Add column types system, refs #2664

datasette/app.py

@@ -354,6 +354,7 @@ def __init__(
         self.immutables = set(immutables or [])
         self.databases = collections.OrderedDict()
         self.actions = {}  # .invoke_startup() will populate this
+        self._column_types = {}  # .invoke_startup() will populate this
         try:
             self._refresh_schemas_lock = asyncio.Lock()
         except RuntimeError as rex:
@@ -692,12 +693,23 @@ async def invoke_startup(self):
                         action_abbrs[action.abbr] = action
                     self.actions[action.name] = action
 
+        # Register column types (classes, not instances)
+        self._column_types = {}
+        for hook in pm.hook.register_column_types(datasette=self):
+            if hook:
+                for ct_cls in hook:
+                    if ct_cls.name in self._column_types:
+                        raise StartupError(f"Duplicate column type name: {ct_cls.name}")
+                    self._column_types[ct_cls.name] = ct_cls
+
         for hook in pm.hook.prepare_jinja2_environment(
             env=self._jinja_env, datasette=self
         ):
             await await_me_maybe(hook)
         # Ensure internal tables and metadata are populated before startup hooks
         await self._refresh_schemas()
+        # Load column_types from config into internal DB
+        await self._apply_column_types_config()
         for hook in pm.hook.startup(datasette=self):
             await await_me_maybe(hook)
         self._startup_invoked = True
@@ -945,6 +957,102 @@ async def set_column_metadata(
             [database_name, resource_name, column_name, key, value],
         )
 
+    # Column types API
+
+    async def _apply_column_types_config(self):
+        """Load column_types from datasette.json config into the internal DB."""
+        import logging
+
+        for db_name, db_conf in (self.config or {}).get("databases", {}).items():
+            for table_name, table_conf in db_conf.get("tables", {}).items():
+                for col_name, ct in table_conf.get("column_types", {}).items():
+                    if isinstance(ct, str):
+                        col_type, config = ct, None
+                    else:
+                        col_type = ct["type"]
+                        config = ct.get("config")
+                    if col_type not in self._column_types:
+                        logging.warning(
+                            "column_types config references unknown type %r "
+                            "for %s.%s.%s",
+                            col_type,
+                            db_name,
+                            table_name,
+                            col_name,
+                        )
+                    await self.set_column_type(
+                        db_name, table_name, col_name, col_type, config
+                    )
+
+    async def get_column_type(self, database: str, resource: str, column: str):
+        """
+        Return a ColumnType instance (with config baked in) for a specific
+        column, or None if no column type is assigned.
+        """
+        row = await self.get_internal_database().execute(
+            "SELECT column_type, config FROM column_types "
+            "WHERE database_name = ? AND resource_name = ? AND column_name = ?",
+            [database, resource, column],
+        )
+        rows = row.rows
+        if not rows:
+            return None
+        ct_name, config = rows[0]
+        ct_cls = self._column_types.get(ct_name)
+        if ct_cls is None:
+            return None
+        return ct_cls(config=json.loads(config) if config else None)
+
+    async def get_column_types(self, database: str, resource: str) -> dict:
+        """
+        Return {column_name: ColumnType instance (with config)}
+        for all columns with assigned types on the given resource.
+        """
+        rows = await self.get_internal_database().execute(
+            "SELECT column_name, column_type, config FROM column_types "
+            "WHERE database_name = ? AND resource_name = ?",
+            [database, resource],
+        )
+        result = {}
+        for row in rows.rows:
+            col_name, ct_name, config = row
+            ct_cls = self._column_types.get(ct_name)
+            if ct_cls is not None:
+                result[col_name] = ct_cls(config=json.loads(config) if config else None)
+        return result
+
+    async def set_column_type(
+        self,
+        database: str,
+        resource: str,
+        column: str,
+        column_type: str,
+        config: dict = None,
+    ) -> None:
+        """Assign a column type. Overwrites any existing assignment."""
+        await self.get_internal_database().execute_write(
+            """INSERT OR REPLACE INTO column_types
+               (database_name, resource_name, column_name, column_type, config)
+               VALUES (?, ?, ?, ?, ?)""",
+            [
+                database,
+                resource,
+                column,
+                column_type,
+                json.dumps(config) if config else None,
+            ],
+        )
+
+    async def remove_column_type(
+        self, database: str, resource: str, column: str
+    ) -> None:
+        """Remove a column type assignment."""
+        await self.get_internal_database().execute_write(
+            "DELETE FROM column_types "
+            "WHERE database_name = ? AND resource_name = ? AND column_name = ?",
+            [database, resource, column],
+        )
+
     def get_internal_database(self):
         return self._internal_database
 

