[Emily Felix] Cuckold, Size, NTR

[fantafan]

Note: The 2nd novel is too new and my process is broken so I had to manually extract the text which was a giant pain. Also explains formatting and maybe the occaisional mistakes.

 

[KidneyCracker]

What file format was the original? original amazon format?

for anyone that doesn't know, epubs are zip files with a different file extension. If you rename the file from book.epub to book.zip, you can open it in any zip program. Within them, you will find some html files, either in the root folder or a sub-folder, and one or more of those will contain the story text, extract those htmil files into a windows folder and open them in a web browser. Then you can copy the txt using Ctrl-A, Ctrl-C.

 

[fantafan]

What file format was the original? original amazon format?

for anyone that doesn't know, epubs are zip files with a different file extension. If you rename the file from book.epub to book.zip, you can open it in any zip program. Within them, you will find some html files, either in the root folder or a sub-folder, and one or more of those will contain the story text, extract those htmil files into a windows folder and open them in a web browser. Then you can copy the txt using Ctrl-A, Ctrl-C.

Yes, Amazon Kindle. The newest releases won't download with Kindle 2.4.0 which is what you need to use with ePubOr at least, which is what I've used to crack the encryption. No solution as of yet.

 

[SmeeJ]

Yes, Amazon Kindle. The newest releases won't download with Kindle 2.4.0 which is what you need to use with ePubOr at least, which is what I've used to crack the encryption. No solution as of yet.

Hope one comes soon. I was generally okay paying for books off Amazon, and then converting them, but I'm less likely to buy now with the extreme DRM. My best solution for converting them now is OCR - take screen shots and copy out the text from each page. Very slow going so only doing this for my favourite authors.

 

[O_P_T]

Yes, Amazon Kindle. The newest releases won't download with Kindle 2.4.0 which is what you need to use with ePubOr at least, which is what I've used to crack the encryption. No solution as of yet.

Yeah the change was on ~ May 1st.

I e-mailed the epubor people and this is what he said.

"it will take months to decrypt v2.7,
if you have kindle device, then you download book in kindle device, we can remove kindle book drm."

Not sure about the last part, since Amazon supposedly disabled the download to kindle back in February. See this thread on the topic.

I don't have a kindle, so I can't try that.

 

[fantafan]

Hope one comes soon. I was generally okay paying for books off Amazon, and then converting them, but I'm less likely to buy now with the extreme DRM. My best solution for converting them now is OCR - take screen shots and copy out the text from each page. Very slow going so only doing this for my favourite authors.

Yup. I'm using SnagIt for now, and its pretty painful. No screenshot required, just drag a square and it OCRs the text and puts it in the clipboard. Does a really good job. Paid program.

 

[fantafan]

Would love a better solution. I'm currently going page by page through a long story on reamstories that I couldn't find elsewhere and it sucks. Tried the obvious--print to pdf, look at web source, etc. OCR page by page is my current solution. I would *never* subscribe to a reamstories thing again given the limitations.

 

[KidneyCracker]

Microsoft has a powertoy named TextExtractor that OCRs screengrabs too. You'd have to test it to compare how reliable it is compared to SnagIt.

I'd like an efficient way to programmatically bypass their "click for more" nonsense and scrape StoriesOnline.net.

 

[fantafan]

Its "click for more" ... "click for more"

 

[O_P_T]

Microsoft has a powertoy named TextExtractor that OCRs screengrabs too. You'd have to test it to compare how reliable it is compared to SnagIt.

I'd like an efficient way to programmatically bypass their "click for more" nonsense and scrape StoriesOnline.net.

I just tried the textExtractor widget. Turns out I'd already installed it with the powertoy package I got.

Pretty simple to use press "win" "shift" and "T" at the same time. drag the cursor to select the text and it's put on your clipboard.

I opened a preview of a book on Amazon and grabbed a few pages and pasted it into Word.

It seemed to get the text correct, but there were two issues that make it less than optimum.

It does not recognize any formatting of the text (e.g., italicizing), so you would loose that.

Second, it put a paragraph mark at the end of each line as it appears on the screen. No extra paragraph marks to indicate a new paragraph on the original text. So, it's going to take a bit of editing to fix that.

I've seen similar issues with some text files from ASSTR, when it was up, in the past, and I haven't figured out a simple way to work around that.

Edit: It will prove useful when I try to figure out what was the original work that was plagiarized by one of the usual suspects, since I'll have text to copy and paste into the multiple site google search trick.

 

[fantafan]

Yeah similar issues with the SnagIt text scan. Click to scan, drag rectangle, wait, paste, then adjust the text per personal preference. Does a reasonable job of not putting carriage returns in incorrectly, but still generally needs some cleanup. Yes you lose formatting like italics. Italics -> *text* would be a nice option but not holding my breath. Very very few mistakes though, even when I zoom out which makes the font very small.

 

[yano2mch]

It does not recognize any formatting of the text (e.g., italicizing), so you would loose that.

