TheCloud/dwarfs
1
0
Fork 0
mirror of https://github.com/mhx/dwarfs.git synced 2025-09-14 14:59:52 -04:00
Code Issues Packages Projects Releases Wiki Activity

Cleanup scanner

Browse Source
This commit is contained in:
Marcus Holland-Moritz 2020-11-29 13:19:58 +01:00
parent 85b1ca1a85
commit bd70ec7d7a
1 changed files with 48 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
src/dwarfs/scanner.cpp
View File

@ -243,6 +243,9 @@ class scanner_ : public scanner::impl {
void scan(filesystem_writer& fsw, const std::string& path, progress& prog); void scan(filesystem_writer& fsw, const std::string& path, progress& prog);
private: private:
std::shared_ptr<entry> scan_tree(const std::string& path, progress& prog);
void order_files(inode_manager& im);
const block_manager::config& cfg_; const block_manager::config& cfg_;
const scanner_options& options_; const scanner_options& options_;
std::shared_ptr<entry_factory> entry_; std::shared_ptr<entry_factory> entry_;
@ -270,12 +273,8 @@ scanner_<LoggerPolicy>::scanner_(logger& lgr, worker_group& wg,
, log_(lgr) {} , log_(lgr) {}
template <typename LoggerPolicy> template <typename LoggerPolicy>
void scanner_<LoggerPolicy>::scan(filesystem_writer& fsw, std::shared_ptr<entry>
const std::string& path, progress& prog) { scanner_<LoggerPolicy>::scan_tree(const std::string& path, progress& prog) {
log_.info() << "scanning " << path;
prog.set_status_function(status_string);
auto root = entry_->create(*os_, path); auto root = entry_->create(*os_, path);
if (root->type() != entry::E_DIR) { if (root->type() != entry::E_DIR) {
@ -367,6 +366,48 @@ void scanner_<LoggerPolicy>::scan(filesystem_writer& fsw,
} }
} }
return root;
}
template <typename LoggerPolicy>
void scanner_<LoggerPolicy>::order_files(inode_manager& im) {
switch (options_.file_order) {
case file_order_mode::NONE:
log_.info() << "keeping inode order";
break;
case file_order_mode::PATH: {
log_.info() << "ordering " << im.count() << " inodes by path name...";
auto ti = log_.timed_info();
im.order_inodes();
ti << im.count() << " inodes ordered";
break;
}
case file_order_mode::SCRIPT:
log_.info() << "ordering " << im.count() << " inodes using script...";
im.order_inodes(script_);
break;
case file_order_mode::SIMILARITY: {
log_.info() << "ordering " << im.count() << " inodes by similarity...";
auto ti = log_.timed_info();
im.order_inodes_by_similarity();
ti << im.count() << " inodes ordered";
break;
}
}
}
template <typename LoggerPolicy>
void scanner_<LoggerPolicy>::scan(filesystem_writer& fsw,
const std::string& path, progress& prog) {
log_.info() << "scanning " << path;
prog.set_status_function(status_string);
auto root = scan_tree(path, prog);
// now scan all files // now scan all files
scan_files_visitor sfv(wg_, *os_, prog); scan_files_visitor sfv(wg_, *os_, prog);
root->accept(sfv); root->accept(sfv);
@ -374,9 +415,6 @@ void scanner_<LoggerPolicy>::scan(filesystem_writer& fsw,
log_.info() << "waiting for background scanners..."; log_.info() << "waiting for background scanners...";
wg_.wait(); wg_.wait();
std::unordered_map<std::string_view, std::vector<file*>, folly::Hash>
file_hash;
log_.info() << "assigning directory and link inodes..."; log_.info() << "assigning directory and link inodes...";
uint32_t first_link_inode = 0; uint32_t first_link_inode = 0;
@ -401,32 +439,7 @@ void scanner_<LoggerPolicy>::scan(filesystem_writer& fsw,
<< prog.duplicate_files << "/" << prog.files_found << prog.duplicate_files << "/" << prog.files_found
<< " duplicate files"; << " duplicate files";
switch (options_.file_order) { order_files(*im);
case file_order_mode::NONE:
log_.info() << "keeping inode order";
break;
case file_order_mode::PATH: {
log_.info() << "ordering " << im->count() << " inodes by path name...";
auto ti = log_.timed_info();
im->order_inodes();
ti << im->count() << " inodes ordered";
break;
}
case file_order_mode::SCRIPT:
log_.info() << "ordering " << im->count() << " inodes using script...";
im->order_inodes(script_);
break;
case file_order_mode::SIMILARITY: {
log_.info() << "ordering " << im->count() << " inodes by similarity...";
auto ti = log_.timed_info();
im->order_inodes_by_similarity();
ti << im->count() << " inodes ordered";
break;
}
}
log_.info() << "assigning file inodes..."; log_.info() << "assigning file inodes...";
im->number_inodes(first_file_inode); im->number_inodes(first_file_inode);