Files
voidandClaude Sonnet 4.6 bc4a6b504f Initial commit: archer, a Go-native MySQL/MariaDB table archiver
Renamed from arcdb: module path, cmd/archer, internal/archererr,
docs, build scripts and ARCHER_* environment variables.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-11-06 22:12:17 +08:00

214 lines
6.1 KiB
Batchfile
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
@echo off
chcp 65001 >nul 2>&1
setlocal EnableDelayedExpansion
REM =====================================================================
REM export-all.cmd
REM
REM 用 archer 把某个 MySQL 库的「所有表」导出为按行数分片 + gzip 压缩的 dump。
REM
REM 用法:
REM scripts\export-all.cmd
REM 使用下面「配置区」的默认值
REM
REM scripts\export-all.cmd DSN OUTDIR SHARD_ROWS CHUNK_SIZE
REM 例: scripts\export-all.cmd ^
REM "root:pass@tcp(10.0.0.5:3306)/app" "D:\dump" 500000 5000
REM
REM 产物:
REM <OUTDIR>\<表名>\000001.<lo>-<hi>.dump.gz 分片文件(按主键连续、不重叠)
REM <OUTDIR>\<表名>\manifest.jsonl 每片的行数/字节/sha256
REM <OUTDIR>\<表名>\state.json 断点续跑状态
REM <OUTDIR>\<表名>.log 该表 archer 的完整输出
REM
REM 断点续跑:
REM 用同样的参数再跑一遍即可,archer 从 state.json 继续;
REM state.json 里 complete=true 的表会被跳过。
REM 想重导某张表,先删掉它的目录。
REM
REM 依赖:
REM dist\archer_windows_amd64.exe (scripts\build.ps1 构建)
REM mysql.exe (只用来枚举表名,需在 PATH 里)
REM =====================================================================
REM ------------------------------ 配置区 ------------------------------
REM 源库连接。改动源库时,SRC 与下面 5 个变量要一起改
REM (mysql 客户端要从后者取参数去枚举表名)。
set "SRC=root:xyTyfxhn4h@tcp(s1.void01.com:3306)/lexis"
set "DB=lexis"
set "DB_HOST=s1.void01.com"
set "DB_PORT=3306"
set "DB_USER=root"
set "DB_PASS=xyTyfxhn4h"
REM 输出根目录(每张表会在其下建一个同名子目录)
set "OUTDIR=%~dp0..\export"
REM 每个分片文件的行数;小于表行数时只会得到 1 个分片
set "SHARD_ROWS=1000000"
REM 每次 SELECT / 事务 / checkpoint 的行数,与分片大小无关
set "CHUNK_SIZE=5000"
REM 输出格式 dump|csv|jsonl 与压缩方式 none|gzip|zstd
set "FMT=dump"
set "COMPRESS=gzip"
REM 只导指定表(逗号分隔);留空表示导全部
set "ONLY_TABLES="
REM 排除表名通配符,如 tmp_* REM_*;留空表示不排除
set "EXCLUDE="
REM --------------------------------------------------------------------
REM 命令行覆盖
if not "%~1"=="" set "SRC=%~1"
if not "%~2"=="" set "OUTDIR=%~2"
if not "%~3"=="" set "SHARD_ROWS=%~3"
if not "%~4"=="" set "CHUNK_SIZE=%~4"
set "EXE=%~dp0..\dist\archer_windows_amd64.exe"
if not exist "%EXE%" (
echo [错误] 找不到 archer: %EXE%
echo 先运行 scripts\build.ps1 构建
goto :fail
)
where mysql.exe >nul 2>&1
if errorlevel 1 (
echo [错误] PATH 中找不到 mysql.exe,无法枚举表名
echo 把 mysql 的 bin 目录加进 PATH,或用 ARCHER 逐表手写命令
goto :fail
)
if not exist "%OUTDIR%" mkdir "%OUTDIR%" >nul 2>&1
if not exist "%OUTDIR%" (
echo [错误] 无法创建输出目录: %OUTDIR%
goto :fail
)
REM mysql 用环境变量传密码,避免出现在命令行/进程列表里
set "MYSQL_PWD=%DB_PASS%"
set "MYSQL_CMD=mysql.exe --protocol=TCP -h%DB_HOST% -P%DB_PORT% -u%DB_USER% --default-character-set=utf8mb4 -N -B"
echo.
echo archer : %EXE%
echo schema : %DB% (%DB_HOST%:%DB_PORT%)
echo out dir : %OUTDIR%
echo shard : %FMT% + %COMPRESS%, %SHARD_ROWS% rows/shard, chunk %CHUNK_SIZE%
echo.
