King_DuckZ/dindexer
1
0
Fork 0
mirror of https://github.com/KingDuckZ/dindexer.git synced 2025-02-22 12:44:55 +00:00
Code Issues Releases Wiki Activity

Fix wrong hashes for directories

Browse source 
set_listing must sort by abs_path, because the path received by
lower_bound() is an absolute path.
Inside hash_dir() the path passed to append_to_vec() is really
the basename of the current item. This will become more clear
in one of the next commits.
This commit is contained in:
King_DuckZ 2016-03-08 18:58:47 +01:00
parent 7b0189a119
commit 63c703f20d
2 changed files with 20 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

26
src/machinery/scantask/hashing.cpp
View file

@ -23,17 +23,22 @@
#include <boost/range/empty.hpp> #include <boost/range/empty.hpp>
#include <boost/utility/string_ref.hpp> #include <boost/utility/string_ref.hpp>
//#define INDEXER_VERBOSE
#if defined(INDEXER_VERBOSE)
# include <iostream>
#endif
namespace mchlib { namespace mchlib {
namespace { namespace {
void append_to_vec (std::vector<char>& parDest, const TigerHash& parHash, const std::string& parString) {
void append_to_vec (std::vector<char>& parDest, const TigerHash& parHash, boost::string_ref parString) {
const auto old_size = parDest.size(); const auto old_size = parDest.size();
parDest.resize(old_size + sizeof(TigerHash) + parString.size()); parDest.resize(old_size + sizeof(TigerHash) + parString.size());
std::copy(parHash.byte_data, parHash.byte_data + sizeof(TigerHash), parDest.begin() + old_size); std::copy(parHash.byte_data, parHash.byte_data + sizeof(TigerHash), parDest.begin() + old_size);
std::copy(parString.begin(), parString.end(), parDest.begin() + old_size + sizeof(TigerHash)); std::copy(parString.begin(), parString.end(), parDest.begin() + old_size + sizeof(TigerHash));
} }
void append_to_vec (std::vector<char>& parDest, boost::string_ref parString) { void append_to_vec (std::vector<char>& parDest, const std::string& parString) {
const auto old_size = parDest.size(); const auto old_size = parDest.size();
parDest.resize(old_size + parString.size()); parDest.resize(old_size + parString.size());
std::copy(parString.begin(), parString.end(), parDest.begin() + old_size); std::copy(parString.begin(), parString.end(), parDest.begin() + old_size);
@ -46,26 +51,29 @@ namespace mchlib {
//is a direct child of current entry //is a direct child of current entry
std::vector<char> dir_blob; std::vector<char> dir_blob;
#if defined(INDEXER_VERBOSE) #if defined(INDEXER_VERBOSE)
std::cout << "Making initial hash for " << parCurrDir << "...\n"; std::cout << "Making initial hash for " << parEntry.abs_path << "...\n";
#endif #endif
PathName curr_dir(parEntry.path);
for (auto it = parList.begin(); it != parList.end(); ++it) { for (auto it = parList.begin(); it != parList.end(); ++it) {
assert(PathName(parEntry.abs_path) == PathName(it->abs_path).pop_right()); assert(PathName(parEntry.abs_path) == PathName(it->abs_path).pop_right());
PathName curr_path(it->path);
const std::string basename = make_relative_path(curr_dir, curr_path).path();
if (it->is_directory) { if (it->is_directory) {
auto cd_list = MutableSetListingView(it); auto cd_list = MutableSetListingView(it);
assert(boost::empty(cd_list) or cd_list.begin()->abs_path != it->abs_path); assert(boost::empty(cd_list) or cd_list.begin()->abs_path != it->abs_path);
hash_dir(*it, cd_list, parIgnoreErrors); hash_dir(*it, cd_list, parIgnoreErrors);
append_to_vec(dir_blob, it->hash, it->path); append_to_vec(dir_blob, it->hash, basename);
} }
else { else {
append_to_vec(dir_blob, it->path); append_to_vec(dir_blob, basename);
} }
} }
tiger_data(dir_blob, parEntry.hash); tiger_data(dir_blob, parEntry.hash);
#if defined(INDEXER_VERBOSE) #if defined(INDEXER_VERBOSE)
std::cout << "Got intermediate hash for dir " << parCurrDir << std::cout << "Got intermediate hash for dir " << parEntry.abs_path <<
": " << tiger_to_string(parEntry.hash) << ": " << tiger_to_string(parEntry.hash) <<
' ' << parEntry.mime_type << '\n'; ' ' << parEntry.mime_type << '\n';
#endif #endif
@ -74,7 +82,7 @@ namespace mchlib {
for (auto it = first_file(parList); it != parList.end(); ++it) { for (auto it = first_file(parList); it != parList.end(); ++it) {
assert(not it->is_directory); assert(not it->is_directory);
#if defined(INDEXER_VERBOSE) #if defined(INDEXER_VERBOSE)
std::cout << "Hashing file " << it->abs_path << "..."; std::cout << "Hashing file " << it->abs_path << "...\n";
#endif #endif
//TODO: notify callback //TODO: notify callback
try { try {
@ -93,7 +101,7 @@ namespace mchlib {
} }
#if defined(INDEXER_VERBOSE) #if defined(INDEXER_VERBOSE)
std::cout << "Final hash for dir " << parCurrDir << " is " << tiger_to_string(parEntry.hash) << '\n'; std::cout << "Final hash for dir " << parEntry.abs_path << " is " << tiger_to_string(parEntry.hash) << '\n';
#endif #endif
parEntry.hash_valid = true; parEntry.hash_valid = true;
} }

6
src/machinery/set_listing.cpp
View file

@ -29,14 +29,14 @@ namespace mchlib {
//to be made. //to be made.
struct FileRecordDataForSearch { struct FileRecordDataForSearch {
FileRecordDataForSearch ( const char* parPath, uint16_t parLevel, bool parIsDir) : FileRecordDataForSearch ( const char* parPath, uint16_t parLevel, bool parIsDir) :
path(parPath), abs_path(parPath),
level(parLevel), level(parLevel),
is_directory(parIsDir) is_directory(parIsDir)
{ {
assert(parPath); assert(parPath);
} }
boost::string_ref path; boost::string_ref abs_path;
uint16_t level; uint16_t level;
bool is_directory; bool is_directory;
}; };
@ -48,7 +48,7 @@ namespace mchlib {
return return
(l.level < r.level) (l.level < r.level)
or (l.level == r.level and l.is_directory and not r.is_directory) or (l.level == r.level and l.is_directory and not r.is_directory)
or (l.level == r.level and l.is_directory == r.is_directory and l.path < r.path) or (l.level == r.level and l.is_directory == r.is_directory and l.abs_path < r.abs_path)
//sort by directory - parent first, children later //sort by directory - parent first, children later
//(level == o.level and is_dir and not o.is_dir) //(level == o.level and is_dir and not o.is_dir)