-
Notifications
You must be signed in to change notification settings - Fork 0
/
manipulate_pdf.py
237 lines (192 loc) · 8.6 KB
/
manipulate_pdf.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
"""Simple script to manipulate images in a PDF file."""
import argparse
import base64
import io
import os
import fitz
import requests
from PIL import Image, ImageFilter, ImageFont, ImageDraw
def get_image_description(image_path, prompt, model, openai_key, max_tokens, is_verbose=False):
"""Gets an image description using OpenAI's API.
Args:
image_path (str): The path to the image.
prompt (str): The prompt to use.
model (str): The model to use.
openai_key (str): The OpenAI key. You can get one at https://platform.openai.com/api-keys
max_tokens (int): The maximum number of tokens to be used (per request, e.g. per image!)
is_verbose (bool): Whether to print debug information.
Returns:
str: The description of the image."""
# Check if the image path is valid
if not os.path.exists(image_path):
print(f"Error: The image path '{image_path}' does not exist.")
return
try:
# Encode the image to base64
with open(image_path, "rb") as image_file:
base64_image = base64.b64encode(image_file.read()).decode("utf-8")
# Make the request to OpenAI's API
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {openai_key}"
}
payload = {
"model": model,
"messages": [{
"role": "user",
"content": [
{"type": "text", "text": prompt},
{"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{base64_image}"}}
]
}],
"max_tokens": max_tokens
}
response = requests.post(url="https://api.openai.com/v1/chat/completions",
headers=headers,
json=payload)
# Get the description from the response
response = response.json()
if is_verbose:
print(f"get_image_description::Response: {response}")
description = response['choices'][0]['message']['content']
return description
except Exception as e:
print("get_image_description::Error:", e)
return ""
except KeyError:
print(f"get_image_description::Key Error in response")
return ""
def text_wrap(text, font, max_width):
"""Wrap text to fit a given width to print on an image.
Args:
text (str): The text to wrap.
font (PIL.ImageFont): The font to use.
max_width (int): The maximum width."""
if text is None or font is None or max_width is None:
print(f"text_wrap::Error: text, font, or max_width is None ({text}, {font}, {max_width})")
return []
lines = []
words = text.split(' ')
i = 0
line = ''
while i < len(words):
test_line = line + words[i] + ' '
bbox = font.getbbox(text)
text_width = bbox[2] - bbox[0]
if text_width > max_width:
lines.append(line.strip())
line = words[i] + ' '
else:
line = test_line
i += 1
lines.append(line.strip())
return lines
def extract_pdf(args):
"""Extracts images from a PDF file and applies some transformations to them.
Args:
args (argparse.Namespace): The arguments from the command line."""
# Create a temporary image path
orig_tmp_path = f"o_tmp.jpg"
tmp_image_path = f"tmp.jpg"
# Check if the OpenAI key is provided (only if the describe flag is set)
if args.describe:
if not args.openai_key:
print("Please provide an OpenAI key with the --openai-key flag.")
return
# Open the PDF file
pdf = args.pdf_file
doc = fitz.open(pdf)
# Iterate over the pages
for page in doc:
if args.verbose:
print(f"Processing page {page.number}")
# Get the images on the page and iterate over them
image_list = page.get_images(full=True)
for img_index, img in enumerate(image_list):
# Create an image to apply the transformations
base_image = doc.extract_image(img[0])
pil_img = Image.open(io.BytesIO(base_image["image"]))
# If the describe flag is set, get the description of the image
description = None
if args.describe:
if args.verbose:
print(f">> Describing image {img_index} on page {page.number}")
# Save the image to a temporary file
try:
pil_img.save(orig_tmp_path, format='JPEG')
except Exception as e:
print(f"extract_pdf::Error saving image to {orig_tmp_path}: {e}")
description = get_image_description(orig_tmp_path,
args.description_prompt,
"gpt-4-vision-preview",
args.openai_key,
args.max_openai_tokens,
is_verbose=args.verbose)
if os.path.exists(orig_tmp_path):
os.remove(orig_tmp_path)
if args.blur > 0:
pil_img = pil_img.filter(ImageFilter.GaussianBlur(args.blur))
if args.verbose:
print(f">> Blurring image {img_index} on page {page.number}")
if args.emboss:
pil_img = pil_img.filter(ImageFilter.EMBOSS)
if args.verbose:
print(f">> Embossing image {img_index} on page {page.number}")
if args.gray:
pil_img = pil_img.convert('L')
if args.verbose:
print(f">> Gray-scaling image {img_index} on page {page.number}")
if args.black:
pil_img = pil_img.convert('1')
if args.verbose:
print(f">> Blackening image {img_index} on page {page.number}")
if args.describe:
draw = ImageDraw.Draw(pil_img)
max_width = pil_img.width
font = ImageFont.truetype('arial.ttf', args.font_size)
lines = text_wrap(description, font, max_width)
y_text = 10
for line in lines:
bbox = font.getbbox(line)
text_height = bbox[3] - bbox[1]
draw.text((10, y_text), line, font=font, fill='white')
# Draw a shadow
draw.text((10 + 1, y_text + 1), line, font=font, fill='black')
y_text += text_height
try:
pil_img.save(tmp_image_path, format='JPEG')
except Exception as e:
print(e)
img_info = page.get_image_info()[img_index]
bbox = img_info['bbox']
page.insert_image(bbox, filename=tmp_image_path, keep_proportion=True)
if os.path.exists(tmp_image_path):
os.remove(tmp_image_path)
doc.save(args.output_file)
doc.close()
def main():
# Create the parser
parser = argparse.ArgumentParser(description='Change images in a PDF file.')
# Add arguments
parser.add_argument('pdf_file', type=str, help='Input')
parser.add_argument('-v', '--verbose', action='store_true', help='Verbose mode')
parser.add_argument('-o', '--output-file', type=str, help='Output file', default='output.pdf')
parser.add_argument('--blur', type=int, help="[0-50] Apply a blur effect to the images of the PDF.",
default=0)
parser.add_argument('--gray', action='store_true', help="Gray scale the images of the PDF")
parser.add_argument('--black', action='store_true', help="Blacken the images of the PDF")
parser.add_argument('--emboss', action='store_true',
help="Apply a emboss effect to the images of the PDF")
parser.add_argument('--describe', action='store_true',
help="Apply a description to the content of the PDF")
parser.add_argument('--openai-key', type=str, help="OpenAI key", default='')
parser.add_argument('--description-prompt', type=str, help="Prompt for the description",
default='Describe the image in less than 20 words. Include the number of people and objects.')
parser.add_argument('--max-openai-tokens', type=int, help="Max tokens", default=300)
parser.add_argument('--font-size', type=int, help="Font size", default=18)
# Parse the arguments
args = parser.parse_args()
# Perform the operation
extract_pdf(args)
if __name__ == "__main__":
main()