2023-10-16 20:26:01 +00:00
|
|
|
import sys
|
2023-11-06 19:44:05 +00:00
|
|
|
|
2023-10-16 20:26:01 +00:00
|
|
|
sys.path.append("../")
|
2023-10-26 18:40:06 +00:00
|
|
|
import src.alttext.alttext as alttext
|
2024-02-15 00:39:56 +00:00
|
|
|
from src.alttext.descengine.bliplocal import BlipLocal
|
2024-03-07 22:30:09 +00:00
|
|
|
from src.alttext.descengine.replicateapi import ReplicateAPI
|
2024-02-15 00:39:56 +00:00
|
|
|
from src.alttext.ocrengine.tesseract import Tesseract
|
|
|
|
from src.alttext.langengine.privategpt import PrivateGPT
|
2024-03-07 22:30:09 +00:00
|
|
|
from src.alttext.langengine.openaiapi import OpenAIAPI
|
2023-11-06 19:44:05 +00:00
|
|
|
import keys
|
2023-10-16 20:26:01 +00:00
|
|
|
|
2023-11-06 19:44:05 +00:00
|
|
|
# HTML BOOK FILEPATHS
|
2024-04-22 23:22:09 +00:00
|
|
|
HTML_ADVENTURES = "../books/pg76-h/pg76-images.html"
|
2023-11-06 19:44:05 +00:00
|
|
|
HTML_BIRD = "../books/pg30221-h/pg30221-images.html"
|
|
|
|
HTML_HUNTING = "../books/pg37122-h/pg37122-images.html"
|
|
|
|
HTML_MECHANIC = "../books/pg71856-h/pg71856-images.html"
|
|
|
|
HTML_INFINITY = "../books/pg71859-h/pg71859-images.html"
|
2023-10-26 18:40:06 +00:00
|
|
|
|
2023-11-06 19:44:05 +00:00
|
|
|
# EPUB BOOK FILEPATHS
|
2023-10-26 18:40:06 +00:00
|
|
|
EPUB1 = "../books/pg71856-images-3.epub"
|
|
|
|
EPUB2 = "../books/pg71908-images-3.epub"
|
|
|
|
EPUB3 = "../books/seuss.epub"
|
|
|
|
|
2023-11-27 23:58:51 +00:00
|
|
|
HOST1 = "http://127.0.0.1:8001"
|
|
|
|
|
2023-11-06 19:44:05 +00:00
|
|
|
|
2023-10-26 18:40:06 +00:00
|
|
|
def testHTML():
|
|
|
|
print("TESTING HTML")
|
2023-11-06 19:44:05 +00:00
|
|
|
alt: alttext.AltTextHTML = alttext.AltTextHTML(
|
2024-03-07 22:30:09 +00:00
|
|
|
# BlipLocal("C:/Users/dacru/Desktop/ALT/image-captioning"),
|
|
|
|
ReplicateAPI(keys.ReplicateEricKey()),
|
2024-02-15 00:39:56 +00:00
|
|
|
Tesseract(),
|
2024-03-07 22:30:09 +00:00
|
|
|
# PrivateGPT(HOST1),
|
|
|
|
OpenAIAPI(keys.OpenAIKey(), "gpt-3.5-turbo"),
|
2023-11-06 19:44:05 +00:00
|
|
|
)
|
2024-02-02 18:38:11 +00:00
|
|
|
|
2024-04-22 23:22:09 +00:00
|
|
|
# imgs = alt.getAllImgs()
|
|
|
|
|
|
|
|
alt.parseFile(HTML_ADVENTURES)
|
|
|
|
img = alt.getImg("images/c01-21.jpg")
|
|
|
|
src = img.attrs["src"]
|
|
|
|
imgData = alt.getImgData(src)
|
|
|
|
chars = alt.genChars(imgData, src)
|
|
|
|
desc = alt.genDesc(imgData, src, alt.getContext(img))
|
|
|
|
altText = alt.genAltText(src)
|
|
|
|
print(chars)
|
|
|
|
print("=====================================")
|
|
|
|
print(desc)
|
|
|
|
print("=====================================")
|
|
|
|
print(altText)
|
2023-11-29 02:17:35 +00:00
|
|
|
|
|
|
|
# desc = alt.genDesc(alt.getImgData(src), src)
|
|
|
|
# print(desc)
|
2024-02-02 18:38:11 +00:00
|
|
|
# associations = alt.genAltAssociations(imgs)
|
|
|
|
# print(associations)
|
2023-11-06 19:44:05 +00:00
|
|
|
|
2023-10-26 18:40:06 +00:00
|
|
|
|
|
|
|
if __name__ == "__main__":
|
2023-11-06 19:44:05 +00:00
|
|
|
testHTML()
|