Performance tweaks for to_camel_case #131
Conversation
| WORD_RE = re.compile(r'^([a-z]+\d*)$', re.IGNORECASE) | ||
|
|
||
| SNAKE_CASE_RE = re.compile(r'^([a-zA-Z]+\d*_[a-zA-Z\d_]*|_+[a-zA-Z\d]+[a-zA-Z\d_]*)$') | ||
| WORD_RE = re.compile(r'^([a-zA-Z]+\d*)$') |
There was a problem hiding this comment.
I believe that both of these could be simplified/combined into a single shorter regular expression, (_[a-zA-Z\d]|[a-zA-Z])\w*, but that would be a larger change and perhaps there's a reason for wanting to separate them. I'm also unclear about whether digits are allowed in the middle of letters, the WORD_RE seems to indicate they're not.
There was a problem hiding this comment.
Thanks, @pamelafox.
@YunchuWang - We can take this experience to simplify out regexes.
Codecov Report
@@ Coverage Diff @@
## dev #131 +/- ##
==========================================
- Coverage 90.09% 86.04% -4.05%
==========================================
Files 51 50 -1
Lines 2959 2903 -56
Branches 396 391 -5
==========================================
- Hits 2666 2498 -168
- Misses 218 329 +111
- Partials 75 76 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
pamelafox
requested review from
vrdmr,
gavin-aguiar and
YunchuWang
as code owners
| WORD_RE = re.compile(r'^([a-z]+\d*)$', re.IGNORECASE) | ||
|
|
||
| SNAKE_CASE_RE = re.compile(r'^([a-zA-Z]+\d*_[a-zA-Z\d_]*|_+[a-zA-Z\d]+[a-zA-Z\d_]*)$') | ||
| WORD_RE = re.compile(r'^([a-zA-Z]+\d*)$') |
There was a problem hiding this comment.
Thanks, @pamelafox.
@YunchuWang - We can take this experience to simplify out regexes.
This PR includes two performance tweaks for the
to_camel_casefunction:It replaces the generator expression passed to
str.join()with a list comprehension. The list comprehension is actually 10-15% faster asstr.join()turns its arguments into a list anyway (if they're not already). More details in this benchmark test: Add benchmark comparing join of gen expression vs join of list comp tonybaloney/anti-patterns#4It removes the re.IGNORECASE flag and instead specifies a-zA-Z instead of a-z. That seems to be 20-25% faster according to benchmark in Add a benchmark for regex related practices tonybaloney/anti-patterns#5