new groff incompatibilities? #99

sten0 · 2023-08-15T20:59:07Z

Hi,

I'm a user of the Debian package (2.0.2+ds1-1), and lintian recently started forwarding these groff warnings:

groff-message an.tmac::508: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

groff-message troff::113: warning: cannot select font 'C'

I'm guessing there was some sort of change in groff that go-md2man now needs to adapt to.

'hope it's an easy fix!

paravoid · 2023-08-17T15:45:15Z

I believe these are artifacts of the groff 1.23 release, so it'll affect all distributions soon. I don't have any insight myself, but several packages of mine have been affected, and I've been following along the bug reports and mailing list threads.

groff-message an.tmac::508: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

This actually results in malformed output, tables not rendered. Thankfully the fix is easy, as described here: add '\" t as the very first line of the output:
See https://lists.debian.org/debian-devel/2023/08/msg00220.html

I tested this with the crun manpage.

groff-message troff::113: warning: cannot select font 'C'

Debian bugs #1041809 and #1043256 provide some context. Basically, go-md2man generates things like \fB\fCstart\fR, and \fC is not valid. There is an alternative described there, to use .EX/.EE, but notably this breaks the groff -Thtml output (which is reported upstream). I don't know if there is an alternative that is able to produce the exact same output in HTML, but without the new warnings.

paravoid · 2023-08-17T23:42:09Z

I reported this in the Debian specifically for the go-md2man package, as bug #1049968, and Cc'ed G. Branden Robinson (groff upstream, among other affiliations), who gave a very long and elaborate response, including specific recommendations for the go-md2man code. You may want to look into it and perhaps talk to him before implementing any fixes here.

g-branden-robinson · 2023-09-11T21:48:25Z

Can I be of any assistance?

thaJeztah · 2023-09-11T22:46:32Z

Contributions are welcome for sure 😅 (looks like I missed the original report, and just now saw this ticket). Also /cc @cpuguy83

It's difficult to portably ask the man(7) macro package for a monospaced font (whether it's referred to as C, CW or CR) without any extra configuration. That's what leads to: warning: cannot select font 'C' This can be reproduced on Fedora 39, which has GNU roff 1.23, with: $ man crun >/dev/null troff:<standard input>:25: warning: cannot select font 'C' troff:<standard input>:134: warning: cannot select font 'C' ... $ man podman >/dev/null troff:<standard input>:15: warning: cannot select font 'C' troff:<standard input>:25: warning: cannot select font 'C' ... $ man toolbox >/dev/null troff:<standard input>:33: warning: cannot select font 'C' troff:<standard input>:43: warning: cannot select font 'C' ... These may look like odd uses of man(1), but it's used by the Toolbx [1] test suite to ensure that 'toolbox help' and 'toolbox --help' do render the toolbox(1) manual, which, like crun(1) and podman(1), is generated by go-md2man(1). The warnings come from the current \\fB\\fC sequence for inline code enclosed in single backticks (ie., `) that uses 'C' to ask for a monospace font. Other than the difficulties with portably asking the man(7) macro package for a monospace font, the \\fB\\fC sequence has some oddities. If a monospace font does get selected by \\fC, then the request for bold with \\fB gets masked, with the possible surprise that switching back to the 'previous font' with \\fP will result in bold. This is what happens when the output is in HyperText Markup Language or PostScript. If the output is a terminal device, \\fC is a NOP because terminals are always monospaced and can't change the font family, so only \\fB is in effect. Therefore, the use of \\fC is only relevant for non-terminal outputs, like HTML and PS. Since HTML and PS use serif fonts by default, an inline switch to monospace looks visually jarring. This is different from other common uses of monospace for inline code, such as on GitLab and GitHub, where sans-serif fonts are used by default. Therefore, just don't attempt to use monospace for inline code enclosed in single backticks (ie., `) and stick to bold everywhere. As suggested by G. Branden Robinson. [1] https://containertoolbx.org/ https://github.com/containers/toolbox cpuguy83#99 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1049968

debarshiray · 2023-10-10T12:14:10Z

Contributions are welcome for sure 😅 (looks like I missed the original report, and just now saw this ticket). Also /cc @cpuguy83

I put something together in #105 based on the conversation between @paravoid and @g-branden-robinson on the Debian bug. I checked the HTML and PS outputs with man -Thtml ... and man -Tps respectively.

