{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from aip import AipOcr\n",
    "\n",
    "\"\"\" 你的 APPID AK SK \"\"\"\n",
    "APP_ID = '17553946'\n",
    "API_KEY = 'IMC1ss3Tyo3vEAdVH6jgcdv2'\n",
    "SECRET_KEY = 'Iua6KdhuDtz159rjzGeFGpqgjhnoU0zZ'\n",
    "\n",
    "client = AipOcr(APP_ID, API_KEY, SECRET_KEY)\n",
    "def get_file_content(filePath):\n",
    "    with open(filePath, 'rb') as fp:\n",
    "        return fp.read()\n",
    "\n",
    "image = get_file_content('1.jpg')\n",
    "\n",
    "\"\"\" 调用通用文字识别, 图片参数为本地图片 \"\"\"\n",
    "#client.basicGeneral(image);\n",
    "\n",
    "\"\"\" 如果有可选参数 \"\"\"\n",
    "options = {}\n",
    "options[\"language_type\"] = \"CHN_ENG\"\n",
    "options[\"detect_direction\"] = \"true\"\n",
    "options[\"detect_language\"] = \"true\"\n",
    "options[\"probability\"] = \"true\"\n",
    "\n",
    "\"\"\" 带参数调用通用文字识别, 图片参数为本地图片 \"\"\"\n",
    "result= client.basicGeneral(image, options)\n",
    "if 'words_result' in result:\n",
    "        print('\\n'.join([w['words'] for w in result['words_result']]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 将指定目录下图片文件进行文字识别"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import os,sys\n",
    "from aip import AipOcr\n",
    "\n",
    "def get_access_token():\n",
    "    client_id = 'IMC1ss3Tyo3vEAdVH6jgcdv2'\n",
    "    client_secret = 'Iua6KdhuDtz159rjzGeFGpqgjhnoU0zZ'   \n",
    "     # client_id 为官网获取的AK， client_secret 为官网获取的SK\n",
    "    host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={}&client_secret={}'.format(\n",
    "        client_id, client_secret)\n",
    "    response = requests.get(host).text\n",
    "    data = json.loads(response)\n",
    "    access_token = data['access_token']\n",
    "    return access_token\n",
    "\n",
    "\"\"\" 你的 APPID AK SK \"\"\"\n",
    "APP_ID = '17553946'\n",
    "API_KEY = 'IMC1ss3Tyo3vEAdVH6jgcdv2'\n",
    "SECRET_KEY = 'Iua6KdhuDtz159rjzGeFGpqgjhnoU0zZ'\n",
    "\n",
    "client = AipOcr(APP_ID, API_KEY, SECRET_KEY)\n",
    "def get_file_content(filePath):\n",
    "    with open(filePath, 'rb') as fp:\n",
    "        return fp.read()\n",
    "\n",
    "options = {}\n",
    "options[\"language_type\"] = \"CHN_ENG\"\n",
    "options[\"detect_direction\"] = \"true\"\n",
    "options[\"detect_language\"] = \"true\"\n",
    "options[\"probability\"] = \"true\"\n",
    "\n",
    "fi_path = os.getcwd()+'/data'\n",
    "fl = os.listdir(fi_path)\n",
    "fl.sort()\n",
    "for fl1 in fl:\n",
    "    file_name = fi_path+'/' + fl1\n",
    "    image = get_file_content(file_name)\n",
    "    result= client.basicGeneral(image, options)\n",
    "    if 'words_result' in result:\n",
    "        print('\\n'.join([w['words'] for w in result['words_result']]))\n",
    "        print('\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 识别图片中表格"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-11-26T08:59:12.967127Z",
     "iopub.status.busy": "2020-11-26T08:59:12.966172Z",
     "iopub.status.idle": "2020-11-26T08:59:28.010488Z",
     "shell.execute_reply": "2020-11-26T08:59:28.006947Z",
     "shell.execute_reply.started": "2020-11-26T08:59:12.967019Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "22917135_2274033\n",
      "已完成\n"
     ]
    }
   ],
   "source": [
    "import requests\n",
    "import json\n",
    "import base64\n",
    "import time\n",
    "\n",
    "def get_access_token():\n",
    "    client_id = 'KwXkGawxh0sjOQdF9Ae9LeLb'\n",
    "    client_secret = 'siprEKMp5UcRTOAngEfIOOe9x6xkqGXq'   \n",
    "     # client_id 为官网获取的AK， client_secret 为官网获取的SK\n",
    "    host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={}&client_secret={}'.format(\n",
    "        client_id, client_secret)\n",
    "    response = requests.get(host).text\n",
    "    data = json.loads(response)\n",
    "    access_token = data['access_token']\n",
    "    return access_token\n",
    "\n",
    "def get_excel(requests_id, access_token):\n",
    "    headers = {'content-type': 'application/x-www-form-urlencoded'}\n",
    "    pargams = {\n",
    "        'request_id': requests_id,\n",
    "        'result_type': 'excel'\n",
    "    }\n",
    "    url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'\n",
