Mercurial > jhg

comparison design.txt @ 9:d6d2a630f4a6

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Access to underlaying file data wrapped into own Access object, implemented with FileChannel and ByteBuffer
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Sat, 25 Dec 2010 04:45:59 +0100
parents 5abe5af181bd
children d46773d89a19
comparison
equal deleted inserted replaced
8:a78c980749e3 9:d6d2a630f4a6
21 21
22 ---------- 22 ----------
23 + support patch from baseRev + few deltas (although done in a way patches are applied one by one instead of accumulated) 23 + support patch from baseRev + few deltas (although done in a way patches are applied one by one instead of accumulated)
24 + command-line samples (-R, filenames) (Log & Cat) to show on any repo 24 + command-line samples (-R, filenames) (Log & Cat) to show on any repo
25 +buildfile + run samples 25 +buildfile + run samples
26 *input stream impl + lifecycle. Step forward with FileChannel and ByteBuffer, although questionable accomplishment (looks bit complicated, cumbersome)
26 27
27 input stream impl + lifecycle 28 calculate sha1 digest for file to see I can deal with nodeid
28 delta merge 29 delta merge
29 Changeset to get index (local revision number) 30 Changeset to get index (local revision number)
30 31
32
33
34 >>>> Effective file read/data access
35 ReadOperation, Revlog does: repo.getFileSystem().run(this.file, new ReadOperation(), long start=0, long end = -1)
36 ReadOperation gets buffer (of whatever size, as decided by FS impl), parses it and then reports if needs more data.
37 This helps to ensure streams are closed after reading, allows caching (if the same file (or LRU) is read few times in sequence)
38 and allows buffer management (i.e. reuse. Single buffer for all reads).
39 Scheduling multiple operations (in future, to deal with writes - single queue for FS operations - no locks?)
40
41 File access:
42 * NIO and mapped files - should be fast. Although seems to give less control on mem usage.
43 * Regular InputStreams and chunked stream on top - allocate List<byte[]>, each (but last) chunk of fixed size (depending on initial file size)
44
45 <<<<<