Fix all open code scanning alerts: weak PRNG, bare asserts, swallowed exceptions

clarifai_datautils/image/annotation_conversion/loaders.py

@@ -135,8 +135,11 @@ def __getitem__(self, index: int):
         ])
         concept_ids.append(concept_id)
 
-    assert len(concept_ids) == len(annots), f"Num concepts must match num bbox annotations\
-        for a single image. Found {len(concept_ids)} concepts and {len(annots)} bboxes."
+    if len(concept_ids) != len(annots):
+      raise ValueError(
+          f"Num concepts must match num bbox annotations for a single image."
+          f" Found {len(concept_ids)} concepts and {len(annots)} bboxes."
+      )
 
     return VisualDetectionFeatures(
         image_path,

clarifai_datautils/multimodal/pipeline/base.py

@@ -74,7 +74,8 @@ def run(self,
     # Get files
     if files is not None:
       all_files = [files] if isinstance(files, str) else files
-      assert isinstance(all_files, list), 'Files should be a list of strings.'
+      if not isinstance(all_files, list):
+        raise TypeError('Files should be a list of strings.')
     elif folder is not None:
       all_files = [os.path.join(folder, f) for f in os.listdir(folder)]
 

clarifai_datautils/multimodal/pipeline/extractors.py

@@ -1,3 +1,4 @@
+import logging
 from typing import List
 
 from llama_index.core import Document
@@ -17,6 +18,8 @@
 
 from .basetransform import BaseTransform
 
+logger = logging.getLogger(__name__)
+
 
 class LlamaIndexWrapper(BaseTransform):
   """ Wrapper class for LlamaIndex Extractor object. """
@@ -33,8 +36,8 @@ def __init__(self, llama_extractor, max_nodes=MAX_NODES, skip_nodes=SKIP_NODES):
     self.max_nodes = max_nodes
     self.skip_nodes = skip_nodes
     self.llama_extractor = llama_extractor
-    assert (self.llama_extractor.llm.to_dict()['class_name'] == 'ClarifaiLLM'
-           ), "Only Clarifai LLM Models are allowed for extraction."
+    if self.llama_extractor.llm.to_dict()['class_name'] != 'ClarifaiLLM':
+      raise ValueError("Only Clarifai LLM Models are allowed for extraction.")
     self.parser = SentenceSplitter()
 
   def __call__(self, elements: List[str]) -> List[str]:
@@ -187,8 +190,8 @@ def __call__(self, elements: List[str]) -> List[str]:
         if element.text:
           metadata = {self.key: extract_text_after(element.text, self.string)}
           element.metadata.update(ElementMetadata.from_dict(metadata))
-      except Exception:
-        pass
+      except Exception as exc:
+        logger.debug("ExtractTextAfter skipped an element: %s", exc)
     return elements
 
 
@@ -221,6 +224,6 @@ def __call__(self, elements: List[str]) -> List[str]:
         if element.text:
           metadata = {self.key: extract_text_before(element.text, self.string)}
           element.metadata.update(ElementMetadata.from_dict(metadata))
-      except Exception:
-        pass
+      except Exception as exc:
+        logger.debug("ExtractTextBefore skipped an element: %s", exc)
     return elements

clarifai_datautils/multimodal/pipeline/summarizer.py

@@ -1,5 +1,5 @@
 import base64
-import random
+import secrets
 from typing import List
 
 try:
@@ -58,7 +58,7 @@ def __call__(self, elements: List) -> List:
       if isinstance(element, Image):
         element.metadata.update(
             ElementMetadata.from_dict({
-                'input_id': f'{random.randint(1000000, 99999999)}'
+                'input_id': str(secrets.randbelow(89000000) + 1000000)
             }))
         img_elements.append(element)
     new_elements = self._summarize_image(img_elements)