Second, it put a paragraph mark at the end of each line as it appears on the screen. No extra paragraph marks to indicate a new paragraph on the original text. So, it's going to take a bit of editing to fix that.

I've seen similar issues with some text files from ASSTR, when it was up, in the past, and I haven't figured out a simple way to work around that.

Not the worst problems to have. Formatting, yeah go side-by-side and try to add formatting back in, that's what i tend to do with my OCR stuff. It usually stands out so it's easy to identify and do. Though i tend to do raw text so i have to add the whole tag ex: <i>italics</i>. Though some custom tags like first letter and header lines have special rules.

As for new or lacking paragraph markers, text length and regex can do probably 80% of the work. ASSRT and newsgroup based stuff had the usual 80 character lines, so lines that were short (say 50 characters or less) usually indicated new paragraphs and you could append newlines. Also the presence or lack of punctuation would specify continuation of a line.

On my OCR stuff i'd tend to have all lines append together and during the scan phase i would glance at the original and then the OCR side-by-side and make sure the paragraphs were single blocks, then let a couple scripts i made do the heavy work.

 

[O_P_T]

Forgot to mention that there's an easy fix for any hyphenation at the end of line, in the on screen version.

I simply did a global search and replace for "-^p" and replaced it with nothing.

 

[KidneyCracker]

notepad++ supports most regex features and you can ask Microsoft's AI co-pilot in MS Edge for help on some pretty funky search and replace schemes. use that to create a macro if you find you have to do the repair frequently.

 

[yano2mch]

notepad++ supports most regex features and you can ask Microsoft's AI co-pilot in MS Edge for help on some pretty funky search and replace schemes. use that to create a macro if you find you have to do the repair frequently.

A lot of programs use PCRE (Perl Compatible Regular Expressions). The search/replace isn't big to do anything too complex, knowing your escape characters and how to use back references greatly helps in a lot of cases. Regular Expressions as a basic skill/tool is pretty easy to learn.

Depending on complexity and how many you need to do, making a series of commands and using sed is probably better. Sed is less complex regex-wise but being able to make a script offsets that for nearly all uses.

Spoiler: sed - stage2 for OCR work

H;
    ${x;
        s:\r::g;
        #hypenated endings...
        s:([a-zA-Z])-\n([a-zA-Z]):\1\2:g;
        
        #Everything else...
        s:([^\n])\n([^\n]):\1 \2:g;
        
        #k basic formatting to fix... Modify to preference.
        s:-?—-?: — :g;
        s:  +: :g;
        s:[“”]:":g;
#        s:[“”]:":g;
#        s:‘:":g;
#        s:’\B:":g;
        s:[‘’]:'"'"':g;
#        s:’:'"'"':g;
        s:©:\©:g
        
        #add html formatting
        s:\n([^\n]+):\n<p>\1</p>:g
        s:\n<p>-----</p>:\n<hr/>:g
        s:\n<p>(C[Hh][Aa][Pp][Tt][Ee][Rr] [a-zA-Z_0-9.\: ?'"'"'-]+|P[Rr][Oo][a-zA-Z]+|E[Pp][Ii][Ll][a-zA-Z]+|ABOUT THE AUTHOR)</p>:\n<center><h3>\1</h3></center>:g
        #close em-phesis tags  <em>test<em> to <em>test</em> (then i dont have to close them manually)
        s:<em>([^\n<>]+)<em>:<em>\1</em>:g
        s:<strong>([^\n<>]+)<strong>:<strong>\1</strong>:g
        
        #forceful header spots instead of chapters
        s:<H([1-9])>([^\n<>]+):<h\1>\2</h\1>:g
        s:<p><h:<center><h:g
        s:(</h[1-9]>)</p>:\1</center>:g
        
        #first line is likely title and author... prep it.
        s:^\n+::
        s:^<p>([^\n]+)</p>\n:<center><h1>\1</h1></center>\n:
        
        #first letter dropcap classes...
        s:<FL>(["'"'"']?\w):<div class="first-letter">\1</div>:g
        
        #super/subscript, extends to the end of the word.
        s:<(su[bp])>(\w+):<\1>\2</\1>:g
        
        #email and link hyperlink
        s:\b(https?\://[^<> \t\r\n]+):<a href="\1">\1</a>:g
        s:\b(\w+@\w+\.\w{3}):<a href="mailto\:\1">\1</a>:g
        
        #common fixes
        s:\blll\b:I'"'"'ll:g
        s:\blI\b:I:g
        s:\bl('"'"')(ll|[dm])\b:I\1\2:g
        s:\bhomy\b:horny:g
        
    p
    }

 

[7hR337R0723]

Its "click for more" ... "click for more"

The Python code below grabs a screenshot and clicks the "next page" button in a loop (you tell it how many pages you want). Doesn't do the OCR part but that can be done later. You click to let it know the region to grab and the location of the next button.

