bfg

I'm struggling on how can I remove values from a column that are different only on the sign. For example:

...

I was trying to create a new view out of 3 or 4 other views/tables.

TableA:

title_id homeTeam 1234 WSV 5678 SSV 7890 NULL 4321 SCC

TableB:

title_id awayTeam 1234 SSV 5678 SFV 7890 NULL 4321 KFC

TableC:

title_id homeTeam 1234 SSV 5678 NULL 7890 AAB 4711 BFG

I would like to generate a new view out of those three which looks like:

title_id Teams 1234 WSV, SSV, SSV 5678 SSV, SFV, N/A 7890 N/A, N/A, AAB 4321 SCC, KFC, N/A 4711 N/A, N/A, BFG

As you can see, NULL should be renamed to N/A, as well if id doesn't exist in one of the other tables. And I would like to get DISTINCT title_id.

@DaleK sure, i tried it like this:

select tableA.title_id, ISNULL(tableA.homeTeam, 'N/A') + ',' + ISNULL(tableB.awayTeam, 'N/A') + ',' + ISNULL(tableC.homeTeam, 'N/A') as Teams from tableA, tableB, tableC;

This leads into an neverending query for my tables which has each more than 300k rows.

Next i tried join:

select tableA.title_id, ISNULL(tableA.homeTeam, 'N/A') + ',' + ISNULL(tableB.awayTeam, 'N/A') + ',' + ISNULL(tableC.homeTeam, 'N/A') as Teams from tableA FULL JOIN tableB ON tableB.title_id = tableA.title_id FULL JOIN tableC ON tableC.title_id = tableA.tile_id

But to be honest i wasnt sure about the number of rows.

...

I have a BitBucket repository that shall be migrated to Github. It is using git lfs for large files and has a few corrupt lfs objects, which hinders the pushing to github (now the origin in below command). The push always returns the error Your push referenced at least X unknown Git LFS objects.

I'm now wondering if I can either ignore the pre-receive hook or drop/delete the lfs objects in question all together. I don't care much about these files, because they are quite old already and by now not even in the repo anymore.

...

We are trying to shrink our git repository to under 500MB due to deployment issues.

To achieve that, we have created a new branch where we have moved all old images, videos and fonts to AWS S3.

I can easily get the list of files with git diff --name-only --diff-filter=D master -- public/assets/.

Now, I have tried to run BFG-repo-cleaner 1.14.0 on each file. But I have 400 files and it is taking ages to delete each files separately (still running as I'm writing this).

git diff --name-only --diff-filter=D master -- public/assets/ | xargs -i basename '{}' | xargs -i bfg --delete-files '{}'

Since each file is distinct, I can not really use a glob pattern, as suggested at Delete multiple files from multiple branch using bfg repo cleaner.

I tried to separate each file with a comma but that resulted in BFG-repo-cleaner telling me:

BFG aborting: No refs to update - no dirty commits found??

Is there a way to provide multiple files to BFG-repo-cleaner without a glob pattern?

PS. The command I tried with multiple files is: git diff --name-only --diff-filter=D master -- public/assets/ | xargs -i basename '{}' | sed -z 's/\n/,/g;s/,$/\n/' | xargs -i bfg --delete-files '{}' && git reflog expire --expire=now --all && git gc --prune=now --aggressive

PPS. The bfg command is on my PATH as a simple bash script with java -jar /tools/BFG-repo-cleaner/bfg-1.14.0.jar "$@"

...

Following my previous question about a way to clear 'hidden' large files in gitlab repo. I proceeded as follows:

Clone repo using mirror flag:

...

suppose I have been given strings "abc", "def" and "ghi", I want to generate all the possible combination of word generated by picking from these strings. for eg for "abc", "def" and "ghi"

we should get

"adg","adh","adi","aeg","aeh","aei","afg","afh","afi", "bdg","bdh","bdi","beg","beh","bei","bfg","bfh","bfi", "cdg","cdh","cdi","ceg","ceh","cei","cfg","cfh","cfi"

How to Do it.

my attampt...

...

I have a repo which is 1.4Gb, altho the actual code is more like 250Mb. I'm trying to use bfg to get the size down.

When running git clone --mirror https://xxx@bitbucket.org/xxx/my-app.git

I'm getting:

...

All answers I see regarding this question (here, here) only ever offer ways to either change the most recent commit(s) or all commits, however I want to change the author of only past commits that match a certain predicate. I have looked at git filter-branch (deprecated), git-filter-repo (its replacement), and bfg, but can't figure out how to make this change.

The way I intuitively tried to approach this is:

1. First find all commits that I'm trying to change.
2. Change author of those commits.

I have other commits in the history with the same author that I do not want to change, I only want to change commits that match a certain predicate.

So for step 1, find all commits whose commit messages have a specific substring:

...