It's difficult to portably ask the man(7) macro package for a monospaced font (whether it's referred to as C, CW or CR) without any extra configuration. That's what leads to: warning: cannot select font 'C' This can be reproduced on Fedora 39, which has GNU roff 1.23, with: $ man crun >/dev/null troff:<standard input>:25: warning: cannot select font 'C' troff:<standard input>:134: warning: cannot select font 'C' ... $ man podman >/dev/null troff:<standard input>:15: warning: cannot select font 'C' troff:<standard input>:25: warning: cannot select font 'C' ... $ man toolbox >/dev/null troff:<standard input>:33: warning: cannot select font 'C' troff:<standard input>:43: warning: cannot select font 'C' ... These may look like odd uses of man(1), but it's used by the Toolbx [1] test suite to ensure that 'toolbox help' and 'toolbox --help' do render the toolbox(1) manual, which, like crun(1) and podman(1), is generated by go-md2man(1). The warnings come from the current \\fB\\fC sequence for inline code enclosed in single backticks (ie., `) that uses 'C' to ask for a monospace font. Other than the difficulties with portably asking the man(7) macro package for a monospace font, the \\fB\\fC sequence has some oddities. If a monospace font does get selected by \\fC, then the request for bold with \\fB gets masked, with the possible surprise that switching back to the 'previous font' with \\fP will result in bold. This is what happens when the manual is rendered in HyperText Markup Language or PostScript: $ man --troff-device html <manual> > foo.html $ man --troff-device ps <manual> > foo.ps If the manual is displayed on a terminal device, \\fC is a NOP because terminals are always monospaced and can't change the font family, so only \\fB is in effect. Therefore, the use of \\fC is only relevant for non-terminal outputs, like HTML and PS. Since HTML and PS use serif fonts by default, an inline switch to monospace looks visually jarring. This is different from other common uses of monospace for inline code, such as on GitLab and GitHub, where sans-serif fonts are used by default. So, just don't attempt to use monospace for inline code enclosed in single backticks (ie., `) and stick to bold everywhere. As suggested by G. Branden Robinson. [1] https://containertoolbx.org/ https://github.com/containers/toolbox cpuguy83#99 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1049968

It's difficult to portably ask the man(7) macro package for a monospaced font (whether it's referred to as C, CW or CR) without any extra configuration. That's what leads to: warning: cannot select font 'C' This can be reproduced on Fedora 39, which has GNU roff 1.23, with: $ man crun >/dev/null troff:<standard input>:25: warning: cannot select font 'C' troff:<standard input>:134: warning: cannot select font 'C' ... $ man podman >/dev/null troff:<standard input>:15: warning: cannot select font 'C' troff:<standard input>:25: warning: cannot select font 'C' ... $ man toolbox >/dev/null troff:<standard input>:33: warning: cannot select font 'C' troff:<standard input>:43: warning: cannot select font 'C' ... These may look like odd uses of man(1), but it's used by the Toolbx [1] test suite to ensure that 'toolbox help' and 'toolbox --help' do render the toolbox(1) manual, which, like crun(1) and podman(1), is generated by go-md2man(1). The warnings come from the current \\fB\\fC sequence for inline code enclosed in single backticks (ie., `) that uses 'C' to ask for a monospace font. Other than the difficulties with portably asking the man(7) macro package for a monospace font, the \\fB\\fC sequence has some oddities. If a monospace font does get selected by \\fC, then the request for bold with \\fB gets masked, with the possible surprise that switching back to the 'previous font' with \\fP will result in bold. This is what happens when the manual is rendered in HyperText Markup Language or PostScript: $ man --troff-device html <manual> > foo.html $ man --troff-device ps <manual> > foo.ps If the manual is displayed on a terminal device, \\fC is a NOP because terminals are always monospaced and can't change the font family. So, only \\fB is in effect. Therefore, the use of \\fC is only relevant for non-terminal outputs, like HTML and PS. Since HTML and PS use serif fonts by default, an inline switch to monospace looks visually jarring. This is different from other common uses of monospace for inline code, such as on GitLab and GitHub, where sans-serif fonts are used by default. Hence, just don't attempt to use monospace for inline code enclosed in single backticks (ie., `) and stick to bold everywhere. As suggested by G. Branden Robinson. [1] https://containertoolbx.org/ https://github.com/containers/toolbox cpuguy83#99 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1049968

