Message ID | 20210119212739.77882-1-rafaeloliveira.cs@gmail.com (mailing list archive) |
---|---|
Headers | show |
Series | teach `worktree list` verbose mode and prunable annotations | expand |
On Tue, Jan 19, 2021 at 4:28 PM Rafael Silva <rafaeloliveira.cs@gmail.com> wrote: > Changes in v3 > ============= > v3 is 1-patch bigger than v2 and it includes the following changes: Thanks. I gave v3 a complete read-through, and it is just about perfect. Aside from a very minor problem with the tests in [5/7] and [7/7] and the documentation enhancement (`core.quotePath`) suggested by Phillip, this is really, really close. (In fact, I would be happy to accept v3 as-is; those minor changes could easily be done on top.) I look forward to giving v4 my Reviewed-by:. > Range-diff against v2: > -: ---------- > 1: 66cd83ba42 worktree: libify should_prune_worktree() > 1: 9d47fcb4a4 ! 2: 81f4824028 worktree: teach worktree to lazy-load "prunable" reason The range-diff seems slightly off. The "libify should_prune_worktree()" patch existed in v1 and v2, but the range-diff suggests that it's new with v3. You probably just need a minor adjustment to your `git format-patch --range-diff` invocation.
In c57b3367be (worktree: teach `list` to annotate locked worktree,
2020-10-11) we taught `git worktree list` to annotate working trees that
is locked by appending "locked" text in order to signalize to the user
that a working tree is locked. During the review, there was some
discussion about additional annotations and information that `list`
command could provide to the user that has long been envisioned and
mentioned in [2], [3] and [4].
This patch series addresses some of these changes by teaching
`worktree list` to show "prunable" annotation, adding verbose mode and
extending the --porcelain format with prunable and locked annotation as
follow up from [1]. Additionally, it addresses one shortcoming for porcelain
format to escape any newline characters (LF or CRLF) and quote the lock
reason to prevent breaking the format that is mentioned in [4] and [1] during
the review cycle.
This patch series includes:
1. The first patch moves the should_prune_worktree() machinery to the top-level
worktree.c exposing the function as general API, that will be reference
by should_prune_worktree() wrapper implemented on the second patch.
The original idea was to not only move should_prune_worktree() but also
refactor to accept a "struct worktree" and load the information directly,
which allows simplify the `prune` command by reusing get_worktrees().
However this seems to also require refactoring get_worktrees() itself
to return "non-valid" working trees that can/should be pruned. This is
also mentioned in [5]. Having the wrapper function makes it easier to add
the prunable annotation without touching the get_worktrees() and the
other worktree sub commands. The refactoring can be addressed in a
future patch, if this turns out to be good idea. One possible approach
is to teach get_worktrees() to take additional flags that will tell
whether to return only valid or all worktrees in GIT_DIR/worktrees/
directory and address its own possible shortcoming.
2. Introduces the worktree_prune_reason() to discovery whether a worktree
is prunable along with two new fields in the `worktree` structure. The
function mimics the workree_lock_reason() API.
3. The third patch changes worktree_lock_reason() to be more gentle for
the main working tree to simply returning NULL instead of aborting the
program via assert() macro. This allow us to simplify the code that
checks if the working tree is locked for default and porcelain format.
This changes is also mentioned in [6].
4. Fix t2402 added in [1] to ensure the locked worktree is properly cleaned up.
5. The fourth patch adds the "locked" attribute for the porcelain format
in order to make both default and --porcelain format consistent.
6. The fifth patch introduces "prunable" annotation for both default
and --porcelain format.
7. The sixth patch introduces verbose mode to expand the `list` default
format and show each annotation reason when its available.
Series is built on top of 66e871b664 (The third batch, 2021-01-15).
