Rename search_template.html

[Ray]

There's nobody confused here,

My comment was in response to your questions regarding the Indexing Word rules, and your suggestions that they affect the parsing of HTML tags and the indexing of text links. They do not. I hope you did not take my comment personally, I was just pointing out that you had made some assumptions with incorrect premises, and I had to clear them up.

I'm trying to figure why this program indexes the way it does. A link sometimes is part of the content and sometimes it isn't. For my purposes it never is. There must be a simple option in configuration to choose whether to index the content in links or not.

What you are asking for is not normal or typical behaviour, and I have explained why. I can understand that it's not exactly what you would like it to do right now, and it is fair enough to say that it would be helpful in your scenario, if there was an option to toggle off the indexing of link text. However, I disagree that there "must" be such an option, or that this should be expected behaviour (as you implied earlier, when you asked "why would anyone...?").

The content of a page is essentially all the text that you can see in the browser. It would be nice if Artificial Intelligence can be employed and it can guess at what counts as navigation menus, or what are headers and footers etc., but this is not technically possible or practical.

In the search engine world, the exclusion of link text from content is a very rare option. In fact, I've never seen it in action before. As I said, I can understand that there would be situations where this would be a handy option, but I'm arguing against the idea that we are being unreasonable in making this assumption. I'm also pointing out that there are reasonable workarounds (explained below).

To further illustrate this, you should note that all link text are included in Google search results. Here's a search on Google for the word "login" where you can see the link text in the description for the facebook site:
http://www.google.com/search?q=Login

I am indexing hundreds of dynamically generated pages and there is no way to manually insert ZOOMSTOP and ZOOMRESTART.

Actually, dynamically generated pages should be easiest to make such changes. Typically, this means that a single script (PHP, ASP, etc.) is responsible for generating many different pages, and that a single line change in the script (the part that creates the "Previous" and "Next" links) will apply the change to ALL pages generated with that script.

Even if there are many scripts to apply the change to, a careful search+replace across your files should be possible (eg. the code to generate the Previous link should be partially similar, if not exactly the same on each script). It is no different to a requirement to change the header or footer of a page. These things should be possible.

[webtrade]

There's nobody confused here, I'm trying to figure why this program indexes the way it does. A link sometimes is part of the content and sometimes it isn't. For my purposes it never is. There must be a simple option in configuration to choose whether to index the content in links or not. I am indexing hundreds of dynamically generated pages and there is no way to manually insert ZOOMSTOP and ZOOMRESTART.

[Ray]

In "Indexing Options" > Indexing Word Rules there is no selection for a single quote.

Single quote is the same as an apostrophe.

The URL's being indexed on all my pages have single quotes surrounding them. Might this be the problem of seemingly stripping the html code and printing text left over in the URL as text in the description? If so, how can I fix this problem without editing every pages(not possible) until a new version of ZOOM fixes this problem?

I don't understand the problem you are describing. But it sounds like you're very confused and it might help if I try to clarify something.

As mentioned before, it may help if you can give us a URL to the page (assuming it is accessible online) so we can see the problem directly and give you more specific advice. At the moment, we can only hope you are asking the right questions (and describing the right problem), for us to give you helpful answers.

First of all, the text description that is used for a link (i.e. what goes in between the anchor tags: <a href="mylink.html">My text description</a>) is indexed unless it is surrounded by <!--ZOOMSTOP--> and <!--ZOOMRESTART-->

You ask this:

Why would anyone want the text from a link in the description of the listing?

The answer is: the text for a link is usually part of the content. And in almost all cases, this should be the default behaviour. For example, you could often have a page like so:

Welcome to my homepage. 
 
I have a <a href="http://en.wikipedia.org/wiki/wombat">pet wombat</a>, 
and his name is <a href="fatso1.jpg">Fatso</a>. 
 
Click <a href="fatso2.jpg">here</a> to see his picture! 
Or <a href="nextpage.html">click here</a> to continue.

If we strip out the link text, this would be how it would appear in the search result description:

Welcome to my homepage. I have a, and his name is. Click to see his picture! Or to continue.

I hope you can see now that it is not meaningful to strip link text out by default.

If there is text that you want to exclude, then you need to wrap it in ZOOMSTOP/ZOOMRESTART tags. If there are links that you wish the spider not to follow, you wrap it in ZOOMSTOPFOLLOW AND ZOOMRESTARTFOLLOW tags.

If you wish to allow the spider to crawl a link entitled "PREVIOUS" but NOT index the text labels, you'd need something like this:

<a href="previous.html"><!--ZOOMSTOP-->PREVIOUS<!--ZOOMRESTART--></a>

I hope that clarifies things a bit.

Is the = (equal) symbol in URL's not being in Indexing Word Rules causing link text to be printed in the description?

The index word rules have nothing to do with how HTML code is handled. It only affects how actual content (words that would appear on a page, or meta description etc.) is indexed, and whether the words are split into multiple words. It does not affect the parsing of HTML tags/code. The equals sign does not have anything to do with this issue.

[webtrade]

It must be the equal sign causing the problem

Why would anyone want the text from a link in the description of the listing? The equal sign not being in Indexing Word Rules I figured out is what is causing the problem of URL text being printed in my description. How do I fix this? How can we add the equal sign to the Indexing Word Rules?

