datajunction 0.0.154__tar.gz → 0.0.156__tar.gz

{datajunction-0.0.154 → datajunction-0.0.156}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datajunction
-Version: 0.0.154
+Version: 0.0.156
 Summary: DataJunction client library for connecting to a DataJunction server
 Project-URL: repository, https://github.com/DataJunction/dj
 Author-email: DataJunction Authors <yian.shang@gmail.com>

{datajunction-0.0.154 → datajunction-0.0.156}/datajunction/__about__.py

@@ -2,4 +2,4 @@
 Version for Hatch
 """
 
-__version__ = "0.0.154"
+__version__ = "0.0.156"

{datajunction-0.0.154 → datajunction-0.0.156}/datajunction/cli.py

@@ -14,7 +14,7 @@ from rich.syntax import Syntax
 from rich.table import Table
 from rich.text import Text
 
-from datajunction import DJBuilder, Project
+from datajunction import DJBuilder
 from datajunction.deployment import DeploymentService
 from datajunction.exceptions import DJClientException, DJDeploymentFailure
 
@@ -1846,6 +1846,7 @@ model: inherit
         """
         tables = [
             "node",
+            "nodenamespace",
             "noderevision",
             "users",
             "materialization",
@@ -1854,6 +1855,7 @@ model: inherit
             "backfill",
             "collection",
             "dimensionlink",
+            "deployments",
         ]
         for table in tables:
             try:
@@ -1869,19 +1871,27 @@ model: inherit
                     )
         logger.info("Finished registering DJ system metadata tables")
 
-        logger.info("Loading DJ system nodes...")
+        logger.info("Pushing DJ system nodes...")
         script_dir = Path(__file__).resolve().parent
         project_dir = script_dir / "seed" / type
-        project = Project.load(str(project_dir))
-        logger.info("Finished loading DJ system nodes.")
-
-        logger.info("Compiling DJ system nodes...")
-        compiled_project = project.compile()
-        logger.info("Finished compiling DJ system nodes.")
-
-        logger.info("Deploying DJ system nodes...")
-        compiled_project.deploy(client=self.builder_client)
-        logger.info("Finished deploying DJ system nodes.")
+        # Read the project's declared prefix and pass it explicitly so the
+        # push flow doesn't apply git-branch namespace derivation to the
+        # bootstrap nodes (they always belong to the canonical system namespace).
+        project_yaml_path = project_dir / "dj.yaml"
+        project_meta: dict = {}
+        if project_yaml_path.exists():  # pragma: no branch
+            import yaml as _yaml
+
+            with open(project_yaml_path, encoding="utf-8") as fh:
+                project_meta = _yaml.safe_load(fh) or {}
+        system_namespace = project_meta.get("namespace") or project_meta.get(
+            "prefix",
+        )
+        self.deployment_service.push(
+            str(project_dir),
+            namespace=system_namespace,
+        )
+        logger.info("Finished pushing DJ system nodes.")
 
 
 def main(builder_client: DJBuilder | None = None):

{datajunction-0.0.154 → datajunction-0.0.156}/datajunction/deployment.py

@@ -136,9 +136,15 @@ class DeploymentService:
             branch=source.get("branch") or branch,
         )
 
