Closed dpanic closed 3 years ago
I actually managed to speedup go-callvis where the performance issue was.
This was problematic part of code, in my fork I removed some of those checks as I don't need them.
// include path prefixes
if len(includePaths) > 0 &&
(inIncludes(caller) || inIncludes(callee)) {
logf("include: %s -> %s", caller, callee)
include = true
}
if !include {
// limit path prefixes
if len(limitPaths) > 0 &&
(!inLimits(caller) || !inLimits(callee)) {
logf("NOT in limit: %s -> %s", caller, callee)
return nil
}
// ignore path prefixes
if len(ignorePaths) > 0 &&
(inIgnores(caller) || inIgnores(callee)) {
logf("IS ignored: %s -> %s", caller, callee)
return nil
}
}
Marking it resolved
Hi,
Is it possible to speedup go-callvis? Do we know what are the botlenecks of project? Is it go-callvis itself or graphiz?