view mercurial/parser.py @ 25561:50a6c3c55db1 stable

parsers: do not cache RevlogError type (issue4451)

Index lookups raise RevlogError when the lookup fails. The previous
implementation was caching a reference to the RevlogError type in a
static variable. This assumed that the "mercurial.error" module was
only loaded once and there was only a single copy of it floating
around in memory. Unfortunately, in some situations - including
certain mod_wsgi configurations - this was not the case: the
"mercurial.error" module could be reloaded. It was possible for a
"RevlogError" reference from the first interpreter to be used by
a second interpreter. While the underlying thing was a
"mercurial.error.RevlogError," the object IDs were different, so
the Python code in revlog.py was failing to catch the exception! This
error has existed since the C index lookup code was implemented in
changeset e8d37b78acfb, which was first released in Mercurial 2.2 in
2012.
http://emptysqua.re/blog/python-c-extensions-and-mod-wsgi/#static-variables-are-shared
contains more details.

This patch removes the caching of the RevlogError type from the
function.

Since pretty much the entire function was refactored and the return
value of the function wasn't used, I changed the function signature
to not return anything.

For reasons unknown to me, we were calling PyErr_SetObject()
with the type of RevlogError and an instance of RevlogError. This
was equivalent to the Python code "raise RevlogError(RevlogError)".
This seemed wonky and completely unnecessary. The Python code only
cares about the type of the exception, not its contents. So I got
rid of this complexity.

This is my first Python C extension patch. Please give extra scrutiny
to it during review.

# parser.py - simple top-down operator precedence parser for mercurial
#
# Copyright 2010 Matt Mackall <[e-mail]>
#
# 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 pairs
# elements is a mapping of types to binding strength, prefix and infix actions
# an action is a tree node name, a tree label, and an optional match
# __call__(program) parses program into a labeled tree

import error
from i18n import _

class parser(object):
    def __init__(self, tokenizer, elements, methods=None):
        self._tokenizer = tokenizer
        self._elements = elements
        self._methods = methods
        self.current = None
    def _advance(self):
        'advance the tokenizer'
        t = self.current
        try:
            self.current = self._iter.next()
        except StopIteration:
            pass
        return t
    def _match(self, m, pos):
        '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 _parse(self, bind=0):
        token, value, pos = self._advance()
        # handle prefix rules on current token
        prefix = self._elements[token][1]
        if not prefix:
            raise error.ParseError(_("not a prefix: %s") % token, pos)
        if len(prefix) == 1:
            expr = (prefix[0], value)
        else:
            if len(prefix) > 2 and prefix[2] == self.current[0]:
                self._match(prefix[2], pos)
                expr = (prefix[0], None)
            else:
                expr = (prefix[0], self._parse(prefix[1]))
                if len(prefix) > 2:
                    self._match(prefix[2], pos)
        # gather tokens until we meet a lower binding strength
        while bind < self._elements[self.current[0]][0]:
            token, value, pos = self._advance()
            e = self._elements[token]
            # check for suffix - next token isn't a valid prefix
            if len(e) == 4 and not self._elements[self.current[0]][1]:
                suffix = e[3]
                expr = (suffix[0], expr)
            else:
                # handle infix rules
                if len(e) < 3 or not e[2]:
                    raise error.ParseError(_("not an infix: %s") % token, pos)
                infix = e[2]
                if len(infix) == 3 and infix[2] == self.current[0]:
                    self._match(infix[2], pos)
                    expr = (infix[0], expr, (None))
                else:
                    expr = (infix[0], expr, self._parse(infix[1]))
                    if len(infix) == 3:
                        self._match(infix[2], pos)
        return expr
    def parse(self, message, lookup=None):
        'generate a parse tree from a message'
        if lookup:
            self._iter = self._tokenizer(message, lookup)
        else:
            self._iter = self._tokenizer(message)
        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, message):
        'parse a message into a parse tree and evaluate if methods given'
        t = self.parse(message)
        if self._methods:
            return self.eval(t)
        return t