简体繁体 English

使用系统VS实现包调用perl脚本

[英]calling perl script with system VS implementing package

原文 2015-08-05 20:49:33 9 1 performance/ perl/ system

Let me start with giving an example of what I'm dealing with first: 让我开始举例说明我首先要处理的事情：

I often call existed Perl scripts from previous engineers to process some data, and then proceed further with my script. 我经常调用以前的工程师提供的现有Perl脚本来处理一些数据，然后进一步处理我的脚本。 I either use system or back-ticks to call other people scripts within my script. 我可以使用system调用或back-ticks在脚本中调用其他人的脚本。

Now, I'm wondering if I rewrite those scripts as packages and use require or use to include those packages in my script, will it increase the processing speed? 现在，我想知道是否将那些脚本重写为包，并使用require或use将这些包包含在我的脚本中，这会提高处理速度吗？ How big of a difference would it be? 会有多大的不同？

1 个解决方案

Benefits: 优点：

It would save the time taken to load the shell, load perl , compile the script and the module it uses. 这样可以节省加载外壳，加载perl ，编译脚本及其使用的模块所需的时间。 That's a couple of seconds minimum , but it could be much larger. 这是最少几秒钟的时间，但可能会更大。
If you had to serialize data to pass to the child, you also save the time taken to serialize and deserialize the data. 如果必须序列化数据以传递给子级，则还可以节省序列化和反序列化数据所花费的时间。
It would allow more flexible interfaces. 它将允许更灵活的接口。
It would make error handling easier and more flexible. 这将使错误处理更加容易和灵活。

Downsides: 缺点：

Since everything is now in the same process, the child can have a much larger effect on the parent. 由于现在一切都在同一过程中，因此孩子对父母的影响可能更大。 eg A crash in the child will take down the parent. 例如，孩子的崩溃将使父母丧生。