From c294261a040f69c0a9efb1bb96e4c6ad98206214 Mon Sep 17 00:00:00 2001
From: Alexander-D-Karpov <alexandr.d.karpov@gmail.com>
Date: Mon, 27 Nov 2023 14:39:28 +0300
Subject: [PATCH] added extract_folder, updated podcasts

---
 extract_folder.py            | 54 ++++++++++++++++++++++++++++++++++++
 podcasts/yandex/compose.yaml |  3 ++
 2 files changed, 57 insertions(+)
 create mode 100644 extract_folder.py

diff --git a/extract_folder.py b/extract_folder.py
new file mode 100644
index 0000000..dbf32a8
--- /dev/null
+++ b/extract_folder.py
@@ -0,0 +1,54 @@
+import os
+import sys
+
+
+def structure_directory_content(input_dir, output_file=None, extensions=None):
+    """
+    This function goes through the input directory recursively and structures
+    all the file contents into one output file based on the given extensions.
+
+    :param input_dir: The input directory to search for files.
+    :param output_file: The output file where the content will be structured.
+                        If None, 'data.txt' or 'data.<extension>' will be used.
+    :param extensions: A list of file extensions to include. If None, all files are included.
+    """
+    if extensions:
+        extensions = [ext.strip() for ext in extensions.split(",") if ext.strip() != ""]
+        # If only one extension is given and no output file is specified, use 'data.<extension>'
+        if not output_file and len(extensions) == 1:
+            output_file = f"data.{extensions[0]}"
+    else:
+        extensions = None
+
+    # If no output file is specified, default to 'data.txt'
+    if not output_file:
+        output_file = "data.txt"
+
+    with open(output_file, "w") as outfile:
+        for root, dirs, files in os.walk(input_dir):
+            for file in files:
+                if extensions is None or any(
+                    file.endswith(f".{ext}") for ext in extensions
+                ):
+                    file_path = os.path.join(root, file)
+                    outfile.write(f"# {os.path.relpath(file_path, input_dir)}\n")
+                    with open(file_path, "r") as infile:
+                        outfile.write(infile.read())
+                        outfile.write("\n\n")
+
+
+if __name__ == "__main__":
+    if len(sys.argv) == 1:
+        input_directory = input("directory path: ")
+        output_filename = input("output file name (optional): ")
+        file_extensions = input("file extensions separated by commas (optional): ")
+        structure_directory_content(
+            input_directory,
+            output_filename if output_filename else None,
+            file_extensions if file_extensions else None,
+        )
+    else:
+        input_directory = sys.argv[1] if len(sys.argv) > 1 else "."
+        output_filename = sys.argv[2] if len(sys.argv) > 2 else None
+        file_extensions = sys.argv[3] if len(sys.argv) > 3 else None
+        structure_directory_content(input_directory, output_filename, file_extensions)
diff --git a/podcasts/yandex/compose.yaml b/podcasts/yandex/compose.yaml
index 81b27d2..cf66bcc 100644
--- a/podcasts/yandex/compose.yaml
+++ b/podcasts/yandex/compose.yaml
@@ -5,3 +5,6 @@ services:
     build: .
     env_file:
       - .env
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
+