Files
BLD_sync/init_full_sync.py

267 lines
10 KiB
Python

import pyodbc
from config import SQL_SERVER_CONN, ACCESS_DRIVER, SYNC_MAPPING, BATCH_SIZE
import db_utils
import time
import os
import ntfy_utils
from log_utils import (LoggerManager, log_success, log_error, log_warning, log_info, log_processing,
log_skip, log_start, log_file, log_database, log_sync, log_complete)
# 初始化日志管理器
LoggerManager("init_full_sync", log_prefix="full_sync")
def format_duration(seconds):
"""格式化时长显示"""
if seconds < 60:
return f"{seconds:.1f}"
elif seconds < 3600:
return f"{seconds/60:.1f}分钟"
else:
return f"{seconds/3600:.1f}小时"
def has_identity_column(sql_cursor, schema, table):
"""检查表是否包含标识列"""
query = """
SELECT COUNT(*)
FROM sys.columns c
JOIN sys.tables t ON c.object_id = t.object_id
JOIN sys.schemas s ON t.schema_id = s.schema_id
WHERE s.name = ? AND t.name = ? AND c.is_identity = 1
"""
sql_cursor.execute(query, (schema, table))
return sql_cursor.fetchone()[0] > 0
def _send_summary_notification(sync_results, success_count, failed_count, total_rows, total_time):
"""发送汇总 ntfy 通知"""
if not sync_results:
return
# 分类成功和失败的表
success_tables = [r for r in sync_results if r['status'] == 'success']
failed_tables = [r for r in sync_results if r['status'] == 'failed']
# 构建消息
lines = []
lines.append("🎯 全量同步完成\n")
# 成功的表
if success_tables:
lines.append(f"✅ 成功 {len(success_tables)} 张表:")
for r in success_tables:
lines.append(f"{r['table']}: {r['rows']:,} 行, {format_duration(r['duration'])}")
lines.append("")
# 失败的表
if failed_tables:
lines.append(f"❌ 失败 {failed_count} 张表:")
for r in failed_tables:
error_brief = r['error'][:50] + "..." if len(r['error']) > 50 else r['error']
lines.append(f"{r['table']}: {error_brief}")
lines.append("")
# 总计
lines.append(f"总计: {total_rows:,} 行 | 用时: {format_duration(total_time)}")
message = "\n".join(lines)
# 根据成功/失败情况决定优先级和标签
if failed_count > 0:
priority = "high"
tags = ["warning", "database"]
else:
priority = "default"
tags = ["white_check_mark", "database"]
ntfy_utils.send_ntfy(
message=message,
title="全量同步完成",
priority=priority,
tags=tags
)
def run_full_sync():
log_info("=" * 70)
log_start( "全量初始化同步")
log_info( f"配置文件数: {len(SYNC_MAPPING)}")
log_info( f"批次大小: {BATCH_SIZE}")
log_info("=" * 70)
sync_start_time = time.time()
total_tables = 0
success_tables = 0
failed_tables = 0
total_rows = 0
# 收集每张表的同步结果,用于最终汇总通知
sync_results = [] # 格式: {'table': str, 'status': 'success'|'failed', 'rows': int, 'duration': float, 'error': str}
try:
sql_conn = db_utils.get_sql_conn()
sql_cursor = sql_conn.cursor()
sql_cursor.fast_executemany = True
log_database( "SQL Server 连接成功")
except Exception as e:
log_error( f"SQL Server 连接失败: {e}")
return
# 第一层循环:遍历配置文件中的所有文件路径
for acc_path, tables_map in SYNC_MAPPING.items():
if not os.path.exists(acc_path):
log_skip( f"文件不存在: {acc_path}")
continue
log_file( f"开始处理: {os.path.basename(acc_path)}")
file_start_time = time.time()
file_success = 0
file_failed = 0
file_rows = 0
try:
acc_conn = db_utils.get_access_conn(acc_path)
acc_cursor = acc_conn.cursor()
log_database( f"Access 文件连接成功: {os.path.basename(acc_path)}")
# 第二层循环:遍历该文件下的所有表映射
for acc_table, target_config in tables_map.items():
total_tables += 1
target_schema = target_config['target_schema']
target_table = target_config['target_table']
full_target_name = db_utils.fmt_table(target_schema, target_table)
log_processing( f"表 [{acc_table}] → [{target_table}]")
table_start_time = time.time()
try:
# 1. 检查 Access 表是否存在
acc_cursor.execute(f"SELECT TOP 1 * FROM [{acc_table}]")
# 2. 获取列结构
columns = [col[0] for col in acc_cursor.description]
log_info( f" 检测到 {len(columns)} 个列")
insert_sql = db_utils.generate_insert_sql(target_schema, target_table, columns)
# 3. TRUNCATE 目标表
sql_cursor.execute(f"TRUNCATE TABLE {full_target_name}")
log_info( f" 已清空目标表")
# 4. 检查并启用 IDENTITY_INSERT
has_identity = has_identity_column(sql_cursor, target_schema, target_table)
identity_enabled = False
if has_identity:
try:
sql_cursor.execute(f"SET IDENTITY_INSERT {full_target_name} ON")
identity_enabled = True
log_info( f" 已启用 IDENTITY_INSERT")
except Exception as id_err:
log_error( f" 无法启用 IDENTITY_INSERT: {id_err}")
raise
# 5. 传输数据
log_info( f" 开始数据传输...")