Ansible's built-in 'package' module doesn't show any details when installing the RPMs. All that can be seen is: TASK [Install RPM packages] fedora-rawhide | changed Therefore, there's no way to know what version of the packages got installed. In this case, not knowing the go-md2man(1) version being used by the CI makes it difficult to know why the tests are failing on Fedora Rawhide and Fedora 39 with: not ok 3 help: Command 'help' in 177ms # (from function `assert_line' in file test/system/libs/bats-assert/src/assert.bash, line 479, # in test file test/system/002-help.bats, line 48) # `assert_line --index 0 --partial "toolbox(1)"' failed # /usr/bin/man # # -- line does not contain substring -- # index : 0 # substring : toolbox(1) # line : troff:<standard input>:33: warning: cannot select font 'C' # -- # It could be either because the CI is still using an older version of go-md2man(1) [1,2], or that there's some other problem. [1] Fedora golang-github-cpuguy83-md2man commit 117806d50e401c19 https://src.fedoraproject.org/rpms/golang-github-cpuguy83-md2man/c/117806d50e401c19 https://src.fedoraproject.org/rpms/golang-github-cpuguy83-md2man/pull-request/3 [2] go-md2man commit d85280db9b54b574 cpuguy83/go-md2man@d85280db9b54b574 cpuguy83/go-md2man#99 containers#1386

paravoid · 2023-10-27T12:30:15Z

Thanks for working on this! I just tested go-md2man 2.0.3 with crun's manpage and:

groff-message an.tmac::508: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

This actually results in malformed output, tables not rendered. Thankfully the fix is easy, as described here: add '\" t as the very first line of the output: See https://lists.debian.org/debian-devel/2023/08/msg00220.html

This is not fixed.

groff-message troff::113: warning: cannot select font 'C'

This is fixed.

debarshiray · 2023-11-02T15:36:20Z

groff-message an.tmac::508: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

This actually results in malformed output, tables not rendered. Thankfully the fix is easy, as described here: add '\" t as the very first line of the output: See https://lists.debian.org/debian-devel/2023/08/msg00220.html

This is not fixed.

Yes, sorry, my bad. I should have been more explicit.

Part of the problem is that I am unable to reproduce the tbl preprocessor failed warning with crun(1). I tried with crun 1.9.2, which was built with GNU roff 1.23, on Fedora 39.

I could try to come up with a fix if you gave me some more explicit hints on how to reproduce.

groff-message troff::113: warning: cannot select font 'C'

This is fixed.

I couldn't reproduce this one easily, either, but it was failing my test suite, which was doing something similar to man toolbox | head or man toolbox | grep ... and running into the warnings.

paravoid · 2023-11-02T15:48:42Z

groff-message an.tmac::508: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

This actually results in malformed output, tables not rendered. Thankfully the fix is easy, as described here: add '\" t as the very first line of the output: See https://lists.debian.org/debian-devel/2023/08/msg00220.html

This is not fixed.

