broom now can remove duplicates and create symlinks !; Updated README; Moved entry groups enum in the entry.hpp file; Broom seems to has become quite a useful tool !

Author: Unbewohnte

README.md

@@ -57,7 +57,7 @@ broom [FLAGS..] [COMMAND] [DIRECTORY]
 
 [COMMANDS]
 
-- `sweep` -> scan for duplicate files, save results in a file and REMOVE empty files
+- `sweep` -> scan for duplicate files, REMOVE empty files and REPLACE other duplicates with symlinks
 - `scan` -> scan and save results in a file without removing anything [DEFAULT]
 
 
@@ -66,9 +66,9 @@ broom [FLAGS..] [COMMAND] [DIRECTORY]
 ### Examples
 
 - `broom scan -od . ~/homework`
-- `broom sweep ~/homework/I/have/a/lot/of/empty/files/here/for/some/reason`
+- `broom sweep ~/homework`
 
-after the scan the results file will be saved in your current working directory, scan results file contains
+after the scan the results file will be saved in your current working directory, unless you specified it to be somewhere else. Scan results file contains
 a list of duplicate files that are grouped together so you can see EXACTLY WHERE each duplicate is in the filesystem.
 
 ---
@@ -80,4 +80,5 @@ GPLv3
 
 ## TODO
 - Make it go  `P` A `R` A `L` L `E` L
-- Output approximate size that could be freed
+- ~~Output approximate size that could be freed~~
+- ~~Remove duplicates and create symlinks~~

build/CMakeLists.txt

@@ -22,5 +22,5 @@ endif()
 
 set(EXECUTABLE_OUTPUT_PATH ../bin)
 
-add_executable(broom ../src/main.cpp ../src/entry.cpp ../src/broom.cpp ../src/group.hpp)
+add_executable(broom ../src/main.cpp ../src/entry.cpp ../src/broom.cpp)
 target_link_libraries(broom Threads::Threads)

src/broom.cpp

@@ -28,7 +28,6 @@ along with broom.  If not, see <https://www.gnu.org/licenses/>.
 
 #include "entry.hpp"
 #include "broom.hpp"
-#include "group.hpp"
 
 namespace broom {
 
@@ -52,15 +51,15 @@ std::vector<entry::Entry> Broom::track(const std::filesystem::path path) {
         );
 
         for (auto dir_entry : std::filesystem::recursive_directory_iterator(path, options)) {
-            if (!dir_entry.is_regular_file()) {
+            if (!dir_entry.is_regular_file() || std::filesystem::is_symlink(dir_entry.path())) {
                 // skip everything that we cannot process so easily
                 continue;
             };
 
             entry::Entry entry(dir_entry.path());
             tracked_entries.push_back(entry);
         }
-    } else if (std::filesystem::is_regular_file(path)) {
+    } else if (std::filesystem::is_regular_file(path) && !std::filesystem::is_symlink(path)) {
         // just a file
         entry::Entry entry(path);
         tracked_entries.push_back(entry);
@@ -186,7 +185,7 @@ uintmax_t Broom::find_empty_files(std::vector<entry::Entry>& tracked_entries) {
     for (entry::Entry& entry : tracked_entries) {
         if (entry.filesize == 0) {
             // empty files can`t be considered as duplicates. assign a group
-            entry.group = group::EMPTY;
+            entry.group = entry::EMPTY;
             found_empty_files++;
         }
     }
@@ -199,7 +198,7 @@ uintmax_t Broom::remove_empty_files(std::vector<entry::Entry>& tracked_entries)
     uintmax_t removed = 0;
 
     tracked_entries.erase(std::remove_if(tracked_entries.begin(), tracked_entries.end(), [&removed](entry::Entry& entry) -> bool {
-        if (entry.group == group::EMPTY) {
+        if (entry.group == entry::EMPTY) {
             try {
                 entry.remove();
                 removed++;
@@ -219,11 +218,11 @@ uintmax_t Broom::remove_empty_files(std::vector<entry::Entry>& tracked_entries)
 // marks every entry without any group as a duplicate
 void Broom::mark_as_duplicates(std::vector<entry::Entry>& tracked_entries) {
     for (entry::Entry& entry : tracked_entries) {
-        if (entry.group == group::EMPTY) {
+        if (entry.group == entry::EMPTY) {
             // do not mess up grouping
             continue;
         }
-        entry.group = group::DUPLICATE;
+        entry.group = entry::DUPLICATE;
     }
 };
 
@@ -252,4 +251,33 @@ std::map<std::string, std::vector<entry::Entry>> Broom::group_duplicates(std::ve
     return duplicate_groups;
 };
 
+// REMOVES every duplicate file in a group except the first one and creates symlinks pointing to the
+// first remaining real file
+void Broom::remove_duplicates_make_symlinks(const std::map<std::string, std::vector<entry::Entry>> grouped_duplicates) {
+    for (const auto& record : grouped_duplicates) {
+        unsigned int i = 0;
+        std::filesystem::path original_file_path;
+
+        for (const auto& duplicate_entry : record.second) {
+            if (i == 0) {
+                // the first duplicate in the group. Save it
+                original_file_path = duplicate_entry.path;
+            } else {
+                // not the first entry; REMOVE it and create a symlink,
+                // pointing to the real file
+                std::filesystem::path removed_duplicate_path = duplicate_entry.path;
+                try {
+                    // remove the entry
+                    duplicate_entry.remove();
+                    // make a symlink
+                    std::filesystem::create_symlink(original_file_path, removed_duplicate_path);
+                } catch(...) {}
+            }
+
+            // serves only the first iteration. It doesn`t matter if it is not incremented after that
+            i++;
+        }
+    }
+};
+
 }

src/broom.hpp

@@ -61,6 +61,10 @@ class Broom {
     // string of pieces. REMOVES EVERYTHING FROM GIVEN TRACKED ENTRIES
     std::map<std::string, std::vector<entry::Entry>> group_duplicates(std::vector<entry::Entry>& tracked_entries);
 
