optimization: caching resultspace environments

[jbohren]

This can save something like 0.2s when starting up if the same resultspace gets loaded twice.

[wjwwood]

Is there a reason and/or a way to get the uncached version? This is the pattern I usually follow:

_foo_cache = {}

def get_foo_cached(arg):
    global _foo_cache
    if arg not in _foo_cache:
        _foo_cache[arg] = get_foo(arg)
    return _foo_cache[arg]


def get_foo(arg):
    # ... Some implementation
    return foo

# ... Somewhere else:

foo = get_foo_cached(arg)

This way it is obvious from the calling code that you are getting a cached value, and it is possible to get an uncached result if desired. What's your opinion?

[jbohren]

I don't really care either way. The only time you'd want to reload it is if
it changes while a single instance of catkin is running.

On Mon, Apr 6, 2015, 18:50 William Woodall notifications@github.com wrote:

Is there a reason and/or a way to get the uncached version? This is the
pattern I usually follow:

_foo_cache = {}
def get_foo_cached(arg):
global _foo_cache
if arg not in _foo_cache:
_foo_cache[arg] = get_foo(arg)
return _foo_cache[arg]

def get_foo(arg):
... Some implementation
return foo

... Somewhere else:

foo = get_foo_cached(arg)

This way it is obvious from the calling code that you are getting a cached
value, and it is possible to get an uncached result if desired. What's your
opinion?

Reply to this email directly or view it on GitHub
#185 (comment).

[jbohren]

@wjwwood Take a look at 866e111, I just added an option to enable/disable caching.

[wjwwood]

lgtm, thanks.