Yes, sorry, my bad. I should have been more explicit.

Part of the problem is that I am unable to reproduce the tbl preprocessor failed warning with crun(1). I tried with crun 1.9.2, which was built with GNU roff 1.23, on Fedora 39.

I could try to come up with a fix if you gave me some more explicit hints on how to reproduce.

On a fresh Debian sid system, e.g. by running podman run --rm -it debian:sid:

root@1fd0128f4302:/# apt update
root@1fd0128f4302:/# apt install -y go-md2man man-db wget
root@1fd0128f4302:/# wget https://raw.githubusercontent.com/containers/crun/main/crun.1.md
root@1fd0128f4302:/# go-md2man -in crun.1.md -out crun.1
root@1fd0128f4302:/# LC_ALL=C.UTF-8 MANROFFSEQ='' MANWIDTH=80 man --warnings -E UTF-8 -l -Tutf8 -Z crun.1 >/dev/null
an.tmac:<standard input>:844: warning: tbl preprocessor failed, or it or soelim was not run; table(s) likely not rendered (TE macro called with TW register undefined)

thaJeztah · 2023-11-02T20:17:49Z

Let me reopen this issue at least if there's things left to fix (for visibility)

paravoid · 2023-12-24T13:39:17Z

Any news here? Thanks :)

cpuguy83 · 2024-02-04T23:42:43Z

@paravoid Something like this? #111

debarshiray mentioned this issue Oct 10, 2023

Silence "warning: cannot select font 'C'" with GNU roff 1.23 #105

Merged

cpuguy83 closed this as completed in #105 Oct 10, 2023

debarshiray mentioned this issue Oct 10, 2023

test/system: Various improvements to the 'help' tests containers/toolbox#1386

Merged

This was referenced Oct 11, 2023

update go-md2man to v2.0.3 docker/cli#4592

Merged

[24.0 backport] update go-md2man to v2.0.3 docker/cli#4594

Merged

[23.0 backport] update go-md2man to v2.0.3 docker/cli#4595

Merged

debarshiray mentioned this issue Oct 12, 2023

test/system: Unbreak the line count checks with Bats >= 1.10.0 containers/toolbox#1387

Merged

thaJeztah reopened this Nov 2, 2023

Luap99 mentioned this issue Nov 27, 2023

f39: make docs: warning: cannot select font 'C' containers/podman#20787

Open

cpuguy83 mentioned this issue Feb 4, 2024

Prepend table preprocessor #111

Merged

cpuguy83 closed this as completed in #111 Feb 26, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

new groff incompatibilities? #99

new groff incompatibilities? #99

sten0 commented Aug 15, 2023

paravoid commented Aug 17, 2023 •

edited

paravoid commented Aug 17, 2023

g-branden-robinson commented Sep 11, 2023

thaJeztah commented Sep 11, 2023

debarshiray commented Oct 10, 2023

paravoid commented Oct 27, 2023

debarshiray commented Nov 2, 2023

paravoid commented Nov 2, 2023 •

edited

thaJeztah commented Nov 2, 2023

paravoid commented Dec 24, 2023

cpuguy83 commented Feb 4, 2024

new groff incompatibilities? #99

new groff incompatibilities? #99

Comments

sten0 commented Aug 15, 2023

paravoid commented Aug 17, 2023 • edited

paravoid commented Aug 17, 2023

g-branden-robinson commented Sep 11, 2023

thaJeztah commented Sep 11, 2023

debarshiray commented Oct 10, 2023

paravoid commented Oct 27, 2023

debarshiray commented Nov 2, 2023

paravoid commented Nov 2, 2023 • edited

thaJeztah commented Nov 2, 2023

paravoid commented Dec 24, 2023

cpuguy83 commented Feb 4, 2024

paravoid commented Aug 17, 2023 •

edited

paravoid commented Nov 2, 2023 •

edited