JS-比较两个数组与indexOf

Question

I'm trying to build a document search that accounts for fragments of words and different ordering of words - for instance, someone might be trying to find the section:

Level 6 Pathology

which they might search for by inputting

Pathology Level 6
Level 6 Path
6 path

I also want to make sure that ALL the search terms are in the result so that

 level 6 path

doesn't open every single sub heading containing 'level' in the document.

To accomplish these requirements, I'm combining the search input strings "level" "6" "path" into every possible combination of space separated strings "level 6 path", "6 level path", "path 6 level" etc.

Code:

function search(){
            event.preventDefault();
            var results = [];
            var searchinput = document.getElementById("searchbar").value.toUpperCase().split((" "), 6);                
            var word = searchinput.join(" ");
            if (searchinput.length > 1){
                var searchinputExpanded = recurCombinations(searchinput.length, searchinput);
                searchinputExpanded.unshift(word);
            } else {
                var searchinputExpanded = searchinput;
            }
            var arrayofdivs = document.querySelectorAll(".collapsible");                
            var j, k, m;

            //Find every combination of the search input so that both
            //Level 6 Path and Path Level 6 will find Level 6 Pathology
            function recurCombinations(len, input){
                var i = 0;
                var l = len - 1;

                if(len === 1){
                    return;
                } else {
                    for(i; i < l; i++){
                        recurCombinations(l, input);
                        l % 2 ? input.swap(i, l) : input.swap(0,l); //even-odd check 
                        results.push(input.join(" ").toString());
                    }
                    recurCombinations(l, input);
                }
                return results.sort();
            }            

            //close any currently expanded divs
            for(m = 0; m < arrayofdivs.length; m++){                    
                arrayofdivs[m].style.backgroundColor = "#eee";
                arrayofdivs[m].style.color = "#2CB2FF";
                arrayofdivs[m].nextElementSibling.style.display = "none";
            }

            //expand any applicable results
            for(j = 0; j < arrayofdivs.length; j++){
                for(k = 0; k < searchinputExpanded.length; k++){                      

                    if(arrayofdivs[j].innerHTML.toUpperCase().indexOf(searchinputExpanded[k]) !== -1){                                                 

                        arrayofdivs[j].style.backgroundColor = "#555";
                        arrayofdivs[j].style.color = "#fff";
                        arrayofdivs[j].nextElementSibling.style.display = "block";
                        arrayofdivs[j].parentNode.style.display = "block";
                        document.getElementById("clearbutton").style.display = "block";                
                    }
                }
            }
        }           

I've got this part working fine but some titles contain more words and the search doesn't account for this

Medical Imaging Level 5

Will be found by

Level 5 Medical

but not

Level 5 Imaging

And in this instance, I think most people are going to search "Imaging level 5" instead of "Medical level 5"

I tried doing every combination of words for possible results - basically:

 everyCombinationOfEveryPossibleResult.indexOf(everyCombinationOfSearchInput(path level 6) !== -1)

but the n! possibilities for each phrase when some possible result phrases are quite long meant it took too long (search input is capped at 6 strings for this reason) and would crash the browser.

Any help would be appreciated.

Answer 1

If I understand correctly, you want to only find strings that contain ALL of the words that are in the search term, but you don't want the order of the search terms to matter. In that case, you can do something like this:

function searchMyArray(arrayToSearch, searchTermsArray){
    return arrayToSearch.filter(function(title){
        return !searchTermsArray.some(function(word){
            return title.indexOf(word) == -1;
        });
    });
}

In the example, arrayToSearch should be an array consisting of all the titles that you want to be searching through, and searchTermsArray should be an array of the search terms that you want to search for (if the search term comes in as a string, you can pretty easily divide it into an array with the String split method). The function returns a list of titles that contains all of the search terms.

The function works by filtering out all of the titles where some word in the search term is not in the title.

Answer 2

The function looks for the first word, if not there then continues into the next haystack option. If it finds the word, then it continues deeper, searching for the next word.

function searchForIndex(needleArr,haystackArr){
 var temp = needleArr;
 for(var i = 0; i < haystackArr.length;i++){
  if (haystackArr[i].indexOf(temp.shift())> -1){
   if (temp.length ==0) {
    return i;
   } else {
    return searchForIndex(temp,haystackArr);
   }
  }
 }
 return -1;
}