    "    url_all = url + \"?access_token=\" + access_token\n",
    "    res = requests.post(url_all, headers=headers, params=pargams)#访问链接获取excel下载页\n",
    "    info_1 = res.json()['result']['ret_msg']\n",
    "    excel_url=res.json()['result']['result_data']\n",
    "    excel_1=requests.get(excel_url).content\n",
    "    with open('识别结果11.xls','wb+') as f:\n",
    "        f.write(excel_1)\n",
    "    print(info_1)\n",
    "\n",
    "\n",
    "request_url = \"https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/request\"\n",
    "# 二进制方式打开图片文件\n",
    "f = open('山东大学强基计划（2020）.jpg', 'rb')\n",
    "img = base64.b64encode(f.read())\n",
    "\n",
    "params = {\"image\":img}\n",
    "access_token = get_access_token()\n",
    "request_url = request_url + \"?access_token=\" + access_token\n",
    "headers = {'content-type': 'application/x-www-form-urlencoded'}\n",
    "response = requests.post(request_url, data=params, headers=headers)\n",
    "if response:\n",
    "    m_xx = response.json()\n",
    "requests_id = m_xx['result'][0]['request_id']  \n",
    "print(requests_id)\n",
    "time.sleep(10)\n",
    "get_excel(requests_id, access_token)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-01-03T00:39:37.591809Z",
     "iopub.status.busy": "2023-01-03T00:39:37.591257Z",
     "iopub.status.idle": "2023-01-03T00:39:37.763387Z",
     "shell.execute_reply": "2023-01-03T00:39:37.761195Z",
     "shell.execute_reply.started": "2023-01-03T00:39:37.591761Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'refresh_token': '25.e1167a22bcc63f2e736768f11eb893f9.315360000.1988066377.282335-17553946', 'expires_in': 2592000, 'session_key': '9mzdWr+jeUXdQXD/259SQnM94l/zWgp0GyY6hh17Kt+kjRLgEQG9OYATKN1tjVBrQIdg8DnKb0OBJpbiVsQ/7ixzXqxduw==', 'access_token': '24.acdee0098fe862a7c598f62121a2baac.2592000.1675298377.282335-17553946', 'scope': 'brain_ocr_meter brain_doc_analysis brain_ocr_webimage_loc vis-ocr_机动车购车发票识别 brain_ocr_vehicle_invoice brain_formula vis-ocr_行程单识别 brain_ocr_air_ticket public vis-ocr_ocr brain_ocr_scope brain_ocr_general brain_ocr_general_basic vis-ocr_business_license brain_ocr_webimage brain_all_scope brain_ocr_idcard brain_ocr_driving_license brain_ocr_vehicle_license vis-ocr_plate_number brain_solution brain_ocr_plate_number brain_ocr_accurate brain_ocr_accurate_basic brain_ocr_receipt brain_ocr_business_license brain_solution_iocr brain_qrcode brain_ocr_handwriting brain_ocr_passport brain_ocr_vat_invoice brain_numbers brain_ocr_business_card brain_ocr_train_ticket brain_ocr_taxi_receipt vis-ocr_household_register vis-ocr_vis-classify_birth_certificate vis-ocr_台湾通行证 vis-ocr_港澳通行证 vis-ocr_机动车检验合格证识别 vis-ocr_车辆vin码识别 vis-ocr_定额发票识别 vis-ocr_保单识别 brain_ocr_vin brain_ocr_quota_invoice brain_ocr_birth_certificate brain_ocr_household_register brain_ocr_HK_Macau_pass brain_ocr_taiwan_pass brain_ocr_vehicle_certificate brain_ocr_insurance_doc wise_adapt lebo_resource_base lightservice_public hetu_basic lightcms_map_poi kaidian_kaidian ApsMisTest_Test权限 vis-classify_flower lpq_开放 cop_helloScope ApsMis_fangdi_permission smartapp_snsapi_base smartapp_mapp_dev_manage iop_autocar oauth_tp_app smartapp_smart_game_openapi oauth_sessionkey smartapp_swanid_verify smartapp_opensource_openapi smartapp_opensource_recapi fake_face_detect_开放Scope vis-ocr_虚拟人物助理 idl-video_虚拟人物助理 smartapp_component smartapp_search_plugin avatar_video_test b2b_tp_openapi b2b_tp_openapi_online smartapp_gov_aladin_to_xcx', 'session_secret': '20a3564eae255945e2a4a45ecc57c608'}\n"
     ]
    }
   ],
   "source": [
    "import requests \n",
    "\n",
    "# client_id 为官网获取的AK， client_secret 为官网获取的SK\n",
    "host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id=IMC1ss3Tyo3vEAdVH6jgcdv2&client_secret=Iua6KdhuDtz159rjzGeFGpqgjhnoU0zZ'\n",
    "response = requests.get(host)\n",
    "if response:\n",
    "    print(response.json())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-11-03T03:24:08.412670Z",
     "iopub.status.busy": "2020-11-03T03:24:08.411768Z",
     "iopub.status.idle": "2020-11-03T03:24:08.813580Z",
     "shell.execute_reply": "2020-11-03T03:24:08.810985Z",
     "shell.execute_reply.started": "2020-11-03T03:24:08.412567Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "已完成\n"
