为什么一个 function 用于 MD5 hash 计算更适合较小的文件，但对于大文件效率低下？

Question

I am working on generating hash values for files as a means of disallowing duplicate files in a small database. As I was researching, I found the following thread: How to generate an MD5 checksum for a file in Android?

Why is the first answer "not efficient" for large files and it is best for small strings, whereas the answer provided by dentex is better-suited for large files? Is it because of the way the solution was programmed, or is there a caveat with MD5 hashing that I am unaware of?

Answer 1

MD5 generates a 128-bit digest.
SHA-1 generates a 160-bit digest.
SHA-2 generates a 224-, 256-, 384- or 512-bit digest.

More bits means more distinct values, means less likelihood of a two distinct inputs generating the same digest.