基于teradata的python模块实现简易的客户端

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Date    : 2018-09-07 20:48:11
# @Author  : Sheldon ([email protected])
# @Version : 0.0.1

import teradata
import pandas as pd


class GenHasnoLen(Exception):
    pass


class TD:
    def __init__(self):
        _udaExec = teradata.UdaExec()
        self.session = _udaExec.connect("${dataSourceName}")
        self._sql = None
        self._len = None

    def __call__(self, sql=None, index_col=None, coerce_float=True, params=None,
                 parse_dates=None, columns=None, chunksize=None, usepd=True, **kwargs):
        if usepd:
            result = pd.read_sql(sql, self.session, index_col, coerce_float, params,
                                 parse_dates, columns, chunksize)
            if chunksize:
                self._len = GenHasnoLen()
            else:
                self._len = len(result)
        else:
            result = self.session.execute(sql, params, **kwargs)
            self._len = result.rowcount
        self._sql = sql
        return result

    def __len__(self):
        return self._len

    def __str__(self):
        info = f"sql语句是；\n{self._sql}\n" \
               f"一共有{self._len}条记录"
        return info

    def _sel_1_tbs_db(self, tablename, *addcol):
        scdbsql = f"SELECT DatabaseName,TableKind,CommentString,CreatorName\n" \
                  f"FROM DBC.Tables\n" \
                  f"WHERE TableName = '{tablename}'"
        temp_dbname = self(scdbsql, usepd=False)
        rc = temp_dbname.rowcount
        if rc > 1:
            rows = []
            print("此表存在于多个数据库中，请从下列库中选择")
            for row in temp_dbname:
                rows.append(row)
                print(row)
            while True:
                indexstr = input("请输入数据库所在行号,按Enter确定," \
                                 "输入非数字退出:\n")
                if not indexstr.isnumeric():
                    return
                index = int(indexstr.strip())
                if 0 < index <= row.rowNum:
                    selrow = rows[index - 1]
                    break
                print("输入错误的行号范围,请重新输入")
        elif rc == 1:
            selrow = next(temp_dbname)
        else:
            return
        return map(lambda col: selrow[col].strip(), ("DatabaseName", *addcol))

    def dblist(self, access=True, dbkind="D"):
        # dbkind数据库类型，D为DATABASE，U为Users,None为所有
        acc_clause = ''
        dbk_clause = ''
        clause = ''
        if access:
            config = self.session.udaexec.config
            externalDSN = config.resolve("${dataSourceName}")
            args = config.section(externalDSN)
            username = args.get("username")
            if username:
                acc_clause = f"AccountName = (\n" \
                             f"\tSELECT DefaultAccount\n" \
                             f"\tFROM DBC.Users\n" \
                             f"{acc_clause}" \
                             f"\tWHERE UserName='{username}'\n" \
                             f")"
        if dbkind:
            assert dbkind in ("D", "U")
            dbk_clause = f"DBKind = '{dbkind}'"
        join_clause = "\nAND ".join(x for x in (acc_clause, dbk_clause) if x)
        if join_clause:
            clause = "\nWHERE " + join_clause
        dblist_sql = f"SELECT DatabaseName,CommentString\n" \
                     f"FROM DBC.Databases{clause}"
        return self(dblist_sql)

    def tbsdb(self, tbname, usepd=True):
        trans_sql = f"SELECT DatabaseName\n" \
                    f"FROM DBC.Databases2 DB2\n" \
                    f"INNER JOIN DBC.Tables2 TB2\n" \
                    f"ON DB2.DatabaseId = TB2.DatabaseId\n" \
                    f"WHERE TVMName = '{tbname}'"
        return self(trans_sql, usepd=usepd)

