pyfirefly's Introduction

pyfirefly

python library for reverse engineered Adobe Firefly API

Installing

pip install pyfirefly

Example Usage

Text to Image

The following example will request for 10 images of "flying cats" from Adobe Firefly and save those images.

import asyncio
import aiohttp
import aiofiles

import pyfirefly
from pyfirefly.utils import ImageOptions

token = 'ey...'  # replace with your bearer token

async def create_save_image(a, prompt, img_options, num):
    result = await a.text_to_image(prompt, **img_options)
    async with aiofiles.open(f'{num}.{result.ext}', mode='wb+') as f:
        await f.write(result.image)


async def demo(prompt, num):
    a = await pyfirefly.Firefly(token)
    img = ImageOptions(image_styles = a.image_styles)
    img.add_styles(['Photo', 'Blurry background', 'Origami'])

    tasks = [create_save_image(a, prompt, img.options, i) for i in range(num)]

    await asyncio.gather(*tasks)


asyncio.run(demo('flying cats', 10))

Glyph to Image

The following example will request for 10 images of a glyph.webp image that has "bundle of colorful electric wires" as the fill effect. example webp image

import asyncio
import aiohttp
import aiofiles

import pyfirefly
from pyfirefly.utils import ImageOptions

token = 'ey...'  # replace with your bearer token

async def create_save_image(a, glyph, img_options, num):
    result = await a.glyph_to_image(glyph, **img_options)
    async with aiofiles.open(f'{num}.{result.ext}', mode='wb+') as f:
        await f.write(result.image)


async def demo(input_fn, num):
    async with aiofiles.open('./glyph.webp', mode='rb') as f:
        glyph = await f.read()

    a = await pyfirefly.Firefly(token)
    img = ImageOptions(text_presets = a.text_presets)
    img.set_text_preset('wires')

    tasks = [create_save_image(a, glyph, img.options, i) for i in range(num)]

    await asyncio.gather(*tasks)


asyncio.run(demo('glyph.webp', 10))