{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append(\"../../FinNLP\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SEC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from finnlp.data_sources.company_announcement.sec import SEC_Announcement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "start_date = \"2020-01-01\"\n",
    "end_date = \"2020-06-01\"\n",
    "stock = \"AAPL\"\n",
    "config = {\n",
    "    \"use_proxy\": \"us_free\",\n",
    "    \"max_retry\": 5,\n",
    "    \"proxy_pages\": 3,\n",
    "}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Checking ips: 100%|██████████| 45/45 [01:42<00:00,  2.28s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Get proxy ips: 45.\n",
      "Usable proxy ips: 44.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading by item...: 100%|██████████| 39/39 [01:39<00:00,  2.54s/it]\n"
     ]
    }
   ],
   "source": [
    "downloader = SEC_Announcement(config)\n",
    "downloader.download_date_range_stock(start_date, end_date, stock = stock)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>_id</th>\n",
       "      <th>ciks</th>\n",
       "      <th>period_ending</th>\n",
       "      <th>root_form</th>\n",
       "      <th>file_num</th>\n",
       "      <th>display_names</th>\n",
       "      <th>xsl</th>\n",
       "      <th>sequence</th>\n",
       "      <th>file_date</th>\n",
       "      <th>biz_states</th>\n",
       "      <th>sics</th>\n",
       "      <th>form</th>\n",
       "      <th>adsh</th>\n",
       "      <th>film_num</th>\n",
       "      <th>biz_locations</th>\n",
       "      <th>file_type</th>\n",
       "      <th>file_description</th>\n",
       "      <th>inc_states</th>\n",
       "      <th>ite</th>\n",
       "      <th>content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000320193-20-000056:wf-form4_158932261319105.xml</td>\n",
       "      <td>[0001631982, 0000320193]</td>\n",
       "      <td>2020-05-08</td>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...</td>\n",
       "      <td>xslF345X03</td>\n",
       "      <td>1</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>[]</td>\n",
       "      <td>[3571]</td>\n",
       "      <td>4</td>\n",
       "      <td>0000320193-20-000056</td>\n",
       "      <td>[]</td>\n",
       "      <td>[, ]</td>\n",
       "      <td>4</td>\n",
       "      <td>FORM 4</td>\n",
       "      <td>[, CA, ]</td>\n",
       "      <td>[]</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0000320193-20-000054:wf-form4_158829658358801.xml</td>\n",
       "      <td>[0001051401, 0000320193]</td>\n",
       "      <td>2020-04-28</td>\n",
       "      <td>4</td>\n",
       "      <td>[001-36743]</td>\n",
       "      <td>[JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...</td>\n",
       "      <td>xslF345X03</td>\n",
       "      <td>1</td>\n",
       "      <td>2020-04-30</td>\n",
       "      <td>[CA]</td>\n",
       "      <td>[3571]</td>\n",
       "      <td>4</td>\n",
       "      <td>0000320193-20-000054</td>\n",
       "      <td>[20838087]</td>\n",
       "      <td>[, Cupertino, CA]</td>\n",
       "      <td>4</td>\n",
       "      <td>FORM 4</td>\n",
       "      <td>[, CA]</td>\n",
       "      <td>[]</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 _id  \\\n",
       "0  0000320193-20-000056:wf-form4_158932261319105.xml   \n",
       "1  0000320193-20-000054:wf-form4_158829658358801.xml   \n",
       "\n",
       "                       ciks period_ending root_form     file_num  \\\n",
       "0  [0001631982, 0000320193]    2020-05-08         4           []   \n",
       "1  [0001051401, 0000320193]    2020-04-28         4  [001-36743]   \n",
       "\n",
       "                                       display_names         xsl sequence  \\\n",
       "0  [KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...  xslF345X03        1   \n",
       "1  [JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...  xslF345X03        1   \n",
       "\n",
       "    file_date biz_states    sics form                  adsh    film_num  \\\n",
       "0  2020-05-12         []  [3571]    4  0000320193-20-000056          []   \n",
       "1  2020-04-30       [CA]  [3571]    4  0000320193-20-000054  [20838087]   \n",
       "\n",