-        if branch:
-            # Use base_namespace from dj.yaml as parent, whether --namespace was
-            # passed explicitly or derived automatically
+        # Apply git config in the normal push flow. Skip only when the
+        # caller passed an explicit ``namespace`` equal to the project's
+        # base namespace — that's the bootstrap / system-seed case where
+        # we'd otherwise set parent to the same namespace and error with
+        # "a namespace cannot be its own parent".
+        skip_git_config = (
+            namespace is not None and deployment_spec["namespace"] == base_namespace
+        )
+        if branch and not skip_git_config:
             parent_namespace = base_namespace or None
             try:
                 self.client._set_namespace_git_config(
@@ -389,10 +395,30 @@ class DeploymentService:
         nodes = []
         warnings = []
         nodes_dir = Path(base_dir)
+        # Skip dirs that may carry stray YAMLs but aren't part of the project
+        # (vendored deps, build artifacts, dot-dirs).
+        skip_dir_names = {
+            "venv",
+            ".venv",
+            "node_modules",
+            "site-packages",
+            "__pycache__",
+            ".git",
+            "build",
+            "dist",
+            ".tox",
+        }
         for path in nodes_dir.rglob("*.yaml"):
             if path.name == "dj.yaml":
                 continue
+            parents = path.relative_to(nodes_dir).parts[:-1]
+            if any(part in skip_dir_names or part.startswith(".") for part in parents):
+                continue
             node_dict = DeploymentService.read_yaml_file(path)
+            if not isinstance(node_dict, dict):
+                # Not a node definition — stray YAML (e.g. a list-typed file
+                # from a vendored package); skip rather than crash.
+                continue
 
             # Check filename matches node name
             node_name = node_dict.get("name", "")
@@ -439,8 +465,10 @@ class DeploymentService:
             seen_names[node_name] = node
         nodes = list(seen_names.values())
 
+        # Accept either `namespace:` (new) or `prefix:` (legacy seed format) in dj.yaml
         deployment_spec = {
-            "namespace": project_metadata.get("namespace", ""),  # fallback to empty
+            "namespace": project_metadata.get("namespace")
+            or project_metadata.get("prefix", ""),
             "nodes": nodes,
             "tags": project_metadata.get("tags", []),
         }

datajunction-0.0.154/datajunction/seed/nodes/date.dimension.yaml → datajunction-0.0.156/datajunction/seed/nodes/date.yaml

@@ -1,3 +1,5 @@
+name: ${prefix}date
+node_type: dimension
 display_name: Date
 description: ''
 query: 'SELECT 1 AS dateint '

datajunction-0.0.156/datajunction/seed/nodes/deployment.yaml

@@ -0,0 +1,89 @@
+name: ${prefix}deployment
+node_type: dimension
+display_name: Deployment
+description: |-
+  One row per deployment. Repo / branch / git-config attributes are extracted
+  inline from the deployment's ``spec`` JSON, so the dim can be sliced by
+  source type, repository, branch, CI system, etc. without a Postgres view.
+query: |-
+  SELECT
+    D.uuid,
+    D.namespace,
+    CAST(D.status AS VARCHAR) AS status,
+    D.created_by_id,
+    D.created_at,
+    D.updated_at,
+    CAST(TO_CHAR(CAST(D.created_at AS date), 'YYYYMMDD') AS integer) AS created_at_date,
+    CAST(TO_CHAR(CAST(DATE_TRUNC('week', D.created_at) AS date), 'YYYYMMDD') AS integer) AS created_at_week,
+    CAST(TO_CHAR(CAST(D.updated_at AS date), 'YYYYMMDD') AS integer) AS updated_at_date,
+    CAST(TO_CHAR(CAST(DATE_TRUNC('week', D.updated_at) AS date), 'YYYYMMDD') AS integer) AS updated_at_week,
+    CAST(json_extract_path_text(D.spec, 'source', 'type') AS VARCHAR) AS source_type,
+    CAST(json_extract_path_text(D.spec, 'source', 'repository') AS VARCHAR) AS repository,
+    CAST(json_extract_path_text(D.spec, 'source', 'branch') AS VARCHAR) AS branch,
+    CAST(json_extract_path_text(D.spec, 'source', 'commit_sha') AS VARCHAR) AS commit_sha,
+    CAST(json_extract_path_text(D.spec, 'source', 'ci_system') AS VARCHAR) AS ci_system,
+    CAST(EXTRACT(EPOCH FROM (D.updated_at - D.created_at)) AS double) AS duration_seconds
+  FROM source.dj_metadata.public.deployments D
+columns:
+- name: uuid
+  attributes:
+  - hidden
+- name: created_by_id
+  attributes:
+  - hidden
+- name: created_at
+  attributes:
+  - hidden
+- name: updated_at
+  attributes:
+  - hidden
+- name: namespace
+  display_name: Namespace
+- name: status
+  display_name: Status
+- name: source_type
+  display_name: Source Type
+- name: repository
+  display_name: Repository
+- name: branch
+  display_name: Branch
+- name: commit_sha
+  display_name: Commit SHA
+  attributes:
+  - hidden
+- name: ci_system
+  display_name: CI System
+- name: duration_seconds
+  display_name: Duration (seconds)
+  attributes:
+  - hidden
+- name: created_at_date
+  display_name: Created At (Day)
+- name: created_at_week
+  display_name: Created At (Week)
+- name: updated_at_date
+  display_name: Updated At (Day)
+- name: updated_at_week
+  display_name: Updated At (Week)
+primary_key:
+- uuid
+dimension_links:
+- type: join
+  dimension_node: ${prefix}user
+  join_type: left
+  join_on: ${prefix}deployment.created_by_id = ${prefix}user.id
+- type: join
+  dimension_node: ${prefix}date
+  join_type: left
+  join_on: ${prefix}deployment.created_at_date = ${prefix}date.dateint
+  role: created_at
+- type: join
+  dimension_node: ${prefix}date
+  join_type: left
+  join_on: ${prefix}deployment.updated_at_date = ${prefix}date.dateint
+  role: updated_at
+- type: join
+  dimension_node: ${prefix}namespace
+  join_type: left
+  join_on: ${prefix}deployment.namespace = ${prefix}namespace.name
+tags: []

