+ /*
+ * We intentionally do not check whether the
+ * escaped character _should_ have been escaped
+ * according to the rules in html_format_path.
+ *
+ * All clients of this parsing function, after a
+ * successful parse, call html_format_path to find
+ * the canonical URI for the same index and return
+ * an HTTP redirect if the provided URI was not
+ * exactly equal to that canonical form. This is
+ * critical when the correction involves adding or
+ * removing a trailing slash (because then
+ * relative hrefs on the generated page can be
+ * computed with respect to the canonical URI
+ * instead of having to remember what the actual
+ * URI was), but also has the useful effect that
+ * if a user attempts to type in (guess) a URI by
+ * hand they don't have to remember the escaping
+ * rules - as long as they type _something_ that
+ * this code can parse into a recognisable
+ * pathname, it will be automatically 301ed into
+ * the canonical form.
+ */