Switch to streamlit based web app
This commit is contained in:
James R. Barlow
+210
-77
@@ -14,94 +14,227 @@ Ghostscript's license as well as OCRmyPDF's.
|
||||
from __future__ import annotations
|
||||
|
||||
import os
|
||||
import shlex
|
||||
from subprocess import run
|
||||
from tempfile import TemporaryDirectory
|
||||
import subprocess
|
||||
import sys
|
||||
import time
|
||||
from functools import partial
|
||||
from operator import getitem
|
||||
from pathlib import Path
|
||||
from shutil import which
|
||||
from tempfile import NamedTemporaryFile
|
||||
|
||||
from flask import Flask, Response, request, send_from_directory
|
||||
from werkzeug.utils import secure_filename
|
||||
import pikepdf
|
||||
import streamlit as st
|
||||
from port_for import get_port
|
||||
from streamlit.components.v1 import iframe
|
||||
|
||||
app = Flask(__name__)
|
||||
app.secret_key = "secret"
|
||||
app.config['MAX_CONTENT_LENGTH'] = 50_000_000
|
||||
app.config.from_envvar("OCRMYPDF_WEBSERVICE_SETTINGS", silent=True)
|
||||
from ocrmypdf._defaults import DEFAULT_ROTATE_PAGES_THRESHOLD
|
||||
|
||||
ALLOWED_EXTENSIONS = {"pdf"}
|
||||
st.title("OCRmyPDF Web Service")
|
||||
|
||||
if not which("ttyd"):
|
||||
st.error("Missing dependency: ttyd. Please install ttyd to the local environment.")
|
||||
sys.exit(1)
|
||||
|
||||
uploaded = st.file_uploader("Upload input PDF or image", type=["pdf"], key="file")
|
||||
|
||||
mode = st.selectbox("Mode", options=["normal", "skip-text", "force-ocr", "redo-ocr"])
|
||||
|
||||
with st.expander("Input options"):
|
||||
invalidate_digital_signatures = st.checkbox(
|
||||
"Invalidate digital signatures", value=False
|
||||
)
|
||||
language = st.selectbox("Language", options=["eng", "deu", "fra", "spa"])
|
||||
|
||||
image_dpi = st.slider(
|
||||
"Image DPI", value=300, key="image_dpi", min_value=1, max_value=5000, step=50
|
||||
)
|
||||
with st.expander("Preprocessing"):
|
||||
skip_big = st.checkbox("Skip OCR on big pages", value=False, key="skip_big")
|
||||
oversample = st.slider("Oversample", min_value=0, max_value=5000, value=0, step=50)
|
||||
rotate_pages = st.checkbox("Rotate pages", value=False, key="rotate")
|
||||
deskew = st.checkbox("Deskew pages", value=False, key="deskew")
|
||||
clean = st.checkbox("Clean pages before OCR", value=False, key="clean")
|
||||
clean_final = st.checkbox("Clean final", value=False, key="clean_final")
|
||||
remove_vectors = st.checkbox("Remove vectors", value=False, key="remove_vectors")
|
||||
|
||||
|
||||
def allowed_file(filename):
|
||||
return "." in filename and filename.rsplit(".", 1)[1].lower() in ALLOWED_EXTENSIONS
|
||||
with st.expander("Output options"):
|
||||
output_type = st.selectbox(
|
||||
"Output type", options=["pdfa", "pdfa", "pdfa-1", "pdfa-2", "pdfa-3", "none"]
|
||||
)
|
||||
|
||||
pdf_renderer = st.selectbox(
|
||||
"PDF rendereer", options=["auto", "hocr", "hocrdebug", "sandwich"]
|
||||
)
|
||||
|
||||
optimize = st.selectbox("Optimize", options=["0", "1", "2", "3"])
|
||||
|
||||
st.selectbox("PDF/A compression", options=["auto", "jpeg", "lossless"])
|
||||
|
||||
with st.expander("Metadata"):
|
||||
title = author = keywords = subject = None
|
||||
if uploaded:
|
||||
with pikepdf.open(uploaded) as pdf, pdf.open_metadata() as meta:
|
||||
st.code(str(meta), language="xml")
|
||||
title = st.text_input("Title", value=meta.get('dc:title', ''))
|
||||
author = st.text_input("Author", value=meta.get('dc:creator', ''))
|
||||
keywords = st.text_input("Keywords", value=meta.get('dc:subject', ''))
|
||||
subject = st.text_input("Subject", value=meta.get('dc:description', ''))
|
||||
|
||||
|
||||
def do_ocrmypdf(file):
|
||||
uploaddir = TemporaryDirectory(prefix="ocrmypdf-upload")
|
||||
downloaddir = TemporaryDirectory(prefix="ocrmypdf-download")
|
||||
with st.expander("Optimization after OCR"):
|
||||
jpeg_quality = st.slider(
|
||||
"JPEG quality", min_value=0, max_value=100, value=75, key="jpeg_quality"
