Chosen: Search doesn't work for items in parantheses: searching for "foo" doesn't find "(Foo)"

[archon810]

Pretty much see topic - parentheses shouldn't be considered word parts and should be ignored during search, just like most other punctuation.

[brennanmceachran]

I believe this has to do with the search regex. I''ve never gone near coffee script, but all you need to do is add an "i" at the end of the regex (i think)

regex = new RegExp('^' + searchText.replace(/[-[]{}()_+?.,\^$|#\s]/g, "\$&"), 'i')
zregex = new RegExp(searchText.replace(/[-[]{}()_+?.,\^$|#\s]/g, "\$&"), 'i')

becomes

regex = new RegExp('^' + searchText.replace(/[-[]{}()_+?.,\^$|#\s]/gi, "\$&"), 'i')
zregex = new RegExp(searchText.replace(/[-[]{}()_+?.,\^$|#\s]/gi, "\$&"), 'i')

Can anyone confirm this or merge it with the master?

[CHgsd]

In regular expressions, the "i" flag deals with case (in)sensitivity. In the listed examples where it was recommended to put an "i" is a regular expression used to sanitize the search string so that it can be fed to another regex.

The "problem" which Chosen has, which I also ran into, is that it only considers word matches starting from the beginning of a word.

For example, "and" will match the following: (bold denotes where a match occurs)

• Anderson
• Cats and dogs
• Robots vs. Androids

But will specifically NOT match: (bold denotes where a match might be expected, but does not occur)

• Land department
• Illegal landfill
• Panda bear

The solution is quite easy and is to remove the '^'+ before searchText resulting in a line which looks like this:
regex = new RegExp(searchText.replace(/[-[\]{}()*+?.,\\^$|#\s]/g, "\\$&"), 'i');

In essence making it the same as zregex. The ^ is used to denote the beginning of the string and thus restricts matches to such.
I am not certain of why the developers decided to perform the search this way, but by looking at the code I would say that it is "by design" and not an accident.

Cheers,

CHgsd

[robpodosek]

Yep, I dug into the source a few days ago and found the offending line, removed the four characters and it worked perfectly.

[sokai]

Thanks CHgsd for your fix! - It works like a charm... :)

Maybe you (or someone else) can give me an answer/hint for the following three things (that is related to the search aspect in this issue):
1st: What for is the mentioned "zregex"?
2nd: With the above fix I can search within whole words. - But how is it possible to highlight all occurrences of the search string? (i.e. if you search for "a" only "America" is highlighted, not "America".)
3rd: How I can define that highlighting should start until I typed in (i.e.) three letters?

Thanks a lot & best regards! :)

[CHgsd]

@sokai:
I cannot guarantee these answers as I am merely a user of Chosen, however they are my best guesses/deductions:

A1: zregex appears to only serve for the highlighting portion of the code. This is what leads me to believe that someone wanted to only match on the beginning of words, otherwise the same regex could be used for both matching and highlighting, no need to separate them. If I were to rewrite this portion of the code I would replace zregex with regex, however removing the carrot is a simpler solution in the short term.

A2: I am not in a position where I can test this right now, but if you want to give this a go, it _maybe_ will work:

Replace this:

              if (searchText.length) {
                startpos = option.html.search(zregex);
                text = option.html.substr(0, startpos + searchText.length) + '</em>' + option.html.substr(startpos + searchText.length);
                text = text.substr(0, startpos) + '<em>' + text.substr(startpos);
              } else {
                text = option.html;
              }

With this:

              if (searchText.length) {
                text = option.html.replace( zregex, function(matched) {
                  return "<em>"+matched+"</em>";
                });
              } else {
                text = option.html;
              }

A3: This is again completely guessing, but perhaps this:

Replace this:

              if (searchText.length) {

With this:

              if ((searchText.length) && (searchText.length >= 3)) {

I hope this helps, please post back if it works so that others can know.

[sokai]

Hey CHgsd,

thanks a lot for your examples!
Sadly I have to confess that I don't use Chosen any longer... - So I can't (ATM) test your hints... :-/

Thank you all the same and best regards!

[sontek]

I have text like "US/Central" and its not finding central, I think this is a similar issue.

[CHgsd]

@sontek You could be running into a couple of different issues. The code examples I provided in this issue should resolve all of them. If they do not please let me know and I will have a deeper look.

[wachunga]

Had same problem. Fixed by applying #212.

[mcanlas]

The options should be documented somewhere.

[Aymkdn]

For people that are looking at this thread, there is now the option search_contains:true that's doing the trick, but it's not documented.

[stof]

@Aymkdn it is documented, together with all options: https://harvesthq.github.io/chosen/options.html

[Aymkdn]

@stof thank you! I never noticed this page.... !

[koenpunt]

So this issue can be closed now?

[Aymkdn]

I think so.