Create patent_downloader.py
Browse files- patent_downloader.py +131 -0
patent_downloader.py
ADDED
@@ -0,0 +1,131 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import List, Union, Optional
|
2 |
+
import os
|
3 |
+
import requests
|
4 |
+
import re
|
5 |
+
import time
|
6 |
+
import shutil
|
7 |
+
import subprocess
|
8 |
+
import pandas as pd
|
9 |
+
from selenium import webdriver
|
10 |
+
from selenium.webdriver.common.keys import Keys
|
11 |
+
from selenium.webdriver.chrome.service import Service
|
12 |
+
from selenium.webdriver.chrome.options import Options
|
13 |
+
from bs4 import BeautifulSoup
|
14 |
+
import chromedriver_autoinstaller
|
15 |
+
|
16 |
+
class PatentDownloader:
|
17 |
+
url = "https://patents.google.com"
|
18 |
+
|
19 |
+
def __init__(self, verbose: bool = False):
|
20 |
+
"""
|
21 |
+
Parameters
|
22 |
+
----------
|
23 |
+
verbose : bool
|
24 |
+
Print additional debug information.
|
25 |
+
"""
|
26 |
+
self.verbose = verbose
|
27 |
+
self.chrome_path = self.install_chrome()
|
28 |
+
|
29 |
+
def install_chrome(self) -> str:
|
30 |
+
"""
|
31 |
+
Download and install Google Chrome dynamically.
|
32 |
+
Returns
|
33 |
+
-------
|
34 |
+
str: Path to the Chrome binary.
|
35 |
+
"""
|
36 |
+
chrome_path = "/usr/bin/google-chrome"
|
37 |
+
|
38 |
+
if not shutil.which("google-chrome"):
|
39 |
+
print("Downloading and installing Google Chrome...")
|
40 |
+
subprocess.run(
|
41 |
+
"wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb -O chrome.deb",
|
42 |
+
shell=True,
|
43 |
+
check=True,
|
44 |
+
)
|
45 |
+
subprocess.run(
|
46 |
+
"apt-get update && apt-get install -y ./chrome.deb",
|
47 |
+
shell=True,
|
48 |
+
check=True,
|
49 |
+
)
|
50 |
+
os.remove("chrome.deb")
|
51 |
+
|
52 |
+
if not shutil.which("google-chrome"):
|
53 |
+
raise ValueError("Google Chrome installation failed!")
|
54 |
+
return chrome_path
|
55 |
+
|
56 |
+
def download(self, patent: Union[str, List[str]], output_path: str = "./",
|
57 |
+
waiting_time: int = 6, remove_kind_codes: Optional[List[str]] = None) -> None:
|
58 |
+
"""
|
59 |
+
Download patent document(s) as PDF.
|
60 |
+
"""
|
61 |
+
if isinstance(patent, list) or os.path.isfile(patent):
|
62 |
+
self.get_pdfs(patent, output_path, waiting_time, remove_kind_codes)
|
63 |
+
else:
|
64 |
+
self.get_pdf(patent, output_path, waiting_time, remove_kind_codes)
|
65 |
+
|
66 |
+
def get_pdf(self, patent: str, output_path: str = "./", waiting_time: int = 6,
|
67 |
+
remove_kind_codes: Optional[List[str]] = None) -> None:
|
68 |
+
"""
|
69 |
+
Download a single patent PDF.
|
70 |
+
"""
|
71 |
+
if remove_kind_codes:
|
72 |
+
for kind_code in remove_kind_codes:
|
73 |
+
patent = re.sub(kind_code + "$", "", patent)
|
74 |
+
|
75 |
+
# Automatically install ChromeDriver
|
76 |
+
chromedriver_autoinstaller.install()
|
77 |
+
|
78 |
+
# Set up Chrome options
|
79 |
+
chrome_options = Options()
|
80 |
+
chrome_options.binary_location = self.chrome_path
|
81 |
+
chrome_options.add_argument("--headless")
|
82 |
+
chrome_options.add_argument("--no-sandbox")
|
83 |
+
chrome_options.add_argument("--disable-dev-shm-usage")
|
84 |
+
|
85 |
+
# Initialize Selenium WebDriver
|
86 |
+
service = Service()
|
87 |
+
driver = webdriver.Chrome(service=service, options=chrome_options)
|
88 |
+
driver.get(self.url)
|
89 |
+
|
90 |
+
try:
|
91 |
+
# Search for the patent
|
92 |
+
element = driver.find_element("css selector", "input[type='search']")
|
93 |
+
|
94 |
+
element.send_keys(patent)
|
95 |
+
element.send_keys(Keys.RETURN)
|
96 |
+
time.sleep(waiting_time)
|
97 |
+
|
98 |
+
# Parse HTML and get the PDF link
|
99 |
+
soup = BeautifulSoup(driver.page_source, "html.parser")
|
100 |
+
pdf_link = self.get_pdf_link(soup, patent)
|
101 |
+
finally:
|
102 |
+
driver.quit()
|
103 |
+
|
104 |
+
# Download the PDF
|
105 |
+
if pdf_link:
|
106 |
+
validate_directory(output_path)
|
107 |
+
pdf_content = requests.get(pdf_link).content
|
108 |
+
with open(os.path.join(output_path, f"{patent}.pdf"), "wb") as file:
|
109 |
+
file.write(pdf_content)
|
110 |
+
print(f">>> Patent {patent} successfully downloaded <<<")
|
111 |
+
else:
|
112 |
+
print(f"Error: PDF link for patent {patent} not found!")
|
113 |
+
|
114 |
+
@staticmethod
|
115 |
+
def get_pdf_link(soup: BeautifulSoup, patent: str) -> Optional[str]:
|
116 |
+
"""
|
117 |
+
Extract the PDF link from parsed HTML.
|
118 |
+
"""
|
119 |
+
pdf_links = [link['href'] for link in soup.find_all('a', href=True) if link['href'].lower().endswith("pdf")]
|
120 |
+
for link in pdf_links:
|
121 |
+
if patent.lower() in link.lower():
|
122 |
+
return link
|
123 |
+
return None
|
124 |
+
|
125 |
+
|
126 |
+
def validate_directory(directory: str) -> None:
|
127 |
+
"""
|
128 |
+
Ensure the output directory exists.
|
129 |
+
"""
|
130 |
+
if not os.path.exists(directory):
|
131 |
+
os.makedirs(directory)
|