_yaml: Remove 'node_get_provenance' and add 'Node.get_provenance'

This replaces the helper method by adding a 'get_provenance' on the node
directly

- Adapt all call sites
- Delay getting provenance wherever possible without major refactor

24 files changed, 115 insertions, 121 deletions

diff --git a/src/buildstream/_basecache.py b/src/buildstream/_basecache.py
index 64ee8fe57..56f6d68dc 100644
--- a/src/buildstream/_basecache.py
+++ b/src/buildstream/_basecache.py
@@ -79,7 +79,7 @@ class BaseCache():
             try:
                 artifacts = config_node.get_sequence(cls.config_node_name, default=[])
             except LoadError:
-                provenance = _yaml.node_get_provenance(config_node, key=cls.config_node_name)
+                provenance = config_node.get_node(cls.config_node_name).get_provenance()
                 raise _yaml.LoadError(_yaml.LoadErrorReason.INVALID_DATA,
                                       "%s: 'artifacts' must be a single 'url:' mapping, or a list of mappings" %
                                       (str(provenance)))
diff --git a/src/buildstream/_cas/casremote.py b/src/buildstream/_cas/casremote.py
index dac92aa2a..f0c84f7b6 100644
--- a/src/buildstream/_cas/casremote.py
+++ b/src/buildstream/_cas/casremote.py
@@ -8,7 +8,6 @@ import uuid
 
 import grpc
 
-from .. import _yaml
 from .._protos.google.rpc import code_pb2
 from .._protos.google.bytestream import bytestream_pb2, bytestream_pb2_grpc
 from .._protos.build.bazel.remote.execution.v2 import remote_execution_pb2, remote_execution_pb2_grpc
@@ -35,7 +34,7 @@ class CASRemoteSpec(namedtuple('CASRemoteSpec', 'url push server_cert client_key
         url = spec_node.get_str('url')
         push = spec_node.get_bool('push', default=False)
         if not url:
-            provenance = _yaml.node_get_provenance(spec_node, 'url')
+            provenance = spec_node.get_node('url').get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: empty artifact cache URL".format(provenance))
 
@@ -54,12 +53,12 @@ class CASRemoteSpec(namedtuple('CASRemoteSpec', 'url push server_cert client_key
             client_cert = os.path.join(basedir, client_cert)
 
         if client_key and not client_cert:
-            provenance = _yaml.node_get_provenance(spec_node, 'client-key')
+            provenance = spec_node.get_node('client-key').get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: 'client-key' was specified without 'client-cert'".format(provenance))
 
         if client_cert and not client_key:
-            provenance = _yaml.node_get_provenance(spec_node, 'client-cert')
+            provenance = spec_node.get_node('client-cert').get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: 'client-cert' was specified without 'client-key'".format(provenance))
 
diff --git a/src/buildstream/_context.py b/src/buildstream/_context.py
index 38ea239c6..c84817f1a 100644
--- a/src/buildstream/_context.py
+++ b/src/buildstream/_context.py
@@ -492,9 +492,10 @@ class Context():
 #    LoadError, when the value is not of the expected type, or is not found.
 #
 def _node_get_option_str(node, key, allowed_options):
-    result = node.get_str(key)
+    result_node = node.get_scalar(key)
+    result = result_node.as_str()
     if result not in allowed_options:
-        provenance = _yaml.node_get_provenance(node, key)
+        provenance = result_node.get_provenance()
         raise LoadError(LoadErrorReason.INVALID_DATA,
                         "{}: {} should be one of: {}".format(
                             provenance, key, ", ".join(allowed_options)))
diff --git a/src/buildstream/_includes.py b/src/buildstream/_includes.py
index 4c1eabb7e..678f24dba 100644
--- a/src/buildstream/_includes.py
+++ b/src/buildstream/_includes.py
@@ -43,7 +43,6 @@ class Includes:
             else:
                 includes = includes_node.as_str_list()
 
-            include_provenance = _yaml.node_get_provenance(node, key='(@)')
             del node['(@)']
 
             for include in reversed(includes):
@@ -53,6 +52,7 @@ class Includes:
                     include_node, file_path, sub_loader = self._include_file(include,
                                                                              current_loader)
                 except LoadError as e:
+                    include_provenance = includes_node.get_provenance()
                     if e.reason == LoadErrorReason.MISSING_FILE:
                         message = "{}: Include block references a file that could not be found: '{}'.".format(
                             include_provenance, include)
@@ -65,6 +65,7 @@ class Includes:
                         raise
 
                 if file_path in included:
+                    include_provenance = includes_node.get_provenance()
                     raise LoadError(LoadErrorReason.RECURSIVE_INCLUDE,
                                     "{}: trying to recursively include {}". format(include_provenance,
                                                                                    file_path))
diff --git a/src/buildstream/_loader/loader.py b/src/buildstream/_loader/loader.py
index e45381fbf..27975dc34 100644
--- a/src/buildstream/_loader/loader.py
+++ b/src/buildstream/_loader/loader.py
@@ -463,7 +463,7 @@ class Loader():
             return meta_element
 
         node = element.node
-        elt_provenance = _yaml.node_get_provenance(node)
+        elt_provenance = node.get_provenance()
         meta_sources = []
 
         sources = node.get_sequence(Symbol.SOURCES, default=[])
diff --git a/src/buildstream/_loader/types.pyx b/src/buildstream/_loader/types.pyx
index cd206cfb4..fe1cea789 100644
--- a/src/buildstream/_loader/types.pyx
+++ b/src/buildstream/_loader/types.pyx
@@ -69,7 +69,7 @@ cdef class Dependency:
                  str default_dep_type=None):
         cdef str dep_type
 
