_cachekey: Remove the 'node_sanitization' done before the json string

Since ujson already sorts the keys, we just need to be able to jsonify
Nodes sensibly. This is done by implementing __json__ on the 'Node' base
class.

This does not break the cache keys.

7 files changed, 26 insertions, 7 deletions

diff --git a/src/buildstream/_cachekey.py b/src/buildstream/_cachekey.py
index e56b582fa..89d47671e 100644
--- a/src/buildstream/_cachekey.py
+++ b/src/buildstream/_cachekey.py
@@ -22,7 +22,6 @@ import hashlib
 
 import ujson
 
-from . import _yaml
 
 # Internal record of the size of a cache key
 _CACHEKEY_SIZE = len(hashlib.sha256().hexdigest())
@@ -63,6 +62,5 @@ def is_key(key):
 #    (str): An sha256 hex digest of the given value
 #
 def generate_key(value):
-    ordered = _yaml.node_sanitize(value)
-    ustring = ujson.dumps(ordered, sort_keys=True, escape_forward_slashes=False).encode('utf-8')
+    ustring = ujson.dumps(value, sort_keys=True, escape_forward_slashes=False).encode('utf-8')
     return hashlib.sha256(ustring).hexdigest()
diff --git a/src/buildstream/_context.py b/src/buildstream/_context.py
index 60bc9ae9b..97956cc6f 100644
--- a/src/buildstream/_context.py
+++ b/src/buildstream/_context.py
@@ -429,7 +429,7 @@ class Context():
         if self._cache_key is None:
 
             # Anything that alters the build goes into the unique key
-            self._cache_key = _cachekey.generate_key(_yaml.new_empty_node())
+            self._cache_key = _cachekey.generate_key({})
 
         return self._cache_key
 
diff --git a/src/buildstream/_project.py b/src/buildstream/_project.py
index cc57a16fa..72cf6306a 100644
--- a/src/buildstream/_project.py
+++ b/src/buildstream/_project.py
@@ -230,7 +230,7 @@ class Project():
 
             # Anything that alters the build goes into the unique key
             # (currently nothing here)
-            self._cache_key = _cachekey.generate_key(_yaml.new_empty_node())
+            self._cache_key = _cachekey.generate_key({})
 
         return self._cache_key
 
diff --git a/src/buildstream/_yaml.pxd b/src/buildstream/_yaml.pxd
index 48bed1dea..45d2d351a 100644
--- a/src/buildstream/_yaml.pxd
+++ b/src/buildstream/_yaml.pxd
@@ -28,6 +28,7 @@ cdef class Node:
     cdef public int column
 
     cpdef Node copy(self)
+    cpdef object strip_node_info(self)
 
     cdef bint _shares_position_with(self, Node target)
     cdef bint _walk_find(self, Node target, list path) except *
diff --git a/src/buildstream/_yaml.pyx b/src/buildstream/_yaml.pyx
index d68f64772..82262f60b 100644
--- a/src/buildstream/_yaml.pyx
+++ b/src/buildstream/_yaml.pyx
@@ -29,6 +29,7 @@ from collections.abc import Mapping, Sequence
 from copy import deepcopy
 
 from ruamel import yaml
+
 from ._exceptions import LoadError, LoadErrorReason
 
 
@@ -83,6 +84,12 @@ cdef class Node:
     cpdef Node copy(self):
         raise NotImplementedError()
 
+    cpdef object strip_node_info(self):
+        raise NotImplementedError()
+
+    def __json__(self):
+        raise ValueError("Nodes should not be allowed when jsonify-ing data", self)
+
 
 cdef class ScalarNode(Node):
 
@@ -132,6 +139,9 @@ cdef class ScalarNode(Node):
             return None
         return str(self.value)
 
+    cpdef object strip_node_info(self):
+        return self.value
+
     cdef bint _walk_find(self, Node target, list path) except *:
         return self._shares_position_with(target)
 
@@ -270,6 +280,12 @@ cdef class MappingNode(Node):
     cpdef object values(self):
         return self.value.values()
 
+    cpdef object strip_node_info(self):
+        cdef str key
+        cdef Node value
+
+        return {key: value.strip_node_info() for key, value in self.value.items()}
+
     def __delitem__(self, str key):
         del self.value[key]
 
@@ -362,6 +378,10 @@ cdef class SequenceNode(Node):
     cpdef list as_str_list(self):
         return [node.as_str() for node in self.value]
 
+    cpdef object strip_node_info(self):
+        cdef Node value
+        return [value.strip_node_info() for value in self.value]
+
     cdef bint _walk_find(self, Node target, list path) except *:
         cdef int i
         cdef Node v
diff --git a/src/buildstream/element.py b/src/buildstream/element.py
index c49a56d09..3030bccc5 100644
--- a/src/buildstream/element.py
+++ b/src/buildstream/element.py
@@ -2166,7 +2166,7 @@ class Element(Plugin):
                 'environment': cache_env,
                 'sources': [s._get_unique_key(workspace is None) for s in self.__sources],
                 'workspace': '' if workspace is None else workspace.get_key(self._get_project()),
-                'public': self.__public,
+                'public': self.__public.strip_node_info(),
                 'cache': 'CASCache'
             }
 
diff --git a/tests/elements/filter/basic/element_plugins/dynamic.py b/tests/elements/filter/basic/element_plugins/dynamic.py
index eb462ceb1..1ec7d4dd3 100644
--- a/tests/elements/filter/basic/element_plugins/dynamic.py
+++ b/tests/elements/filter/basic/element_plugins/dynamic.py
@@ -11,7 +11,7 @@ class DynamicElement(Element):
         pass
 
     def get_unique_key(self):
-        return {'split-rules': self.split_rules}
+        return {'split-rules': self.split_rules.strip_node_info()}
 
     def configure_sandbox(self, sandbox):
         pass