在克隆中进行更改时保持 Mercurial 子存​​储库最新的工作流程? [英] Workflow to keep Mercurial subrepositories up to date when changes are made in clones?

问题描述

我已经阅读了有关 Mercurial 子存​​储库的所有问题，但我仍然不确定它是如何工作的.我们在 IIS 上使用 hgweb.cgi，所以我们有 http://到我们所有存储库的路径.

I have read every question on Mercurial subrepositories that I could and still not sure how it works. We are using hgweb.cgi on IIS so we have http:// paths to all of our repos.

这是基本布局:

/Libraries - http://server/Libraries
/Project1 - http://server/Project1
/Project1/LibrariesSubrepo - http://server/Project1/LibrariesSubrepo (clone of libraries)
/Project2 - http://server/Project1
/Project2/LibrariesSubrepo - http://server/Project2/LibrariesSubrepo (clone of libraries)

在 .hgsub 文件中，我将路径设置为:

In the .hgsub files, I have the path setup as:

LibrariesSubrepo = LibrariesSubrepo

如果我想处理这些项目，我会将它们从服务器克隆到我的工作站，这样我就有了 Project1 和 Project2 的克隆，它也会自动拉取子存储库.

If I want to work on the projects, I clone them from the server to my workstation so I have a clone of Project1 and Project2 and it automatically pulls the subrepos as well.

在我本地工作站的 project1 中，我对/Project1/LibrariesSubrepo 中的文件进行了一些更改.如何确保这些更改返回到服务器上库的源存储库?我是否必须手动将子存储库中的更改推送到子存储库的源存储库?

In project1 on my local workstation, I make some changes to files in the /Project1/LibrariesSubrepo. How do I ensure that those changes make it back to the source repo for Libraries on the server? Do I have to manually push the changes from the subrepo to the source repo of the subrepo ?

我希望能够在本地对 Project1 存储库进行更改，然后提交/推送这些更改，然后在 Project2 中，我应该能够从应该将 Project2/LibrariesSubrepo 更新到最新的服务器中提取更改版本.

I would like to be able to make changes in the Project1 repo locally, then commit/push those changes, then in Project2, I should be able to pull the changes from the server which should update the Project2/LibrariesSubrepo to the latest version.

推荐答案

你绝对可以手动从 Project1 中的 LibrariesSubrepo 推送到 Libraries> 服务器上的存储库，然后手动从 Libraries 拉入 Project2 克隆中的 LibrariesSubrepo.

You can definitely push manually from your LibrariesSubrepo in Project1 to the Libraries repository on the server and then manually pull from Libraries into LibrariesSubrepo in a Project2 clone.

但是，您也可以链接服务器上的子仓库:

However, you can also link the subrepos on the server:

1. 共享扩展让您可以创建两个或多个共享一个.hg 文件夹.在服务器上执行此操作，以便服务器上子存储库的三个实例共享 .hg 文件夹.

1. The share extension lets you create two or more repositories that share a .hg folder. Do this on the server so that the three instances of the subrepo on the server share the .hg folders.

符号链接 .hg 文件夹——更老套，只有在你可以创建符号链接的服务器上才有效(Unix 服务器或 Windows Vista 及更高版本，具有正确的权限).

Symlink the .hg folders -- more oldschol and it only works if you're on a server where you can create symlink (Unix servers or Windows Vista and later with the right priviledge).

在服务器上只保留一个子存储库，但在 hgweb 配置中将其公开 3 次.像这样:

Keep just one subrepo on the server but expose it three times in the hgweb configuration. Like this:

[paths]
Project1                  = /repos/Project1
Project1/LibrariesSubrepo = /repos/Libraries
Project2                  = /repos/Project2
Project2/LibrariesSubrepo = /repos/Libraries

在所有情况下，服务器上实际上只有三个存储库:Project1、Project2 和 Libraries.

In all cases, there are actually only three repositories on the server: Project1, Project2, and Libraries.

这篇关于在克隆中进行更改时保持 Mercurial 子存​​储库最新的工作流程?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！