Hunt flow pair extraction bug on master

CMakeLists.txt

@@ -45,37 +45,34 @@ if(NOT PCAP_FOUND)
     set(PCAP_INCLUDE_DIRS "")
 endif()
 
-# Find glog
-find_package(glog REQUIRED)
-
-# Source files
+# Source files for the original executable
 set(PKT2FLOW_SOURCES
-    src/pkt2flow.cpp
-    src/flow_db.cpp
-    src/utilities.cpp
+    src/pkt2flow.c
+    src/flow_db.c
+    src/utilities.c
+    src/main.c
 )
 
-# Create the main executable
+# Create the original executable
 add_executable(pkt2flow ${PKT2FLOW_SOURCES})
 
-# Link libraries
-target_link_libraries(pkt2flow 
+# Link libraries for both executables
+target_link_libraries(pkt2flow
     ${PCAP_LIBRARIES}
-    glog::glog
 )
 
-# Include directories
-target_include_directories(pkt2flow PRIVATE 
-    ${CMAKE_CURRENT_SOURCE_DIR}/src
+# Include directories for both executables
+target_include_directories(pkt2flow PRIVATE
+    ${CMAKE_CURRENT_SOURCE_DIR}/include
     ${PCAP_INCLUDE_DIRS}
 )
 
-# Set compile definitions
+# Set compile definitions for both executables
 target_compile_definitions(pkt2flow PRIVATE _GNU_SOURCE)
 
 # Installation
 include(GNUInstallDirs)
-install(TARGETS pkt2flow 
+install(TARGETS pkt2flow
     RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
 )
 
@@ -87,68 +84,94 @@ option(BUILD_TESTS "Build unit tests" ON)
 if(BUILD_TESTS)
     # Find GoogleTest
     find_package(GTest REQUIRED)
-
+    find_package(Glog REQUIRED)
+
     # Test source files
     set(TEST_SOURCES
         tests/test_flow_db.cpp
         tests/test_utilities.cpp
+        tests/test_pkt2flow_base.cpp
+        tests/test_pkt2flow_integration.cpp
+        tests/test_pkt2flow_tcp_integration.cpp
+        tests/test_pkt2flow_udp_integration.cpp
+        tests/test_pkt2flow_udplite_integration.cpp
         tests/test_main.cpp
     )
-    
+
     # Create test executable
     add_executable(pkt2flow_tests ${TEST_SOURCES})
-    
+
     # Link test libraries
     target_link_libraries(pkt2flow_tests
         ${PCAP_LIBRARIES}
         glog::glog
         GTest::gtest
         GTest::gtest_main
     )
-    
+
     # Include directories for tests
-    target_include_directories(pkt2flow_tests PRIVATE 
-        ${CMAKE_CURRENT_SOURCE_DIR}/src
+    target_include_directories(pkt2flow_tests PRIVATE
+        ${CMAKE_CURRENT_SOURCE_DIR}/include
         ${PCAP_INCLUDE_DIRS}
     )
-    
+
     # Set compile definitions for tests
     target_compile_definitions(pkt2flow_tests PRIVATE _GNU_SOURCE)
-    
+
     # Add test
     add_test(NAME pkt2flow_unit_tests COMMAND pkt2flow_tests)
-    
+
     # Create a library from the source files for testing
     add_library(pkt2flow_lib STATIC
-        src/flow_db.cpp
-        src/utilities.cpp
+        src/flow_db.c
+        src/utilities.c
     )
-    
-    target_include_directories(pkt2flow_lib PUBLIC 
-        ${CMAKE_CURRENT_SOURCE_DIR}/src
+
+    target_include_directories(pkt2flow_lib PUBLIC
+        ${CMAKE_CURRENT_SOURCE_DIR}/include
         ${PCAP_INCLUDE_DIRS}
     )
-    
-    target_link_libraries(pkt2flow_lib 
+
+    target_link_libraries(pkt2flow_lib
         ${PCAP_LIBRARIES}
         glog::glog
     )
-    
+
     target_compile_definitions(pkt2flow_lib PRIVATE _GNU_SOURCE)
-
-    # Link the library to tests
-    target_link_libraries(pkt2flow_tests pkt2flow_lib)
+
+    # Create a library that includes the main pkt2flow.c for testing (excluding main)
+    add_library(pkt2flow_test_lib STATIC
+        src/pkt2flow.c
+    )
+
+    # Exclude main function from the test library
+    target_compile_definitions(pkt2flow_test_lib PRIVATE _GNU_SOURCE TESTING_MODE)
+
+    target_include_directories(pkt2flow_test_lib PUBLIC
+        ${CMAKE_CURRENT_SOURCE_DIR}/include
+        ${PCAP_INCLUDE_DIRS}
+    )
+
+    target_link_libraries(pkt2flow_test_lib
+        ${PCAP_LIBRARIES}
+        glog::glog
+    )
+
+    target_compile_definitions(pkt2flow_test_lib PRIVATE _GNU_SOURCE)
+
+    # Link both libraries to tests
+    target_link_libraries(pkt2flow_tests pkt2flow_lib pkt2flow_test_lib)
 endif()
 
