Boom. It works.

modified:   Dockerfile
	modified:   app.py

Dockerfile

@@ -31,6 +31,11 @@ COPY . .
 
 RUN pip install --no-cache-dir -r requirements.lock
 
+RUN playwright install --with-deps && \
+    python -c "from unstructured.nlp.tokenize import download_nltk_packages; download_nltk_packages()" && \
+    python -c "import nltk;nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger_eng')" && \
+    python -c "from unstructured.partition.model_init import initialize; initialize()"
+
 EXPOSE 8001
 
 CMD ["python", "app.py"]

app.py

@@ -35,19 +35,17 @@ from langchain_openai import ChatOpenAI
 from llama_index.core.schema import Document as LlamaDocument
 from llama_parse import LlamaParse as _LlamaParse
 from llama_parse.utils import Language, ResultType
-#from megaparse_sdk.schema.languages import Language as MP_Language
 from pydantic import BaseModel, Field, ValidationError, validator
 
+from unstructured.partition.auto import partition
+
 
 
 # Load environment variables
 load_dotenv()
 
 
-# ---------------------- Langauges --------------------------
-
-from enum import Enum
-
+# ---------------------- Langs --------------------------
 
 class Language(str, Enum):
     BAZA = "abq"
@@ -136,6 +134,26 @@ class Language(str, Enum):
 
 MP_Language = Language
 
+class FileExtension(str, Enum):
+    """Supported file extension enumeration."""
+
+    PDF = ".pdf"
+    DOCX = ".docx"
+    DOC = ".doc"
+    TXT = ".txt"
+    OTF = ".odt"
+    EPUB = ".epub"
+    HTML = ".html"
+    XML = ".xml"
+    CSV = ".csv"
+    XLSX = ".xlsx"
+    XLS = ".xls"
+    PPTX = ".pptx"
+    PPT = ".ppt"
+    JSON = ".json"
+    MD = ".md"
+    MARKDOWN = ".markdown"
+
 # --------------------- Exception Classes ---------------------
 
 class HTTPModelNotSupported(HTTPException):