[3.2.9-3.3.x][RC] Glossary 3.0.20

[3Di]

i cant see the extension installed, help

Read the first post for instructions.

[pierredu]

I can't give a better answer.

[motte]

A really great idea would be if it is also stated who added and edited an entry.

[pierredu]

Fine idea.
I'll take a note for next version.
In the mean time, you could use the cat column to store the author name.

[wamfiya]

Another cool improvement might be the use of the bbcode editor for the 'definition' field.

[pierredu]

Another good idea.

[gbergna]

Installed the latest version available on github (3.0.23) with phpbb version 3.3.12 and everything works perfectly.

However, I would like to propose one important change, in my opinion.

- in a post, I may have text where a word is repeated very often anchored to the glossary link. It would be sufficient to anchor the link just once for each type of word in that post without continuing to anchor links to the same words.

For example:
1) a user posts a response
2) in the answer, the words "meteorite" are present 4 times and the words "orbit" 5 times. I therefore find myself having 9 words with 9 links in that answer.
3) It would therefore be sufficient to anchor to the link "once the word meteorite" and "once the word orbit"

This would allow you to have a much tidier text on the forum from a visual and also functional point of view. There is no point in having too many words with links anchored to them.

[pierredu]

Hello,

This is a fine idea. I'll try to implement it.

[gbergna]

Installed the latest version available on github (3.0.23) with phpbb version 3.3.12 and everything works perfectly.

However, I would like to propose one important change, in my opinion.

- in a post, I may have text where a word is repeated very often anchored to the glossary link. It would be sufficient to anchor the link just once for each type of word in that post without continuing to anchor links to the same words.

For example:
1) a user posts a response
2) in the answer, the words "meteorite" are present 4 times and the words "orbit" 5 times. I therefore find myself having 9 words with 9 links in that answer.
3) It would therefore be sufficient to anchor to the link "once the word meteorite" and "once the word orbit"

This would allow you to have a much tidier text on the forum from a visual and also functional point of view. There is no point in having too many words with links anchored to them.

Probably, I have find a solution. Change the function

Code: Select all

private function glossary_pass($texte)

in listener.php like this:

Code: Select all

	private function glossary_pass($texte)
	{
		static $glossterms;
	
		$anchored_terms = []; // Reset per ogni chiamata della funzione
	
		if (!isset($glossterms) || !is_array($glossterms))
		{
			$this->compute_glossary_list();
			$glossterms = $this->cache->get('_glossterms');
		}
		if (sizeof($glossterms))
		{
			$acro = $code = $quote = $img = $link = $script = false;
			$rech = $glossterms['rech'];
			$remp = $glossterms['remp'];
			preg_match_all('#[][><][^][><]*|[^][><]+#', $texte, $matches);
			$parts = $matches[0];
			if (empty($parts))
			{
				return '';
			}
			foreach ($parts as $index => $part)
			{
				// Code
				if (strstr($part, '[code'))
				{
					$code = true;
				}
				if (!empty($code) && strstr($part, '[/code'))
				{
					$code = false;
				}
				// quote 3.2.0
				if (strstr($part, '[quote'))
				{
					$quote = true;
				}
				if (!empty($quote) && strstr($part, '[/quote'))
				{
					$quote = false;
				}
				// Pictures
				if (strstr($part, '[img'))
				{
					$img = true;
				}
				if (!empty($img) && strstr($part, '[/img'))
				{
					$img = false;
				}
				// <a> - <a> links
				if (strstr($part, '<a '))
				{
					$link = true;
				}
				if (!empty($link) && strstr($part, '</a'))
				{
					$link = false;
				}
				// [url] - [url] links
				if (strstr($part, '[url'))
				{
					$link = true;
				}
				if (!empty($link) && strstr($part, '[/url'))
				{
					$link = false;
				}
				// Script
				if (strstr($part, '<script '))
				{
					$script = true;
				}
				if (!empty($script) && strstr($part, '</script'))
				{
					$script = false;
				}
				if (!($part[0] == '<') && !($part[0] == '[') &&
					empty($code) && empty($quote) &&
					empty($img) && empty($link) && empty($script))
				{
					foreach ($rech as $key => $pattern)
					{
						if (preg_match($pattern, $part, $matches))
						{
							$term = strtolower($matches[1]); // Convertire a minuscolo per confronto case-insensitive
							// Check if this term has already been anchored
							if (!in_array($term, $anchored_terms))
							{
								$part = preg_replace($pattern, $remp[$key], $part, 1);
								$anchored_terms[] = $term; // Aggiungiamo il termine all'array dei termini ancorati
							}
							else
							{
								// Term has already been anchored, do not replace again
								$part = preg_replace($pattern, '$1', $part, 1);
							}
						}
					}
					$parts[$index] = $part;
				}
			}
			unset($part);
			return implode("", $parts);
		}
		else
		{
			return ($texte);
		}
	} // glossary_pass

[gbergna]

Another tips... change the database table "variants" from varchar(80) to varchar(500) so in this way, we can put more words into the "terms of variant" fot ancora them

[pierredu]

Thanks for your contribution.
I'll have a look at it later on. I'm now in holidays.

[Hervé]

Hi,
Filling the form for each term is tedious, is it possible to add terms in batch ?

It seams that the terms are not case sensitive, for instance if I add "PAR" (Photosynthetically Active Radiation) in upper case.
The words "par" in lower case which means "by" and is very common in french are also highlighted !

[pierredu]

You can do that in the database, directly.
There are indeed false positives.
I don't think there is a solution for such a problem.

[Hervé]

Directly in the database, of course, but not given to anybody.
I could rather write a php script which inserts the data in the DB.

About the lowercase/uppercase, I guess there is something somewhere that identifies a world to highlight and doesn't care if it's the same case.
Would it be possible to modify this process so that it becomes case sensitive?
Some search functions of php exist with case sensitive or insensitive version.
Of course if I want to detect for example the word "word" I will have to specify "word, Word" so that it detects both.

Otherwise a trick so that if I specify "PAR" it does not highlight "par" but only "PAR"?

[Hervé]

I have tried with "P.A.R." but it is not detected in the message.
I've found a solution : "P A R".

In which module does the glossary search take place ?