     ]
    }
   ],
   "source": [
    "import requests\n",
    "import json\n",
    "import base64\n",
    "access_token = '24.277af0e5357bc3958c5b3cac49662a80.2592000.1606961184.282335-22917135'\n",
    "headers = {'content-type': 'application/x-www-form-urlencoded'}\n",
    "pargams = {\n",
    "    'request_id': '22917135_2227436',\n",
    "    'result_type': 'excel'\n",
    "}\n",
    "\n",
    "\n",
    "url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'\n",
    "url_all = url + \"?access_token=\" + access_token\n",
    "res = requests.post(url_all, headers=headers, params=pargams)#访问链接获取excel下载页\n",
    "info_1 = res.json()['result']['ret_msg']\n",
    "excel_url=res.json()['result']['result_data']\n",
    "excel_1=requests.get(excel_url).content\n",
    "with open('识别结果12.xls','wb+') as f:\n",
    "    f.write(excel_1)\n",
    "print(info_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-11-03T08:43:37.180693Z",
     "iopub.status.busy": "2020-11-03T08:43:37.179800Z",
     "iopub.status.idle": "2020-11-03T08:43:37.656055Z",
     "shell.execute_reply": "2020-11-03T08:43:37.653668Z",
     "shell.execute_reply.started": "2020-11-03T08:43:37.180589Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "list"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import requests\n",
    "import json\n",
    "import base64\n",
    "import demjson\n",
    "access_token = '24.277af0e5357bc3958c5b3cac49662a80.2592000.1606961184.282335-22917135'\n",
    "headers = {'content-type': 'application/x-www-form-urlencoded'}\n",
    "pargams = {\n",
    "    'request_id': '22917135_2227436',\n",
    "    'result_type': 'json'\n",
    "}\n",
    "\n",
    "\n",
    "url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'\n",
    "url_all = url + \"?access_token=\" + access_token\n",
    "res = requests.post(url_all, headers=headers, params=pargams)#访问链接获取excel下载页\n",
    "#info_1 = res.json()['result']['ret_msg']\n",
    "excel_1=res.json()['result']['result_data']#['forms'][0]['body']\n",
    "type(excel_1)\n",
    "#excel_new = demjson.decode(excel_1)\n",
    "#for m_col in excel_new['forms'][0]['body']:\n",
    "#    print(m_col)\n",
    "m_xx =json.loads(excel_1)\n",
    "\n",
    "#with open('识别结果12.json','w') as fl:\n",
    "#    json.dump(m_xx['forms'][0],fl,ensure_ascii=False)\n",
    "#print(info_1)\n",
    "#print(json.dumps(m_xx['forms'][0],ensure_ascii=False))\n",
    "print(m_xx['forms'][0]['body'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 识别保存为json文件"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-11-03T08:48:26.082167Z",
     "iopub.status.busy": "2020-11-03T08:48:26.081266Z",
     "iopub.status.idle": "2020-11-03T08:48:26.430039Z",
     "shell.execute_reply": "2020-11-03T08:48:26.428236Z",
     "shell.execute_reply.started": "2020-11-03T08:48:26.082060Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "已完成\n"
     ]
    }
   ],
   "source": [
    "import requests\n",
    "import json\n",
    "import base64\n",
    "import demjson\n",
    "access_token = '24.277af0e5357bc3958c5b3cac49662a80.2592000.1606961184.282335-22917135'\n",
    "headers = {'content-type': 'application/x-www-form-urlencoded'}\n",
    "pargams = {\n",
    "    'request_id': '22917135_2227436',\n",
    "    'result_type': 'json'\n",
    "}\n",
    "\n",
    "\n",
    "url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'\n",
    "url_all = url + \"?access_token=\" + access_token\n",
    "res = requests.post(url_all, headers=headers, params=pargams)#访问链接获取excel下载页\n",
    "#info_1 = res.json()['result']['ret_msg']\n",
    "excel_1=res.json()['result']['result_data']#['forms'][0]['body']\n",
    "type(excel_1)\n",
    "#excel_new = demjson.decode(excel_1)\n",
    "#for m_col in excel_new['forms'][0]['body']:\n",
    "#    print(m_col)\n",
    "m_xx =json.loads(excel_1)\n",
    "\n",
    "with open('识别结果12.json','w') as fl:\n",
    "    json.dump(m_xx['forms'][0],fl,ensure_ascii=False)\n",
    "print(info_1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  },
  "toc-autonumbering": true,
  "toc-showmarkdowntxt": false,
  "toc-showtags": false
 },
 "nbformat": 4,
 "nbformat_minor": 4
}