Files
FastSync/src/client/scanner.c
T
TapTap 20a26e2f5a
CI / build-and-test (push) Successful in 53s
CI / sanitizer (push) Failing after 1m1s
CI / clang-tidy (push) Successful in 5m37s
CI / build-and-test (pull_request) Successful in 53s
CI / sanitizer (pull_request) Failing after 1m2s
CI / clang-tidy (pull_request) Successful in 33s
feat: update AI automation agents, add reviewer agent, fix leaks, enhance CI
Agents:
- cmake-expert: fix stale CMake version (4.1 -> 3.22), add OpenSSL/xxHash deps
- integrator: fix stale test.py references to tests/integration/, update CI docs
- test-writer: update integration test patterns for modular test structure
- reviewer: new comprehensive PR reviewer (code, build, CI, docs, quality)

Bug fixes:
- delta.c: set instructions[i].type = DELTA_INSTR_LITERAL in deserialize
- scanner.c: free ArrayList when directory_scanner_next returns NULL

CI:
- Add sanitizer job (ASan + UBSan) that catches real leaks
- Add clang-tidy job with proper warning/error detection
- Remove || true masking (fixes now make sanitizer useful)
2026-07-19 21:04:30 +02:00

161 lines
4.6 KiB
C

#include "scanner.h"
#include "array_list.h"
#include "chunk.h"
#include "file.h"
#include "queue.h"
#include "utils.h"
#include <dirent.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/stat.h>
#include <unistd.h>
DirectoryScanner *directory_scanner_create(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) {
DirectoryScanner *scanner = malloc(sizeof(DirectoryScanner));
scanner->directories = queue_create(100, free);
scanner->current_dir = NULL;
scanner->current_path = NULL;
scanner->use_metadata = use_metadata;
scanner->chunk_size = chunk_size > 0 ? chunk_size : DESIRED_CHUNK_SIZE;
scanner->exclude_patterns = exclude_patterns;
scanner->exclude_count = exclude_count;
scanner->include_patterns = include_patterns;
scanner->include_count = include_count;
scanner->max_size = max_size;
scanner->min_size = min_size;
queue_enqueue(scanner->directories, str_dup(root_directory));
return scanner;
}
void directory_scanner_destroy(DirectoryScanner *scanner) {
if (scanner == NULL)
return;
if (scanner->current_dir) {
closedir(scanner->current_dir);
scanner->current_dir = NULL;
}
free(scanner->current_path);
queue_destroy(scanner->directories);
free(scanner);
}
static Chunk *chunk_data_to_chunk(ArrayList *chunk_data) {
void **chunk_items = array_list_to_array(chunk_data);
Chunk *chunk = chunk_create((File **)chunk_items, chunk_data->size);
free(chunk_items);
chunk_data->item_destroyer = NULL;
array_list_delete(chunk_data);
return chunk;
}
static int open_next_directory(DirectoryScanner *scanner) {
if (scanner->current_dir) {
closedir(scanner->current_dir);
scanner->current_dir = NULL;
}
free(scanner->current_path);
if (queue_is_empty(scanner->directories))
return 0;
scanner->current_path = (char *)queue_dequeue(scanner->directories);
scanner->current_dir = opendir(scanner->current_path);
if (scanner->current_dir == NULL) {
perror("Could not open directory");
free(scanner->current_path);
scanner->current_path = NULL;
return 0;
}
return 1;
}
Chunk *directory_scanner_next(DirectoryScanner *scanner) {
ArrayList *chunk_data = array_list_create(file_destroy);
unsigned long long chunk_data_size = 0;
while (1) {
if (scanner->current_dir == NULL) {
if (!open_next_directory(scanner))
break;
}
struct dirent *entry = readdir(scanner->current_dir);
if (entry == NULL) {
closedir(scanner->current_dir);
scanner->current_dir = NULL;
free(scanner->current_path);
scanner->current_path = NULL;
continue;
}
if (strcmp(entry->d_name, ".") == 0 || strcmp(entry->d_name, "..") == 0)
continue;
char *cur_path = path_cat(scanner->current_path, entry->d_name);
struct stat stats;
if (stat(cur_path, &stats) != 0) {
free(cur_path);
continue;
}
if (S_ISDIR(stats.st_mode)) {
queue_enqueue(scanner->directories, (void *)cur_path);
} else {
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;
}
}
if ((scanner->max_size > 0 && (unsigned long long)stats.st_size > scanner->max_size) ||
(scanner->min_size > 0 && (unsigned long long)stats.st_size < scanner->min_size)) {
free(cur_path);
continue;
}
File *file = file_create(cur_path);
if (file == NULL) {
free(cur_path);
continue;
}
file->data->size = stats.st_size;
if (scanner->use_metadata)
file->metadata = file_metadata_create(&stats);
array_list_add(chunk_data, file);
chunk_data_size += file->data->size;
if (chunk_data_size > scanner->chunk_size) {
free(cur_path);
return chunk_data_to_chunk(chunk_data);
}
free(cur_path);
}
}
if (chunk_data->size > 0)
return chunk_data_to_chunk(chunk_data);
chunk_data->item_destroyer = NULL;
array_list_delete(chunk_data);
return NULL;
}