From ef3ee358d108dff6f74fc505021d618a2ba5555f Mon Sep 17 00:00:00 2001
From: Grische <2787581+grische@users.noreply.github.com>
Date: Tue, 11 Jun 2024 17:11:34 +0200
Subject: [PATCH] fix(database/influxdb): replace invisible chars and
 whitespaces with space (#235)

This fixes issues where users use '\n' in tag values (e.g. the owner)
that is incompatible with InfluxDB's Line protocol.
---
 database/influxdb/database.go      | 24 ++++++++++++++++++
 database/influxdb/database_test.go | 40 +++++++++++++++++++++++++++++-
 2 files changed, 63 insertions(+), 1 deletion(-)

diff --git a/database/influxdb/database.go b/database/influxdb/database.go
index cd636be6..b28e31e7 100644
--- a/database/influxdb/database.go
+++ b/database/influxdb/database.go
@@ -1,8 +1,10 @@
 package influxdb
 
 import (
+	"strings"
 	"sync"
 	"time"
+	"unicode"
 
 	"github.com/bdlm/log"
 	"github.com/influxdata/influxdb1-client/models"
@@ -94,6 +96,25 @@ func Connect(configuration map[string]interface{}) (database.Connection, error)
 	return db, nil
 }
 
+func sanitizeValues(tags models.Tags) models.Tags {
+	// https://docs.influxdata.com/influxdb/v2/reference/syntax/line-protocol/
+	// Line protocol does not support the newline character \n in tag or field values.
+	// To be safe, remove all non-printable characters and spaces except ASCII space and U+0020.
+	for _, tag := range tags {
+		cleaned_value := strings.Map(func(r rune) rune {
+			if unicode.IsPrint(r) {
+				return r
+			}
+			return ' '
+		}, string(tag.Value))
+
+		if cleaned_value != string(tag.Value) {
+			tags.SetString(string(tag.Key), cleaned_value)
+		}
+	}
+	return tags
+}
+
 func (conn *Connection) addPoint(name string, tags models.Tags, fields models.Fields, t ...time.Time) {
 	if configTags := conn.config.Tags(); configTags != nil {
 		for tag, valueInterface := range configTags {
@@ -107,6 +128,9 @@ func (conn *Connection) addPoint(name string, tags models.Tags, fields models.Fi
 			}
 		}
 	}
+
+	tags = sanitizeValues(tags)
+
 	point, err := client.NewPoint(name, tags.Map(), fields, t...)
 	if err != nil {
 		log.Panicf("could not save points: %s", err)
diff --git a/database/influxdb/database_test.go b/database/influxdb/database_test.go
index dce80029..673fa752 100644
--- a/database/influxdb/database_test.go
+++ b/database/influxdb/database_test.go
@@ -7,7 +7,7 @@ import (
 	"time"
 
 	"github.com/influxdata/influxdb1-client/models"
-	"github.com/influxdata/influxdb1-client/v2"
+	client "github.com/influxdata/influxdb1-client/v2"
 
 	"github.com/stretchr/testify/assert"
 )
@@ -98,3 +98,41 @@ func TestAddPoint(t *testing.T) {
 		connection.addPoint("name", models.Tags{}, nil, time.Now())
 	})
 }
+
+func TestAddPointWithInvalidCharacters(t *testing.T) {
+	assert := assert.New(t)
+
+	connection := &Connection{
+		config: map[string]interface{}{},
+		points: make(chan *client.Point, 1),
+	}
+
+	tagsOrigin := models.Tags{}
+	tagsOrigin.SetString("owner", "\u00a0this owner\nuses invalid chars\t")
+
+	connection.addPoint("name", tagsOrigin, models.Fields{"clients.total": 10}, time.Now())
+	point := <-connection.points
+	assert.NotNil(point)
+	tags := point.Tags()
+	assert.NotNil(tags)
+	assert.Equal(tags["owner"], " this owner uses invalid chars ")
+}
+
+func TestAddPointWithValidCharacters(t *testing.T) {
+	assert := assert.New(t)
+
+	connection := &Connection{
+		config: map[string]interface{}{},
+		points: make(chan *client.Point, 1),
+	}
+
+	tagsOrigin := models.Tags{}
+	tagsOrigin.SetString("owner", "📶this owner uses only\u0020valid chars🛜")
+
+	connection.addPoint("name", tagsOrigin, models.Fields{"clients.total": 10}, time.Now())
+	point := <-connection.points
+	assert.NotNil(point)
+	tags := point.Tags()
+	assert.NotNil(tags)
+	assert.Equal(tags["owner"], "📶this owner uses only\u0020valid chars🛜")
+}