Untracking by contents

src/broom.cpp

@@ -32,28 +32,8 @@ Broom::Broom(Options options) {
     m_sweeping = options.sweeping;
 };
 
-Broom::~Broom() {};
-
-// Print current statistics
-void Broom::print_statistics() {
-    std::cout
-    << "| sweeped " << m_sweeped_files << " files" << std::endl
-    << "| with a total size of " << m_sweeped_size << " bytes" << std::endl
-    << std::endl;
-};
-
-// returns amount of regular files in path, searching recursively.
-// Throws an invalid_argument error in case path does not exist
-uintmax_t amount_of_files(const std::filesystem::path path) {
-    if (!std::filesystem::exists(path)) {
-        throw std::invalid_argument("\"" + path.string() + "\"" + " does not exist !");
-    };
-
-    if (!std::filesystem::is_directory(path)) {
-        throw std::invalid_argument("\"" + path.string() + "\"" + " is not a directory");
-    };
-
-    return std::distance(std::filesystem::directory_iterator(path), std::filesystem::directory_iterator{});
+Broom::~Broom() {
+    m_tracked_entries.clear();
 };
 
 // recursively track every file that lies in given path. Throws an invalid_argument
@@ -90,7 +70,7 @@ void Broom::track(const std::filesystem::path path) {
         auto tracking_time = std::chrono::high_resolution_clock::now();
 
         std::cout
-        << "Tracking took "
+        << "[BENCHMARK] Tracking took "
         << std::chrono::duration_cast<std::chrono::milliseconds>(tracking_time - t0).count()
         << " ms" << std::endl;
     };
@@ -102,16 +82,16 @@ uintmax_t Broom::untrack_unique_sizes() {
     // key: size, value: amount of occurences
     std::map<uintmax_t, uintmax_t> sizes_map;
 
-    for (Entry& entry : m_tracked_entries) {
+    for (auto entry_iter = m_tracked_entries.begin(); entry_iter != m_tracked_entries.end(); entry_iter++) {
         // check if size of this entry is already in the map
         // if yes --> increment occurences counter
         // if not --> add it to the map with a counter of 1
-        entry.get_size();
+        entry_iter->get_size();
 
-        auto iterator = sizes_map.find(entry.filesize);
+        auto iterator = sizes_map.find(entry_iter->filesize);
         if (iterator == sizes_map.end()) {
             // there is no such size
-            sizes_map.insert({entry.filesize, 1});
+            sizes_map.insert({entry_iter->filesize, 1});
         } else {
             // there is such size
             sizes_map[iterator->first]++;
@@ -130,85 +110,105 @@ uintmax_t Broom::untrack_unique_sizes() {
         // std::cout << "duplicate fsize: " << iter->first << " occurences: " << iter->second << std::endl;
 
         return false;
-    }));
+    }), m_tracked_entries.end());
+
+    // std::cout << "Size after untracking by size: " << m_tracked_entries.size() << std::endl;
 
 
     return untracked;
 };
 
-// removes entries with unique first and last 20 bytes. Returns amount of
+// removes entries with the same content-pieces. Returns amount of
 // files that are no longer being tracked
-// uintmax_t Broom::untrack_unique_contents() {
-//     // contents, occurences
-//     std::map<char[CHECKSUM_SIZE], uintmax_t> contents_map;
-//     std::map<char[CHECKSUM_SIZE], uintmax_t>::iterator iterator;
-//
-//     for (Entry& entry : m_tracked_entries) {
-//         // the same logic:
-//         // check if contents of this entry is already in the map
-//         // if yes --> increment occurences counter
-//         // if not --> add it to the map with a counter of 1
-//
-//         iterator = contents_map.find(entry.checksum);
-//
-//         if (iterator == contents_map.end()) {
-//             // add it to the map
-//             contents_map.insert(std::pair<char[CHECKSUM_SIZE], uintmax_t>(entry.checksum, 1));
-//         } else {
-//             // increment occurences counter
-//             uintmax_t occurences = contents_map[iterator->first];
-//             contents_map[iterator->first] = occurences++;
-//         };
-//     };
-//
-//     uintmax_t untracked = 0;
-//     for (std::pair<const char[CHECKSUM_SIZE], uintmax_t> contents_entry : contents_map) {
-//         if (contents_entry.second > 1) {
-//             // not a unique size. Keep such entries
-//         } else {
-//             // a unique one. Untrack such an entry
-//             std::remove_if(m_tracked_entries.begin(), m_tracked_entries.end(), [contents_entry](Entry e) -> bool {
-//                 return (e.compare_checksums(contents_entry.first));
-//             });
-//             untracked++;
-//         };
-//     };
-//
-//     return untracked;
-// };
+uintmax_t Broom::untrack_unique_contents() {
+    // contents, occurences
+    std::map<std::string, uintmax_t> contents_map;
+    std::map<std::string, uintmax_t>::iterator map_iter;
+
+    for (auto entry_iter = m_tracked_entries.begin(); entry_iter != m_tracked_entries.end();) {
+        // the same logic:
+        // check if contents of this entry is already in the map
+        // if yes --> increment occurences counter
+        // if not --> add it to the map with a counter of 1
+
+        try{
+            // can get "permission denied" when opening file
+            entry_iter->get_pieces();
+        } catch(const std::ifstream::failure& e) {
+            // there is nothing we can do. Untrack this entry
+            // std::cerr << e.what();
+            entry_iter = m_tracked_entries.erase(entry_iter);
+            continue;
+        }
+
+        map_iter = contents_map.find(entry_iter->pieces);
+        if (map_iter == contents_map.end()) {
+            // add it to the map
+            contents_map.insert({entry_iter->pieces, 1});
+            // std::cout << "First time seeing this piece: " << entry_iter->pieces << std::endl;
+        } else {
+            // increment occurences counter
+            contents_map[map_iter->first]++;
+        };
+
+        entry_iter++;
+    };
+
+    uintmax_t untracked = 0;
+    m_tracked_entries.erase(std::remove_if(m_tracked_entries.begin(), m_tracked_entries.end(), [&untracked, contents_map](Entry entry) -> bool {
+        auto iter = contents_map.find(entry.pieces);
+        if (iter->second == 1) {
+            // unique
+            untracked++;
+            return true;
+        } else {
+            return false;
+        }
+    }), m_tracked_entries.end());
+
+    return untracked;
+};
 
