Bug fixes in key phrase and pre embedding cleaner (#47)

* Bug fixes in key phrase and pre embedding cleaner

adi_function_app/key_phrase_extraction.py

@@ -1,84 +1,107 @@
 # Copyright (c) Microsoft Corporation.
 # Licensed under the MIT License.
-
 import logging
 import json
 import os
 from azure.ai.textanalytics.aio import TextAnalyticsClient
 from azure.core.exceptions import HttpResponseError
-from azure.core.credentials import AzureKeyCredential
-import asyncio
 from azure.identity import DefaultAzureCredential
-from environment import IdentityType, get_identity_type
+from tenacity import retry
+from tenacity.stop import stop_after_attempt
+from tenacity.wait import wait_exponential
+import asyncio
 
 MAX_TEXT_ELEMENTS = 5120
 
 
-def split_document(document: str, max_size: int) -> list[str]:
-    """Split a document into chunks of max_size.
+def split_document(document, max_size):
+    """Split a document into chunks of max_size and filter out any empty strings
 
     Args:
         document (str): The document to split.
-        max_size (int): The maximum size of each chunk."""
-    return [document[i : i + max_size] for i in range(0, len(document), max_size)]
+        max_size (int): The maximum size of each chunk.
 
-
-async def extract_key_phrases_from_text(
-    data: list[str], max_key_phrase_count: int, retries_left: int = 3
+    Returns:
+        list: The list of document chunks."""
+    return [
+        document[i : i + max_size]
+        for i in range(0, len(document), max_size)
+        if len(document[i : i + max_size]) > 0
+    ]
+
+
+@retry(
+    reraise=True,
+    stop=stop_after_attempt(3),
+    wait=wait_exponential(multiplier=1, min=1, max=10),
+)
+async def extract_key_phrases_from_batch(
+    batch_data: list[str], max_key_phrase_count: int
 ) -> list[str]:
-    """Extract key phrases from the text.
+    """Extract key phrases from text using Azure AI services.
 
     Args:
-        data (list[str]): The text data.
-        max_key_phrase_count (int): The maximum number of key phrases to return.
+        batch_data (list[str]): The list of text to process.
+        max_key_phrase_count(int): no of keywords to return
 
     Returns:
-        list[str]: The key phrases extracted from the text."""
-    logging.info("Python HTTP trigger function processed a request.")
+        list: The list of key phrases."""
 
     key_phrase_list = []
 
-    if get_identity_type() == IdentityType.SYSTEM_ASSIGNED:
-        credential = DefaultAzureCredential()
-    elif get_identity_type() == IdentityType.USER_ASSIGNED:
-        credential = DefaultAzureCredential(
-            managed_identity_client_id=os.environ.get("FunctionApp__ClientId")
-        )
-    else:
-        credential = AzureKeyCredential(os.environ.get("AIService__Language__Key"))
     text_analytics_client = TextAnalyticsClient(
-        endpoint=os.environ.get("AIService__Language__Endpoint"),
-        credential=credential,
+        endpoint=os.environ["AIService__Services__Endpoint"],
+        credential=DefaultAzureCredential(
+            managed_identity_client_id=os.environ.get("FunctionApp__ClientId")
+        ),
     )
 
     async with text_analytics_client:
         try:
-            # Split large documents
-            split_documents = []
-            for doc in data:
-                if len(doc) > MAX_TEXT_ELEMENTS:
-                    split_documents.extend(split_document(doc, MAX_TEXT_ELEMENTS))
-                else:
-                    split_documents.append(doc)
-
-            result = await text_analytics_client.extract_key_phrases(split_documents)
-            for idx, doc in enumerate(result):
+            result = await text_analytics_client.extract_key_phrases(batch_data)
+            for doc in result:
                 if not doc.is_error:
                     key_phrase_list.extend(doc.key_phrases[:max_key_phrase_count])
                 else:
-                    raise Exception(f"Document {idx} error: {doc.error}")
+                    raise Exception(f"Document error: {doc.error}")
         except HttpResponseError as e:
-            if e.status_code == 429 and retries_left > 0:  # Rate limiting error
-                wait_time = 2**retries_left  # Exponential backoff
-                logging.info(
-                    "%s Rate limit exceeded. Retrying in %s seconds...", e, wait_time
-                )
-                await asyncio.sleep(wait_time)
-                return await extract_key_phrases_from_text(
-                    data, max_key_phrase_count, retries_left - 1
-                )
-            else:
-                raise Exception(f"An error occurred: {e}") from e
+            logging.error("An error occurred: %s", e)
+            raise e
+
+    return key_phrase_list
+
+
+async def extract_key_phrases_from_text(
+    data: list[str], max_key_phrase_count: int
+) -> list[str]:
+    """Extract key phrases from text using Azure AI services.
+
+    Args:
+        data (list[str]): The list of text to process.
+        max_key_phrase_count(int): no of keywords to return"""
+    logging.info("Python HTTP trigger function processed a request.")
+    key_phrase_list = []
+
+    split_documents = []
+    for doc in data:
+        if len(doc) > MAX_TEXT_ELEMENTS:
+            split_documents.extend(split_document(doc, MAX_TEXT_ELEMENTS))
+        elif len(doc) > 0:
+            split_documents.append(doc)
+
+    # Filter out any empty documents
+    split_documents = [doc for doc in split_documents if len(doc) > 0]
+
+    for i in range(0, len(split_documents), 10):
+        key_phrase_list.extend(
+            await extract_key_phrases_from_batch(
+                split_documents[i : i + 10], max_key_phrase_count
+            )
+        )
+
+        if len(key_phrase_list) > max_key_phrase_count:
+            key_phrase_list = key_phrase_list[:max_key_phrase_count]
+            break
 
