Confluence: preserve page hierarchy for tree filtering

[TobiasGoerke]

Why

Confluence pages are currently synchronized using only their page titles. This loses the page hierarchy and makes it impossible to use existing include/exclude filters to select a page tree.

For example, users cannot reliably select only a subtree without synchronizing the entire space.

What

Add an optional hierarchical structure mode:

confluence:ABC?structure=hierarchical

In this mode, page paths include their full Confluence hierarchy:

Knowledge Base and FAQ/FAQ - Using OpenCode.txt

The root page itself is represented as:

Knowledge Base and FAQ.txt

Existing filters can select a subtree:

source: "confluence:ABC?structure=hierarchical"
filter:
  include:
    - "Knowledge Base and FAQ/**"

The default flat mode remains unchanged.
Thus, there is full backwards compatibility.

Changes

• Add structure=hierarchical source configuration.
• Build paths from Confluence page parentId relationships.
• Resolve non-numeric space keys to numeric Confluence space IDs.
• Keep page content lookup correct for hierarchical paths.
• Reject ambiguous page paths instead of syncing the wrong page.
• Add mocked tests for parsing, both modes, filtering, page reads, key resolution, collisions, and repeated manifest builds.
• Update the Confluence documentation.

Verification

• 11 tests passed.
• Python compilation passed.
• git diff --check passed.