-        self.provenance = _yaml.node_get_provenance(dep)
+        self.provenance = dep.get_provenance()
 
         if type(dep) is _yaml.ScalarNode:
             self.name = dep.as_str()
@@ -88,7 +88,7 @@ cdef class Dependency:
                 if dep_type is None or dep_type == <str> Symbol.ALL:
                     dep_type = None
                 elif dep_type not in [Symbol.BUILD, Symbol.RUNTIME]:
-                    provenance = _yaml.node_get_provenance(dep, key=Symbol.TYPE)
+                    provenance = dep.get_scalar(Symbol.TYPE).get_provenance()
                     raise LoadError(LoadErrorReason.INVALID_DATA,
                                     "{}: Dependency type '{}' is not 'build', 'runtime' or 'all'"
                                     .format(provenance, dep_type))
diff --git a/src/buildstream/_options/optionarch.py b/src/buildstream/_options/optionarch.py
index 75be1a8ff..612ca2aa0 100644
--- a/src/buildstream/_options/optionarch.py
+++ b/src/buildstream/_options/optionarch.py
@@ -17,7 +17,6 @@
 #  Authors:
 #        Tristan Van Berkom <tristan.vanberkom@codethink.co.uk>
 
-from .. import _yaml
 from .._exceptions import LoadError, LoadErrorReason, PlatformError
 from .._platform import Platform
 from .optionenum import OptionEnum
@@ -55,7 +54,7 @@ class OptionArch(OptionEnum):
                     # Do not terminate the loop early to ensure we validate
                     # all values in the list.
             except PlatformError as e:
-                provenance = _yaml.node_get_provenance(node.get_sequence('values').scalar_at(index))
+                provenance = node.get_sequence('values').scalar_at(index).get_provenance()
                 prefix = ""
                 if provenance:
                     prefix = "{}: ".format(provenance)
diff --git a/src/buildstream/_options/optionenum.py b/src/buildstream/_options/optionenum.py
index 477fe1ed9..d1a7a85c9 100644
--- a/src/buildstream/_options/optionenum.py
+++ b/src/buildstream/_options/optionenum.py
@@ -17,7 +17,6 @@
 #  Authors:
 #        Tristan Van Berkom <tristan.vanberkom@codethink.co.uk>
 
-from .. import _yaml
 from .._exceptions import LoadError, LoadErrorReason
 from .option import Option, OPTION_SYMBOLS
 
@@ -50,16 +49,17 @@ class OptionEnum(Option):
         if not self.values:
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: No values specified for {} option '{}'"
-                            .format(_yaml.node_get_provenance(node), self.OPTION_TYPE, self.name))
+                            .format(node.get_provenance(), self.OPTION_TYPE, self.name))
 
         # Allow subclass to define the default value
         self.value = self.load_default_value(node)
 
     def load_value(self, node, *, transform=None):
-        self.value = node.get_str(self.name)
+        value_node = node.get_scalar(self.name)
+        self.value = value_node.as_str()
         if transform:
             self.value = transform(self.value)
-        self.validate(self.value, _yaml.node_get_provenance(node, self.name))
+        self.validate(self.value, value_node)
 
     def set_value(self, value):
         self.validate(value)
@@ -68,17 +68,20 @@ class OptionEnum(Option):
     def get_value(self):
         return self.value
 
-    def validate(self, value, provenance=None):
+    def validate(self, value, node=None):
         if value not in self.values:
-            prefix = ""
-            if provenance:
+            if node is not None:
+                provenance = node.get_provenance()
                 prefix = "{}: ".format(provenance)
+            else:
+                prefix = ""
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}Invalid value for {} option '{}': {}\n"
                             .format(prefix, self.OPTION_TYPE, self.name, value) +
                             "Valid values: {}".format(", ".join(self.values)))
 
     def load_default_value(self, node):
-        value = node.get_str('default')
-        self.validate(value, _yaml.node_get_provenance(node, 'default'))
+        value_node = node.get_scalar('default')
+        value = value_node.as_str()
+        self.validate(value, value_node)
         return value
diff --git a/src/buildstream/_options/optionflags.py b/src/buildstream/_options/optionflags.py
index 1d86361d8..80dd1b55d 100644
--- a/src/buildstream/_options/optionflags.py
+++ b/src/buildstream/_options/optionflags.py
@@ -17,7 +17,6 @@
 #  Authors:
 #        Tristan Van Berkom <tristan.vanberkom@codethink.co.uk>
 
