68 lines
2.3 KiB
Python
68 lines
2.3 KiB
Python
# python client_mock.py --pdf_path=/home/iriley/Documents/pdfs/unscanned/06.pdf --operations=table-parsing
|
|
import argparse
|
|
import json
|
|
from multiprocessing.sharedctypes import Value
|
|
import requests
|
|
|
|
from cv_analysis.utils.preprocessing import open_pdf
|
|
|
|
|
|
def parse_args():
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("--pdf_path", required=True, help="path to PDF file")
|
|
parser.add_argument(
|
|
"--first_page", type=int, required=False, default=0, help="page number from which to start (starts at 0)"
|
|
)
|
|
parser.add_argument(
|
|
"--last_page",
|
|
type=int,
|
|
required=False,
|
|
default=None,
|
|
help="page number at which to stop (non-inclusive); specify None to go to the end",
|
|
)
|
|
parser.add_argument(
|
|
"--operations",
|
|
type=str,
|
|
required=False,
|
|
help="Comma-separated list of operations, any of the following: \ntable-parsing\nredaction-detection\
|
|
\nfigure-detection\nlayout-detection",
|
|
default="table-parsing",
|
|
)
|
|
args = parser.parse_args()
|
|
|
|
return args
|
|
|
|
|
|
def main(args):
|
|
# files = {"name": (
|
|
# "name",
|
|
# open(args.pdf_path, "rb"),
|
|
# "file object corresponding to pdf file",
|
|
# {"operations": args.operations.split(",")}
|
|
# )
|
|
# }
|
|
operations = args.operations.split(",")
|
|
for operation in operations:
|
|
print("****************************")
|
|
print(f"{' '+operation+' ':^27}")
|
|
print("****************************")
|
|
if operation == "table-parsing":
|
|
response = requests.post("http://127.0.0.1:5000/tables", data=open(args.pdf_path, "rb"))
|
|
elif operation == "redaction-detection":
|
|
response = requests.post("http://127.0.0.1:5000/redactions", data=open(args.pdf_path, "rb"))
|
|
elif operation == "figure-detection":
|
|
response = requests.post("http://127.0.0.1:5000/figures", data=open(args.pdf_path, "rb"))
|
|
elif operation == "layout-parsing":
|
|
response = requests.post("http://127.0.0.1:5000/layout", data=open(args.pdf_path, "rb"))
|
|
else:
|
|
raise ValueError("{args.operation} is not a valid value.")
|
|
response.raise_for_status()
|
|
predictions = response.json()
|
|
|
|
print(json.dumps(predictions, indent=2))
|
|
|
|
|
|
if __name__ == "__main__":
|
|
args = parse_args()
|
|
main(args)
|