Generate mysql database structure documentation with python

pip install pymysql
pip install document

# -*- coding: utf-8 -*-
import pymysql
from docx import Document
from  import Pt
from  import qn

db = (host='127.0.0.1', # Database server IP
                         port=3306,
                         user='root',
                         passwd='123456',
                         db='test_db') # database name)
# Query the corresponding field information based on the table name.
def query(tableName):
    # Open a database connection
    cur = ()
    sql = "select b.COLUMN_NAME,b.COLUMN_TYPE,b.COLUMN_COMMENT from (select * from information_schema.`TABLES`  where TABLE_SCHEMA='test_db') a right join(select * from information_schema.`COLUMNS` where TABLE_SCHEMA='test_db_test') b on a.TABLE_NAME = b.TABLE_NAME where a.TABLE_NAME='" + tableName+"'"
    (sql)
    data = ()
    
    return data
# query the current library below all the table names, table name: tableName; table name + comments (for filling the word document): concat(TABLE_NAME, '(', TABLE_COMMENT, ')')
def queryTableName():
    cur = ()
    sql = "select TABLE_NAME,concat(TABLE_NAME,'(',TABLE_COMMENT,')') from information_schema.`TABLES`  where TABLE_SCHEMA='test_db_test'"
    (sql)
    data = ()
    return data
# Generate word structure for each table, output to word document
def generateWord(singleTableData,document,tableName):
    p=document.add_paragraph()
    p.paragraph_format.line_spacing=1.5 # Set the paragraph to 1.5 times the line spacing.
    p.paragraph_format.space_after=Pt(0) # Set paragraph 0 pounds after paragraph
    #document.add_paragraph(tableName,style='ListBullet')
    r=p.add_run('\n'+tableName)
    =u'Song Style'
    =Pt(12)
    table = document.add_table(rows=len(singleTableData)+1, cols=3,style='Table Grid')
    =Pt(11)
    =u'Calibri'
    # Setting the header style
    # Only three table headers are generated here, which can be modified by actual requirements
    for i in ((0,'NAME'),(1,'TYPE'),(2,'COMMENT')):
        run = (0,i[0]).paragraphs[0].add_run(i[1])
         = 'Calibri'
         = Pt(11)
        r = run._element
        (qn('w:eastAsia'), 'Song Style')
    
    for i in range(len(singleTableData)):
        # Set the style of the data in the table
        for j in range(len(singleTableData[i])):
            run = (i+1,j).paragraphs[0].add_run(singleTableData[i][j])
             = 'Calibri'
             = Pt(11)
            r = run._element
            (qn('w:eastAsia'), 'Song Style')
        #(i+1, 0).text=singleTableData[i][1]
        #(i+1, 1).text=singleTableData[i][2]
        #(i+1, 2).text=singleTableData[i][3]
    

if __name__ == '__main__':
    # Define a document
    document = Document()
    # Setting the default style of fonts
    ['Normal']. = u'Song Style'
    ['Normal']._element.(qn('w:eastAsia'), u'Song Style')
    # Get all table name information and table comment information under the current library
    tableList = queryTableName()
    # Loop through the database to get the table field details and call generateWord to generate word data
    # Due to the rush of time, my side of the choice is to query the database directly, the execution of more than 100 queries, you can optimize, query all the table structure, in the code inside the structure of each table to be split
    for singleTableName in tableList:
        data = query(singleTableName[0])
        generateWord(data,document,singleTableName[1])
    #Save to Document
    ('Database Design.docx')