Let's mimic git.git auto tag fetch behavior.

Related to https://github.com/libgit2/libgit2sharp/pull/927/files#r23450279

git fetch against a newly initialized repository with the following config

[core]
    repositoryformatversion = 0
    filemode = false
    bare = false
    logallrefupdates = true
    symlinks = false
    ignorecase = true
    hideDotFiles = dotGitOnly
[remote "origin"]
    url = https://github.com/libgit2/TestGitRepository
    fetch = +refs/heads/*:refs/remotes/origin/*

outputs

$ git fetch
remote: Counting objects: 68, done.
remote: Total 68 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (68/68), done.
From https://github.com/libgit2/TestGitRepository
 * [new branch]      first-merge -> origin/first-merge
 * [new branch]      master     -> origin/master
 * [new branch]      no-parent  -> origin/no-parent
remote: Counting objects: 1, done.
remote: Total 1 (delta 0), reused 1 (delta 0)
Unpacking objects: 100% (1/1), done.
 * [new tag]         annotated_tag -> annotated_tag
 * [new tag]         blob       -> blob
 * [new tag]         commit_tree -> commit_tree

git is noticing that the server did not automatically send the tags and is performing a second fetch. This second fetch is completely different from the first one

issue 中没有指定源文件或测试。首先定位 fetch 的实现及其现有的 fetch 测试，然后在服务器不会自动发送标签时，将初始 fetch 与 Git 后续的标签 fetch 进行比较。完成的标准是匹配所演示的两次 fetch 行为，并报告新获取的标签。