调用http接口ocr返回文本信息提取json字段写入txt

import json
import requests
import cv2 as cv
import os
import re
import random
import glob as gb
import time

'''
(1)json.dumps()函数是将一个Python数据类型列表进行json格式的编码（可以这么理解，json.dumps()函数是将字典转化为字符串）
(2)json.loads()函数是将json格式数据转换为字典（可以这么理解，json.loads()函数是将字符串转化为字典）
'''


def post_requests(url, data, file_path):
    files = {'file': open(str(file_path), 'rb')}

    angle_result = requests.post(url, files=files, data=data)
    # angle_result_utf = angle_result.content.decode('utf-8')
    angle_result_utf = angle_result.text
    # print(json.loads(angle_result_utf))
    # print(angle_result_utf)
    return json.loads(angle_result_utf)

def get_images(root):
    '''
    get images's path and name
    '''
    files = []
    for ext in ['jpg', 'png', 'jpeg', 'JPG']:
        files.extend(gb.glob(os.path.join(root, '*.{}'.format(ext))))
    name = []
    for i in range(len(files)):
        name.append(files[i].split('/')[-1])
    return files, name

def img_path_parse(data_path):
    # class_names = os.listdir(data_path)
    img_path = []
    class_dir_path = os.path.join(data_path)
    # c. 获取子文件夹中的所有图像名称
    image_names = os.listdir(class_dir_path)
    # d. 遍历所有图像，分别提取图像特征信息
    for image_name in image_names:
        # 1. 获取图像的具体路径
        image_path = os.path.join(class_dir_path, image_name)
        #image_name = image_name.split('.')[-2].split('/')[-1]
        # 2. 基于图像路径获取图像的特征数据，要求格式为: [features,]
        img_path.append(image_path)
    random.shuffle(img_path)
    return img_path

def readFilename(path, allfile):
    filelist = os.listdir(path)

    for filename in filelist:
        filepath = os.path.join(path, filename)
        if os.path.isdir(filepath):
            readFilename(filepath, allfile)
        else:
            allfile.append(filepath)
    return allfile


def imglist(path):
    img = []
    label = []
    for file in os.listdir(path):
        file_img = os.path.join(path, file)
        if os.path.isdir(file_img):
            img = img + os.listdir(file_img)
    num = len(img)
    for i in range(num):
        imgs = img[i]
        #name1 = imgs.split(".")[-2]  # 获取后缀之前的元素
        name1 = imgs.split('.')[-2].split('/')[-1]
        name2 = name1.split('%')[-1]  # 获取标签
        label = label + [int(name2)]
    assert len(img) == len(label)


    #print("train img:", len(img), "......train label", len(label))


if __name__ == '__main__':
    #img_path = "D:\\.....\\test1.jpg"
    #img_path = "E:\\trade\\贸易合同工期页\\"

    data_path = "E:\\trade\\贸易合同价款页"
    image_path = img_path_parse(data_path)
    #print(image_path)









    # for fname in os.listdir(img_path):
    #     if '.jpg' in fname or '.JPG' in fname or '.png' in fname:
    #         img_path = os.path.join(img_path, fname)


    for name in image_path:
        angle_result_utf = post_requests("http://172.16.:9009/icr/......", {}, name)
        #print(img_path)


        allfile1 = []
        allfile1 = readFilename(data_path, allfile1)
        allname1 = []

        # print(name)
        name = name
        t = name.split(".")[0].split("\\")[-1]

        t1 = allname1.append(t)

        #print(t)

        i = 0

        # file_handle = open('./{}.txt'.format(t), mode='a', encoding = 'utf-8')
        with open('E:/trade/gts/贸易合同价款页/{}.txt'.format(t), mode='a', encoding='utf-8') as f:
            for t in angle_result_utf["document"]["blocks"]:
                data = angle_result_utf["document"]["blocks"][i]["lines"][0]["text"]
                boxes = angle_result_utf["document"]["blocks"][i]["lines"][0]["position"]["vertices"]
                box2 = ", ".join(repr(e) for e in boxes)

                # print(box2)
                box2 = re.sub("\D", ' ', box2)
                box2 = ','.join(x for x in box2.split() if x)

                strl = box2 + ',' + data
                i = i + 1

                f.write(strl + '\n')
                # print(strl)
            # print(i)
            # f.close()