changelog: add class to represent parsed changelog revisions
Currently, changelog entries are parsed into their respective
components at read time. Many operations are only interested
in a subset of fields of a changelog entry. The parsing and
storing of all the fields adds avoidable overhead.
This patch introduces the "changelogrevision" class. It takes
changelog raw text and exposes the parsed results as attributes.
The code for parsing changelog entries has been moved into its
construction function. changelog.read() has been modified to use
the new class internally while maintaining its existing API.
Future patches will make revision parsing lazy.
We implement the construction function of the new class with
__new__ instead of __init__ so we can use a named tuple to
represent the empty revision. This saves overhead and complexity
of coercing later versions of this class to represent an empty
instance.
While we are here, we add a method on changelog to obtain an
instance of the new type.
The overhead of constructing the new class regresses performance
of revsets accessing this data:
author(mpm)
0.896565
0.929984
desc(bug)
0.887169
0.935642 105%
date(2015)
0.878797
0.908094
extra(rebase_source)
0.865446
0.922624 106%
author(mpm) or author(greg)
1.801832
1.902112 105%
author(mpm) or desc(bug)
1.812438
1.860977
date(2015) or branch(default)
0.968276
1.005824
author(mpm) or desc(bug) or date(2015) or extra(rebase_source)
3.656193
3.743381
Once lazy parsing is implemented, these revsets will all be faster
than before. There is no performance change on revsets that do not
access this data. There /could/ be a performance regression on
operations that perform several changelog reads. However, I can't
think of anything outside of revsets and `hg log` (basically the
same as a revset) that would be impacted.
# parser.py - simple top-down operator precedence parser for mercurial
#
# 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.
# see http://effbot.org/zone/simple-top-down-parsing.htm and
# http://eli.thegreenplace.net/2010/01/02/top-down-operator-precedence-parsing/
# for background
# takes a tokenizer and elements
# tokenizer is an iterator that returns (type, value, pos) tuples
# elements is a mapping of types to binding strength, primary, prefix, infix
# and suffix actions
# an action is a tree node name, a tree label, and an optional match
# __call__(program) parses program into a labeled tree
from __future__ import absolute_import
from .i18n import _
from . import error
class parser(object):
def __init__(self, elements, methods=None):
self._elements = elements
self._methods = methods
self.current = None
def _advance(self):
'advance the tokenizer'
t = self.current
self.current = next(self._iter, None)
return t
def _hasnewterm(self):
'True if next token may start new term'
return any(self._elements[self.current[0]][1:3])
def _match(self, m):
'make sure the tokenizer matches an end condition'
if self.current[0] != m:
raise error.ParseError(_("unexpected token: %s") % self.current[0],
self.current[2])
self._advance()
def _parseoperand(self, bind, m=None):
'gather right-hand-side operand until an end condition or binding met'
if m and self.current[0] == m:
expr = None
else:
expr = self._parse(bind)
if m:
self._match(m)
return expr
def _parse(self, bind=0):
token, value, pos = self._advance()
# handle prefix rules on current token, take as primary if unambiguous
primary, prefix = self._elements[token][1:3]
if primary and not (prefix and self._hasnewterm()):
expr = (primary, value)
elif prefix:
expr = (prefix[0], self._parseoperand(*prefix[1:]))
else:
raise error.ParseError(_("not a prefix: %s") % token, pos)
# gather tokens until we meet a lower binding strength
while bind < self._elements[self.current[0]][0]:
token, value, pos = self._advance()
# handle infix rules, take as suffix if unambiguous
infix, suffix = self._elements[token][3:]
if suffix and not (infix and self._hasnewterm()):
expr = (suffix[0], expr)
elif infix:
expr = (infix[0], expr, self._parseoperand(*infix[1:]))
else:
raise error.ParseError(_("not an infix: %s") % token, pos)
return expr
def parse(self, tokeniter):
'generate a parse tree from tokens'
self._iter = tokeniter
self._advance()
res = self._parse()
token, value, pos = self.current
return res, pos
def eval(self, tree):
'recursively evaluate a parse tree using node methods'
if not isinstance(tree, tuple):
return tree
return self._methods[tree[0]](*[self.eval(t) for t in tree[1:]])
def __call__(self, tokeniter):
'parse tokens into a parse tree and evaluate if methods given'
t = self.parse(tokeniter)
if self._methods:
return self.eval(t)
return t
def buildargsdict(trees, funcname, keys, keyvaluenode, keynode):
"""Build dict from list containing positional and keyword arguments
Invalid keywords or too many positional arguments are rejected, but
missing arguments are just omitted.
