wip

Author: Alex-PLACET

integration_tests/arrow_file_to_stream.cpp

@@ -4,16 +4,19 @@
 #include <iostream>
 #include <vector>
 
-#include "integration_tools.hpp"
+#include <sparrow_ipc/deserialize.hpp>
+#include <sparrow_ipc/memory_output_stream.hpp>
+#include <sparrow_ipc/serializer.hpp>
+#include "sparrow_ipc/stream_file_serializer.hpp"
 
 /**
- * @brief Reads a JSON file containing record batches and outputs the serialized Arrow IPC stream to a file.
+ * @brief Reads an Arrow IPC file and outputs the serialized Arrow IPC stream to a file.
  *
- * This program takes a JSON file path and an output file path as command-line arguments,
- * parses the record batches from the JSON data, serializes them into Arrow IPC stream format,
+ * This program takes an Arrow IPC file path and an output file path as command-line arguments,
+ * deserializes the record batches from the Arrow file, serializes them into Arrow IPC stream format,
  * and writes the binary stream to the specified output file.
  *
- * Usage: arrow_file_to_stream <json_file_path> <output_arrow_file>
+ * Usage: arrow_file_to_stream <arrow_file_path> <output_stream_file>
  *
  * @param argc Number of command-line arguments
  * @param argv Array of command-line arguments
@@ -23,18 +26,40 @@ int main(int argc, char* argv[])
 {
     if (argc != 3)
     {
-        std::cerr << "Usage: " << argv[0] << " <json_file_path> <output_arrow_file>\n";
-        std::cerr << "Reads a JSON file and outputs the serialized Arrow IPC stream to a file.\n";
+        std::cerr << "Usage: " << argv[0] << " <arrow_file_path> <output_stream_file>\n";
+        std::cerr << "Reads an Arrow IPC file and outputs the serialized Arrow IPC stream to a file.\n";
         return EXIT_FAILURE;
     }
 
-    const std::filesystem::path json_path(argv[1]);
+    const std::filesystem::path input_path(argv[1]);
     const std::filesystem::path output_path(argv[2]);
 
     try
     {
-        const std::vector<uint8_t> stream_data = integration_tools::json_file_to_stream(json_path);
+        // Read the Arrow file
+        std::ifstream input_file(input_path, std::ios::binary);
+        if (!input_file)
+        {
+            std::cerr << "Error: Could not open input file: " << input_path << "\n";
+            return EXIT_FAILURE;
+        }
+        
+        const std::vector<uint8_t> file_data(
+            (std::istreambuf_iterator<char>(input_file)),
+            (std::istreambuf_iterator<char>())
+        );
+        input_file.close();
+        
+        // Deserialize record batches from Arrow file format
+        auto batches = sparrow_ipc::deserialize_file(file_data);
+        
+        // Serialize to Arrow stream format
+        std::vector<uint8_t> stream_data;
+        sparrow_ipc::memory_output_stream mem_stream(stream_data);
+        sparrow_ipc::serializer serializer(mem_stream);
+        serializer << batches << sparrow_ipc::end_stream;
 
+        // Write to output file
         std::ofstream output_file(output_path, std::ios::binary);
         if (!output_file)
         {