Bug#1041847: man-db: ftbfs on all arch due to test failed

2023-07-24 Thread Bo YU
Hi!

On Mon, Jul 24, 2023 at 7:37 PM Colin Watson  wrote:
>
> On Mon, Jul 24, 2023 at 06:05:43PM +0800, Bo YU wrote:
> > The package build failed due to like:
>
...
> I can't easily do an upstream release immediately, because there are a
> number of changes that ought to go through a call for updated
> translations.  I'll cherry-pick the fix, though.

It was fixed quite fast, many thanks!

BR,
Bo
>
> --
> Colin Watson (he/him)  [cjwat...@debian.org]



Bug#1041847: man-db: ftbfs on all arch due to test failed

2023-07-24 Thread Colin Watson
On Mon, Jul 24, 2023 at 06:05:43PM +0800, Bo YU wrote:
> The package build failed due to like:

Ah, thanks - I somehow had it in my head that this only affected
autopkgtests so was a bit less urgent.

> I found the MR has fixed the issue:
> https://gitlab.com/man-db/man-db/-/issues/25

Yes, but this also fixed an issue on Darwin that isn't relevant to
Debian, so we only need one of the two commits there.

> This failure may block the official rebootstrap for riscv64. 
> So we can cherry-pick your change or upgrade the version of
> man-db(if upstream has plan) as soon as possible.

The upstream and Debian maintainers are the same person :-)

I can't easily do an upstream release immediately, because there are a
number of changes that ought to go through a call for updated
translations.  I'll cherry-pick the fix, though.

-- 
Colin Watson (he/him)  [cjwat...@debian.org]



Bug#1041847: man-db: ftbfs on all arch due to test failed

2023-07-24 Thread Bo YU
Package: man-db
Version: 2.11.2-2
Severity: important
Tags: ftbfs, patch
User: debian-ri...@lists.debian.org
Usertags: riscv64
X-Debbugs-Cc: debian-ri...@lists.debian.org

Dear Maintainer,

The package build failed due to like:

```
FAIL: man1/whatis.1
===

man -E UTF-8 -l ./man1/whatis.1 produced error output:
troff::34: warning [p 1, 3.2i]: cannot break line
troff::50: warning [p 1, 5.8i]: cannot break line
troff::55: warning [p 1, 6.7i]: cannot break line
troff::86: warning [p 2, 0.3i]: cannot break line
FAIL man1/whatis.1 (exit status: 1)

FAIL: man8/mandb.8
==
...
```

I found the MR has fixed the issue:
https://gitlab.com/man-db/man-db/-/issues/25

This failure may block the official rebootstrap for riscv64. 
So we can cherry-pick your change or upgrade the version of
man-db(if upstream has plan) as soon as possible.



-- 
Regards,
--
  Bo YU

diff -Nru man-db-2.11.2/debian/changelog man-db-2.11.2/debian/changelog
--- man-db-2.11.2/debian/changelog  2023-03-13 06:23:59.0 +0800
+++ man-db-2.11.2/debian/changelog  2023-07-24 17:40:58.0 +0800
@@ -1,3 +1,10 @@
+man-db (2.11.2-2.1) UNRELEASED; urgency=medium
+
+  * Non-maintainer upload.
+  * Cherry-pick commit to fix BTBFS issue in sid.(Closes: #-1)
+
+ -- Bo YU   Mon, 24 Jul 2023 17:40:58 +0800
+
 man-db (2.11.2-2) unstable; urgency=medium
 