[1] https://lore.kernel.org/git/20200928154953.30396-1-rafaeloliveira.cs@gmail.com/
[2] https://lore.kernel.org/git/CAPig+cQF6V8HNdMX5AZbmz3_w2WhSfA4SFfNhQqxXBqPXTZL+w@mail.gmail.com/
[3] https://lore.kernel.org/git/CAPig+cSGXqJuaZPhUhOVX5X=LMrjVfv8ye_6ncMUbyKox1i7QA@mail.gmail.com/
[4] https://lore.kernel.org/git/CAPig+cTitWCs5vB=0iXuUyEY22c0gvjXvY1ZtTT90s74ydhE=A@mail.gmail.com/
[5] https://lore.kernel.org/git/CACsJy8ChM99n6skQCv-GmFiod19mnwwH4j-6R+cfZSiVFAxjgA@mail.gmail.com/
[6] https://lore.kernel.org/git/xmqq8sctlgzx.fsf@gitster.c.googlers.com/
Changes in v4
=============
Thanks Eric Sunshine and Phillip Wood for the review and suggestions.
* Added documentation to explain that the lock reason is quoted with
the same rules as described for `core.quotePath`.
* The `worktree unlock` issued in the test cleanup is splitted and
executed after each `worktree lock` to ensure the unlock is only
done after we know each locked command was successful.
* Fix a couple of grammos in the [4/7] commit message.
Changes in v3
=============
v3 is 1-patch bigger than v2 and it includes the changes:
* Dropped CQUOTE_NODQ flag in the the locked reason to return a string
enclosed with double quotes if the text reason contains especial
characters, like newlines.
* fix in t2402 to ensure the locked worktree is properly cleaned up,
is move to its own patch.
* In worktree_prune_reason(), the `path` variable is initialize
with NULL to make the code easier to follow.
* The removal of `is_main_worktree()` before `worktree_lock_reason()`
is moved to the patch that actually changes the API to be more gentle
with the main worktree instead of refactoring the code in the next
patches that adds the annotations.
* Drop the `*reason` in `(reason && *reason)` as we know that
worktree_prune_reason() is either going to return a non-empty string
or NULL which makes the code easier to follow.
* The --verbose test for the locked annotation now tests that "locked"
annotation stays on the same line when there is no locked reason.
* Small documentation updates and make the "git worktree --verbose"
example a little consistent with the worktree path.
Changes in v2
=============
v2 changes considerably from v1 taking into account several comments
suggested by Eric Sunshine and Phillip Wood (thanks guys for the
insightful comments and patience reviewing my patches :) ).
* The biggest change are the way the series is organized. In v1 all the
code was implemented in the fourth patch, all the tests and documentation
updates was added in sixth and seventh patch respectfully, in v2 each
patch introduces the annotations and verbose mode together with theirs
respective test and documentation updates.
* Several rewrite of the commit messages
* In v1 the fifth patch was introducing a new function to escape newline
characters for the "locked" attribute. However, we already have this
feature from the quote.h API. So the new function is dropped and the
changes are squashed into v2's fourth patch.
* The new prunable annotation and locked annotation that was introduced
by [1] was refactor to not poke around the worktree private fields.
* Refactoring of the worktree_prune_reason() to cache the prune_reason
and refactor to early return the use cases following the more common
pattern used on Git's codebase.
* Few documentation rewrites, most notably the `--verbose` and `--expire`
doc sentences for the list command are moved to its own line to clearly
separate the description from the others commands instead of continuing
on the same paragraph.
* The `git unlock <worktree>` used in the test's cleanup is moved to after
we know the `git worktree locked` is executed successfully. This ensures
the unlock doesn't error in the cleanup stage which will make it harder
to debug the tests.
Rafael Silva (7):
worktree: libify should_prune_worktree()
worktree: teach worktree to lazy-load "prunable" reason
worktree: teach worktree_lock_reason() to gently handle main worktree
t2402: ensure locked worktree is properly cleaned up
worktree: teach `list --porcelain` to annotate locked worktree
worktree: teach `list` to annotate prunable worktree
worktree: teach `list` verbose mode
Documentation/git-worktree.txt | 74 ++++++++++++++++++++--
builtin/worktree.c | 110 +++++++++++----------------------
t/t2402-worktree-list.sh | 96 ++++++++++++++++++++++++++++
worktree.c | 91 ++++++++++++++++++++++++++-
worktree.h | 23 +++++++
5 files changed, 314 insertions(+), 80 deletions(-)
Range-diff against v3:
1: fc4252c129 = 1: 66cd83ba42 worktree: libify should_prune_worktree()
2: 664d84155b = 2: 81f4824028 worktree: teach worktree to lazy-load "prunable" reason
3: cf334da2f0 = 3: c62ecf60d6 worktree: teach worktree_lock_reason() to gently handle main worktree
4: ecb11793d9 ! 4: d2ea467927 t2402: ensure locked worktree is properly cleaned up
@@ Metadata
## Commit message ##
t2402: ensure locked worktree is properly cleaned up
- In c57b3367be (worktree: teach `list` to annotate locked worktree,
+ c57b3367be (worktree: teach `list` to annotate locked worktree,
2020-10-11) introduced a new test to ensure locked worktrees are listed
with "locked" annotation. However, the test does not clean up after
itself as "git worktree prune" is not going to remove the locked worktree
in the first place. This not only leaves the test in an unclean state it
- also potentially breaks following tests that relies on the
+ also potentially breaks following tests that rely on the
"git worktree list" output.
