Compare commits

..

No commits in common. "588766ad0aa55e5573c878d88bbbd5fedc037665" and "3927f8e6c09687373f2d2653d07fef19f0e3aaff" have entirely different histories.

33 changed files with 318 additions and 767 deletions

View File

@ -1,45 +0,0 @@
{
"mcpServers": {
"conport": {
"command": "C:\\Users\\theis\\context-portal\\.venv\\Scripts\\python.exe",
"args": [
"C:\\Users\\theis\\context-portal\\src\\context_portal_mcp\\main.py",
"--mode",
"stdio",
"--workspace_id",
"${workspaceFolder}"
],
"alwaysAllow": [
"get_product_context",
"update_product_context",
"get_active_context",
"update_active_context",
"log_decision",
"get_decisions",
"search_decisions_fts",
"log_progress",
"get_progress",
"update_progress",
"delete_progress_by_id",
"log_system_pattern",
"get_system_patterns",
"log_custom_data",
"get_custom_data",
"delete_custom_data",
"search_project_glossary_fts",
"export_conport_to_markdown",
"import_markdown_to_conport",
"link_conport_items",
"search_custom_data_value_fts",
"get_linked_items",
"batch_log_items",
"get_item_history",
"delete_decision_by_id",
"delete_system_pattern_by_id",
"get_conport_schema",
"get_recent_activity_summary",
"semantic_search_conport"
]
}
}
}

View File

@ -1,3 +0,0 @@
{
"customModes": []
}

View File

@ -16,7 +16,6 @@ This document outlines the key features of the Asset Processor Tool.
* Saves maps in appropriate formats (JPG, PNG, EXR) based on complex rules involving map type (`FORCE_LOSSLESS_MAP_TYPES`), resolution (`RESOLUTION_THRESHOLD_FOR_JPG`), bit depth, and source format.
* Calculates basic image statistics (Min/Max/Mean) for a reference resolution.
* Calculates and stores the relative aspect ratio change string in metadata (e.g., `EVEN`, `X150`, `Y125`).
* **Low-Resolution Fallback:** If enabled (`ENABLE_LOW_RESOLUTION_FALLBACK`), automatically saves an additional "LOWRES" variant of source images if their largest dimension is below a configurable threshold (`LOW_RESOLUTION_THRESHOLD`). This "LOWRES" variant uses the original image dimensions and is saved in addition to any standard resolution outputs.
* **Channel Merging:** Combines channels from different maps into packed textures (e.g., NRMRGH) based on preset rules (`MAP_MERGE_RULES` in `config.py`).
* **Metadata Generation:** Creates a `metadata.json` file for each asset containing details about maps, category, archetype, aspect ratio change, processing settings, etc.
* **Output Organization:** Creates a clean, structured output directory (`<output_base>/<supplier>/<asset_name>/`).

View File

@ -13,18 +13,6 @@ The `app_settings.json` file is structured into several key sections, including:
* `ASSET_TYPE_DEFINITIONS`: Defines known asset types (like Surface, Model, Decal) and their properties.
* `MAP_MERGE_RULES`: Defines how multiple input maps can be merged into a single output map (e.g., combining Normal and Roughness into one).
### Low-Resolution Fallback Settings
These settings control the generation of low-resolution "fallback" variants for source images:
* `ENABLE_LOW_RESOLUTION_FALLBACK` (boolean, default: `true`):
* If `true`, the tool will generate an additional "LOWRES" variant for source images whose largest dimension is smaller than the `LOW_RESOLUTION_THRESHOLD`.
* This "LOWRES" variant uses the original dimensions of the source image and is saved in addition to any other standard resolution outputs (e.g., 1K, PREVIEW).
* If `false`, this feature is disabled.
* `LOW_RESOLUTION_THRESHOLD` (integer, default: `512`):
* Defines the pixel dimension (for the largest side of an image) below which the "LOWRES" fallback variant will be generated (if enabled).
* For example, if set to `512`, any source image smaller than 512x512 (e.g., 256x512, 128x128) will have a "LOWRES" variant created.
### LLM Predictor Settings
For users who wish to utilize the experimental LLM Predictor feature, the following settings are available in `config/llm_settings.json`:

View File

@ -58,7 +58,6 @@ The `<output_base_directory>` (the root folder where processing output starts) i
Each asset directory contains the following:
* Processed texture maps (e.g., `WoodFloor_Albedo_4k.png`, `MetalPanel_Normal_2k.exr`). The exact filenames depend on the `OUTPUT_FILENAME_PATTERN`. These are the resized, format-converted, and bit-depth adjusted texture files.
* **LOWRES Variants:** If the "Low-Resolution Fallback" feature is enabled and a source image's dimensions are below the configured threshold, an additional variant with "LOWRES" as its resolution token (e.g., `MyTexture_COL_LOWRES.png`) will be saved. This variant uses the original dimensions of the source image.
* Merged texture maps (e.g., `WoodFloor_Combined_4k.png`). The exact filenames depend on the `OUTPUT_FILENAME_PATTERN`. These are maps created by combining channels from different source maps based on the configured merge rules.
* Model files (if present in the source asset).
* `metadata.json`: A JSON file containing detailed information about the asset and the processing that was performed. This includes details about the maps (resolutions, formats, bit depths, and for roughness maps, a `derived_from_gloss_filename: true` flag if it was inverted from an original gloss map), merged map details, calculated image statistics, aspect ratio change information, asset category and archetype, the source preset used, and a list of ignored source files. This file is intended for use by downstream tools or scripts (like the Blender integration scripts).

View File

@ -12,9 +12,6 @@ The tool's configuration is loaded from several JSON files, providing a layered
1. **Application Settings (`config/app_settings.json`):** This JSON file defines the core global default settings, constants, and rules that apply generally across different asset sources (e.g., the global `OUTPUT_DIRECTORY_PATTERN` and `OUTPUT_FILENAME_PATTERN`, standard image resolutions, map merge rules, output format rules, Blender paths, temporary directory prefix, initial scaling mode, merge dimension mismatch strategy). See the [User Guide: Output Structure](../01_User_Guide/09_Output_Structure.md#available-tokens) for a list of available tokens for these patterns.
* *Note:* `ASSET_TYPE_DEFINITIONS` and `FILE_TYPE_DEFINITIONS` are no longer stored here; they have been moved to dedicated files.
* It also includes settings for new features like the "Low-Resolution Fallback":
* `ENABLE_LOW_RESOLUTION_FALLBACK` (boolean): Enables or disables the generation of "LOWRES" variants for small source images. Defaults to `true`.
* `LOW_RESOLUTION_THRESHOLD` (integer): The pixel dimension threshold (largest side) below which a "LOWRES" variant is created if the feature is enabled. Defaults to `512`.
2. **User Settings (`config/user_settings.json`):** This optional JSON file allows users to override specific settings defined in `config/app_settings.json`. If this file exists, its values for corresponding keys will take precedence over the base application settings. This file is primarily managed through the GUI's Application Preferences Editor.

