_artifact.py: Make cache query explicit

Cache query can be fairly expensive as it checks the presence of all
artifact blobs. Make this more explicit with a `query_cache()` method,
instead of implicitly querying the cache on the first call to
`cached()`.

3 files changed, 25 insertions, 22 deletions

diff --git a/src/buildstream/_artifact.py b/src/buildstream/_artifact.py
index d4a716ff1..b63cff6e5 100644
--- a/src/buildstream/_artifact.py
+++ b/src/buildstream/_artifact.py
@@ -540,7 +540,7 @@ class Artifact:
 
         return dependency_refs
 
-    # cached():
+    # query_cache():
     #
     # Check whether the artifact corresponding to the stored cache key is
     # available. This also checks whether all required parts of the artifact
@@ -550,11 +550,7 @@ class Artifact:
     # Returns:
     #     (bool): Whether artifact is in local cache
     #
-    def cached(self):
-
-        if self._cached is not None:
-            return self._cached
-
+    def query_cache(self):
         context = self._context
 
         artifact = self._load_proto()
@@ -587,6 +583,18 @@ class Artifact:
         self._cached = True
         return True
 
+    # cached()
+    #
+    # Return whether the artifact is available in the local cache. This must
+    # be called after `query_cache()` or `set_cached()`.
+    #
+    # Returns:
+    #     (bool): Whether artifact is in local cache
+    #
+    def cached(self):
+        assert self._cached is not None
+        return self._cached
+
     # cached_logs()
     #
     # Check if the artifact is cached with log files.
@@ -600,15 +608,6 @@ class Artifact:
         # If the artifact is cached, its log files are available as well.
         return self._element._cached()
 
-    # reset_cached()
-    #
-    # Allow the Artifact to query the filesystem to determine whether it
-    # is cached or not.
-    #
-    def reset_cached(self):
-        self._proto = None
-        self._cached = None
-
     # set_cached()
     #
     # Mark the artifact as cached without querying the filesystem.
diff --git a/src/buildstream/_frontend/widget.py b/src/buildstream/_frontend/widget.py
index 0d5379f43..626611dac 100644
--- a/src/buildstream/_frontend/widget.py
+++ b/src/buildstream/_frontend/widget.py
@@ -357,6 +357,8 @@ class LogLine(Widget):
                 else:
                     if element.get_kind() == "junction":
                         line = p.fmt_subst(line, "state", "junction", fg="magenta")
+                    elif not element._can_query_cache():
+                        line = p.fmt_subst(line, "state", "waiting", fg="blue")
                     elif element._cached_failure():
                         line = p.fmt_subst(line, "state", "failed", fg="red")
                     elif element._cached_success():
diff --git a/src/buildstream/element.py b/src/buildstream/element.py
index 945ca5cf7..0f2a01ced 100644
--- a/src/buildstream/element.py
+++ b/src/buildstream/element.py
@@ -1179,9 +1179,6 @@ class Element(Plugin):
     #            the artifact cache
     #
     def _cached(self):
-        if not self.__artifact:
-            return False
-
         return self.__artifact.cached()
 
     # _cached_remotely():
@@ -1297,7 +1294,7 @@ class Element(Plugin):
     #
     def _can_query_cache(self):
         # cache cannot be queried until strict cache key is available
-        return self.__strict_cache_key is not None
+        return self.__artifact is not None
 
     # _initialize_state()
     #
@@ -1585,7 +1582,9 @@ class Element(Plugin):
     def __should_schedule(self):
         # We're processing if we're already scheduled, we've
         # finished assembling or if we're waiting to pull.
-        processing = self.__assemble_scheduled or self.__assemble_done or self._pull_pending()
+        processing = (
+            self.__assemble_scheduled or self.__assemble_done or (self._can_query_cache() and self._pull_pending())
+        )
 
         # We should schedule a build when
         return (
@@ -1651,7 +1650,7 @@ class Element(Plugin):
             self.__artifact.set_cached()
             self.__cached_successfully = True
         else:
-            self.__artifact.reset_cached()
+            self.__artifact.query_cache()
 
         # When we're building in non-strict mode, we may have
         # assembled everything to this point without a strong cache
@@ -1898,7 +1897,7 @@ class Element(Plugin):
 
         # Artifact may become cached after pulling, so let it query the
         # filesystem again to check
-        self.__artifact.reset_cached()
+        self.__artifact.query_cache()
 
         # We may not have actually pulled an artifact - the pull may
         # have failed. We might therefore need to schedule assembly.
@@ -2570,6 +2569,7 @@ class Element(Plugin):
             return None
 
         artifact = Artifact(self, self._get_context(), strong_key=workspace.last_build)
+        artifact.query_cache()
 
         if not artifact.cached():
             return None
@@ -3279,12 +3279,14 @@ class Element(Plugin):
             strict_key=self.__strict_cache_key,
             weak_key=self.__weak_cache_key,
         )
+        strict_artifact.query_cache()
         if context.get_strict() or strict_artifact.cached():
             self.__artifact = strict_artifact
         else:
             self.__artifact = Artifact(
                 self, context, strict_key=self.__strict_cache_key, weak_key=self.__weak_cache_key
             )
+            self.__artifact.query_cache()
 
         if not context.get_strict() and self.__artifact.cached():
             # In non-strict mode, strong cache key becomes available when