Let's fix that by unlocking the worktree before the "prune" command.
5: 1fb68562f6 ! 5: 1b7661a0b3 worktree: teach `list --porcelain` to annotate locked worktree
@@ Commit message
with its value depending whether the value is available or not. Thus
documenting the case of the new "locked" attribute.
+ Helped-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Helped-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Rafael Silva <rafaeloliveira.cs@gmail.com>
@@ Documentation/git-worktree.txt: worktree /path/to/other-linked-worktree
+branch refs/heads/locked-with-reason
+locked reason why is locked
+
++------------
++
++If the lock reason contains "unusual" characters such as newline, they
++are escaped and the entire reason is quoted as explained for the
++configuration variable `core.quotePath` (see linkgit:git-config[1]).
++For Example:
++
++------------
++$ git worktree list --porcelain
++...
++locked "reason\nwhy is locked"
++...
------------
EXAMPLES
@@ t/t2402-worktree-list.sh: test_expect_success '"list" all worktrees with locked
+ # unlocked worktree should not be annotated with "locked"
+ git worktree add --detach unlocked &&
+ git worktree lock locked1 &&
++ test_when_finished "git worktree unlock locked1" &&
+ git worktree lock locked2 --reason "with reason" &&
-+ test_when_finished "git worktree unlock locked1 && git worktree unlock locked2" &&
++ test_when_finished "git worktree unlock locked2" &&
+ git worktree list --porcelain >out &&
+ grep "^locked" out >actual &&
+ test_cmp expect actual
@@ t/t2402-worktree-list.sh: test_expect_success '"list" all worktrees with locked
+ git worktree add --detach locked_lf &&
+ git worktree add --detach locked_crlf &&
+ git worktree lock locked_lf --reason "$(printf "locked\nreason")" &&
++ test_when_finished "git worktree unlock locked_lf" &&
+ git worktree lock locked_crlf --reason "$(printf "locked\r\nreason")" &&
-+ test_when_finished "git worktree unlock locked_lf && git worktree unlock locked_crlf" &&
++ test_when_finished "git worktree unlock locked_crlf" &&
+ git worktree list --porcelain >out &&
+ grep "^locked" out >actual &&
+ test_cmp expect actual
6: 834584f826 = 6: c29ad7ffb1 worktree: teach `list` to annotate prunable worktree
7: 35ad456fbb ! 7: 7938950d04 worktree: teach `list` verbose mode
@@ t/t2402-worktree-list.sh: test_expect_success '"list" all worktrees with prunabl
+ git worktree add locked1 --detach &&
+ git worktree add locked2 --detach &&
+ git worktree lock locked1 &&
++ test_when_finished "git worktree unlock locked1" &&
+ git worktree lock locked2 --reason "with reason" &&
-+ test_when_finished "git worktree unlock locked1 && git worktree unlock locked2" &&
++ test_when_finished "git worktree unlock locked2" &&
+ echo "$(git -C locked2 rev-parse --show-toplevel) $(git rev-parse --short HEAD) (detached HEAD)" >expect &&
+ printf "\tlocked: with reason\n" >>expect &&
+ git worktree list --verbose >out &&
Eric Sunshine <sunshine@sunshineco.com> writes: > On Tue, Jan 19, 2021 at 4:28 PM Rafael Silva > <rafaeloliveira.cs@gmail.com> wrote: >> Changes in v3 >> ============= >> v3 is 1-patch bigger than v2 and it includes the following changes: > > Thanks. I gave v3 a complete read-through, and it is just about perfect. > > Aside from a very minor problem with the tests in [5/7] and [7/7] and > the documentation enhancement (`core.quotePath`) suggested by Phillip, > this is really, really close. (In fact, I would be happy to accept v3 > as-is; those minor changes could easily be done on top.) > > I look forward to giving v4 my Reviewed-by:. > Thanks again for reviewing the series. I've just sent the v4 and I'm looking forward to your review. >> Range-diff against v2: >> -: ---------- > 1: 66cd83ba42 worktree: libify should_prune_worktree() >> 1: 9d47fcb4a4 ! 2: 81f4824028 worktree: teach worktree to lazy-load "prunable" reason > > The range-diff seems slightly off. The "libify > should_prune_worktree()" patch existed in v1 and v2, but the > range-diff suggests that it's new with v3. You probably just need a > minor adjustment to your `git format-patch --range-diff` invocation. Thanks for pointing this out, I didn't notice this from the v3 range-diff.