datasette/column_types.py

@@ -0,0 +1,44 @@
+class ColumnType:
+    """
+    Base class for column types.
+
+    Subclasses must define ``name`` and ``description`` as class attributes:
+
+    - ``name``: Unique identifier string. Lowercase, no spaces.
+      Examples: "markdown", "file", "email", "url", "point", "image".
+    - ``description``: Human-readable label for admin UI dropdowns.
+      Examples: "Markdown text", "File reference", "Email address".
+
+    Instantiate with an optional ``config`` dict to bind per-column
+    configuration::
+
+        ct = MyColumnType(config={"key": "value"})
+        ct.config  # {"key": "value"}
+    """
+
+    name: str
+    description: str
+
+    def __init__(self, config=None):
+        self.config = config
+
+    async def render_cell(self, value, column, table, database, datasette, request):
+        """
+        Return an HTML string to render this cell value, or None to
+        fall through to the default render_cell plugin hook chain.
+        """
+        return None
+
+    async def validate(self, value, datasette):
+        """
+        Validate a value before it is written. Return None if valid,
+        or a string error message if invalid.
+        """
+        return None
+
+    async def transform_value(self, value, datasette):
+        """
+        Transform a value before it appears in JSON API output.
+        Return the transformed value. Default: return unchanged.
+        """
+        return value

datasette/default_column_types.py

@@ -0,0 +1,78 @@
+import json
+import re
+
+import markupsafe
+
+from datasette import hookimpl
+from datasette.column_types import ColumnType
+
+
+class UrlColumnType(ColumnType):
+    name = "url"
+    description = "URL"
+
+    async def render_cell(self, value, column, table, database, datasette, request):
+        if not value or not isinstance(value, str):
+            return None
+        escaped = markupsafe.escape(value.strip())
+        return markupsafe.Markup(f'<a href="{escaped}">{escaped}</a>')
+
+    async def validate(self, value, datasette):
+        if value is None or value == "":
+            return None
+        if not isinstance(value, str):
+            return "URL must be a string"
+        if not re.match(r"^https?://\S+$", value.strip()):
+            return "Invalid URL"
+        return None
+
+
+class EmailColumnType(ColumnType):
+    name = "email"
+    description = "Email address"
+
+    async def render_cell(self, value, column, table, database, datasette, request):
+        if not value or not isinstance(value, str):
+            return None
+        escaped = markupsafe.escape(value.strip())
+        return markupsafe.Markup(f'<a href="mailto:{escaped}">{escaped}</a>')
+
+    async def validate(self, value, datasette):
+        if value is None or value == "":
+            return None
+        if not isinstance(value, str):
+            return "Email must be a string"
+        if not re.match(r"^[^@\s]+@[^@\s]+\.[^@\s]+$", value.strip()):
+            return "Invalid email address"
+        return None
+
+
+class JsonColumnType(ColumnType):
+    name = "json"
+    description = "JSON data"
+
+    async def render_cell(self, value, column, table, database, datasette, request):
+        if value is None:
+            return None
+        try:
+            parsed = json.loads(value) if isinstance(value, str) else value
+            formatted = json.dumps(parsed, indent=2)
+            escaped = markupsafe.escape(formatted)
+            return markupsafe.Markup(f"<pre>{escaped}</pre>")
+        except (json.JSONDecodeError, TypeError):
+            return None
+
+    async def validate(self, value, datasette):
+        if value is None or value == "":
+            return None
+        if isinstance(value, str):
+            try:
+                json.loads(value)
+            except json.JSONDecodeError:
+                return "Invalid JSON"
+        return None
+
+
+@hookimpl
+def register_column_types(datasette):
+    return [UrlColumnType, EmailColumnType, JsonColumnType]

datasette/hookspecs.py

@@ -55,7 +55,17 @@ def publish_subcommand(publish):
 
