book2text KeyError: 'Title'

[WarmCatUK]

Came across your project when attempting to create a similar tool for summarising novels I have read but forgotten the story. Personally I found that models I tried with would sometimes ignore my specific formatting requests and do their own thing.
My method simply chunked the book after it had been converted to text, but I liked your method of converting an ePub.

Anyway, stuck at the first hurdle.
I've tried with a few books, converted to epub via calibre.

CSV file created: /Users/wayne/Documents/Development/ollama-ebook-summary/out/Book-1---Pawn-of-Prophecy---David-Eddings.csv Traceback (most recent call last): File "/Users/wayne/Documents/Development/ollama-ebook-summary/book2text.py", line 209, in <module> main(input_file, output_dir, output_csv) File "/Users/wayne/Documents/Development/ollama-ebook-summary/book2text.py", line 193, in main process_csv(output_csv) File "/Users/wayne/Documents/Development/ollama-ebook-summary/lib/chunking.py", line 101, in process_csv title = re.sub(r'^[0-9]+-', '', row['Title']) ~~~^^^^^^^^^ KeyError: 'Title'

[uvwxy]

Stumbled across the same, changed the capital Title and Text to title and text respectively:

index 4c5ee7e..7ed3ed6 100644
--- a/lib/chunking.py
+++ b/lib/chunking.py
@@ -98,8 +98,8 @@ def process_csv(input_file):
             previous_chunk = None
             
             for row in reader:
-                title = re.sub(r'^[0-9]+-', '', row['Title'])
-                text = row['Text']
+                title = re.sub(r'^[0-9]+-', '', row['title'])
+                text = row['text']
                 char_count = len(text)
 
                 # If text is empty, store title and continue

[cognitivetech]

thanks @uvwxy feel free to send a pull-request, I thought I fixed this... due to using many different iterations of this project from home