Improve search relevance

FORK.md

@@ -86,6 +86,17 @@ To help keep things trackable, the following branch naming conventions are used:
   By using the syntax ` ```mermaid ` in any markdown file, dbt entities can enhance
   their catalog page with any mermaid.js diagram.
 
+- Prioritise exact file name matches over fuzzy matches when conducting a search.
+
+  - Upstream PR: https://github.com/dbt-labs/dbt-docs/pull/503
+  - Upstream issue: https://github.com/dbt-labs/dbt-docs/issues/217
+  - Fork PR: https://github.com/PicnicSupermarket/dbt-docs/pull/8
+
+  Adds an additional step to the relevance calculation process. Now, a weight 
+  value is also computed and stored for the name field. Finally, the results are 
+  sorted first by overallNameWeight and then by overallWeight.
+
+
 ## Changelog entry example
 
 ```

src/app/components/search/search.html

@@ -17,11 +17,11 @@ <h1>
                 <small><span>{{ results.length }}</span> search results</small>
             </h1>
             <input type="checkbox" id="name" ng-model="checkboxStatus.show_names" ng-click="filterResults(results, checkboxStatus)" style = "margin-left:10px;margin-right:5px">
-            <label for="name" style="margin-right:25px">Name</label> 
+            <label for="name" style="margin-right:25px">Name</label>
             <input type="checkbox" id="desc" ng-model="checkboxStatus.show_descriptions" ng-click = "filterResults(results, checkboxStatus)" style = "margin-right:5px">
-            <label for="desc" style="margin-right:25px;">Description</label> 
+            <label for="desc" style="margin-right:25px;">Description</label>
             <input type="checkbox" id="column" ng-model="checkboxStatus.show_columns" ng-click = "filterResults(results, checkboxStatus)" style = "margin-right:5px">
-            <label for="column" style="margin-right:25px;">Column</label> 
+            <label for="column" style="margin-right:25px;">Column</label>
             <input type="checkbox" id="column_description" ng-model="checkboxStatus.show_column_descriptions" ng-click = "filterResults(results, checkboxStatus)" style = "margin-right:5px">
             <label for="column_description" style="margin-right:25px;">Column Description</label>
             <input type="checkbox" id="code" ng-model="checkboxStatus.show_code" ng-click = "filterResults(results, checkboxStatus)" style = "margin-right:5px">
@@ -33,7 +33,7 @@ <h1>
     <div class="app-details">
         <div class="app-frame app-pad">
             <div class="results">
-                <div ng-repeat="result in results | filter:limit_search | orderBy:'overallWeight':true track by result.model.unique_id"
+                <div ng-repeat="result in results | filter:limit_search track by result.model.unique_id"
                      data-ui-state="getState(result.model)" data-ui-state-params="{unique_id: result.model.unique_id}"
                      ng-click="onSelect()"
                      class="result search-result a">

src/app/components/search/search.js

@@ -63,10 +63,9 @@ angular
                 if(!_.some(_.values(checkboxStatus))){
                     return results;
                 }
-
                 let finalResults = [];
                 let fileIDs = [];
-                
+
                 const {show_names, show_descriptions, show_columns, show_column_descriptions, show_code, show_tags} = checkboxStatus;
                 _.each(results, function(result){
                     _.each(result.matches, function(match){
@@ -94,18 +93,18 @@ angular
             });
 
             scope.shorten = function(text) {
-                if(text != null && text.trim().length > 0 && scope.query != null && scope.query.trim().length > 0){  
-                    let modified = text.replace(/\s+/g, ' '); 
-                    //choose the first word in the search as the anchor for shortening. 
+                 if(text != null && text.trim().length > 0 && scope.query != null && scope.query.trim().length > 0){
+                    let modified = text.replace(/\s+/g, ' ');
+                    //choose the first word in the search as the anchor for shortening.