[webtrade]

Is the = (equal) symbol in URL's not being in Indexing Word Rules causing link text to be printed in the description?

[webtrade]

The apostrophe should be the same as a single quote?

[webtrade]

In "Indexing Options" > Indexing Word Rules there is no selection for a single quote. The URL's being indexed on all my pages have single quotes surrounding them. Might this be the problem of seemingly stripping the html code and printing text left over in the URL as text in the description? If so, how can I fix this problem without editing every pages(not possible) until a new version of ZOOM fixes this problem?

[webtrade]

That was a typo. It indexes where it is supposed to. It just strips the link code around PREVIOUS and NEXT links and thinks it is text and part of the description.

[Ray]

The version is OK I just bought this program 2 days ago on Monday. I use this combination:
<body>
<!--ZOOMSTOP-->
<!--ZOOMSTOPFOLLOW-->
code not to be indexed
<!--ZOOMRESTARTFOLLOW-->
<!--ZOOMRESTART-->
content to be indexed (dynamically generated section)
<!--ZOOMSTOP-->
<!--ZOOMSTOP-->
code not to be indexed
<!--ZOOMRESTARTFOLLOW-->
<!--ZOOMRESTART-->
</body>
</html>

Was this a typo in your post? According to your example above, you have two ZOOMSTOP tags immediately after each other, as opposed to where I think you meant to have a ZOOMSTOPFOLLOW tag. If this is actually the case on your web page, then this would certainly cause the "code not to be indexed" part to be followed.

If your webpage is online, you might save some time by just pasting us the URL to the page and we can take a look.

[webtrade]

The version is OK I just bought this program 2 days ago on Monday. I use this combination:
<body>
<!--ZOOMSTOP-->
<!--ZOOMSTOPFOLLOW-->
code not to be indexed
<!--ZOOMRESTARTFOLLOW-->
<!--ZOOMRESTART-->
content to be indexed (dynamically generated section)
<!--ZOOMSTOP-->
<!--ZOOMSTOP-->
code not to be indexed
<!--ZOOMRESTARTFOLLOW-->
<!--ZOOMRESTART-->
</body>
</html>

[Ray]

My other problem. I use the ZOOMSTOP ZOOMRESTART combo around dynamically generated code. Inside the dynamically generated code are 3 links. One link has the label PREVIOUS another has the label NEXT and the other has the link of the item I want indexed. I am able to have the software not index the PREVIOUS and NEXT links but the labels show up at the start of the description as I stated earlier. Somehow it's stripping the link tags around PREVIOUS and NEXT and the program thinks these labels are text.

This sounds like you are using the ZOOMSTOPFOLLOW and ZOOMRESTARTFOLLOW tags instead of ZOOMSTOP/ZOOMRESTART. You might want to check this (an explanation of their differences can be found here, with more examples in the Users Guide).

Another possibility is if you are using an older build: there was a bug related to ZOOMSTOP which was recently fixed. Check your version and build (click "Help" -> "About") and make sure you are using the latest available here:
http://www.wrensoft.com/zoom/whatsnew.html

One more question, I might as well ask it here. Is there a search term that can be searched that will list all items besides the star and question mark " *? " ?

"**" should work (the multiple stars are required to defeat the 'minimum word length to skip' setting), or "*e*" for all pages containing the letter 'e'. Generally, we don't recommend doing this for anything but debugging purposes. It is a slow, exhaustive search.

[webtrade]

One more thing

One more question, I might as well ask it here. Is there a search term that can be searched that will list all items besides the star and question mark " *? " ?

[webtrade]

Includes Working

I have the includes working perfectly. This problem is solved. Thanks.

My other problem. I use the ZOOMSTOP ZOOMRESTART combo around dynamically generated code. Inside the dynamically generated code are 3 links. One link has the label PREVIOUS another has the label NEXT and the other has the link of the item I want indexed. I am able to have the software not index the PREVIOUS and NEXT links but the labels show up at the start of the description as I stated earlier. Somehow it's stripping the link tags around PREVIOUS and NEXT and the program thinks these labels are text.

There is no way to add the ZOOMSTOP ZOOMRESTART to not include the PREVIOUS and NEXT links because there are thousands of pages being generated.

[David]

If you are creating your own search engine, you need to change the action variable

Yes, if you have created your own seach form, then you need to make sure that form points to the correct script.

You also need to specify the link back URL in the Zoom configuration window. As detailed in the FAQ.

You can avoid text being indexed using the ZOOMSTOP tags.

[webtrade]

Includes working now in search results pages

Excellent. Now I am able to have php includes work with the search script. One note needs to be added to the support page at: http://www.wrensoft.com/zoom/support/faq_ssi.html
If you are creating your own search engine, you need to change the action variable in the Form tag to the new script path or URL to keep the script from going to the embedded script search.php.

Now my other problem causing me to use only titles in the listing without descriptions. The pages I am indexing are dynamically generated simple product pages. The listing I'm trying to index is surrounded by two links labeled PREVIOUS and NEXT next to each other. I was able to configure out the indexing of the PREVIOUS and NEXT links but the software is including the labels PREVIOUS and NEXT in the description. This is an example of every description that appears : "PREVIOUS NEXT Anything about the item. Product information and description continues in the description".
The PREVIOUS NEXT labels show up in the description as text and are not links. I don't want PREVIOUS and NEXT to show up as either text or links in the description.