You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As the ThingTalk compiler grows in complexity, we should benchmark & profile the time spent doing:
parsing
typechecking
compilation to JS code
compilation of JS code
This way, we can decide whether it is meaningful to have a compiled code cache for frequently used commands, and at what level (JS code or JS function), and whether JSIR-level optimizations (eg dead code elimination, common subexpression elimination) are meaningful, when the JS backend compiler cannot do them easily.
The text was updated successfully, but these errors were encountered:
As the ThingTalk compiler grows in complexity, we should benchmark & profile the time spent doing:
This way, we can decide whether it is meaningful to have a compiled code cache for frequently used commands, and at what level (JS code or JS function), and whether JSIR-level optimizations (eg dead code elimination, common subexpression elimination) are meaningful, when the JS backend compiler cannot do them easily.
The text was updated successfully, but these errors were encountered: