NFT_GIF_GENERATOOOR

Generate a collection of NFTs as GIFs👾 (In Progress..)

Local Setup 💻

To develop and test locally, a virtual environment will need to be created

$ pip3 install virtualenv
$ cd nft_gif_generatooor/
$ virtualenv env
$ source env/bin/activate

To install required packages

$ pip3 install requests

To see what packages are in your virtual environment

$ pip list

To deactivate

$ deactivate

File Structure 🌳

gif_layers

gif_layers
| -- background
| ----- background_1.gif
| ----- background_2.gif
| ----- background_3.gif
| -- body
| ----- body_1.gif
| ----- body_2.gif
| ----- body_3.gif
| ----- body_4.gif
| -- hair
| ----- hair_1.gif
| ----- hair_2
| -------- hair_2_front.gif
| -------- hair_2_back.gif
| ----- hair_3.gif
| -- face
| ----- face_1.gif
| ----- face_2.gif

• No spaces!
• Subdirectory names must match layer names in LAYERS_ORDER in order to be processed
• Nested folders within those subdirectories will be parsed for their GIF files as well (Ex: hair_2). This is done so that GIF layers that need to be built in a different order than LAYERS_ORDER are supported as well. See Config Constants & Options for more

How To Run 🏃‍♀️

1. Create a folder called 'gif_layers' and input all trait folders with their named attributes [follow syntax above].
2. From the nft_gif_generatooor directory, run the below command for a 1 time conversion of all GIF files into their separate PNG files required to produce the final GIFs. These files will be stored in the GIF_DECONSTRUCTED_LAYERS_PATH, which will be built upon the first run.

python3 src/convert_gif_to_png.py

3. From the same directory, once the frames of all GIFs have been parsed into PNG's & saved into their subdirectories, run the below command to build.

python3 src/main.py

Config Constants & Options 🎛

• NUMBER_OF_FRAMES - This is a REQUIRED constant. Each GIF should be using the same # of frames to generate in order to properly overlap the GIFs into the final NFT. An error will be thrown from the convert_gif_to_png.py script
• GENERATION_COUNT - The total number of NFTs that will be randomly generated
• SAVE_AS_MP4 - Toggle between saving the output file as a .gif or .mp4 using this boolean
• METADATA_NAME - Name of the collection that will be placed in the 'name' key in the metadata JSON object
• PARSE_UNDERSCORE - Boolean to turn ON if underscores should be replaced with spaces in attribute value names when being saved in the
• FRAME_RATE - FPS of the final output.

Metadata Format 📀

• name: METADATA_NAME,
• description: METADATA_DESCRIPTION,
• edition: Numeric value of NFT edition (0, 1, 2, 3, etc.),
• date: Date in UNIX,
• thumbnail": "",
• attributes : List of {attribute type: attribute value} objects.