TheCloud/dwarfs
1
0
Fork 0
mirror of https://github.com/mhx/dwarfs.git synced 2025-09-16 07:46:27 -04:00
Code Issues Packages Projects Releases Wiki Activity

Clean up inode manager

Browse Source
This commit is contained in:
Marcus Holland-Moritz 2020-12-07 22:20:17 +01:00
parent 63c9e9a3c3
commit e3bbeac5d6
1 changed files with 176 additions and 169 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

135
src/dwarfs/inode_manager.cpp
View File

@ -161,7 +161,72 @@ class inode_manager_ : public inode_manager::impl {
void order_inodes(std::shared_ptr<script> scr, file_order_mode file_order,
uint32_t first_inode,
inode_manager::inode_cb const& fn) override {
inode_manager::inode_cb const& fn) override;
void
for_each_inode(std::function<void(std::shared_ptr<inode> const&)> const& fn)
const override {
for (const auto& ino : inodes_) {
fn(ino);
}
}
private:
void order_inodes_by_path() {
std::vector<std::string> paths;
std::vector<size_t> index(inodes_.size());
paths.reserve(inodes_.size());
for (auto const& ino : inodes_) {
paths.emplace_back(ino->any()->path());
}
std::iota(index.begin(), index.end(), size_t(0));
std::sort(index.begin(), index.end(),
[&](size_t a, size_t b) { return paths[a] < paths[b]; });
std::vector<std::shared_ptr<inode>> tmp;
tmp.reserve(inodes_.size());
for (size_t ix : index) {
tmp.emplace_back(inodes_[ix]);
}
inodes_.swap(tmp);
}
void order_inodes_by_similarity() {
std::sort(
inodes_.begin(), inodes_.end(),
[](const std::shared_ptr<inode>& a, const std::shared_ptr<inode>& b) {
auto ash = a->similarity_hash();
auto bsh = b->similarity_hash();
return ash < bsh ||
(ash == bsh && (a->size() > b->size() ||
(a->size() == b->size() &&
a->any()->path() < b->any()->path())));
});
}
void order_inodes_by_nilsimsa(inode_manager::inode_cb const& fn,
uint32_t inode_no);
void number_inodes(size_t first_no) {
for (auto& i : inodes_) {
i->set_num(first_no++);
}
}
std::vector<std::shared_ptr<inode>> inodes_;
log_proxy<LoggerPolicy> log_;
};
template <typename LoggerPolicy>
void inode_manager_<LoggerPolicy>::order_inodes(
std::shared_ptr<script> scr, file_order_mode file_order,
uint32_t first_inode, inode_manager::inode_cb const& fn) {
switch (file_order) {
case file_order_mode::NONE:
log_.info() << "keeping inode order";
@ -211,46 +276,9 @@ class inode_manager_ : public inode_manager::impl {
}
}
void order_inodes_by_path() {
std::vector<std::string> paths;
std::vector<size_t> index(inodes_.size());
paths.reserve(inodes_.size());
for (auto const& ino : inodes_) {
paths.emplace_back(ino->any()->path());
}
std::iota(index.begin(), index.end(), size_t(0));
std::sort(index.begin(), index.end(),
[&](size_t a, size_t b) { return paths[a] < paths[b]; });
std::vector<std::shared_ptr<inode>> tmp;
tmp.reserve(inodes_.size());
for (size_t ix : index) {
tmp.emplace_back(inodes_[ix]);
}
inodes_.swap(tmp);
}
void order_inodes_by_similarity() {
std::sort(
inodes_.begin(), inodes_.end(),
[](const std::shared_ptr<inode>& a, const std::shared_ptr<inode>& b) {
auto ash = a->similarity_hash();
auto bsh = b->similarity_hash();
return ash < bsh ||
(ash == bsh && (a->size() > b->size() ||
(a->size() == b->size() &&
a->any()->path() < b->any()->path())));
});
}
void order_inodes_by_nilsimsa(inode_manager::inode_cb const& fn,
uint32_t inode_no) {
template <typename LoggerPolicy>
void inode_manager_<LoggerPolicy>::order_inodes_by_nilsimsa(
inode_manager::inode_cb const& fn, uint32_t inode_no) {
auto finalize_inode = [&](auto& ino) {
ino->set_num(inode_no++);
fn(ino);
@ -302,8 +330,7 @@ class inode_manager_ : public inode_manager::impl {
auto& db = cache[b];
return da.similarity > db.similarity ||
(da.similarity == db.similarity &&
(da.size > db.size ||
(da.size == db.size && da.path < db.path)));
(da.size > db.size || (da.size == db.size && da.path < db.path)));
};
size_t depth = 0;
@ -316,8 +343,7 @@ class inode_manager_ : public inode_manager::impl {
if (index.size() > max_depth) {
while (depth < max_depth && depth + depth_step < index.size()) {
std::partial_sort(index.begin() + depth,
index.begin() + depth + depth_step, index.end(),
cmp);
index.begin() + depth + depth_step, index.end(), cmp);
depth += depth_step;
if (cache[index[0]].similarity - cache[index[depth - 1]].similarity >
sim_thresh_depth) {
@ -373,25 +399,6 @@ class inode_manager_ : public inode_manager::impl {
}
}
void number_inodes(size_t first_no) {
for (auto& i : inodes_) {
i->set_num(first_no++);
}
}
void
for_each_inode(std::function<void(std::shared_ptr<inode> const&)> const& fn)
const override {
for (const auto& ino : inodes_) {
fn(ino);
}
}
private:
std::vector<std::shared_ptr<inode>> inodes_;
log_proxy<LoggerPolicy> log_;
};
inode_manager::inode_manager(logger& lgr)
: impl_(make_unique_logging_object<impl, inode_manager_, logger_policies>(
lgr)) {}