Txt to json

script/benchmark/benchmarkTxtToJson.py

@@ -0,0 +1,104 @@
+import json
+import hashlib
+import re
+import sys
+import os
+
+def parse_benchmark_file(input_file):
+    with open(input_file, 'r') as f:
+        data = f.read()
+
+    sections = data.split('AcceleratorType:')
+    results = []
+
+    for section in sections:
+        section = section.strip()
+        if not section:
+            continue
+
+        # Extract metadata information
+        meta = {
+            "creator": "unknown",
+            "datetime": "unknown",
+            "hostname": "unknown",
+            "moreinterestingmetadata": "unknown"
+        }
+
+        env = {}
+        workdivs = {}
+
+        # Extract properties
+        for line in section.splitlines():
+            line = line.strip()
+            if not line:
+                continue
+
+            if re.match(r'^[A-Za-z]', line) and ':' in line:
+                key, value = map(str.strip, line.split(':', 1))
+                if key.startswith("WorkDiv"):
+                    # Store workdivs in a dictionary for later assignment to kernels
+                    kernel_name = key.replace("WorkDiv", "").strip()
+                    workdivs[kernel_name] = value
+                else:
+                    env[key] = value
+
+        # Add the kind information
+        env["kind"] = "Babelstream"
+
+        # Extract kernel data
+        kernels_section = re.search(r'Kernels\s+Bandwidths\(GB/s\).+', section, re.DOTALL)
+        if kernels_section:
+            kernel_lines = kernels_section.group(0).splitlines()[1:]
+            for kernel_line in kernel_lines:
+                parts = kernel_line.split()
+                if len(parts) >= 6:
+                    kernel_name = parts[0].replace("Kernel", "")  # Remove "Kernel" from the name if present
+                    kernel_data = {
+                        "name": kernel_name,
+                        "Bandwidth(GB/s)": parts[1],
+                        "MinTime(s)": parts[2],
+                        "MaxTime(s)": parts[3],
+                        "AvgTime(s)": parts[4],
+                        "DataUsage(MB)": parts[5],
+                        "WorkDiv": workdivs.get(kernel_name, ""),  # Fetch the correct workdiv
+                        "AcceleratorType": env.get("AcceleratorType", ""),
+                        "NumberOfRuns": env.get("NumberOfRuns", ""),
+                        "Precision": env.get("Precision", ""),
+                        "DataSize(items)": env.get("DataSize(items)", ""),
+                        "DeviceName": env.get("DeviceName", ""),
+                        "kind": env.get("kind", "")
+                    }
+
+                    # Generate a unique ID for each kernel data point and rename it to 'id_test'
+                    kernel_id = hashlib.md5(json.dumps(kernel_data, sort_keys=True).encode('utf-8')).hexdigest()
+                    kernel_data["id_test"] = kernel_id
+
+                    results.append({
+                        "id_test": kernel_id,
+                        "meta": meta,
+                        "env": env,
+                        **kernel_data
+                    })
+
+    return results
+
+def convert_to_json(input_file, output_file):
+    data = parse_benchmark_file(input_file)
+    with open(output_file, 'w') as f:
+        json.dump(data, f, indent=4)
+
+if __name__ == "__main__":
+    if len(sys.argv) != 3:
+        print("Usage: python script.py <input_file> <output_file>")
+        sys.exit(1)
+
+    input_file = sys.argv[1]
+    output_file = sys.argv[2]
+
+    if not os.path.isfile(input_file):
+        print(f"Error: The input file '{input_file}' does not exist.")
+        sys.exit(1)
+
+    print(f"Processing input file '{input_file}' to output file '{output_file}'")
+    convert_to_json(input_file, output_file)
+    print(f"Conversion completed. JSON saved to '{output_file}'")