Remove PyMuPDF

conda/environments/all_cuda-121_arch-x86_64.yaml

@@ -120,7 +120,6 @@ dependencies:
 - pip:
   - --find-links https://data.dgl.ai/wheels-test/repo.html
   - --find-links https://data.dgl.ai/wheels/cu121/repo.html
-  - PyMuPDF==1.23.*
   - databricks-cli < 0.100
   - databricks-connect
   - dgl==2.0.0

conda/environments/dev_cuda-121_arch-x86_64.yaml

@@ -98,7 +98,6 @@ dependencies:
 - yapf=0.40.1
 - zlib=1.2.13
 - pip:
-  - PyMuPDF==1.23.*
   - databricks-cli < 0.100
   - databricks-connect
   - milvus==2.3.5

conda/environments/examples_cuda-121_arch-x86_64.yaml

@@ -67,7 +67,6 @@ dependencies:
 - pip:
   - --find-links https://data.dgl.ai/wheels-test/repo.html
   - --find-links https://data.dgl.ai/wheels/cu121/repo.html
-  - PyMuPDF==1.23.*
   - databricks-cli < 0.100
   - databricks-connect
   - dgl==2.0.0

dependencies.yaml

@@ -371,7 +371,6 @@ dependencies:
           - &python-docx python-docx==1.1.0
           - pip
           - pip:
-              - &PyMuPDF PyMuPDF==1.23.*
               - pytest-kafka==0.6.0
 
   example-dfp-prod:
@@ -420,7 +419,6 @@ dependencies:
             - faiss-gpu==1.7.*
             - google-search-results==2.4
             - nemollm==0.3.5
-            - *PyMuPDF
 
   model-training-tuning:
     common:

examples/llm/vdb_upload/module/content_extractor_module.py

@@ -22,7 +22,6 @@
 from typing import Dict
 from typing import List
 
-import fitz
 import fsspec
 import mrc
 import mrc.core.operators as ops
@@ -171,12 +170,7 @@ def wrapper(input_info: ConverterInputInfo, *args, **kwargs):
 
 @_converter_error_handler
 def _pdf_to_text_converter(input_info: ConverterInputInfo) -> str:
-    text = ""
-    pdf_document = fitz.open(stream=input_info.io_bytes, filetype="pdf")
-    for page_num in range(pdf_document.page_count):
-        page = pdf_document[page_num]
-        text += page.get_text()
-    return text
+    raise NotImplementedError("PDF to text conversion is not implemented.")
 
 
 @_converter_error_handler