datajunction-0.0.154/datajunction/seed/nodes/dimension_link.dimension.yaml → datajunction-0.0.156/datajunction/seed/nodes/dimension_link.yaml

@@ -1,3 +1,5 @@
+name: ${prefix}dimension_link
+node_type: dimension
 display_name: Dimension Link
 description: ''
 query: |-
@@ -6,7 +8,16 @@ query: |-
     dimension_id,
     node_revision_id
   FROM source.dj_metadata.public.dimensionlink
-columns: []
+columns:
+- name: id
+  attributes:
+  - hidden
+- name: dimension_id
+  attributes:
+  - hidden
+- name: node_revision_id
+  attributes:
+  - hidden
 primary_key:
 - id
 dimension_links:

datajunction-0.0.156/datajunction/seed/nodes/distinct_node_authors.yaml

@@ -0,0 +1,20 @@
+name: ${prefix}distinct_node_authors
+node_type: metric
+display_name: Distinct Node Authors
+description: |-
+  Count of distinct users who originally created nodes. When sliced by
+  ``created_at_day`` / ``created_at_week`` etc., gives daily / weekly /
+  monthly active authors.
+query: |-
+  SELECT COUNT(DISTINCT created_by_id)
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Activity
+  subgroup: Authors & Edits
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: null
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/distinct_node_editors.yaml

@@ -0,0 +1,20 @@
+name: ${prefix}distinct_node_editors
+node_type: metric
+display_name: Distinct Node Editors
+description: |-
+  Count of distinct users who authored a node revision. Slice by
+  ``created_at_day`` / ``created_at_week`` to get daily / weekly / monthly
+  active editors (people iterating on nodes, not just first-time creators).
+query: |-
+  SELECT COUNT(DISTINCT created_by_id)
+   FROM ${prefix}node_revision
+tags: []
+custom_metadata:
+  group: Activity
+  subgroup: Authors & Edits
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: null
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/dj.yaml

@@ -0,0 +1,33 @@
+name: DJ System Metadata
+description: This project contains DJ system metadata modeled in DJ itself.
+prefix: system.dj
+build:
+  priority:
+  # Dimension nodes (sources of breakdowns)
+  - repo
+  - namespace
+  - nodes
+  - node_revision
+  - dimension_link
+  - materialization
+  - deployment
+  - node_type
+  - is_active
+  - user
+  - date
+  # Metric nodes
+  - number_of_nodes
+  - number_of_namespaces
+  - number_of_materializations
+  - number_of_deployments
+  - median_deployment_duration_seconds
+  - distinct_node_authors
+  - distinct_node_editors
+  - median_revisions_per_node
+  - median_upstream_count
+  - median_downstream_count
+  - median_dim_links_per_node
+  - median_dimension_indegree
+  - number_of_orphan_nodes
+  - number_of_unused_dimensions
+  - number_of_branches

datajunction-0.0.154/datajunction/seed/nodes/is_active.dimension.yaml → datajunction-0.0.156/datajunction/seed/nodes/is_active.yaml

