`strip` option preserves inner text of removed tags — request for content-removing option

[Taeknology]

Environment

• markdownify: 1.2.2
• Python: 3.12.3

Current Behavior

strip=["script"] removes the tag but preserves its inner text as plain text:

```python
from markdownify import markdownify as md

html = '

Hello

<script>alert("js noise")</script>

World

'
print(md(html, strip=["script"]))

Hello

alert("js noise") ← script content remains as plain text

World

```

Root Cause

process_tag() collects and joins all child text before checking
should_convert_tag(). By the time get_conv_fn_cached() returns None
for a stripped tag, the child text has already been assembled:

```python
def process_tag(self, node, ...):
# Children processed unconditionally — strip has no effect here
child_strings = [self.process_element(el, ...) for el in children_to_convert]
text = ''.join(child_strings)

# should_convert_tag() checked only here — too late to suppress children
convert_fn = self.get_conv_fn_cached(node.name)  # returns None for stripped tags
if convert_fn is not None:
    text = convert_fn(node, text, ...)

return text  # child text always returned

```

Proposed Solution

Add a decompose parameter that suppresses child processing entirely:

```python
md(html, decompose=["script", "style", "noscript"])

→ Hello\n\nWorld

```

Implementation: early return '' in process_tag when the tag is in the
decompose list.

Workaround

```python
from bs4 import BeautifulSoup
from markdownify import markdownify as md

soup = BeautifulSoup(html, "html.parser")
for tag in soup(["script", "style", "noscript"]):
tag.decompose()
md(str(soup), heading_style="ATX")
```

Use Case

This affects web scraping pipelines where <script> and <style> content
is meaningless as plain text — analytics snippets, minified JS bundles, and
inline CSS all leak into the markdown output when using strip.