You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
britney2-ubuntu/installability/builder.py

399 lines
15 KiB

# -*- coding: utf-8 -*-
# Copyright (C) 2012 Niels Thykier <niels@thykier.net>
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
from collections import defaultdict
from contextlib import contextmanager
from britney_util import ifilter_except, iter_except
from installability.solver import InstallabilitySolver
class _RelationBuilder(object):
"""Private helper class to "build" relations"""
def __init__(self, itbuilder, binary):
self._itbuilder = itbuilder
self._binary = binary
binary_data = itbuilder._package_table[binary]
self._new_deps = set(binary_data[0])
self._new_breaks = set(binary_data[1])
def add_dependency_clause(self, or_clause):
"""Add a dependency clause
The clause must be a sequence of (name, version, architecture)
tuples. The clause is an OR clause, i.e. any tuple in the
sequence can satisfy the relation. It is irrelevant if the
dependency is from the "Depends" or the "Pre-Depends" field.
Note that is the sequence is empty, the dependency is assumed
to be unsatisfiable.
The binaries in the clause are not required to have been added
to the InstallabilityTesterBuilder when this method is called.
However, they must be added before the "build()" method is
called.
"""
itbuilder = self._itbuilder
clause = itbuilder._intern_set(or_clause)
binary = self._binary
okay = False
for dep_tuple in clause:
okay = True
rdeps, _, rdep_relations = itbuilder._reverse_relations(dep_tuple)
rdeps.add(binary)
rdep_relations.add(clause)
self._new_deps.add(clause)
if not okay:
itbuilder._broken.add(binary)
def add_breaks(self, broken_binary):
"""Add a Breaks-clause
Marks the given binary as being broken by the current
package. That is, the given package satisfies a relation
in either the "Breaks" or the "Conflicts" field. The binary
given must be a (name, version, architecture)-tuple.
The binary is not required to have been added to the
InstallabilityTesterBuilder when this method is called. However,
it must be added before the "build()" method is called.
"""
itbuilder = self._itbuilder
self._new_breaks.add(broken_binary)
reverse_relations = itbuilder._reverse_relations(broken_binary)
reverse_relations[1].add(self._binary)
def _commit(self):
itbuilder = self._itbuilder
data = (itbuilder._intern_set(self._new_deps),
itbuilder._intern_set(self._new_breaks))
itbuilder._package_table[self._binary] = data
class InstallabilityTesterBuilder(object):
"""Builder to create instances of InstallabilityTester"""
def __init__(self):
self._package_table = {}
self._reverse_package_table = {}
self._essentials = set()
self._testing = set()
self._internmap = {}
self._broken = set()
def add_binary(self, binary, essential=False, in_testing=False,
frozenset=frozenset):
"""Add a new binary package
Adds a new binary package. The binary must be given as a
(name, version, architecture)-tuple. Returns True if this
binary is new (i.e. has never been added before) or False
otherwise.
Keyword arguments:
* essential - Whether this package is "Essential: yes".
* in_testing - Whether this package is in testing.
The frozenset argument is a private optimisation.
Cave-at: arch:all packages should be "re-mapped" to given
architecture. That is, (pkg, version, "all") should be
added as:
for arch in architectures:
binary = (pkg, version, arch)
it.add_binary(binary)
The resulting InstallabilityTester relies on this for
correctness!
"""
# Note, even with a dup, we need to do these
if in_testing:
self._testing.add(binary)
if essential:
self._essentials.add(binary)
if binary not in self._package_table:
# Allow binaries to be added multiple times (happens
# when sid and testing have the same version)
self._package_table[binary] = (frozenset(), frozenset())
return True
return False
@contextmanager
def relation_builder(self, binary):
"""Returns a _RelationBuilder for a given binary [context]
This method returns a context-managed _RelationBuilder for a
given binary. So it should be used in a "with"-statment,
like:
with it.relation_builder(binary) as rel:
rel.add_dependency_clause(dependency_clause)
rel.add_breaks(pkgtuple)
...
The binary given must be a (name, version, architecture)-tuple.
Note, this method is optimised to be called at most once per
binary.
"""
if binary not in self._package_table:
raise ValueError("Binary %s/%s/%s does not exist" % binary)
rel = _RelationBuilder(self, binary)
yield rel
rel._commit()
def _intern_set(self, s, frozenset=frozenset):
"""Freeze and intern a given sequence (set variant of intern())
Given a sequence, create a frozenset copy (if it is not
already a frozenset) and intern that frozen set. Returns the
interned set.
At first glance, interning sets may seem absurd. However,
it does enable memory savings of up to 600MB when applied
to the "inner" sets of the dependency clauses and all the
conflicts relations as well.
"""
if type(s) == frozenset:
fset = s
else:
fset = frozenset(s)
if fset in self._internmap:
return self._internmap[fset]
self._internmap[fset] = fset
return fset
def _reverse_relations(self, binary, set=set):
"""Return the reverse relations for a binary
Fetch the reverse relations for a given binary, which are
created lazily.
"""
if binary in self._reverse_package_table:
return self._reverse_package_table[binary]
rel = [set(), set(), set()]
self._reverse_package_table[binary] = rel
return rel
def build(self):
"""Compile the installability tester
This method will compile an installability tester from the
information given and (where possible) try to optimise a
few things.