     return key_phrase_list
 
@@ -105,26 +128,40 @@ async def process_key_phrase_extraction(
             "errors": None,
             "warnings": None,
         }
-        extracted_record["data"]["key_phrases"] = await extract_key_phrases_from_text(
+        extracted_record["data"]["keyPhrases"] = await extract_key_phrases_from_text(
             [record["data"]["text"]], max_key_phrase_count
         )
-    except Exception as inner_e:
-        logging.error("key phrase extraction Error: %s", inner_e)
-        logging.error(
-            "Failed to extract key phrase. Check function app logs for more details of exact failure."
-        )
-        return {
-            "recordId": record["recordId"],
-            "data": {},
-            "errors": [
-                {
-                    "message": "Failed to extract key phrase. Check function app logs for more details of exact failure."
-                }
-            ],
-            "warnings": None,
-        }
-    else:
-        json_str = json.dumps(extracted_record, indent=4)
-
-        logging.info(f"key phrase extraction output: {json_str}")
-        return extracted_record
+    except Exception as e:
+        logging.error("key phrase extraction Error: %s", e)
+        await asyncio.sleep(10)
+        try:
+            extracted_record = {
+                "recordId": record["recordId"],
+                "data": {},
+                "errors": None,
+                "warnings": None,
+            }
+            extracted_record["data"][
+                "keyPhrases"
+            ] = await extract_key_phrases_from_text(
+                [record["data"]["text"]], max_key_phrase_count
+            )
+        except Exception as inner_e:
+            logging.error("key phrase extraction Error: %s", inner_e)
+            logging.error(
+                "Failed to extract key phrase. Check function app logs for more details of exact failure."
+            )
+            return {
+                "recordId": record["recordId"],
+                "data": {},
+                "errors": [
+                    {
+                        "message": "Failed to extract key phrase. Check function app logs for more details of exact failure."
+                    }
+                ],
+                "warnings": None,
+            }
+    json_str = json.dumps(extracted_record, indent=4)
+
+    logging.info(f"key phrase extraction output: {json_str}")
+    return extracted_record

adi_function_app/pre_embedding_cleaner.py

@@ -2,13 +2,7 @@
 # Licensed under the MIT License.
 import logging
 import json
-import nltk
 import re
-from nltk.tokenize import word_tokenize
-
-nltk.download("punkt")
-nltk.download("stopwords")
-nltk.download("punkt_tab")
 
 
 def get_section(cleaned_text: str) -> list:
@@ -69,38 +63,28 @@ def clean_text(src_text: str) -> str:
         str: The clean text."""
 
     try:
+        logging.info(f"Input text: {src_text}")
+        if len(src_text) == 0:
+            logging.error("Input text is empty")
+            raise ValueError("Input text is empty")
+
         # Define specific patterns for each tag
         tag_patterns = {
-            "figurecontent": r"<!--.*?FigureContent=(.*?)-->",
+            "figurecontent": r"<!-- FigureContent=(.*?)-->",
             "figure": r"<figure>(.*?)</figure>",
             "figures": r"\(figures/\d+\)(.*?)\(figures/\d+\)",
             "figcaption": r"<figcaption>(.*?)</figcaption>",
         }
         cleaned_text = remove_markdown_tags(src_text, tag_patterns)
 
-        # remove html tags
-        cleaned_text = re.sub(r"<.*?>", "", cleaned_text)
-
-        # Replace newline characters with spaces
-        cleaned_text = re.sub(r"\n", " ", cleaned_text)
-
-        # Replace multiple whitespace characters with a single space
-        cleaned_text = re.sub(r"\s+", " ", cleaned_text)
-
-        # remove stopwords
-        tokens = word_tokenize(cleaned_text, "english")
-        stop_words = nltk.corpus.stopwords.words("english")
-        filtered_tokens = [word for word in tokens if word not in stop_words]
-        cleaned_text = " ".join(filtered_tokens)
-
-        # remove special characters
-        cleaned_text = re.sub(r"[^a-zA-Z\s]", "", cleaned_text)
-
-        # remove extra white spaces
-        cleaned_text = " ".join([word for word in cleaned_text.split()])
+        # Updated regex to keep Unicode letters, punctuation, whitespace, currency symbols, and percentage signs,
+        # while also removing non-printable characters
+        cleaned_text = re.sub(r"[^\p{L}\p{P}\s\p{Sc}%\x20-\x7E]", "", cleaned_text)
 
-        # case normalization
-        cleaned_text = cleaned_text.lower()
+        logging.info(f"Cleaned text: {cleaned_text}")
+        if len(cleaned_text) == 0:
+            logging.error("Cleaned text is empty")
+            raise ValueError("Cleaned text is empty")
     except Exception as e:
         logging.error(f"An error occurred in clean_text: {e}")
         return ""

adi_function_app/requirements.txt

@@ -9,7 +9,7 @@ pandas
 azure-identity
 openpyxl
 regex
-nltk==3.9.1
+tenacity
 bs4
 azure-search
 azure-search-documents