Git tracking block of code moved/removed in a file

Question

I have an old commit md5hash on myfile.extension with SOME CHANGE in the commit body (not the commit title/metadata).

How can I generate a list of commits with SOME CHANGE modified (not just present) in the commits leading up to HEAD from md5hash without having to inspect each diff? (of which there are unfortunately many in the current case.)

I tried git rev-list --all | xargs git grep 'SOME CHANGE' but this seems to find all the commits with SOME CHANGE which is in the file.

git blame seems useless as the lines have changed and SOME CHANGE has moved.

Ciro

Answer 1

I think the answer you're looking for is git --no-pager log --pretty="%H" -G"SOME CHANGE" -- myfile.extension.

At first I thought of git log -S but it only covers add/remove. git log -G would probably be close to what you want. Here you can see the difference between -S and -G, and I've included the full commit history as well so you can see what's not covered. Read the commit messages for a description of what I did in the body.

# git --no-pager log --oneline -S"SOME CHANGE"
12e24ed Remove text
9427ffc Add the text
# git --no-pager log --oneline -G"SOME CHANGE"
12e24ed Remove text
6a33653 Change other text on same line
ac09bbb Append other text to same line
484b447 Move the text two lines down
9427ffc Add the text
# git --no-pager log --oneline
12e24ed Remove text
9c7f7d5 Change text on adjacent line
6a33653 Change other text on same line
ac09bbb Append other text to same line
484b447 Move the text two lines down
377936f Add other text on adjacent line
9427ffc Add the text
1929648 Initial commit

To get it with just the hashes:

# git --no-pager log --pretty="%H" -G"SOME CHANGE"
12e24ed749e499bc2d8920c5d8a3ca98a6422e3f
6a336532210ca85dea86968c34cef516345b8ab4
ac09bbb5c95bbea65e7c99c730653d27f90397f4
484b4478e4cb16c839dac558f3e958683b428a64
9427ffc7dd60a3cfb1d9880083e6262faea0eefb

Decarburize

Answer 2

I think you are looking for:

git log -S<string>

Look for differences that change the number of occurrences of the specified string (i.e. addition/deletion) in a file. Intended for the scripter’s use.

It is useful when you’re looking for an exact block of code (like a struct), and want to know the history of that block since it first came into being: use the feature iteratively to feed the interesting block in the preimage back into -S, and keep going until you get the very first version of the block.

Pleasurable

Answer 3

This would still require a bit of wading through results, but this may get you closer:

git rev-list --all | xargs git show | egrep '(^ ?[+-].*(SOME CHANGE)|^commit)' | egrep -B1 '^ ?[+-]' | uniq

This shows the patch for each commit, shows the sha it's looking at, and then spits out the line of code that matches. The relevant sha for the match is above the match.

You could further the pipeline to cleanup the results even more if this is a viable way to go.

Mcclenaghan