Zimit is a scraper allowing to create ZIM file from any Web site.
Zimit adheres to openZIM's Contribution Guidelines.
Zimit has implemented openZIM's Python bootstrap, conventions and policies v1.0.1.
While we would like to support as many websites as possible, making an offline archive of any website with a versatile tool obviously has some limitations.
Most capabilities and known limitations are documented in warc2zim README. There are also some limitations in Browsertrix Crawler (used to fetch the website) and wombat (used to properly replay dynamic web requests), but these are not (yet?) clearly documented.
Zimit runs a fully automated browser-based crawl of a website property and produces a ZIM of the crawled content. Zimit runs in a Docker container.
The system:
- runs a website crawl with Browsertrix Crawler, which produces WARC files
- converts the crawled WARC files to a single ZIM using warc2zim
The zimit.py
is the entrypoint for the system.
After the crawl is done, warc2zim is used to write a zim to the /output
directory, which should be mounted as a volume to not loose the ZIM created when container stops.
Using the --keep
flag, the crawled WARCs and few other artifacts will also be kept in a temp directory inside /output
zimit
is intended to be run in Docker. Docker image is published at https://github.com/orgs/openzim/packages/container/package/zimit.
The image accepts the following parameters, as well as any of the warc2zim ones; useful for setting metadata, for instance:
- Required:
--url URL
- the url to be crawled - Required:
--name
- Name of ZIM file --output
- output directory (defaults to/output
)--limit U
- Limit capture to at most U URLs--behaviors
- Control which browsertrix behaviors are ran (defaults toautoplay,autofetch,siteSpecific
, addingautoscroll
to the list is possible to automatically scroll the pages and fetch resources which are lazy loaded)--exclude <regex>
- skip URLs that match the regex from crawling. Can be specified multiple times. An example is--exclude="(\?q=|signup-landing\?|\?cid=)"
, where URLs that contain either?q=
orsignup-landing?
or?cid=
will be excluded.--workers N
- number of crawl workers to be run in parallel--wait-until
- Puppeteer setting for how long to wait for page load. See page.goto waitUntil options. The default isload
, but for static sites,--wait-until domcontentloaded
may be used to speed up the crawl (to avoid waiting for ads to load for example).--keep
- if set, keep the WARC files in a temp directory inside the output directory
Example command:
docker run ghcr.io/openzim/zimit zimit --help
docker run ghcr.io/openzim/zimit warc2zim --help
docker run -v /output:/output ghcr.io/openzim/zimit zimit --url URL --name myzimfile
Note: Image automatically filters out a large number of ads by using the 3 blocklists from anudeepND. If you don't want this filtering, disable the image's entrypoint in your container (docker run --entrypoint="" ghcr.io/openzim/zimit ...
).
To re-build the Docker image locally run:
docker build -t ghcr.io/openzim/zimit .
The Zimit contributor's team maintains a page with most Frequently Asked Questions.
While Zimit 1.x relied on a Service Worker to display the ZIM content, this is not anymore the case since Zimit 2.x which does not have any special requirements anymore.
It should also be noted that a first version of a generic HTTP scraper was created in 2016 during the Wikimania Esino Lario Hackathon.
That version is now considered outdated and archived in 2016
branch.