Monitoring commits on all branches
Marc Branchaud
marcnarc at gmail.com
Mon Nov 23 19:43:36 UTC 2020
On 2020-11-22 8:47 p.m., Dan Langille wrote:
>> On Nov 22, 2020, at 7:58 PM, Dan Langille <dan at langille.org> wrote:
>>
>> Today I managed to sit down and go through the multiple scripts involved in pulling in a commit.
>>
>> https://news.freshports.org/2020/11/22/git-commit-processing-how-is-it-done/
>>
>> Some changes were made....
>>
>>> On Nov 21, 2020, at 2:06 PM, Marc Branchaud <marcnarc at gmail.com> wrote:
>>>
>>> (trimmed some of the quoting)
>>>
>>> On 2020-11-21 9:11 a.m., Dan Langille wrote:
>>>> On Thu, Nov 19, 2020, at 5:30 PM, Marc Branchaud wrote:
>>>>> Watch out with that "git checkout branch" part: This is where a lot of
>>>>> people get tripped up when they move from svn to git. It doesn't help
>>>>> that git tries to do some hand-holding here, but really git users would
>>>>> benefit from simply understanding how branch names are just labels for
>>>>> commit SHA IDs. Without that understanding, people end up going down
>>>>> the rabbit-hole that is the "git pull" command, a wretched hive of scum
>>>>> and villainy if there ever was one.
>>>> I understand the concern and I think I follow. Within the confines of
>>>> 'only wanting the files for reference, never local modification', pull and
>>>> fetch might be the same for me. I'll look at fetch more.
>>>
>>> "git fetch" is just the "get updates from the remote repo" part of "git pull".
>>>
>>> The thing with pull is that it wants to "help" you update your local branches the "right" way. But the "right" way really depends on how you work and how your project works. So most advice to "do a 'git pull'" is only useful for the most general of cases.
>>
>> Commits on the quarterly branch are now being automatically processed.
Glad to hear you've made some progress!
>> Now that the code is not doing 'git checkout master', which made sense on master, things are going better.
>>
>> git fetch
>> git checkout master
>> git rebase $REMOTE/master
>>
>> Now its just doing:
>>
>> git pull
>> git rev-list <HASH>..HEAD
>>
>> This seems to be working.
>>
>> However, that may change if I start doing local tags.
>
> I just tried this.
>
> While I understand the issue around
>
> [dan at devgit-ingress01:~/src/freebsd/freebsd-ports] $ git fetch
> remote: Enumerating objects: 220, done.
> remote: Counting objects: 100% (194/194), done.
> remote: Compressing objects: 100% (60/60), done.
> remote: Total 111 (delta 60), reused 100 (delta 49), pack-reused 0
> Receiving objects: 100% (111/111), 18.38 KiB | 4.59 MiB/s, done.
> Resolving deltas: 100% (60/60), completed with 36 local objects.
> From https://github.com/freebsd/freebsd-ports
> 24a927a0ef5d..bba44fef62c0 master -> origin/master
> e3989b2f380e..9944c9cd2c35 branches/2020Q4 -> origin/branches/2020Q4
> 7c20845bd227..430121ca5ac7 svn_head -> origin/svn_head
>
>
> OK, I see how I might be able to use that output, I can see which branches have new commits.
>
>
> [dan at devgit-ingress01:~/src/freebsd/freebsd-ports] $ git merge
> Updating 24a927a0ef5d..bba44fef62c0
> Fast-forward
FYI, the "Fast-forward" here is git's way of saying "The branch you
currently have checked-out did not add any new commits compared to the
remote repo's version of the branch, so I'm just going to make your
checked-out branch point to the same commit as on the remote repo."
A "fast-forward" merge means that git did not do any patch-like
operations to combine different sets of code changes.