 # Build example in examples/
 add_executable(example_pkt2flow examples/example_pkt2flow.c)
 target_link_libraries(example_pkt2flow pkt2flow_lib)
-target_include_directories(example_pkt2flow PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/src)
+target_include_directories(example_pkt2flow PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/include)
 
 # CPack configuration for packaging
 include(CPack)
 set(CPACK_PACKAGE_NAME "pkt2flow")
 set(CPACK_PACKAGE_VERSION "${PROJECT_VERSION}")
 set(CPACK_PACKAGE_DESCRIPTION_SUMMARY "${PROJECT_DESCRIPTION}")
 set(CPACK_PACKAGE_VENDOR "chenxm")
-set(CPACK_PACKAGE_CONTACT "chenxm35@gmail.com")
+set(CPACK_PACKAGE_CONTACT "chenxm35@gmail.com")

README.md

@@ -81,19 +81,6 @@ Options:
 ```
 This will split all TCP and UDP flows from `input.pcap` into separate files in the `output_flows/` directory.
 
-## Logging
-
-Google glog is used for structured logging. Logs are written to:
-- Console (stderr) for important messages
-- Log files in `./logs/` for detailed debugging
-
-Control verbosity with environment variables:
-```bash
-export GLOG_v=2  # Increase verbosity
-export GLOG_log_dir=/custom/log/path
-./pkt2flow input.pcap
-```
-
 ## Development
 
 ### Running Tests

examples/example_pkt2flow.c

@@ -3,41 +3,41 @@
 #include <string.h>
 
 int main(void) {
-    // Initialize the hash table
-    init_hash_table();
+  // Initialize the hash table
+  init_hash_table();
 
-    // Create a 6-tuple for a flow (IPv4, TCP, src: 1.2.3.4:1234, dst: 5.6.7.8:80)
-    struct af_6tuple tuple;
-    memset(&tuple, 0, sizeof(tuple));
-    tuple.af_family = AF_INET;
-    tuple.protocol = 6; // TCP
-    tuple.ip1.v4.s_addr = 0x04030201; // 1.2.3.4
-    tuple.ip2.v4.s_addr = 0x08070605; // 5.6.7.8
-    tuple.port1 = 1234;
-    tuple.port2 = 80;
-    tuple.is_vlan = 0;
+  // Create a 6-tuple for a flow (IPv4, TCP, src: 1.2.3.4:1234, dst: 5.6.7.8:80)
+  struct af_6tuple tuple;
+  memset(&tuple, 0, sizeof(tuple));
+  tuple.af_family = AF_INET;
+  tuple.protocol = 6;               // TCP
+  tuple.ip1.v4.s_addr = 0x04030201; // 1.2.3.4
+  tuple.ip2.v4.s_addr = 0x08070605; // 5.6.7.8
+  tuple.port1 = 1234;
+  tuple.port2 = 80;
+  tuple.is_vlan = 0;
 
-    // Register the flow
-    struct ip_pair *pair = register_ip_pair(tuple);
-    if (pair) {
-        printf("Flow registered: src=%x:%u dst=%x:%u\n",
-            pair->af_6tuple.ip1.v4.s_addr, pair->af_6tuple.port1,
-            pair->af_6tuple.ip2.v4.s_addr, pair->af_6tuple.port2);
-    }
+  // Register the flow
+  struct ip_pair *pair = register_ip_pair(tuple);
+  if (pair) {
+    printf("Flow registered: src=%x:%u dst=%x:%u\n",
+           pair->af_6tuple.ip1.v4.s_addr, pair->af_6tuple.port1,
+           pair->af_6tuple.ip2.v4.s_addr, pair->af_6tuple.port2);
+  }
 
-    // Find the flow
-    struct ip_pair *found = find_ip_pair(tuple);
-    if (found) {
-        printf("Flow found!\n");
-    } else {
-        printf("Flow not found!\n");
-    }
+  // Find the flow
+  struct ip_pair *found = find_ip_pair(tuple);
+  if (found) {
+    printf("Flow found!\n");
+  } else {
+    printf("Flow not found!\n");
+  }
 
-    // Reset the flow's packet dump file
-    reset_pdf(&pair->pdf);
+  // Reset the flow's packet dump file
+  reset_pdf(&pair->pdf);
 
-    // Free the hash table
-    free_hash_table();
