优化autoFrontmatter代码

autoFrontmatter 是用来为文章自动生成frontMatter的，可以更好让我们的文章链接固定。
其实Vdoing主题本身是有自动生成frontMatter功能的，但是需要在dev或者build的时候生效，但是我们所有的编译交予gihub的自动工作流程，因此我写了这个脚本在执行deploy的时候生成frontMatter.

以前的代码比较粗糙，这次借着优化博客仓库的机会顺便将此代码优化一下。

优化内容：

• 一些功能行函数独立出来
• 引入正则表达式

详细代码如下：

import contextlib
import os
import re
import random
import time
import glob

# 作者名及主页链接
AUTHOR = 'Shanya'
LINK = 'https://github.com/shanyaliux'


# 判断一个字符串是否为数字
def isNumber(s):
    with contextlib.suppress(ValueError):
        float(s)
        return True
    with contextlib.suppress(TypeError, ValueError):
        import unicodedata
        unicodedata.numeric(s)
        return True
    return False


# 生成一个新的permalink
def getNewPermalink():
    return hex(random.randint(1118481, 16777215))[2:]


# 得到所有的已有的permalink
def getExistsPermalinks(docsList):
    permalinks = []
    for doc in docsList:
        fileName = doc.split(os.sep)[-1]

        if doc.split(os.sep)[1] != '_posts' and (not isNumber(fileName.split('.')[0]) or fileName.split('.')[-2] == 'friends'):
            continue

        with open(doc, encoding="utf-8", errors='ignore') as f:

            content = f.read()
            result = re.finditer(r'---+', content)

            positions = [i.span() for i in result]

            if not positions:
                continue

            temp = content[positions[0][1] + 1: positions[0][1] + 6]
            if temp != 'title':
                continue

            frontMatter = content[positions[0][1]: positions[1][0]]
            permalinkSpan = re.search(r'permalink', frontMatter).span()
            categoriesSpan = re.search(r'categories', frontMatter).span()
            permalink = frontMatter[permalinkSpan[1] + 1: categoriesSpan[0] - 1]
            permalink = permalink.strip('/').split('/')[-1]
            permalinks.append(permalink)
    return permalinks


# 生成新的FrontMatter
def getFrontMatter(title, ct, permalink):
    date = time.strftime('%Y-%m-%d %H:%M:%S', time.localtime())

    frontMatter = '---\n' + f'title: {title}\n'
    frontMatter += f'date: {date}\n'
    frontMatter += f'permalink: /pages/{permalink}/\n'
    frontMatter += f'categories: \n - {ct}\n'
    frontMatter += f'tags: \n - {ct}\n'
    frontMatter += f'author: \n name: {AUTHOR}\n link: {LINK}\n'
    frontMatter += '---\n\n'

    return frontMatter


if __name__=="__main__":

    docsPath = r'docs'
    docsList = glob.glob(f'{docsPath}/**/*.md', recursive=True)

    permalinks = getExistsPermalinks(docsList)

    for doc in docsList:
        fileName = doc.split(os.sep)[-1]

        if doc.split(os.sep)[1] != '_posts' and (not isNumber(fileName.split('.')[0]) or fileName.split('.')[-2] == 'friends'):
            continue

        with open(doc, 'r+', encoding="utf-8", errors='ignore') as f:

            content = f.read()
            result = re.finditer(r'---+', content)
            if positions := [i.span() for i in result]:
                temp = content[positions[0][1] + 1: positions[0][1] + 6]
                if temp == 'title':
                    continue

            print(doc)

            if doc.split(os.sep)[1] != '_posts':
                ct = doc.split(os.sep)[1].split('.')[1]
                title = fileName.split('.')[1]
            else:
                ct = doc.split(os.sep)[2]
                title = doc.split(os.sep)[-1].split('.')[0]
            permalink = getNewPermalink()
            while permalink in permalinks:
                permalink = getNewPermalink()
            frontMatter = getFrontMatter(title, ct, permalink)
            f.seek(0, 0)
            f.write(frontMatter + content)

    print('done!')