-from .. import _yaml
 from .._exceptions import LoadError, LoadErrorReason
 from .option import Option, OPTION_SYMBOLS
 
@@ -51,17 +50,19 @@ class OptionFlags(Option):
         if not self.values:
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: No values specified for {} option '{}'"
-                            .format(_yaml.node_get_provenance(node), self.OPTION_TYPE, self.name))
+                            .format(node.get_provenance(), self.OPTION_TYPE, self.name))
 
-        self.value = node.get_sequence('default', default=[]).as_str_list()
-        self.validate(self.value, _yaml.node_get_provenance(node, 'default'))
+        value_node = node.get_sequence('default', default=[])
+        self.value = value_node.as_str_list()
+        self.validate(self.value, value_node)
 
     def load_value(self, node, *, transform=None):
-        self.value = node.get_sequence(self.name).as_str_list()
+        value_node = node.get_sequence(self.name)
+        self.value = value_node.as_str_list()
         if transform:
             self.value = [transform(x) for x in self.value]
         self.value = sorted(self.value)
-        self.validate(self.value, _yaml.node_get_provenance(node, self.name))
+        self.validate(self.value, value_node)
 
     def set_value(self, value):
         # Strip out all whitespace, allowing: "value1, value2 , value3"
@@ -76,12 +77,14 @@ class OptionFlags(Option):
     def get_value(self):
         return ",".join(self.value)
 
-    def validate(self, value, provenance=None):
+    def validate(self, value, node=None):
         for flag in value:
             if flag not in self.values:
-                prefix = ""
-                if provenance:
+                if node is not None:
+                    provenance = node.get_provenance()
                     prefix = "{}: ".format(provenance)
+                else:
+                    prefix = ""
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}Invalid value for flags option '{}': {}\n"
                                 .format(prefix, self.name, value) +
diff --git a/src/buildstream/_options/optionpool.py b/src/buildstream/_options/optionpool.py
index cd4bfb6dc..3b33c0349 100644
--- a/src/buildstream/_options/optionpool.py
+++ b/src/buildstream/_options/optionpool.py
@@ -74,7 +74,7 @@ class OptionPool():
             try:
                 opt_type = _OPTION_TYPES[opt_type_name]
             except KeyError:
-                p = _yaml.node_get_provenance(option_definition, 'type')
+                p = option_definition.get_scalar('type').get_provenance()
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}: Invalid option type '{}'".format(p, opt_type_name))
 
@@ -90,11 +90,11 @@ class OptionPool():
     #    node (dict): The loaded YAML options
     #
     def load_yaml_values(self, node, *, transform=None):
-        for option_name in node.keys():
+        for option_name, option_value in node.items():
             try:
                 option = self._options[option_name]
             except KeyError as e:
-                p = _yaml.node_get_provenance(node, option_name)
+                p = option_value.get_provenance()
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}: Unknown option '{}' specified"
                                 .format(p, option_name)) from e
@@ -256,7 +256,7 @@ class OptionPool():
         # assertion in a given dictionary, and not lose an assertion due to
         # it being overwritten by a later assertion which might also trigger.
         if assertion is not None:
-            p = _yaml.node_get_provenance(node, '(!)')
+            p = node.get_scalar('(!)').get_provenance()
             raise LoadError(LoadErrorReason.USER_ASSERTION,
                             "{}: {}".format(p, assertion.strip()))
 
@@ -266,7 +266,7 @@ class OptionPool():
             for condition in conditions:
                 tuples = list(condition.items())
                 if len(tuples) > 1:
-                    provenance = _yaml.node_get_provenance(condition)
+                    provenance = condition.get_provenance()
                     raise LoadError(LoadErrorReason.INVALID_DATA,
                                     "{}: Conditional statement has more than one key".format(provenance))
 
@@ -275,11 +275,11 @@ class OptionPool():
                     apply_fragment = self._evaluate(expression)
                 except LoadError as e:
                     # Prepend the provenance of the error
-                    provenance = _yaml.node_get_provenance(condition)
+                    provenance = condition.get_provenance()
                     raise LoadError(e.reason, "{}: {}".format(provenance, e)) from e
 
                 if type(value) is not _yaml.MappingNode:  # pylint: disable=unidiomatic-typecheck
-                    provenance = _yaml.node_get_provenance(condition)
+                    provenance = condition.get_provenance()
                     raise LoadError(LoadErrorReason.ILLEGAL_COMPOSITE,
                                     "{}: Only values of type 'dict' can be composed.".format(provenance))
 
diff --git a/src/buildstream/_project.py b/src/buildstream/_project.py
index 45eae03f9..f075aeb51 100644
--- a/src/buildstream/_project.py
+++ b/src/buildstream/_project.py
@@ -268,14 +268,14 @@ class Project():
         full_path = self._absolute_directory_path / path
 
         if full_path.is_symlink():
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID_KIND,
                             "{}: Specified path '{}' must not point to "
                             "symbolic links "
                             .format(provenance, path_str))
 
