convert_ocr_result.py

# pdfs 下的识别结果转成 ocr 的格试 image/md5.txt json/md5.json ， 拷贝到这里 hubggingface/project/ocr/ocrServer/data ，就不用重新识别了

# see huggingface/project/ocr/tools/flask_auto_selection.py

import numpy as np
import cv2
import json
import base64
import os
import shutil
import hashlib
import glob
import re

def md5(fname):
    hash_md5 = hashlib.md5()
    with open(fname, "rb") as f:
        for chunk in iter(lambda: f.read(4096), b""):
            hash_md5.update(chunk)
    return hash_md5.hexdigest()

if __name__ == '__main__':

    dir_book = 'pdfs/jp/徐一平日本语句型辞典'

    dir = 'data'
    if os.path.exists(dir):
        shutil.rmtree(dir)
        os.makedirs(os.path.join(dir, 'img'))
        os.makedirs(os.path.join(dir, 'json'))
        os.makedirs(os.path.join(dir, os.path.basename(dir_book)))

    


    
    pths = glob.glob(dir_book + '/*.jpg', recursive=False)

    pths = sorted(pths, key=lambda p:(
        match := re.findall(r'.+?(\d+)\.jpg', p),
        number := int(match[0]),
        number # 如果第一项相等就会比较第二项
    ))

    for idx, pth in enumerate(pths):

        m5 = md5(pth)
        p_j = pth.replace('.jpg', '.json')
        if not os.path.exists(p_j):
            raise Exception(f'##### error: json not found. {p_j}')

        imgData = np.fromfile(pth, dtype=np.uint8)
        img = cv2.imdecode(imgData, cv2.IMREAD_UNCHANGED)
        img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
        # 把img 对象编码为jpg 格式
        success, encoded_image = cv2.imencode(".jpg", img)
        img_bytes = encoded_image.tobytes()
        base64_str = base64.b64encode(img_bytes).decode('ascii')

        img_bytes_restored = base64.b64decode(base64_str)
        imgdata_restored = np.frombuffer(img_bytes_restored, dtype=np.uint8) # .reshape(img.shape)
        img_restored = cv2.imdecode(imgdata_restored, cv2.IMREAD_UNCHANGED)


        pth_img = os.path.join(dir, 'img/{}.txt'.format(m5))
        pth_json = os.path.join(dir, 'json/{}.json'.format(m5))
        pth_book = os.path.join(dir, os.path.basename(dir_book))

        with open(pth_img, 'w', encoding='utf-8') as f:
            f.write(base64_str)

        dest = shutil.copy(p_j, pth_json)

        dest = shutil.copy(pth, pth_book)  

        if not os.path.exists(pth_img):  # 没有相应的图片，可能被删除了
            # raise Exception(f'Warnnig: no image {pth_img}')
            print( f'Warnnig: no image {pth_img}' )
            pass

        if not os.path.exists(pth_json):
            # raise Exception(f'Warnnig: no image {pth_json}')
            pass

        # with open(pth_img, "r", encoding="utf-8") as fp:
        #     imgdata = fp.read()
        #     imgdata = base64.b64decode(imgdata)
        #     imgdata = np.frombuffer(imgdata, np.uint8)
        #     img = cv2.imdecode(imgdata, cv2.IMREAD_UNCHANGED)

            # cv2.imshow('img', img)
            # cv2.waitKey(0)

        print( f'one task done. {idx+1} / {len(pths)}' )

        break

    print( 'all task done.' )