                     //Escaping in case the first token is "*" or another reserved regex character
-                    let first_token = escapeRegExp(getQueryTokens(scope.query)[0]); 
+                    let first_token = escapeRegExp(getQueryTokens(scope.query)[0]);
                     let indexOfInstance = modified.search(new RegExp(first_token));
                     let startIndex = (indexOfInstance - 75) < 0 ? 0 : indexOfInstance - 75;
                     let endIndex = (indexOfInstance + 75) > modified.length ? modified.length : indexOfInstance + 75;
                     let shortened = "..." + modified.substring(startIndex, endIndex) + "...";
                     return shortened;
                  }
-                return text;
+                 return text;
             }
 
             scope.highlight = function(text) {
@@ -116,7 +115,7 @@ angular
                 //e.g. "hello WORLD" changes to "(hello)|(world)"
                 let query_segments = getQueryTokens(scope.query);
                 let escaped_segments = query_segments.map(segment => escapeRegExp(segment));
-                let highlight_words = "(" + escaped_segments.join(")|(") + ")"; 
+                let highlight_words = "(" + escaped_segments.join(")|(") + ")";
                 return $sce.trustAsHtml(text.replace(new RegExp(highlight_words, 'gi'), '<span class="search-result-match">$&</span>'));
             }
 

src/app/main/index.js

@@ -122,7 +122,7 @@ angular
         var state_changed = true;
         if (from_state == to_state && prev_node_id == cur_node_id) {
             state_changed = false;
-        } 
+        }
 
         if (state_changed && params.unique_id) {
             var tree = projectService.updateSelected(params.unique_id);
@@ -145,75 +145,6 @@ angular
         }
     });
 
-    $scope.$watch('search.query', function(q) {
-        $scope.search.results = assignSearchRelevance(projectService.search(q));
-    });
-
-    function assignSearchRelevance(results){
-        if($scope.search.query === "")
-            return results;
-        let criteriaArr = {
-            "name": 10,
-            "tags": 5,
-            "description": 3,
-            "raw_code": 2,
-            "columns": 1
-        };
-        _.each(results, function(result){
-            result.overallWeight = 0;
-            _.each(Object.keys(criteriaArr), function(criteria){
-                if(result.model[criteria] != undefined){
-                    let count = 0;
-                    let body = result.model[criteria];
-                    let query = ($scope.search.query).toLowerCase();
-                    if(criteria === "columns"){
-                        _.each(body, function(column){
-                            // there a spark bug where columns are missign from the catalog.  That
-                            // needs to be fixed outside of docs but this if != null check will
-                            // allow docs to continue to function now and also when the bug is
-                            // fixed.
-                            // relevant issue: https://github.com/dbt-labs/dbt-spark/issues/295
-                            if (column.name) {
-                                let columnName = column.name.toLowerCase();
-                                let index = 0;
-                                while(index != -1){
-                                    index = columnName.indexOf(query, index);
-                                    if (index != -1) {
-                                        count++; index++;
-                                    }
-                                }
-                            }
-                        });
-                    }
-                    else if(criteria === "tags"){
-                        _.each(body, function(tag){
-                            let tagName = tag.toLowerCase();
-                            let index = 0;
-                            while(index != -1){
-                                index = tagName.indexOf(query, index);
-                                if (index != -1) {
-                                    count++; index++;
-                                }
-                            }
-                        });
-                    }
-                    else{
-                        body = body.toLowerCase();
-                        let index = 0;
-                        while(index != -1){
-                            index = body.indexOf(query, index);
-                            if(index != -1){
-                                count++; index++;
-                            }
-                        }
-                    }
-                    result.overallWeight += (count * criteriaArr[criteria]);
-                }
-            });
-        }); 
-        return results;
-    }
-
     /*
     INITIALIZE THE APPLICATION:
         1. Set the selected model (if there is one) via the url

src/app/services/project_service.js

@@ -3,6 +3,7 @@ const angular = require('angular');
 const $ = require('jquery');
 const _ = require('lodash');
 const { getQuoteChar } = require('./compat');
+const { assignSearchRelevance } = require('./project_service_utils');
 
 import merge from 'deepmerge';
 
@@ -334,7 +335,7 @@ angular
     }
 
     service.search = function(q) {
-        if (q.length == 0) {
+        if (q.length === 0) {
             return _.map(service.project.searchable, function(model) {
                 return {
                     model: model,
@@ -353,7 +354,7 @@ angular
                 });
             }
         });
-        return res;
+        return assignSearchRelevance(res, q);
     }
 
     function clean_project_macros(macros, adapter) {

src/app/services/project_service_utils.js

@@ -0,0 +1,89 @@
+const _ = require('lodash');
+
+
+function assignSearchRelevance(results, q) {
+        if(q === "") {
+            return results;
+        }
+       let criteriaArr = {
+            "name": 10,
+            "tags": 5,
+            "description": 3,
+            "raw_code": 2,
+            "columns": 1
+        };
+
+        _.each(results, function(result){
+            result.overallWeight = 0;
+            result.overallNameWeight = 0;
+            _.each(Object.keys(criteriaArr), function(criteria){
+                if(result.model[criteria] !== undefined){
+                    let count = 0;
+                    let body = result.model[criteria];
+                    let query = (q).toLowerCase();
+                    if(criteria === "columns"){
+                        _.each(body, function(column){
+                            // there a spark bug where columns are missign from the catalog.  That
+                            // needs to be fixed outside of docs but this if != null check will
+                            // allow docs to continue to function now and also when the bug is
+                            // fixed.
