TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Advice: how to convert SQL data?

2 点作者 adatta02超过 16 年前
I need some advice/a clever idea. Basically, I have a legacy SQL database that I inherited from our client's old development shop and I need to move all the data over to our new database.<p>The two schemas are vastly different - the old one is a MYISAM based system with no FKs and is missing some serious integrity.<p>My question is - is there some easy way to massage the old data into our new schema?<p>What I've been working on now is basically to build PHP objects for their old data and then write something to generate SQL statements for our new schema. Since we are using Symfony for the new project, I might also try and hook into Symfony and let it figure out the SQL. This will probably work but its really slow and tedious.

4 条评论

fizx超过 16 年前
Not nearly enough info for anyone to help you with, and not interesting enough for the crowd here to likely care about. You should probably seek paid help from someone with experience.
bayareaguy超过 16 年前
Since the schemas are vastly different there's probably no alternative to building a custom ETL process. If you already have a workable solution to convert the data then you should probably just get to work and do what you can to keep your unknowns to a minimum even if it is a slow and tedious process. But if you don't really know if what you're doing will work then be honest about that upfront - you'll need to factor in additional time to compensate for mistakes. Try doing your conversions in a series of small steps with the appropriate testing along the way.<p>I've seen projects that sound like yours go way over budget because of well-meaning attempts to use "silver bullet" tools (often championed by non-stakeholders) that either took too long to figure out how to use properly or didn't end up working as well as expected.
akronim超过 16 年前
how much data are you moving?
DanielBMarkham超过 16 年前
This is hard work. There are no magic bullets.<p>Some kind of ETL is all you got. Look forward to lots of exceptions.