       "       biz_locations file_type file_description inc_states ite  \\\n",
       "0               [, ]         4           FORM 4   [, CA, ]  []   \n",
       "1  [, Cupertino, CA]         4           FORM 4     [, CA]  []   \n",
       "\n",
       "                                             content  \n",
       "0  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "1  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = downloader.dataframe\n",
    "# df = df.drop_duplicates()\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(21, 20)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>file_date</th>\n",
       "      <th>display_names</th>\n",
       "      <th>content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>[KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-04-30</td>\n",
       "      <td>[JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-04-17</td>\n",
       "      <td>[O'BRIEN DEIRDRE  (CIK 0001767094), Apple Inc....</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-04-17</td>\n",
       "      <td>[KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>[Maestri Luca  (CIK 0001513362), Apple Inc.  (...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2020-04-03</td>\n",
       "      <td>[WILLIAMS JEFFREY E  (CIK 0001496686), Apple I...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2020-04-03</td>\n",
       "      <td>[Maestri Luca  (CIK 0001513362), Apple Inc.  (...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2020-02-28</td>\n",
       "      <td>[WAGNER SUSAN  (CIK 0001059235), Apple Inc.  (...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2020-02-28</td>\n",
       "      <td>[LEVINSON ARTHUR D  (CIK 0001214128), Apple In...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2020-02-28</td>\n",
       "      <td>[JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...</td>\n",
       "      <td>SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    file_date                                      display_names  \\\n",
       "0  2020-05-12  [KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...   \n",
       "1  2020-04-30  [JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...   \n",
       "2  2020-04-17  [O'BRIEN DEIRDRE  (CIK 0001767094), Apple Inc....   \n",
       "3  2020-04-17  [KONDO CHRIS  (CIK 0001631982), Apple Inc.  (A...   \n",
       "4  2020-04-09  [Maestri Luca  (CIK 0001513362), Apple Inc.  (...   \n",
       "5  2020-04-03  [WILLIAMS JEFFREY E  (CIK 0001496686), Apple I...   \n",
       "6  2020-04-03  [Maestri Luca  (CIK 0001513362), Apple Inc.  (...   \n",
       "7  2020-02-28  [WAGNER SUSAN  (CIK 0001059235), Apple Inc.  (...   \n",
       "8  2020-02-28  [LEVINSON ARTHUR D  (CIK 0001214128), Apple In...   \n",
       "9  2020-02-28  [JUNG ANDREA  (CIK 0001051401), Apple Inc.  (A...   \n",
       "\n",
       "                                             content  \n",
       "0  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "1  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "2  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "3  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "4  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "5  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "6  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "7  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "8  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  \n",
       "9  SEC Form 4 \\n   FORM 4UNITED STATES SECURITIES...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_columns = [\"file_date\", \"display_names\", \"content\"]\n",
    "df[selected_columns].head(10)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Juchao"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from finnlp.data_sources.company_announcement.juchao import Juchao_Announcement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "start_date = \"2020-01-01\"\n",
    "end_date = \"2020-06-01\"\n",
    "stock = \"000001\"\n",
    "config = {\n",
    "    \"use_proxy\": \"china_free\",\n",
    "    \"max_retry\": 5,\n",
    "    \"proxy_pages\": 3,\n",
    "}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Gathering free ips by pages...: 100%|██████████| 3/3 [00:05<00:00,  1.86s/it]\n",