         if path.parts and path.parts[0] == '..':
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID,
                             "{}: Specified path '{}' first component must "
                             "not be '..'"
@@ -287,7 +287,7 @@ class Project():
             else:
                 full_resolved_path = full_path.resolve(strict=True)  # pylint: disable=unexpected-keyword-arg
         except FileNotFoundError:
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.MISSING_FILE,
                             "{}: Specified path '{}' does not exist"
                             .format(provenance, path_str))
@@ -296,14 +296,14 @@ class Project():
             full_resolved_path == self._absolute_directory_path)
 
         if not is_inside:
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID,
                             "{}: Specified path '{}' must not lead outside of the "
                             "project directory"
                             .format(provenance, path_str))
 
         if path.is_absolute():
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID,
                             "{}: Absolute path: '{}' invalid.\n"
                             "Please specify a path relative to the project's root."
@@ -312,20 +312,20 @@ class Project():
         if full_resolved_path.is_socket() or (
                 full_resolved_path.is_fifo() or
                 full_resolved_path.is_block_device()):
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID_KIND,
                             "{}: Specified path '{}' points to an unsupported "
                             "file kind"
                             .format(provenance, path_str))
 
         if check_is_file and not full_resolved_path.is_file():
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID_KIND,
                             "{}: Specified path '{}' is not a regular file"
                             .format(provenance, path_str))
 
         if check_is_dir and not full_resolved_path.is_dir():
-            provenance = _yaml.node_get_provenance(node)
+            provenance = node.get_provenance()
             raise LoadError(LoadErrorReason.PROJ_PATH_INVALID_KIND,
                             "{}: Specified path '{}' is not a directory"
                             .format(provenance, path_str))
@@ -627,9 +627,10 @@ class Project():
                         ignore_unknown=True)
 
         # Use separate file for storing source references
-        self.ref_storage = pre_config_node.get_str('ref-storage')
+        ref_storage_node = pre_config_node.get_scalar('ref-storage')
+        self.ref_storage = ref_storage_node.as_str()
         if self.ref_storage not in [ProjectRefStorage.INLINE, ProjectRefStorage.PROJECT_REFS]:
-            p = _yaml.node_get_provenance(pre_config_node, 'ref-storage')
+            p = ref_storage_node.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Invalid value '{}' specified for ref-storage"
                             .format(p, self.ref_storage))
diff --git a/src/buildstream/_projectrefs.py b/src/buildstream/_projectrefs.py
index cee74dc11..10dd000b2 100644
--- a/src/buildstream/_projectrefs.py
+++ b/src/buildstream/_projectrefs.py
@@ -63,7 +63,7 @@ class ProjectRefs():
     def load(self, options):
         try:
             self._toplevel_node = _yaml.load(self._fullpath, shortname=self._base_name, copy_tree=True)
-            provenance = _yaml.node_get_provenance(self._toplevel_node)
+            provenance = self._toplevel_node.get_provenance()
             self._toplevel_save = provenance.toplevel
 
             # Process any project options immediately
diff --git a/src/buildstream/_variables.pyx b/src/buildstream/_variables.pyx
index ba8338f23..eb2deb553 100644
--- a/src/buildstream/_variables.pyx
+++ b/src/buildstream/_variables.pyx
@@ -139,7 +139,7 @@ cdef class Variables:
             for var in expstr[1::2]:
                 if var not in self._expstr_map:
                     line = "  unresolved variable '{unmatched}' in declaration of '{variable}' at: {provenance}"
-                    provenance = _yaml.node_get_provenance(self.original, key)
+                    provenance = expstr.get_provenance()
                     summary.append(line.format(unmatched=var, variable=key, provenance=provenance))
         if summary:
             raise LoadError(LoadErrorReason.UNRESOLVED_VARIABLE,
@@ -153,7 +153,7 @@ cdef class Variables:
                     continue
                 if var in visited:
                     raise LoadError(LoadErrorReason.RECURSIVE_VARIABLE,
-                                    "{}: ".format(_yaml.node_get_provenance(self.original, var)) +
+                                    "{}: ".format(self.original.get_scalar(var).get_provenance()) +
                                     ("Variable '{}' expands to contain a reference to itself. " +
                                      "Perhaps '{}' contains '%{{{}}}").format(var, visited[-1], var))
                 visited.append(var)
diff --git a/src/buildstream/_yaml.pxd b/src/buildstream/_yaml.pxd
index 7b5209077..9c67f1ad2 100644
--- a/src/buildstream/_yaml.pxd
+++ b/src/buildstream/_yaml.pxd
@@ -27,6 +27,7 @@ cdef class Node:
     cdef int column
 
     cpdef Node copy(self)
+    cpdef ProvenanceInformation get_provenance(self)
     cpdef object strip_node_info(self)
 
     cpdef void _assert_fully_composited(self) except *
@@ -89,6 +90,3 @@ cdef class ProvenanceInformation:
     cdef public int col, line
     cdef public object project, toplevel
     cdef public bint is_synthetic
-
-
-cpdef ProvenanceInformation node_get_provenance(Node node, str key=*)
diff --git a/src/buildstream/_yaml.pyx b/src/buildstream/_yaml.pyx
index a84c05759..945a82378 100644
--- a/src/buildstream/_yaml.pyx
+++ b/src/buildstream/_yaml.pyx
@@ -94,6 +94,9 @@ cdef class Node:
     cpdef Node copy(self):
         raise NotImplementedError()
 
