Don't use fallback file_rule to clean real targets

3 files changed, 22 insertions, 3 deletions

diff --git a/libbuild2/algorithm.cxx b/libbuild2/algorithm.cxx
index e15fcbc..9b6fd4e 100644
--- a/libbuild2/algorithm.cxx
+++ b/libbuild2/algorithm.cxx
@@ -967,9 +967,21 @@ namespace build2
         //    not seem like it will be easy to fix (we don't know whether
         //    someone else will execute this target).
         //
-        // @@ What if we always do match & execute together? After all,
+        //    What if we always do match & execute together? After all,
         //    if a group can be resolved in apply(), then it can be
-        //    resolved in match()!
+        //    resolved in match()! Feels a bit drastic.
+        //
+        //    But, this won't be a problem if the target returns noop_recipe.
+        //    And perhaps it's correct to fail if it's not noop_recipe but
+        //    nobody executed it? Maybe not.
+        //
+        //    Another option would be to have a count for such "matched but
+        //    may not be executed" targets and then make sure target_count
+        //    is less than that at the end. Though this definitelt makes it
+        //    less exact (since we can end up executed this target but not
+        //    some other). Maybe we can increment and decrement such targets
+        //    in a separate count (i.e., mark their recipe as special or some
+        //    such).
         //
 
         // Apply (locked).
diff --git a/libbuild2/rule.cxx b/libbuild2/rule.cxx
index 6dad685..c573339 100644
--- a/libbuild2/rule.cxx
+++ b/libbuild2/rule.cxx
@@ -56,10 +56,13 @@ namespace build2
     // are not doing anything for this action so not checking if the file
     // exists seems harmless.
     //
+    // But we also don't want to match real targets and not cleaning their
+    // output files.
+    //
     switch (a)
     {
     case perform_clean_id:
-      return true;
+      return t.decl != target_decl::real;
     default:
       {
         // While normally we shouldn't do any of this in match(), no other
diff --git a/libbuild2/target.hxx b/libbuild2/target.hxx
index 76c01b6..809bf8b 100644
--- a/libbuild2/target.hxx
+++ b/libbuild2/target.hxx
@@ -129,6 +129,10 @@ namespace build2
   // @@ We have cases (like pkg-config extraction) where it should probably be
   //    prereq_file rather than implied (also audit targets.insert<> calls).
   //
+  // @@ Also, synthesized dependency declarations (e.g., in cc::link_rule) are
+  //    fuzzy: they feel more `real` than `implied`. Maybe introduce
+  //    `synthesized` in-between?
+  //
   enum class target_decl: uint8_t
   {
     prereq_new,  // Created from prerequisite (create_new_target()).