Re: git diff: add option for omitting the contents of deletes

[Jeff King <peff@peff.net>]

On Tue, Mar 08, 2011 at 11:25:04PM +0200, Mart Sõmermaa wrote:

> "It also might make sense to apply the similar principle to shorten the output
> with -B when a rewrite patch is expressed as a single hunk patch that removes
> everything old and then adds everthing new."
> 
> I have to admit that I've never used -B, only -M, so please don't mind that
> its semantics and exact behavior are a bit foreign to me.
> 
> After running t/t4130-apply-criss-cross-rename.sh,
> `git diff -M -B` outputs the following:
> 
> diff --git a/file2 b/file1
> similarity index 100%
> rename from file2
> rename to file1
> diff --git a/file1 b/file2
> similarity index 100%
> rename from file1
> rename to file2
> 
> Can you bring a similar example of changes in the output after
> the above-mentioned similar principle has been implemented for -B?

Try:

  $ git init
  $ perl -e 'print "a\n" for (1 .. 1000)' >file
  $ git add file && git commit -m one
  $ perl -e 'print "b\n" for (1 .. 1000)' >file

Now you can see that -B breaks notes it as a rewrite:

  $ git diff --stat --summary -B
   file | 2000 +++++++++++++++++++++++++++----------------------------
   1 files changed, 1000 insertions(+), 1000 deletions(-)
   rewrite file (100%)

And the diff is long:

  $ git diff -B
  diff --git a/file b/file
  dissimilarity index 100%
  index 5cfafaa..a7b871a 100644
  --- a/file
  +++ b/file
  @@ -1,1000 +1,1000 @@
  -a
  -a
  [... x 1000]
  +b
  +b
  [... x 1000]

But we could perhaps drop the actual 1000-line hunks (or maybe even just
the deletion half).

-Peff