From 8aaa3f56f5949e9fd456bff653b906e87d533c20 Mon Sep 17 00:00:00 2001
From: Haiyan Meng <haiyanmeng@google.com>
Date: Tue, 3 Dec 2019 10:31:36 -0800
Subject: [PATCH] Set the ElasticSearch index creation configuration

Currently, the `kustomize` index in ElasticSearch is using dynamic
mapping, which sets the types of all the fields to `text`. However,
`text` field type is good for full-text value matching, and not good for
exact-value matching.  For exact-value matching, the `keyword` filed
type should be used.
---
 api/internal/crawl/index/elasticsearch.go | 44 +++++++++++++++++++----
 api/internal/crawl/index/kustomize.go     | 18 ++++++++++
 2 files changed, 56 insertions(+), 6 deletions(-)

diff --git a/api/internal/crawl/index/elasticsearch.go b/api/internal/crawl/index/elasticsearch.go
index 0d5dc6993..8893dc2e8 100644
--- a/api/internal/crawl/index/elasticsearch.go
+++ b/api/internal/crawl/index/elasticsearch.go
@@ -13,6 +13,40 @@ import (
 	"github.com/elastic/go-elasticsearch/v6/esapi"
 )
 
+const IndexConfig = `
+{
+	"mappings": {
+		"_doc": {
+			"properties": {
+				"repositoryUrl": {
+					"type": "keyword"
+				},
+				"filePath": {
+					"type": "keyword"
+				},
+				"defaultBranch": {
+					"type": "keyword"
+				},
+				"document": {
+					"type": "text"
+				},
+				"creationTime": {
+					"type": "date"
+				},
+				"kinds": {
+					"type": "text"
+				},
+				"identifiers": {
+					"type": "text"
+				},
+				"values": {
+					"type": "text"
+				}
+			}
+		}
+	}
+}`
+
 // TODO(damienr74) Split index into reader and writer?
 type index struct {
 	ctx    context.Context
@@ -118,21 +152,19 @@ func (idx *index) UpdateSetting(settings []byte) error {
 		res, err, ignoreResponseBody)
 }
 
-// Create an index providing both the mappings and the settings.
-func (idx *index) CreateIndex(mappings []byte, settings []byte) error {
-	request := byteJoin(`{ "mappings":`, mappings, `, "settings":`, settings, `}`)
+// Create an index providing the config for both the mappings and the settings.
+func (idx *index) CreateIndex(config []byte) error {
 	op := idx.client.Indices.Create
 	res, err := op(
 		idx.name,
-		op.WithBody(bytes.NewReader(request)),
+		op.WithBody(bytes.NewReader(config)),
 		op.WithContext(idx.ctx),
 		op.WithHuman(),
 		op.WithPretty(),
-		op.WithIncludeTypeName(true),
 	)
 
 	return idx.responseErrorOrNil(
-		fmt.Sprintf("could not create index with config '%s'", request),
+		fmt.Sprintf("could not create index with config '%s'", config),
 		res, err, ignoreResponseBody)
 }
 
diff --git a/api/internal/crawl/index/kustomize.go b/api/internal/crawl/index/kustomize.go
index 68bfe4cd5..79833d3fd 100644
--- a/api/internal/crawl/index/kustomize.go
+++ b/api/internal/crawl/index/kustomize.go
@@ -96,6 +96,24 @@ func NewKustomizeIndex(ctx context.Context) (*KustomizeIndex, error) {
 	if err != nil {
 		return nil, err
 	}
+
+	indicesExistsOp := idx.client.Indices.Exists
+	resp, err := indicesExistsOp([]string{"kustomize"},
+		indicesExistsOp.WithContext(idx.ctx),
+		indicesExistsOp.WithPretty())
+	if err != nil {
+		return nil, err
+	}
+
+	if resp.StatusCode == 200 {
+		fmt.Printf("The kustomize index already exists\n")
+	} else {
+		fmt.Printf("Creating the kustomize index\n")
+		if err := idx.CreateIndex([]byte(IndexConfig)); err != nil {
+			return nil, err
+		}
+	}
+
 	return &KustomizeIndex{idx}, nil
 }