mammoth-cli 2.1.5__tar.gz → 2.1.7__tar.gz

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mammoth-cli
-Version: 2.1.5
+Version: 2.1.7
 Summary: Command-line interface for the Mammoth Analytics platform
 Project-URL: homepage, https://mammoth.io
 Project-URL: repository, https://github.com/EdgeMetric/mammothsdk

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/mammoth_cli/__init__.py

@@ -6,7 +6,7 @@ not implement a second HTTP client and does not call private SDK members.
 
 from __future__ import annotations
 
-__version__ = "2.1.5"
+__version__ = "2.1.7"
 
 # The versioned machine output envelope contract.
 SCHEMA_VERSION = 1

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/mammoth_cli/app.py

@@ -30,6 +30,7 @@ from typer.core import TyperGroup
 from mammoth_cli import __version__
 from mammoth_cli.commands import BESPOKE
 from mammoth_cli.commands.registry import HANDLERS, Handler
+from mammoth_cli.commands.schema import IN_PLACE_RECIPE, edits_view_in_place
 from mammoth_cli.context import profiles
 from mammoth_cli.errors.envelope import EXIT_USAGE, CliError, not_implemented_error
 from mammoth_cli.manifest.loader import command_by_id, load_commands
@@ -929,6 +930,8 @@ def _command_help(command_id: str, record: dict[str, Any] | None) -> str | None:
         summary = doc.strip().splitlines()[0].strip().replace("``", "'")
         if summary:
             parts.append(summary)
+    if record is not None and edits_view_in_place(record):
+        parts.append(IN_PLACE_RECIPE)
     example = (record or {}).get("agent_example")
     if example:
         parts.append(f"Example: {example}")

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/mammoth_cli/commands/schema.py

@@ -469,6 +469,28 @@ _SCOPE_REQUIREMENTS: dict[str, dict[str, Any]] = {
 }
 
 _MAX_FIND_RESULTS = 20
+# View transform/task commands add a step to the GIVEN view. An agent asked for
+# a NEW dataset must know that before calling, so the statement is carried by
+# every such command's contract (schema get ``preconditions``, find, --help).
+_IN_PLACE_PREFIXES = ("view.transform.", "view.task.")
+IN_PLACE_RECIPE = (
+    "EDITS THE GIVEN VIEW IN PLACE: this adds a step to the view and changes that view "
+    "and its dataset's output. To make a NEW dataset and leave the source untouched: "
+    "`mammoth view create SOURCE_DATASET_ID` (working view), do the steps on that working "
+    'view, then `mammoth view export dataset WORKING_VIEW_ID --input \'{"dataset_name": "..."}\'`.'
+)
+_IN_PLACE_KEYWORDS = (
+    "new dataset create make in place edits given view working view source untouched"
+)
+
+
+def edits_view_in_place(record: dict[str, Any]) -> bool:
+    """True for a view transform/task command that changes an existing view."""
+    return bool(record.get("edits_target")) and str(record.get("command_id", "")).startswith(
+        _IN_PLACE_PREFIXES
+    )
+
+
 # Outranks any word-overlap score: a command named by its full path comes first.
 _NAMED_COMMAND_BOOST = 10_000
 # Bag-of-words scoring cannot tell "create a NEW VIEW from an existing
@@ -944,7 +966,9 @@ def _compact_contract(record: dict[str, Any]) -> dict[str, Any]:
             "the intended postcondition."
         )
     restrictions = record.get("known_restrictions")
-    if restrictions is None:
+    if edits_view_in_place(record):
+        restrictions = f"{IN_PLACE_RECIPE} {restrictions or ''}".strip()
+    elif restrictions is None:
         required_positionals = [
             str(item.get("metavar") or item.get("name"))
             for item in record.get("positionals", [])
@@ -1664,6 +1688,7 @@ def find_schemas(
             (30, f"{record.get('human_example', '')} {record.get('agent_example', '')}"),
             (20, positional_help),
             (15, _operation_hints_by_command().get(command_id, "")),
+            (15, _IN_PLACE_KEYWORDS if edits_view_in_place(record) else ""),
             (60, _COMMAND_DISCOVERY_PURPOSES.get(command_id, "")),
             (3, _GROUP_DISCOVERY_PURPOSES.get(command_path.split()[0], "")),
         )

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/mammoth_cli/commands/view.py

@@ -1038,6 +1038,10 @@ def key_overlap(left_counts: dict[Any, int], right_counts: dict[Any, int]) -> di
         "unmatched_keys": _unmatched(left_counts, right_keys),
         "right_rows_checked": right_total,
         "right_match_rate": round(right_matched / right_total, 3) if right_total else None,
+        "right_unmatched_rows": right_total - right_matched,
+        "right_unmatched_key_count": sum(
+            1 for key in right_counts if key in (None, "") or key not in left_keys
+        ),
         "right_unmatched_keys": _unmatched(right_counts, left_keys),
     }
 