View File

@ -50,44 +50,27 @@ These stages are executed sequentially once for each asset before the core item
### Core Item Processing Loop
The [`PipelineOrchestrator`](processing/pipeline/orchestrator.py:36) iterates through the `context.processing_items` list (populated by the [`PrepareProcessingItemsStage`](processing/pipeline/stages/prepare_processing_items.py:10)). Each `item` in this list is now either a [`ProcessingItem`](rule_structure.py:0) (representing a specific variant of a source map, e.g., Color at 1K, or Color at LOWRES) or a [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16).
The [`PipelineOrchestrator`](processing/pipeline/orchestrator.py:36) iterates through the `context.processing_items` list (populated by the [`PrepareProcessingItemsStage`](processing/pipeline/stages/prepare_processing_items.py:10)). For each item (either a [`FileRule`](rule_structure.py:5) for a regular map or a [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16) for a merged map), the following stages are executed sequentially:
1. **[`PrepareProcessingItemsStage`](processing/pipeline/stages/prepare_processing_items.py:10)** (`processing/pipeline/stages/prepare_processing_items.py`):
* **Responsibility**: (Executed once before the loop) This stage is now responsible for "exploding" each relevant [`FileRule`](rule_structure.py:5) into one or more [`ProcessingItem`](rule_structure.py:0) objects.
* For each [`FileRule`](rule_structure.py:5) that represents an image map:
* It loads the source image data and determines its original dimensions and bit depth.
* It creates standard [`ProcessingItem`](rule_structure.py:0)s for each required output resolution (e.g., "1K", "PREVIEW"), populating them with a copy of the source image data and the respective `resolution_key`.
* If the "Low-Resolution Fallback" feature is enabled (`ENABLE_LOW_RESOLUTION_FALLBACK` in config) and the source image's largest dimension is below `LOW_RESOLUTION_THRESHOLD`, it creates an additional [`ProcessingItem`](rule_structure.py:0) with `resolution_key="LOWRES"`, using the original image data and dimensions.
* It also adds [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16)s derived from global `map_merge_rules`.
* **Context Interaction**: Reads `context.files_to_process` and `context.config_obj`. Populates `context.processing_items` with a list of [`ProcessingItem`](rule_structure.py:0) and [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16) objects. Initializes `context.intermediate_results`.
* **Responsibility**: (Executed once before the loop) Creates the `context.processing_items` list by combining [`FileRule`](rule_structure.py:5)s from `context.files_to_process` and [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16)s derived from the global `map_merge_rules` configuration. It correctly accesses `map_merge_rules` from `context.config_obj` and validates each merge rule for the presence of `output_map_type` and a dictionary for `inputs`. Initializes `context.intermediate_results`.
* **Context Interaction**: Reads from `context.files_to_process` and `context.config_obj` (accessing `map_merge_rules`). Populates `context.processing_items` and initializes `context.intermediate_results`.
For each `item` in `context.processing_items`:
2. **Transformations (Implicit or via a dedicated stage - formerly `RegularMapProcessorStage` logic):**
* **Responsibility**: If the `item` is a [`ProcessingItem`](rule_structure.py:0), its `image_data` (loaded by `PrepareProcessingItemsStage`) may need transformations (Gloss-to-Rough, Normal Green Invert). This logic, previously in `RegularMapProcessorStage`, might be integrated into `PrepareProcessingItemsStage` before `ProcessingItem` creation, or handled by a new dedicated transformation stage that operates on `ProcessingItem.image_data`. The `item.map_type_identifier` would be updated if a transformation like Gloss-to-Rough occurs.
* **Context Interaction**: Modifies `item.image_data` and `item.map_type_identifier` within the [`ProcessingItem`](rule_structure.py:0) object.
2. **[`RegularMapProcessorStage`](processing/pipeline/stages/regular_map_processor.py:18)** (`processing/pipeline/stages/regular_map_processor.py`):
* **Responsibility**: (Executed per [`FileRule`](rule_structure.py:5) item) Checks if the `FileRule.item_type` starts with "MAP_". If not, the item is skipped. Otherwise, it loads the image data for the file, determines its potentially suffixed internal map type (e.g., "MAP_COL-1"), applies in-memory transformations (Gloss-to-Rough, Normal Green Invert) using the shared utility function [`apply_common_map_transformations`](processing/utils/image_processing_utils.py), and returns the processed image data and details in a [`ProcessedRegularMapData`](processing/pipeline/asset_context.py:23) object. The `internal_map_type` in the output reflects any transformations (e.g., "MAP_GLOSS" becomes "MAP_ROUGH").
* **Context Interaction**: Reads from the input [`FileRule`](rule_structure.py:5) (checking `item_type`) and [`Configuration`](configuration.py:68). Returns a [`ProcessedRegularMapData`](processing/pipeline/asset_context.py:23) object which is stored in `context.intermediate_results`.
3. **[`MergedTaskProcessorStage`](processing/pipeline/stages/merged_task_processor.py:68)** (`processing/pipeline/stages/merged_task_processor.py`):
* **Responsibility**: (Executed if `item` is a [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16)) Same as before: validates inputs, loads source map data (likely from `ProcessingItem`s in `context.processing_items` or a cache populated from them), applies transformations, merges channels, and returns [`ProcessedMergedMapData`](processing/pipeline/asset_context.py:35).
* **Context Interaction**: Reads [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16), potentially `context.processing_items` (or a cache derived from it) for input image data. Returns [`ProcessedMergedMapData`](processing/pipeline/asset_context.py:35).
* **Responsibility**: (Executed per [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16) item) Validates that all input map types specified in the merge rule start with "MAP_". If not, the task is failed. It dynamically loads input images by looking up the required input map types (e.g., "MAP_NRM") in `context.processed_maps_details` and using the temporary file paths from their `saved_files_info`. It applies in-memory transformations to inputs using [`apply_common_map_transformations`](processing/utils/image_processing_utils.py), handles dimension mismatches (with fallback creation if configured and `source_dimensions` are available), performs the channel merging operation, and returns the merged image data and details in a [`ProcessedMergedMapData`](processing/pipeline/asset_context.py:35) object. The `output_map_type` of the merged map must also be "MAP_" prefixed in the configuration.
* **Context Interaction**: Reads from the input [`MergeTaskDefinition`](processing/pipeline/asset_context.py:16) (checking input map types), `context.workspace_path`, `context.processed_maps_details` (for input image data), and [`Configuration`](configuration.py:68). Returns a [`ProcessedMergedMapData`](processing/pipeline/asset_context.py:35) object which is stored in `context.intermediate_results`.
4. **[`InitialScalingStage`](processing/pipeline/stages/initial_scaling.py:14)** (`processing/pipeline/stages/initial_scaling.py`):
* **Responsibility**: (Executed per item)
* If `item` is a [`ProcessingItem`](rule_structure.py:0): Takes `item.image_data`, `item.current_dimensions`, and `item.resolution_key` as input. If `item.resolution_key` is "LOWRES", POT scaling is skipped. Otherwise, applies POT scaling if configured.
* If `item` is from a `MergeTaskDefinition` (i.e., `processed_data` from `MergedTaskProcessorStage`): Applies POT scaling as before.
* **Context Interaction**: Takes [`InitialScalingInput`](processing/pipeline/asset_context.py:46) (now including `resolution_key`). Returns [`InitialScalingOutput`](processing/pipeline/asset_context.py:54) (also including `resolution_key`), which updates `context.intermediate_results`. The `current_image_data` and `current_dimensions` for saving are taken from this output.
* **Responsibility**: (Executed per item) Applies initial scaling (e.g., Power-of-Two downscaling) to the image data from the previous processing stage based on the `initial_scaling_mode` configuration.
* **Context Interaction**: Takes a [`InitialScalingInput`](processing/pipeline/asset_context.py:46) (containing image data and config) and returns an [`InitialScalingOutput`](processing/pipeline/asset_context.py:54) object, which updates the item's entry in `context.intermediate_results`.
5. **[`SaveVariantsStage`](processing/pipeline/stages/save_variants.py:15)** (`processing/pipeline/stages/save_variants.py`):
* **Responsibility**: (Executed per item) Saves the (potentially scaled) `current_image_data`.
* **Context Interaction**:
* Takes [`SaveVariantsInput`](processing/pipeline/asset_context.py:61).
* `internal_map_type` is set from `item.map_type_identifier` (for `ProcessingItem`) or `processed_data.output_map_type` (for merged).
* `output_filename_pattern_tokens['resolution']` is set to the `resolution_key` obtained from `scaled_data_output.resolution_key` (which originates from `item.resolution_key` for `ProcessingItem`s, or is `None` for merged items that get all standard resolutions).
* `image_resolutions` argument for `SaveVariantsInput`:
* If `resolution_key == "LOWRES"`: Set to `{"LOWRES": width_of_lowres_data}`.
* If `resolution_key` is a standard key (e.g., "1K"): Set to `{resolution_key: configured_dimension}`.
* For merged items (where `resolution_key` from scaling is likely `None`): Set to the full `config.image_resolutions` map to generate all applicable standard sizes.
* Returns [`SaveVariantsOutput`](processing/pipeline/asset_context.py:79). Orchestrator stores details in `context.processed_maps_details`.
* **Responsibility**: (Executed per item) Takes the final processed image data (potentially scaled) and configuration, and calls a utility to save the image to temporary files in various resolutions and formats as defined by the configuration.
* **Context Interaction**: Takes a [`SaveVariantsInput`](processing/pipeline/asset_context.py:61) object (which includes the "MAP_" prefixed `internal_map_type`). It uses the `get_filename_friendly_map_type` utility to convert this to a "standard type" (e.g., "COL") for output naming. Returns a [`SaveVariantsOutput`](processing/pipeline/asset_context.py:79) object containing details about the saved temporary files. The orchestrator stores these details, including the original "MAP_" prefixed `internal_map_type`, in `context.processed_maps_details` for the item.
### Post-Item Stages

View File

@ -26,7 +26,7 @@
},
{
"file_path": "BoucleChunky001_DISP_1K_METALNESS.png",
"item_type": "EXTRA",
"item_type": "MAP_DISP",
"target_asset_name_override": "BoucleChunky001"
},
{

View File

@ -96,8 +96,6 @@ class InfoSummaryFilter(logging.Filter):
"verify: processingengine.process called",
": effective supplier set to",
": metadata initialized.",
"path",
"\\asset_processor",
": file rules queued for processing",
"successfully loaded base application settings",
"successfully loaded and merged asset_type_definitions",
@ -110,6 +108,12 @@ class InfoSummaryFilter(logging.Filter):
"worker thread: finished processing for rule:",
"task finished signal received for",
# Autotest step markers (not global summaries)
"step 1: loading zip file:",
"step 2: selecting preset:",
"step 4: retrieving and comparing rules...",
"step 5: starting processing...",
"step 7: checking output path:",
"output path check completed.",
]
def filter(self, record):
@ -584,59 +588,63 @@ class AutoTester(QObject):
logger.error(f"Value mismatch for field '{key}' in {item_type_name} ({current_context}): Actual='{actual_value}', Expected='{expected_value}'.")
item_match = False
return item_match
return item_match
def _compare_list_of_rules(self, actual_list: List[Dict[str, Any]], expected_list: List[Dict[str, Any]], item_type_name: str, parent_context: str, item_key_field: str) -> bool:
"""
Compares a list of actual rule items against a list of expected rule items.
Items are matched by a key field (e.g., 'asset_name' or 'file_path').
Order independent for matching, but logs count mismatches.
"""
list_match = True
if not isinstance(actual_list, list) or not isinstance(expected_list, list):
logger.error(f"Type mismatch for list of {item_type_name}s in {parent_context}. Expected lists.")
return False
if len(actual_list) != len(expected_list):
logger.error(f"Mismatch in number of {item_type_name}s for {parent_context}. Actual: {len(actual_list)}, Expected: {len(expected_list)}.")
list_match = False # Count mismatch is an error
# If counts differ, we still try to match what we can to provide more detailed feedback,
# but the overall list_match will remain False.
actual_items_map = {item.get(item_key_field): item for item in actual_list if item.get(item_key_field) is not None}
# Keep track of expected items that found a match to identify missing ones more easily
matched_expected_keys = set()
for expected_item in expected_list:
expected_key_value = expected_item.get(item_key_field)
if expected_key_value is None:
logger.error(f"Expected {item_type_name} in {parent_context} is missing key field '{item_key_field}'. Cannot compare this item: {expected_item}")
list_match = False # This specific expected item cannot be processed
continue
actual_item = actual_items_map.get(expected_key_value)
if actual_item:
matched_expected_keys.add(expected_key_value)
if not self._compare_rule_item(actual_item, expected_item, item_type_name, parent_context):
list_match = False # Individual item comparison failed
else:
logger.error(f"Expected {item_type_name} with {item_key_field} '{expected_key_value}' not found in actual items for {parent_context}.")
list_match = False
# Identify actual items that were not matched by any expected item
# This is useful if len(actual_list) >= len(expected_list) but some actual items are "extra"
for actual_key_value, actual_item_data in actual_items_map.items():
if actual_key_value not in matched_expected_keys:
logger.debug(f"Extra actual {item_type_name} with {item_key_field} '{actual_key_value}' found in {parent_context} (not in expected list or already matched).")
if len(actual_list) != len(expected_list): # If counts already flagged a mismatch, this is just detail
pass
else: # Counts matched, but content didn't align perfectly by key
def _compare_list_of_rules(self, actual_list: List[Dict[str, Any]], expected_list: List[Dict[str, Any]], item_type_name: str, parent_context: str, item_key_field: str) -> bool:
"""
Compares a list of actual rule items against a list of expected rule items.
Items are matched by a key field (e.g., 'asset_name' or 'file_path').
Order independent for matching, but logs count mismatches.
"""
list_match = True # Corrected indentation
if not isinstance(actual_list, list) or not isinstance(expected_list, list):
logger.error(f"Type mismatch for list of {item_type_name}s in {parent_context}. Expected lists.")
return False
if len(actual_list) != len(expected_list):
logger.error(f"Mismatch in number of {item_type_name}s for {parent_context}. Actual: {len(actual_list)}, Expected: {len(expected_list)}.")
list_match = False # Count mismatch is an error
# If counts differ, we still try to match what we can to provide more detailed feedback,
# but the overall list_match will remain False.
actual_items_map = {item.get(item_key_field): item for item in actual_list if item.get(item_key_field) is not None}
# Keep track of expected items that found a match to identify missing ones more easily
matched_expected_keys = set()
for expected_item in expected_list:
expected_key_value = expected_item.get(item_key_field)
if expected_key_value is None:
logger.error(f"Expected {item_type_name} in {parent_context} is missing key field '{item_key_field}'. Cannot compare this item: {expected_item}")
list_match = False # This specific expected item cannot be processed
continue
actual_item = actual_items_map.get(expected_key_value)
if actual_item:
matched_expected_keys.add(expected_key_value)
if not self._compare_rule_item(actual_item, expected_item, item_type_name, parent_context):
list_match = False # Individual item comparison failed
else:
logger.error(f"Expected {item_type_name} with {item_key_field} '{expected_key_value}' not found in actual items for {parent_context}.")
list_match = False
return list_match
# Identify actual items that were not matched by any expected item
# This is useful if len(actual_list) >= len(expected_list) but some actual items are "extra"
for actual_key_value, actual_item_data in actual_items_map.items():
if actual_key_value not in matched_expected_keys:
logger.debug(f"Extra actual {item_type_name} with {item_key_field} '{actual_key_value}' found in {parent_context} (not in expected list or already matched).")
if len(actual_list) != len(expected_list): # If counts already flagged a mismatch, this is just detail
pass
else: # Counts matched, but content didn't align perfectly by key
list_match = False
return list_match # Corrected indentation
def _compare_rules(self, actual_rules_data: Dict[str, Any], expected_rules_data: Dict[str, Any]) -> bool: # Corrected structure: moved out
item_match = False
return item_match
def _compare_rules(self, actual_rules_data: Dict[str, Any], expected_rules_data: Dict[str, Any]) -> bool:
"""

