Simple SEO URL (OLD version) [support thread]

[jeking]

Check again, see if you like something like this:

http://demo.rubikintegration.com/zencart/product_info/c-big-linked-22/p-a-bug-s-life-multi-pak-special-2003-collectors-edition-34

Yes, that's more like it.

[yellow1912]

@skaimauve: so you want the id first? Anyone else wants this link format:

http://host/store/product_info/c2-vinyl/p4-led-zep

Instead of:

http://demo.rubikintegration.com/zencart/product_info/c-big-linked-22/p-a-bug-s-life-multi-pak-special-2003-collectors-edition-34

[yellow1912]

Alpha 2.6 is now released with new fixes and new features:

1. EZ page links are built correctly now
2. Allowing you to set the category depth level as far as you want.
Example: assuming you a category tree like this: 10_11_13_15
You can set it so that the names of the 3 last categories appear on the link:
c-category11name-category13name-category15name-10_11_13_15

For upgrader: there is a database patch in this new update, make sure you read the doc first.

[skaimauve]

@skaimauve: so you want the id first?

Well, I am toying with different options.The great thing is that your code is that it's clear and it allows for customization. All I need is to suit the client's preference.

In my case, the client does not like Amazon links:

http://www.amazon.com/Led-Zeppelin-II/dp/B000002J03

CRE Loaded SEO links are acceptable:

Example of Category URL: 
http://host.com/Mens/c30/index.html
Example of Product URL: 
http://host.com/Mens/c30/p53/poloshirt/product_info.html

All it needs is to be readable!

[yellow1912]

Im glad it helped you, there are plenty of other things you need to take care of b4 using it on a live site:

A proper 301 direction is need, especially if the site is not new
Also, the way we put the cpath and pid right, while making the link shorter, expose it to various possible errors, but good thing is that these errors will not occur unless the users play with the link, and even, so no harm will be done. This mod does not effect ZC internal security mechanism in anyway, so as long as ZC properly sanitizes all the inputs as it is doing now, there is nothing to worry about.

If you happen to develop something on the current code, dont forget to post it here and I'll gladly merge that into the current one if possible.

[skaimauve]

Alpha 2.6 is now released with new fixes and new features:

1. EZ page links are built correctly now
2. Allowing you to set the category depth level as far as you want.
Example: assuming you a category tree like this: 10_11_13_15
You can set it so that the names of the 3 last categories appear on the link:
c-category11name-category13name-category15name-10_11_13_15

Nicely done! Just remember to clear the cache after changing the category depth level. Maybe the code should do that for you.

[yellow1912]

Right now, to build a category tree name with x cats for the FIRST time the mod has to query database x times. Even though this is done only ONCE, it is not efficient.
I should either try to make use of the name already retrieved in files, or use just 1 query. There is a a problem with trying to use 1 single query string tho, how should we select....in() yet still make the category names in the exact order.

[skaimauve]

Right now, to build a category tree name with x cats for the FIRST time the mod has to query database x times. Even though this is done only ONCE, it is not efficient.

My take on this is:
1. it is done once, therefore it is efficient, considering
2. it is looping though simple queries, therefore less prone to bugs

If someone has a site with x cats, where x is large, they should have the power to run these Simple SEO URL queries, cause the rest of ZenCart will require a lot of power anyway.

And again, if this is Simple SEO URL, we should have relatively compact code.

[yellow1912]

Thanks skaimauve, I thought so as well.

Also, I have made minor changes to the code, previously the content in files are stored like this:
example: c-name1-name2-19_20

From now on, we will simply store the names only: name1-name2

This allows us to re-use the names if we ever need to.

[skaimauve]

... these errors will not occur unless the users play with the link, and even, so no harm will be done..

Maybe you could check that the link exists in the cache before processing it, in a way similar to the code I have submitted.

  if(file_get_contents($this->cache_folder.$file_name) !== false)

Just a thought.