      "Checking ips: 100%|██████████| 45/45 [00:48<00:00,  1.09s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "获取到的代理ip数量: 45 。Get proxy ips: 45.\n",
      "能用的代理数量： 6。Usable proxy ips: 6.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1bb13261e75147929b30222347ab9cc5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading by page...:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "011ea7c465ad4e1aaccf09714b8e3e19",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Getting the text data...:   0%|          | 0/42 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "downloader = Juchao_Announcement(config)\n",
    "downloader.download_date_range_stock(start_date, end_date, stock = stock, get_content = True, delate_pdf = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>secCode</th>\n",
       "      <th>secName</th>\n",
       "      <th>orgId</th>\n",
       "      <th>announcementId</th>\n",
       "      <th>announcementTitle</th>\n",
       "      <th>announcementTime</th>\n",
       "      <th>adjunctUrl</th>\n",
       "      <th>adjunctSize</th>\n",
       "      <th>adjunctType</th>\n",
       "      <th>...</th>\n",
       "      <th>important</th>\n",
       "      <th>batchNum</th>\n",
       "      <th>announcementContent</th>\n",
       "      <th>orgName</th>\n",
       "      <th>tileSecName</th>\n",
       "      <th>shortTitle</th>\n",
       "      <th>announcementTypeName</th>\n",
       "      <th>secNameList</th>\n",
       "      <th>PDF_path</th>\n",
       "      <th>Content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>000001</td>\n",
       "      <td>平安银行</td>\n",
       "      <td>gssz0000001</td>\n",
       "      <td>1207862647</td>\n",
       "      <td>关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告</td>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>finalpage/2020-05-27/1207862647.PDF</td>\n",
       "      <td>148</td>\n",
       "      <td>PDF</td>\n",
       "      <td>...</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td></td>\n",
       "      <td>None</td>\n",
       "      <td>平安银行</td>\n",
       "      <td>关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>removed</td>\n",
       "      <td>证券代码： 000001               证券简称：平安银行          ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>None</td>\n",
       "      <td>000001</td>\n",
       "      <td>平安银行</td>\n",
       "      <td>gssz0000001</td>\n",
       "      <td>1207843688</td>\n",
       "      <td>2019年年度权益分派实施公告</td>\n",
       "      <td>2020-05-22</td>\n",
       "      <td>finalpage/2020-05-22/1207843688.PDF</td>\n",
       "      <td>214</td>\n",
       "      <td>PDF</td>\n",
       "      <td>...</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td></td>\n",
       "      <td>None</td>\n",
       "      <td>平安银行</td>\n",
       "      <td>2019年年度权益分派实施公告</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>removed</td>\n",
       "      <td>1  证券代码： 000001               证券简称：平安银行       ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     id secCode secName        orgId announcementId  \\\n",
       "0  None  000001    平安银行  gssz0000001     1207862647   \n",
       "1  None  000001    平安银行  gssz0000001     1207843688   \n",
       "\n",
       "                 announcementTitle announcementTime  \\\n",
       "0  关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告       2020-05-27   \n",
       "1                  2019年年度权益分派实施公告       2020-05-22   \n",
       "\n",
       "                            adjunctUrl  adjunctSize adjunctType  ...  \\\n",
       "0  finalpage/2020-05-27/1207862647.PDF          148         PDF  ...   \n",
       "1  finalpage/2020-05-22/1207843688.PDF          214         PDF  ...   \n",
       "\n",
       "  important batchNum announcementContent orgName tileSecName  \\\n",
       "0      None     None                        None        平安银行   \n",
       "1      None     None                        None        平安银行   \n",
       "\n",
       "                        shortTitle announcementTypeName secNameList PDF_path  \\\n",
       "0  关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告                 None        None  removed   \n",
       "1                  2019年年度权益分派实施公告                 None        None  removed   \n",
