Mercurial > hg
view mercurial/filesetlang.py @ 42044:bb271ec2fbfb
compression: introduce a `storage.revlog.zstd.level` configuration
This option control the zstd compression level used when compressing revlog
chunk. The usage of zstd for revlog compression has not graduated from
experimental yet, but we intend to fix that soon.
The option name for the compression level is more straight forward to pick, so
this changesets comes first. Having a dedicated option for each compression
engine is useful because they don't support the same range of values.
I ran the same measurement as for the zlib compression level (in the parent
changesets). The variation in repository size is stay mostly in the same (small)
range. The "read/write" performance see smallish variation, but are overall much
better than zlib. Write performance show the same tend of having better write
performance for when reaching high-end compression.
Again, we don't intend to change the default zstd compression level (currently:
3) in this series. However this is worth investigating in the future.
The Performance comparison of zlib vs zstd is quite impressive. The repository
size stay in the same range, but the performance are much better in all
situations.
Comparison summary
==================
We are looking at:
- performance range for zlib
- performance range for zstd
- comparison of default zstd (level-3) to default zlib (level 6)
- comparison of the slowest zstd time to the fastest zlib time
Read performance:
-----------------
| zlib | zstd | cmp | f2s
mercurial | 0.170159 - 0.189219 | 0.144127 - 0.149624 | 80% | 88%
pypy | 2.679217 - 2.768691 | 1.532317 - 1.705044 | 60% | 63%
netbeans | 122.477027 - 141.620281 | 72.996346 - 89.731560 | 58% | 73%
mozilla | 147.867662 - 170.572118 | 91.700995 - 105.853099 | 56% | 71%
Write performance:
------------------
| zlib | zstd | cmp | f2s
mercurial | 53.250304 - 56.2936129 | 40.877025 - 45.677286 | 75% | 86%
pypy | 460.721984 - 476.589918 | 270.545409 - 301.002219 | 63% | 65%
netbeans | 520.560316 - 715.930400 | 370.356311 - 428.329652 | 55% | 82%
mozilla | 739.803002 - 987.056093 | 505.152906 - 591.930683 | 57% | 80%
Raw data
--------
repo alg lvl .hg/store size 00manifest.d read write
mercurial zlib 1 49,402,813 5,963,475 0.170159 53.250304
mercurial zlib 6 47,197,397 5,875,730 0.182820 56.264320
mercurial zlib 9 47,121,596 5,849,781 0.189219 56.293612
mercurial zstd 1 49,737,084 5,966,355 0.144127 40.877025
mercurial zstd 3 48,961,867 5,895,208 0.146376 42.268142
mercurial zstd 5 48,200,592 5,938,676 0.149624 43.162875
mercurial zstd 10 47,833,520 5,913,353 0.145185 44.012489
mercurial zstd 15 47,314,604 5,728,679 0.147686 45.677286
mercurial zstd 20 47,330,502 5,830,539 0.145789 45.025407
mercurial zstd 22 47,330,076 5,830,539 0.143996 44.690460
pypy zlib 1 370,830,572 28,462,425 2.679217 460.721984
pypy zlib 6 340,112,317 27,648,747 2.768691 467.537158
pypy zlib 9 338,360,736 27,639,003 2.763495 476.589918
pypy zstd 1 362,377,479 27,916,214 1.532317 270.545409
pypy zstd 3 354,137,693 27,905,988 1.686718 294.951509
pypy zstd 5 342,640,043 27,655,774 1.705044 301.002219
pypy zstd 10 334,224,327 27,164,493 1.567287 285.186239
pypy zstd 15 329,000,363 26,645,965 1.637729 299.561332
pypy zstd 20 324,534,039 26,199,547 1.526813 302.149827
pypy zstd 22 324,530,595 26,198,932 1.525718 307.821218
netbeans zlib 1 1,281,847,810 165,495,457 122.477027 520.560316
netbeans zlib 6 1,205,284,353 159,161,207 139.876147 715.930400
netbeans zlib 9 1,197,135,671 155,034,586 141.620281 678.297064
netbeans zstd 1 1,259,581,737 160,840,613 72.996346 370.356311
netbeans zstd 3 1,232,978,122 157,691,551 81.622317 396.733087
netbeans zstd 5 1,208,034,075 160,246,880 83.080549 364.342626
netbeans zstd 10 1,188,624,176 156,083,417 79.323935 403.594602
netbeans zstd 15 1,176,973,589 153,859,477 89.731560 428.329652
netbeans zstd 20 1,162,958,258 151,147,535 82.842667 392.335349
netbeans zstd 22 1,162,707,029 151,150,220 82.565695 402.840655
mozilla zlib 1 2,775,497,186 298,527,987 147.867662 751.263721
mozilla zlib 6 2,596,856,420 286,597,671 170.572118 987.056093
mozilla zlib 9 2,587,542,494 287,018,264 163.622338 739.803002
mozilla zstd 1 2,723,159,348 286,617,532 91.700995 570.042751
mozilla zstd 3 2,665,055,001 286,152,013 95.240155 561.412805
mozilla zstd 5 2,607,819,817 288,060,030 101.978048 505.152906
