Document section detection is too greedy

[gdavison]

When using -enable-contents-check, the section detector only looks for a match at the start of the section header.

For example, a heading like ## Imported Certificates will trigger an error, even when there's a section ## Import later in the document.

Several possible solutions could be used:

1. Adding a comment to ignore a rule
2. Only trigger a header rule if there's no exactly matching header in the document
3. Content analysis on the body of the section, especially for sections like Import which have a standard format