Revert "First pass at optionally truncating long string values"

This reverts commit 8bde832.

morpheus/modules/output/write_to_vector_db.py

@@ -148,8 +148,7 @@ def on_completed():
             try:
                 if accum_stats.data:
                     merged_df = cudf.concat(accum_stats.data)
-
-                    service.insert_dataframe(name=key, df=merged_df, truncate_long_strings=True)
+                    service.insert_dataframe(name=key, df=merged_df)
                     final_df_references.append(accum_stats.data)
             except Exception as e:
                 logger.error("Unable to upload dataframe entries to vector database: %s", e)
@@ -214,10 +213,7 @@ def on_data(msg: typing.Union[ControlMessage, MultiResponseMessage, MultiMessage
                             merged_df = cudf.concat(accum_stats.data)
 
                             # pylint: disable=not-a-mapping
-                            service.insert_dataframe(name=key,
-                                                     df=merged_df,
-                                                     truncate_long_strings=True,
-                                                     **resource_kwargs)
+                            service.insert_dataframe(name=key, df=merged_df, **resource_kwargs)
                             # Reset accumulator stats
                             accum_stats.data.clear()
                             accum_stats.last_insert_time = current_time

morpheus/service/vdb/milvus_vector_db_service.py

@@ -32,9 +32,6 @@
 IMPORT_EXCEPTION = None
 IMPORT_ERROR_MESSAGE = "MilvusVectorDBResourceService requires the milvus and pymilvus packages to be installed."
 
-# https://milvus.io/docs/limitations.md#Length-of-a-string
-MAX_STRING_LENGTH = 65535
-
 try:
     import pymilvus
     from pymilvus.orm.mutation import MutationResult
@@ -278,19 +275,14 @@ def insert(self, data: list[list] | list[dict], **kwargs: dict[str, typing.Any])
 
         return self._insert_result_to_dict(result=result)
 
-    def insert_dataframe(self,
-                         df: typing.Union[cudf.DataFrame, pd.DataFrame],
-                         truncate_long_strings: bool = False,
-                         **kwargs: dict[str, typing.Any]) -> dict:
+    def insert_dataframe(self, df: typing.Union[cudf.DataFrame, pd.DataFrame], **kwargs: dict[str, typing.Any]) -> dict:
         """
         Insert a dataframe entires into the vector database.
 
         Parameters
         ----------
         df : typing.Union[cudf.DataFrame, pd.DataFrame]
             Dataframe to be inserted into the collection.
-        truncate_long_strings : bool, optional
-            When true, truncate strings values that are longer than the max length supported by Milvus (65535).
         **kwargs : dict[str, typing.Any]
             Extra keyword arguments specific to the vector database implementation.
 
@@ -303,19 +295,6 @@ def insert_dataframe(self,
         if isinstance(df, cudf.DataFrame):
             df = df.to_pandas()
 
-        if truncate_long_strings:
-            for col in df:
-                str_series = df[col]
-                if str_series.dtype == "object":
-                    max_len = str_series.str.len().max()
-                    if max_len > MAX_STRING_LENGTH:
-                        logger.warning(("Column '%s' has a string length of %d, larger than the max of %d"
-                                        "supported by Milvus, truncating"),
-                                       col,
-                                       max_len,
-                                       MAX_STRING_LENGTH)
-                        df[col] = str_series.str.slice(0, MAX_STRING_LENGTH)
-
         # Ensure that there are no None values in the DataFrame entries.
         for field_name, dtype in self._fillna_fields_dict.items():
             if dtype in (pymilvus.DataType.VARCHAR, pymilvus.DataType.STRING):