Optimized fuzz.py #136
Optimized fuzz.py #136
Conversation
|
Sorry for not looking at this sooner. Is this a speed optimization, and if so, do you have benchmarks? |
| # sort, join, and strip | ||
| sorted_sect = (" ".join(sorted(intersection))).strip() | ||
| combined_1to2 = (" ".join([sorted_sect, " ".join(sorted(diff1to2))])).strip() | ||
| combined_2to1 = " ".join([sorted_sect, " ".join(sorted(diff2to1))]).strip() |
There was a problem hiding this comment.
This is pretty much the same, but not completely. In this changed code whitespace from sorted_sect is stripped before the strings are combined, while it was done afterwards in the original code. This does not matter in this case, since the original strip for sorted_sect was not required, since " ".join(sorted(intersection)) can never have whitespaces in begin or end. From a performance standpoint this does not really matter, but is probably slightly slower, since it creates some intermediate lists and calls the join function more often.
| if not utils.validate_string(p1): | ||
| return 0 | ||
| if not utils.validate_string(p2): | ||
| if not (utils.validate_string(p1) and utils.validate_string(p2)): |
There was a problem hiding this comment.
Again no performance implication. Simply a preference, whether these two if statements should be joined
|
@josegonzalez It's a line count optimisation. However in my opinion at least the string joining hurts readability. |
Simplified logic using De Morgan's Laws. Combined some lines to reduce string concatenation costs and make coding decisions more uniform. Hope they help.