+    cpdef ProvenanceInformation get_provenance(self):
+        return ProvenanceInformation(self)
+
     cpdef object strip_node_info(self):
         raise NotImplementedError()
 
@@ -171,7 +174,7 @@ cdef class ScalarNode(Node):
         elif self.value in ('False', 'false'):
             return False
         else:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             path = provenance.toplevel._find(self)[-1]
             raise LoadError(LoadErrorReason.INVALID_DATA,
                 "{}: Value of '{}' is not of the expected type '{}'"
@@ -181,7 +184,7 @@ cdef class ScalarNode(Node):
         try:
             return int(self.value)
         except ValueError:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             path = provenance.toplevel._find(self)[-1]
             raise LoadError(LoadErrorReason.INVALID_DATA,
                 "{}: Value of '{}' is not of the expected type '{}'"
@@ -205,8 +208,8 @@ cdef class ScalarNode(Node):
         if target_value is not None and type(target_value) is not ScalarNode:
             raise CompositeError(path,
                                  "{}: Cannot compose scalar on non-scalar at {}".format(
-                                    node_get_provenance(self),
-                                    node_get_provenance(target_value)))
+                                    self.get_provenance(),
+                                    target_value.get_provenance()))
 
         target.value[key] = self
 
@@ -263,7 +266,7 @@ cdef class MappingNode(Node):
         try:
             self._composite(target, [])
         except CompositeError as e:
-            source_provenance = node_get_provenance(self)
+            source_provenance = self.get_provenance()
             error_prefix = ""
             if source_provenance:
                 error_prefix = "{}: ".format(source_provenance)
@@ -292,7 +295,7 @@ cdef class MappingNode(Node):
 
         if value is _sentinel:
             if default is _sentinel:
-                provenance = node_get_provenance(self)
+                provenance = self.get_provenance()
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}: Dictionary did not contain expected key '{}'".format(provenance, key))
 
@@ -308,7 +311,7 @@ cdef class MappingNode(Node):
         value = self.get(key, default, MappingNode)
 
         if type(value) is not MappingNode and value is not None:
-            provenance = node_get_provenance(value)
+            provenance = value.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not of the expected type 'Mapping'"
                             .format(provenance, key))
@@ -322,12 +325,12 @@ cdef class MappingNode(Node):
             if allow_none:
                 return None
 
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Dictionary did not contain expected key '{}'".format(provenance, key))
 
         if allowed_types and type(value) not in allowed_types:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not one of the following: {}.".format(
                                 provenance, key, ", ".join(allowed_types)))
@@ -341,7 +344,7 @@ cdef class MappingNode(Node):
             if value is None:
                 value = ScalarNode.__new__(ScalarNode, self.file_index, 0, next_synthetic_counter(), None)
             else:
-                provenance = node_get_provenance(value)
+                provenance = value.get_provenance()
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}: Value of '{}' is not of the expected type 'Scalar'"
                                 .format(provenance, key))
@@ -352,7 +355,7 @@ cdef class MappingNode(Node):
         value = self.get(key, default, SequenceNode)
 
         if type(value) is not SequenceNode and value is not None:
-            provenance = node_get_provenance(value)
+            provenance = value.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not of the expected type 'Sequence'"
                             .format(provenance, key))
@@ -403,7 +406,7 @@ cdef class MappingNode(Node):
 
         for key in self.value:
             if key not in valid_keys_set:
-                provenance = node_get_provenance(self, key=key)
+                provenance = self.get_node(key).get_provenance()
                 raise LoadError(LoadErrorReason.INVALID_DATA,
                                 "{}: Unexpected key: {}".format(provenance, key))
 
@@ -445,8 +448,8 @@ cdef class MappingNode(Node):
                     # Else composing on top of normal dict or a scalar, so raise...
                     raise CompositeError(path,
                                          "{}: Cannot compose lists onto {}".format(
-                                             node_get_provenance(self),
-                                             node_get_provenance(target_value)))
+                                             self.get_provenance(),
+                                             target_value.get_provenance()))
         else:
             # We're composing a dict into target now
             if key not in target.value:
@@ -514,7 +517,7 @@ cdef class MappingNode(Node):
                 has_keys = True
 
         if has_keys and has_directives:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Dictionary contains array composition directives and arbitrary keys"
                             .format(provenance))
@@ -566,7 +569,7 @@ cdef class MappingNode(Node):
             # never existed in the underlying data
             #
             if key in ('(>)', '(<)', '(=)'):
-                provenance = node_get_provenance(value)
+                provenance = value.get_provenance()
                 raise LoadError(LoadErrorReason.TRAILING_LIST_DIRECTIVE,
                                 "{}: Attempt to override non-existing list".format(provenance))
 
