text-splitters: fix stale header metadata in ExperimentalMarkdownSyntaxTextSplitter (#31622)

**Description:**

Previously, when transitioning from a deeper Markdown header (e.g., ###)
to a shallower one (e.g., ##), the
ExperimentalMarkdownSyntaxTextSplitter retained the deeper header in the
metadata.

This commit updates the `_resolve_header_stack` method to remove headers
at the same or deeper levels before appending the current header. As a
result, each chunk now reflects only the active header context.

Fixes unexpected metadata leakage across sections in nested Markdown
documents.

Additionally, test cases have been updated to:
- Validate correct header resolution and metadata assignment.
- Cover edge cases with nested headers and horizontal rules.

**Issue:** 
Fixes [#31596](https://github.com/langchain-ai/langchain/issues/31596)

**Dependencies:**
None

**Twitter handle:** -> [_RaghuKapur](https://twitter.com/_RaghuKapur)

**LinkedIn:** ->
[https://www.linkedin.com/in/raghukapur/](https://www.linkedin.com/in/raghukapur/)

libs/text-splitters/langchain_text_splitters/markdown.py

@@ -376,10 +376,10 @@ class ExperimentalMarkdownSyntaxTextSplitter:
 
     def _resolve_header_stack(self, header_depth: int, header_text: str) -> None:
         for i, (depth, _) in enumerate(self.current_header_stack):
-            if depth == header_depth:
-                self.current_header_stack[i] = (header_depth, header_text)
-                self.current_header_stack = self.current_header_stack[: i + 1]
-                return
+            if depth >= header_depth:
+                # Truncate everything from this level onward
+                self.current_header_stack = self.current_header_stack[:i]
+                break
         self.current_header_stack.append((header_depth, header_text))
 
     def _resolve_code_chunk(self, current_line: str, raw_lines: List[str]) -> str: