manifest: add manifestlog.get to obtain subdirectory instances Previously manifestlog only allowed obtaining root level manifests. Future patches will need direct access to subdirectory manifests as part of changegroup creation, so let's add a get() function that knows how to deal with subdirectories.

manifest: throw LookupError if node not in revlog When accessing a manifest via manifestlog[node], let's verify that the node actually exists and throw a LookupError if it doesn't. This matches the old read behavior, so we don't accidentally return invalid manifestctxs. We do this in manifestlog instead of in the manifestctx/treemanifestctx constructors because the treemanifest code currently relies on the fact that certain code paths can produce treemanifests without touching the revlogs (and it has tests that verify things work if certain revlogs are missing entirely, so they break if we add validation that tries to read them).

revlog: optimize _chunkraw when startrev==endrev In many cases, _chunkraw() is called with startrev==endrev. When this is true, we can avoid an extra index lookup and some other minor operations. On the mozilla-unified repo, `hg perfrevlogchunks -c` says this has the following impact: ! read w/ reused fd ! wall 0.371846 comb 0.370000 user 0.350000 sys 0.020000 (best of 27) ! wall 0.337930 comb 0.330000 user 0.300000 sys 0.030000 (best of 30) ! read batch w/ reused fd ! wall 0.014952 comb 0.020000 user 0.000000 sys 0.020000 (best of 197) ! wall 0.014866 comb 0.010000 user 0.000000 sys 0.010000 (best of 196) So, we've gone from ~25x slower than batch to ~22.5x slower. At this point, there's probably not much else we can do except implement an optimized function in the index itself, including in C.

revlog: inline start() and end() for perf reasons When I implemented `hg perfrevlogchunks`, one of the things that stood out was N * _chunk() calls was ~38x slower than 1 _chunks() call. Specifically, on the mozilla-unified repo: N*_chunk: 0.528997s 1*_chunks: 0.013735s This repo has 352,097 changesets. So the average time per changeset comes out to: N*_chunk: 1.502us 1*_chunks: 0.039us If you extrapolate these numbers to a repository with 1M changesets, that comes out to 1.502s versus 0.039s, which is significant. At these latencies, Python attribute lookups and function calls matter. So, this patch inlines some code to cut down on that overhead. The impact of this patch on N*_chunk() calls is clear: ! wall 0.528997 comb 0.520000 user 0.500000 sys 0.020000 (best of 19) ! wall 0.367723 comb 0.370000 user 0.350000 sys 0.020000 (best of 27) So, we go from ~38x slower to ~27x. A nice improvement. But there's still a long way to go. It's worth noting that functionality like revsets perform changelog lookups one revision at a time. So this code path is worth optimizing.