@@ -1,3 +1,5 @@
+name: ${prefix}is_active
+node_type: dimension
 display_name: Is Active
 description: ''
 query: |-
@@ -8,7 +10,10 @@ query: |-
       (true, 'Active'),
       (false, 'Deactivated')
   AS t (active_id, active_status);
-columns: []
+columns:
+- name: active_id
+  attributes:
+  - hidden
 primary_key:
 - active_id
 dimension_links: []

datajunction-0.0.154/datajunction/seed/nodes/materialization.dimension.yaml → datajunction-0.0.156/datajunction/seed/nodes/materialization.yaml

@@ -1,3 +1,5 @@
+name: ${prefix}materialization
+node_type: dimension
 display_name: Materialization
 description: ''
 query: |-
@@ -9,7 +11,25 @@ query: |-
     M.strategy,
     M.schedule
   FROM source.dj_metadata.public.materialization M
-columns: []
+columns:
+- name: id
+  attributes:
+  - hidden
+- name: node_revision_id
+  attributes:
+  - hidden
+- name: name
+  display_name: Materialization Name
+  attributes:
+  - hidden
+- name: job
+  display_name: Orchestrator Job
+  attributes:
+  - hidden
+- name: strategy
+  display_name: Strategy
+- name: schedule
+  display_name: Schedule
 primary_key:
 - id
 dimension_links:

datajunction-0.0.156/datajunction/seed/nodes/median_deployment_duration_seconds.yaml

@@ -0,0 +1,23 @@
+name: ${prefix}median_deployment_duration_seconds
+node_type: metric
+display_name: Median Deployment Duration (seconds)
+description: |-
+  Median elapsed time from ``created_at`` to ``updated_at`` per deployment.
+  Slice by ``deployment.status`` to compare successful vs failed deploys,
+  by ``deployment.source_type`` to compare git vs local, or by
+  ``deployment.created_at_week`` to trend over time.
+query: |-
+  SELECT MEDIAN(duration_seconds)
+   FROM ${prefix}deployment
+tags: []
+custom_metadata:
+  group: Activity
+  subgroup: Deployments
+  suggested_compare_by:
+  - system.dj.deployment.status
+required_dimensions: []
+direction: lower_is_better
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/median_dim_links_per_node.yaml

@@ -0,0 +1,26 @@
+name: ${prefix}median_dim_links_per_node
+node_type: metric
+display_name: Median Dimension Links per Linkable Node
+description: |-
+  Median number of dimension links defined on each node's current revision,
+  restricted to nodes that *can* define links (transforms and dimensions —
+  sources, metrics, and cubes inherit their dim links through their parents,
+  so including them would just dilute the median with zeros).
+
+  Slice by ``nodes.type`` to compare transforms vs dimensions, or by
+  ``namespace.top_level`` to see which areas have richer dim graphs.
+query: |-
+  SELECT MEDIAN(IF(type IN ('TRANSFORM', 'DIMENSION'), dim_link_count, NULL))
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Quality
+  subgroup: Lineage Shape
+  suggested_compare_by:
+  - system.dj.node_type.type
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/median_dimension_indegree.yaml

@@ -0,0 +1,23 @@
+name: ${prefix}median_dimension_indegree
+node_type: metric
+display_name: Median Dimension Indegree
+description: |-
+  Median number of distinct upstream nodes that link TO each dimension node
+  via their current revision. Higher = more reused dimensions. Computed only
+  over dimension nodes — non-dimension types are excluded so they don't
+  drag the median to zero.
+query: |-
+  SELECT MEDIAN(IF(type = 'DIMENSION', dim_indegree, NULL))
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Quality
+  subgroup: Lineage Shape
+  suggested_compare_by:
+  - system.dj.namespace.top_level
+required_dimensions: []
+direction: higher_is_better
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/median_downstream_count.yaml

