Remove any local browsertrix-crawler files after the WACZ has been
copied to storage. Note, until this issue has a release on DockerHub the
local files won't be able to be deleted since Docker on Linux creates
the files as root:
https://github.com/webrecorder/browsertrix-crawler/issues/170
The code will catch this exception and log a warning instead of failing
and losing the work that has been completed.
This commit adds a browsertrix profile option to the configuration. In
order to not require the passing of the browsertrix config to every
Archiver, the Archiver constructors (include the base) were modified to
accept a Storage and Config instance. Some of the constructors them pick
out the pieces they need from the Config, in addition to calling the
parent constructor. In order to avoid a circular import that this
created the Config object now defines the default hash function to use,
rather than having it be a static property of the Archiver class.
The [browsertrix-crawler] utility is a browser-based crawler that can
crawl one or more pages. browsertrix-crawler creates archives in the
[WACZ] format which is essentially a standardized ZIP file (similar to DOCX, EPUB, JAR, etc) which can then be replayed using the [ReplayWeb.page] web
component, or unzipped to get the original WARC data (the ISO standard
format used by the Internet Archive Wayback Machine).
This PR adds browsertrix-crawler to archiver classes where screenshots are made made. The WACZ is uploaded to storage and then added to a new column in the spreadsheet. A column can be added that will display the WACZ, loaded from cloud storage (S3, digitalocean, etc) using the client side ReplayWeb page. You can see an example of the spreadsheet here:
https://docs.google.com/spreadsheets/d/1Tk-iJWzT9Sx2-YccuPttL9HcMdZEnhv_OR7Bc6tfeu8/edit#gid=0
browsertrix-crawler requires Docker to be installed. If Docker is not
installed an error message will be logged and things continue as normal.
[browsertrix-crawler]: https://github.com/webrecorder/browsertrix-crawler
[WACZ]: https://specs.webrecorder.net/wacz/latest/
[ReplayWeb.page]: https://replayweb.page