@@ -1097,13 +1101,31 @@ def join_dry_run_preview(
     }
 
 
-def with_join_check(data: Any, before: Any, after: dict[str, Any], document: dict[str, Any]) -> Any:
+_RIGHT_CHECK_FIELDS = (
+    "right_rows_checked",
+    "right_match_rate",
+    "right_unmatched_rows",
+    "right_unmatched_key_count",
+    "right_unmatched_keys",
+)
+
+
+def with_join_check(
+    data: Any,
+    before: Any,
+    after: dict[str, Any],
+    document: dict[str, Any],
+    right: dict[str, Any] | None = None,
+) -> Any:
     """Add ``join_check`` (row counts, columns added, match rate) to a join result.
 
     ``unmatched_rows`` counts rows where every added column is blank: for a
     LEFT join these are rows whose key found no match in the other view. It is
     counted over the whole view when ``after`` carries ``unmatched_total``
     (see :func:`join_after_snapshot`), else over the sampled first page.
+    ``right`` is a :func:`join_dry_run_preview` of the same join: its right-side
+    fields (keys only the other view has) are copied in, so a LEFT join does not
+    hide them.
     """
     if not isinstance(data, dict) or not isinstance(before, dict):
         return data
@@ -1156,6 +1178,14 @@ def with_join_check(data: Any, before: Any, after: dict[str, Any], document: dic
                 f"{before_n - after_n} rows had no match and were dropped "
                 "(an INNER join keeps matched rows only)."
             )
+    if right and right.get("checked"):
+        check.update({k: right[k] for k in _RIGHT_CHECK_FIELDS})
+        if right["right_unmatched_key_count"]:
+            notes.append(
+                f"{right['right_unmatched_key_count']} key(s) in the other view found no "
+                "partner here (right_unmatched_keys lists the first ones); report them "
+                "as well as the unmatched rows on this side."
+            )
     if notes:
         check["notes"] = notes
     return {**data, "join_check": check}
@@ -2713,6 +2743,7 @@ def view_task_add(invocation: Invocation) -> HandlerResult:
         reject_pipeline_reference_errors,
         reject_task_runtime_error,
         require_expected_task_count,
+        with_in_place_note,
     )
 
     with open_service(invocation) as (service, auth):
@@ -2723,7 +2754,7 @@ def view_task_add(invocation: Invocation) -> HandlerResult:
         reject_task_runtime_error(
             service, dataview_id, kwargs.get("dataset_id"), data, submitted_at
         )
-    return data, _meta(invocation, auth.workspace_id, None)
+    return with_in_place_note(data, dataview_id), _meta(invocation, auth.workspace_id, None)
 
 
 def view_task_delete(invocation: Invocation) -> HandlerResult:

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/mammoth_cli/commands/view_ops.py

@@ -375,6 +375,25 @@ def _dispatch_view(
     return data, _meta(invocation, auth.workspace_id)
 
 
+def with_in_place_note(data: Any, view_id: int) -> Any:
+    """Tell the caller this write changed view ``view_id`` itself, and how to avoid that.
+
+    An agent asked for a NEW dataset edited the source's own view; the way to
+    leave the source untouched is a working view, then ``view export dataset``.
+    """
+    if not isinstance(data, dict):
+        return data
+    return {
+        **data,
+        "in_place_note": (
+            f"This edited existing view {view_id} in place. If the user asked for a NEW "
+            "dataset or the source must stay untouched, do not edit it: run 'mammoth view "
+            "create DATASET_ID' for a working view, make the steps on that view, then "
+            "'mammoth view export dataset' it into the new dataset."
+        ),
+    }
+
+
 def _flag_unsettled_pipeline(service: Any, dataset_id: int, view_id: int, data: Any) -> None:
     """Wait for the pipeline after a write with no row check, and record what it finds.
 
@@ -1193,6 +1212,7 @@ def view_transform_join(invocation: Invocation) -> HandlerResult:
             state,
             join_after_snapshot(service, dataset_id, view_id, project_id, state),
             document,
+            _join_preview(service, dataset_id, view_id, document, project_id),
         )
 
     try:

{mammoth_cli-2.1.5 → mammoth_cli-2.1.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mammoth-cli"
-version = "2.1.5"
+version = "2.1.7"
 description = "Command-line interface for the Mammoth Analytics platform"
 authors = [{name = "Mammoth Analytics", email = "support@mammoth.io"}]
 readme = "README.md"