@@ -613,7 +616,7 @@ cdef class SequenceNode(Node):
         value = self.value[index]
 
         if type(value) is not MappingNode:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             path = ["[{}]".format(p) for p in provenance.toplevel._find(self)] + ["[{}]".format(index)]
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not of the expected type '{}'"
@@ -624,7 +627,7 @@ cdef class SequenceNode(Node):
         cdef value = self.value[index]
 
         if allowed_types and type(value) not in allowed_types:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not one of the following: {}.".format(
                                 provenance, index, ", ".join(allowed_types)))
@@ -635,7 +638,7 @@ cdef class SequenceNode(Node):
         value = self.value[index]
 
         if type(value) is not ScalarNode:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             path = ["[{}]".format(p) for p in provenance.toplevel._find(self)] + ["[{}]".format(index)]
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not of the expected type '{}'"
@@ -646,7 +649,7 @@ cdef class SequenceNode(Node):
         value = self.value[index]
 
         if type(value) is not SequenceNode:
-            provenance = node_get_provenance(self)
+            provenance = self.get_provenance()
             path = ["[{}]".format(p) for p in provenance.toplevel._find(self)] + ["[{}]".format(index)]
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: Value of '{}' is not of the expected type '{}'"
@@ -675,9 +678,9 @@ cdef class SequenceNode(Node):
                 target_value._is_composite_list()):
             raise CompositeError(path,
                                  "{}: List cannot overwrite {} at: {}"
-                                 .format(node_get_provenance(self),
+                                 .format(self.get_provenance(),
                                          key,
-                                         node_get_provenance(target_value)))
+                                         target_value.get_provenance()))
         # Looks good, clobber it
         target.value[key] = self
 
@@ -763,7 +766,7 @@ cdef int next_synthetic_counter():
     return __counter
 
 
-# Returned from node_get_provenance
+# Returned from Node.get_provenance
 cdef class ProvenanceInformation:
 
     def __init__(self, Node nodeish):
@@ -1126,24 +1129,6 @@ cpdef Node load_data(str data, int file_index=_SYNTHETIC_FILE_INDEX, str file_na
     return contents
 
 
-# node_get_provenance()
-#
-# Gets the provenance for a node
-#
-# Args:
-#   node (Node): a dictionary
-#   key (str): key in the dictionary
-#
-# Returns: The Provenance of the dict, member or list element
-#
-cpdef ProvenanceInformation node_get_provenance(Node node, str key=None):
-    if key is None:
-        # Retrieving the provenance for this node directly
-        return ProvenanceInformation(node)
-
-    return ProvenanceInformation((<MappingNode> node).value.get(key))
-
-
 # new_synthetic_file()
 #
 # Create a new synthetic mapping node, with an associated file entry
@@ -1251,7 +1236,7 @@ def assert_symbol_name(str symbol_name, str purpose, *, Node ref_node=None, bint
 
         message = "Invalid symbol name for {}: '{}'".format(purpose, symbol_name)
         if ref_node:
-            provenance = node_get_provenance(ref_node)
+            provenance = ref_node.get_provenance()
             if provenance is not None:
                 message = "{}: {}".format(provenance, message)
 
diff --git a/src/buildstream/element.py b/src/buildstream/element.py
index 0de9208bb..53d26528b 100644
--- a/src/buildstream/element.py
+++ b/src/buildstream/element.py
@@ -515,7 +515,7 @@ class Element(Plugin):
         try:
             return self.__variables.subst(value)
         except LoadError as e:
-            provenance = _yaml.node_get_provenance(node, key=member_name)
+            provenance = node.get_scalar(member_name).get_provenance()
             raise LoadError(e.reason, '{}: {}'.format(provenance, e), detail=e.detail) from e
 
     def node_subst_list(self, node, member_name):
@@ -537,7 +537,7 @@ class Element(Plugin):
             try:
                 ret.append(self.__variables.subst(value.as_str()))
             except LoadError as e:
-                provenance = _yaml.node_get_provenance(value)
+                provenance = value.get_provenance()
                 raise LoadError(e.reason, '{}: {}'.format(provenance, e), detail=e.detail) from e
         return ret
 
@@ -2605,8 +2605,13 @@ class Element(Plugin):
         variables._assert_fully_composited()
 
         for var in ('project-name', 'element-name', 'max-jobs'):
-            provenance = _yaml.node_get_provenance(variables, var)
-            if provenance and not provenance.is_synthetic:
+            node = variables.get_node(var, allow_none=True)
+
+            if node is None:
+                continue
+
+            provenance = node.get_provenance()
+            if not provenance.is_synthetic:
                 raise LoadError(LoadErrorReason.PROTECTED_VARIABLE_REDEFINED,
                                 "{}: invalid redefinition of protected variable '{}'"
                                 .format(provenance, var))
diff --git a/src/buildstream/plugin.py b/src/buildstream/plugin.py
index 2e5cc9678..92e52504d 100644
--- a/src/buildstream/plugin.py
+++ b/src/buildstream/plugin.py
@@ -116,7 +116,6 @@ import sys
 from contextlib import contextmanager
 from weakref import WeakValueDictionary
 
