diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 09:06:44 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 09:06:44 +0000 |
commit | ed5640d8b587fbcfed7dd7967f3de04b37a76f26 (patch) | |
tree | 7a5f7c6c9d02226d7471cb3cc8fbbf631b415303 /solenv/gdb/boost/lib/unordered.py | |
parent | Initial commit. (diff) | |
download | libreoffice-ed5640d8b587fbcfed7dd7967f3de04b37a76f26.tar.xz libreoffice-ed5640d8b587fbcfed7dd7967f3de04b37a76f26.zip |
Adding upstream version 4:7.4.7.upstream/4%7.4.7upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'solenv/gdb/boost/lib/unordered.py')
-rw-r--r-- | solenv/gdb/boost/lib/unordered.py | 129 |
1 files changed, 129 insertions, 0 deletions
diff --git a/solenv/gdb/boost/lib/unordered.py b/solenv/gdb/boost/lib/unordered.py new file mode 100644 index 000000000..ee58d0481 --- /dev/null +++ b/solenv/gdb/boost/lib/unordered.py @@ -0,0 +1,129 @@ +# -*- tab-width: 4; indent-tabs-mode: nil; py-indent-offset: 4 -*- +# Helper classes for working with Boost.Unordered. +# +# Copyright (C) 2012 Red Hat, Inc., David Tardon <dtardon@redhat.com> +# +# This file is part of boost-gdb-printers. +# +# 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 3 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. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see <http://www.gnu.org/licenses/>. + +import gdb +import six + +class Unordered(object): + '''Common representation of Boost.Unordered types''' + + def __init__(self, value, extractor): + self.value = value + self.extractor = extractor + self.node_type = self._node_type() + + def __len__(self): + table = self.value['table_'] + if table['buckets_']: + return int(table['size_']) + else: + return 0 + + def __iter__(self): + table = self.value['table_'] + buckets = table['buckets_'] + if buckets: + first = table['cached_begin_bucket_'] + last = buckets + table['bucket_count_'] + else: + first = last = None + return self._iterator(first, last, self.node_type, self.extractor) + + def empty(self): + return not self.value['table_']['buckets_'] + + def _node_type(self): + hash_table = self.value['table_'].type.fields()[0] + assert hash_table.is_base_class + hash_buckets = hash_table.type.fields()[0] + assert hash_buckets.is_base_class + node_type = gdb.lookup_type("%s::node" % hash_buckets.type) + assert node_type != None + return node_type + + class _iterator(six.Iterator): + '''Iterator for Boost.Unordered types''' + + def __init__(self, first_bucket, last_bucket, node_type, extractor): + self.bucket = first_bucket + self.last_bucket = last_bucket + self.node = self.bucket + self.node_type = node_type + self.value_type = self._value_type() + self.extractor = extractor + + def __iter__(self): + return self + + def __next__(self): + if self.node: + self.node = self.node.dereference()['next_'] + + # we finished the current bucket: go on to the next non-empty one + if not self.node: + while not self.node and self.bucket != self.last_bucket: + self.bucket += 1 + self.node = self.bucket.dereference()['next_'] + + # sorry, no node available + if not self.node or self.node == self.bucket: + raise StopIteration() + + mapped = self._value() + return (self.extractor.key(mapped), self.extractor.value(mapped)) + + def _value(self): + assert self.node != self.bucket # bucket node has no value + assert self.node != None + node = self.node.dereference().cast(self.node_type) + return node['data_'].cast(self.value_type) + + def _value_type(self): + value_base = self.node_type.fields()[1] + assert value_base.is_base_class + return value_base.type.template_argument(0) + +class Map(Unordered): + + def __init__(self, value): + super(Map, self).__init__(value, self._extractor()) + + class _extractor(object): + + def key(self, node): + return node['first'] + + def value(self, node): + return node['second'] + +class Set(Unordered): + + def __init__(self, value): + super(Set, self).__init__(value, self._extractor()) + + class _extractor(object): + + def key(self, node): + return None + + def value(self, node): + return node + +# vim:set filetype=python shiftwidth=4 softtabstop=4 expandtab: |