 
 @hookspec
-def render_cell(row, value, column, table, pks, database, datasette, request):
+def render_cell(
+    row,
+    value,
+    column,
+    table,
+    pks,
+    database,
+    datasette,
+    request,
+    column_type,
+):
     """Customize rendering of HTML table cell values"""
 
 
@@ -74,6 +84,11 @@ def register_actions(datasette):
     """Register actions: returns a list of datasette.permission.Action objects"""
 
 
+@hookspec
+def register_column_types(datasette):
+    """Return a list of ColumnType instances"""
+
+
 @hookspec
 def register_routes(datasette):
     """Register URL routes: return a list of (regex, view_function) pairs"""

datasette/plugins.py

@@ -25,6 +25,7 @@
     "datasette.default_permissions",
     "datasette.default_permissions.tokens",
     "datasette.default_actions",
+    "datasette.default_column_types",
     "datasette.default_magic_parameters",
     "datasette.blob_renderer",
     "datasette.default_menu_links",

datasette/utils/internal_db.py

@@ -103,6 +103,15 @@ async def initialize_metadata_tables(db):
             value text,
             unique(database_name, resource_name, column_name, key)
         );
+
+        CREATE TABLE IF NOT EXISTS column_types (
+            database_name TEXT NOT NULL,
+            resource_name TEXT NOT NULL,
+            column_name TEXT NOT NULL,
+            column_type TEXT NOT NULL,
+            config TEXT,
+            PRIMARY KEY (database_name, resource_name, column_name)
+        );
             """))
 
 

datasette/views/database.py

@@ -1205,6 +1205,7 @@ async def display_rows(datasette, database, request, rows, columns):
                 database=database,
                 datasette=datasette,
                 request=request,
+                column_type=None,
             ):
                 candidate = await await_me_maybe(candidate)
                 if candidate is not None:

datasette/views/row.py

@@ -179,26 +179,41 @@ async def template_data():
 
         if "render_cell" in extras:
             # Call render_cell plugin hook for each cell
+            ct_map = await self.ds.get_column_types(database, table)
             rendered_rows = []
             for row in rows:
                 rendered_row = {}
                 for value, column in zip(row, columns):
-                    # Call render_cell plugin hook
+                    ct = ct_map.get(column)
                     plugin_display_value = None
-                    for candidate in pm.hook.render_cell(
-                        row=row,
-                        value=value,
-                        column=column,
-                        table=table,
-                        pks=resolved.pks,
-                        database=database,
-                        datasette=self.ds,
-                        request=request,
-                    ):
-                        candidate = await await_me_maybe(candidate)
+                    # Try column type render_cell first
+                    if ct:
+                        candidate = await ct.render_cell(
+                            value=value,
+                            column=column,
+                            table=table,
+                            database=database,
+                            datasette=self.ds,
+                            request=request,
+                        )
                         if candidate is not None:
                             plugin_display_value = candidate
-                            break
+                    if plugin_display_value is None:
+                        for candidate in pm.hook.render_cell(
+                            row=row,
+                            value=value,
+                            column=column,
+                            table=table,
+                            pks=resolved.pks,
+                            database=database,
+                            datasette=self.ds,
+                            request=request,
+                            column_type=ct,
+                        ):
+                            candidate = await await_me_maybe(candidate)
+                            if candidate is not None:
+                                plugin_display_value = candidate
+                                break
                     if plugin_display_value:
                         rendered_row[column] = str(plugin_display_value)
                 rendered_rows.append(rendered_row)
@@ -352,6 +367,15 @@ async def post(self, request):
 
         update = data["update"]
 
+        # Validate column types
+        from datasette.views.table import _validate_column_types
+
+        ct_errors = await _validate_column_types(
+            self.ds, resolved.db.name, resolved.table, [update]
+        )
+        if ct_errors:
+            return _error(ct_errors, 400)
+
         alter = data.get("alter")
         if alter and not await self.ds.allowed(
             action="alter-table",