Some time since the last time I looked at it, Google Books improved their search results interface, for books which are not viewable in full.
Or at least they did for some books, at least as viewable by me. I know sometimes Google rolls out a test feature to just some people; and I’ve learned that sometimes it’s something I thought was new, but was just applied in very specific circumstances.
But anyway, what I used to see from GBS search results was simply a textual results list, showing hits for query in context just in ‘ascii’ excerpts — and just as one little box on book metadata page.
What I now see, at least in one example, is actual scanned pages, with band markers on the vertical scrollbar indicating at what points the matches were found; clicking on them shows highlighted matches on actual scanned images. Very nice! Although I wonder if it will confuse some users.
Now Umlaut, has been providing it’s own search box, directing to results on an actual GBS page upon submission. I did this just by reverse engineering the URLs that GBS used for searches, and combining with the book URLs returned by the GBS Data API.
Quite nicely, although they changed their interface, and I think the nature of the URLs returned by the Data API have also changed since I wrote the code — the simple procedure te code uses to create a direct link to search-within-a-book results still works quite well with no changes, showing you the new interface, upon following a direct link like this one.
If you pay attention you’ll see there’s an anchor (aka “fragment identifier”) on the URL Umlaut directs you to, that no longer is neccesary or useful, but in the old version of the page targetted the result list section.