"Your own personal internet archive" (网站存档 / 爬虫)
ArchiveBox takes a list of website URLs you want to archive, and creates a local, static, browsable HTML clone of the content from those websites (it saves HTML, JS, media files, PDFs, images and more).
You can use it to preserve access to websites you care about by storing them locally offline. ArchiveBox imports lists of URLs, renders the pages in a headless, authenticated, user-scriptable browser, and then archives the content in multiple redundant common formats (HTML, PDF, PNG, WARC) that will last long after the originals disappear off the internet. It automatically extracts assets and media from pages and saves them in easily-accessible folders, with out-of-the-box support for extracting git repositories, audio, video, subtitles, images, PDFs, and more.
pip install archivebox mkdir data && cd data archivebox init archivebox add 'https://example.com' archivebox add --depth=1 'https://example.com/table-of-contents.html' echo 'any text with https://example.com URLs in it' | archivebox add # open the static data/index.html, or use the interactive web GUI: archivebox server 0.0.0.0:8000
After installing archivebox, just pass some new links to the
archivebox add command to start your collection.
ArchiveBox is written in Python 3.7 and uses wget, Chrome headless, youtube-dl, pywb, and other common UNIX tools to save each page you add in multiple redundant formats. It doesn’t require a constantly running server or backend (though it does include an optional one), just open the generated
data/index.html in a browser to view the archive or run
archivebox server to use the interactive Web UI. It can import and export links as JSON (among other formats), so it’s easy to script or hook up to other APIs. If you run it on a schedule and import from browser history or bookmarks regularly, you can sleep soundly knowing that the slice of the internet you care about will be automatically preserved in multiple, durable long-term formats that will be accessible for decades (or longer).
ArchiveBox is written in
python3.7 and has 4 main binary dependencies:
To get started, you can install them manually using your system’s package manager, use the automated helper script, or use the official Docker container. These dependencies are optional if disabled in settings.
# Docker mkdir data && cd data docker run -v $PWD:/data -it nikisweeting/archivebox init docker run -v $PWD:/data -it nikisweeting/archivebox add 'https://example.com' docker run -v $PWD:/data -it nikisweeting/archivebox manage createsuperuser docker run -v $PWD:/data -it -p 8000:8000 nikisweeting/archivebox server 0.0.0.0:8000 open http://127.0.0.1:8000
# Docker Compose # first download: https://github.com/pirate/ArchiveBox/blob/master/docker-compose.yml docker-compose run archivebox init docker-compose run archivebox add 'https://example.com' docker-compose run archivebox manage createsuperuser docker-compose up open http://127.0.0.1:8000
# Bare Metal # Use apt on Ubuntu/Debian, brew on mac, or pkg on BSD # You may need to add a ppa with a more recent version of nodejs apt install python3 python3-pip python3-dev git curl wget youtube-dl chromium-browser # Install Node + NPM curl -s https://deb.nodesource.com/gpgkey/nodesource.gpg.key | apt-key add - \ && echo 'deb https://deb.nodesource.com/node_14.x $(lsb_release -cs) main' >> /etc/apt/sources.list \ && apt-get update -qq \ && apt-get install -qq -y --no-install-recommends nodejs # Make a directory to hold your collection mkdir data && cd data # (doesn't have to be called data) # Install python package (or do this in a .venv if you want) pip install --upgrade archivebox # Install node packages (needed for SingleFile, Readability, and Puppeteer) npm install --prefix data 'git+https://github.com/pirate/ArchiveBox.git' archivebox init archivebox add 'https://example.com' # add URLs via args or stdin # or import an RSS/JSON/XML/TXT feed/list of links curl https://getpocket.com/users/USERNAME/feed/all | archivebox add archivebox add --depth=1 https://example.com/table-of-contents.html
Once you’ve added your first links, open
data/index.html in a browser to view the static archive.
You can also start it as a server with a full web UI to manage your links:
archivebox manage createsuperuser archivebox server
You can visit
http://127.0.0.1:8000 in your browser to access it.
Because modern websites are complicated and often rely on dynamic content, ArchiveBox archives the sites in several different formats beyond what public archiving services like Archive.org and Archive.is are capable of saving. Using multiple methods and the market-dominant browser to execute JS ensures we can save even the most complex, finicky websites in at least a few high-quality, long-term data formats.
ArchiveBox imports a list of URLs from stdin, remote URL, or file, then adds the pages to a local archive folder using wget to create a browsable HTML clone, youtube-dl to extract media, and a full instance of Chrome headless for PDF, Screenshot, and DOM dumps, and more…
archivebox add adds only new, unique links into your collection on each run. Because it will ignore duplicates and only archive each link the first time you add it, you can schedule it to run on a timer and re-import all your feeds multiple times a day. It will run quickly even if the feeds are large, because it’s only archiving the newest links since the last run. For each link, it runs through all the archive methods. Methods that fail will save
None and be automatically retried on the next run, methods that succeed save their output into the data folder and are never retried/overwritten by subsequent runs. Support for saving multiple snapshots of each site over time will be added soon (along with the ability to view diffs of the changes between runs).