-from . import _yaml
 from . import utils
 from ._exceptions import PluginError, ImplError
 from ._message import Message, MessageType
@@ -360,7 +359,7 @@ class Plugin():
         Returns:
             (str): A string describing the provenance of the node and member
         """
-        provenance = _yaml.node_get_provenance(node, key=member_name)
+        provenance = node.get_node(member_name).get_provenance()
         return str(provenance)
 
     def node_get_project_path(self, node, *,
diff --git a/src/buildstream/sandbox/_sandboxremote.py b/src/buildstream/sandbox/_sandboxremote.py
index ca3c5c3d9..76b1a6270 100644
--- a/src/buildstream/sandbox/_sandboxremote.py
+++ b/src/buildstream/sandbox/_sandboxremote.py
@@ -114,7 +114,7 @@ class SandboxRemote(Sandbox):
         def require_node(config, keyname):
             val = config.get_mapping(keyname, default=None)
             if val is None:
-                provenance = _yaml.node_get_provenance(remote_config, key=keyname)
+                provenance = remote_config.get_provenance()
                 raise _yaml.LoadError(_yaml.LoadErrorReason.INVALID_DATA,
                                       "{}: '{}' was not present in the remote "
                                       "execution configuration (remote-execution). "
@@ -146,7 +146,7 @@ class SandboxRemote(Sandbox):
             if 'execution-service' not in remote_config:
                 exec_config = _yaml.Node.from_dict({'url': remote_config['url']})
             else:
-                provenance = _yaml.node_get_provenance(remote_config, key='url')
+                provenance = remote_config.get_node('url').get_provenance()
                 raise _yaml.LoadError(_yaml.LoadErrorReason.INVALID_DATA,
                                       "{}: 'url' and 'execution-service' keys were found in the remote "
                                       "execution configuration (remote-execution). "
@@ -164,7 +164,7 @@ class SandboxRemote(Sandbox):
         for config_key, config in zip(service_keys, service_configs):
             # Either both or none of the TLS client key/cert pair must be specified:
             if ('client-key' in config) != ('client-cert' in config):
-                provenance = _yaml.node_get_provenance(remote_config, key=config_key)
+                provenance = remote_config.get_node(config_key).get_provenance()
                 raise _yaml.LoadError(_yaml.LoadErrorReason.INVALID_DATA,
                                       "{}: TLS client key/cert pair is incomplete. "
                                       "You must specify both 'client-key' and 'client-cert' "
diff --git a/src/buildstream/source.py b/src/buildstream/source.py
index a0a53fba1..e8cadbab1 100644
--- a/src/buildstream/source.py
+++ b/src/buildstream/source.py
@@ -307,7 +307,7 @@ class Source(Plugin):
     """
 
     def __init__(self, context, project, meta, *, alias_override=None, unique_id=None):
-        provenance = _yaml.node_get_provenance(meta.config)
+        provenance = meta.config.get_provenance()
         super().__init__("{}-{}".format(meta.element_name, meta.element_index),
                          context, project, provenance, "source", unique_id=unique_id)
 
@@ -961,9 +961,9 @@ class Source(Plugin):
         for key, action in actions.items():
             # Obtain the top level node and its file
             if action == 'add':
-                provenance = _yaml.node_get_provenance(node)
+                provenance = node.get_provenance()
             else:
-                provenance = _yaml.node_get_provenance(node, key=key)
+                provenance = node.get_node(key).get_provenance()
 
             toplevel_node = provenance.toplevel
 
diff --git a/tests/frontend/buildcheckout.py b/tests/frontend/buildcheckout.py
index 07ebe7e6a..d3eec0d21 100644
--- a/tests/frontend/buildcheckout.py
+++ b/tests/frontend/buildcheckout.py
@@ -462,7 +462,7 @@ def test_inconsistent_junction(cli, tmpdir, datafiles, ref_storage):
     # Assert that we have the expected provenance encoded into the error
     element_node = _yaml.load(element_path, shortname='junction-dep.bst')
     ref_node = element_node.get_sequence('depends').mapping_at(0)
-    provenance = _yaml.node_get_provenance(ref_node)
+    provenance = ref_node.get_provenance()
     assert str(provenance) in result.stderr
 
 
diff --git a/tests/frontend/fetch.py b/tests/frontend/fetch.py
index ba79dcfd1..7ea357ac2 100644
--- a/tests/frontend/fetch.py
+++ b/tests/frontend/fetch.py
@@ -172,5 +172,5 @@ def test_inconsistent_junction(cli, tmpdir, datafiles, ref_storage):
     # Assert that we have the expected provenance encoded into the error
     element_node = _yaml.load(element_path, shortname='junction-dep.bst')
     ref_node = element_node.get_sequence('depends').mapping_at(0)
-    provenance = _yaml.node_get_provenance(ref_node)
+    provenance = ref_node.get_provenance()
     assert str(provenance) in result.stderr
diff --git a/tests/frontend/show.py b/tests/frontend/show.py
index a0421bd37..756fe1786 100644
--- a/tests/frontend/show.py
+++ b/tests/frontend/show.py
@@ -335,7 +335,7 @@ def test_inconsistent_junction(cli, tmpdir, datafiles, ref_storage, workspaced):
         # Assert that we have the expected provenance encoded into the error
         element_node = _yaml.load(element_path, shortname='junction-dep.bst')
         ref_node = element_node.get_sequence('depends').mapping_at(0)
-        provenance = _yaml.node_get_provenance(ref_node)
+        provenance = ref_node.get_provenance()
         assert str(provenance) in dep_result.stderr
 