 
 // find all duplicates among tracked entries, stop tracking uniques
 void Broom::find_duplicates() {
-    if (m_benchmarking) {
-        auto t0 = std::chrono::high_resolution_clock::now();
+    auto t0 = std::chrono::high_resolution_clock::now();
+
+    // print how many files are being tracked
+    uintmax_t global_untracked = m_tracked_entries.size();
+    std::cout << "[INFO] Tracking " << m_tracked_entries.size() << " files" << std::endl;
 
-        untrack_unique_sizes();
+    // untrack by size
+    uintmax_t untracked_by_size = untrack_unique_sizes();
+    global_untracked += untracked_by_size;
+    std::cout << "[INFO] Untracked " << untracked_by_size << " unique sizes" << std::endl;
 
-        auto sizes_untrack_time = std::chrono::high_resolution_clock::now();
+    auto sizes_untrack_time = std::chrono::high_resolution_clock::now();
 
+    if (m_benchmarking) {
         std::cout
-        << "Untracking by size took "
+        << "[BENCHMARK] Untracking by size took "
         << std::chrono::duration_cast<std::chrono::milliseconds>(sizes_untrack_time - t0).count()
         << " ms" << std::endl;
-    } else {
-        size_t startsize = m_tracked_entries.size();
-        std::cout << "Tracking " << startsize << std::endl;
+    };
 
-        uintmax_t global_untracked = 0;
+    // untrack by contents
+    uintmax_t untracked_by_contents = untrack_unique_contents();
+    global_untracked += untracked_by_contents;
 
-        // uintmax_t untracked_by_contents = untrack_unique_contents();
-        // global_untracked += untracked_by_contents;
-        // std::cout << "Untracked " << untracked_by_contents << " unique contents" << std::endl;
+    auto contents_untrack_time = std::chrono::high_resolution_clock::now();
 
+    if (m_benchmarking) {
+        std::cout
+        << "[BENCHMARK] Untracking by contents took "
+        << std::chrono::duration_cast<std::chrono::milliseconds>(contents_untrack_time - sizes_untrack_time).count()
+        << " ms" << std::endl;
+    };
 
-        uintmax_t untracked_by_size = untrack_unique_sizes();
-        global_untracked += untracked_by_size;
-        std::cout << "Untracked " << untracked_by_size << " unique sizes" << std::endl;
+    std::cout << "[INFO] Untracked " << untracked_by_contents << " unique contents" << std::endl;
 
+    std::cout << "[INFO] Duplicates: " << m_tracked_entries.size() << std::endl;
 
-        std::cout << "Duplicates: " << startsize - global_untracked << std::endl;
-    };
 };
 
 // remove ALL duplicate files

src/broom.hpp

@@ -48,13 +48,6 @@ public:
     Broom(Options options);
     ~Broom();
 
-    // Print current statistics
-    void print_statistics();
-
-    // returns amount of regular files in path, searching recursively.
-    // Throws an invalid_argument error in case path does not exist
-    uintmax_t amount_of_files(const std::filesystem::path path);
-
     // recursively track every file that lies in given path. Throws an invalid_argument
     // error in case path does not exist
     void track(const std::filesystem::path path);
@@ -66,7 +59,7 @@ public:
     // that are no longer being tracked
     uintmax_t untrack_unique_sizes();
 
-    // removes entries with unique first and last 20 bytes. Returns amount of
+    // removes entries with the same content-pieces. Returns amount of
     // files that are no longer being tracked
     uintmax_t untrack_unique_contents();
 

src/entry.cpp

@@ -19,6 +19,11 @@ along with broom.  If not, see <https://www.gnu.org/licenses/>.
 
