From 3035afc9482ce5f6f73df5cd13dc53ca0dd87bdb Mon Sep 17 00:00:00 2001
From: Baptiste Rajaut <rajaut@bayesia.com>
Date: Tue, 9 Apr 2024 17:35:13 +0200
Subject: [PATCH] feat: added a script launcher fix : the canary scripts now
 output their whole file path

---
 canary_launcher.py | 70 ++++++++++++++++++++++++++++++++++++++++++++++
 docx_canary.py     |  2 +-
 pptx_canary.py     |  2 +-
 xlsx_canary.py     |  2 +-
 4 files changed, 73 insertions(+), 3 deletions(-)
 create mode 100644 canary_launcher.py

diff --git a/canary_launcher.py b/canary_launcher.py
new file mode 100644
index 0000000..d7d19be
--- /dev/null
+++ b/canary_launcher.py
@@ -0,0 +1,70 @@
+import os
+import subprocess
+import json
+import tempfile
+import time
+
+# Mapping file extensions to corresponding Python scripts
+file_mapping = {
+    '.docx': 'docx_canary.py',
+    '.sql': 'mysql_canary.py',
+    '.pptx': 'pptx_canary.py',
+    '.xlsx': 'xlsx_canary.py',
+}
+
+def process_file(file_path, temp_json_path, generate_json):
+    _, extension = os.path.splitext(file_path)
+
+    if extension in file_mapping:
+        script_name = file_mapping[extension]
+        command = f"python {script_name} --input \"{file_path}\""
+        # Add JSON output option if JSON generation is requested
+        if generate_json:
+            temp_output = f"{temp_json_path}/{int(time.time() * 1000)}.json"
+            command += f" --json \"{temp_output}\""
+        print(f"Executing: {command}")
+        try:
+            subprocess.run(command, shell=True, check=True)
+        except subprocess.CalledProcessError as e:
+            print(f"Error executing {script_name} on {file_path}: {e}")
+
+def explore_directory(directory, temp_json_path, generate_json):
+    for root, _, files in os.walk(directory):
+        for file in files:
+            file_path = os.path.join(root, file)
+            process_file(file_path, temp_json_path, generate_json)
+
+if __name__ == "__main__":
+    import argparse
+
+    parser = argparse.ArgumentParser(description="Explore a directory and execute scripts based on file extensions.")
+    parser.add_argument('--directory', '-d', type=str, required=True, help="Path to the directory to explore.")
+    parser.add_argument('--json', '-j', type=str, help="Output path for the combined JSON file.")
+    args = parser.parse_args()
+
+    generate_json = bool(args.json)
+
+    if generate_json:
+        with tempfile.TemporaryDirectory() as tmp_dir:
+            explore_directory(args.directory, tmp_dir, generate_json)
+
+            # Combine all the JSON results into one file
+            combined_results = []
+            for root, dirs, files in os.walk(tmp_dir):
+                for file in files:
+                    if file.endswith('.json'):
+                        file_path = os.path.join(root, file)
+                        with open(file_path, 'r') as f:
+                            try:
+                                json_data = json.load(f)
+                                combined_results.extend(json_data if isinstance(json_data, list) else [json_data])
+                            except json.JSONDecodeError as e:
+                                print(f"Error decoding JSON from {file_path}: {e}")
+
+            with open(args.json, 'w') as f:
+                json.dump(combined_results, f, indent=4)
+
+            print(f"All results have been combined into {args.json}")
+    else:
+        # Just execute the scripts without generating JSON output
+        explore_directory(args.directory, None, generate_json)
diff --git a/docx_canary.py b/docx_canary.py
index 8693747..e9401de 100644
--- a/docx_canary.py
+++ b/docx_canary.py
@@ -81,7 +81,7 @@ def main():
             "md5": md5,
             "sha1": sha1
         }
-        write_to_json(args.json, args.input.split('/')[-1], data_to_export)
+        write_to_json(args.json, args.input, data_to_export)
         print(f"\nResults have been written to {args.json}")
 
 if __name__ == "__main__":
diff --git a/pptx_canary.py b/pptx_canary.py
index a729f71..e1d195f 100644
--- a/pptx_canary.py
+++ b/pptx_canary.py
@@ -80,7 +80,7 @@ def main():
             "md5": md5,
             "sha1": sha1
         }
-        write_to_json(args.json, args.input.split('/')[-1], data_to_export)
+        write_to_json(args.json, args.input, data_to_export)
         print(f"\nResults have been written to {args.json}")
 
 if __name__ == "__main__":
diff --git a/xlsx_canary.py b/xlsx_canary.py
index 72baf38..ee7fcd4 100644
--- a/xlsx_canary.py
+++ b/xlsx_canary.py
@@ -102,7 +102,7 @@ def main():
             "md5": md5,
             "sha1": sha1
         }
-        write_to_json(args.json, args.input.split('/')[-1], data_to_export)
+        write_to_json(args.json, args.input, data_to_export)
         print(f"\nResults have been written to {args.json}")
 
 if __name__ == "__main__":