Update the exports to rely on the bulk endpoint for speed

src/pyetm/services/scenario_runners/fetch_output_curves.py

@@ -1,8 +1,8 @@
 import io
-from typing import Any, Dict
+from typing import Any, Dict, List, Optional
+import pandas as pd
 from pyetm.services.scenario_runners.base_runner import BaseRunner
 from pyetm.services.scenario_runners.fetch_curves_generic import (
-    GenericCurveBulkRunner,
     GenericCurveDownloadRunner,
 )
 from ..service_result import ServiceResult
@@ -26,7 +26,6 @@ def run(
 class FetchAllOutputCurvesRunner(BaseRunner[Dict[str, io.StringIO]]):
     """Download all known output curves."""
 
-    # Known curve types from the Rails controller
     CURVE_TYPES = [
         "merit_order",
         "electricity_price",
@@ -44,10 +43,63 @@ class FetchAllOutputCurvesRunner(BaseRunner[Dict[str, io.StringIO]]):
     def run(
         client: BaseClient,
         scenario: Any,
+        curve_types: Optional[List[str]] = None,
     ) -> ServiceResult[Dict[str, io.StringIO]]:
-        return GenericCurveBulkRunner.run(
-            client,
-            scenario,
-            FetchAllOutputCurvesRunner.CURVE_TYPES,
-            curve_type="output",
-        )
+        """
+        Uses the bulk endpoint to fetch output curves as a single CSV,
+        then splits them into their carrier groups
+
+        Args:
+            client: The API client
+            scenario: The scenario object
+            curve_types: Optional list of curve types to fetch. If None, fetches all default curves.
+        """
+        try:
+            types_to_fetch = curve_types or FetchAllOutputCurvesRunner.CURVE_TYPES
+
+            path = f"/scenarios/{scenario.id}/bulk_output_curves"
+            params = {"curve_types": ",".join(types_to_fetch)}
+            resp = client.session.get(
+                path, params=params, headers={"Accept": "text/csv"}
+            )
+
+            if not resp.ok:
+                return ServiceResult.fail([f"{resp.status_code}: {resp.text}"])
+
+            try:
+                csv_text = resp.content.decode("utf-8")
+                df = pd.read_csv(io.StringIO(csv_text), index_col=0)
+            except Exception as e:
+                return ServiceResult.fail([f"Failed to parse bulk CSV: {e}"])
+
+            results: Dict[str, io.StringIO] = {}
+            warnings: list[str] = []
+
+            groups: Dict[str, list[str]] = {}
+            for col in df.columns:
+                base = str(col)
+                for sep in (":", "/"):
+                    if sep in base:
+                        base = base.split(sep, 1)[0]
+                        break
+                groups.setdefault(base, []).append(col)
+
+            for base, cols in groups.items():
+                try:
+                    sub = df[cols].dropna(how="all")
+                    buf = io.StringIO()
+                    sub.to_csv(buf, index=True)
+                    buf.seek(0)
+                    results[base] = buf
+                except Exception as e:
+                    warnings.append(f"{base}: Failed to prepare CSV: {e}")
+
+            if results:
+                return ServiceResult.ok(data=results, errors=warnings or None)
+            else:
+                return ServiceResult.fail(warnings or ["No curves present in CSV"])
+
+        except (PermissionError, ValueError, ConnectionError) as e:
+            return ServiceResult.fail([str(e)])
+        except Exception as e:
+            return ServiceResult.fail([str(e)])