+    // REMOVES every duplicate file in a group except the first one and creates symlinks pointing to the
+    // first remaining real file
+    void remove_duplicates_make_symlinks(const std::map<std::string, std::vector<entry::Entry>> grouped_duplicates);
+
     // creates a list of duplicate, empty files and puts it into a file
     void create_scan_results_list(const std::map<std::string, std::vector<entry::Entry>> grouped_duplicates, const std::filesystem::path dir = ".", const std::string filename = "scan_results.txt");
 };

src/entry.cpp

@@ -95,7 +95,7 @@ void Entry::get_pieces() {
 };
 
 // Remove entry from the disk
-void Entry::remove() {
+void Entry::remove() const {
     std::filesystem::remove(path);
 };
 

src/entry.hpp

@@ -26,10 +26,14 @@ along with broom.  If not, see <https://www.gnu.org/licenses/>.
 #include <iomanip>
 #include <string>
 
-#include "group.hpp"
-
 
 namespace entry {
+
+enum Group {
+    DUPLICATE,
+    EMPTY,
+};
+
 // 3 pieces (beginning, middle and end of the file)
 const uint8_t PIECE_SIZE = 75;
 const uint8_t PIECES_AMOUNT = 3;
@@ -40,7 +44,7 @@ class Entry {
     std::filesystem::path path; // set via constructor
     uintmax_t filesize; // set via constructor
     std::string pieces; // 3 hex-represented pieces of file; set only via a method call to not stress the disk
-    group::Group group; // set externally
+    Group group; // set externally
 
     Entry(const std::filesystem::path entry_path);
     ~Entry();
@@ -51,7 +55,7 @@ class Entry {
     void get_pieces();
 
     // REMOVE entry from the disk
-    void remove();
+    void remove() const;
 };
 
 }

src/group.hpp

@@ -28,7 +28,7 @@ along with broom.  If not, see <https://www.gnu.org/licenses/>.
 #include "broom.hpp"
 
 // Broom version number
-#define VERSION "v0.2.3"
+#define VERSION "v0.3.0"
 
 void print_help() {
     std::cout
@@ -39,7 +39,7 @@ void print_help() {
     << "-od | --output-directory -> path to the directory to save results file in" << std::endl << std::endl
 
     << "[COMMANDS]" << std::endl
-    << "sweep -> scan for duplicate files, save results in a file and REMOVE empty files" << std::endl
+    << "sweep -> scan for duplicate files, REMOVE empty files and REPLACE other duplicates with symlinks" << std::endl
     << "scan -> scan and save results in a file without removing anything [DEFAULT]" << std::endl << std::endl
 
     << "[DIRECTORY]" << std::endl
@@ -170,26 +170,35 @@ int main(int argc, char* argv[]) {
 
         std::cout << "[INFO] " << tracked_entries.size() << " files left being tracked" << std::endl;
 
-        auto grouped_duplicates = broom.group_duplicates(tracked_entries);
-
-        if (grouped_duplicates.size() == 0) {
+        if (tracked_entries.size() == 0) {
+            // No duplicates at all !
             std::cout << "[INFO] Nothing I can help with ! Congratulations !" << std::endl;
             return 0;
         }
 
-        // now only files with a non-unique size and contents are being tracked
-        // are they REALLY duplicates ?
-        // better to leave the REALL cleanup for the user, saving these entries in a file, than doing a blind and possibly destructive purge
-        broom.create_scan_results_list(grouped_duplicates, results_file_dir_path);
-        std::cout << "[INFO] Created scan results file" << std::endl;
+        // make duplicate groups from all this mess that tracked_entries right now are
+        auto grouped_duplicates = broom.group_duplicates(tracked_entries);
 
-        // output a little information about how much space could be freed if every duplicate
-        // in the group will be deleted but one
         double could_be_freed = 0;
         for (auto& record : grouped_duplicates) {
             could_be_freed += record.second[0].filesize * (record.second.size() - 1);
         }
-        std::cout <<"[INFO] " << could_be_freed / 1024 / 1024 << " MB could be freed" << std::endl;
+
+        if (!sweeping) {
+            // output a little information about how much space could be freed if every duplicate
+            // in the group will be deleted but one
+            std::cout <<"[INFO] " << could_be_freed / 1024 / 1024 << " MB could be freed" << std::endl;
+
+            broom.create_scan_results_list(grouped_duplicates, results_file_dir_path);
+            std::cout << "[INFO] Created scan results file" << std::endl;
+
+        } else {
+            // remove duplicates and create symlinks
+            std::cout << "[INFO] Removing duplicates and creating symlinks..." << std::endl;
+            broom.remove_duplicates_make_symlinks(grouped_duplicates);
+
+            std::cout <<"[INFO] Freed approximately " << could_be_freed / 1024 / 1024 << " MB (May be incorrect)" << std::endl;
+        }
 
     } catch(const std::exception& e) {
         std::cerr