REM ---------------------------- 表清单 ----------------------------
if defined ONLY_TABLES (
set "TABLE_LIST=%ONLY_TABLES:,= %"
set "TABLE_LIST=!TABLE_LIST: =!|"
) else (
set "TABLE_LIST="
for /f "usebackq delims=" %%t in (`%MYSQL_CMD% -e "SELECT table_name FROM information_schema.tables WHERE table_schema='%DB%' AND table_type='BASE TABLE' ORDER BY table_name" 2^>nul`) do (
set "TABLE_LIST=!TABLE_LIST!%%t|"
)
)
if "!TABLE_LIST!"=="" (
echo [错误] 没有取到任何表,检查库名/账号/网络
goto :fail
)
REM ---------------------------- 逐表导出 ----------------------------
set "OK=0"
set "SKIP=0"
set "FAILED="
set /a IDX=0
for %%t in ("!TABLE_LIST:|=" "!") do (
set "T=%%~t"
if not "!T!"=="" (
call :one_table "!T!"
)
)
echo.
if defined FAILED (
echo 失败的表: %FAILED%
echo.
echo 完成: 成功 %OK%,跳过 %SKIP%,失败若干
goto :fail
)
echo 完成: 成功 %OK%,跳过 %SKIP%,失败 0
endlocal
exit /b 0
REM ====================================================================
REM 单表导出
REM ====================================================================
:one_table
set "T=%~1"
set /a IDX+=1
if defined EXCLUDE (
REM 简单通配:把 * 替换成批处理的 .*
set "PAT=!EXCLUDE:*=.!"
for %%p in (!PAT!) do (
for /f "delims=" %%m in ("!T!") do (
if "%%m"=="%%m" (
REM 用 findstr 做通配匹配
)
)
)
echo !T! | findstr /r /i /c:"^!EXCLUDE:\*=.!$" >nul 2>&1
if not errorlevel 1 (
echo [!IDX!] !T! 跳过(被 EXCLUDE 匹配)
goto :eof
)
)
set "TDIR=%OUTDIR%\!T!"
REM 已完成则跳过
if exist "!TDIR!\state.json" (
findstr /c:"\"complete\": true" "!TDIR!\state.json" >nul 2>&1
if not errorlevel 1 (
echo [!IDX!] !T! 跳过(已完成)
set /a SKIP+=1
goto :eof
)
echo [!IDX!] !T! 续跑
)
set "LOG=%OUTDIR%\!T!.log"
set "START=%TIME%"
"%EXE%" ^
--source "%SRC%" ^
--table "!T!" ^
--where "1=1" ^
--file "!TDIR!" ^
--format %FMT% ^
--compress %COMPRESS% ^
--shard-by count ^
--shard-rows %SHARD_ROWS% ^
--chunk-size %CHUNK_SIZE% ^
--statistics > "!LOG!" 2>&1
if errorlevel 1 (
echo [!IDX!] !T! 失败,日志: !LOG!
for /f "delims=" %%l in ('findstr /c:"archer:" "!LOG!" 2^>nul') do echo %%l
set "FAILED=!FAILED! !T!"
goto :eof
)
REM 统计分片数
set "SHARDS=0"
if exist "!TDIR!\manifest.jsonl" (
for /f %%n in ('type "!TDIR!\manifest.jsonl" 2^>nul ^| find /c /v ""') do set "SHARDS=%%n"
)
echo [!IDX!] !T! 完成: !SHARDS! 片
set /a OK+=1
goto :eof
REM ====================================================================
:fail
endlocal
exit /b 1