LuisMayo/objection_engine

Objection! rendering engine 👨🏼‍⚖️

Code that allows you to convert chains of comments into ace attorney scenes. It's meant to be used by bots or other apps. List of users:

This is a fork of a wonderful Reddit bot

Getting Started

Prerequisites

Python 3. Python 3.8 and 3.9 and 3.10 have been proven to work.
FFMPEG instalation. In most Linux distros it should be available in the default package manager. In Windows systems it'd include downloading a pre-compiled zip folder, extracting it and adding the /bin folder into the system path

Optional

Google Translation API Credentials: These are needed for language support other than English. Other languages may work even without this credentials as the system will fallback to TextBlob's translation system.
Libraqm: Improves text rendering on right-to-left languages. In windows refer to the faq.md

Installing

Clone the repository

git clone https://github.com/LuisMayo/objection_engine

Install dependencies (in case any problems are encountered please check faq.md) You can use either pip

python -m pip install .

Or poetry

poetry install

(optional) In case you want language support outside English install polyglot and its dependencies: (if on windows check faq.md)

python3 -m pip install pyICU pycld2 morfessor polyglot

Check the exmaple python example.py

Using it as a library

Add this git repository as a library: pip install objection_engine

Import it into your python file

import objection_engine
# You can also import the components like this
from objection_engine.renderer import render_comment_list
from objection_engine.beans.comment import Comment

Create a list of comments

foo = [objection_engine.comment.Comment(), objection_engine.comment.Comment(text_content='Second comment',  user_name="Second user")]

Render the list
```
objection_engine.renderer.render_comment_list(foo)
```
For a list of arguments to the class and method check both https://github.com/LuisMayo/objection_engine/blob/main/renderer.py and https://github.com/LuisMayo/objection_engine/blob/main/beans/comment.py

There is a complete example in https://github.com/LuisMayo/objection-engine-testing

Rendering a video using Docker

cp example.py docker-entrypoint.py

docker build --tag objection-engine .
docker run --rm \
  --volume $(pwd)/docker-entrypoint.py:/app/entrypoint.py:ro \
  --volume $(pwd)/assets:/app/assets \
  --volume $(pwd)/outputs:/app/outputs \
  objection-engine

The video will be in the /outputs directory.

You can download Polyglot models by setting oe_polyglot_models environment variable and preserve the data by mounting /root/polyglot_data:

docker run --rm \
  --volume $(pwd)/docker-entrypoint.py:/app/entrypoint.py:ro \
  --volume $(pwd)/assets:/app/assets \
  --volume $(pwd)/outputs:/app/outputs \
  --volume $(pwd)/polyglot_data:/root/polyglot_data \
  --env oe_polyglot_models="de fr" \
  objection-engine

Settings

The following environment variables are honored by objection_engine:

oe_bypass_sentiment: If on any value other than the empty string, the sentiment analysis is bypassed
oe_sentiment_model: If "polyglot", polyglot will be used for sentiment analisis, if empty or "hf" it would be a huggingface model. When using polyglot models remember to do python -m polyglot download TASK:sentiment2
~~oe_stats_server~~ (not longer supported as of Objection engine 3.3.0): If present, it will be used as the URL to post statistics to. The server responsible should be similar to https://github.com/LuisMayo/simple-server-counter
oe_polyglot_models: (docker only) If on polyglot model(s), the data for the model will be downloaded when starting the container.
OE_DIRECT_H264_ENCODING: If "true" then it will directly encode the videos in H264 with OpenCV, instead of encoding in mp4v and re-encoding later. This is faster than the default, but it requires running on windows or having a self-compiled OpenCV build in your system. If you don't know what any of this means, don't enable it.
Contributing

Since this is a tiny project we don't have strict rules about contributions. Just open a Pull Request to fix any of the project issues or any improvement you have percieved on your own. Any contributions which improve or fix the project will be accepted as long as they don't deviate too much from the project objectives. If you have doubts about whether the PR would be accepted or not you can open an issue before coding to ask for my opinion.