fix(html): escape square brackets and backslashes in image alt text

[bodapatisaikrishna]

Problem

When converting HTML images to Markdown, image alt text is taken directly from the element attribute and formatted as ![<alt>](<src>). In Markdown/CommonMark syntax, the image description is delimited by square brackets [...].

If the alt text contains unescaped square brackets:

1. An unescaped closing bracket (e.g. alt="Photo ] 1") prematurely closes the image description delimiter, resulting in ![Photo ] 1](url) which fails Markdown image parsing.
2. An unbalanced opening bracket (e.g. alt="Photo [1") prevents Markdown parsers from matching the image delimiter or misinterprets subsequent text as a nested link (![Photo [1](url) -> <p>![Photo <a href="url">1</a></p>).
3. Link-like text inside alt (e.g. alt="Click [here](url)") breaks CommonMark image parsing because image descriptions cannot contain nested links.

Solution

In _CustomMarkdownify.convert_img, backslash-escape literal backslashes and square brackets (\ -> \\, [ -> \[, ] -> \]) before assembling the ![<alt>](<src>) Markdown image syntax:

• Preserves the verbatim content when parsed back into HTML.
• Keeps plain inline text conversion unaltered when images are converted as inline text.

Testing

Added regression tests in test_html_converter.py covering:

• Unbalanced closing brackets (])
• Unbalanced opening brackets ([)
• Balanced brackets ([Q3])
• Link-like patterns ([here](url))
• Literal backslashes (\)
• Preserving raw alt text in inline mode