Fix missing space between block and inline sibling elements in JSoupTextExtractor

[nicoscandolo]

Fixes #1925

Problem

When a block-level element (e.g. <h3>) is immediately followed by an inline element (e.g. <a>, <span>), the extracted text concatenates both without any space. For instance, this HTML from a real page (https://www.acai-island.com/contact):

<h3>Email</h3>
<a href="mailto:acaiisland1300@gmail.com">acaiisland1300@gmail.com</a>

Produces Emailacaiisland1300@gmail.com instead of Email acaiisland1300@gmail.com.

Root cause

In tail(), the space after a block element is only appended when nextSibling() instanceof TextNode. This misses cases where the next sibling is an inline Element.

Fix

Changed the condition to nextSibling() != null, so a space is appended after any block element regardless of the sibling type. The existing lastCharIsWhitespace guard prevents duplicate spaces.

Tests

Added two test cases covering block+inline combinations (<h3>+<a> and <div>+<span>).

[rzo1]

Thanks for the fix. Overall, it looks good to me.

The only thing I would like to change is, that tests/comments embed a real small business's website and email address (acaiisland1300@gmail.com, acai-island.com). For an Apache codebase that lives forever, I'd suggest anonymizing to info@example.com / generic text. The issue reference (#1925) already preserves the real-world provenance.

In addition, the existing comment above the condition still reads "a space between block tags and immediately following text nodes" - it could be updated to "following siblings" to match the new behavior.

[dpol1]

LGTM. lastCharIsWhitespace already guards against double spaces so widening to != null is safe. comment + tests good.

[jnioche]

thanks @nicoscandolo and our reviewers