View File

@ -46,10 +46,7 @@
"TEMP_DIR_PREFIX": "_PROCESS_ASSET_",
"INITIAL_SCALING_MODE": "POT_DOWNSCALE",
"MERGE_DIMENSION_MISMATCH_STRATEGY": "USE_LARGEST",
"ENABLE_LOW_RESOLUTION_FALLBACK": true,
"LOW_RESOLUTION_THRESHOLD": 512,
"general_settings": {
"invert_normal_map_green_channel_globally": false,
"app_version": "Pre-Alpha"
"invert_normal_map_green_channel_globally": false
}
}

View File

@ -190,7 +190,7 @@
],
"is_grayscale": false,
"keybind": "E",
"standard_type": "EXTRA"
"standard_type": ""
},
"FILE_IGNORE": {
"bit_depth_rule": "",

View File

@ -4,7 +4,6 @@ from pathlib import Path
import logging
import re
import collections.abc
from typing import Optional
log = logging.getLogger(__name__)
@ -13,7 +12,7 @@ APP_SETTINGS_PATH = BASE_DIR / "config" / "app_settings.json"
LLM_SETTINGS_PATH = BASE_DIR / "config" / "llm_settings.json"
ASSET_TYPE_DEFINITIONS_PATH = BASE_DIR / "config" / "asset_type_definitions.json"
FILE_TYPE_DEFINITIONS_PATH = BASE_DIR / "config" / "file_type_definitions.json"
USER_SETTINGS_PATH = BASE_DIR / "config" / "user_settings.json"
USER_SETTINGS_PATH = BASE_DIR / "config" / "user_settings.json" # New path for user settings
SUPPLIERS_CONFIG_PATH = BASE_DIR / "config" / "suppliers.json"
PRESETS_DIR = BASE_DIR / "Presets"
@ -650,24 +649,6 @@ class Configuration:
"""Returns the LLM request timeout in seconds from LLM settings."""
return self._llm_settings.get('llm_request_timeout', 120)
@property
def app_version(self) -> Optional[str]:
"""Returns the application version from general_settings."""
gs = self._core_settings.get('general_settings')
if isinstance(gs, dict):
return gs.get('app_version')
return None
@property
def enable_low_resolution_fallback(self) -> bool:
"""Gets the setting for enabling low-resolution fallback."""
return self._core_settings.get('ENABLE_LOW_RESOLUTION_FALLBACK', True)
@property
def low_resolution_threshold(self) -> int:
"""Gets the pixel dimension threshold for low-resolution fallback."""
return self._core_settings.get('LOW_RESOLUTION_THRESHOLD', 512)
@property
def FILE_TYPE_DEFINITIONS(self) -> dict:
return self._file_type_definitions

Binary file not shown.

View File

@ -126,15 +126,12 @@ class SupplierSearchDelegate(QStyledItemDelegate):
"""Loads the list of known suppliers from the JSON config file."""
try:
with open(SUPPLIERS_CONFIG_PATH, 'r') as f:
suppliers_data = json.load(f) # Renamed variable for clarity
if isinstance(suppliers_data, list):
suppliers = json.load(f)
if isinstance(suppliers, list):
# Ensure all items are strings
return sorted([str(s) for s in suppliers_data if isinstance(s, str)])
elif isinstance(suppliers_data, dict): # ADDED: Handle dictionary case
# If it's a dictionary, extract keys as supplier names
return sorted([str(key) for key in suppliers_data.keys() if isinstance(key, str)])
else: # MODIFIED: Updated warning message
log.warning(f"'{SUPPLIERS_CONFIG_PATH}' does not contain a valid list or dictionary of suppliers. Starting fresh.")
return sorted([str(s) for s in suppliers if isinstance(s, str)])
else:
log.warning(f"'{SUPPLIERS_CONFIG_PATH}' does not contain a valid list. Starting fresh.")
return []
except FileNotFoundError:
log.info(f"'{SUPPLIERS_CONFIG_PATH}' not found. Starting with an empty supplier list.")

View File

@ -20,8 +20,7 @@ script_dir = Path(__file__).parent
project_root = script_dir.parent
PRESETS_DIR = project_root / "Presets"
TEMPLATE_PATH = PRESETS_DIR / "_template.json"
APP_SETTINGS_PATH_LOCAL = project_root / "config" / "app_settings.json" # Retain for other settings if used elsewhere
FILE_TYPE_DEFINITIONS_PATH = project_root / "config" / "file_type_definitions.json"
APP_SETTINGS_PATH_LOCAL = project_root / "config" / "app_settings.json"
log = logging.getLogger(__name__)
@ -64,19 +63,18 @@ class PresetEditorWidget(QWidget):
"""Loads FILE_TYPE_DEFINITIONS keys from app_settings.json."""
keys = []
try:
if FILE_TYPE_DEFINITIONS_PATH.is_file():
with open(FILE_TYPE_DEFINITIONS_PATH, 'r', encoding='utf-8') as f:
if APP_SETTINGS_PATH_LOCAL.is_file():
with open(APP_SETTINGS_PATH_LOCAL, 'r', encoding='utf-8') as f:
settings = json.load(f)
# The FILE_TYPE_DEFINITIONS key is at the root of file_type_definitions.json
ftd = settings.get("FILE_TYPE_DEFINITIONS", {})
keys = list(ftd.keys())
log.debug(f"Successfully loaded {len(keys)} FILE_TYPE_DEFINITIONS keys from {FILE_TYPE_DEFINITIONS_PATH}.")
log.debug(f"Successfully loaded {len(keys)} FILE_TYPE_DEFINITIONS keys.")
else:
log.error(f"file_type_definitions.json not found at {FILE_TYPE_DEFINITIONS_PATH} for PresetEditorWidget.")
log.error(f"app_settings.json not found at {APP_SETTINGS_PATH_LOCAL} for PresetEditorWidget.")
except json.JSONDecodeError as e:
log.error(f"Failed to parse file_type_definitions.json in PresetEditorWidget: {e}")
log.error(f"Failed to parse app_settings.json in PresetEditorWidget: {e}")
except Exception as e:
log.error(f"Error loading FILE_TYPE_DEFINITIONS keys from {FILE_TYPE_DEFINITIONS_PATH} in PresetEditorWidget: {e}")
log.error(f"Error loading FILE_TYPE_DEFINITIONS keys in PresetEditorWidget: {e}")
return keys
def _init_ui(self):

View File

@ -552,13 +552,6 @@ class UnifiedViewModel(QAbstractItemModel):
supplier_col_index = self.createIndex(existing_source_row, self.COL_SUPPLIER, existing_source_rule)
self.dataChanged.emit(supplier_col_index, supplier_col_index, [Qt.DisplayRole, Qt.EditRole])
# Always update the preset_name from the new_source_rule, as this reflects the latest prediction context
if existing_source_rule.preset_name != new_source_rule.preset_name:
log.debug(f" Updating preset_name for SourceRule '{source_path}' from '{existing_source_rule.preset_name}' to '{new_source_rule.preset_name}'")
existing_source_rule.preset_name = new_source_rule.preset_name
# Note: preset_name is not directly displayed in the view, so no dataChanged needed for a specific column,
# but if it influenced other display elements, dataChanged would be emitted for those.
# --- Merge AssetRules ---
existing_assets_dict = {asset.asset_name: asset for asset in existing_source_rule.assets}

12
main.py
View File

@ -4,7 +4,6 @@ import time
import os
import logging
from pathlib import Path
import re # Added for checking incrementing token
from concurrent.futures import ProcessPoolExecutor, as_completed
import subprocess
import shutil
@ -239,14 +238,9 @@ class ProcessingTask(QRunnable):
# output_dir should already be a Path object
pattern = getattr(config, 'output_directory_pattern', None)
if pattern:
# Only call get_next_incrementing_value if the pattern contains an incrementing token
if re.search(r"\[IncrementingValue\]|#+", pattern):
log.debug(f"Incrementing token found in pattern '{pattern}'. Calculating next value for dir: {output_dir}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"Calculated next incrementing value for {output_dir}: {next_increment_str}")
else:
log.debug(f"No incrementing token found in pattern '{pattern}'. Skipping increment calculation.")
next_increment_str = None # Or a default like "00" if downstream expects a string, but None is cleaner if handled.
log.debug(f"Calculating next incrementing value for dir: {output_dir} using pattern: {pattern}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"Calculated next incrementing value for {output_dir}: {next_increment_str}")
else:
log.warning(f"Cannot calculate incrementing value: 'output_directory_pattern' not found in configuration for preset {config.preset_name}")
except Exception as e:

View File

@ -195,25 +195,17 @@ def _process_archive_task(archive_path: Path, output_dir: Path, processed_dir: P
# Assuming config object has 'output_directory_pattern' attribute/key
pattern = getattr(config, 'output_directory_pattern', None) # Use getattr for safety
if pattern:
if re.search(r"\[IncrementingValue\]|#+", pattern):
log.debug(f"[Task:{archive_path.name}] Incrementing token found in pattern '{pattern}'. Calculating next value for dir: {output_dir}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"[Task:{archive_path.name}] Calculated next incrementing value: {next_increment_str}")
else:
log.debug(f"[Task:{archive_path.name}] No incrementing token found in pattern '{pattern}'. Skipping increment calculation.")
next_increment_str = None
log.debug(f"[Task:{archive_path.name}] Calculating next incrementing value for dir: {output_dir} using pattern: {pattern}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"[Task:{archive_path.name}] Calculated next incrementing value: {next_increment_str}")
else:
# Check if config is a dict as fallback (depends on load_config implementation)
if isinstance(config, dict):
pattern = config.get('output_directory_pattern')
if pattern:
if re.search(r"\[IncrementingValue\]|#+", pattern):
log.debug(f"[Task:{archive_path.name}] Incrementing token found in pattern '{pattern}' (from dict). Calculating next value for dir: {output_dir}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"[Task:{archive_path.name}] Calculated next incrementing value (from dict): {next_increment_str}")
else:
log.debug(f"[Task:{archive_path.name}] No incrementing token found in pattern '{pattern}' (from dict). Skipping increment calculation.")
next_increment_str = None
log.debug(f"[Task:{archive_path.name}] Calculating next incrementing value for dir: {output_dir} using pattern (from dict): {pattern}")
next_increment_str = get_next_incrementing_value(output_dir, pattern)
log.info(f"[Task:{archive_path.name}] Calculated next incrementing value (from dict): {next_increment_str}")
else:
log.warning(f"[Task:{archive_path.name}] Cannot calculate incrementing value: 'output_directory_pattern' not found in configuration dictionary.")
else:

View File

@ -1,4 +1,3 @@
import dataclasses # Added import
from dataclasses import dataclass
from pathlib import Path
from typing import Dict, List, Optional
@ -28,7 +27,6 @@ class ProcessedRegularMapData:
original_bit_depth: Optional[int]
original_dimensions: Optional[Tuple[int, int]] # (width, height)
transformations_applied: List[str]
resolution_key: Optional[str] = None # Added field
status: str = "Processed"
error_message: Optional[str] = None
@ -47,10 +45,9 @@ class ProcessedMergedMapData:
@dataclass
class InitialScalingInput:
image_data: np.ndarray
initial_scaling_mode: str # Moved before fields with defaults
original_dimensions: Optional[Tuple[int, int]] # (width, height)
resolution_key: Optional[str] = None # Added field
# Configuration needed
initial_scaling_mode: str
# Output for InitialScalingStage
@dataclass
@ -58,7 +55,6 @@ class InitialScalingOutput:
scaled_image_data: np.ndarray
scaling_applied: bool
final_dimensions: Tuple[int, int] # (width, height)
resolution_key: Optional[str] = None # Added field
# Input for SaveVariantsStage
@dataclass

View File

@ -8,7 +8,7 @@ from typing import List, Dict, Optional, Any, Union # Added Any, Union
import numpy as np # Added numpy
from configuration import Configuration
from rule_structure import SourceRule, AssetRule, FileRule, ProcessingItem # Added ProcessingItem
from rule_structure import SourceRule, AssetRule, FileRule # Added FileRule
# Import new context classes and stages
from .asset_context import (
@ -200,224 +200,145 @@ class PipelineOrchestrator:
current_image_data: Optional[np.ndarray] = None # Track current image data ref
try:
# The 'item' is now expected to be a ProcessingItem or MergeTaskDefinition
if isinstance(item, ProcessingItem):
item_key = f"{item.source_file_info_ref}_{item.map_type_identifier}_{item.resolution_key}"
item_log_prefix = f"Asset '{asset_name}', ProcItem '{item_key}'"
log.info(f"{item_log_prefix}: Starting processing.")
# Data for ProcessingItem is already loaded by PrepareProcessingItemsStage
current_image_data = item.image_data
current_dimensions = item.current_dimensions
item_resolution_key = item.resolution_key
# Transformations (like gloss to rough, normal invert) are assumed to be applied
# by RegularMapProcessorStage if it's still used, or directly in PrepareProcessingItemsStage
# before creating the ProcessingItem, or a new dedicated transformation stage.
# For now, assume item.image_data is ready for scaling/saving.
# Store initial ProcessingItem data as "processed_data" for consistency if RegularMapProcessor is bypassed
# This is a simplification; a dedicated transformation stage would be cleaner.
# For now, we assume transformations happened before or within PrepareProcessingItemsStage.
# The 'processed_data' variable here is more of a placeholder for what would feed into scaling.
# Create a simple ProcessedRegularMapData-like structure for logging/details if needed,
# or adapt the final_details population later.
# For now, we'll directly use 'item' fields.
# 2. Scale (Optional)
scaling_mode = getattr(context.config_obj, "INITIAL_SCALING_MODE", "NONE")
# Pass the item's resolution_key to InitialScalingInput
scale_input = InitialScalingInput(
image_data=current_image_data,
original_dimensions=current_dimensions,
initial_scaling_mode=scaling_mode,
resolution_key=item_resolution_key # Pass the key
)
# Add _source_file_path for logging within InitialScalingStage if available
setattr(scale_input, '_source_file_path', item.source_file_info_ref)
log.debug(f"{item_log_prefix}: Calling InitialScalingStage. Input res_key: {scale_input.resolution_key}")
scaled_data_output = self._scaling_stage.execute(scale_input)
current_image_data = scaled_data_output.scaled_image_data
current_dimensions = scaled_data_output.final_dimensions # Dimensions after scaling
# The resolution_key from item is passed through by InitialScalingOutput
output_resolution_key = scaled_data_output.resolution_key
log.debug(f"{item_log_prefix}: InitialScalingStage output. Scaled: {scaled_data_output.scaling_applied}, New Dims: {current_dimensions}, Output ResKey: {output_resolution_key}")
context.intermediate_results[item_key] = scaled_data_output
# 3. Save Variants
if current_image_data is None or current_image_data.size == 0:
log.warning(f"{item_log_prefix}: Skipping save stage because image data is empty.")
context.processed_maps_details[item_key] = {"status": "Skipped", "notes": "No image data to save", "stage": "SaveVariantsStage"}
continue
log.debug(f"{item_log_prefix}: Preparing to save variant with resolution key '{output_resolution_key}'...")
output_filename_tokens = {
'asset_name': asset_name,
'output_base_directory': context.engine_temp_dir,
'supplier': context.effective_supplier or 'UnknownSupplier',
'resolution': output_resolution_key # Use the key from the item/scaling stage
}
# Determine image_resolutions argument for save_image_variants
save_specific_resolutions = {}
if output_resolution_key == "LOWRES":
# For LOWRES, the "resolution value" is its actual dimension.
# image_saving_utils needs a dict like {"LOWRES": 64} if current_dim is 64x64
# Assuming current_dimensions[0] is width.
save_specific_resolutions = {"LOWRES": current_dimensions[0] if current_dimensions else 0}
log.debug(f"{item_log_prefix}: Preparing to save LOWRES variant. Dimensions: {current_dimensions}. Save resolutions arg: {save_specific_resolutions}")
elif output_resolution_key in context.config_obj.image_resolutions:
save_specific_resolutions = {output_resolution_key: context.config_obj.image_resolutions[output_resolution_key]}
else:
log.warning(f"{item_log_prefix}: Resolution key '{output_resolution_key}' not found in config.image_resolutions and not LOWRES. Saving might fail or use full res.")
# Fallback: pass all configured resolutions, image_saving_utils will try to match by size.
# This might not be ideal if the key is truly unknown.
# Or, more strictly, fail here if key is unknown and not LOWRES.
# For now, let image_saving_utils handle it by passing all.
save_specific_resolutions = context.config_obj.image_resolutions
save_input = SaveVariantsInput(
image_data=current_image_data,
internal_map_type=item.map_type_identifier,
source_bit_depth_info=[item.bit_depth] if item.bit_depth is not None else [8], # Default to 8 if not set
output_filename_pattern_tokens=output_filename_tokens,
image_resolutions=save_specific_resolutions, # Pass the specific resolution(s)
file_type_defs=getattr(context.config_obj, "FILE_TYPE_DEFINITIONS", {}),
output_format_8bit=context.config_obj.get_8bit_output_format(),
output_format_16bit_primary=context.config_obj.get_16bit_output_formats()[0],
output_format_16bit_fallback=context.config_obj.get_16bit_output_formats()[1],
png_compression_level=context.config_obj.png_compression_level,
jpg_quality=context.config_obj.jpg_quality,
output_filename_pattern=context.config_obj.output_filename_pattern,
resolution_threshold_for_jpg=getattr(context.config_obj, "resolution_threshold_for_jpg", None)
)
saved_data = self._save_stage.execute(save_input)
if saved_data and saved_data.status.startswith("Processed"):
item_status = saved_data.status
log.info(f"{item_log_prefix}: Item successfully processed and saved. Status: {item_status}")
context.processed_maps_details[item_key] = {
"status": item_status,
"saved_files_info": saved_data.saved_files_details,
"internal_map_type": item.map_type_identifier,
"resolution_key": output_resolution_key,
"original_dimensions": item.original_dimensions,
"final_dimensions": current_dimensions, # Dimensions after scaling
"source_file": item.source_file_info_ref,
# 1. Process (Load/Merge + Transform)
if isinstance(item, FileRule):
if item.item_type == 'EXTRA':
log.debug(f"{item_log_prefix}: Skipping image processing for EXTRA FileRule '{item.file_path}'.")
# Add a basic entry to processed_maps_details to acknowledge it was seen
context.processed_maps_details[item.file_path] = {
"status": "Skipped (EXTRA file)",
"internal_map_type": "EXTRA",
"source_file": str(item.file_path)
}
else:
error_msg = saved_data.error_message if saved_data else "Save stage returned None"
log.error(f"{item_log_prefix}: Failed during save stage. Error: {error_msg}")
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Save Error: {error_msg}", "stage": "SaveVariantsStage"}
asset_had_item_errors = True
item_status = "Failed"
continue # Skip to the next item
item_key = item.file_path # Use file_path string as key
log.debug(f"{item_log_prefix}: Processing FileRule '{item.file_path}'...")
processed_data = self._regular_processor_stage.execute(context, item)
elif isinstance(item, MergeTaskDefinition):
# --- This part needs similar refactoring for resolution_key if merged outputs can be LOWRES ---
# --- For now, assume merged tasks always produce standard resolutions ---
item_key = item.task_key
item_log_prefix = f"Asset '{asset_name}', MergeTask '{item_key}'"
log.info(f"{item_log_prefix}: Processing MergeTask.")
# 1. Process Merge Task
item_key = item.task_key # Use task_key string as key
log.info(f"{item_log_prefix}: Executing MergedTaskProcessorStage for MergeTask '{item_key}'...") # Log call
processed_data = self._merged_processor_stage.execute(context, item)
if not processed_data or processed_data.status != "Processed":
error_msg = processed_data.error_message if processed_data else "Merge processor returned None"
log.error(f"{item_log_prefix}: Failed during merge processing. Error: {error_msg}")
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Merge Error: {error_msg}", "stage": "MergedTaskProcessorStage"}
asset_had_item_errors = True
continue
context.intermediate_results[item_key] = processed_data
current_image_data = processed_data.merged_image_data
current_dimensions = processed_data.final_dimensions
# Log status/error from merge processor
if processed_data:
log.info(f"{item_log_prefix}: MergedTaskProcessorStage result - Status: {processed_data.status}, Error: {processed_data.error_message}")
else:
log.warning(f"{item_log_prefix}: MergedTaskProcessorStage returned None for MergeTask '{item_key}'.")
else:
log.warning(f"{item_log_prefix}: Unknown item type '{type(item)}'. Skipping.")
item_key = f"unknown_item_{item_index}"
context.processed_maps_details[item_key] = {"status": "Skipped", "notes": f"Unknown item type {type(item)}"}
asset_had_item_errors = True
continue # Next item
# 2. Scale Merged Output (Optional)
# Merged tasks typically don't have a single "resolution_key" like LOWRES from source.
# They produce an image that then gets downscaled to 1K, PREVIEW etc.
# So, resolution_key for InitialScalingInput here would be None or a default.
scaling_mode = getattr(context.config_obj, "INITIAL_SCALING_MODE", "NONE")
# Check for processing failure
if not processed_data or processed_data.status != "Processed":
error_msg = processed_data.error_message if processed_data else "Processor returned None"
log.error(f"{item_log_prefix}: Failed during processing stage. Error: {error_msg}")
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Processing Error: {error_msg}", "stage": processed_data.__class__.__name__ if processed_data else "UnknownProcessor"}
asset_had_item_errors = True
continue # Next item
# Store intermediate result & get current image data
context.intermediate_results[item_key] = processed_data
current_image_data = processed_data.processed_image_data if isinstance(processed_data, ProcessedRegularMapData) else processed_data.merged_image_data
current_dimensions = processed_data.original_dimensions if isinstance(processed_data, ProcessedRegularMapData) else processed_data.final_dimensions
# 2. Scale (Optional)
scaling_mode = getattr(context.config_obj, "INITIAL_SCALING_MODE", "NONE")
if scaling_mode != "NONE" and current_image_data is not None and current_image_data.size > 0:
if isinstance(item, MergeTaskDefinition): # Log scaling call for merge tasks
log.info(f"{item_log_prefix}: Calling InitialScalingStage for MergeTask '{item_key}' (Mode: {scaling_mode})...")
log.debug(f"{item_log_prefix}: Applying initial scaling (Mode: {scaling_mode})...")
scale_input = InitialScalingInput(
image_data=current_image_data,
original_dimensions=current_dimensions,
initial_scaling_mode=scaling_mode,
resolution_key=None # Merged outputs are not "LOWRES" themselves before this scaling
original_dimensions=current_dimensions, # Pass original/merged dims
initial_scaling_mode=scaling_mode
)
setattr(scale_input, '_source_file_path', f"MergeTask_{item_key}") # For logging
log.debug(f"{item_log_prefix}: Calling InitialScalingStage for merged data.")
scaled_data_output = self._scaling_stage.execute(scale_input)
current_image_data = scaled_data_output.scaled_image_data
current_dimensions = scaled_data_output.final_dimensions
# Merged items don't have a specific output_resolution_key from source,
# they will be saved to all applicable resolutions from config.
# So scaled_data_output.resolution_key will be None here.
context.intermediate_results[item_key] = scaled_data_output
# 3. Save Merged Variants
if current_image_data is None or current_image_data.size == 0:
log.warning(f"{item_log_prefix}: Skipping save for merged task, image data is empty.")
context.processed_maps_details[item_key] = {"status": "Skipped", "notes": "No merged image data to save", "stage": "SaveVariantsStage"}
continue
output_filename_tokens = {
'asset_name': asset_name,
'output_base_directory': context.engine_temp_dir,
'supplier': context.effective_supplier or 'UnknownSupplier',
# 'resolution' token will be filled by image_saving_utils for each variant
}
# For merged tasks, we usually want to generate all standard resolutions.
# The `resolution_key` from the item itself is not applicable here for the `resolution` token.
# The `image_saving_utils.save_image_variants` will iterate through `context.config_obj.image_resolutions`.
save_input = SaveVariantsInput(
image_data=current_image_data,
internal_map_type=processed_data.output_map_type,
source_bit_depth_info=processed_data.source_bit_depths,
output_filename_pattern_tokens=output_filename_tokens,
image_resolutions=context.config_obj.image_resolutions, # Pass all configured resolutions
file_type_defs=getattr(context.config_obj, "FILE_TYPE_DEFINITIONS", {}),
output_format_8bit=context.config_obj.get_8bit_output_format(),
output_format_16bit_primary=context.config_obj.get_16bit_output_formats()[0],
output_format_16bit_fallback=context.config_obj.get_16bit_output_formats()[1],
png_compression_level=context.config_obj.png_compression_level,
jpg_quality=context.config_obj.jpg_quality,
output_filename_pattern=context.config_obj.output_filename_pattern,
resolution_threshold_for_jpg=getattr(context.config_obj, "resolution_threshold_for_jpg", None)
)
saved_data = self._save_stage.execute(save_input)
if saved_data and saved_data.status.startswith("Processed"):
item_status = saved_data.status
log.info(f"{item_log_prefix}: Merged task successfully processed and saved. Status: {item_status}")
context.processed_maps_details[item_key] = {
"status": item_status,
"saved_files_info": saved_data.saved_files_details,
"internal_map_type": processed_data.output_map_type,
"final_dimensions": current_dimensions,
}
else:
error_msg = saved_data.error_message if saved_data else "Save stage for merged task returned None"
log.error(f"{item_log_prefix}: Failed during save stage for merged task. Error: {error_msg}")
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Save Error (Merged): {error_msg}", "stage": "SaveVariantsStage"}
asset_had_item_errors = True
item_status = "Failed"
# Update intermediate result and current image data reference
context.intermediate_results[item_key] = scaled_data_output # Overwrite previous intermediate
current_image_data = scaled_data_output.scaled_image_data # Use scaled data for saving
log.debug(f"{item_log_prefix}: Scaling applied: {scaled_data_output.scaling_applied}. New Dims: {scaled_data_output.final_dimensions}")
else:
log.warning(f"{item_log_prefix}: Unknown item type in loop: {type(item)}. Skipping.")
# Ensure some key exists to prevent KeyError if item_key was not set
unknown_item_key = f"unknown_item_at_index_{item_index}"
context.processed_maps_details[unknown_item_key] = {"status": "Skipped", "notes": f"Unknown item type {type(item)}"}
log.debug(f"{item_log_prefix}: Initial scaling skipped (Mode: NONE or empty image).")
# Create dummy output if scaling skipped, using current dims
final_dims = current_dimensions if current_dimensions else (current_image_data.shape[1], current_image_data.shape[0]) if current_image_data is not None else (0,0)
scaled_data_output = InitialScalingOutput(scaled_image_data=current_image_data, scaling_applied=False, final_dimensions=final_dims)
# 3. Save Variants
if current_image_data is None or current_image_data.size == 0:
log.warning(f"{item_log_prefix}: Skipping save stage because image data is empty.")
context.processed_maps_details[item_key] = {"status": "Skipped", "notes": "No image data to save", "stage": "SaveVariantsStage"}
# Don't mark as asset error, just skip this item's saving
continue # Next item
if isinstance(item, MergeTaskDefinition): # Log save call for merge tasks
log.info(f"{item_log_prefix}: Calling SaveVariantsStage for MergeTask '{item_key}'...")
log.debug(f"{item_log_prefix}: Saving variants...")
# Prepare input for save stage
internal_map_type = processed_data.final_internal_map_type if isinstance(processed_data, ProcessedRegularMapData) else processed_data.output_map_type
source_bit_depth = [processed_data.original_bit_depth] if isinstance(processed_data, ProcessedRegularMapData) and processed_data.original_bit_depth is not None else processed_data.source_bit_depths if isinstance(processed_data, ProcessedMergedMapData) else [8] # Default bit depth if unknown
# Construct filename tokens (ensure temp dir is used)
output_filename_tokens = {
'asset_name': asset_name,
'output_base_directory': context.engine_temp_dir, # Save variants to temp dir
# Add other tokens from context/config as needed by the pattern
'supplier': context.effective_supplier or 'UnknownSupplier',
}
# Log the value being read for the threshold before creating the input object
log.info(f"ORCHESTRATOR_DEBUG: Reading RESOLUTION_THRESHOLD_FOR_JPG from config for SaveVariantsInput: {getattr(context.config_obj, 'RESOLUTION_THRESHOLD_FOR_JPG', None)}")
save_input = SaveVariantsInput(
image_data=current_image_data, # Use potentially scaled data
internal_map_type=internal_map_type,
source_bit_depth_info=source_bit_depth,
output_filename_pattern_tokens=output_filename_tokens,
# Pass config values needed by save stage
image_resolutions=context.config_obj.image_resolutions,
file_type_defs=getattr(context.config_obj, "FILE_TYPE_DEFINITIONS", {}),
output_format_8bit=context.config_obj.get_8bit_output_format(),
output_format_16bit_primary=context.config_obj.get_16bit_output_formats()[0],
output_format_16bit_fallback=context.config_obj.get_16bit_output_formats()[1],
png_compression_level=context.config_obj.png_compression_level,
jpg_quality=context.config_obj.jpg_quality,
output_filename_pattern=context.config_obj.output_filename_pattern,
resolution_threshold_for_jpg=getattr(context.config_obj, "resolution_threshold_for_jpg", None) # Corrected case
)
saved_data = self._save_stage.execute(save_input)
# Log saved_data for merge tasks
if isinstance(item, MergeTaskDefinition):
log.info(f"{item_log_prefix}: SaveVariantsStage result for MergeTask '{item_key}' - Status: {saved_data.status if saved_data else 'N/A'}, Saved Files: {len(saved_data.saved_files_details) if saved_data else 0}")
# Check save status and finalize item result
if saved_data and saved_data.status.startswith("Processed"):
item_status = saved_data.status # e.g., "Processed" or "Processed (No Output)"
log.info(f"{item_log_prefix}: Item successfully processed and saved. Status: {item_status}")
# Populate final details for this item
final_details = {
"status": item_status,
"saved_files_info": saved_data.saved_files_details, # List of dicts from save util
"internal_map_type": internal_map_type,
"original_dimensions": processed_data.original_dimensions if isinstance(processed_data, ProcessedRegularMapData) else None,
"final_dimensions": scaled_data_output.final_dimensions if scaled_data_output else current_dimensions,
"transformations": processed_data.transformations_applied if isinstance(processed_data, ProcessedRegularMapData) else processed_data.transformations_applied_to_inputs,
# Add source file if regular map
"source_file": str(processed_data.source_file_path) if isinstance(processed_data, ProcessedRegularMapData) else None,
}
# Log final details addition for merge tasks
if isinstance(item, MergeTaskDefinition):
log.info(f"{item_log_prefix}: Adding final details to context.processed_maps_details for MergeTask '{item_key}'. Details: {final_details}")
context.processed_maps_details[item_key] = final_details
else:
error_msg = saved_data.error_message if saved_data else "Save stage returned None"
log.error(f"{item_log_prefix}: Failed during save stage. Error: {error_msg}")
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Save Error: {error_msg}", "stage": "SaveVariantsStage"}
asset_had_item_errors = True
continue
item_status = "Failed" # Ensure item status reflects failure
except Exception as e:
log.exception(f"Asset '{asset_name}', Item Loop Index {item_index}: Unhandled exception: {e}")
log.exception(f"{item_log_prefix}: Unhandled exception during item processing loop: {e}")
# Ensure details are recorded even on unhandled exception
if item_key is not None:
context.processed_maps_details[item_key] = {"status": "Failed", "notes": f"Unhandled Loop Error: {e}", "stage": "OrchestratorLoop"}

View File

@ -1,5 +1,5 @@
import logging
from typing import Tuple, Optional # Added Optional
from typing import Tuple
import cv2 # Assuming cv2 is available for interpolation flags
import numpy as np
@ -7,93 +7,77 @@ import numpy as np
from .base_stage import ProcessingStage
# Import necessary context classes and utils
from ..asset_context import InitialScalingInput, InitialScalingOutput
# ProcessingItem is no longer created here, so its import can be removed if not used otherwise.
# For now, keep rule_structure import if other elements from it might be needed,
# but ProcessingItem itself is not directly instantiated by this stage anymore.
# from rule_structure import ProcessingItem
from ...utils import image_processing_utils as ipu
import numpy as np
import cv2 # Added cv2 for interpolation flags (already used implicitly by ipu.resize_image)
log = logging.getLogger(__name__)
class InitialScalingStage(ProcessingStage):
"""
Applies initial Power-of-Two (POT) downscaling to image data if configured
and if the item is not already a 'LOWRES' variant.
Applies initial scaling (e.g., Power-of-Two downscaling) to image data
if configured via the InitialScalingInput.
"""
def execute(self, input_data: InitialScalingInput) -> InitialScalingOutput:
"""
Applies POT scaling based on input_data.initial_scaling_mode,
unless input_data.resolution_key is 'LOWRES'.
Passes through the resolution_key.
Applies scaling based on input_data.initial_scaling_mode.
"""
# Safely access source_file_path for logging, if provided by orchestrator via underscore attribute
source_file_path = getattr(input_data, '_source_file_path', "UnknownSourcePath")
log_prefix = f"InitialScalingStage (Source: {source_file_path}, ResKey: {input_data.resolution_key})"
log.debug(f"{log_prefix}: Mode '{input_data.initial_scaling_mode}'. Received resolution_key: '{input_data.resolution_key}'")
log.debug(f"Initial Scaling Stage: Mode '{input_data.initial_scaling_mode}'.")
image_to_scale = input_data.image_data
current_dimensions_wh = input_data.original_dimensions # Dimensions of the image_to_scale
original_dims_wh = input_data.original_dimensions
scaling_mode = input_data.initial_scaling_mode
output_resolution_key = input_data.resolution_key # Pass through the resolution key
scaling_applied = False
final_image_data = image_to_scale # Default to original if no scaling happens
if image_to_scale is None or image_to_scale.size == 0:
log.warning(f"{log_prefix}: Input image data is None or empty. Skipping POT scaling.")
log.warning("Initial Scaling Stage: Input image data is None or empty. Skipping.")
# Return original (empty) data and indicate no scaling
return InitialScalingOutput(
scaled_image_data=np.array([]),
scaling_applied=False,
final_dimensions=(0, 0),
resolution_key=output_resolution_key
final_dimensions=(0, 0)
)
if not current_dimensions_wh:
log.warning(f"{log_prefix}: Original dimensions not provided for POT scaling. Using current image shape.")
h_pre_pot_scale, w_pre_pot_scale = image_to_scale.shape[:2]
if original_dims_wh is None:
log.warning("Initial Scaling Stage: Original dimensions not provided. Using current image shape.")
h_pre_scale, w_pre_scale = image_to_scale.shape[:2]
original_dims_wh = (w_pre_scale, h_pre_scale)
else:
w_pre_pot_scale, h_pre_pot_scale = current_dimensions_wh
final_image_data = image_to_scale # Default to original if no scaling happens
scaling_applied = False
w_pre_scale, h_pre_scale = original_dims_wh
# Skip POT scaling if the item is already a LOWRES variant or scaling mode is NONE
if output_resolution_key == "LOWRES":
log.info(f"{log_prefix}: Item is a 'LOWRES' variant. Skipping POT downscaling.")
elif scaling_mode == "NONE":
log.info(f"{log_prefix}: Mode is NONE. No POT scaling applied.")
elif scaling_mode == "POT_DOWNSCALE":
pot_w = ipu.get_nearest_power_of_two_downscale(w_pre_pot_scale)
pot_h = ipu.get_nearest_power_of_two_downscale(h_pre_pot_scale)
if (pot_w, pot_h) != (w_pre_pot_scale, h_pre_pot_scale):
log.info(f"{log_prefix}: Applying POT Downscale from ({w_pre_pot_scale},{h_pre_pot_scale}) to ({pot_w},{pot_h}).")
if scaling_mode == "POT_DOWNSCALE":
pot_w = ipu.get_nearest_power_of_two_downscale(w_pre_scale)
pot_h = ipu.get_nearest_power_of_two_downscale(h_pre_scale)
if (pot_w, pot_h) != (w_pre_scale, h_pre_scale):
log.info(f"Initial Scaling: Applying POT Downscale from ({w_pre_scale},{h_pre_scale}) to ({pot_w},{pot_h}).")
# Use INTER_AREA for downscaling generally
resized_img = ipu.resize_image(image_to_scale, pot_w, pot_h, interpolation=cv2.INTER_AREA)
if resized_img is not None:
final_image_data = resized_img
scaling_applied = True
log.debug(f"{log_prefix}: POT Downscale applied successfully.")
log.debug("Initial Scaling: POT Downscale applied successfully.")
else:
log.warning(f"{log_prefix}: POT Downscale resize failed. Using pre-POT-scaled data.")
log.warning("Initial Scaling: POT Downscale resize failed. Using original data.")
# final_image_data remains image_to_scale
else:
log.info(f"{log_prefix}: Image already POT or smaller. No POT scaling needed.")
log.info("Initial Scaling: POT Downscale - Image already POT or smaller. No scaling needed.")
# final_image_data remains image_to_scale
elif scaling_mode == "NONE":
log.info("Initial Scaling: Mode is NONE. No scaling applied.")
# final_image_data remains image_to_scale
else:
log.warning(f"{log_prefix}: Unknown INITIAL_SCALING_MODE '{scaling_mode}'. Defaulting to NONE (no scaling).")
log.warning(f"Initial Scaling: Unknown INITIAL_SCALING_MODE '{scaling_mode}'. Defaulting to NONE.")
# final_image_data remains image_to_scale
# Determine final dimensions
if final_image_data is not None and final_image_data.size > 0:
final_h, final_w = final_image_data.shape[:2]
final_dims_wh = (final_w, final_h)
else:
final_dims_wh = (0,0)
if final_image_data is None: # Ensure it's an empty array for consistency if None
final_image_data = np.array([])
final_h, final_w = final_image_data.shape[:2]
final_dims_wh = (final_w, final_h)
return InitialScalingOutput(
scaled_image_data=final_image_data,
scaling_applied=scaling_applied,
final_dimensions=final_dims_wh,
resolution_key=output_resolution_key # Pass through the resolution key
final_dimensions=final_dims_wh
)

View File

@ -148,15 +148,12 @@ class MetadataInitializationStage(ProcessingStage):
context.asset_metadata['processing_start_time'] = datetime.datetime.now().isoformat()
context.asset_metadata['status'] = "Pending"
app_version_value = None
if context.config_obj and hasattr(context.config_obj, 'app_version'):
app_version_value = context.config_obj.app_version
if app_version_value:
context.asset_metadata['version'] = app_version_value
if context.config_obj and hasattr(context.config_obj, 'general_settings') and \
hasattr(context.config_obj.general_settings, 'app_version'):
context.asset_metadata['version'] = context.config_obj.general_settings.app_version
else:
logger.warning("App version not found using config_obj.app_version. Setting version to 'N/A'.")
context.asset_metadata['version'] = "N/A"
logger.warning("App version not found in config_obj.general_settings. Setting version to 'N/A'.")
context.asset_metadata['version'] = "N/A" # Default or placeholder
if context.incrementing_value is not None:
context.asset_metadata['incrementing_value'] = context.incrementing_value

View File

@ -1,69 +1,21 @@
import logging
from typing import List, Union, Optional, Tuple, Dict # Added Dict
from pathlib import Path # Added Path
from typing import List, Union, Optional
from .base_stage import ProcessingStage
from ..asset_context import AssetProcessingContext, MergeTaskDefinition
from rule_structure import FileRule, ProcessingItem # Added ProcessingItem
from processing.utils import image_processing_utils as ipu # Added ipu
from rule_structure import FileRule # Assuming FileRule is imported correctly
log = logging.getLogger(__name__)
class PrepareProcessingItemsStage(ProcessingStage):
"""
Identifies and prepares a unified list of ProcessingItem and MergeTaskDefinition objects
to be processed in subsequent stages. Performs initial validation and explodes
FileRules into specific ProcessingItems for each required output variant.
Identifies and prepares a unified list of items (FileRule, MergeTaskDefinition)
to be processed in subsequent stages. Performs initial validation.
"""
def _get_target_resolutions(self, source_w: int, source_h: int, config_resolutions: dict, file_rule: FileRule) -> Dict[str, int]:
"""
Determines the target output resolutions for a given source image.
Placeholder logic: Uses all config resolutions smaller than or equal to source, plus PREVIEW if smaller.
Needs to be refined to consider FileRule.resolution_override and actual project requirements.
"""
# For now, very basic logic:
# If FileRule has a resolution_override (e.g., (1024,1024)), that might be the *only* target.
# This needs to be clarified. Assuming override means *only* that size.
if file_rule.resolution_override and isinstance(file_rule.resolution_override, tuple) and len(file_rule.resolution_override) == 2:
# How to get a "key" for an arbitrary override? For now, skip if overridden.
# This part of the design (how overrides interact with standard resolutions) is unclear.
# Let's assume for now that if resolution_override is set, we don't generate standard named resolutions.
# This is likely incorrect for a full implementation.
log.warning(f"FileRule '{file_rule.file_path}' has resolution_override. Standard resolution key generation skipped (needs design refinement).")
return {}
target_res = {}
max_source_dim = max(source_w, source_h)
for key, res_val in config_resolutions.items():
if key == "PREVIEW": # Always consider PREVIEW if its value is smaller
if res_val < max_source_dim : # Or just always include PREVIEW? For now, if smaller.
target_res[key] = res_val
elif res_val <= max_source_dim:
target_res[key] = res_val
# Ensure PREVIEW is included if it's defined and smaller than the smallest other target, or if no other targets.
# This logic is still a bit naive.
if "PREVIEW" in config_resolutions and config_resolutions["PREVIEW"] < max_source_dim:
if not target_res or config_resolutions["PREVIEW"] < min(v for k,v in target_res.items() if k != "PREVIEW" and isinstance(v,int)):
target_res["PREVIEW"] = config_resolutions["PREVIEW"]
elif "PREVIEW" in config_resolutions and not target_res : # if only preview is applicable
if config_resolutions["PREVIEW"] <= max_source_dim:
target_res["PREVIEW"] = config_resolutions["PREVIEW"]
if not target_res and max_source_dim > 0 : # If no standard res is smaller, but image exists
log.debug(f"No standard resolutions from config are <= source dimension {max_source_dim}. Only LOWRES (if applicable) or PREVIEW (if smaller) might be generated.")
log.debug(f"Determined target resolutions for source {source_w}x{source_h}: {target_res}")
return target_res
def execute(self, context: AssetProcessingContext) -> AssetProcessingContext:
"""
Populates context.processing_items with ProcessingItem and MergeTaskDefinition objects.
Populates context.processing_items with FileRule and MergeTaskDefinition objects.
"""
asset_name_for_log = context.asset_rule.asset_name if context.asset_rule else "Unknown Asset"
log.info(f"Asset '{asset_name_for_log}': Preparing processing items...")
@ -73,135 +25,72 @@ class PrepareProcessingItemsStage(ProcessingStage):
context.processing_items = []
return context
# Output list will now be List[Union[ProcessingItem, MergeTaskDefinition]]
items_to_process: List[Union[ProcessingItem, MergeTaskDefinition]] = []
items_to_process: List[Union[FileRule, MergeTaskDefinition]] = []
preparation_failed = False
config = context.config_obj
# --- Process FileRules into ProcessingItems ---
# --- Add regular files ---
if context.files_to_process:
# Validate source path early for regular files
source_path_valid = True
if not context.source_rule or not context.source_rule.input_path:
log.error(f"Asset '{asset_name_for_log}': SourceRule or SourceRule.input_path is not set.")
log.error(f"Asset '{asset_name_for_log}': SourceRule or SourceRule.input_path is not set. Cannot process regular files.")
source_path_valid = False
preparation_failed = True
preparation_failed = True # Mark as failed if source path is missing
context.status_flags['prepare_items_failed_reason'] = "SourceRule.input_path missing"
elif not context.workspace_path or not context.workspace_path.is_dir():
log.error(f"Asset '{asset_name_for_log}': Workspace path '{context.workspace_path}' is invalid.")
log.error(f"Asset '{asset_name_for_log}': Workspace path '{context.workspace_path}' is not a valid directory. Cannot process regular files.")
source_path_valid = False
preparation_failed = True
preparation_failed = True # Mark as failed if workspace path is bad
context.status_flags['prepare_items_failed_reason'] = "Workspace path invalid"
if source_path_valid:
for file_rule in context.files_to_process:
log_prefix_fr = f"Asset '{asset_name_for_log}', FileRule '{file_rule.file_path}'"
# Basic validation for FileRule itself
if not file_rule.file_path:
log.warning(f"{log_prefix_fr}: Skipping FileRule with empty file_path.")
continue
item_type = file_rule.item_type_override or file_rule.item_type
if not item_type or item_type == "EXTRA" or not item_type.startswith("MAP_"):
log.debug(f"{log_prefix_fr}: Item type is '{item_type}'. Not creating map ProcessingItems.")
# Optionally, create a different kind of ProcessingItem for EXTRAs if they need pipeline processing
continue
source_image_path = context.workspace_path / file_rule.file_path
if not source_image_path.is_file():
log.error(f"{log_prefix_fr}: Source image file not found at '{source_image_path}'. Skipping this FileRule.")
preparation_failed = True # Individual file error can contribute to overall stage failure
context.status_flags.setdefault('prepare_items_file_errors', []).append(str(source_image_path))
continue
# Load image data to get dimensions and for LOWRES variant
# This data will be passed to subsequent stages via ProcessingItem.
# Consider caching this load if RegularMapProcessorStage also loads.
# For now, load here as dimensions are needed for LOWRES decision.
log.debug(f"{log_prefix_fr}: Loading image from '{source_image_path}' to determine dimensions and prepare items.")
source_image_data = ipu.load_image(str(source_image_path))
if source_image_data is None:
log.error(f"{log_prefix_fr}: Failed to load image from '{source_image_path}'. Skipping this FileRule.")
preparation_failed = True
context.status_flags.setdefault('prepare_items_file_errors', []).append(f"Failed to load {source_image_path}")
continue
orig_h, orig_w = source_image_data.shape[:2]
original_dimensions_wh = (orig_w, orig_h)
source_bit_depth = ipu.get_image_bit_depth(str(source_image_path)) # Get bit depth from file
source_channels = ipu.get_image_channels(source_image_data)
log.warning(f"Asset '{asset_name_for_log}': Skipping FileRule with empty file_path.")
continue # Skip this specific rule, but don't fail the whole stage
items_to_process.append(file_rule)
log.debug(f"Asset '{asset_name_for_log}': Added {len(context.files_to_process)} potential FileRule items.")
else:
log.warning(f"Asset '{asset_name_for_log}': Skipping addition of all FileRule items due to invalid source/workspace path.")
# Determine standard resolutions to generate
# This logic needs to be robust and consider file_rule.resolution_override, etc.
# Using a placeholder _get_target_resolutions for now.
target_resolutions = self._get_target_resolutions(orig_w, orig_h, config.image_resolutions, file_rule)
# --- Add merged tasks ---
# --- Add merged tasks from global configuration ---
# merged_image_tasks are expected to be loaded into context.config_obj
# by the Configuration class from app_settings.json.
merged_tasks_list = getattr(context.config_obj, 'map_merge_rules', None)
for res_key, _res_val in target_resolutions.items():
pi = ProcessingItem(
source_file_info_ref=str(source_image_path), # Using full path as ref
map_type_identifier=item_type,
resolution_key=res_key,
image_data=source_image_data.copy(), # Give each PI its own copy
original_dimensions=original_dimensions_wh,
current_dimensions=original_dimensions_wh,
bit_depth=source_bit_depth,
channels=source_channels,
status="Pending"
)
items_to_process.append(pi)
log.debug(f"{log_prefix_fr}: Created standard ProcessingItem: {pi.map_type_identifier}_{pi.resolution_key}")
# Create LOWRES variant if applicable
if config.enable_low_resolution_fallback and max(orig_w, orig_h) < config.low_resolution_threshold:
# Check if a LOWRES item for this source_file_info_ref already exists (e.g. if target_resolutions was empty)
# This check is important if _get_target_resolutions might return empty for small images.
# A more robust way is to ensure LOWRES is distinct from standard resolutions.
# Avoid duplicate LOWRES if _get_target_resolutions somehow already made one (unlikely with current placeholder)
is_lowres_already_added = any(p.resolution_key == "LOWRES" and p.source_file_info_ref == str(source_image_path) for p in items_to_process if isinstance(p, ProcessingItem))
if not is_lowres_already_added:
pi_lowres = ProcessingItem(
source_file_info_ref=str(source_image_path),
map_type_identifier=item_type,
resolution_key="LOWRES",
image_data=source_image_data.copy(), # Fresh copy for LOWRES
original_dimensions=original_dimensions_wh,
current_dimensions=original_dimensions_wh,
bit_depth=source_bit_depth,
channels=source_channels,
status="Pending"
)
items_to_process.append(pi_lowres)
log.info(f"{log_prefix_fr}: Created LOWRES ProcessingItem because {orig_w}x{orig_h} < {config.low_resolution_threshold}px threshold.")
else:
log.debug(f"{log_prefix_fr}: LOWRES item for this source already added by target resolution logic. Skipping duplicate LOWRES creation.")
elif config.enable_low_resolution_fallback:
log.debug(f"{log_prefix_fr}: Image {orig_w}x{orig_h} not below LOWRES threshold {config.low_resolution_threshold}px.")
else: # Source path not valid
log.warning(f"Asset '{asset_name_for_log}': Skipping creation of ProcessingItems from FileRules due to invalid source/workspace path.")
# --- Add MergeTaskDefinitions --- (This part remains largely the same)
merged_tasks_list = getattr(config, 'map_merge_rules', None)
if merged_tasks_list and isinstance(merged_tasks_list, list):
log.debug(f"Asset '{asset_name_for_log}': Found {len(merged_tasks_list)} merge tasks in global config.")
for task_idx, task_data in enumerate(merged_tasks_list):
if isinstance(task_data, dict):
task_key = f"merged_task_{task_idx}"
# Basic validation for merge task data: requires output_map_type and an inputs dictionary
if not task_data.get('output_map_type') or not isinstance(task_data.get('inputs'), dict):
log.warning(f"Asset '{asset_name_for_log}', Task Index {task_idx}: Skipping merge task due to missing 'output_map_type' or valid 'inputs'. Task data: {task_data}")
continue
log.warning(f"Asset '{asset_name_for_log}', Task Index {task_idx}: Skipping merge task due to missing 'output_map_type' or valid 'inputs' dictionary. Task data: {task_data}")
continue # Skip this specific task
log.debug(f"Asset '{asset_name_for_log}', Preparing Merge Task Index {task_idx}: Raw task_data: {task_data}")
merge_def = MergeTaskDefinition(task_data=task_data, task_key=task_key)
log.debug(f"Asset '{asset_name_for_log}': Created MergeTaskDefinition object: {merge_def}")
log.info(f"Asset '{asset_name_for_log}': Successfully CREATED MergeTaskDefinition: Key='{merge_def.task_key}', OutputType='{merge_def.task_data.get('output_map_type', 'N/A')}'")
items_to_process.append(merge_def)
log.info(f"Asset '{asset_name_for_log}': Added MergeTaskDefinition: Key='{merge_def.task_key}', OutputType='{merge_def.task_data.get('output_map_type', 'N/A')}'")
else:
log.warning(f"Asset '{asset_name_for_log}': Item at index {task_idx} in config.map_merge_rules is not a dict. Skipping. Item: {task_data}")
# ... (rest of merge task handling) ...
log.warning(f"Asset '{asset_name_for_log}': Item at index {task_idx} in config_obj.merged_image_tasks is not a dictionary. Skipping. Item: {task_data}")
# The log for "Added X potential MergeTaskDefinition items" will be covered by the final log.
elif merged_tasks_list is None:
log.debug(f"Asset '{asset_name_for_log}': 'merged_image_tasks' not found in config_obj. No global merge tasks to add.")
elif not isinstance(merged_tasks_list, list):
log.warning(f"Asset '{asset_name_for_log}': 'merged_image_tasks' in config_obj is not a list. Skipping global merge tasks. Type: {type(merged_tasks_list)}")
else: # Empty list
log.debug(f"Asset '{asset_name_for_log}': 'merged_image_tasks' in config_obj is empty. No global merge tasks to add.")
if not items_to_process and not preparation_failed: # Check preparation_failed too
log.info(f"Asset '{asset_name_for_log}': No valid items (ProcessingItem or MergeTaskDefinition) found to process.")
if not items_to_process:
log.info(f"Asset '{asset_name_for_log}': No valid items found to process after preparation.")
log.debug(f"Asset '{asset_name_for_log}': Final items_to_process before assigning to context: {items_to_process}")
context.processing_items = items_to_process
context.intermediate_results = {} # Initialize intermediate results storage

