make order of lines in top_level.txt deterministic

without this, every build produces a different result, as the lines are
ordered randomly.

This makes builds reproducible.
If you build a package two times, you get the same result, as you would
expect. Basically the same thing was already done for requirements and
entry_points.txt

It is useful e.g. if you want to rebuild a package that you downloaded,
to ensure that no-one tampered with it, or even if you just want to
examine the differences between two builds/versions of one package.

1 files changed, 1 insertions, 1 deletions

diff --git a/setuptools/command/egg_info.py b/setuptools/command/egg_info.py
index 72493d0b..06764a17 100755
--- a/setuptools/command/egg_info.py
+++ b/setuptools/command/egg_info.py
@@ -389,7 +389,7 @@ def write_toplevel_names(cmd, basename, filename):
             for k in cmd.distribution.iter_distribution_names()
         ]
     )
-    cmd.write_file("top-level names", filename, '\n'.join(pkgs) + '\n')
+    cmd.write_file("top-level names", filename, '\n'.join(sorted(pkgs)) + '\n')
 
 
 def overwrite_arg(cmd, basename, filename):