Centralize default parser as DEFAULT_PDF_PARSER constant

pageindex/client.py

@@ -8,7 +8,7 @@ from pathlib import Path
 from .page_index import page_index
 from .page_index_md import md_to_tree
 from .retrieve import get_document, get_document_structure, get_page_content
-from .utils import ConfigLoader, read_pdf_pages, remove_fields
+from .utils import ConfigLoader, DEFAULT_PDF_PARSER, read_pdf_pages, remove_fields
 
 META_INDEX = "_meta.json"
 
@@ -31,7 +31,7 @@ class PageIndexClient:
     For agent-based QA, see examples/agentic_vectorless_rag_demo.py.
     """
     def __init__(self, api_key: str = None, model: str = None, retrieve_model: str = None,
-                 workspace: str = None, pdf_parser: str = "PyPDF2"):
+                 workspace: str = None, pdf_parser: str = DEFAULT_PDF_PARSER):
         if api_key:
             os.environ["OPENAI_API_KEY"] = api_key
         elif not os.getenv("OPENAI_API_KEY") and os.getenv("CHATGPT_API_KEY"):

pageindex/page_index.py

@@ -1063,7 +1063,7 @@ async def tree_parser(page_list, opt, doc=None, logger=None):
     return toc_tree
 
 
-def page_index_main(doc, opt=None, pdf_parser="PyPDF2"):
+def page_index_main(doc, opt=None, pdf_parser=DEFAULT_PDF_PARSER):
     logger = JsonLogger(doc)
 
     is_valid_pdf = (
@@ -1112,7 +1112,7 @@ def page_index_main(doc, opt=None, pdf_parser="PyPDF2"):
 
 def page_index(doc, model=None, toc_check_page_num=None, max_page_num_each_node=None, max_token_num_each_node=None,
                if_add_node_id=None, if_add_node_summary=None, if_add_doc_description=None, if_add_node_text=None,
-               pdf_parser="PyPDF2"):
+               pdf_parser=DEFAULT_PDF_PARSER):
 
     user_opt = {
         arg: value for arg, value in locals().items()

pageindex/retrieve.py

@@ -1,9 +1,9 @@
 import json
 
 try:
-    from .utils import get_number_of_pages, read_pdf_pages, remove_fields
+    from .utils import DEFAULT_PDF_PARSER, get_number_of_pages, read_pdf_pages, remove_fields
 except ImportError:
-    from utils import get_number_of_pages, read_pdf_pages, remove_fields
+    from utils import DEFAULT_PDF_PARSER, get_number_of_pages, read_pdf_pages, remove_fields
 
 
 # ── Helpers ──────────────────────────────────────────────────────────────────
@@ -41,7 +41,7 @@ def _get_pdf_page_content(doc_info: dict, page_nums: list[int]) -> list[dict]:
             {'page': p, 'content': page_map[p]}
             for p in page_nums if p in page_map
         ]
-    parser = doc_info.get('pdf_parser') or 'PyPDF2'
+    parser = doc_info.get('pdf_parser') or DEFAULT_PDF_PARSER
     all_pages = read_pdf_pages(doc_info['path'], pdf_parser=parser)
     total = len(all_pages)
     valid_pages = [p for p in page_nums if 1 <= p <= total]

pageindex/utils.py

@@ -385,9 +385,10 @@ def add_preface_if_needed(data):
 
 
 SUPPORTED_PDF_PARSERS = ("PyPDF2", "pypdfium2", "PyMuPDF")
+DEFAULT_PDF_PARSER = SUPPORTED_PDF_PARSERS[0]
 
 
-def read_pdf_pages(doc, pdf_parser="PyPDF2"):
+def read_pdf_pages(doc, pdf_parser=DEFAULT_PDF_PARSER):
     """Return a list of per-page text strings using the selected parser.
 
     `doc` may be a file path (str/Path) or a BytesIO. `pdf_parser` is one of
@@ -438,7 +439,7 @@ def read_pdf_pages(doc, pdf_parser="PyPDF2"):
     )
 
 
-def get_page_tokens(pdf_path, model=None, pdf_parser="PyPDF2"):
+def get_page_tokens(pdf_path, model=None, pdf_parser=DEFAULT_PDF_PARSER):
     pages = read_pdf_pages(pdf_path, pdf_parser=pdf_parser)
     return [(text, litellm.token_counter(model=model, text=text)) for text in pages]
 

run_pageindex.py

@@ -3,7 +3,7 @@ import os
 import json
 from pageindex import *
 from pageindex.page_index_md import md_to_tree
-from pageindex.utils import ConfigLoader
+from pageindex.utils import ConfigLoader, DEFAULT_PDF_PARSER
 
 if __name__ == "__main__":
     # Set up argument parser
@@ -67,7 +67,7 @@ if __name__ == "__main__":
         opt = ConfigLoader().load({k: v for k, v in user_opt.items() if v is not None})
 
         # Process the PDF
-        toc_with_page_number = page_index_main(args.pdf_path, opt, pdf_parser=args.pdf_parser or "PyPDF2")
+        toc_with_page_number = page_index_main(args.pdf_path, opt, pdf_parser=args.pdf_parser or DEFAULT_PDF_PARSER)
         print('Parsing done, saving to file...')
         
         # Save results