[New-bugs-announce] [issue31324] support._match_test() used by test.bisect is very inefficient

STINNER Victor report at bugs.python.org
Fri Sep 1 11:11:02 EDT 2017

New submission from STINNER Victor:

support._match_test() uses a nested loop calling fnmatch.fnmatchcase(). This function creates a temporary regular expression object. The cache of the re module works around the performance... if the length of support.match_tests fits into the cache. But when the list is longer, the function becomes dead slow...

def _match_test(test):
    global match_tests
    if match_tests is None:
        return True 
    test_id = test.id()                                                                            
    for match_test in match_tests:
        if fnmatch.fnmatchcase(test_id, match_test):                                               
            return True
        for name in test_id.split("."):                                                            
            if fnmatch.fnmatchcase(name, match_test):                                              
                return True
    return False

Maybe we should build a giant regex matching test_id at each, but cache the regex since support.match_tests can be modified anytime. I implemented this once, but I lost the code :-)

Currently, it's possible to match 3 things:

* test method name: test_exit
* test class name: SysModuleTest
* full test id: test.test_sys.SysModuleTest.test_exit

It's also possible to use "*" joker character in a test name. I would like to keep these convenient CLI.

messages: 301122
nosy: haypo, serhiy.storchaka
priority: normal
severity: normal
status: open
title: support._match_test() used by test.bisect is very inefficient
versions: Python 3.7

Python tracker <report at bugs.python.org>

More information about the New-bugs-announce mailing list