mozilla zstd 10 2,558,761,085 283,967,648 104.113481 497.771202
mozilla zstd 15 2,526,216,060 275,581,300 105.853099 591.930683
mozilla zstd 20 2,485,114,806 266,478,859 95.268795 576.515389
mozilla zstd 22 2,484,869,080 266,456,505 94.429282 572.785537
author | Pierre-Yves David <pierre-yves.david@octobus.net> |
---|---|
date | Wed, 27 Mar 2019 18:35:59 +0100 |
parents | e79a69af1593 |
children | 2372284d9457 |
line wrap: on
line source
# filesetlang.py - parser, tokenizer and utility for file set language # # Copyright 2010 Matt Mackall <mpm@selenic.com> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. from __future__ import absolute_import from .i18n import _ from . import ( error, parser, pycompat, ) # common weight constants for static optimization # (see registrar.filesetpredicate for details) WEIGHT_CHECK_FILENAME = 0.5 WEIGHT_READ_CONTENTS = 30 WEIGHT_STATUS = 10 WEIGHT_STATUS_THOROUGH = 50 elements = { # token-type: binding-strength, primary, prefix, infix, suffix "(": (20, None, ("group", 1, ")"), ("func", 1, ")"), None), ":": (15, None, None, ("kindpat", 15), None), "-": (5, None, ("negate", 19), ("minus", 5), None), "not": (10, None, ("not", 10), None, None), "!": (10, None, ("not", 10), None, None), "and": (5, None, None, ("and", 5), None), "&": (5, None, None, ("and", 5), None), "or": (4, None, None, ("or", 4), None), "|": (4, None, None, ("or", 4), None), "+": (4, None, None, ("or", 4), None), ",": (2, None, None, ("list", 2), None), ")": (0, None, None, None, None), "symbol": (0, "symbol", None, None, None), "string": (0, "string", None, None, None), "end": (0, None, None, None, None), } keywords = {'and', 'or', 'not'} symbols = {} globchars = ".*{}[]?/\\_" def tokenize(program): pos, l = 0, len(program) program = pycompat.bytestr(program) while pos < l: c = program[pos] if c.isspace(): # skip inter-token whitespace pass elif c in "(),-:|&+!": # handle simple operators yield (c, None, pos) elif (c in '"\'' or c == 'r' and program[pos:pos + 2] in ("r'", 'r"')): # handle quoted strings if c == 'r': pos += 1 c = program[pos] decode = lambda x: x else: decode = parser.unescapestr pos += 1 s = pos while pos < l: # find closing quote d = program[pos] if d == '\\': # skip over escaped characters pos += 2 continue if d == c: yield ('string', decode(program[s:pos]), s) break pos += 1 else: raise error.ParseError(_("unterminated string"), s) elif c.isalnum() or c in globchars or ord(c) > 127: # gather up a symbol/keyword s = pos pos += 1 while pos < l: # find end of symbol d = program[pos] if not (d.isalnum() or d in globchars or ord(d) > 127): break pos += 1 sym = program[s:pos] if sym in keywords: # operator keywords yield (sym, None, s) else: yield ('symbol', sym, s) pos -= 1 else: raise error.ParseError(_("syntax error"), pos) pos += 1 yield ('end', None, pos) def parse(expr): p = parser.parser(elements) tree, pos = p.parse(tokenize(expr)) if pos != len(expr): raise error.ParseError(_("invalid token"), pos) return parser.simplifyinfixops(tree, {'list', 'or'}) def getsymbol(x): if x and x[0] == 'symbol': return x[1] raise error.ParseError(_('not a symbol')) def getstring(x, err): if x and (x[0] == 'string' or x[0] == 'symbol'): return x[1] raise error.ParseError(err) def getkindpat(x, y, allkinds, err): kind = getsymbol(x) pat = getstring(y, err) if kind not in allkinds: raise error.ParseError(_("invalid pattern kind: %s") % kind) return '%s:%s' % (kind, pat) def getpattern(x, allkinds, err): if x and x[0] == 'kindpat': return getkindpat(x[1], x[2], allkinds, err) return getstring(x, err) def getlist(x): if not x: return [] if x[0] == 'list': return list(x[1:]) return [x] def getargs(x, min, max, err): l = getlist(x) if len(l) < min or len(l) > max: raise error.ParseError(err) return l def _analyze(x): if x is None: return x op = x[0] if op in {'string', 'symbol'}: return x if op == 'kindpat': getsymbol(x[1]) # kind must be a symbol t = _analyze(x[2]) return (op, x[1], t) if op == 'group': return _analyze(x[1]) if op == 'negate': raise error.ParseError(_("can't use negate operator in this context")) if op == 'not': t = _analyze(x[1]) return (op, t) if op == 'and': ta = _analyze(x[1]) tb = _analyze(x[2]) return (op, ta, tb) if op == 'minus': return _analyze(('and', x[1], ('not', x[2]))) if op in {'list', 'or'}: ts = tuple(_analyze(y) for y in x[1:]) return (op,) + ts if op == 'func': getsymbol(x[1]) # function