> audio/hpsjam/Makefile | 2 +-
> audio/hpsjam/distinfo | 6 +-
> devel/Makefile | 1 +
> devel/php80-intl/files/patch-icu68 | 222 ----------------------------------------
> devel/py-pytest-datadir/Makefile | 27 +++++
> devel/py-pytest-datadir/distinfo | 3 +
> devel/py-pytest-datadir/pkg-descr | 4 +
> graphics/tiled/Makefile | 7 +-
> graphics/tiled/distinfo | 6 +-
> graphics/tiled/pkg-plist | 4 +
> irc/irssi/Makefile | 81 +++++++--------
> irc/irssi/distinfo | 6 +-
> irc/irssi/files/patch-Makefile.in | 22 ----
> irc/irssi/files/patch-configure.ac | 15 ---
> irc/irssi/files/patch-meson.build | 22 ++++
> irc/irssi/files/patch-perl-Makefile | 22 ----
> irc/irssi/pkg-plist | 10 +-
> lang/php80/Makefile | 2 +-
> lang/php80/distinfo | 6 +-
> mail/mutt/Makefile | 2 +-
> mail/mutt/distinfo | 6 +-
> multimedia/mpv/Makefile | 33 +++---
> multimedia/mpv/distinfo | 10 +-
> multimedia/mpv/pkg-plist | 4 +-
> net-im/signal-cli/Makefile | 13 +--
> net-im/signal-cli/distinfo | 14 +--
> net-im/signal-cli/pkg-plist | 4 +-
> net/fb303/Makefile | 16 ++-
> net/fb303/distinfo | 6 +-
> net/fb303/files/patch-build_fbcode__builder_CMake_FBThriftLibrary.cmake | 2 +-
> net/fb303/files/patch-fb303_thrift_CMakeLists.txt | 22 ----
> net/fb303/pkg-plist | 4 +-
> net/geoipupdate/Makefile | 23 +++--
> net/geoipupdate/distinfo | 18 +++-
> net/geoipupdate/files/patch-Makefile | 13 ++-
> net/netatalk3/Makefile | 2 +-
> net/netatalk3/files/patch-etc_afpd_volume.c | 20 ++++
> net/netatalk3/files/patch-libatalk_vfs_extattr.c | 19 ++++
> sysutils/rpi-firmware/Makefile | 4 +-
> sysutils/rpi-firmware/distinfo | 6 +-
> sysutils/rpi-firmware/pkg-plist | 4 +-
> 41 files changed, 265 insertions(+), 448 deletions(-)
> delete mode 100644 devel/php80-intl/files/patch-icu68
> create mode 100644 devel/py-pytest-datadir/Makefile
> create mode 100644 devel/py-pytest-datadir/distinfo
> create mode 100644 devel/py-pytest-datadir/pkg-descr
> delete mode 100644 irc/irssi/files/patch-Makefile.in
> delete mode 100644 irc/irssi/files/patch-configure.ac
> create mode 100644 irc/irssi/files/patch-meson.build
> delete mode 100644 irc/irssi/files/patch-perl-Makefile
> delete mode 100644 net/fb303/files/patch-fb303_thrift_CMakeLists.txt
> create mode 100644 net/netatalk3/files/patch-etc_afpd_volume.c
> create mode 100644 net/netatalk3/files/patch-libatalk_vfs_extattr.c
>
> That output might also be useful for knowing what ports are affected, but really, this is what brings everything in.
You can get also that output when comparing any two commits in the repo,
with "git diff --stat" (or "git diff --numstat" for a more parsable format):
git diff --numstat 24a927a0ef5d origin/branches/2020Q4
> '24a927a0ef5d' is taken from the 'git fetch' output but would also be the LAST_COMMIT for this branch.