@@ -0,0 +1,22 @@
+name: ${prefix}median_downstream_count
+node_type: metric
+display_name: Median Downstream Count
+description: |-
+  Median number of distinct downstream nodes per node (current revisions only).
+  Useful for understanding fanout — high values mean many consumers depend on
+  the typical node.
+query: |-
+  SELECT MEDIAN(child_count)
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Quality
+  subgroup: Lineage Shape
+  suggested_compare_by:
+  - system.dj.node_type.type
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/median_revisions_per_node.yaml

@@ -0,0 +1,21 @@
+name: ${prefix}median_revisions_per_node
+node_type: metric
+display_name: Median Revisions per Node
+description: |-
+  Median number of revisions across nodes. Slice by ``nodes.type`` to see
+  which entity types churn the most.
+query: |-
+  SELECT MEDIAN(revision_count)
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Quality
+  subgroup: Lineage Shape
+  suggested_compare_by:
+  - system.dj.node_type.type
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/median_upstream_count.yaml

@@ -0,0 +1,22 @@
+name: ${prefix}median_upstream_count
+node_type: metric
+display_name: Median Upstream Count
+description: |-
+  Median number of direct upstream parents across nodes (current revisions
+  only). Sliced by ``nodes.type``, surfaces how interconnected each entity
+  type tends to be.
+query: |-
+  SELECT MEDIAN(parent_count)
+   FROM ${prefix}nodes
+tags: []
+custom_metadata:
+  group: Quality
+  subgroup: Lineage Shape
+  suggested_compare_by:
+  - system.dj.node_type.type
+required_dimensions: []
+direction: null
+unit: null
+significant_digits: 2
+min_decimal_exponent: null
+max_decimal_exponent: null