         dep_result.assert_main_error(ErrorDomain.LOAD, LoadErrorReason.SUBPROJECT_INCONSISTENT)
diff --git a/tests/frontend/track.py b/tests/frontend/track.py
index b8ab0d4eb..a628043d8 100644
--- a/tests/frontend/track.py
+++ b/tests/frontend/track.py
@@ -277,7 +277,7 @@ def test_inconsistent_junction(cli, tmpdir, datafiles, ref_storage):
     # Assert that we have the expected provenance encoded into the error
     element_node = _yaml.load(element_path, shortname='junction-dep.bst')
     ref_node = element_node.get_sequence('depends').mapping_at(0)
-    provenance = _yaml.node_get_provenance(ref_node)
+    provenance = ref_node.get_provenance()
     assert str(provenance) in result.stderr
 
 
@@ -316,7 +316,7 @@ def test_junction_element(cli, tmpdir, datafiles, ref_storage):
     # Assert that we have the expected provenance encoded into the error
     element_node = _yaml.load(element_path, shortname='junction-dep.bst')
     ref_node = element_node.get_sequence('depends').mapping_at(0)
-    provenance = _yaml.node_get_provenance(ref_node)
+    provenance = ref_node.get_provenance()
     assert str(provenance) in result.stderr
 
     # Now track the junction itself
diff --git a/tests/internals/yaml.py b/tests/internals/yaml.py
index 8ceed9d15..9e62cdcb3 100644
--- a/tests/internals/yaml.py
+++ b/tests/internals/yaml.py
@@ -24,8 +24,8 @@ def test_load_yaml(datafiles):
     assert loaded.get_str('kind') == 'pony'
 
 
-def assert_provenance(filename, line, col, node, key=None):
-    provenance = _yaml.node_get_provenance(node, key=key)
+def assert_provenance(filename, line, col, node):
+    provenance = node.get_provenance()
 
     assert isinstance(provenance, _yaml.ProvenanceInformation)
 
@@ -56,7 +56,7 @@ def test_member_provenance(datafiles):
 
     loaded = _yaml.load(filename)
     assert loaded.get_str('kind') == 'pony'
-    assert_provenance(filename, 2, 13, loaded, 'description')
+    assert_provenance(filename, 2, 13, loaded.get_scalar('description'))
 
 
 @pytest.mark.datafiles(os.path.join(DATA_DIR))
@@ -108,7 +108,7 @@ def test_node_get(datafiles):
     assert len(children) == 7
 
     child = base.get_sequence('children').mapping_at(6)
-    assert_provenance(filename, 20, 8, child, 'mood')
+    assert_provenance(filename, 20, 8, child.get_scalar('mood'))
 
     extra = base.get_mapping('extra')
     with pytest.raises(LoadError) as exc:
@@ -256,7 +256,7 @@ def test_list_composition(datafiles, filename, tmpdir,
     child = children.mapping_at(index)
 
     assert child.get_str('mood') == mood
-    assert_provenance(prov_file, prov_line, prov_col, child, 'mood')
+    assert_provenance(prov_file, prov_line, prov_col, child.get_node('mood'))
 
 
 # Test that overwriting a list with an empty list works as expected.
@@ -378,7 +378,7 @@ def test_list_composition_twice(datafiles, tmpdir, filename1, filename2,
     child = children.mapping_at(index)
 
     assert child.get_str('mood') == mood
-    assert_provenance(prov_file, prov_line, prov_col, child, 'mood')
+    assert_provenance(prov_file, prov_line, prov_col, child.get_node('mood'))
 
     #####################
     # Round 2 - Fight !
@@ -395,7 +395,7 @@ def test_list_composition_twice(datafiles, tmpdir, filename1, filename2,
     child = children.mapping_at(index)
 
     assert child.get_str('mood') == mood
-    assert_provenance(prov_file, prov_line, prov_col, child, 'mood')
+    assert_provenance(prov_file, prov_line, prov_col, child.get_node('mood'))
 
 
 @pytest.mark.datafiles(os.path.join(DATA_DIR))
@@ -494,7 +494,7 @@ def test_node_find_target(datafiles, case):
     # are not the same nodes as in `prov.toplevel`
     loaded = _yaml.load(filename, copy_tree=True)
 
-    prov = _yaml.node_get_provenance(loaded)
+    prov = loaded.get_provenance()
 
     toplevel = prov.toplevel