import pyautogui
import time
import os

# --- Configuration (You might adjust these defaults) ---
DEFAULT_OUTPUT_FOLDER = "ocr_screenshots"
DEFAULT_PAGE_LOAD_DELAY = 2  # Seconds to wait for the next page to load
DEFAULT_ACTION_DELAY = 0.5 # Seconds to wait between mouse actions

# --- Helper function to get coordinates ---
def get_coordinates(prompt_message):
    """
    Prompts the user to move their mouse to a position and press Enter
    to capture the coordinates.
    """
    input(f"{prompt_message}. Move your mouse to the desired position and press Enter...")
    x, y = pyautogui.position()
    print(f"Captured coordinates: X={x}, Y={y}\n")
    return x, y

# --- Main Script ---
def main():
    print("Welcome to the Automated Screenshotter for OCR!")
    print("----------------------------------------------")

    # pyautogui.FAILSAFE = True # Optional: Move mouse to top-left corner to abort

    # 1. Get user input
    num_pages = int(input("How many pages do you want to capture? "))
    output_folder = input(f"Enter output folder name (default: {DEFAULT_OUTPUT_FOLDER}): ") or DEFAULT_OUTPUT_FOLDER
    page_load_delay = float(input(f"Enter delay after clicking 'Next' (seconds, default: {DEFAULT_PAGE_LOAD_DELAY}): ") or DEFAULT_PAGE_LOAD_DELAY)
    action_delay = float(input(f"Enter delay between mouse actions (seconds, default: {DEFAULT_ACTION_DELAY}): ") or DEFAULT_ACTION_DELAY)

    print("\n--- Coordinate Setup ---")
    print("You will now be asked to identify key screen locations.")
    print("Make sure the old program window is visible and in the position it will be during capture.")

    # 2. Get screenshot region coordinates
    print("\nStep 1: Define Screenshot Area")
    x1, y1 = get_coordinates("Identify the TOP-LEFT corner of the content area to screenshot")
    x2, y2 = get_coordinates("Identify the BOTTOM-RIGHT corner of the content area to screenshot")

    # Calculate width and height for the screenshot region
    screenshot_width = x2 - x1
    screenshot_height = y2 - y1

    if screenshot_width <= 0 or screenshot_height <= 0:
        print("Error: Screenshot region coordinates are invalid (e.g., top-left is to the right of bottom-right).")
        return

    # 3. Get "Next Page" button coordinates
    print("\nStep 2: Identify 'Next Page' Button")
    button_x, button_y = get_coordinates("Identify the CENTER of the 'Next Page' button")

    # 4. Create output folder if it doesn't exist
    os.makedirs(output_folder, exist_ok=True)
    print(f"\nScreenshots will be saved in: '{os.path.abspath(output_folder)}'")

    print("\n--- Starting Capture Process ---")
    print("You have 5 seconds to switch to the old program and ensure the first page is visible.")
    print("DO NOT move the mouse or use the keyboard during the capture process.")
    time.sleep(5)

    for i in range(num_pages):
        current_page_num = i + 1
        print(f"Processing page {current_page_num} of {num_pages}...")

        # a. Take screenshot
        try:
            # Region is defined by (left, top, width, height)
            screenshot = pyautogui.screenshot(region=(x1, y1, screenshot_width, screenshot_height))
            time.sleep(action_delay) # Brief pause
        except Exception as e:
            print(f"Error taking screenshot for page {current_page_num}: {e}")
            print("Aborting.")
            return

        # b. Save screenshot
        filename = os.path.join(output_folder, f"page_{current_page_num:04d}.png") # e.g., page_0001.png
        try:
            screenshot.save(filename)
            print(f"  Saved: {filename}")
        except Exception as e:
            print(f"Error saving screenshot {filename}: {e}")
            # Decide if you want to continue or abort
            # continue

        # c. Click "Next Page" button (if not the last page)
        if current_page_num < num_pages:
            try:
                pyautogui.moveTo(button_x, button_y, duration=0.2) # Move mouse smoothly
                time.sleep(action_delay)
                pyautogui.click()
                print(f"  Clicked 'Next Page' button.")
            except Exception as e:
                print(f"Error clicking 'Next Page' button: {e}")
                print("Aborting.")
                return

            # d. Wait for the next page to load
            print(f"  Waiting {page_load_delay} seconds for page to load...")
            time.sleep(page_load_delay)
        else:
            print("  Last page processed.")

    print("\n--- Capture Complete! ---")
    print(f"All {num_pages} pages captured and saved to '{output_folder}'.")

if __name__ == "__main__":
    try:
        main()
    except KeyboardInterrupt:
        print("\nProcess interrupted by user. Exiting.")
    except Exception as e:
        print(f"\nAn unexpected error occurred: {e}")

 

[drdoom39]

Some more books:

 

[fantafan]

The remainder that can be downloaded

 

[KidneyCracker]

cuckolded on camera 3 is available for free on bookfunnel. I can't get it because my account is locked. just beware that the epub is probably seeded so it identifies who shared it.

 

[fantafan]

Okay, I'm not clear how to get a link for the book or even the author from bookfunnel. Wanna share? I created a reader account...