Notes

import webbrowser
import json
import boto3
import io
from io import BytesIO
from pdf2image import convert_from_path
from pprint import pprint

def pdf_to_image(pdf_path):
# Convert PDF to image using pdf2image
images = convert_from_path(pdf_path, fmt='png')
return images

def get_rows_columns_map(table_result, blocks_map):
rows = {}
scores = []
for relationship in table_result['Relationships']:
if relationship['Type'] == 'CHILD':
for child_id in relationship['Ids']:
cell = blocks_map[child_id]
if cell['BlockType'] == 'CELL':
row_index = cell['RowIndex']
col_index = cell['ColumnIndex']
if row_index not in rows:
# create new row
rows[row_index] = {}

# get confidence score
scores.append(str(cell['Confidence']))

# get the text value
rows[row_index][col_index] = get_text(cell, blocks_map)
return rows, scores

def get_text(result, blocks_map):
text = ''
if 'Relationships' in result:
for relationship in result['Relationships']:
if relationship['Type'] == 'CHILD':
for child_id in relationship['Ids']:
word = blocks_map[child_id]
if word['BlockType'] == 'WORD':
if "," in word['Text'] and word['Text'].replace(",", "").isnumeric():
text += '"' + word['Text'] + '"' + ' '
else:
text += word['Text'] + ' '
if word['BlockType'] == 'SELECTION_ELEMENT':
if word['SelectionStatus'] =='SELECTED':
text += 'X '
return text

def get_table_text_results(image_bytes):
session = boto3.Session(profile_name=None)
client = session.client('textract', region_name='us-east-2', aws_access_key_id="AKIATBS6HHWP4WL7Q6XI", aws_secret_access_key="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")

response = client.analyze_document(Document={'Bytes': image_bytes}, FeatureTypes=['TABLES'])

blocks = response['Blocks']
blocks_map = {}
table_blocks = []

for block in blocks:
blocks_map[block['Id']] = block
if block['BlockType'] == "TABLE":
table_blocks.append(block)

if len(table_blocks) <= 0:
return ""

text = ''
for index, table in enumerate(table_blocks):
rows, scores = get_rows_columns_map(table, blocks_map)
text += generate_table_text(rows, scores)
text += 'nn'

return text

def generate_table_text(rows, scores):
text = ''

for row_index, cols in rows.items():
for col_index, cell_text in cols.items():
text += f'{cell_text} '
text += 'n'

# text += 'nn Confidence Scores % (Table Cell) n'
# cols_count = 0
# for score in scores:
# cols_count += 1
# text += f'{score},'
# if cols_count == len(scores):
# text += 'n'
# cols_count = 0

text += 'nnn'

return text


def main(pdf_path, output_file='output.txt'):
images = pdf_to_image(pdf_path)

all_text = ''

for i, image in enumerate(images):
image_bytes = BytesIO()
image.save(image_bytes, format='PNG')
image_bytes = image_bytes.getvalue()

table_text = get_table_text_results(image_bytes)
table_text = table_text[:table_text.find('Confidence Scores % (Table Cell)')]
all_text += table_text

with open(output_file, "wt") as fout:
fout.write(all_text)

print('Text OUTPUT FILE: ', all_text)
return all_text

if __name__ == "__main__":
pdf_path = '/content/06-2023 3590.pdf'
main(pdf_path)