fix: features and enhancements — issues #70, #36, #34, #33, #32, #57, #40, #37

src/client/scanner.c

@@ -16,6 +16,17 @@ DirectoryScanner* directory_scanner_create(char* root_directory, bool use_metada
                                            int exclude_count, char** include_patterns,
                                            int include_count, unsigned long long max_size,
                                            unsigned long long min_size) {
+  return directory_scanner_create_full(root_directory, use_metadata, chunk_size,
+                                       exclude_patterns, exclude_count,
+                                       include_patterns, include_count,
+                                       max_size, min_size, true);
+}
+
+DirectoryScanner* directory_scanner_create_full(char* root_directory, bool use_metadata,
+                                           unsigned long long chunk_size, char** exclude_patterns,
+                                           int exclude_count, char** include_patterns,
+                                           int include_count, unsigned long long max_size,
+                                           unsigned long long min_size, bool follow_symlinks) {
   DirectoryScanner* scanner = malloc(sizeof(DirectoryScanner));
   if (scanner == NULL)
     return NULL;
@@ -79,6 +90,7 @@ DirectoryScanner* directory_scanner_create(char* root_directory, bool use_metada
   scanner->include_count = include_count;
   scanner->max_size = max_size;
   scanner->min_size = min_size;
+  scanner->follow_symlinks = follow_symlinks;
   queue_enqueue(scanner->directories, str_dup(root_directory));
   return scanner;
 }
@@ -159,13 +171,76 @@ Chunk* directory_scanner_next(DirectoryScanner* scanner) {
 
     char* cur_path = path_cat(scanner->current_path, entry->d_name);
     struct stat stats;
-    if (stat(cur_path, &stats) != 0) {
+    // Use lstat to detect symlinks
+    if (lstat(cur_path, &stats) != 0) {
       free(cur_path);
       continue;
     }
 
+    // If follow_symlinks is enabled and this is a symlink, resolve it
+    if (scanner->follow_symlinks && S_ISLNK(stats.st_mode)) {
+      struct stat target_stats;
+      if (stat(cur_path, &target_stats) != 0) {
+        // Broken symlink, skip
+        free(cur_path);
+        continue;
+      }
+      stats = target_stats;
+    }
+
     if (S_ISDIR(stats.st_mode)) {
       queue_enqueue(scanner->directories, (void*)cur_path);
+    } else if (S_ISLNK(stats.st_mode)) {
+      // Handle symlink (not following)
+      bool excluded = false;
+      for (int i = 0; i < scanner->exclude_count; i++) {
+        if (glob_match(scanner->exclude_patterns[i], entry->d_name)) {
+          excluded = true;
+          break;
+        }
+      }
+      if (excluded) {
+        free(cur_path);
+        continue;
+      }
+
+      if (scanner->include_count > 0) {
+        bool included = false;
+        for (int i = 0; i < scanner->include_count; i++) {
+          if (glob_match(scanner->include_patterns[i], entry->d_name)) {
+            included = true;
+            break;
+          }
+        }
+        if (!included) {
+          free(cur_path);
+          continue;
+        }
+      }
+
+      File* file = file_create(cur_path);
+      if (file == NULL) {
+        free(cur_path);
+        continue;
+      }
+      file->type = FILE_TYPE_SYMLINK;
+      // Read link target
+      char link_buf[4096];
+      ssize_t link_len = readlink(cur_path, link_buf, sizeof(link_buf) - 1);
+      if (link_len >= 0) {
+        link_buf[link_len] = '\0';
+        file->link_target = str_dup(link_buf);
+      }
+      file->data->size = 0;
+      if (scanner->use_metadata)
+        file->metadata = file_metadata_create(&stats);
+      array_list_add(chunk_data, file);
+      chunk_data_size += 1; // small size for symlinks
+      if (chunk_data_size > scanner->chunk_size) {
+        free(cur_path);
+        return chunk_data_to_chunk(chunk_data);
+      }
+      free(cur_path);
     } else {
       bool excluded = false;
       for (int i = 0; i < scanner->exclude_count; i++) {