    def definition(self, tablename: str, tbkind: str = "T"):
        kind2type = {"T": "TABLE", "O": "TABLE", "V": "VIEW", "M": "MACRO",
                     "P": "PROCEDURE", "E": "PROCEDURE", "R": "FUNCTION",
                     "F": "FUNCTION", "G": "TRIGGER", "I": "JOIN INDEX",
                     "N": "HASH INDEX"}
        if '.' in tablename:
            dtname = tablename
        else:
            dbname, tbkind = self._sel_1_tbs_db(tablename, "TableKind")
            if not dbname:
                return
            dtname = f"{dbname}.{tablename}"
        def_cursor = self(f"SHOW {kind2type[tbkind]} {dtname}", usepd=False)
        row_lst = next(def_cursor)
        if row_lst:
            return row_lst[0].replace("\r", "\n")

    def tablelist(self, dbname: str = None):
        clause = '' if dbname is None else f"\nWHERE TableName = '{dbname}'"
        tbsql = f"SELECT DISTINCT DatabaseName,TableName,TableKind,CommentString,CreatorName\n" \
                f"FROM DBC.Tables" \
                f"{clause}"
        return self(tbsql)

    def columnlist(self, tablename: str):
        colstr = "ColumnName,ColumnTitle,ColumnType,ColumnLength,DecimalTotalDigits,DecimalFractionalDigits,ColumnFormat,DefaultValue,Nullable,CommentString"
        base = f"SELECT {colstr}\n" \
               f"FROM DBC.Columns\n"
        if "." in tablename:
            db, table = tablename.split(".")
            suffix = f"WHERE DatabaseName='{db}' AND TableName='{table}'"
        else:
            colstr = "DatabaseName" + colstr
            suffix = f"WHERE TableName='{tablename}'\n ORDER BY DatabaseName"
        sql = base + suffix
        print(sql)
        return self(sql)

    def indexs(self, dtname):
        dbname, tablename = dtname.split(".")
        indexsql = f"SELECT IndexName,ColumnName,IndexType,UniqueFlag,IndexNumber,ColumnPosition\n" \
                   f"FROM DBC.Indices\n" \
                   f"WHERE DatabaseName='{dbname}' AND TableName='{tablename}'"
        return self(indexsql)

    def search(self, sth, dbname: str = None, user: str = None):
        if dbname:
            dbclause = f" AND DatabaseName='{dbname}'"
        else:
            dbclause = ''
        if user:
            userclause = f" AND CreatorName='{user}'"
        else:
            userclause = ''
        scsql = f"SELECT TRIM(DatabaseName) || '.' || TRIM(TableName) AS TbPath,TableKind,CommentString\n" \
                f"FROM DBC.Tables\n" \
                f"WHERE (TableName LIKE '%{sth}%' OR CommentString LIKE '%{sth}%'){dbclause+userclause}\n" \
                f"ORDER BY CASE 2*LEAST(1,INDEX(TableName,'{sth}'))+LEAST(1,INDEX(CommentString,'{sth}'))\n" \
                f"\t\tWHEN 3 THEN 1.0*LENGTH('{sth}')/LENGTH(TRIM(TableName))+1.0*LENGTH('{sth}')/LENGTH(TRIM(CommentString))\n" \
                f"\t\tWHEN 2 THEN 1.0*LENGTH('{sth}')/LENGTH(TRIM(TableName))\n" \
                f"\t\tELSE 1.0*LENGTH('{sth}')/LENGTH(TRIM(CommentString))\n" \
                f"\t\tEND DESC\n"
        return self(scsql)


if __name__ == "__main__":
    app = TD()
    print(app.indexs("FRT_DATA.POS_DTL"))
    app('SELECT TOP 10 * FROM DBC.Tables')
    print(app.definition("FRT_DATA.POS_DTL"))
    print(app.definition("TRD_INFO"))
    print(app.dblist())
    print(app.search('外卖'))
    print(app)

使用方法：

• 安装依赖库
• 安装ODBC驱动
• 在python脚本同级目录，新建配置文件：udaexec.ini，自行修改相应的配置项，配置内容如下：

# Application Configuration
[CONFIG]
appName=TDClient
version=2
logConsole=False
dataSourceName=TDPROD
table=DBC.DBCInfo 
 
# Default Data Source Configuration
[DEFAULT]
method=odbc
charset=ASCII
 
# Data Source Definition
[TDPROD]
system=10.213.5.6
username=Sheldon
password=pw12345