Movie-Database

Gathers information about movies and TV series stored in a local directory and compiles it into static searchable HTML documents.

Demo

Script output loaded with some randomly chosen sample data is available here. I've tried to include several foreign titles to ensure that they're tagged correctly. Commit used for these reports is now significantly behind this repo.

Screenshots

As of 20 August 2017

List of Content

Movie Details

Usage

Download the repository here, set the appropriate permissions and launch Media-Database via:

$ git clone https://github.com/at1as/Media-Database.git
$ cd Media-Database
$ vim conf.json
$ sudo python3 run.py

Titles will be fetched from IMDB the first time the script is run, and then be saved in _data/movie_data.json, which the templates are built from. On subsequent runs, only newly added items to the directory will have their data fetched from IMDB.

Dry Run

Run with the --dry-run flag to rebuild site from an existing json dump in _data/movie_data.json, without adding new entries or needing access to the physical files.

$ python run --dry-run

Environment Configuration

Set your environment in conf.json:

• max_quantity [Integer] => the maximum number of entries to retrieve from IMDB. Play nice; ramp this number up slowly as you build your local database
• asset_location [String] => the absolute path to the folder containing the files
• exclude_files [Array of Strings] => Exclude these files from retrieval (even if their extensions are included)
• include_extensions [Array of Strings] => files will these extensions (or folders) will be retrieved from IMDB. These should all be lower case

Removing Entries

For items tagged incorrectly, remove them using the bundled script

$ python scripts/remove_entry.py --movie "<movie_title>" "<YYYY>"
$ python scripts/remove_entry.py --series "<series_title>" "<YYYY>"

Year is optional. If not provided, first entry with a matching name will removed.

See $ python scripts/remove_entry.py --help for more details

Playback

Some movie listing pages will show an embedded video player. This currently only works in Safari with H264 content which is not a subdirectory

Apache

As the generated output is static content, it can easily be set up for access to generated files across an internal network. Symlink the Apache Documents directory to the root of this repo (Media-Database), and then access the page at /Media-Database/_output/index.html.

Usage Notes

General Notes

• To view the site in your browser : $ open _output/index.html
• All dependencies are listed in setup.py
• Script will only search files in the folder specified by asset_location. Will not search subdirectories (see below for details)
• Only works for movies, television and documentaries (documentaries will be listed under the movies section, anime will be listed under Series)

Movie Asset Locations

Movies should be titled as Inception (2010).mp4. If multiple files exist (split movie or separate subtitle files), this format is also acceptable:

Inception (2010)/
|── Inception (2010).mp4
└── Inception (2010).srt

These pathes are relative to movie location set it conf.json

Series Asset Locations

• Series should be titled by the name and year, such as Firefly (2002) where the year is the start date of the series
• Subtitles should be stored in the same directory as the episodes and should have the exact same title with a different extension
• Episodes may optionally contain titles after the series title and episodes number

Firefly (2002)/
└── Firefly Season 1
    └── Firefly S01E01.mp4
    └── Firefly S01E01.srt
    └── Firefly S01E02 The Train Job.mp4
    └── Firefly S01E02 The Train Job.srt

These pathes are relative to series location set it conf.json

Filters

To use the search filters, items should be comma separated and can be negated with a preceding "!"

For example, in the case of the language input:

• "English, French, German" => Returns movies with languages matching English, French, and German
• "English, !Klingon, Welsh" => Returns movies with languages matching both English and Welsh, but filters out any matches with the presence of Klingon
• "Ger, P" => Returns movies with languages matching Ger... (i.e., German), P... (i.e., Portuguese, Persian, etc)

Disclaimer

• This product works by scraping the IMDB website. Therefore, problems may arise due to even minor changes to the website layout
• Retrieval is only as good as your movie titles, and IMDB search. Meticulously named files will virtually always be correct, but there may still be some misses
• Media-Database was written and tested on Python 2.7.10 on macOS 10.11

TODO

• FIXME: locally stored glyphicons won't render in Firefox unless they're stored in a child directory
• Add support for other clients (themoviedb, etc)
• Search by Alternative Title
• Improvements to Scraper. Currently small schema changes break the scraper. There are some small changes to the XPaths being used to make this slightly more robust (see comments in src/scrapers/imdb.py)