datax-admin/app/crud/data_table.py

import re
import time
import uuid
from typing import List
from app import models, schemas
from sqlalchemy.orm import Session
from app.common.hive import hiveDs
from configs.globals import g

def import_datalake(db: Session, item: schemas.ImportDataLake):
    # 校验数据湖中是否存在此库表
    ailab_table_names = hiveDs.list_tables()
    if not item.table_name in ailab_table_names: raise Exception(f'数据湖中不存在表{item.table_name}')
    data_table = models.DataTable(**{
        'database_name': item.database_name,
        'table_name': item.table_name,
        'table_path': item.table_path,
        'source': 0,
        'type': 2,
        'project_id': g.project_id,
        'create_time': int(time.time()),
        'user_id': g.user_id
    })
    db.add(data_table)
    db.commit()
    db.refresh(data_table)
    return data_table

def update_datalake(db: Session, datalake_id: int, item: schemas.ImportDataLake):
    ailab_table_names = hiveDs.list_tables()
    if not item.table_name in ailab_table_names: raise Exception(f'数据湖中不存在表{item.table_name}')
    db_item: models.DataTable = db.query(models.DataTable).filter(models.DataTable.id == datalake_id).first()
    if not db_item: raise Exception('未找到此数据')
    db_item.database_name = item.database_name
    db_item.table_name = item.table_name
    db_item.table_path = item.table_path
    db.commit()
    db.flush()
    db.refresh(db_item)
    return db_item

def delete_datalake(db: Session, datalake_id: int):
    db_item: models.DataTable = db.query(models.DataTable).filter(models.DataTable.id == datalake_id).first()
    if not db_item: raise Exception('未找到此数据')
    db.delete(db_item)
    db.commit()
    return db_item


def share_ailab(db: Session, item: schemas.ShareAilab):
    # 校验ailab中是否存在此表
    ailab_table_names = hiveDs.list_tables()
    table_names = item.table_names
    project_ids = item.project_ids
    db_items = []
    for table_name in table_names:
        if not table_name in ailab_table_names: raise Exception(f'数据湖中不存在表{table_name}')
        data_tables: List[models.DataTable] = db.query(models.DataTable)\
            .filter(models.DataTable.table_name==table_name)\
            .filter(models.DataTable.type == 0).all()
        p_ids = [d.project_id for d in data_tables]
        for project_id in project_ids:
            if project_id in p_ids:
                continue
            data_table = models.DataTable(**{
                'table_name': table_name,
                'source': 1,
                'type': 0,
                'project_id': project_id,
                'create_time': int(time.time()),
                'user_id': g.user_id
            })
            db_items.append(data_table)
    res = []
    for db_item in db_items:
        db.add(db_item)
        db.commit()
        db.refresh(db_item)
        res.append(db_item.to_dict())
    return res

def create_table(db: Session, item: schemas.CreateAilab):
    table_name = item.table_name
    columns = item.columns
    partition_column = item.partition_column
    # 校验表名是否重复
    table_names = hiveDs.list_tables()
    if table_name in table_names: raise Exception(f'表 {table_name} 命名重复')
    create_table_str = f"create table if not exists `{hiveDs.database_name}`.`{table_name}` ("
    if len(columns) == 0: raise Exception('表中至少有一个字段')
    for column in columns:
        create_table_str += f"`{column.column_name}` {column.Column_type},"
    create_table_str = create_table_str[:-1] + ")"
    if partition_column:
        create_table_str += f" PARTITIONED BY (`{partition_column}` string)"
    create_table_str += " ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' STORED AS TEXTFILE"
    hiveDs._execute_create_sql([create_table_str])
    res = hiveDs._execute_sql([f"show tables like '{table_name}'"])
    db_item = models.DataTable(**{
        'database_name': hiveDs.database_name,
        'table_name': table_name,
        'source': 1,
        'type': 1,
        'project_id': g.project_id,
        'create_time': int(time.time()),
        'user_id': g.user_id
    })
    if len(res[0][0])>0:
        db.add(db_item)
        db.commit()
        db.refresh(db_item)
        return {
            'status': True,
            'sql': create_table_str
        }
    else:
        raise Exception("建表失败")

def filter_by_name(table_name):
    return not bool(re.search(r'(?:job.*_tmp|task.*_tmp)',table_name))

def get_ailab_table(db: Session, project_id: int):
    project: models.Project = db.query(models.Project)\
        .filter(models.Project.id==project_id).first()
    if not project:
        raise Exception('项目不存在')
    res = []
    if project.type == 0:
        table_list = hiveDs.list_tables()
        for table_name in table_list:
            if filter_by_name(table_name): res.append(table_name)
    else:
        data_tables: List[models.DataTable] = db.query(models.DataTable)\
            .filter(models.DataTable.project_id==project_id)\
            .filter(models.DataTable.source==1)\
            .order_by(models.DataTable.create_time.desc()).all()
        res = [t.table_name for t in data_tables]
    return res

def get_ailab_table_schema(db: Session, table_name: str):
    table_schema = hiveDs.get_table_schema(table_name)
    return table_schema

def get_preview_ailab_table(db: Session, table_name: str):
    preview_data = hiveDs.get_preview_data(table_name)
    return preview_data

def get_lake_table(db: Session, project_id: int):
    data_tables: List[models.DataTable] = db.query(models.DataTable)\
            .filter(models.DataTable.project_id==project_id)\
            .filter(models.DataTable.source==0)\
            .order_by(models.DataTable.create_time.desc()).all()
    return data_tables

def get_lake_table_info(db: Session, dl_id: int):
    data_table: models.DataTable = db.query(models.DataTable)\
            .filter(models.DataTable.id==dl_id).first()
    if not data_table: raise Exception('未找到该数据湖表')
    return data_table