"""
package_table = self._package_table
reverse_package_table = self._reverse_package_table
intern_set = self._intern_set
safe_set = set()
broken = self._broken
not_broken = ifilter_except(broken)
check = set(broken)
def safe_set_satisfies(t):
"""Check if t's dependencies can be satisfied by the safe set"""
if not package_table[t][0]:
# If it has no dependencies at all, then it is safe. :)
return True
for depgroup in package_table[t][0]:
if not any(dep for dep in depgroup if dep in safe_set):
return False
return True
# Merge reverse conflicts with conflicts - this saves some
# operations in _check_loop since we only have to check one
# set (instead of two) and we remove a few duplicates here
# and there.
#
# At the same time, intern the rdep sets
for pkg in reverse_package_table:
if pkg not in package_table:
raise RuntimeError("%s/%s/%s referenced but not added!" % pkg)
deps, con = package_table[pkg]
rdeps, rcon, rdep_relations = reverse_package_table[pkg]
if rcon:
if not con:
con = intern_set(rcon)
else:
con = intern_set(con | rcon)
package_table[pkg] = (deps, con)
reverse_package_table[pkg] = (intern_set(rdeps), con,
intern_set(rdep_relations))
# Check if we can expand broken.
for t in not_broken(iter_except(check.pop, KeyError)):
# This package is not known to be broken... but it might be now
isb = False
for depgroup in package_table[t][0]:
if not any(not_broken(depgroup)):
# A single clause is unsatisfiable, the
# package can never be installed - add it to
# broken.
isb = True
break
if not isb:
continue
broken.add(t)
if t not in reverse_package_table:
continue
check.update(reverse_package_table[t][0] - broken)
if broken:
# Since a broken package will never be installable, nothing that depends on it
# will ever be installable. Thus, there is no point in keeping relations on
# the broken package.
seen = set()
empty_set = frozenset()
null_data = (frozenset([empty_set]), empty_set)
for b in (x for x in broken if x in reverse_package_table):
for rdep in (r for r in not_broken(reverse_package_table[b][0])
if r not in seen):
ndep = intern_set((x - broken) for x in package_table[rdep][0])
package_table[rdep] = (ndep, package_table[rdep][1] - broken)
seen.add(rdep)
# Since they won't affect the installability of any other package, we might as
# as well null their data. This memory for these packages, but likely there
# will only be a handful of these "at best" (fsvo of "best")
for b in broken:
package_table[b] = null_data
if b in reverse_package_table:
del reverse_package_table[b]
# Now find an initial safe set (if any)
check = set()
for pkg in package_table:
if package_table[pkg][1]:
# has (reverse) conflicts - not safe
continue
if not safe_set_satisfies(pkg):
continue
safe_set.add(pkg)
if pkg in reverse_package_table:
# add all rdeps (except those already in the safe_set)
check.update(reverse_package_table[pkg][0] - safe_set)
# Check if we can expand the initial safe set
for pkg in iter_except(check.pop, KeyError):
if package_table[pkg][1]:
# has (reverse) conflicts - not safe
continue
if safe_set_satisfies(pkg):
safe_set.add(pkg)
if pkg in reverse_package_table:
# add all rdeps (except those already in the safe_set)
check.update(reverse_package_table[pkg][0] - safe_set)
eqv_table = self._build_eqv_packages_table(package_table,
reverse_package_table)
return InstallabilitySolver(package_table,
reverse_package_table,
self._testing, self._broken,
self._essentials, safe_set,
eqv_table)
def _build_eqv_packages_table(self, package_table,
reverse_package_table,
frozenset=frozenset):
"""Attempt to build a table of equivalent packages
This method attempts to create a table of packages that are
equivalent (in terms of installability). If two packages (A
and B) are equivalent then testing the installability of A is
the same as testing the installability of B. This equivalency
also applies to co-installability.
The example cases:
* aspell-*
* ispell-*
Cases that do *not* apply:
* MTA's
The theory:
The packages A and B are equivalent iff:
reverse_depends(A) == reverse_depends(B) AND
conflicts(A) == conflicts(B) AND
depends(A) == depends(B)
Where "reverse_depends(X)" is the set of reverse dependencies
of X, "conflicts(X)" is the set of negative dependencies of X
(Breaks and Conflicts plus the reverse ones of those combined)
and "depends(X)" is the set of strong dependencies of X
(Depends and Pre-Depends combined).
To be honest, we are actually equally interested another
property as well, namely substitutability. The package A can
always used instead of B, iff:
reverse_depends(A) >= reverse_depends(B) AND
conflicts(A) <= conflicts(B) AND
depends(A) == depends(B)
(With the same definitions as above). Note that equivalency
is just a special-case of substitutability, where A and B can
substitute each other (i.e. a two-way substituation).
Finally, note that the "depends(A) == depends(B)" for
substitutability is actually not a strict requirement. There
are cases where those sets are different without affecting the
property.
"""
# Despite talking about substitutability, the method currently
# only finds the equivalence cases. Lets leave
# substitutability for a future version.
find_eqv_table = defaultdict(list)
eqv_table = {}
for pkg in reverse_package_table:
rdeps = reverse_package_table[pkg][2]
if not rdeps:
# we don't care for things without rdeps (because
# it is not worth it)
continue
deps, con = package_table[pkg]
ekey = (deps, con, rdeps)
find_eqv_table[ekey].append(pkg)
for pkg_list in find_eqv_table.values():
if len(pkg_list) < 2:
continue
eqv_set = frozenset(pkg_list)
for pkg in pkg_list:
eqv_table[pkg] = eqv_set
return eqv_table