On Wed, Jan 27, 2021 at 3:03 AM Rafael Silva <rafaeloliveira.cs@gmail.com> wrote: > Changes in v4 > * Added documentation to explain that the lock reason is quoted with > the same rules as described for `core.quotePath`. > * The `worktree unlock` issued in the test cleanup is splitted and > executed after each `worktree lock` to ensure the unlock is only > done after we know each locked command was successful. > * Fix a couple of grammos in the [4/7] commit message. I just gave v4 a complete read-through and it looks great. All review comments on previous rounds have been addressed, and I didn't find anything important to comment on in v4. Very nicely done. Please consider this entire series: Reviewed-by: Eric Sunshine <sunshine@sunshineco.com> Junio: Perhaps the summary of this series in the merge message can be refined a bit, as it only talks about verbose mode, but this series does a bit more than that. Heres a proposal: `git worktree list` now annotates worktrees as prunable, shows locked and prunable attributes in --porcelain mode, and gained a --verbose option.
Eric Sunshine <sunshine@sunshineco.com> writes: > On Wed, Jan 27, 2021 at 3:03 AM Rafael Silva > <rafaeloliveira.cs@gmail.com> wrote: >> Changes in v4 >> * Added documentation to explain that the lock reason is quoted with >> the same rules as described for `core.quotePath`. >> * The `worktree unlock` issued in the test cleanup is splitted and >> executed after each `worktree lock` to ensure the unlock is only >> done after we know each locked command was successful. >> * Fix a couple of grammos in the [4/7] commit message. > > I just gave v4 a complete read-through and it looks great. All review > comments on previous rounds have been addressed, and I didn't find > anything important to comment on in v4. Very nicely done. > > Please consider this entire series: > > Reviewed-by: Eric Sunshine <sunshine@sunshineco.com> > Thanks for the review and for all the insightful comments. > Junio: Perhaps the summary of this series in the merge message can be > refined a bit, as it only talks about verbose mode, but this series > does a bit more than that. Heres a proposal: > > `git worktree list` now annotates worktrees as prunable, shows > locked and prunable attributes in --porcelain mode, and gained > a --verbose option. I believe this message summarize nicely all the changes implemented by this series.
Eric Sunshine <sunshine@sunshineco.com> writes: > On Wed, Jan 27, 2021 at 3:03 AM Rafael Silva > <rafaeloliveira.cs@gmail.com> wrote: >> Changes in v4 >> * Added documentation to explain that the lock reason is quoted with >> the same rules as described for `core.quotePath`. >> * The `worktree unlock` issued in the test cleanup is splitted and >> executed after each `worktree lock` to ensure the unlock is only >> done after we know each locked command was successful. >> * Fix a couple of grammos in the [4/7] commit message. > > I just gave v4 a complete read-through and it looks great. All review > comments on previous rounds have been addressed, and I didn't find > anything important to comment on in v4. Very nicely done. > > Please consider this entire series: > > Reviewed-by: Eric Sunshine <sunshine@sunshineco.com> > > Junio: Perhaps the summary of this series in the merge message can be > refined a bit, as it only talks about verbose mode, but this series > does a bit more than that. Heres a proposal: > > `git worktree list` now annotates worktrees as prunable, shows > locked and prunable attributes in --porcelain mode, and gained > a --verbose option. Nicely done. Thanks for excellent contribution, both of you.