"""
if len(trees) > len(keys):
raise error.ParseError(_("%(func)s takes at most %(nargs)d arguments")
% {'func': funcname, 'nargs': len(keys)})
args = {}
# consume positional arguments
for k, x in zip(keys, trees):
if x[0] == keyvaluenode:
break
args[k] = x
# remainder should be keyword arguments
for x in trees[len(args):]:
if x[0] != keyvaluenode or x[1][0] != keynode:
raise error.ParseError(_("%(func)s got an invalid argument")
% {'func': funcname})
k = x[1][1]
if k not in keys:
raise error.ParseError(_("%(func)s got an unexpected keyword "
"argument '%(key)s'")
% {'func': funcname, 'key': k})
if k in args:
raise error.ParseError(_("%(func)s got multiple values for keyword "
"argument '%(key)s'")
% {'func': funcname, 'key': k})
args[k] = x[2]
return args
def unescapestr(s):
try:
return s.decode("string_escape")
except ValueError as e:
# mangle Python's exception into our format
raise error.ParseError(str(e).lower())
def _prettyformat(tree, leafnodes, level, lines):
if not isinstance(tree, tuple) or tree[0] in leafnodes:
lines.append((level, str(tree)))
else:
lines.append((level, '(%s' % tree[0]))
for s in tree[1:]:
_prettyformat(s, leafnodes, level + 1, lines)
lines[-1:] = [(lines[-1][0], lines[-1][1] + ')')]
def prettyformat(tree, leafnodes):
lines = []
_prettyformat(tree, leafnodes, 0, lines)
output = '\n'.join((' ' * l + s) for l, s in lines)
return output
def simplifyinfixops(tree, targetnodes):
"""Flatten chained infix operations to reduce usage of Python stack
>>> def f(tree):
... print prettyformat(simplifyinfixops(tree, ('or',)), ('symbol',))
>>> f(('or',
... ('or',
... ('symbol', '1'),
... ('symbol', '2')),
... ('symbol', '3')))
(or
('symbol', '1')
('symbol', '2')
('symbol', '3'))
>>> f(('func',
... ('symbol', 'p1'),
... ('or',
... ('or',
... ('func',
... ('symbol', 'sort'),
... ('list',
... ('or',
... ('or',
... ('symbol', '1'),
... ('symbol', '2')),
... ('symbol', '3')),
... ('negate',
... ('symbol', 'rev')))),
... ('and',
... ('symbol', '4'),
... ('group',
... ('or',
... ('or',
... ('symbol', '5'),
... ('symbol', '6')),
... ('symbol', '7'))))),
... ('symbol', '8'))))
(func
('symbol', 'p1')
(or
(func
('symbol', 'sort')
(list
(or
('symbol', '1')
('symbol', '2')
('symbol', '3'))
(negate
('symbol', 'rev'))))
(and
('symbol', '4')
(group
(or
('symbol', '5')
('symbol', '6')
('symbol', '7'))))
('symbol', '8')))
"""
if not isinstance(tree, tuple):
return tree
op = tree[0]
if op not in targetnodes:
return (op,) + tuple(simplifyinfixops(x, targetnodes) for x in tree[1:])
# walk down left nodes taking each right node. no recursion to left nodes
# because infix operators are left-associative, i.e. left tree is deep.
# e.g. '1 + 2 + 3' -> (+ (+ 1 2) 3) -> (+ 1 2 3)
simplified = []
x = tree
while x[0] == op:
l, r = x[1:]
simplified.append(simplifyinfixops(r, targetnodes))
x = l
simplified.append(simplifyinfixops(x, targetnodes))
simplified.append(op)
return tuple(reversed(simplified))