Viewer cleanup

README.md

@@ -61,6 +61,8 @@ You can also bulk convert many PDFS with a glob pattern:
 python -m olmocr.pipeline ./localworkspace --pdfs tests/gnarly_pdfs/*.pdf
 ```
 
+#### Viewing Results
+
 Once that finishes, output is stored as [Dolma](https://github.com/allenai/dolma)-style JSONL inside of the `./localworkspace/results` directory.
 
 ```bash

olmocr/viewer/dolmaviewer.py

@@ -11,20 +11,14 @@
 from concurrent.futures import ThreadPoolExecutor, as_completed
 import markdown2
 
-from olmocr.s3_utils import get_s3_bytes
+from olmocr.s3_utils import get_s3_bytes, parse_s3_path
 from olmocr.data.renderpdf import render_pdf_to_base64webp
 
 def read_jsonl(path):
     with smart_open.smart_open(path, 'r', encoding='utf-8') as f:
         for line in f:
             yield line.strip()
 
-def parse_s3_path(path):
-    # s3://bucket_name/key_name
-    path = path[5:]  # Remove 's3://'
-    bucket_name, key_name = path.split('/', 1)
-    return bucket_name, key_name
-
 def generate_presigned_url(s3_client, bucket_name, key_name):
     try:
         response = s3_client.generate_presigned_url('get_object',