⚠️ This project is not actively maintained. ⚠️ EDGI no longer uses it for any ongoing work, and it depends on some no-longer-supported tools. Please let us know if you are using it to manage your data from Versionista.

web-monitoring-versionista-scraper

This is a version of versionista-outputter that has been rewritten in Node.js and JSDom.

Why? Speed is important here. Scraping Versionista can take a long time. We don’t need the overhead of a browser (like loading and executing images, CSS, and JavaScript) because all the necessary content is in the inital HTML payload. Parallelizing operations is also a little easier (for me, at least) in Node than in Ruby—and we absolutely ought to be doing more in parallel.

Installation

You’ll need Node.js. Then you should be able to globally install this with:

$ npm install -g https://github.com/edgi-govdata-archiving/web-monitoring-versionista-scraper.git

Then run it like so:

$ scrape-versionista --email EMAIL --password PASSWORD --after '2017-03-22' --format csv --output './scrape/versions.csv'

You can also split output into multiple files (by site) with the --group-by-site option:

$ scrape-versionista --email EMAIL --password PASSWORD --after '2017-03-22' --format csv --output './scrape/versions.csv' --group-by-site

Alternatively, you can clone this repo, then:

$ yarn install
# Or if you don't have yarn:
$ npm install

# And run it:
$ ./bin/scrape-versionista --email EMAIL --password PASSWORD --after '2017-03-22' --format csv --output './scrape/versions.csv'

Usage

This has the same basic capabilities as versionista-outputter, but can also save the versioned HTML (and diffs).

For basic info:

$ scrape-versionista --help

Options

• --email STRING Required! The E-mail address of Versionista Account. You can also use an env var instead: VERSIONISTA_EMAIL

• --password STRING Required! The password of Versionista Account. You can also use an env var instead: VERSIONISTA_PASSWORD

• --after DATE|HOURS Only check versions captured after this date. It can be an ISO 8601 date string like 2017-03-01T00:00:00Z or a number, representing hours before the current time.

• --before DATE|HOURS Only check versions captured before this date. It can be an ISO 8601 date string like 2017-03-01T00:00:00Z or a number, representing hours before the current time.

• --format FORMAT The output format. One of: csv, json, json-stream. [default: json]

• --output FILEPATH Write output to this file instead of directly to your console on stdout.

• --save-content If set, the raw HTML of each captured version will also be saved. Files are written to the working directory or, if --output is specified, the same directory as the output file.

• --save-diffs If set, the HTML of diffs between a version and its previous version will also be saved. Files are written to the working directory or, if --output is specified, the same directory as the output file.

• --latest-version-only If set, only the latest version (of the versions matching --after/--before times) for each page is captured.

• --group-by-site If set, a separate output file will be generated for each site. Files are placed in the same directory as --output, so the actual filename specified in --output will never be created.

Examples

ALL the options!

$ scrape-versionista --email '[email protected]' --password somepassword --after '2017-02-01' --before '2017-03-01' --format csv --output './scrape/versions.csv' --save-content --save-diffs

Use environment variables for credentials:

$ export VERSIONISTA_EMAIL='[email protected]'
$ export VERSIONISTA_PASSWORD=somepassword
$ scrape-versionista --after '2017-02-01' --before '2017-03-01' --format csv --output './scrape/versions.csv' --save-content --save-diffs

Specifying time as hours ago instead of a date:

# Starting 5 hours ago
$ scrape-versionista --after 5
# Decimals are accepted, so you can start 30 minutes ago, too
$ scrape-versionista --after 0.5

Other Scripts

The bin directory contains several other scripts besides scrape-versionista. They’re all closely related and perform helper tasks that are important in EDGI’s workflow around Versionista. You can use the --help option with all of them to see details about arguments, options, and usage.

• scrape-versionista-and-email runs scrape-versionista, then compresses the results into a single .tar.gz archive and e-mails them to a specified address.

• scrape-versionista-and-upload runs scrape-versionista, uploads the resulting files to Amazon S3 and Google Cloud Storage, and finally imports them into an instance of web-monitoring-db.

• upload-to-google uploads a directory’s contents to Google Cloud Storage. (Used as part of scrape-versionista-and-upload.)

• upload-to-s3 uploads a directory’s contents to Amazon S3. (Used as part of scrape-versionista-and-upload.)

• import-to-db sends the contents of a JSON-stream file listing versions that was generated by scrape-versionista to an instance of web-monitoring-db. (Used as part of scrape-versionista-and-upload.)

• query-db-and-email queries a web-monitoring-db instance for pages that were updated with new versions during a given time frame and e-mails a compressed .tar.gz archive of the results to a specified address. Results are CSV files — one per combination of tags specified with the --group-by option.

  NOTE: this will soon be deprecated in favor of web-monitoring-task-sheets.

• get-versionista-metadata and get-versionista-page-chunk are for advanced usage loading extremely large amounts of data from Versionista. See backfilling-data.md for usage instructions.

Deployment

For details about how this tool is deployed to automatically scrape Versionista in production, see deployment.md.

Code of Conduct

This repository falls under EDGI's Code of Conduct.

Contributing Guidelines

We love improvements to our tools! EDGI has general guidelines for contributing to all of our organizational repos.

License & Copyright

Copyright (C) 2017 Environmental Data and Governance Initiative (EDGI) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3.0.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

See the LICENSE file for details.