pocket-stats 0.2.1

Tools for the Pocket reading app https://app.getpocket.com/

Pocket Stats

A tool to analyze your Pocket reading list (https://app.getpocket.com/).

• Pocket Stats
  • Prerequisites
  • Installation
    • Pip
    • Git Clone
  • Usage
    • Data Querying
      • Fetch data from the Pocket server and cache it
      • Load cached data
      • Extract useful information
    • Visualization
      • Word Cloud
      • Article Count timeseries
      • Word Count distribution
      • Reading Speed & Reading Time
      • Top Domains
      • Language & Favorite
  • Test
  • Deployment
  • Contribute
  • Authors
  • License
  • Acknowledgments

Prerequisites

• A Pocket account.
• Pocket consumer key.
• Pocket access token. You can use fxneumann's OneClickPocket tool to automate this.
• Python 3.6+.

Installation

Pip

    pip install pocket-stats

Git Clone

    git clone https://github.com/nlbao/pocket_stats.git
    cd pocket_stats
    make setup

Usage

Set necessary environment variables:

    export POCKET_STATS_CONSUMER_KEY='<your_pocket_consumer_key>'
    export POCKET_STATS_ACCESS_TOKEN='<your_pocket_access_token>'

Data Querying

Fetch data from the Pocket server and cache it

Command line:

    # read only
    python -m pocket_stats fetch-data --offset 0 --limit 2

    # to write ALL the items to cache
    python -m pocket_stats fetch-data --overwrite_cache

Or in python code:

    from pocket_stats.data import fetch_data

    # fetch the first 100 items, won't overwrite cache
    data = fetch_data(offset=0, limit=100, overwrite_cache=False)

    # fetch ALL items, will overwrite cache
    data = fetch_data(ovewrite_cache=True)

The default location of cache file is ~/pocket-tools.cache, you can change it using the POCKET_STATS_CACHE_FILE environment variable.

Load cached data

    from pocket_stats.data import load_cache
    data = load_cache()

Extract useful information

    from pocket_stats.data import <your_function_names>

• Count word in all the titles:

    >>> count_words_in_title(data)
    Counter({'-': 5, '|': 5, 'python': 3, 'problem': 2, 'strace': 1, 'wow': 1, 'much': 1, 'syscall': 1, 'martin': 1, 'heinz': 1, 'personal': 1, 'website': 1, '&': 1, 'blog': 1, 'call': 1, 'programmer,': 1})

• Number of words in each article:

    >>> get_word_counts(data)
    [2207, 0, 5449, 4721, 3245, 805, 1849, 4087, 0, 538, 5054, 21, 866, 266, 1146, 213, 823, 3551, 787, 0]

• Reading time of each article:

    >>> get_reading_time(data)
    [9.80888888888889, 24.217777777777776, 20.982222222222223, 14.422222222222222, 3.577777777777778, 8.217777777777778, 18.164444444444445, 2.391111111111111, 22.462222222222223, 0.09333333333333334, 3.848888888888889, 1.1822222222222223, 5.093333333333334, 0.9466666666666667, 3.6577777777777776, 15.782222222222222, 3.497777777777778]

• Number of newly added articles per day:

    >>> get_added_time_series(data) 
                All articles
    2020-07-04             5
    2020-07-03             8
    2020-07-02             2
    2020-07-01             5

• Number of newly archived articles per day:

    >>> get_archived_time_series(data)
                Archived articles
    2020-07-04             2

• Number of articles per domain:

    >>> get_domain_counts(data)
    Counter({'kalzumeus.com': 3, 'bogleheads.org': 2, 'github.io': 2, 'brendangregg.com': 1, 'martinheinz.dev': 1, 'awealthofcommonsense.com': 1, 'jlcollinsnh.com': 1, 'callan.com': 1, 'engineerseekingfire.com': 1, 'arxiv.org': 1, 'popularmechanics.com': 1, 'dolpages.com': 1, 'economist.com': 1, 'romantomjak.com': 1, 'digitalocean.com': 1, 'deepnote.com': 1})

• Number of aritlces per language:

    >>> get_language_counts(data)
    Counter({'en': 17, 'unknown': 3})

• Number of favorite articles and its percent:

    >>> get_favorite_count(data)
    {'count': 2, 'percent': 0.1}

Visualization

Start the application (webserver) by running the commands below:

    python -m pocket_stats webapp

    # You will see something like this:
    # Dash is running on http://127.0.0.1:8050/

Other parameters:

    python -m pocket_stats webapp --help

    # Usage: __main__.py webapp [OPTIONS]
    # Options:
    # --debug         Debug mode
    # --port INTEGER  Port of the web server. Default = 8050.
    # --help          Show this message and exit.

Enter http://127.0.0.1:8050/ in your web browser. The site contains multiple components:

Word Cloud

Article Count timeseries

Word Count distribution

Stacked histograms

Reading Speed & Reading Time

Stacked histograms

Top Domains

Stacked bar charts.

Language & Favorite

Test

    make check

Deployment

You can deploy the app.py as a webserver.

Example: https://dash.plotly.com/deployment.

Contribute

Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.

Authors

• Bao Nguyen.
• contributors who participated in this project.

License

MIT License - see the LICENSE.md file for details.

Acknowledgments

• Pocket Python client: https://github.com/rakanalh/pocket-api
• fxneumann's OneClickPocket tool: http://reader.fxneumann.de/plugins/oneclickpocket/auth.php