+  // Free the hash table
+  free_hash_table();
 
-    return 0;
-} 
+  return 0;
+}

format_code.sh

@@ -61,7 +61,7 @@ echo -e "${YELLOW}Using clang-format version: $(clang-format --version | head -n
 
 # Define source directories and file patterns
 SOURCE_DIRS=("src" "tests" "examples")
-FILE_PATTERNS=("*.cpp" "*.h" "*.hpp")
+FILE_PATTERNS=("*.cpp" "*.h" "*.hpp" "*.c")
 
 # Collect all files to format
 FILES_TO_FORMAT=()

src/pkt2flow.h → include/pkt2flow.h

@@ -1,7 +1,8 @@
 
 /* pkt2flow
+ * Xiaming Chen (chen_xm@sjtu.edu.cn)
  *
- * Copyright (c) 2012 Xiaming Chen <chen_xm@sjtu.edu.cn>
+ * Copyright (c) 2012
  * Copyright (c) 2014 Sven Eckelmann <sven@narfation.org>
  *
  * Permission is hereby granted, free of charge, to any person
@@ -30,32 +31,28 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-
 #ifndef PKT2FLOW_H
 #define PKT2FLOW_H
 
 #include <netinet/in.h>
 #include <netinet/ip6.h>
+#include <pcap/pcap.h>
 #include <stdint.h>
 
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-#define __SOURCE_VERSION__ "1.2"
-#define __AUTHOR__ "X. Chen (chenxm35@gmail.com)"
+#define __SOURCE_VERSION__ "1.4"
+#define __AUTHOR__ "Xiaming Chen (chenxm35@gmail.com)"
 #define __GLOBAL_NAME__ "pkt2flow"
 #define FLOW_TIMEOUT 1800 // seconds
 #define HASH_MULTIPLIER 37
 #define HASH_TBL_SIZE 48611
 
 #define BIT(bitnr) (1ULL << (bitnr))
-#ifndef PKT2FLOW_INLINE_ISSET_BITS
-#define PKT2FLOW_INLINE_ISSET_BITS
-static inline int isset_bits(uint64_t x, uint64_t bitmask) {
-  return (x & bitmask) == bitmask;
-}
-#endif
+#define isset_bits(x, bitmask)                                                 \
+  ({                                                                           \
+    uint32_t _bitmask = (uint32_t)(bitmask);                                   \
+    uint32_t _x = (uint32_t)(x);                                               \
+    (_bitmask & _x) == _bitmask;                                               \
+  })
 
 enum dump_allow_flags {
   DUMP_OTHER_ALLOWED = BIT(0),
@@ -73,6 +70,7 @@ enum pkt_dump_file_status {
 struct pkt_dump_file {
   char *file_name;
   unsigned long pkts;
+  pcap_dumper_t *dumper;
 
   enum pkt_dump_file_status status;
   unsigned long start_time;
@@ -106,6 +104,27 @@ struct ip_pair {
 /* pkt2flow.c */
 extern struct ip_pair *pairs[];
 
+/*
+ * Open the trace file
+ */
+void open_trace_file(void);
+
+/*
+ * Close the trace file
+ */
+void close_trace_files(void);
+
+/*
+ * Handle Ethernet packet processing
+ */
+int pcap_handle_ethernet(struct af_6tuple *af_6tuple,
+  const struct pcap_pkthdr *hdr, const u_char *pkt);
+
+/*
+* Create the full file path for a packet dump file
+*/
+char *resemble_file_path(struct pkt_dump_file *pdf);
+
 /* utilities.c */
 
 /*
@@ -149,8 +168,24 @@ struct ip_pair *register_ip_pair(struct af_6tuple af_6tuple);
  */
 void reset_pdf(struct pkt_dump_file *f);
 
-#ifdef __cplusplus
-}
-#endif
+/*
+ * Set the dump allowed flags for controlling which packet types to process
+ */
+void set_dump_allowed(uint32_t flags);
+
+/*
+ * Set the input file to process
+ */
+void set_readfile(const char *filename);
+
+/*
+ * Set the output directory
+ */
+void set_outputdir(const char *dir);
+
+/*
+ * Process the trace file
+ */
+void process_trace(void);
 
-#endif /* PKT2FLOW_H */
+#endif /* PKT2FLOW_H */