From 81eadd46720e384c0f598fba0f580eaee346a4f1 Mon Sep 17 00:00:00 2001 From: msramalho <19508417+msramalho@users.noreply.github.com> Date: Tue, 8 Nov 2022 14:22:13 +0000 Subject: [PATCH] disable browsertrix on docker, see #66 --- archivers/base_archiver.py | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/archivers/base_archiver.py b/archivers/base_archiver.py index 7037e4d..5ef2b7e 100644 --- a/archivers/base_archiver.py +++ b/archivers/base_archiver.py @@ -37,6 +37,7 @@ class Archiver(ABC): self.driver = config.webdriver self.hash_algorithm = config.hash_algorithm self.browsertrix = config.browsertrix_config + self.is_docker = config.is_docker def __str__(self): return self.__class__.__name__ @@ -206,6 +207,11 @@ class Archiver(ABC): if not self.browsertrix.enabled: logger.debug(f"Browsertrix WACZ generation is not enabled, skipping.") return + if self.is_docker: + # TODO: figure out support for browsertrix in docker + # see: https://github.com/bellingcat/auto-archiver/issues/66 + logger.warning(f"Browsertrix WACZ is not yet supported when using DOCKER.") + return logger.debug(f"getting wacz for {url}") key = self._get_key_from_url(url, ".wacz", append_datetime=True)