datajunction-0.0.156/datajunction/seed/nodes/namespace.yaml

@@ -0,0 +1,148 @@
+name: ${prefix}namespace
+node_type: dimension
+display_name: Namespace
+description: |-
+  One row per distinct DJ namespace. Repo-related attributes are derived by
+  walking the nodenamespace tree:
+
+  - ``repo_container`` is the longest-prefix namespace that holds a
+    ``github_repo_path`` (the "repo container", e.g. ``ads``).
+  - ``branch_namespace`` is the longest-prefix namespace under that
+    container that has its own ``git_branch`` set.
+  - ``trunk_namespace`` is conventionally ``repo_container.default_branch``
+    (e.g. ``ads.main``).
+query: |-
+  WITH ns AS (
+    -- Every distinct namespace that DJ knows about: every namespace
+    -- containing a node, plus every nodenamespace row (so container /
+    -- empty namespaces like ``customer_service`` are still represented).
+    SELECT DISTINCT namespace
+    FROM source.dj_metadata.public.node
+    WHERE namespace IS NOT NULL
+    UNION
+    SELECT namespace
+    FROM source.dj_metadata.public.nodenamespace
+  ),
+  deployed_ns AS (
+    SELECT DISTINCT namespace
+    FROM source.dj_metadata.public.deployments
+  ),
+  repo_containers AS (
+    SELECT
+      NN.namespace AS container,
+      NN.github_repo_path,
+      NN.default_branch,
+      NN.git_only,
+      NN.namespace || '.' || NN.default_branch AS trunk_namespace
+    FROM source.dj_metadata.public.nodenamespace NN
+    WHERE NN.parent_namespace IS NULL
+      AND NN.github_repo_path IS NOT NULL
+  ),
+  branch_rows AS (
+    SELECT
+      NN.namespace AS branch_namespace,
+      NN.parent_namespace,
+      NN.git_branch
+    FROM source.dj_metadata.public.nodenamespace NN
+    WHERE NN.parent_namespace IS NOT NULL
+      AND NN.git_branch IS NOT NULL
+  ),
+  ns_repo_match AS (
+    SELECT
+      ns.namespace,
+      rc.container,
+      rc.github_repo_path,
+      rc.default_branch,
+      rc.git_only,
+      rc.trunk_namespace,
+      ROW_NUMBER() OVER (
+        PARTITION BY ns.namespace ORDER BY LENGTH(rc.container) DESC
+      ) AS rn
+    FROM ns
+    JOIN repo_containers rc
+      ON ns.namespace = rc.container OR ns.namespace LIKE rc.container || '.%'
+  ),
+  ns_repo AS (
+    SELECT namespace, container, github_repo_path, default_branch, git_only,
+           trunk_namespace
+    FROM ns_repo_match
+    WHERE rn = 1
+  ),
+  ns_branch_match AS (
+    SELECT
+      ns.namespace,
+      br.branch_namespace,
+      br.git_branch,
+      ROW_NUMBER() OVER (
+        PARTITION BY ns.namespace ORDER BY LENGTH(br.branch_namespace) DESC
+      ) AS rn
+    FROM ns
+    JOIN branch_rows br
+      ON ns.namespace = br.branch_namespace
+      OR ns.namespace LIKE br.branch_namespace || '.%'
+  ),
+  ns_branch AS (
+    SELECT namespace, branch_namespace, git_branch
+    FROM ns_branch_match
+    WHERE rn = 1
+  )
+  SELECT
+    ns.namespace                                                AS name,
+    CASE WHEN deployed_ns.namespace IS NOT NULL
+         THEN true ELSE false END                               AS has_deployments,
+    CAST(SPLIT_PART(ns.namespace, '.', 1) AS VARCHAR)           AS top_level,
+    CAST(LENGTH(ns.namespace) - LENGTH(REPLACE(ns.namespace, '.', '')) + 1 AS integer) AS depth,
+    ns_repo.github_repo_path                                    AS github_repo_path,
+    ns_repo.container                                           AS repo_container,
+    ns_repo.trunk_namespace                                     AS trunk_namespace,
+    COALESCE(ns_branch.git_branch, ns_repo.default_branch)      AS git_branch,
+    CASE WHEN ns_repo.github_repo_path IS NOT NULL
+         THEN true ELSE false END                               AS is_repo_backed,
+    CASE WHEN ns.namespace = ns_repo.container
+         THEN true ELSE false END                               AS is_repo_container,
+    CASE WHEN ns.namespace = ns_repo.trunk_namespace
+         THEN true ELSE false END                               AS is_trunk_namespace,
+    CASE
+      WHEN ns_branch.branch_namespace IS NOT NULL
+        AND ns_branch.branch_namespace != ns_repo.trunk_namespace
+      THEN true ELSE false
+    END                                                         AS is_branch_namespace
+  FROM ns
+  LEFT JOIN deployed_ns ON ns.namespace = deployed_ns.namespace
+  LEFT JOIN ns_repo ON ns.namespace = ns_repo.namespace
+  LEFT JOIN ns_branch ON ns.namespace = ns_branch.namespace
+columns:
+- name: name
+  display_name: Namespace
+- name: top_level
+  display_name: Top-level Namespace
+- name: depth
+  display_name: Namespace Depth
+- name: has_deployments
+  display_name: Has Deployments
+- name: github_repo_path
+  display_name: Repository
+  attributes:
+  - hidden
+- name: repo_container
+  display_name: Repo Container
+- name: trunk_namespace
+  display_name: Trunk Namespace
+- name: git_branch
+  display_name: Git Branch
+- name: is_repo_backed
+  display_name: Is Repo-Backed
+- name: is_repo_container
+  display_name: Is Repo Container
+- name: is_trunk_namespace
+  display_name: Is Trunk Namespace
+- name: is_branch_namespace
+  display_name: Is Branch Namespace
+primary_key:
+- name
+dimension_links:
+- type: join
+  dimension_node: ${prefix}repo
+  join_type: left
+  join_on: ${prefix}namespace.github_repo_path = ${prefix}repo.repository
+tags: []