##// END OF EJS Templates
wireprotov2: add phases to "changesetdata" command...
wireprotov2: add phases to "changesetdata" command This commit teaches the "changesetdata" wire protocol command to emit the phase state for each changeset. This is a different approach from existing phase transfer in a few ways. Previously, if there are no new revisions (or we're not using bundle2), we perform a "listkeys" request to retrieve phase heads. And when revision data is being transferred with bundle2, phases data is encoded in a standalone bundle2 part. In both cases, phases data is logically decoupled from the changeset data and is encountered/applied after changeset revision data is received. The new wire protocol purposefully tries to more tightly associate changeset metadata (phases, bookmarks, obsolescence markers, etc) with the changeset revision and index data itself, rather than have it live as a separate entity that must be fetched and processed separately. I reckon that one reason we didn't do this before was it was difficult to add new data types/fields without breaking existing consumers. By using CBOR maps to transfer changeset data and putting clients in control of what fields are requested / present in those maps, we can easily add additional changeset data while maintaining backwards compatibility. I believe this to be a superior approach to the problem. That being said, for performance reasons, we may need to resort to alternative mechanisms for transferring data like phases. But for now, I think giving the wire protocol the ability to transfer changeset metadata next to the changeset itself is a powerful feature because it is a raw, changeset-centric data API. And if you build simple APIs for accessing the fundamental units of repository data, you enable client-side experimentation (partial clone, etc). If it turns out that we need specialized APIs or mechanisms for transferring data like phases, we can build in those APIs later. For now, I'd like to see how far we can get on simple APIs. It's worth noting that when phase data is being requested, the server will also emit changeset records for nodes in the bases specified by the "noderange" argument. This is to ensure that phase-only updates for nodes the client has are available to the client, even if no new changesets will be transferred. Differential Revision: https://phab.mercurial-scm.org/D4483

File last commit:

r38915:e79a69af default
r39668:c1aacb0d default
Show More
filesetlang.py
330 lines | 10.2 KiB | text/x-python | PythonLexer
Yuya Nishihara
fileset: extract language processing part to new module (API)...
r38841 # 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,
)
Yuya Nishihara
fileset: introduce weight constants for readability...
r38899 # 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
Yuya Nishihara
fileset: extract language processing part to new module (API)...
r38841 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
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862 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)
Yuya Nishihara
fileset: drop 'group' node from tree to be evaluated...
r38863 if op == 'group':
return _analyze(x[1])
Yuya Nishihara
fileset: reject 'negate' node early while transforming parsed tree...
r38864 if op == 'negate':
raise error.ParseError(_("can't use negate operator in this context"))
if op == 'not':
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862 t = _analyze(x[1])
return (op, t)
Yuya Nishihara
fileset: optimize 'x and not y' to 'x - y'...
r38868 if op == 'and':
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862 ta = _analyze(x[1])
tb = _analyze(x[2])
return (op, ta, tb)
Yuya Nishihara
fileset: optimize 'x and not y' to 'x - y'...
r38868 if op == 'minus':
return _analyze(('and', x[1], ('not', x[2])))
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862 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)
Yuya Nishihara
fileset: insert hints where status should be computed...
r38915 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)
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862 def analyze(x):
"""Transform raw parsed tree to evaluatable tree which can be fed to
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 optimize() or getmatch()
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862
All pseudo operations should be mapped to real operations or functions
defined in methods or symbols table respectively.
"""
Yuya Nishihara
fileset: insert hints where status should be computed...
r38915 t = _analyze(x)
_h, t = _insertstatushints(t)
return _mergestatushints(t, instatus=False)
Yuya Nishihara
fileset: add phase to transform parsed tree...
r38862
Yuya Nishihara
fileset: optimize 'x and not y' to 'x - y'...
r38868 def _optimizeandops(op, ta, tb):
if tb is not None and tb[0] == 'not':
return ('minus', ta, tb[1])
return (op, ta, tb)
Yuya Nishihara
fileset: combine union of basic patterns into single matcher...
r38901 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
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 def _optimize(x):
if x is None:
return 0, x
op = x[0]
Yuya Nishihara
fileset: insert hints where status should be computed...
r38915 if op == 'withstatus':
w, t = _optimize(x[1])
return w, (op, t, x[2])
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 if op in {'string', 'symbol'}:
Yuya Nishihara
fileset: introduce weight constants for readability...
r38899 return WEIGHT_CHECK_FILENAME, x
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 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)
Yuya Nishihara
fileset: reorder 'and' expression to evaluate basic patterns first...
r38867 if op == 'and':
wa, ta = _optimize(x[1])
wb, tb = _optimize(x[2])
if wa <= wb:
Yuya Nishihara
fileset: optimize 'x and not y' to 'x - y'...
r38868 return wa, _optimizeandops(op, ta, tb)
Yuya Nishihara
fileset: reorder 'and' expression to evaluate basic patterns first...
r38867 else:
Yuya Nishihara
fileset: optimize 'x and not y' to 'x - y'...
r38868 return wb, _optimizeandops(op, tb, ta)
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 if op == 'or':
Yuya Nishihara
fileset: combine union of basic patterns into single matcher...
r38901 ws, ts = _optimizeunion(x[1:])
if len(ts) == 1:
return ws[0], ts[0] # 'or' operation is fully optimized out
Yuya Nishihara
fileset: reorder 'or' expression by weight
r38900 ts = tuple(it[1] for it in sorted(enumerate(ts),
key=lambda it: ws[it[0]]))
Yuya Nishihara
fileset: add stub for weight-based optimization...
r38865 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
Yuya Nishihara
fileset: extract language processing part to new module (API)...
r38841 def prettyformat(tree):
return parser.prettyformat(tree, ('string', 'symbol'))