|
||||
)
|
||||
png_quality = st.slider(
|
||||
"PNG quality", min_value=0, max_value=100, value=75, key="png_quality"
|
||||
)
|
||||
jbig2_lossy = st.checkbox("JBIG2 lossy (dangerous)", value=False, key="jbig2_lossy")
|
||||
jbig2_threshold = st.number_input("JBIG2 threshold", value=0, key="jbig2_threshold")
|
||||
|
||||
filename = secure_filename(file.filename)
|
||||
up_file = os.path.join(uploaddir.name, filename)
|
||||
file.save(up_file)
|
||||
with st.expander("Advanced options"):
|
||||
jobs = st.slider(
|
||||
"Threads",
|
||||
min_value=1,
|
||||
max_value=os.cpu_count(),
|
||||
value=os.cpu_count(),
|
||||
key="threads",
|
||||
)
|
||||
pages = st.text_input(
|
||||
"Pages", value="", help="Comma-separated list of pages to process"
|
||||
)
|
||||
max_image_mpixels = st.number_input(
|
||||
"Max image size",
|
||||
value=250.0,
|
||||
min_value=0.0,
|
||||
help="Maximum image size in megapixels",
|
||||
)
|
||||
rotate_pages_threshold = st.number_input(
|
||||
"Rotate pages threshold",
|
||||
value=DEFAULT_ROTATE_PAGES_THRESHOLD,
|
||||
min_value=0.0,
|
||||
max_value=1000.0,
|
||||
help="Threshold for automatic page rotation",
|
||||
)
|
||||
fast_web_view = st.number_input(
|
||||
"Fast web view",
|
||||
value=1.0,
|
||||
min_value=0.0,
|
||||
help="Linearize files above this size in MB",
|
||||
)
|
||||
continue_on_soft_render_error = st.checkbox(
|
||||
"Continue on soft render error", value=True
|
||||
)
|
||||
verbose_labels = ["quiet", "default", "debug", "debug_all"]
|
||||
verbose = st.selectbox(
|
||||
"Verbosity level",
|
||||
options=[-1, 0, 1, 2],
|
||||
index=1,
|
||||
format_func=partial(getitem, verbose_labels),
|
||||
)
|
||||
|
||||
down_file = os.path.join(downloaddir.name, filename)
|
||||
if uploaded:
|
||||
args = []
|
||||
if mode and mode != 'normal':
|
||||
args.append(f"--{mode}")
|
||||
if language:
|
||||
args.append(f"--language={language}")
|
||||
if not uploaded.name.lower().endswith(".pdf") and image_dpi:
|
||||
args.append(f"--image-dpi={image_dpi}")
|
||||
if skip_big:
|
||||
args.append("--skip-big")
|
||||
if oversample:
|
||||
args.append(f"--oversample={oversample}")
|
||||
if rotate_pages:
|
||||
args.append("--rotate-pages")
|
||||
if deskew:
|
||||
args.append("--deskew")
|
||||
if clean:
|
||||
args.append("--clean")
|
||||
if clean_final:
|
||||
args.append("--clean-final")
|
||||
if remove_vectors:
|
||||
args.append("--remove-vectors")
|
||||
if output_type:
|
||||
args.append(f"--output-type={output_type}")
|
||||
if pdf_renderer:
|
||||
args.append(f"--pdf-renderer={pdf_renderer}")
|
||||
if optimize:
|
||||
args.append(f"--optimize={optimize}")
|
||||
if title:
|
||||
args.append(f"--title={title}")
|
||||
if author:
|
||||
args.append(f"--author={author}")
|
||||
if keywords:
|
||||
args.append(f"--keywords={keywords}")
|
||||
if subject:
|
||||
args.append(f"--subject={subject}")
|
||||
if pages:
|
||||
args.append(f"--pages={pages}")
|
||||
if max_image_mpixels:
|
||||
args.append(f"--max-image-mpixels={max_image_mpixels}")
|
||||
if rotate_pages_threshold:
|
||||
args.append(f"--rotate-pages-threshold={rotate_pages_threshold}")
|
||||
if fast_web_view:
|
||||
args.append(f"--fast-web-view={fast_web_view}")
|
||||
if continue_on_soft_render_error:
|
||||
args.append("--continue-on-soft-render-error")
|
||||
if verbose:
|
||||
args.append(f"--verbose={verbose}")
|
||||
if optimize > '0' and jpeg_quality:
|
||||
args.append(f"--jpeg-quality={jpeg_quality}")
|
||||
if optimize > '0' and png_quality:
|
||||
args.append(f"--png-quality={png_quality}")
|
||||
if jbig2_lossy:
|
||||
args.append("--jbig2-lossy")
|
||||
if jbig2_threshold:
|
||||
args.append(f"--jbig2-threshold={jbig2_threshold}")
|
||||
if jobs:
|
||||
args.append(f"--jobs={jobs}")
|
||||
input_file = NamedTemporaryFile(delete=True, suffix=f"_{uploaded.name}")
|
||||
input_file.write(uploaded.getvalue())
|
||||
input_file.flush()
