This is the scripts needed to split the Apache Brooklyn repository into smaller modules, as agreed by vote on the community's dev mailing list.
To try it out:
git clone https://github.com/ahgittin/brooklyn-repo-split.git
cd brooklyn-repo-split
git submodule init
git submodule update
./all.sh
Then look for the results, all the new repos in new-repos,
carved up including carved up history
Running this script can take a couple of hours. You can run parts of it; see below.
The file env.sh sets local environment variables. Adjust to suit your individual identity.
The script runs several stages, as indicated by the commands 1-*.sh, 2-*.sh, ... .
For the most part these can be run individually.
There are a few other scripts *.sh and programs *.rb used by the scripts above.
The files *-whitelist.txt files list the initial whitelists to use for each project.
If any projects are restructured these should be kept consistent with script 1-rearrange-incubator.sh.
Files *.gen.txt are auto-generated.
In particular the files *-whitelist.full.gen.txt contain the final expanded whitelists,
auto-generated by step 3.
These are checked in as it is useful to cache these and to observe changes
(but these may need updating after any change to the incubator project).
The files big-to-small.* are a script and output of a program which records the biggest files in the history.
Step 1 is done and applied to incubator-brooklyn, and now skipped as part of all.sh.
Step 2 is required locally at least once.
Step 3 has been run and the results checked in here (*-whitelist.full.gen.txt).
Step 4
Steps 2-4 will need re-run as more changes are pushed to incubator.
The steps are now pretty quick, with 1 taking <1m, 2 taking 20m, 3 taking 3m, on my box. Step 4 takes the longest, about 15m per sub-repo or 1h30m total.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent