marieai · rsteele5 · Aug 27, 2025 · Aug 27, 2025 · Sep 2, 2025 · Aug 14, 2025
diff --git a/marie/api/__init__.py b/marie/api/__init__.py
@@ -6,14 +6,11 @@
 
 import cv2
 import numpy as np
-from docarray import DocList
 
 from marie.api.docs import AssetKeyDoc
 from marie.logging_core.predefined import default_logger
 from marie.storage import StorageManager
 from marie.utils.base64 import base64StringToBytes
-from marie.utils.docs import docs_from_asset, frames_from_docs
-from marie.utils.image_utils import ensure_max_page_size
 from marie.utils.utils import FileSystem, ensure_exists
 
 logger = default_logger
@@ -279,118 +276,17 @@ def parse_payload_to_docs_sync(
             key = "uri"
         del payload[key]
 
-    ref_id = value_from_payload_or_args(payload, "ref_id",
-                                        default=value_from_payload_or_args(payload, "doc_id", default=""))
-    ref_type = value_from_payload_or_args(payload, "ref_type",
-                                          default=value_from_payload_or_args(payload, "doc_type", default=""))
+    ref_id = value_from_payload_or_args(
+        payload,
+        "ref_id",
+        default=value_from_payload_or_args(payload, "doc_id", default=""),
+    )
+    ref_type = value_from_payload_or_args(
+        payload,
+        "ref_type",
+        default=value_from_payload_or_args(payload, "doc_type", default=""),
+    )
     asset_doc = AssetKeyDoc(asset_key=asset_uri, pages=pages)
     parameters = {"queue_id": queue_id, "ref_id": ref_id, "ref_type": ref_type}
 
     return parameters, asset_doc
-
-
-def parse_parameters(parameters: dict, strict: bool = True) -> tuple:
-    """
-    Parses a dictionary of parameters and extracts relevant information such as job_id, ref_id, ref_type, queue_id,
-    and payload. This function includes error handling for missing values based on the `strict` flag.
-
-    Args:
-        parameters (dict): A dictionary of parameters containing keys like 'job_id', 'ref_id', 'ref_type',
-                           'queue_id', and 'payload'.
-        strict (bool, optional): If set to True, raises a ValueError when required parameters are missing.
-                                 Defaults to True.
-
-    Returns:
-        tuple: A tuple containing extracted values in the following order:
-               (job_id, ref_id, ref_type, queue_id, payload).
-
-    Raises:
-        ValueError: If a required parameter is missing and the `strict` flag is set to True.
-    """
-    if parameters is None or "job_id" not in parameters:
-        logger.error(f"Job ID is not present in parameters")
-        if strict:
-            raise ValueError("Job ID is not present in parameters")
-
-    job_id = parameters.get("job_id", "0000-0000-0000-0000")
-
-    logger.debug("Parsing Parameters")
-    for key, value in parameters.items():
-        logger.debug("The value of {} is {}".format(key, value))
-
-    ref_id = parameters.get("ref_id")
-    if ref_id is None and strict:
-        raise ValueError("ref_id is not present in parameters")
-    ref_type = parameters.get("ref_type", "not_defined")
-    queue_id: str = parameters.get("queue_id", "0000-0000-0000-0000")
-
-    payload = parameters.get("payload")
-    if payload is None:
-        logger.error("Empty Payload")
-        if strict:
-            raise ValueError("Empty Payload")
-
-    return job_id, ref_id, ref_type, queue_id, payload
-
-
-def get_frames_from_docs(
-    docs: DocList[AssetKeyDoc], pages: list[int] = None
-) -> list[np.ndarray]:
-    """
-    Extracts and processes frames from a single document.
-
-    This function is responsible for extracting frames from a provided document, ensuring
-    that the document adheres to specific constraints (only a single document is supported).
-    It checks and retrieves the document's frames from specific pages if specified, or all pages
-    if none are given. Additionally, it ensures that the frames comply with a maximum page size
-    constraint, adjusting their size as necessary and logging relevant warnings.
-
-    Parameters:
-        docs (DocList[AssetKeyDoc]): A list containing a single document from which to extract frames.
-        pages (list[int], optional): A list of page indices to extract frames from. If not provided,
-                                     frames from all pages of the document will be extracted.
-
-    Returns:
-        list[numpy.ndarray]: A list of processed frames extracted from the specified pages of the
-                             document.
-
-    Raises:
-        ValueError: If no documents are found in the input, or if multiple documents are provided.
-    """
-    if len(docs) == 0:
-        raise ValueError("Expected single document. No documents found")
-    if len(docs) > 1:
-        raise ValueError("Expected single document. Multiple documents found.")
-
-    doc: AssetKeyDoc = docs[0]
-    logger.debug(f"Document asset key: {doc.asset_key}")
-    pages = doc.pages if pages is None else pages
-    docs = docs_from_asset(doc.asset_key, pages)
-    src_frames = frames_from_docs(docs)
-    changed, frames = ensure_max_page_size(src_frames)
-    if changed:
-        logger.warning(f"Page size of frames was changed ")
-        for i, (s, f) in enumerate(zip(src_frames, frames)):
-            logger.warning(f"Frame[{i}] changed : {s.shape} -> {f.shape}")
-
-    return frames
-
-
-def get_payload_features(
-    payload,
-    name=None,
-    f_type=None,
-) -> list:
-    if "features" not in payload:
-        return []
-
-    features = []
-    for feature in payload["features"]:
-        if not isinstance(feature, dict):
-            continue
-        if name and feature.get("name") != name:
-            continue
-        if f_type and feature.get("type") != f_type:
-            continue
-        features.append(feature)
-    return features
diff --git a/marie/executor/extract/util.py → marie/executor/asset_util.py b/marie/executor/extract/util.py → marie/executor/asset_util.py
diff --git a/marie/executor/classifier/__init__.py b/marie/executor/classifier/__init__.py
@@ -1 +1,3 @@
-from .document_classifier_executor import DocumentClassificationExecutor
+from marie.executor.classifier.document_classifier_executor import (
+    DocumentClassificationExecutor,
+)