       "\n",
       "                                             Content  \n",
       "0  证券代码： 000001               证券简称：平安银行          ...  \n",
       "1  1  证券代码： 000001               证券简称：平安银行       ...  \n",
       "\n",
       "[2 rows x 25 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = downloader.dataframe\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42, 25)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>announcementTime</th>\n",
       "      <th>shortTitle</th>\n",
       "      <th>Content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告</td>\n",
       "      <td>证券代码： 000001               证券简称：平安银行          ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-05-22</td>\n",
       "      <td>2019年年度权益分派实施公告</td>\n",
       "      <td>1  证券代码： 000001               证券简称：平安银行       ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-05-20</td>\n",
       "      <td>关于获准发行小微企业贷款专项金融债券的公告</td>\n",
       "      <td>证券代码： 000001             证券简称：平安银行            ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>监事会决议公告</td>\n",
       "      <td>1  证券代码： 000001                证券简称： 平安银行     ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-05-15</td>\n",
       "      <td>2019年年度股东大会决议公告</td>\n",
       "      <td>1  证券代码： 000001               证券简称：平安银行       ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2020-05-15</td>\n",
       "      <td>2019年年度股东大会的法律意见书</td>\n",
       "      <td>北京总部   电话 : (86 -10) 8519 -1300  传真 : (86 -10...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2020-04-30</td>\n",
       "      <td>中信证券股份有限公司、平安证券股份有限公司关于公司关联交易有关事项的核查意见</td>\n",
       "      <td>1 中信证券股份有限公司 、平安证券股份有限 公司 关于平安银行股份有限公司  关联交易 有...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2020-04-30</td>\n",
       "      <td>独立董事独立意见</td>\n",
       "      <td>1   平安银行股份有限公司独立董事独立意见   根据《关于在上市公司建立独立董事制度的指导...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2020-04-30</td>\n",
       "      <td>关联交易公告</td>\n",
       "      <td>1  证券代码： 000001               证券简称：平安银行       ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2020-04-21</td>\n",
       "      <td>2020年第一季度报告全文</td>\n",
       "      <td>证券代码： 000001            证券简称：平安银行             ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  announcementTime                              shortTitle  \\\n",
       "0       2020-05-27         关于2020年第一期小型微型企业贷款专项金融债券发行完毕的公告   \n",
       "1       2020-05-22                         2019年年度权益分派实施公告   \n",
       "2       2020-05-20                   关于获准发行小微企业贷款专项金融债券的公告   \n",
       "3       2020-05-16                                 监事会决议公告   \n",
       "4       2020-05-15                         2019年年度股东大会决议公告   \n",
       "5       2020-05-15                       2019年年度股东大会的法律意见书   \n",
       "6       2020-04-30  中信证券股份有限公司、平安证券股份有限公司关于公司关联交易有关事项的核查意见   \n",
       "7       2020-04-30                                独立董事独立意见   \n",
       "8       2020-04-30                                  关联交易公告   \n",
       "9       2020-04-21                           2020年第一季度报告全文   \n",
       "\n",
       "                                             Content  \n",
       "0  证券代码： 000001               证券简称：平安银行          ...  \n",
       "1  1  证券代码： 000001               证券简称：平安银行       ...  \n",
       "2  证券代码： 000001             证券简称：平安银行            ...  \n",
       "3  1  证券代码： 000001                证券简称： 平安银行     ...  \n",
       "4  1  证券代码： 000001               证券简称：平安银行       ...  \n",
       "5   北京总部   电话 : (86 -10) 8519 -1300  传真 : (86 -10...  \n",
       "6  1 中信证券股份有限公司 、平安证券股份有限 公司 关于平安银行股份有限公司  关联交易 有...  \n",
       "7  1   平安银行股份有限公司独立董事独立意见   根据《关于在上市公司建立独立董事制度的指导...  \n",
       "8  1  证券代码： 000001               证券简称：平安银行       ...  \n",
       "9  证券代码： 000001            证券简称：平安银行             ...  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_columns = [\"announcementTime\", \"shortTitle\",\"Content\"]\n",
    "df[selected_columns].head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "finrl",
   "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": "3.7.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "afd6dc03c9be451573fc2885de79a969af6a24a159f11a3ead741ab7a9ff405f"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}