Fix incorrect grouping when identical text nodes occur

[nquenouille]

Problem: When processing continued elements, tag spans could extend too far if identical text content appeared multiple times (e.g. repeated words).

Cause: The code used index-of(current-group(), $final) to determine the position of a node within a group.
However, index-of() compares atomic values, not node identity.
As a result, multiple nodes with identical string values were treated as equal, and the last matching position was selected.

Solution: Replaced the index-of()-based lookup with a node identity comparison using "is", ensuring that the correct node instance is selected even when text content is identical.

Result:

• Correct closing of tag spans
• No unintended merging across identical text nodes
• Stable behavior regardless of duplicate string values

This change avoids value-based comparison for node sequences and ensures consistent behavior when duplicate text nodes are present.

With many thanks!