脚本-字体woff做映射表

Copy Highlighter-hljs
from fontTools.ttLib import TTFont
import ddddocr
from io import BytesIO
from PIL import Image, ImageDraw, ImageFont
 
 
def convert_cmap_to_image(cmap_code, font_path):
    img_size = 1024
    img = Image.new("1", (img_size, img_size), 255)  # 创建一个黑白图像对象
    draw = ImageDraw.Draw(img)  # 创建图对象
    font = ImageFont.truetype(font_path, img_size)  # 加载字体文件
    character = chr(cmap_code)  # 将 cmap code转换为字符
    bbox = draw.textbbox((0, 0), character, font=font)  # 获取文本在图像中的边界框，bbox获取的是四个值
    width = bbox[2] - bbox[0]  # 文本宽度
    height = bbox[3] - bbox[1]  # 文本高度
    draw.text(((img_size - width) // 2, (img_size - height) // 2), character, font=font)  # 绘制文本，居中显示
    return img
 
 
def extract_text_from_font(font_path):
    font = TTFont(font_path)  # 加载字体文件
    # font.saveXML("font.xml")  # 保存XML文件(可以这么操作，但是这里我们不用)
    ocr = ddddocr.DdddOcr(beta=True, show_ad=False)  # 实例化
    font_map = {}
    for cmap_code, glyph in font['cmap'].getBestCmap().items():
        image = convert_cmap_to_image(cmap_code, font_path)
        bytes_io = BytesIO()
        image.save(bytes_io, format='PNG')
        # 识别图像
        text = ocr.classification(bytes_io.getvalue()) 
        # 生成对应名字的图片
        # image.save(f'{text}.png', format='PNG')
        # 最后生成映射表
        # font_map[cmap_code] = text
        font_map[glyph.replace("uni", "&#x").lower()] = text # 根据案例来实现的
    return font_map
 
 
# font_file_path = "./font.woff"
# print(extract_text_from_font(font_file_path))