Search Results for Image Files

[AcoustiKatie]

Alright, I have been scouring this site to try and find the answer, but I have been unsuccessful so far.

The Situation:

We are using the enterprise edition and I am trying to get a search set up on one of our sites first before attempting anything too intricate. Right now I have it set to search all .htm, .html, .txt, .jpg, .png, .gif, .tiff, and .pdf files. (Yes, I have the appropriate plugins for images/pdfs)

We also have categories set up for PDF Files (.pdf) and Image Files (.jpg, .png, .gif, and .tiff).

The Problem:

When doing a search, everything is working for the most part. I have thumbnails associated with PDFs and Images to visually make them stand out from the searches. I can search by my categories (pdf and images) just fine. But, all of the results that are images link to the html page that they are located on. Have I messed up some setting? Am I missing something? Is it possible to have the search result of an image file open the image?

Search: http://www.acousticalsurfaces.com/zoom_search/
(Mind you, this isn't "live" on the site yet as we are still tweaking. That is why you will see the crude "search over there" item.)

Thanks for reading, and especially if you have anything that can help me.

[Ray]

Images linked via <img src="mypic.jpg"> tags are considered to be "embedded" links, such that the search result will point to the page containing this tag, rather than the image file itself.

But if the image file was linked to independently (via <a href="mypic.jpg"> style links), then they will be indexed individually. This is to reflect the fact that the website itself points the user to the image file individually rather than hosting it on a page.

This behaviour only occurs in Spider Mode where links are used to find pages. In Offline Mode, all image files are indexed independently because we don't evaluate the links. If you have no dynamically generated pages (e.g. PHP pages) you might want to consider using Offline Mode and indexing the files from your drive directly. It's alot faster too.

Another setup you can consider with the latter, is to then skip all files with "_thumb" in the filename (under "Skip Options") so they won't be indexed. But setup your thumbnails for image files (under "Scan options", double click ".jpg"->"Configure images" and change from "Display same icon..." to "Display different thumbnails"). And set it up so that it always looks for a thumbnail of the same file with "_thumb" in the name. This will display the actual image thumbnail in the search results. Although I noticed not all of your thumbnails are named in the same format, this seems to be the majority of them. You can also check "Spider options"->"Check thumbnails exists on website before using URL" to avoid broken links.

[AcoustiKatie]

Ray, thank you for clearing that up for me! We do not have dynamically generated pages, so I am going to go for the offline mode and see if that helps.

Also - the site is a mess when it comes to organizing in any rhyme or reason. I walked into this a year ago and have been trying to instill some sort of method to our madness. This is the primary reason we are using a generic image to represent any images in the results. If it were possible to use the generic image ONLY if there was not another thumbnail present, that might be more intriguing. We are more interested in providing a visual clue that this will bring the user to an image vs a page than we are in actually showing the thumbnail of the page they are going to. Maybe in my free time I will organize the thumbnails

[AcoustiKatie]

Actually...I retract my beginning statement. We do have dynamic content. We have our blog which is full of dynamic stuff. Is there a way to merge a spider and offline index? Or am I just trying to do way too much? We already have a search specifically on our blog, but I liked the idea of being able to find blog posts from our main website, too.

Maybe for now I will just sacrifice the images in lieu of the blog posts. So much time, so little to do.... (Willy Wonka)

[Ray]

Unfortunately you can not merge a spider and offline mode index at this point (without a more intricate setup like using MasterNode, but that's overkill for this situation).

What I described above is still possible with just using Spider Mode, I was just suggesting Offline because it was easier.

Before I do proceed, it's worth pointing out that most people want the image search results to point to the page that the image is found on. Because otherwise, the user clicks through to the image and they have no further interaction available - there's no text description, no button to buy now, etc. Whereas being taken to the page that the image was found on gives them that.

Now if you still want an image search which points to the image files directly, you can add them as additional start points. You will need to enable Directory Listing on your server such that a URL like http://www.mysite.com/images/ will take you to a page of links to the files in that folder. Then specify each of the image folders as a start point (click "More" next to the start Spider URL). You'd want the image start points before the actual website ones so it doesn't use the web page links (it won't index the same file twice and it'll ignore them the second time around). Like I said, it's a more complicated setup.

[AcoustiKatie]

Ray, Thanks so much for all of your help. At this point, we are just going to omit images in the search and I may work on a separate image search at a later point. It was mostly for my benefit anyway. I can't find anything on that site...and I'm the one who is supposed to be updating it. If only they would let me give it a makeover...