   * Add debconf translations:
diff -Nru man-db-2.11.2/debian/.gitignore man-db-2.11.2/debian/.gitignore
--- man-db-2.11.2/debian/.gitignore 2023-03-13 06:23:59.0 +0800
+++ man-db-2.11.2/debian/.gitignore 1970-01-01 07:30:00.0 +0730
@@ -1,6 +0,0 @@
-build
-files
-man-db
-templates
-*.debhelper*
-*.substvars
diff -Nru man-db-2.11.2/debian/patches/fix-sid-ftbfs.patch 
man-db-2.11.2/debian/patches/fix-sid-ftbfs.patch
--- man-db-2.11.2/debian/patches/fix-sid-ftbfs.patch1970-01-01 
07:30:00.0 +0730
+++ man-db-2.11.2/debian/patches/fix-sid-ftbfs.patch2023-07-24 
17:39:52.0 +0800
@@ -0,0 +1,244 @@
+Description: fix ftbfs on sid 
+ Cherry-pick the commit from upstream: 
https://gitlab.com/man-db/man-db/-/issues/25 
+Last-Update: 2023-07-24
+---
+This patch header follows DEP-3: http://dep.debian.net/deps/dep3/
+--- a/configure.ac
 b/configure.ac
+@@ -420,6 +420,7 @@
+ AM_GNU_GETTEXT_VERSION([0.18.3])
+ AC_SUBST([LINGUAS])
+ AM_ICONV
++AC_SUBST([HAVE_ICONV], [$am_func_iconv])
+ MAN_PO4A
+ MAN_LINGUAS
+ 
+--- a/man/check-man
 b/man/check-man
+@@ -31,7 +31,7 @@
+ # Indeed, even for other languages we're going to get "cannot adjust line"
+ # if %manpath_config_file% expands to something long. Hmm. We'll just ignore
+ # this across the board for now.
+-errors="$(echo "$errors" | $EGREP -v "(cannot adjust line|can't break 
line)")" || true
++errors="$(echo "$errors" | $EGREP -v "(cannot adjust line|(can't|cannot break 
line)")" || true
+ 
+ if [ "$code" != 0 ]; then
+   echo "man -E UTF-8 -l $1 failed with exit status $code and error 
output:"
+--- a/src/tests/Makefile.am
 b/src/tests/Makefile.am
+@@ -23,7 +23,8 @@
+   MANDIR_LAYOUT=$(MANDIR_LAYOUT); export MANDIR_LAYOUT; \
+   abs_top_builddir=$(abs_top_builddir); export 
abs_top_builddir; \
+   OVERRIDE_DIR="$(override_dir)"; export OVERRIDE_DIR; \
+-  troff_is_groff=$(troff_is_groff); export troff_is_groff;
++  troff_is_groff=$(troff_is_groff); export troff_is_groff; \
++  HAVE_ICONV=$(HAVE_ICONV); export HAVE_ICONV;
+ # Each test must use the configure-detected shell, not necessarily /bin/sh.
+ AM_LOG_FLAGS = $(SHELL)
+ ALL_TESTS = \
+--- a/src/tests/manconv-coding-tags
 b/src/tests/manconv-coding-tags
+@@ -19,7 +19,11 @@
+ EOF
+ <"$tmpdir/1.exp" tail -n +2 | iconv -f UTF-8 -t ISO-8859-1 >>"$tmpdir/1.inp"
+ run $MANCONV -f UTF-8 -t UTF-8 <"$tmpdir/1.inp" >"$tmpdir/1.out"
+-expect_files_equal 'simple coding tag' "$tmpdir/1.exp" "$tmpdir/1.out"
++if [ "$HAVE_ICONV" = yes ]; then
++  expect_files_equal 'simple coding tag' "$tmpdir/1.exp" "$tmpdir/1.out"
++else
++  report_skip 'simple coding tag'
++fi
+ 
+ cat >"$tmpdir/2.exp" <<'EOF'
+ '\" -*- mode: troff; coding: UTF-8 -*-
+@@ -30,7 +34,12 @@
+ EOF
+ <"$tmpdir/2.exp" tail -n +2 | iconv -f UTF-8 -t ISO-8859-1 >>"$tmpdir/2.inp"
+ run $MANCONV -f UTF-8 -t UTF-8 <"$tmpdir/2.inp" >"$tmpdir/2.out"
+-expect_files_equal 'mode and coding tags' "$tmpdir/2.exp" "$tmpdir/2.out"
++if [ "$HAVE_ICONV" = yes ]; then
++  expect_files_equal 'mode and coding tags' "$tmpdir/2.exp" 
"$tmpdir/2.out"
++else
++  report_skip 'mode and coding tags'
++fi
++
+ 
+ cat >"$tmpdir/3.exp" <<'EOF'
+ '\" -*- mode: troff; coding: UTF-8 -*-
+@@ -41,7 +50,12 @@
+ EOF
+ <"$tmpdir/3.exp" tail -n +2 | iconv -f UTF-8 -t ISO-8859-1 >>"$tmpdir/3.inp"
+ run $MANCONV -f UTF-8 -t UTF-8 <"$tmpdir/3.inp" >"$tmpdir/3.out"
+-expect_files_equal 'iso-latin-1 coding alias' "$tmpdir/3.exp" "$tmpdir/3.out"
++if [ "$HAVE_ICONV" = yes ]; then
++