Package Details: doctest 2.4.1-1

Git Clone URL: (read-only, click to copy)
Package Base: doctest
Description: The lightest feature rich C++ single header testing framework
Upstream URL:
Keywords: c c++ framework testing unit
Licenses: MIT
Submitter: jetm
Maintainer: justicehunter
Last Packager: justicehunter
Votes: 7
Popularity: 0.69
First Submitted: 2016-05-22 20:38
Last Updated: 2020-11-11 09:21

Dependencies (1)

Required by (10)

Sources (1)

Pinned Comments

justicehunter commented on 2021-12-03 09:11

I guess I was too late to update, there is a version in community-testing

Latest Comments

1 2 Next › Last »

justicehunter commented on 2021-12-03 09:11

I guess I was too late to update, there is a version in community-testing

ysblokje commented on 2021-12-01 09:20

There's a better way of getting cmake to just build using whatever generator you have chosen. Use cmake --build <path>

justicehunter commented on 2020-07-27 19:33

@svenihoney: Done. Thank you for the suggestion.

svenihoney commented on 2020-07-17 10:39

If you stick to make as the build tool, please add -G"Unix Makefiles" to the PKGBUILD. I have set the CMAKE_GENERATOR=Ninja environment variable to use Ninja as default build tool, and with this set the package build fails. Thx.

justicehunter commented on 2020-07-10 16:31

@mmha: Tested it with arm, it works & changed the architecture to any. I am a newbie at managing packages, so sorry for spaghetti commits.

jetm commented on 2019-05-26 20:57

@mmha: I have updated with the PKGBUILD with your suggestion. Thank you.

I am going to hold on make the two recommendations. First, I don't know if this works in ARM; I unable to test it. It should work, though. 2nd point, this is a little package, not that much gain with ninja as a make replacement.


mmha commented on 2019-05-25 13:04

The installation fails with:

error: failed to commit transaction (conflicting files)
doctest: /usr/lib64 exists in filesystem (owned by filesystem)

This can be fixed by passing -DCMAKE_INSTALL_LIBDIR=lib to CMake.

I also have two recommendations:

  1. Set the arch to arch=('any') to make the package work on ARM, too.
  2. Use ninja instead of make. No one likes build log spaghetti from parallel builds. It's also slightly faster.

jetm commented on 2019-01-30 15:48

@muttleyxd: Updated the package with the new hash. I had the hash updated in my local repo, but never was submitted. Thank you for your perseverance.

muttleyxd commented on 2019-01-30 15:34

Well, I've cleaned it few times, I'm pretty sure it downloads correct file, I tried it both on my private and work PCs, maybe it got silently updated on GitHub?

$ git clone
Cloning into 'doctest'...
remote: Enumerating objects: 60, done.
remote: Counting objects: 100% (60/60), done.
remote: Compressing objects: 100% (46/46), done.
remote: Total 60 (delta 14), reused 60 (delta 14)
Unpacking objects: 100% (60/60), done.

$  cd doctest/
$  ls

$  makepkg
==> Making package: doctest 2.2.2-1 (Wed Jan 30 16:32:47 2019)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Downloading 2.2.2.tar.gz...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   121    0   121    0     0     40      0 --:--:--  0:00:03 --:--:--    40
100 2148k    0 2148k    0     0   519k      0 --:--:--  0:00:04 --:--:-- 3091k
==> Validating source files with sha256sums...
    2.2.2.tar.gz ... FAILED
==> ERROR: One or more files did not pass the validity check!

$ md5sum 2.2.2.tar.gz
5e569a10924fac5130dff4f3e246d87f  2.2.2.tar.gz

$ sha256sum 2.2.2.tar.gz 
4ea79f2a7d1ee268310e99d4e721fb83bcc63af5ee46583085612fcaf19df446  2.2.2.tar.gz

jetm commented on 2019-01-29 15:20

@muttleyxd: Unable to reproduce the issue. From my side it builds fine. You might want to clean your AUR cache using the tool you normally use to install packages from AUR.