feat: add regex extraction

README.md

@@ -25,6 +25,9 @@ $ curl "http://localhost:7979/probe?module=default&target=http://localhost:8000/
 # HELP example_global_value Example of a top-level global value scrape in the json
 # TYPE example_global_value untyped
 example_global_value{environment="beta",location="planet-mars"} 1234
+# HELP example_regex_value Example of a regex value scrapes in the json
+# TYPE example_regex_value gauge
+example_regex_value{environment="beta"} 1000
 # HELP example_timestamped_value_count Example of a timestamped value scrape in the json
 # TYPE example_timestamped_value_count untyped
 example_timestamped_value_count{environment="beta"} 2

cmd/main.go

@@ -121,7 +121,7 @@ func probeHandler(w http.ResponseWriter, r *http.Request, logger *slog.Logger, c
 
 	registry := prometheus.NewPedanticRegistry()
 
-	metrics, err := exporter.CreateMetricsList(config.Modules[module])
+	metrics, err := exporter.CreateMetricsList(config.Modules[module], logger)
 	if err != nil {
 		logger.Error("Failed to create metrics list from config", "err", err)
 	}

cmd/main_test.go

@@ -470,3 +470,28 @@ func TestBodyPostQuery(t *testing.T) {
 		target.Close()
 	}
 }
+func TestInvalidRegexConfig(t *testing.T) {
+	_, err := config.LoadConfig("../test/config/invalidConfig.yml")
+	if err != nil {
+		return
+	}
+
+	target := httptest.NewServer(http.FileServer(http.Dir("../test")))
+	defer target.Close()
+
+	c, err := config.LoadConfig("../test/config/invalidConfig.yml")
+	if err != nil {
+		t.Fatalf("Failed to load config file")
+	}
+
+	req := httptest.NewRequest("GET", "http://example.com/foo"+"?module=default&target="+target.URL+"/serve/good.json", nil)
+	recorder := httptest.NewRecorder()
+	logBuffer := strings.Builder{}
+	promslogConfig := &promslog.Config{Writer: &logBuffer}
+	logger := promslog.New(promslogConfig)
+	probeHandler(recorder, req, logger, c)
+
+	if logBuffer.Len() == 0 {
+		t.Fatal("Expected error log for invalid regex, but got none")
+	}
+}

config/config.go

@@ -30,6 +30,7 @@ type Metric struct {
 	EpochTimestamp  string
 	Help            string
 	Values          map[string]string
+	IncludeRegex    string
 	AllowMissingKey bool `yaml:"allow_missing_key,omitempty"`
 }
 

examples/config.yml

@@ -30,6 +30,13 @@ modules:
         active: 1         # static value
         count: '{.count}' # dynamic value
         boolean: '{.some_boolean}'
+    - name: example_regex_value
+      valuetype: gauge
+      help: Example of a regex value scrapes in the json
+      path: '{ .available_memory }'
+      labels:
+        environment: beta # static label
+      regex: ^\d*\.?\d*   # only match digits
     - name: example_missing_key
       path: '{ .missing_key }'
       allow_missing_key: true

examples/data.json

@@ -21,5 +21,6 @@
             "state": "ACTIVE"
         }
     ],
-    "location": "mars"
+    "location": "mars",
+    "available_memory": "1000 MB"
 }

exporter/collector.go

@@ -17,6 +17,7 @@ import (
 	"bytes"
 	"encoding/json"
 	"log/slog"
+	"regexp"
 	"time"
 
 	"github.com/prometheus-community/json_exporter/config"
@@ -38,6 +39,7 @@ type JSONMetric struct {
 	LabelsJSONPaths        []string
 	ValueType              prometheus.ValueType
 	EpochTimestampJSONPath string
+	IncludeRegex           *regexp.Regexp
 	AllowMissingKey        bool
 }
 
@@ -59,7 +61,12 @@ func (mc JSONMetricCollector) Collect(ch chan<- prometheus.Metric) {
 			if missing {
 				continue
 			}
-
+		if m.IncludeRegex != nil {
+			if value = m.IncludeRegex.FindString(value); value == "" {
+				mc.Logger.Error("No matching for this pattern", "pattern", m.IncludeRegex.String())
+				continue
+			}
+		}
 			if floatValue, err := SanitizeValue(value); err == nil {
 				metric := prometheus.MustNewConstMetric(
 					m.Desc,

exporter/util.go

@@ -22,6 +22,7 @@ import (
 	"math"
 	"net/http"
 	"net/url"
+	"regexp"
 	"strconv"
 	"strings"
 	"text/template"
@@ -73,7 +74,7 @@ func SanitizeIntValue(s string) (int64, error) {
 	return value, errors.New(resultErr)
 }
 
-func CreateMetricsList(c config.Module) ([]JSONMetric, error) {
+func CreateMetricsList(c config.Module, logger *slog.Logger) ([]JSONMetric, error) {
 	var (
 		metrics   []JSONMetric
 		valueType prometheus.ValueType
@@ -90,10 +91,18 @@ func CreateMetricsList(c config.Module) ([]JSONMetric, error) {
 		switch metric.Type {
 		case config.ValueScrape:
 			var variableLabels, variableLabelsValues []string
+			var err error
+			var re *regexp.Regexp
 			for k, v := range metric.Labels {
 				variableLabels = append(variableLabels, k)
 				variableLabelsValues = append(variableLabelsValues, v)
 			}
+			if metric.IncludeRegex != "" {
+				if re, err = regexp.Compile(metric.IncludeRegex); err != nil {
+					logger.Error("Invalid regex expression", "err", err)
+					continue
+				}
+			}
 			jsonMetric := JSONMetric{
 				Type: config.ValueScrape,
 				Desc: prometheus.NewDesc(
@@ -106,6 +115,7 @@ func CreateMetricsList(c config.Module) ([]JSONMetric, error) {
 				LabelsJSONPaths:        variableLabelsValues,
 				ValueType:              valueType,
 				EpochTimestampJSONPath: metric.EpochTimestamp,
+				IncludeRegex:           re,
 				AllowMissingKey:        metric.AllowMissingKey,
 			}
 			metrics = append(metrics, jsonMetric)

test/config/good.yml

@@ -22,3 +22,9 @@ modules:
         active: 1      # static value
         count: '{.count}' # dynamic value
         boolean: '{.some_boolean}'
+
+    - name: example_regex_value
+      path: '{.availableMemory}'
+      help: Example of a regex filtered value scrape in the json
+      valuetype: gauge
+      includeregex: ^\d*\.?\d*

test/config/invalidConfig.yml

@@ -0,0 +1,9 @@
+---
+modules:
+  default:
+    metrics:
+      - name: available_memory
+        path: '{.availableMemory}'
+        help: Available memory in MB
+        valuetype: gauge
+        includeregex: ^\d*\.?\d*)  # Invalid regular expression, parentheses do not match

test/response/good.txt

@@ -1,6 +1,9 @@
 # HELP example_global_value Example of a top-level global value scrape in the json
 # TYPE example_global_value gauge
 example_global_value{environment="beta",location="planet-mars"} 1234
+# HELP example_regex_value Example of a regex filtered value scrape in the json
+# TYPE example_regex_value gauge
+example_regex_value 5
 # HELP example_value_active Example of sub-level value scrapes from a json
 # TYPE example_value_active counter
 example_value_active{environment="beta",id="id-A"} 1

test/serve/good.json

@@ -20,5 +20,6 @@
             "state": "ACTIVE"
         }
     ],
-    "location": "mars"
+    "location": "mars",
+    "availableMemory": "5 MB"
 }