterns, then this whole conversion might be in the way. But it is there already, so let's go with it. Thank you for making it more uniform, this is exactly the kind of changes we need! |
telephon
changed the title
|
Glad to help! |
This is a fix for issue #2489. I pulled most of the description from PstepDur. While testing out the class I also noticed that it handles
levelsbeing a list but notdurs, and I couldn't see a reason why not, so I added a single line toPstep.initto take care of that.