jukebox-software/read-datasheet.py

28 lines
1.1 KiB
Python
Executable File

#!/usr/bin/env python3
# Parse Belden catalog techdata datasheets
import camelot
import numpy as np
def parse(filename):
tables = camelot.read_pdf(filename, pages="1-end", flavor='lattice', backend="poppler", split_text=False, line_scale=100, process_background=True, resolution=600, interations=1, layout_kwargs={'detect_vertical': False, 'char_margin': 0.5}, shift_text=['r', 't'])
print("Total tables extracted:", tables.n)
n = 0
for table in tables:
table.df.replace('', np.nan, inplace=True)
table.df.dropna(inplace=True, how="all")
table.df.dropna(inplace=True, axis="columns", how="all")
table.df.replace(np.nan, '', inplace=True)
if not table.df.empty:
table.to_html("table" + str(n) + ".html")
print("\nTable " + str(n))
print(table.df)
#camelot.plot(table, kind='grid').savefig("test" + str(n) + ".png")
n=n+1
camelot.plot(tables[0], kind='contour').savefig("test.png")
#tables.export('foo.csv', f='csv')
if __name__ == "__main__":
parse("test.pdf")