Add processor to workspace state

.gitignore

@@ -1,3 +1,5 @@
+.mise.toml
+
 .vunnel.yaml
 .grype-db.yaml
 .grype.yaml

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 # the default is to run these on commit + push
 default_stages:
-  - push
+  - pre-push
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
@@ -10,47 +10,47 @@ repos:
     # Prevent giant files from being committed.
     - id: check-added-large-files
       stages:
-        - push
+        - pre-push
 
     # Checks for a common error of placing code before the docstring.
     - id: check-docstring-first
       stages:
-        - push
+        - pre-push
 
     # Attempts to load all yaml files to verify syntax.
     - id: check-yaml
       stages:
-        - push
+        - pre-push
 
     # Attempts to load all json files to verify syntax.
     - id: check-json
       stages:
-        - push
+        - pre-push
 
     # Makes sure files end in a newline and only a newline.
     - id: end-of-file-fixer
       stages:
-        - push
+        - pre-push
 
     # Trims trailing whitespace.
     - id: trailing-whitespace
       stages:
-        - push
+        - pre-push
 
     # Check for files that contain merge conflict strings.
     - id: check-merge-conflict
       stages:
-        - push
+        - pre-push
 
     # Simply check whether files parse as valid python.
     - id: check-ast
       stages:
-        - push
+        - pre-push
 
     # Check for files with names that would conflict on a case-insensitive filesystem like MacOS HFS+ or Windows FAT.
     - id: check-case-conflict
       stages:
-        - push
+        - pre-push
 
   # why use the local repo instead of hosted hooks? so that dependencies are centrally managed through poetry
   - repo: local

schema/provider-workspace-state/schema-1.0.3.json

@@ -0,0 +1,84 @@
+{
+  "$schema": "http://json-schema.org/draft-04/schema#",
+  "type": "object",
+  "title": "provider-workspace-state",
+  "description": "describes the filesystem state of a provider workspace directory",
+  "properties": {
+    "provider": {
+      "type": "string"
+    },
+    "urls": {
+      "type": "array",
+      "items": [
+        {
+          "type": "string"
+        }
+      ]
+    },
+    "store": {
+      "type": "string"
+    },
+    "timestamp": {
+      "type": "string"
+    },
+    "listing": {
+      "type": "object",
+      "properties": {
+        "digest": {
+          "type": "string"
+        },
+        "path": {
+          "type": "string"
+        },
+        "algorithm": {
+          "type": "string"
+        }
+      },
+      "required": [
+        "digest",
+        "path",
+        "algorithm"
+      ]
+    },
+    "version": {
+      "type": "integer",
+      "description": "version describing the result data shape + the provider processing behavior semantics"
+    },
+    "distribution_version": {
+      "type": "integer",
+      "description": "version describing purely the result data shape"
+    },
+    "processor": {
+      "type": "string",
+      "description": "the application that created the workspace, formatted as 'name@version'"
+    },
+    "schema": {
+      "type": "object",
+      "properties": {
+        "version": {
+          "type": "string"
+        },
+        "url": {
+          "type": "string"
+        }
+      },
+      "required": [
+        "version",
+        "url"
+      ]
+    },
+    "stale": {
+      "type": "boolean",
+      "description": "set to true if the workspace is stale and cannot be used for an incremental update"
+    }
+  },
+  "required": [
+    "provider",
+    "urls",
+    "store",
+    "timestamp",
+    "listing",
+    "version",
+    "schema"
+  ]
+}

src/vunnel/schema.py

@@ -4,7 +4,7 @@
 from dataclasses import dataclass
 
 # Note: this metadata.json file currently is not allowed to have a breaking change
-PROVIDER_WORKSPACE_STATE_SCHEMA_VERSION = "1.0.2"
+PROVIDER_WORKSPACE_STATE_SCHEMA_VERSION = "1.0.3"
 
 PROVIDER_ARCHIVE_LISTING_SCHEMA_VERSION = "1.0.0"
 MATCH_EXCLUSION_SCHEMA_VERSION = "1.0.0"

src/vunnel/workspace.py

@@ -17,6 +17,7 @@
 from vunnel.utils import hasher
 
 if TYPE_CHECKING:
+    import types
     from collections.abc import Generator
 
 METADATA_FILENAME = "metadata.json"
@@ -38,10 +39,31 @@ class State(DataClassDictMixin):
     timestamp: datetime.datetime
     version: int = 1
     distribution_version: int = 1
+    processor: str | None = None
     listing: Optional[File] = None  # noqa:UP007  # why use Optional? mashumaro does not support this on python 3.9
     schema: schema_def.Schema = field(default_factory=schema_def.ProviderStateSchema)
     stale: bool = False
 
+    def __post_init__(self) -> None:
+        if not self.processor:
+            metadata: types.ModuleType
+            package_name = "vunnel"
+            version = "dev"
+            try:
+                from importlib import metadata
+            except ImportError:
+                # Python < 3.8
+                import importlib_metadata as metadata
+
+            try:
+                if not metadata:
+                    raise metadata.PackageNotFoundError
+                version = metadata.version(package_name)
+            except metadata.PackageNotFoundError:
+                version = "unknown"
+
+            self.processor = f"{package_name}@{version}"
+
     @staticmethod
     def read(root: str) -> State:
         metadata_path = os.path.join(root, METADATA_FILENAME)