Upload 5 files

README.md

@@ -1,5 +1,17 @@
----
-license: other
-license_name: apglv3
-license_link: https://www.gnu.org/licenses/gpl-3.0.en.html
----
+## What is this
+
+2500+ Screenplays scraped from the scriptlab for science purposes
+
+## Different folders
+
+`pdfs` includes screenplays as PDFs, this is the raw scraped data.
+
+`text2` includes PDFs converted into .txt directly, some pdfs have formatting issues and some are outright blank or have obfuscated text.
+
+`texts3` includes OCRd PDFs which mitigate the previous issues but add issues with slight character or formatting errors (like inconsistent newlines)
+
+Note: Originally we were going to do OCR, however, some morons put image decompression bombs and I can't be bothered to fight that, for now around 1/6 pdfs with DRMs won't be available as .txt format, feel free to convert them to .txt yourself and send a merge request though!
+
+Obviously the TAR files are compressed versions of the folders.
+
+It could be helpful to use an LLM to clean it up! Make sure the LLM doesn't change any of the actual contents. Otherwise I don't have enough time to do it but still want to release this golden trove of data.

convert_pdfs.py

@@ -0,0 +1,46 @@
+import fitz  # PyMuPDF
+import os
+import pytesseract
+from PIL import Image
+from tqdm import tqdm
+
+# Directory containing PDFs
+input_dir = "pdfs"
+# Directory to save text files
+output_dir = "text3"
+
+# Create output directory if it doesn't exist
+os.makedirs(output_dir, exist_ok=True)
+
+def extract_text_from_image(page):
+    # Convert PDF page to an image
+    pix = page.get_pixmap()
+    img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
+    # Use OCR to extract text from the image
+    text = pytesseract.image_to_string(img)
+    return text
+
+# Loop through all PDF files in the input directory
+for pdf_file in tqdm(os.listdir(input_dir)):
+    if pdf_file.endswith(".pdf"):
+        # Open the PDF file
+        pdf_path = os.path.join(input_dir, pdf_file)
+        doc = fitz.open(pdf_path)
+        
+        # Extract text from each page
+        text = ""
+        for page_num in range(len(doc)):
+            page = doc.load_page(page_num)
+            page_text = page.get_text()
+            if not page_text.strip():
+                # If no text is found, try OCR
+                page_text = extract_text_from_image(page)
+            text += page_text
+
+        # Save the extracted text to a .txt file
+        base_name = os.path.splitext(pdf_file)[0]
+        txt_path = os.path.join(output_dir, f"{base_name}.txt")
+        with open(txt_path, "w", encoding="utf-8") as txt_file:
+            txt_file.write(text)
+
+print("Conversion complete!")

ocr_pdfs.py

@@ -0,0 +1,26 @@
+import os
+import subprocess
+from tqdm import tqdm
+
+# Define the directories
+pdfs_folder = "pdfs"
+texts_folder = "texts3"
+ocr_pdfs_folder = "ocr_d_pdfs"
+
+# Create the texts3 folder if it doesn't exist
+if not os.path.exists(texts_folder):
+    os.makedirs(texts_folder)
+
+if not os.path.exists(ocr_pdfs_folder):
+    os.makedirs(ocr_pdfs_folder)
+
+# Iterate over all files in the pdfs folder
+for filename in tqdm(os.listdir(pdfs_folder)):
+    if filename.endswith(".pdf"):
+        pdf_path = os.path.join(pdfs_folder, filename)
+        text_filename = filename.replace(".pdf", ".txt")
+        output_pdf = os.path.join(ocr_pdfs_folder, filename)
+        text_path = os.path.join(texts_folder, text_filename)
+        
+        # Run the ocrmypdf command
+        subprocess.run(["ocrmypdf", pdf_path, output_pdf, "--sidecar", text_path, "--force-ocr"])

remove_dupes.py

@@ -0,0 +1,40 @@
+import os
+import hashlib
+from tqdm import tqdm
+
+def calculate_hash(file_path, chunk_size=1024):
+    """Calculate the MD5 hash of a file."""
+    hash_md5 = hashlib.md5()
+    with open(file_path, "rb") as f:
+        for chunk in iter(lambda: f.read(chunk_size), b""):
+            hash_md5.update(chunk)
+    return hash_md5.hexdigest()
+
+def find_and_remove_duplicates(directory):
+    """Find and remove duplicate files in the given directory."""
+    if not os.path.isdir(directory):
+        print(f"The directory {directory} does not exist.")
+        return
+
+    file_hashes = {}
+    files_to_remove = []
+
+    # Walk through the directory and calculate hashes
+    for root, _, files in tqdm(os.walk(directory)):
+        for file in files:
+            file_path = os.path.join(root, file)
+            file_hash = calculate_hash(file_path)
+            if file_hash in file_hashes:
+                files_to_remove.append(file_path)
+            else:
+                file_hashes[file_hash] = file_path
+
+    # Remove duplicate files with a progress bar
+    for file_path in tqdm(files_to_remove, desc="Removing duplicates"):
+        os.remove(file_path)
+
+    print(f"Removed {len(files_to_remove)} duplicate files.")
+
+if __name__ == "__main__":
+    directory = input("Enter the directory to check for duplicates: ")
+    find_and_remove_duplicates(directory)

tar_command.sh

@@ -0,0 +1 @@
+tar -cavf texts3.tar.zst ./text3