>
> [dan at devgit-ingress01:~/src/freebsd/freebsd-ports] $ git rev-list 24a927a0ef5d..HEAD
> bba44fef62c062bf942dd5312128b43dbd813dcf
> 6a1e09447b223d078f78f7a7a9072e9842343c76
> dffe7ea4b193785b7309dd161e76e9d626e12b96
> f1a85af087281cfc6cd117301d7932472bd51adc
> 2cbf3d698dc5e2d265c8f58447f5dea3e0a9b7ee
> e9e1f60c3603041bcca9dda1e95918dbfcd3e480
> 352ce94919051329b591ef3534541f3138e52be7
> 0140eb70289386ad0d2b5a660ae1108a2c6eed45
> e273657ca09b1d2c862b24eb1154789f24e69ea6
> 7a79ec535067cd83fde3b8800509680cf1e60a27
> 182883a91022b14581cd951a733fd7ca1538ba73
> 494e0eafb5c57aadf36e6ef911f922169152cf1a
> c1ff2a4159b9fd8d2199bb908a4facfb9b0e6399
>
>
> Next, I tried a tag, but went over to the freebsd source repo, because it has more recent commits.
>
> Command output is not shown.
>
> [dan at devgit-ingress01:~/src/freebsd/freebsd] $ git tag freshports_last_commit
> [dan at devgit-ingress01:~/src/freebsd/freebsd] $ git fetch
> [dan at devgit-ingress01:~/src/freebsd/freebsd] $ git merge
> [dan at devgit-ingress01:~/src/freebsd/freebsd] $ git rev-list freshports_last_commit..HEAD
> 5902bf651a44a052969477e3ed4ee0c382dd7b90
> a0e591191ab3d906d2d3d1a6985d424513c69c70
> 8a81e9904d4c3bf58328345f958f5fa5ac3d0801
> 9c6918af8b7927e77d829b7226c456415c029bdf
> 6006818f9acd67f5b45d262ec949f0403b34100d
> bdbb0128dce50974d8140ad511a1a895db205ece
> bf39ccedb1130d2041d937ab6bfa0e6d32f10b68
> df7e32b0f26fd7500b7c1360071ed323fb5b7639
> 236ac76160e57a74747dc4453961fe5bbd18d7fb
> 4b194b2e1ccbeb53c7b1a22169fdbcb79ef985ea
> e8a0e692d11de506b18de046891293a5d03f15e2
> 45011cba53a65bbcaeb91dba13798d49387f34b3
> 2af377ccd66ed48f2ac8c3677fddfd6ad2608a7b
> d97705ee7cb3e6405dd60a1d6f03d9737754593a
> dbc3ee07f871a067e0d72d3463992370879a4372
> 7f2206c5a9f89d5e47cce05b79cfeead2b6598f5
> 2241a9b4d5b35fb0fb9ea0eb0f87dc7f0463edc3
> b766790d1f5c8304463537f537a6955f1e0df442
> cbc574bf4544f824705626383f9bef33d9a26089
> 3cc7c8abf47a2558fa0f1179905814ee167257b9
> 5844c0bcff59fd35dc2e0906bb696b17289e5e5c
> e6e213967ba710191f3d5d912c9ecf70d7e26b8a
> 1dab84551f51bf73e9653559dd7c4aff547719b4
> 50db1dd05447122f5d2c1f2b60e9bab771294bea
>
> After the last commit is processed, this gets done:
>
> $ git tag freshports_last_commit
> fatal: tag 'freshports_last_commit' already exists
>
> Oops:
>
> $ git tag -f freshports_last_commit
> Updated tag 'freshports_last_commit' (was 4a90ed0de4e)
>
> A compromise might be just logging the last commit somewhere so it can be derived easily if all is lost.
>
> Mind you, this might be the same thing.
>
> SELECT hash FROM commit_log order by id desc limit 1
>
> .. assuming everything gets inserted into the database in the correct order.
>
> That might work. Just go back a few dozens commits and use that value. If we reprocess commits, no bit deal. They get ignored if already in the database.
You should do whatever works for you, of course. Storing your script's
state within your git repo is just one option.
M.
More information about the freebsd-git
mailing list