|
||||
input_file.seek(0)
|
||||
args.append(str(input_file.name))
|
||||
output_file = NamedTemporaryFile(delete=True, suffix=".pdf")
|
||||
args.append(str(output_file.name))
|
||||
|
||||
cmd_args = [arg for arg in shlex.split(request.form["params"])]
|
||||
if "--sidecar" in cmd_args:
|
||||
return Response("--sidecar not supported", 501, mimetype='text/plain')
|
||||
st.session_state['running'] = (
|
||||
'run_button' in st.session_state and st.session_state.run_button
|
||||
)
|
||||
if st.button(
|
||||
"Run OCRmyPDF",
|
||||
disabled=st.session_state.get("running", False),
|
||||
key='run_button',
|
||||
):
|
||||
st.session_state['running'] = True
|
||||
args = [sys.executable, '-m', "ocrmypdf"] + args
|
||||
cmdline = " ".join(args)
|
||||
st.code(cmdline, language="bash")
|
||||
|
||||
ocrmypdf_args = ["ocrmypdf", *cmd_args, up_file, down_file]
|
||||
proc = run(ocrmypdf_args, capture_output=True, encoding="utf-8", check=False)
|
||||
if proc.returncode != 0:
|
||||
stderr = proc.stderr
|
||||
return Response(stderr, 400, mimetype='text/plain')
|
||||
port = get_port((5000, 7000))
|
||||
ttyd_args = ['ttyd', '--port', str(port), '--once', '--readonly']
|
||||
|
||||
return send_from_directory(downloaddir.name, filename)
|
||||
ttyd_proc = subprocess.Popen(
|
||||
ttyd_args + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE
|
||||
)
|
||||
iframe(src=f"http://localhost:{port}", height=400)
|
||||
|
||||
while ttyd_proc.poll() is None:
|
||||
ttyd_proc.poll()
|
||||
time.sleep(1)
|
||||
|
||||
@app.route("/", methods=["GET", "POST"])
|
||||
def upload_file():
|
||||
if request.method == "POST":
|
||||
if "file" not in request.files:
|
||||
return Response("No file in POST", 400, mimetype='text/plain')
|
||||
file = request.files["file"]
|
||||
if file.filename == "":
|
||||
return Response("Empty filename", 400, mimetype='text/plain')
|
||||
if not allowed_file(file.filename):
|
||||
return Response("Invalid filename", 400, mimetype='text/plain')
|
||||
if file and allowed_file(file.filename):
|
||||
return do_ocrmypdf(file)
|
||||
return Response("Some other problem", 400, mimetype='text/plain')
|
||||
|
||||
return """
|
||||
<!doctype html>
|
||||
<title>OCRmyPDF webservice</title>
|
||||
<h1>Upload a PDF (debug UI)</h1>
|
||||
<form method=post enctype=multipart/form-data>
|
||||
<label for="args">Command line parameters</label>
|
||||
<input type=textbox name=params>
|
||||
<label for="file">File to upload</label>
|
||||
<input type=file name=file>
|
||||
<input type=submit value=Upload>
|
||||
</form>
|
||||
<h4>Notice</h2>
|
||||
<div style="font-size: 70%; max-width: 34em;">
|
||||
<p>This is a webservice wrapper for OCRmyPDF.</p>
|
||||
<p>Copyright 2019 James R. Barlow</p>
|
||||
<p>This program is free software: you can redistribute it and/or modify
|
||||
it under the terms of the GNU Affero General Public License as published by
|
||||
the Free Software Foundation, either version 3 of the License, or
|
||||
(at your option) any later version.
|
||||
</p>
|
||||
<p>This program is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
GNU General Public License for more details.
|
||||
</p>
|
||||
<p>
|
||||
You should have received a copy of the GNU Affero General Public License
|
||||
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
||||
</p>
|
||||
</div>
|
||||
"""
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
app.run(host='0.0.0.0', port=5000)
|
||||
if ttyd_proc.returncode == 0:
|
||||
if Path(output_file.name).stat().st_size == 0:
|
||||
st.error("No output PDF file was generated")
|
||||
else:
|
||||
st.download_button(
|
||||
label="Download input PDF",
|
||||
data=input_file.read(),
|
||||
file_name=uploaded.name,
|
||||
mime="application/pdf",
|
||||
)
|
||||
else:
|
||||
st.error(f"ttyd failed with exit code {ttyd_proc.returncode}")
|
||||
st.session_state['running'] = False
|
||||
|
||||
Reference in New Issue
Block a user