name must be a symbol ta = _analyze(x[2]) return (op, x[1], ta) raise error.ProgrammingError('invalid operator %r' % op) def _insertstatushints(x): """Insert hint nodes where status should be calculated (first path) This works in bottom-up way, summing up status names and inserting hint nodes at 'and' and 'or' as needed. Thus redundant hint nodes may be left. Returns (status-names, new-tree) at the given subtree, where status-names is a sum of status names referenced in the given subtree. """ if x is None: return (), x op = x[0] if op in {'string', 'symbol', 'kindpat'}: return (), x if op == 'not': h, t = _insertstatushints(x[1]) return h, (op, t) if op == 'and': ha, ta = _insertstatushints(x[1]) hb, tb = _insertstatushints(x[2]) hr = ha + hb if ha and hb: return hr, ('withstatus', (op, ta, tb), ('string', ' '.join(hr))) return hr, (op, ta, tb) if op == 'or': hs, ts = zip(*(_insertstatushints(y) for y in x[1:])) hr = sum(hs, ()) if sum(bool(h) for h in hs) > 1: return hr, ('withstatus', (op,) + ts, ('string', ' '.join(hr))) return hr, (op,) + ts if op == 'list': hs, ts = zip(*(_insertstatushints(y) for y in x[1:])) return sum(hs, ()), (op,) + ts if op == 'func': f = getsymbol(x[1]) # don't propagate 'ha' crossing a function boundary ha, ta = _insertstatushints(x[2]) if getattr(symbols.get(f), '_callstatus', False): return (f,), ('withstatus', (op, x[1], ta), ('string', f)) return (), (op, x[1], ta) raise error.ProgrammingError('invalid operator %r' % op) def _mergestatushints(x, instatus): """Remove redundant status hint nodes (second path) This is the top-down path to eliminate inner hint nodes. """ if x is None: return x op = x[0] if op == 'withstatus': if instatus: # drop redundant hint node return _mergestatushints(x[1], instatus) t = _mergestatushints(x[1], instatus=True) return (op, t, x[2]) if op in {'string', 'symbol', 'kindpat'}: return x if op == 'not': t = _mergestatushints(x[1], instatus) return (op, t) if op == 'and': ta = _mergestatushints(x[1], instatus) tb = _mergestatushints(x[2], instatus) return (op, ta, tb) if op in {'list', 'or'}: ts = tuple(_mergestatushints(y, instatus) for y in x[1:]) return (op,) + ts if op == 'func': # don't propagate 'instatus' crossing a function boundary ta = _mergestatushints(x[2], instatus=False) return (op, x[1], ta) raise error.ProgrammingError('invalid operator %r' % op) def analyze(x): """Transform raw parsed tree to evaluatable tree which can be fed to optimize() or getmatch() All pseudo operations should be mapped to real operations or functions defined in methods or symbols table respectively. """ t = _analyze(x) _h, t = _insertstatushints(t) return _mergestatushints(t, instatus=False) def _optimizeandops(op, ta, tb): if tb is not None and tb[0] == 'not': return ('minus', ta, tb[1]) return (op, ta, tb) def _optimizeunion(xs): # collect string patterns so they can be compiled into a single regexp ws, ts, ss = [], [], [] for x in xs: w, t = _optimize(x) if t is not None and t[0] in {'string', 'symbol', 'kindpat'}: ss.append(t) continue ws.append(w) ts.append(t) if ss: ws.append(WEIGHT_CHECK_FILENAME) ts.append(('patterns',) + tuple(ss)) return ws, ts def _optimize(x): if x is None: return 0, x op = x[0] if op == 'withstatus': w, t = _optimize(x[1]) return w, (op, t, x[2]) if op in {'string', 'symbol'}: return WEIGHT_CHECK_FILENAME, x if op == 'kindpat': w, t = _optimize(x[2]) return w, (op, x[1], t) if op == 'not': w, t = _optimize(x[1]) return w, (op, t) if op == 'and': wa, ta = _optimize(x[1]) wb, tb = _optimize(x[2]) if wa <= wb: return wa, _optimizeandops(op, ta, tb) else: return wb, _optimizeandops(op, tb, ta) if op == 'or': ws, ts = _optimizeunion(x[1:]) if len(ts) == 1: return ws[0], ts[0] # 'or' operation is fully optimized out ts = tuple(it[1] for it in sorted(enumerate(ts), key=lambda it: ws[it[0]])) return max(ws), (op,) + ts if op == 'list': ws, ts = zip(*(_optimize(y) for y in x[1:])) return sum(ws), (op,) + ts if op == 'func': f = getsymbol(x[1]) w = getattr(symbols.get(f), '_weight', 1) wa, ta = _optimize(x[2]) return w + wa, (op, x[1], ta) raise error.ProgrammingError('invalid operator %r' % op) def optimize(x): """Reorder/rewrite evaluatable tree for optimization All pseudo operations should be transformed beforehand. """ _w, t = _optimize(x) return t def prettyformat(tree): return parser.prettyformat(tree, ('string', 'symbol'))