git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Request for change 610e2b9240 reversal

From
René Scharfe <l.s.r@web.de>
Date
Nov 10, 2020, 11:38 UTC
Message-ID
<171b9a99-d48a-babe-168b-a7cd410e4635@web.de>
In-Reply-To
<CA+phgpFS_MOKOib5+yAE7U2QPMqMD_t+yXhR29UG9aSO4pgk+A@mail.gmail.com>
Am 08.11.20 um 12:15 schrieb Jean-Yves Avenard:
Show 17 quoted lines
> Hi
>
> At Mozilla we use a mercurial repository, however many developers are
> using a git repository; we maintain two git mirrors, mozilla-unified
> accessed via the cinnabar plugin and a geck-dev native git mirror.
>
> Our repositories contain a .git-blame-ignore-revs that is used for
> both repositories.
> https://searchfox.org/mozilla-central/source/.git-blame-ignore-revs
>
> That git was ignoring invalid entries (for the currently in use repo)
> is a requirement for our use.
> Following the merge of 610e2b9240  jc/blame-ignore-fix later we have
> lost the ability to run git blame on any of our files.
>
> Could we get that change reversed?
> If it ain't broken, don't fix it as they say.
That looks like one tidily maintained ignore list!

I saw that coming ([1], [2]), but didn't communicate my doubts clearly enough and undermined them when I wrote that I'd run into the same issue due to laziness. So I feel some responsibility for your trouble. Your actual use case is not lazy -- it uses the fundamental fact that Git object IDs are practically unique worldwide, without synchronization between repositories, i.e. they form a single flat namespace.

610e2b9240 (blame: validate and peel the object names on the ignore list, 2020-09-24) introduced two changes: Support for peeling tags and rejection of non-commits after peeling. Here's a patch for silently ignoring such non-commits read from a file.

René

[1] https://public-inbox.org/git/40488753-c179-4ce2-42d0-e57b5b1ec6cd@web.de/ [2] https://public-inbox.org/git/1fa730c4-eaef-2f32-e1b4-716a27ed4646@web.de/

-- >8 --
Subject: [PATCH] blame: silently ignore invalid ignore file objects

Since 610e2b9240 (blame: validate and peel the object names on the ignore list, 2020-09-24) git blame reports checks if objects specified with --ignore-rev and in files loaded with --ignore-revs-file and config option blame.ignoreRevsFile are actual objects and dies if they aren't. The intent is to report typos to the user.

This also breaks the ability to use a single ignore file for multiple repositories. Typos are presumably less likely in files than on the command line, so alerting is less useful here. Restore that feature by skipping non-commits without dying.

Reported-by: Jean-Yves Avenard <jyavenard@mozilla.com>
Signed-off-by: René Scharfe <l.s.r@web.de>
---
 oidset.c                     | 5 +++--
 t/t8013-blame-ignore-revs.sh | 4 ++--
 2 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/oidset.c b/oidset.c
index 2d0ab76fb5..5aac633c1f 100644
--- a/oidset.c
+++ b/oidset.c
@@ -72,9 +72,10 @@ void oidset_parse_file_carefully(struct oidset *set, const char *path,
 		if (!sb.len)
 			continue;

-		if (parse_oid_hex(sb.buf, &oid, &p) || *p != '\0' ||
-		    (fn && fn(&oid, cbdata)))
+		if (parse_oid_hex(sb.buf, &oid, &p) || *p != '\0')
 			die("invalid object name: %s", sb.buf);
+		if (fn && fn(&oid, cbdata))
+			continue;
 		oidset_insert(set, &oid);
 	}
 	if (ferror(fp))
diff --git a/t/t8013-blame-ignore-revs.sh b/t/t8013-blame-ignore-revs.sh
index 24ae5018e8..b18633dee1 100755
--- a/t/t8013-blame-ignore-revs.sh
+++ b/t/t8013-blame-ignore-revs.sh
@@ -39,10 +39,10 @@ test_expect_success 'validate --ignore-rev' '
 	test_must_fail git blame --ignore-rev X^{tree} file
 '

-# Ensure bogus --ignore-revs-file requests are caught
+# Ensure bogus --ignore-revs-file requests are silently accepted
 test_expect_success 'validate --ignore-revs-file' '
 	git rev-parse X^{tree} >ignore_x &&
-	test_must_fail git blame --ignore-revs-file ignore_x file
+	git blame --ignore-revs-file ignore_x file
 '

 for I in X XT
--
2.29.2
Previous: Jean-Yves AvenardNext: Barret Rhoden
Message 2 of 4 in “Request for change 610e2b9240 reversal”
  1. Jean-Yves AvenardNov 8, 2020
  2. René ScharfeNov 10, 2020
  3. Barret RhodenNov 10, 2020
  4. Junio C HamanoNov 10, 2020

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.