From f62cb5fb43972f1ddf94560ec500c1318a73c9e6 Mon Sep 17 00:00:00 2001 From: Nick Sweeting Date: Tue, 7 May 2024 05:03:01 -0700 Subject: [PATCH] change wget to use stricter ascii filepath normalization --- archivebox/config.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/archivebox/config.py b/archivebox/config.py index 22da3700..758ea651 100644 --- a/archivebox/config.py +++ b/archivebox/config.py @@ -72,7 +72,7 @@ CONFIG_SCHEMA: Dict[str, ConfigDefaultDict] = { 'TIMEOUT': {'type': int, 'default': 60}, 'MEDIA_TIMEOUT': {'type': int, 'default': 3600}, 'OUTPUT_PERMISSIONS': {'type': str, 'default': '644'}, - 'RESTRICT_FILE_NAMES': {'type': str, 'default': 'windows'}, + 'RESTRICT_FILE_NAMES': {'type': str, 'default': 'ascii'}, # TODO: move this to be a default WGET_ARGS 'URL_DENYLIST': {'type': str, 'default': r'\.(css|js|otf|ttf|woff|woff2|gstatic\.com|googleapis\.com/css)(\?.*)?$', 'aliases': ('URL_BLACKLIST',)}, # to avoid downloading code assets as their own pages 'URL_ALLOWLIST': {'type': str, 'default': None, 'aliases': ('URL_WHITELIST',)},