satif-ai 0.2.5__tar.gz → 0.2.7__tar.gz

{satif_ai-0.2.5 → satif_ai-0.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: satif-ai
-Version: 0.2.5
+Version: 0.2.7
 Summary: AI Agents for Satif
 License: MIT
 Author: Bryan Djafer

{satif_ai-0.2.5 → satif_ai-0.2.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "satif-ai"
-version = "0.2.5"
+version = "0.2.7"
 description = "AI Agents for Satif"
 authors = [
     {name = "Bryan Djafer", email = "bryan.djafer@syncpulse.fr"}

{satif_ai-0.2.5 → satif_ai-0.2.7}/satif_ai/code_builders/transformation.py

@@ -1,8 +1,9 @@
 import base64
 import os
 import re
+from collections import defaultdict
 from pathlib import Path
-from typing import Dict, List, Optional, Union
+from typing import Any, Dict, List, Optional, Union
 
 from agents import Agent, Runner, function_tool
 from agents.mcp.server import MCPServer
@@ -18,6 +19,37 @@ OUTPUT_TARGET_FILES: Optional[Dict[Union[str, Path], str]] = None
 SCHEMA_ONLY: Optional[bool] = None
 
 
+def _format_comparison_output(
+    comparison_result: Dict[str, Any],
+    schema_only_mode: Optional[bool],
+    source_file_display_name: str,
+    target_file_display_name: str,
+) -> str:
+    """
+    Formats the comparison result string, with special handling for schema_only mode
+    where files are equivalent due to being empty.
+    """
+    base_message_prefix = f"Comparison for {source_file_display_name} [SOURCE] with {target_file_display_name} [TARGET]:"
+
+    if schema_only_mode is True and comparison_result.get("are_equivalent") is True:
+        details = comparison_result.get("details", {})
+        row_comparison = details.get("row_comparison", {})
+
+        row_count1 = row_comparison.get("row_count1")
+        row_count2 = row_comparison.get("row_count2")
+
+        if (
+            isinstance(row_count1, (int, float))
+            and row_count1 == 0
+            and isinstance(row_count2, (int, float))
+            and row_count2 == 0
+        ):
+            return f"{base_message_prefix} Files have the same headers but are both empty (no data rows). This should not happen. Please verify the instructions and try again."
+
+    # Default formatting if the special condition isn't met
+    return f"{base_message_prefix} {comparison_result}"
+
+
 @function_tool
 async def execute_transformation(code: str) -> str:
     """Executes the transformation code on the input and returns the
@@ -33,7 +65,7 @@ async def execute_transformation(code: str) -> str:
     generated_output_path = code_transformer.export(INPUT_SDIF_PATH)
 
     comparisons = []
-    comparator_kwargs = {}
+    comparator_kwargs = {"check_structure_only": True}
     if SCHEMA_ONLY:
         comparator_kwargs["check_structure_only"] = True
 
@@ -53,9 +85,13 @@ async def execute_transformation(code: str) -> str:
                 comparison = comparator.compare(
                     generated_file_path, output_base_file, **comparator_kwargs
                 )
-                comparisons.append(
-                    f"Comparison for {generated_file_path} [SOURCE] with {output_target_file_name} [TARGET]: {comparison}"
+                formatted_message = _format_comparison_output(
+                    comparison,
+                    SCHEMA_ONLY,
+                    generated_file_path,
+                    output_target_file_name,
                 )
+                comparisons.append(formatted_message)
             else:
                 comparisons.append(
                     f"Error: {output_target_file_name} not found in the generated output"
@@ -69,9 +105,13 @@ async def execute_transformation(code: str) -> str:
             comparison = comparator.compare(
                 generated_output_path, output_file, **comparator_kwargs
             )
-            comparisons.append(
-                f"Comparison for {generated_output_path} [SOURCE] with {output_target_file_name} [TARGET]: {comparison}"
+            formatted_message = _format_comparison_output(
+                comparison,
+                SCHEMA_ONLY,
+                str(generated_output_path),
+                output_target_file_name,
             )
+            comparisons.append(formatted_message)
         else:
             comparisons.append(
                 "Error: Single output file generated but multiple target files expected"
@@ -112,6 +152,7 @@ class TransformationAsyncCodeBuilder(AsyncCodeBuilder):
         output_sdif: Optional[Path] = None,  # This will now be relative or None
         instructions: Optional[str] = None,
         schema_only: bool = False,
+        representer_options_for_build: Optional[Dict[str, Any]] = None,
     ) -> str:
         global INPUT_SDIF_PATH, OUTPUT_TARGET_FILES, SCHEMA_ONLY
         # INPUT_SDIF_PATH is used by execute_transformation tool, needs to be accessible from where that tool runs.
@@ -179,16 +220,21 @@ class TransformationAsyncCodeBuilder(AsyncCodeBuilder):
         else:
             OUTPUT_TARGET_FILES = {}
 
-        output_representation = {}
+        output_representation = defaultdict(dict)
         if OUTPUT_TARGET_FILES:
             for file_key_abs_path in list(OUTPUT_TARGET_FILES.keys()):
                 agent_facing_name = OUTPUT_TARGET_FILES[file_key_abs_path]
-                print(f"Representing {agent_facing_name} from {file_key_abs_path}!!")
+                print(f"Representing {agent_facing_name} from {file_key_abs_path}")
                 try:
                     # Representer uses the absolute path (file_key_abs_path) to read the example file.
-                    output_representation[agent_facing_name] = get_representer(
-                        file_key_abs_path
-                    ).represent(file_key_abs_path)
+                    representer = get_representer(file_key_abs_path)
+                    representation, used_params = representer.represent(
+                        file_key_abs_path, **(representer_options_for_build or {})
+                    )
+                    output_representation[agent_facing_name] = {
+                        "representation": representation,
+                        "used_params": used_params,
+                    }
                 except Exception as e:
                     print(
                         f"Warning: Could not get representation for {agent_facing_name} (path {file_key_abs_path}): {e}"
@@ -213,10 +259,11 @@ class TransformationAsyncCodeBuilder(AsyncCodeBuilder):
                 "output_schema": output_schema_text,
                 "output_sample": output_sample_text
                 if not SCHEMA_ONLY
-                else "Sample not available. Transform according to the schema only. No data is present in the output example.",
+                else "Sample not available.",
                 "output_representation": str(
                     output_representation
                 ),  # Representation keyed by agent-facing name
+                "instructions": instructions,
             },
         )
         agent = Agent(