I am a fresher and I am about to participate in a contest this weekend. The problem is about archiving and retrieving a large HTML dataset and I have no idea about it. My friend suggested to me to use a web archive and common crawl. Please suggest to me a way to convert the HTML dataset into a web archive and how to index them. Thanks in advance.
2 Answers
The WARC format is a widely used standard, definitely a good decisions to archive web pages. Also the HTTP headers are contained in the WARC file. As a consequence, you need a crawler to create a WARC file. If the HTML pages are provided as a collection of files, you would need to crawl the file system (ev. via a local HTTP server) to get the content into a WARC file.
Everything else depends on the concrete task: there are many tools and libraries
to crawl and export the content as WARC: the simplest is
wget --warc-file
but there are many moreto read WARC files and process the content.
See The WARC Ecosystem for a collection of tools. If you just need a serious WARC file to start with, fetch one from Common Crawl, e.g., https://commoncrawl.s3.amazonaws.com/crawl-data/CC-MAIN-2016-30/segments/1469257824853.47/warc/CC-MAIN-20160723071024-00101-ip-10-185-27-174.ec2.internal.warc.gz

- 2,049
- 10
- 10
You could use Heritrix
Crawler to crawl the websites you require. This can be automated via CURL
requests simply writing a Shell
Script.
Once you have crawled the websites, you can install OpenWayBack
, to 'play back' archived websites in your browser.
The OpenWayback
comes with a tool: CDX-Indexer
which could be used for indexing
the crawled Websites.
The current OpenWayBack
is not under development and you may use WaybackPY
for the playing of the WARCs.

- 11,530
- 2
- 71
- 51