acc_cursor.execute(f"SELECT * FROM [{acc_table}]")
table_rows = 0
start_time = time.time()
last_log_time = start_time
while True:
rows = acc_cursor.fetchmany(BATCH_SIZE)
if not rows: break
sql_cursor.executemany(insert_sql, rows)
table_rows += len(rows)
# 每5秒或每10000行记录一次进度
current_time = time.time()
if (current_time - last_log_time >= 5) or (table_rows % 10000 == 0):
elapsed = current_time - start_time
rate = table_rows / elapsed if elapsed > 0 else 0
log_info( f" 进度: {table_rows:,} 行 | 速率: {rate:,.0f} 行/秒")
last_log_time = current_time
# 6. 关闭 IDENTITY_INSERT
if identity_enabled:
try:
sql_cursor.execute(f"SET IDENTITY_INSERT {full_target_name} OFF")
log_info( f" 已关闭 IDENTITY_INSERT")
except Exception as id_err:
log_warning( f" 关闭 IDENTITY_INSERT 时警告: {id_err}")
sql_conn.commit()
# 显示最终统计
table_time = time.time() - table_start_time
final_rate = table_rows / table_time if table_time > 0 else 0
log_info( f"表 [{target_table}] 完成: {table_rows:,} 行 | 速率: {final_rate:,.0f} 行/秒 | 用时: {format_duration(table_time)}")
# 收集成功结果
sync_results.append({
'table': target_table,
'status': 'success',
'rows': table_rows,
'duration': table_time,
'error': None
})
success_tables += 1
file_success += 1
total_rows += table_rows
file_rows += table_rows
except Exception as tbl_err:
failed_tables += 1
file_failed += 1
error_msg = str(tbl_err)
log_info( f"表 [{acc_table}] 同步失败: {error_msg}")
# 收集失败结果
sync_results.append({
'table': target_table,
'status': 'failed',
'rows': 0,
'duration': time.time() - table_start_time,
'error': error_msg
})
sql_conn.rollback()
# 确保清理 IDENTITY_INSERT 状态
try:
sql_cursor.execute(f"SET IDENTITY_INSERT {full_target_name} OFF")
except:
pass
acc_conn.close()
# 文件级别汇总
file_time = time.time() - file_start_time
summary = f"文件 [{os.path.basename(acc_path)}] 完成: "
summary += f"成功 {file_success} 张表"
if file_failed > 0:
summary += f" | 失败 {file_failed} 张表"
summary += f" | 共 {file_rows:,} 行 | 用时: {format_duration(file_time)}"
log_sync( summary)
except Exception as file_err:
log_error( f"文件 [{os.path.basename(acc_path)}] 处理失败: {file_err}", exc_info=True)
sql_conn.close()
# 总结统计
sync_time = time.time() - sync_start_time
log_info("\n" + "=" * 70)
log_info( "全量同步任务结束")
log_info("-" * 70)
log_info( f"总表数: {total_tables} 张 | 成功: {success_tables} 张 | 失败: {failed_tables}")
log_info( f"总行数: {total_rows:,}")
log_info( f"总用时: {format_duration(sync_time)}")
if total_rows > 0 and sync_time > 0:
log_info( f"整体平均速率: {total_rows/sync_time:,.0f} 行/秒")
log_info("=" * 70)
# 发送汇总 ntfy 通知
_send_summary_notification(sync_results, success_tables, failed_tables, total_rows, sync_time)
if __name__ == "__main__":
run_full_sync()