In#104512 we madepathlib.Path.glob() use a "walk-and-filter" strategy for expanding** wildcards in patterns: when we encounter a** segment, we immediately consume subsequent segments and use them to build a regex that is used to filter results. This saves a bunch ofscandir() calls.

However! We actually build a regex for theentire pattern given toglob(), rather than just the segments following** wildcards. And so when evaluating a pattern likedir*/**/file*, thedir* part is needlessly matched twice against each path.@zooba noted this in areview comment at the time.

We should be able to improve performance by building anre.Pattern only for segments following** wildcards, and not the entireglob() pattern.

Linked PRs

• GH-115060: Speed uppathlib.Path.glob() by removing redundant regex matching #115061
• GH-115060: Speed uppathlib.Path.glob() by skipping directory scanning #116152
• GH-115060: Speed uppathlib.Path.glob() by not scanning literal parts #117732
• GH-115060: Speed uppathlib.Path.glob() by omitting initialstat() #117831