因为 git 是为源代码设计的，它默认的 diff 算法将一行视为最小的不可分割单元。
我正在尝试编辑一些在第 80 列自动换行的 Markdown 文件。添加一个句子可能会导致该段落的其余部分被标记为已更改。
有没有办法让 Git 使用更适合文本的差异算法？我需要一个将单词或句子视为不可分割的单元而不是行的工具吗？
最佳答案
您可以试试 git diff --word-diff。
$ git diff --word-diff
diff --git a/test.txt b/test.txt
index 54585bb..a8cd97e 100644
--- a/test.txt
+++ b/test.txt
@@ -1,7 +1,7 @@
Because git is designed for source code, its diff algorithms {+are bibbity +}
{+bobbity boo+} treat a line as the minimum indivisible unit. I am trying to edit
some markdown files that are word wrapped at column 80. Adding a sentence can
cause the rest of the paragraph to be marked as changed.
Is there a way to have Git use a diff algorithm more suited to text? One that
treats words or sentences as indivisible units rather then lines?
No newline at end of file
关于git - 在 Git 中使用替代差异算法，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/5999495/