17 lines
463 B
Python
17 lines
463 B
Python
|
import pytest
|
||
|
|
from pdf_parser import PdfParser
|
||
|
|
|
||
|
|
def test_pdf_parser_text_extraction():
|
||
|
|
pdf_parser = PdfParser('path_to_test_pdf')
|
||
|
|
text = pdf_parser.extract_text()
|
||
|
|
|
||
|
|
assert isinstance(text, str)
|
||
|
|
assert len(text) > 0
|
||
|
|
|
||
|
|
def test_pdf_parser_table_extraction():
|
||
|
|
pdf_parser = PdfParser('path_to_test_pdf')
|
||
|
|
tables = pdf_parser.extract_table()
|
||
|
|
|
||
|
|
assert isinstance(tables, list)
|
||
|
|
assert all(isinstance(table, pd.DataFrame) for table in tables)
|