The function returns a struct, DocumentMetadata, which contains fields
with document metadata, including the entries from the /Info dictionary
of a PDF.

Implement the function in PDFiumEngine. Add a test and check in a test
PDF.

Remove PDFEngine::GetMetadata() as it only had one caller and its
functionality would be redundant with the new GetDocumentMetadata().
Simultaneously, privatize the old GetDocumentMetadata(), which only
fetched unparsed field entries one by one and rename to
GetMetadataByField().

Bug: 93619
Change-Id: I36ce50fa2bf2e4a5eba94efa54cfd3e3bbf72f72
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2096103
Commit-Queue: Daniel Hosseinian <dhoss@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#753385}