OpenEuphoria

Optimization

• Posted by Jason Gade <jaygade at yahoo.com> Feb 09, 2006
• 435 views

I know I've brought this up before, but I want to ask again.

Does Euphoria optimize read-modify-write situation?

I know that internally Euphoria passes by reference and only allocates a new
sequence if elements are modified.

But does the interpreter recognize the following:

sequence foo
foo = bar(foo)

Is this optimized as a pattern where foo itself can be modified instead of
allocating and returning a new sequence?

Recognizing this optimization may be difficult, but I think that it would be
very useful. I suppose where the variable are foo[1] or foo[1][2] or whatever
would make things even more difficult. Maybe just for top-level references?

Of course you could also have:

sequence foo
foo = bar(baz, x, y, z, fun(foo), foo[1..a], d, foo, i, j)

Which would be truly horrid, but still...

If Euphoria does this automatically it should be in the docs.

Unless you'd actually implement a var_id() that is orthogonal with
routine_id()...

--
"Any programming problem can be solved by adding a level of indirection."
--anonymous
"Any performance problem can be solved by removing a level of indirection."
--M. Haertel
j.