make TPAC handle HEAD requests better

At least one link-checker has been observed sending a bunch of HTTP HEAD requests to an Evergreen catalog at a high rate. Many of the links being checked run catalog searches, and it is possible for the crawling to overwhelm available storage and search backends.

To avoid this, one possibility would be for TPac's mod_perl handlers to check for HEAD requests and refuse to run catalog searches. Possible behaviors include:

1. unconditionally return HTTP 200
2. process /eg/opac/record requests normally except for refusing to run a catalog search
3. process /eg/opac/record HEAD requests request only to the point of determining whether the bib is visible to the OPAC, but dispense with template processing or copy retrieval

Evergreen master