Extract Myanmar text from PDF. This script will convert Win Innwa to Unicode using Burglish

extract-myanmar-text.py

import os
import time

import fitz  # PyMuPDF
from docx import Document
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.chrome.service import Service
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.common.by import By
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.support.ui import WebDriverWait
from webdriver_manager.chrome import ChromeDriverManager

driver = webdriver.Chrome(service=Service(ChromeDriverManager().install()))

def set_font_selection(driver, panel_type, font_name):
    """Set font for either source (input) or target (output) panel"""
    # Determine which dropdown to click
    dropdown_selector = "ol#source > li > p" if panel_type == "source" else "ol#target > li > p"

    # Click the dropdown
    dropdown = driver.find_element(By.CSS_SELECTOR, dropdown_selector)
    ActionChains(driver).move_to_element(dropdown).click().perform()
    time.sleep(0.5)  # Allow dropdown animation

    # IMPORTANT: xpath is required to locate the element
    # Select the font
    font_xpath = f"""
    //ol[@id='{panel_type}']//li[
        @id='{font_name}' or
        @data-detail='{font_name}' or
        text()='{font_name}'
    ]
    """
    font_element = WebDriverWait(driver, 10).until(
        EC.presence_of_element_located((By.XPATH, font_xpath)))
    font_element.click()
    time.sleep(0.3)

def trigger_conversion(driver):
    """Click the convert button"""
    switch_btn = driver.find_element(By.CSS_SELECTOR, "#targetOption > .icon-convert-normal")
    ActionChains(driver).move_to_element(switch_btn).click().perform()
    time.sleep(0.5)  # Allow conversion to complete

def convert_text(driver, text):
    """Input text and get converted result"""
    # Clear and input text
    input_area = driver.find_element(By.ID, "sourceText")
    input_area.clear()
    driver.execute_script("arguments[0].value = arguments[1];", input_area, text)

    # Trigger conversion
    trigger_conversion(driver)

    # Get output
    return driver.find_element(By.ID, "targetText").get_attribute("value")

def auto_convert_pdf(pdf_path, output_docx):
    # 1. Extract text from PDF
    pdf_texts = []
    with fitz.open(pdf_path) as pdf:
        for page in pdf:
            pdf_texts.append(page.get_text())

    # 2. Configure Selenium
    options = Options()
    options.add_argument("--headless=new")
    options.add_argument("--disable-gpu")
    options.add_argument("--no-sandbox")
    options.add_argument("--disable-extensions")
    options.add_argument("--blink-settings=imagesEnabled=false")
    driver = webdriver.Chrome(options=options)
    driver.get("https://scriptive.github.io/burglish/")
    driver.maximize_window()

    # 3. Set conversion options
    set_font_selection(driver, "source", "WinInnwa")  # Input font
    set_font_selection(driver, "target", "Myanmar3")  # Output font

    # 4. Process each page
    doc = Document()
    for i, text in enumerate(pdf_texts):
        # Input text
        converted_text = convert_text(driver, text)

        # Add to DOCX
        doc.add_paragraph(f"Page {i+1}")
        doc.add_paragraph(converted_text)
        if i < len(pdf_texts) - 1:
            doc.add_page_break()

    # 5. Cleanup
    driver.quit()
    doc.save(output_docx)
    print(f"Converted DOCX saved to {os.path.abspath(output_docx)}")

# Usage
auto_convert_pdf("../input/body.pdf", "../output/body.docx")