View File

@ -37,7 +37,7 @@ class RegularMapProcessorStage(ProcessingStage):
"""
final_internal_map_type = initial_internal_map_type # Default
base_map_type_match = re.match(r"(MAP_[A-Z]+)", initial_internal_map_type)
base_map_type_match = re.match(r"(MAP_[A-Z]{3})", initial_internal_map_type)
if not base_map_type_match or not asset_rule or not asset_rule.files:
return final_internal_map_type # Cannot determine suffix without base type or asset rule files
@ -47,7 +47,7 @@ class RegularMapProcessorStage(ProcessingStage):
peers_of_same_base_type = []
for fr_asset in asset_rule.files:
fr_asset_item_type = fr_asset.item_type_override or fr_asset.item_type or "UnknownMapType"
fr_asset_base_match = re.match(r"(MAP_[A-Z]+)", fr_asset_item_type)
fr_asset_base_match = re.match(r"(MAP_[A-Z]{3})", fr_asset_item_type)
if fr_asset_base_match and fr_asset_base_match.group(1) == true_base_map_type:
peers_of_same_base_type.append(fr_asset)
@ -197,17 +197,10 @@ class RegularMapProcessorStage(ProcessingStage):
result.final_internal_map_type = final_map_type # Update if Gloss->Rough changed it
result.transformations_applied = transform_notes
# --- Determine Resolution Key for LOWRES ---
if config.enable_low_resolution_fallback and result.original_dimensions:
w, h = result.original_dimensions
if max(w, h) < config.low_resolution_threshold:
result.resolution_key = "LOWRES"
log.info(f"{log_prefix}: Image dimensions ({w}x{h}) are below threshold ({config.low_resolution_threshold}px). Flagging as LOWRES.")
# --- Success ---
result.status = "Processed"
result.error_message = None
log.info(f"{log_prefix}: Successfully processed regular map. Final type: '{result.final_internal_map_type}', ResolutionKey: {result.resolution_key}.")
log.info(f"{log_prefix}: Successfully processed regular map. Final type: '{result.final_internal_map_type}'.")
except Exception as e:
log.exception(f"{log_prefix}: Unhandled exception during processing: {e}")

View File

@ -23,17 +23,8 @@ class SaveVariantsStage(ProcessingStage):
Calls isu.save_image_variants with data from input_data.
"""
internal_map_type = input_data.internal_map_type
# The input_data for SaveVariantsStage doesn't directly contain the ProcessingItem.
# It receives data *derived* from a ProcessingItem by previous stages.
# For debugging, we'd need to pass more context or rely on what's in output_filename_pattern_tokens.
resolution_key_from_tokens = input_data.output_filename_pattern_tokens.get('resolution', 'UnknownResKey')
log_prefix = f"Save Variants Stage (Type: {internal_map_type}, ResKey: {resolution_key_from_tokens})"
log_prefix = f"Save Variants Stage (Type: {internal_map_type})"
log.info(f"{log_prefix}: Starting.")
log.debug(f"{log_prefix}: Input image_data shape: {input_data.image_data.shape if input_data.image_data is not None else 'None'}")
log.debug(f"{log_prefix}: Input source_bit_depth_info: {input_data.source_bit_depth_info}")
log.debug(f"{log_prefix}: Configured image_resolutions for saving: {input_data.image_resolutions}")
log.debug(f"{log_prefix}: Output filename pattern tokens: {input_data.output_filename_pattern_tokens}")
# Initialize output object with default failure state
result = SaveVariantsOutput(
@ -73,11 +64,11 @@ class SaveVariantsStage(ProcessingStage):
"resolution_threshold_for_jpg": input_data.resolution_threshold_for_jpg, # Added
}
log.debug(f"{log_prefix}: Calling save_image_variants utility with args: {save_args}")
log.debug(f"{log_prefix}: Calling save_image_variants utility.")
saved_files_details: List[Dict] = isu.save_image_variants(**save_args)
if saved_files_details:
log.info(f"{log_prefix}: Save utility completed successfully. Saved {len(saved_files_details)} variants: {[details.get('filepath') for details in saved_files_details]}")
log.info(f"{log_prefix}: Save utility completed successfully. Saved {len(saved_files_details)} variants.")
result.saved_files_details = saved_files_details
result.status = "Processed"
result.error_message = None

View File

@ -194,16 +194,6 @@ def get_image_bit_depth(image_path_str: str) -> Optional[int]:
print(f"Error getting bit depth for {image_path_str}: {e}")
return None
def get_image_channels(image_data: np.ndarray) -> Optional[int]:
"""Determines the number of channels in an image."""
if image_data is None:
return None
if len(image_data.shape) == 2: # Grayscale
return 1
elif len(image_data.shape) == 3: # Color
return image_data.shape[2]
return None # Unknown shape
def calculate_image_stats(image_data: np.ndarray) -> Optional[Dict]:
"""
Calculates min, max, mean for a given numpy image array.

View File

@ -1,44 +0,0 @@
# Project Brief: Asset Processor Tool
## 1. Main Goal & Purpose
The primary goal of the Asset Processor Tool is to provide **CG artists and 3D content teams with a friendly, fast, and flexible interface to process and organize 3D asset source files into a standardized library format.** It automates repetitive and complex tasks involved in preparing assets from various suppliers for use in production pipelines.
## 2. Key Features & Components
* **Automated Asset Processing:** Ingests 3D asset source files (texture sets, models, etc.) from `.zip`, `.rar`, `.7z` archives, or folders.
* **Preset-Driven Workflow:** Utilizes configurable JSON presets to interpret different asset sources (e.g., from various online vendors or internal standards), defining rules for file classification and processing.
* **Comprehensive File Operations:**
* **Classification:** Automatically identifies map types (Color, Normal, Roughness, etc.), models, and other file categories based on preset rules.
* **Image Processing:** Performs tasks like image resizing (to standard resolutions like 1K, 2K, 4K, avoiding upscaling), glossiness-to-roughness conversion, normal map green channel inversion (OpenGL/DirectX handling), alpha channel extraction, bit-depth adjustments, and low-resolution fallback generation for small source images.
* **Channel Merging:** Combines channels from different source maps into packed textures (e.g., Normal + Roughness + Metallic into a single NRMRGH map).
* **Metadata Generation:** Creates a detailed `metadata.json` file for each processed asset, containing information about maps, categories, processing settings, and more, for downstream tool integration.
* **Flexible Output Organization:** Generates a clean, structured output directory based on user-configurable naming patterns and tokens.
* **Multiple User Interfaces:**
* **Graphical User Interface (GUI):** The primary interface, designed to be user-friendly, offering drag-and-drop functionality, an integrated preset editor, a live preview table for rule validation and overrides, and clear processing controls.
* **Directory Monitor:** An automated script that watches a specified folder for new asset archives and processes them based on preset names embedded in the archive filename.
* **Command-Line Interface (CLI):** Intended for batch processing and scripting (currently with limited core functionality).
* **Optional Blender Integration:** Can automatically run Blender scripts post-processing to create PBR node groups and materials in specified `.blend` files, linking to the newly processed textures.
* **Hierarchical Rule System:** Allows for dynamic, granular overrides of preset configurations at the source, asset, or individual file level via the GUI.
* **Experimental LLM Prediction:** Includes an option to use a Large Language Model for file interpretation and rule prediction.
## 3. Target Audience
* **CG Artists:** Individual artists looking for an efficient way to manage and prepare their personal or downloaded asset libraries.
* **3D Content Creation Teams:** Studios or groups needing a standardized pipeline for processing and organizing assets from multiple sources.
* **Technical Artists/Pipeline Developers:** Who may extend or integrate the tool into broader production workflows.
## 4. Overall Architectural Style & Key Technologies
* **Core Language:** Python
* **GUI Framework:** PySide6
* **Configuration:** Primarily JSON-based (application settings, user overrides, type definitions, supplier settings, presets, LLM settings).
* **Processing Architecture:** A modular, staged processing pipeline orchestrated by a central engine. Each stage performs a discrete task on an `AssetProcessingContext` object.
* **Key Libraries:** OpenCV (image processing), NumPy (numerical operations), py7zr/rarfile (archive handling), watchdog (directory monitoring).
* **Design Principles:** Modularity, configurability, and user-friendliness (especially for the GUI).
## 5. Foundational Information
* The tool aims to significantly reduce manual effort and ensure consistency in asset preparation.
* It is designed to be adaptable to various asset sources and pipeline requirements through its extensive configuration options and preset system.
* The output `metadata.json` is key for enabling further automation and integration with other tools or digital content creation (DCC) applications.

View File

@ -1,7 +1,6 @@
import dataclasses
import json
from typing import List, Dict, Any, Tuple, Optional
import numpy as np # Added for ProcessingItem
@dataclasses.dataclass
class FileRule:
file_path: str = None
@ -11,12 +10,8 @@ class FileRule:
resolution_override: Tuple[int, int] = None
channel_merge_instructions: Dict[str, Any] = dataclasses.field(default_factory=dict)
output_format_override: str = None
processing_items: List['ProcessingItem'] = dataclasses.field(default_factory=list) # Added field
def to_json(self) -> str:
# Need to handle ProcessingItem serialization if it contains non-serializable types like np.ndarray
# For now, assume asdict handles it or it's handled before calling to_json for persistence.
# A custom asdict_factory might be needed for robust serialization.
return json.dumps(dataclasses.asdict(self), indent=4)
@classmethod
@ -59,43 +54,4 @@ class SourceRule:
data = json.loads(json_string)
# Manually deserialize nested AssetRule objects
data['assets'] = [AssetRule.from_json(json.dumps(asset_data)) for asset_data in data.get('assets', [])]
# Need to handle ProcessingItem deserialization if it was serialized
# For now, from_json for FileRule doesn't explicitly handle processing_items from JSON.
return cls(**data)
@dataclasses.dataclass
class ProcessingItem:
"""
Represents a specific version of an image map to be processed and saved.
This could be a standard resolution (1K, 2K), a preview, or a special
variant like 'LOWRES'.
"""
source_file_info_ref: str # Reference to the original SourceFileInfo or unique ID of the source image
map_type_identifier: str # The internal map type (e.g., "MAP_COL", "MAP_ROUGH")
resolution_key: str # The resolution identifier (e.g., "1K", "PREVIEW", "LOWRES")
image_data: np.ndarray # The actual image data for this item
original_dimensions: Tuple[int, int] # (width, height) of the source image for this item
current_dimensions: Tuple[int, int] # (width, height) of the image_data in this item
target_filename: str = "" # Will be populated by SaveVariantsStage
is_extra: bool = False # If this item should be treated as an 'extra' file
bit_depth: Optional[int] = None
channels: Optional[int] = None
file_extension: Optional[str] = None # Determined during saving based on format
processing_applied_log: List[str] = dataclasses.field(default_factory=list)
status: str = "Pending" # e.g., Pending, Processed, Failed
error_message: Optional[str] = None
# __getstate__ and __setstate__ might be needed if we pickle these objects
# and np.ndarray causes issues. For JSON, image_data would typically not be serialized.
def __getstate__(self):
state = self.__dict__.copy()
# Don't pickle image_data if it's large or not needed for state
if 'image_data' in state: # Or a more sophisticated check
del state['image_data'] # Example: remove it
return state
def __setstate__(self, state):
self.__dict__.update(state)
# Potentially re-initialize or handle missing 'image_data'
if 'image_data' not in self.__dict__:
self.image_data = None # Or load it if a path was stored instead
return cls(**data)