[XFA] Tags Generator (support threads, resources, medias, UBS, AMS) - XF2

[XFA] Tags Generator (support threads, resources, medias, UBS, AMS) - XF2 [Paid] 3.2.0

No permission to buy (€15.00)
Overview Updates (5) Reviews (4) History Discussion
While to me one thing is clear - the addon badly perceived Cyrillic. It also picks up the user name, BB-codes and other garbage, up to https protocol.
 
XFA said:
Not sure to understand ?
Click to expand...
Right now this addon support core content types (resource, threads, media), right?

There are addons that create content types (example Showcase from @Bob) that strictly follow xf development standards. How much difficult could be to automatically support addons like that?
 
imno007 said:
If you currently have zero tags on the resources, I would think the number you need in that field is 1, as in "less than 1." Have you tried that?
Click to expand...

That's it !

sbj said:
Any interest expanding the tag generator into a "custom thread fields" generator? Maybe a separate addon? @XFA
Click to expand...

Custom thread fields generator ? Hum I don't see how that would work.
Anyways, I don't have time to create new add-ons that are not in my portfolios and not custom jobs.

xBt said:
While to me one thing is clear - the addon badly perceived Cyrillic. It also picks up the user name, BB-codes and other garbage, up to https protocol.
Click to expand...

Wrt. Cyrillic, what makes you think that ?

It detects word occurences so I guess quotes and stuff like that could be detected, is that what you meant ?

Kintaro said:
Right now this addon support core content types (resource, threads, media), right?

There are addons that create content types (example Showcase from @Bob) that strictly follow xf development standards. How much difficult could be to automatically support addons like that?
Click to expand...

It's impossible to automatically support addons as I tend to add some specificities to each add-on but that's not a bad idea maybe there could be some solution.
 
XFA said:
Wrt. Cyrillic, what makes you think that ?
Click to expand...

A friend gave an old backup of his forum, there about two hundred resources and threads without tags. Title - Latin, description - Cyrillic. In one of fifty cases, the addon offers Cyrillic.

The rest of the time, displays:
1. The Title of the thread or resource
2. Site domain
3. Name of the user who posted the resource
4. BB-codes
5. https Protocol
6. If you exclude a title from the search, it may return a null result.
7. Etc.

But not Cyrillic.
This is a significant drawback, in my opinion.
 
xBt said:
A friend gave an old backup of his forum, there about two hundred resources and threads without tags. Title - Latin, description - Cyrillic. In one of fifty cases, the addon offers Cyrillic.

The rest of the time, displays:
1. The Title of the thread or resource
2. Site domain
3. Name of the user who posted the resource
4. BB-codes
5. https Protocol
6. If you exclude a title from the search, it may return a null result.
7. Etc.

But not Cyrillic.
This is a significant drawback, in my opinion.
Click to expand...

I may have to do the count in a different way to cope with cyrillic characters.

Do you have a test forum where that backup is, where you could give me access to the ftp and admin to try to fix that ?
 
XFA said:
Do you have a test forum where that backup is, where you could give me access to the ftp and admin to try to fix that ?
Click to expand...
No, as a test, I use VMware with CentOS 7 installed and configured. I can deploy a new forum and open ports. Then you'll have access, only is it legal, in terms of XenForo license?
 
xBt said:
No, as a test, I use VMware with CentOS 7 installed and configured. I can deploy a new forum and open ports. Then you'll have access, only is it legal, in terms of XenForo license?
Click to expand...

You are allowed to setup a test forum but you should htaccess protect it.
 
There is a problem with German Umlauts: ä, ö ,ü ,ß have a size larger than one. If you choose to use words with at least 4 chars you will get also "für" and other words with 3 chars and an Umlaut. In french language you should have the same problem with all that ´`^, i guess.

It would be nice if the list of not used words could be sorted and cleaned while saving. Is there an idea after which count of words the script would have problems? There thousands of words i dont want as tags.

Tags that are blacklisted needs to be in the window for stopwords?
Makes this sense? It would be better if these sblacklisted tags are just not used to fill the form?
Maybe i dont understand the workflow.

Instead of "Add tags" there should be a second button: Add and next

Option needed: The words in title should be always there as tags

My auto-functions stucks always at the first page, it is not usable. Go page by page is a hard job, because it starts always again at page one after safing.

The stop-words makes no sense for me. They should be compared with the blacklisted list instead, but while searching, not while try to save.

At the moment this is too much hand-word around.


A good workflow from my view:

Set min words
Title complete
Set how many times a word must be there (in text)
Run and fetch everything that is not in the blacklist

Then we use the list of tags to delete and blacklist words.

Then we start again.

Now we have the problem not to know which tags are new; means we have to watch the whole list again.
So the new tags should be temporarily saved in another list to be shown at top or only. Empty "new list" deletes the list.

So we walk trough all threads some times till we have cleaned all not needed now blacklisted tags.
 
Last edited:
Robert9 said:
There is a problem with German Umlauts: ä, ö ,ü ,ß have a size larger than one. If you choose to use words with at least 4 chars you will get also "für" and other words with 3 chars and an Umlaut. In french language you should have the same problem with all that ´`^, i guess.
Click to expand...

You mean für is counted like 4 chars ?
I don't know about the french language, tested in english as a matter of fact.

Robert9 said:
It would be nice if the list of not used words could be sorted and cleaned while saving. Is there an idea after which count of words the script would have problems? There thousands of words i dont want as tags.
Click to expand...

Not sure to understand what you meant by sorted and cleaned ?

I don't have a clear target of when this could become a problem, this is searching in an array but thousands of words might not be a problem.

Clément
 
I have changed the text above meanwhile.
If you cant understand it i can try it again. :)

This add-on could be a mighty tool, but i believe it needs some changes to be that powerful i see it.

In short my workflow how it should be from my view (again):

Min chars: x
Min words: y
Tag the whole title (yes/no)

1.
Before start searching, fetch all stop-words. Show the rest or go next.

2. Step
All new found tags should be in a temp table.
Now we check them and blacklist the ones we dont like. They are deleted in the threads also.
Now we delete the temporary list of new tags.

We start the search again with 1.
New tags will come to the threads.
We do again step 2.

This we do some times.

3. While a lot of threads are searched again and again though there are no new tags, the number of posts should be automatically
changed then from #1 to #x to #x to #2*

But this method had a problem, because we can do it only one time as long we dont save that we have seen x to 2*x also.
So it would be better to take random x posts here. We can do this as long we have enough tags. This is not perfect solved, but better.

Best solution would be to save that all posts are used; means we have to save the number of used posts in the thread. If there is nothing new, we dont have to touch it again.
 
Also it would be nice to have an option: show empty ones yes/no

At the moment i walk by hand through the threads, then i use the BACK-BUTTON to see the last used page again and press NEXT, because after saving i am again at page 1.

Because the list of stopwords is passed by browser, this means that you have to start from the first page = form with words.
This list must be fetched from the DB directly before searching, then you can add a word, press, save and go on with the last or next page.
 
I am sorry but there is far too much text to read here.
I am not lazy but I like things to be straightforward.
 
You must log in or register to reply here.
Top Bottom