简体繁体 English

MVC：将C ++模型插入基于Web的视图

[英]MVC: Plugging a C++ Model into a Web-based View

原文 2015-06-22 23:32:34 8 1 javascript/ php/ jquery/ html/ c++

Background 背景

I am writing a tool using MVC where the model is written in C++, controller in JavaScript/jQuery and the view in HTML/CSS. 我正在使用MVC编写工具，其中模型是用C ++编写的，控制器是用JavaScript / jQuery编写的，视图是用HTML / CSS编写的。 The reason the model is in C++ is for eventual portability to a thick client from the "speed build" on web. 该模型使用C ++的原因是为了最终从Web上的“速度构建”将其移植到胖客户端。

Issue 问题

My major roadblock at this point is connecting my C++ model to the HTML form that is currently collecting input data from the user. 此时，我的主要障碍是将C ++模型连接到当前正在从用户收集输入数据的HTML表单中。

I have looked into Emscripten for compiling the C++ directly into JavaScript, but I don't like the idea of having my underlying model be visible to the end user in any form . 我已经研究过Emscripten ，可以将C ++直接编译为JavaScript，但是我不喜欢让我的底层模型以任何形式对最终用户可见的想法。 This is due to the model being hooked up to a MySQL server handling a lot of sensitive data. 这是由于该模型已连接到处理大量敏感数据的MySQL服务器。 I want to minimize risk of the end user accidentally (or purposely) messing with the function call invoking the model. 我想最大程度地降低最终用户意外（或故意）弄乱调用模型的函数调用的风险。

I have done quite a bit of research on PHP/C++ linking, but all the resources I could find have been severely outdated and/or convoluted. 我已经对PHP / C ++链接做了很多研究，但是我发现的所有资源都已经严重过时和/或令人费解。

I'd ideally like to have the the data pushed to the server using a POST method before invoking the model so I can have maximum control over input validation both on the end user and server side. 理想情况下，我希望在调用模型之前使用POST方法将数据推送到服务器，以便在最终用户和服务器端都可以最大程度地控制输入验证。

I've also thought about using PHP to push to the MySQL server then having the C++ pull down from the server, but this seems a little obfuscated and raises the issue of when/how to wake the C++ up. 我还考虑过使用PHP推送到MySQL服务器，然后将C ++从服务器中拉出，但这似乎有点令人困惑，并引发了何时/如何唤醒C ++的问题。

Question 题

Is there a good way to maintain security of my model (have it run entirely server-side behind a POST method) as well as the understandability of how this was implemented? 有没有很好的方法来维护我的模型的安全性（它已完全在POST方法后在服务器端运行）以及如何实现该模型？ I am a temp employee and my project will pass to other hands for the port to a thick client from web. 我是临时工，我的项目将移交给其他人，以从Web移植到胖客户端。

Speed is also a concern, but I'm assuming (or hoping) that the only bottleneck will be the pipe/script between the POST to the server and the C++ since most of the heavy-lifting will be occurring in the model. 速度也是一个问题，但是我假设（或希望）唯一的瓶颈将是到服务器的POST和C ++之间的管道/脚本，因为大多数繁重的工作都会在模型中发生。

1 个解决方案

What about not putting any business logic in your model and instead break the model up into an Entity and Service ? 不将任何业务逻辑放入模型中，而是将模型分解为Entity和Service怎么办？

You could then use emscripten on the Entity while your Service holds all of your MySql connection logic. 然后，当Service包含所有MySql连接逻辑时，您可以在Entity上使用emscripten 。