Matching a query string to a string to get a boolean answer whether it matches or not is a pain. Take for instance the following query, roughly matching each string related to an aviation disaster:
(plane OR helicopter OR boeing) AND (crash OR fire OR accident)
.
Parsing this sentence recursively, with each bracket and boolean AND
/OR
logic becomes a bug prone process.
Introducing Boolean-parser
This function converts a boolean query to a 2 dimensional array with all possibilities. This allows you to quickly and easily see what scenarios will equal true in a complex boolean conditional.
This tool is great when constructing complex search filters that need to be
matched with text. Creating search tools that need to show up more refined
results than that would be possible with a simple AND
combination.
View boolean-parser-js on GitHub
Examples:
Input | Output |
---|---|
a AND b |
[[a, b]] |
a OR b |
[[a], [b]] |
a AND b AND c |
[[a, b, c]] |
a AND b OR c |
[[a, b], [c]] |
a AND (b OR c) |
[[a, b], [a, c]] |
a AND (b OR c) AND (d OR e) |
[[a, b, d], [a, b, e], [a, c, d], [a, c, e]] |
Whereas a
, b
and c
represent words, forming a complex query pattern.
This function works recursively trough all brackets and generates an array of all possible combinations of a matching query.
Long term example
Input:
((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j
Output:
[[a,b,d,e,f],
[a,c,d,e,f],
[a,b,d,e,g],
[a,c,d,e,g],
[a,b,d,e,h],
[a,c,d,e,h],
[i],
[j]]
Philosophy
The output is meant to be easily parsed to check for matches. There are more efficient ways to check matches to this query by only checking each term once, though this method is one that is easier to maintain and limits risk of side effects. Especially when considering recursively nested queries involving many brackets and AND/OR combinations.
Installing
npm install boolean-parser
Usage
var booleanParser = require('boolean-parser');
var searchPhrase = '((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j';
var parsedQuery = booleanParser.parseBooleanQuery(searchPhrase);
// Returns:
// [['a','b','d','e','f'],
// ['a','b','d','e','g'],
// ['a','b','d','e','h'],
// ['a','c','d','e','f'],
// ['a','c','d','e','g'],
// ['a','c','d','e','h'],
// ['i'],['j']]
How does this library work
- Parse string to an Array of OR items (strings). Everything that’s in between brackets will be treated as one word and will later be recursively parsed.
- Go trough each string in that Array, and parse it to an array of AND items.
- Recursively call the
parseBooleanQuery
function on whatever is in between brackets. These recursive calls will return an array of all possible combinations within those brackets. (AnOR
array ofAND
combinations) - Create an empty array called
nestedPaths
. And add all nested combinations that are in between brackets to that array. For instance, with the query:((a AND (b OR c)) AND (d AND e) AND (f OR g OR h OR j)) AND x AND y AND z
Path will look like the following.
// nestedPath = [ [ [a,b], [a,c] ], [ [d,e] ], [ [f], [g], [h], [j] ] ]
- Then push the remaining non-bracket AND terms to this array.
// nestedPath = [ [ [a,b], [a,c] ], [ [d,e] ], [ [f], [g], [h], [j] ] [ [x,y,z] ] ]
- Then using the
orsAndMerge
, all those AND paths in those OR paths will be combined with the other OR combinations. In:[ [ [ a ], [ b ] ], [ [ c, d ], [ e ] ], [ [ f ] ] ]
Out:
[ [ a, c, d, f ], [ b, c, d, f ], [ a, e, f ], [ b, e, f ] ]
- Then we concatenate all those
OR
paths that were in between thoseOR
terms to one Array using themergeOrs
function. In:[ [ [ a, b ], [ c ] ], [ [ d ] ], [ [ e ], [ f, g ] ] ]
Out:
[ [ a, b ], [ c ], [ d ], [ e ], [ f, g ] ]