summaryrefslogtreecommitdiff
path: root/Debian
diff options
context:
space:
mode:
authorJoey Hess <joey@gnu.kitenet.net>2009-04-14 15:13:03 -0400
committerJoey Hess <joey@gnu.kitenet.net>2009-04-14 15:13:03 -0400
commit2a57fa03aefa2396ee49aa769f8cb0605ec3933b (patch)
treeb586c50fcf1b765e7d8dbc3686af7d36d7f2cdf7 /Debian
parentdee85c67c6730518bc8f7dc29e1d557a9f2718af (diff)
more comments
Diffstat (limited to 'Debian')
-rw-r--r--Debian/Debhelper/Dh_Buildsystem_Basic.pm16
1 files changed, 15 insertions, 1 deletions
diff --git a/Debian/Debhelper/Dh_Buildsystem_Basic.pm b/Debian/Debhelper/Dh_Buildsystem_Basic.pm
index 58d073b..b6eec4b 100644
--- a/Debian/Debhelper/Dh_Buildsystem_Basic.pm
+++ b/Debian/Debhelper/Dh_Buildsystem_Basic.pm
@@ -3,6 +3,9 @@
# Copyright: © 2008-2009 Modestas Vainius
# License: GPL-2+
+# XXX JEH maybe rename this class to Debian::Debhelper::Dh_Buildsystem?
+# XXX JEH also it seems the functions in Dh_Buildsystems could be merged
+# into this same file.
package Debian::Debhelper::Dh_Buildsystem_Basic;
use strict;
@@ -51,11 +54,18 @@ sub DEFAULT_BUILD_DIRECTORY {
# is_auto_buildable in the constructor, and only have the constructor
# succeed if it can handle the source? That would also eliminate the
# delayed warning mess in enforce_in_source_building.
+#
+# (In turn that could be used to remove the pre_action, since that's the
+# only use of it -- the post_action is currently unused too. It could be
+# argued that these should be kept in case later buildsystems need them
+# though.)
+#
# AFAICS, there is only one reason you need an instance of the object
# if it can't build -- to list build systems. But that only needs
# DESCRIPTION and NAME, which could be considered to be class methods,
# rather than object methods -- no need to construct an instance of the
# class before calling those.
+#
# I see that if --buildsystem is manually specified to override,
# the is_auto_buildable test is completely skipped. So if this change were
# made, you'd not be able to skip the test, and some --buildsystem choices
@@ -157,6 +167,7 @@ sub get_rel2builddir_path {
sub _mkdir {
my ($cls, $dir)=@_;
+ # XXX JEH is there any reason not to just doit("mkdir") ?
if (-e $dir && ! -d $dir) {
error("error: unable to create '$dir': object already exists and is not a directory");
}
@@ -172,6 +183,8 @@ sub _mkdir {
sub _cd {
my ($cls, $dir)=@_;
+ # XXX JEH I think this should verbose_print("cd $dir")
+ # then the verbose_prints in doit_in_builddir are unnecessary.
if (! $dh{NO_ACT}) {
chdir $dir or error("error: unable to chdir to $dir");
}
@@ -180,7 +193,7 @@ sub _cd {
# Creates a build directory. Returns 1 if the directory was created
# or 0 if it already exists or there is no need to create it.
sub mkdir_builddir {
- my $self=shift;
+ my $self=shift;a
if ($self->get_builddir()) {
return $self->_mkdir($self->get_builddir());
}
@@ -209,6 +222,7 @@ sub doit_in_builddir {
# In case of outside-source tree building, whole build directory
# gets wiped (if it exists) and 1 is returned. Otherwise, nothing
# is done and 0 is returned.
+# XXX JEH only makefile.pm uses this, move it there?
sub clean_builddir {
my $self=shift;
if ($self->get_builddir()) {