 #include "entry.hpp"
 
+#include <iostream>
+#include <cerrno>
+#include <cstring>
+
+
 // A wrapper for every file with all necessary information
 Entry::Entry(const std::filesystem::path entry_path) {
     // path
@@ -29,54 +34,50 @@ Entry::~Entry() {};
 
 // sets this entry`s filesize
 void Entry::get_size() {
-    // filesize
     filesize = std::filesystem::file_size(path);
 };
 
-// calculates and sets this entry`s checksum
-void Entry::get_checksum() {
-    // checksum
+// reads 2 pieces from the beginning and the end of a file, converts them into
+// a convenient hex-encoded string
+void Entry::get_pieces() {
     std::fstream entry_file;
     entry_file.open(path);
 
     if (!entry_file.is_open()) {
-        throw std::ifstream::failure("Could not open \"" + path.filename().string() + "\"");
+        throw std::ifstream::failure("Could not open \"" + path.string() + "\"; reason: " + std::string(std::strerror(errno)) + "\n");
     }
 
     // TODO(Properly test it)
-    if (filesize <= CHECKSUM_SIZE) {
-        entry_file.read(checksum, CHECKSUM_SIZE);
+    char pieces_buffer[PIECE_SIZE * 2];
+    if (filesize <= PIECE_SIZE * 2) {
+        // can`t take whole 2 pieces !
+        // read the whole file then
+        entry_file.read(pieces_buffer, filesize);
     } else {
-        char start_buf[CHUNK_SIZE];
-        entry_file.read(start_buf, CHUNK_SIZE);
-
-        entry_file.seekg(CHUNK_SIZE, std::ios::end);
-        char end_buf[CHUNK_SIZE];
-        entry_file.read(end_buf, CHUNK_SIZE);
-
-        for (uint8_t i = 0; i < CHECKSUM_SIZE; i++) {
-            if (i < CHUNK_SIZE) {
-                checksum[i] = start_buf[i];
-            }
-            else if (i > CHUNK_SIZE) {
-                checksum[i] = end_buf[i - CHUNK_SIZE];
-            };
+        // read CHUNK_SIZE bytes from the beginning of the file
+        char start_buf[PIECE_SIZE];
+        entry_file.read(start_buf, PIECE_SIZE);
+        for (uint8_t i = 0; i < PIECE_SIZE; i++) {
+            pieces_buffer[i] = start_buf[i];
         };
-    };
 
+        // jump to the last CHUNK_SIZE bytes of the file and read the as well
+        entry_file.seekg(PIECE_SIZE, std::ios::end);
+        char end_buf[PIECE_SIZE];
+        entry_file.read(end_buf, PIECE_SIZE);
+        for (uint8_t i = PIECE_SIZE; i < PIECE_SIZE * 2; i++) {
+            pieces_buffer[i] = end_buf[i - PIECE_SIZE];
+        };
+    };
     entry_file.close();
-};
-
 
-// Compare this entry`s checksum with the other one.
-// If the checksums are the same -> returns true, else -> false
-bool Entry::compare_checksums(const char other_checksum[CHECKSUM_SIZE]) {
-    for (uint8_t i = 0; i < CHECKSUM_SIZE; i++) {
-        if (checksum[i] != other_checksum[i]) {
-            return false;
-        };
+    // make a convenient hex string out of pure bytes
+    std::stringstream pieces_hex;
+    for (uint8_t i = 0; i < PIECE_SIZE * 2; i++) {
+        pieces_hex << std::hex << static_cast<unsigned>(pieces_buffer[i]);
     };
-    return true;
+
+    pieces = pieces_hex.str();
 };
 
 // Remove entry from the disk

src/entry.hpp

@@ -22,18 +22,18 @@ along with broom.  If not, see <https://www.gnu.org/licenses/>.
 
 #include <filesystem>
 #include <fstream>
+#include <sstream>
+#include <iomanip>
 
-// 2 chunks (beginning and end of the file)
-const uint8_t CHUNK_SIZE = 24;
-const uint8_t CHECKSUM_SIZE = CHUNK_SIZE * 2;
+// 2 pieces (beginning and end of the file)
+const uint8_t PIECE_SIZE = 24;
 
 // A wrapper for every file with all necessary information
 class Entry {
 public:
     std::filesystem::path path;
     uintmax_t filesize;
-    char checksum[CHECKSUM_SIZE];
-
+    std::string pieces; // 2 hex-represented pieces of file (beginning and end)
 
     Entry(const std::filesystem::path entry_path);
     ~Entry();
@@ -41,12 +41,9 @@ public:
     // sets this entry`s filesize
     void get_size();
 
-    // calculates and sets this entry`s checksum
-    void get_checksum();
-
-    // Compare this entry`s checksum with the other one.
-    // If the checksums are the same -> returns true, else -> false
-    bool compare_checksums(const char other_checksum[CHECKSUM_SIZE]);
+    // reads 2 pieces from the beginning and the end of a file, converts them into
+    // a convenient hex-encoded string
+    void get_pieces();
 
     // REMOVE entry from the disk
     void remove();