sanspie/scripts
1
0
Fork 0
mirror of https://github.com/Alexander-D-Karpov/scripts.git synced 2025-06-30 09:33:04 +03:00
Code Issues Packages Projects Releases Wiki Activity

updated extract_folder script

Browse Source
This commit is contained in:
Alexander Karpov 2025-03-06 22:20:41 +03:00
parent cb6e37274c
commit 06cd5b7bea
1 changed files with 26 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
extract_folder.py
View File

@ -2,24 +2,38 @@ import os
import sys import sys
import pathspec import pathspec
def read_gitignore(input_dir): def read_gitignore(input_dir):
""" """
Reads the .gitignore file in the input directory and returns a pathspec object. Reads the .gitignore file in the input directory and returns a pathspec object.
""" """
# Default patterns to always ignore
default_patterns = [
"migrations/*.py", # Ignore Django migration files
"node_modules/", # Ignore node_modules directory
]
gitignore_path = os.path.join(input_dir, ".gitignore") gitignore_path = os.path.join(input_dir, ".gitignore")
if os.path.isfile(gitignore_path): if os.path.isfile(gitignore_path):
with open(gitignore_path, "r") as gitignore_file: with open(gitignore_path, "r") as gitignore_file:
return pathspec.PathSpec.from_lines("gitwildmatch", gitignore_file) patterns = default_patterns + list(gitignore_file)
return pathspec.PathSpec.from_lines("gitwildmatch", patterns)
else: else:
return pathspec.PathSpec.from_lines("gitwildmatch", []) return pathspec.PathSpec.from_lines("gitwildmatch", default_patterns)
def should_ignore_path(path, gitignore_spec):
"""
Additional check for paths that should be ignored.
"""
# Check if the path contains node_modules or migrations directory
if "node_modules" in path.split(os.sep) or \
("migrations" in path.split(os.sep) and path.endswith(".py")):
return True
return gitignore_spec.match_file(path)
def structure_directory_content(input_dir, output_file=None, extensions=None): def structure_directory_content(input_dir, output_file=None, extensions=None):
""" """
This function goes through the input directory recursively and structures This function goes through the input directory recursively and structures
all the file contents into one output file based on the given extensions. all the file contents into one output file based on the given extensions.
:param input_dir: The input directory to search for files. :param input_dir: The input directory to search for files.
:param output_file: The output file where the content will be structured. :param output_file: The output file where the content will be structured.
If None, 'data.txt' or 'data.<extension>' will be used. If None, 'data.txt' or 'data.<extension>' will be used.
@ -39,16 +53,18 @@ def structure_directory_content(input_dir, output_file=None, extensions=None):
with open(output_file, "w") as outfile: with open(output_file, "w") as outfile:
for root, dirs, files in os.walk(input_dir): for root, dirs, files in os.walk(input_dir):
# Filter files and directories using the enhanced ignore check
files = [ files = [
f f for f in files
for f in files if not should_ignore_path(os.path.join(root, str(f)), gitignore_spec)
if not gitignore_spec.match_file(os.path.join(root, str(f)))
] ]
# Filter directories in-place
dirs[:] = [ dirs[:] = [
d d for d in dirs
for d in dirs if not should_ignore_path(os.path.join(root, str(d)), gitignore_spec)
if not gitignore_spec.match_file(os.path.join(root, str(d)))
] ]
for file in files: for file in files:
if extensions is None or any( if extensions is None or any(
file.endswith(f".{ext}") for ext in extensions file.endswith(f".{ext}") for ext in extensions
@ -65,7 +81,6 @@ def structure_directory_content(input_dir, output_file=None, extensions=None):
except UnicodeDecodeError: except UnicodeDecodeError:
continue continue
if __name__ == "__main__": if __name__ == "__main__":
if len(sys.argv) == 1: if len(sys.argv) == 1:
input_directory = input("directory path: ") input_directory = input("directory path: ")