根据url链接下载pdf，并提取图中文本

最新推荐文章于 2023-07-17 11:49:34 发布

有洁癖的懒羊羊

最新推荐文章于 2023-07-17 11:49:34 发布

阅读量374

点赞数

分类专栏： python基础文章标签： pdf python java

本文链接：https://blog.csdn.net/liyunyang2000/article/details/130709172

版权

python基础专栏收录该内容

9 篇文章 0 订阅

订阅专栏

该代码实现了一个从URL下载PDF文件，将其转换为PNG图像，然后使用pytesseract进行中文OCR识别的流程。文件名通过hashlib进行MD5编码以确保唯一性，整个过程涉及文件操作、网络请求和图像处理技术。

摘要由CSDN通过智能技术生成

# -*- coding: utf-8 -*-
import fitz
import hashlib
import requests as req
import pytesseract
from PIL import Image
import os

title = 'aaaa'
if not os.path.isdir('./xxxx_test'):
    os.mkdir('./xxxx_test')
files_path = './xxxx_test'
rename_title = hashlib.md5(title.encode('utf-8'))
title_md5 = rename_title.hexdigest()
the_path = os.path.join(files_path, title_md5)
the_path = f'{the_path}.pdf'


def download_file(link, the_path):
    """下载文件."""
    try:
        response = req.get(link, stream=True, timeout=500)
        chunk_size = 1024
        with open(the_path, "wb") as file:
            for data in response.iter_content(chunk_size=chunk_size):
                file.write(data)
                # count += len(data)
            return True
    except req.RequestException:
        print('下载文件失败')
        return False


def convert_img(the_path, png_dir):
    doc = fitz.open(the_path)
    pdf_name = os.path.splitext(the_path)[0]
    print(pdf_name)
    print(doc.pageCount)
    i = 0

    for pg in range(doc.pageCount):
        i = i + 1
        png_name = str(i)+".png"
        print(i)
        png_path = os.path.join(png_dir, png_name)
        page = doc[pg]
        rotate = int(0)
        zoom_x = 2.0
        zoom_y = 2.0
        trans = fitz.Matrix(zoom_x, zoom_y).prerotate(rotate)
        pm = page.get_pixmap(matrix=trans, alpha=False)
        pm.save(png_path)


def get_pdf_txt(png_dir):
    num = len(os.listdir(png_dir))
    text =''
    for i in range(num):
        i = i+1
        png_name = str(i) + ".png"
        png_path = os.path.join(png_dir, png_name)
        image = Image.open(png_path)
        # pytesseract.pytesseract.tesseract_cmd = 'xxxxxxxxxxxxxxxxx'
        text += pytesseract.image_to_string(image, lang='chi_sim')
        os.remove(png_name)
    print(text)


url = 'xxxxxxxx'
state = download_file(url, the_path)
if not os.path.isdir('./xxxx_png'):
    os.mkdir('./xxxx_png')
if not os.path.isdir(f'./xxxx_png/{title}'):
    os.mkdir(f'./xxxx_png/{title}')
png_dir = f'./xxxx_png/{title}'
convert_img(the_path, png_dir)
get_pdf_txt(png_dir)