All the archived links are stored by date bookmarked in
./archive/<timestamp>, and everything is indexed nicely with JSON & HTML files. The intent is for all the content to be viewable with common software in 50 - 100 years without needing to run ArchiveBox in a VM.
echo 'http://example.com' | archivebox add archivebox add 'https://example.com/some/page' archivebox add < ~/Downloads/firefox_bookmarks_export.html archivebox add < any_text_with_urls_in_it.txt archivebox add --depth=1 'https://example.com/some/downloads.html' archivebox add --depth=1 'https://news.ycombinator.com#2020-12-12'
See the Usage: CLI page for documentation and examples.
index.jsonHTML and JSON index files containing metadata and details
titletitle of the site
favicon.icofavicon of the site
example.com/page-name.htmlwget clone of the site, with .html appended if not present
warc/<timestamp>.gzgzipped WARC of all the resources fetched while archiving
output.pdfPrinted PDF of site using headless chrome
screenshot.png1440x900 screenshot of site using headless chrome
output.htmlDOM Dump of the HTML after rendering using headless chrome
archive.org.txtA link to the saved site on archive.org
media/all audio/video files + playlists, including subtitles & metadata with youtube-dl
git/clone of any repository found on github, bitbucket, or gitlab links
It does everything out-of-the-box by default, but you can disable or tweak individual archive methods via environment variables or config file.
If you’re importing URLs with secret tokens in them (e.g Google Docs, CodiMD notepads, etc), you may want to disable some of these methods to avoid leaking private URLs to 3rd party APIs during the archiving process. See the Security Overview page for more details.
Vast treasure troves of knowledge are lost every day on the internet to link rot. As a society, we have an imperative to preserve some important parts of that treasure, just like we preserve our books, paintings, and music in physical libraries long after the originals go out of print or fade into obscurity.
Whether it’s to resist censorship by saving articles before they get taken down or edited, or just to save a collection of early 2010’s flash games you love to play, having the tools to archive internet content enables to you save the stuff you care most about before it disappears.
The balance between the permanence and ephemeral nature of content on the internet is part of what makes it beautiful. I don’t think everything should be preserved in an automated fashion, making all content permanent and never removable, but I do think people should be able to decide for themselves and effectively archive specific content that they care about.
▶ Check out our community page for an index of web archiving initiatives and projects.
The aim of ArchiveBox is to go beyond what the Wayback Machine and other public archiving services can do, by adding a headless browser to replay sessions accurately, and by automatically extracting all the content in multiple redundant formats that will survive being passed down to historians and archivists through many generations.
ArchiveBox differentiates itself from similar projects by being a simple, one-shot CLI interface for users to ingest bulk feeds of URLs over extended periods, as opposed to being a backend service that ingests individual, manually-submitted URLs from a web UI. However, we also have the option to add urls via a web interface through our Django frontend.
Unlike crawler software that starts from a seed URL and works outwards, or public tools like Archive.org designed for users to manually submit links from the public internet, ArchiveBox tries to be a set-and-forget archiver suitable for archiving your entire browsing history, RSS feeds, or bookmarks,
including private/authenticated content that you wouldn’t otherwise share with a centralized service (do not do this until v0.5 is released with some security fixes). Also by having each user store their own content locally, we can save much larger portions of everyone’s browsing history than a shared centralized service would be able to handle.
Because ArchiveBox is designed to ingest a firehose of browser history and bookmark feeds to a local disk, it can be much more disk-space intensive than a centralized service like the Internet Archive or Archive.today. However, as storage space gets cheaper and compression improves, you should be able to use it continuously over the years without having to delete anything. In my experience, ArchiveBox uses about 5gb per 1000 articles, but your milage may vary depending on which options you have enabled and what types of sites you’re archiving. By default, it archives everything in as many formats as possible, meaning it takes more space than a using a single method, but more content is accurately replayable over extended periods of time. Storage requirements can be reduced by using a compressed/deduplicated filesystem like ZFS/BTRFS, or by setting
SAVE_MEDIA=False to skip audio & video files.
Whether you want to learn which organizations are the big players in the web archiving space, want to find a specific open-source tool for your web archiving need, or just want to see where archivists hang out online, our Community Wiki page serves as an index of the broader web archiving community. Check it out to learn about some of the coolest web archiving projects and communities on the web!
You can also access the docs locally by looking in the
You can build the docs by running:
cd ArchiveBox pipenv install --dev sphinx-apidoc -o docs archivebox cd docs/ make html # then open docs/_build/html/index.html