Indexing external data to .html files

[flayio]

Hi,

I'm currently evaluating the free edition, and the problem is: I do not want the indexed keywords appearing inside the .html files.

So I'd like to know if it's possible to index external data (from .tsv or .csv files) and link it to the respective .html files.

Thanks in advance for any help.

[David]

I do not want the indexed keywords appearing inside the .html files.

I am not sure exactly what you mean. But you can turn off the indexing of meta keywords from the "Indexing options" tab in the Zoom configuration window. Does this solve the problem?

[flayio]

Does this solve the problem?

No...

What I want to do is:

Use ZSE to index a .txt file and link the results to a .html file.

Example:

Let's say that test.txt has the word apple. I want to index this file, and link the result to test.html file (that does not contain the word apple). So, when a user searches for apple, ZSE will show test.html...

[David]

In the current software release you can't do this for a HTML file. (But you can for binary files, like PDF, DOC, etc..)

You could however edit the HTML file and add some Zoom specific meta data, like this,
<meta name="ZOOMWORDS" content="apple orange grapefriut">

[flayio]

So, if you don't mind, I'd like to suggest you to implement that feature in next releases - in this way, users could keep .html files small and, at the same time, index more than 400 characters (zoomwords limit).

[David]

I think the limit is 1000 characters. Our general thinking was that if you really had this much meta data, some or all of it should be placed in the documents main content. Can you give an example of why this isn't reasonable?

[flayio]

Please forgive me, I thought I read in the manual that the zoomwords limit was 400 characters...

Anyway, a good example is using ZSE to index images keywords: you can easily reach the 400 characters limit to describe them!

[David]

But images are not HTML files? And I thought the problem was with HTML files? The limit is 1000 characters in V4 of Zoom, but you are correct that the users guide is out of date and states 400 characters (which was true in the past). We'll fix the guide.

In V5 of Zoom we'll also better support seaching for images and .desc files for images.

[flayio]

I know images are not HTML files.

I just want to:

1. create photo galleries (HTML files) using templates from imaging software like Photoshop.

2. use Zoom V5 to index the HTML files and make my image database searchable.

Have you ever visited a photo library like Getty or Corbis? That's exactly what I want to do (but on a much, much smaller scale, of course).

[David]

I see, so you probably don't want to search for the image file directly (and have search results pointing to the image).

Both Getty and Corbis use a large backend database with structured data and don't have any image related meta data on their HTML pages. So you won't get something that looks exactly like Getty or Corbis.

If you have 1 image per HTML page then 1000 characters (~200 key words) would seem enough to describe any image?

Having the keywords on the page itself (rather than in an external file or SQL database) will help Google and other external search engines find the page.

You could also use other tricky HTML techniques like including the descriptions of the photos in the body of the HTML page but initially hiding it from the user via DHTML / Javascript. (or even something as unsophisticated as using white text on a white background).

[flayio]

If you have 1 image per HTML page then 1000 characters (~200 key words) would seem enough to describe any image?).

Yes, I'll have one image per HTML page, and as I said previously, I thought the zoomwords limit was 400 characters - having 1000 characters available seems to be pretty enough...

Having the keywords on the page itself (rather than in an external file or SQL database) will help Google and other external search engines find the page.?).

Correct - but I just want to make the image caption and 4 or 5 keywords 'googlable'; the rest I would prefer to keep hidden inside .ZDAT files. Why? Three reasons: prevent keywords stealing (yes, this exists), keep the HTML file light and have the ability to edit keywords without touching the .HTML files.

You could also use other tricky HTML techniques like including the descriptions of the photos in the body of the HTML page but initially hiding it from the user via DHTML / Javascript. (or even something as unsophisticated as using white text on a white background).

Sure I could, but my DHTML / Javascript knowledge is near zero - and besides that, I really prefer to have a clean code.

[David]

Some valid reasons. If we get a couple of other people asking for the same feature we look at implementing it. But at the moment we have pretty much finalised the functionality list for V5.

[flayio]

Some valid reasons. If we get a couple of other people asking for the same feature we look at implementing it. But at the moment we have pretty much finalised the functionality list for V5.

Ok - thank you for the interest in my needs and congratulations, ZSE is a really great tool.