+                            // relevant issue: https://github.com/dbt-labs/dbt-spark/issues/295
+                            if (column.name) {
+                                let columnName = column.name.toLowerCase();
+                                let index = 0;
+                                while(index !== -1){
+                                    index = columnName.indexOf(query, index);
+                                    if (index !== -1) {
+                                        count++; index++;
+                                    }
+                                }
+                            }
+                        });
+                    }
+                    else if(criteria === "name"){
+                        const calculateNameMatchWeight = (body, query) => {
+                            if (body === query) return 10;
+                            const lowerBody = body.toLowerCase();
+                            if (lowerBody.startsWith(query)) return 5;
+                            if (lowerBody.endsWith(query)) return 3;
+                            if (lowerBody.includes(query)) return 1;
+                            return 0;
+                        };
+
+                        count += calculateNameMatchWeight(body, (q).toLowerCase());
+                        result.overallNameWeight += (count * criteriaArr[criteria]);
+
+                    }
+                    else if(criteria === "tags"){
+                        _.each(body, function(tag){
+                            let tagName = tag.toLowerCase();
+                            let index = 0;
+                            while(index != -1){
+                                index = tagName.indexOf(query, index);
+                                if (index != -1) {
+                                    count++; index++;
+                                }
+                            }
+                        });
+                    }
+                    else{
+                        body = body.toLowerCase();
+                        let index = 0;
+                        while(index != -1){
+                            index = body.indexOf(query, index);
+                            if(index != -1){
+                                count++; index++;
+                            }
+                        }
+                    }
+                    result.overallWeight += (count * criteriaArr[criteria]);
+                }
+            });
+        });
+        results.sort((a, b) => b.overallNameWeight - a.overallNameWeight || b.overallWeight - a.overallWeight);
+        return results
+    }
+
+module.exports = {
+    assignSearchRelevance,
+}

src/app/services/project_service_utils.test.js

@@ -0,0 +1,42 @@
+const proj_utils = require('./project_service_utils')
+
+describe('Project Service Tests', () => {
+  describe('assignSearchRelevance', () => {
+    let results;
+
+    beforeEach(() => {
+      results = [
+        { model: { name: 'dm_test', tags: ["dm", "test", "test model"], columns: {"id": {"name": "id"}}, raw_code: "SELECT test from test" }, overallWeight: 0, overallNameWeight: 0 },
+        { model: { name: 'ft_test_person', tags: ["test", "ft", "person"], columns: {"id": {"name": "id"}}, raw_code: "SELECT test, test from test" }, overallWeight: 0, overallNameWeight: 0 },
+        { model: { name: 'test_event', tags: ["test", "event"], columns: {"test": {"name": "test"}} , raw_code: "SELECT id from abc" }, overallWeight: 0, overallNameWeight: 0 },
+        { model: { name: 'test_log', tags: ["test", "log"], overallWeight: 0, overallNameWeight: 0 }},
+        { model: { name: 'test', tags: [], columns: {} }, overallWeight: 0, overallNameWeight: 0 },
+        { model: { name: 'n/a', tags: [], columns: {} }, overallWeight: 0, overallNameWeight: 0 },
+      ];
+    });
+
+    it('should prioritize exact name matches', () => {
+      proj_utils.assignSearchRelevance(results, 'test');
+      expect(results[0].model.name).toBe('test');
+      expect(results[0].overallNameWeight).toBe(100);
+      expect(results[0].overallWeight).toBe(100);
+
+      expect(results[1].model.name).toBe('test_event');
+      expect(results[1].overallNameWeight).toBe(50);
+      expect(results[1].overallWeight).toBe(56);
+
+      expect(results[2].model.name).toBe('test_log');
+      expect(results[2].overallNameWeight).toBe(50);
+      expect(results[2].overallWeight).toBe(55);
+
+      expect(results[3].model.name).toBe('dm_test');
+      expect(results[3].overallNameWeight).toBe(30);
+      expect(results[3].overallWeight).toBe(44);
+
+      expect(results[4].model.name).toBe('ft_test_person');
+      expect(results[4].overallNameWeight).toBe(10);
+      expect(results[4].overallWeight).toBe(21);
+    });
+
+  });
+});