threads / patch / 43573

patchRe: [PATCH 1/1] Bypass expensive content comparsion during rename detection

Subject: Re: [PATCH 1/1] Bypass expensive content comparsion during rename detection

## tl;dr

2 messages between Dec 14, 2006 and Dec 14, 2006. Diffs are folded; open one to read it.

replies: 1people: 2as markdown or json

Junio C Hamano· Dec 14, 2006, 11:26 UTC · lore

estimate_similarity() is called after it has determined that one and two do not match exactly, and it relies on populate_filespec with at least size-only has been called on them so that it can reject filepair with vastly different sizes, still without looking at (thus without loading) the contents.

Shawn Pearce· Dec 14, 2006, 12:13 UTC · re: Junio C Hamano · lore
Junio C Hamano <junio@hera.kernel.org> wrote:
Show 5 quoted lines
> estimate_similarity() is called after it has determined that one
> and two do not match exactly, and it relies on populate_filespec
> with at least size-only has been called on them so that it can
> reject filepair with vastly different sizes, still without
> looking at (thus without loading) the contents.

*ouch* That hurts. OK, it makes a lot of sense, but its crazy. About as crazy as the comment I put in my revised patch... ;-)

← back to recent threads