Search/get-searchxng.go

138 lines
3.8 KiB
Go
Raw Permalink Normal View History

2024-06-08 21:06:56 +00:00
package main
import (
"encoding/json"
"fmt"
"log"
"net/http"
"strings"
"sync"
"time"
)
const SEARX_SPACE_URL = "https://searx.space/data/instances.json"
type SearXInstance struct {
URL string `json:"url"`
Status int `json:"status_code"`
SSLGrade string `json:"grade"`
}
type SearXInstanceMetadata struct {
Timestamp int64 `json:"timestamp"`
}
type SearXInstanceResponse struct {
Metadata SearXInstanceMetadata `json:"metadata"`
Instances map[string]map[string]interface{} `json:"instances"`
}
var searxInstances []SearXInstance
var searxInstanceLastFetched time.Time
var searxInstanceFetchLock sync.Mutex
var backupInstances = []SearXInstance{
{URL: "https://searx.ox2.fr/", Status: 200, SSLGrade: "A+"},
{URL: "https://search.datura.network/", Status: 200, SSLGrade: "A+"},
{URL: "https://searx.foss.family/", Status: 200, SSLGrade: "A+"},
// Add more backup instances as needed
}
func fetchSearXInstances() error {
resp, err := http.Get(SEARX_SPACE_URL)
if err != nil {
return fmt.Errorf("error fetching SearX instances: %w", err)
}
defer resp.Body.Close()
if resp.StatusCode != http.StatusOK {
return fmt.Errorf("unexpected status code: %d", resp.StatusCode)
}
var instanceResp SearXInstanceResponse
if err := json.NewDecoder(resp.Body).Decode(&instanceResp); err != nil {
return fmt.Errorf("error decoding SearX instance response: %w", err)
}
searxInstances = make([]SearXInstance, 0)
for url, instanceData := range instanceResp.Instances {
if httpData, ok := instanceData["http"].(map[string]interface{}); ok {
if status, ok := httpData["status_code"].(float64); ok && int(status) == 200 {
if grade, ok := httpData["grade"].(string); ok && grade != "" {
searxInstances = append(searxInstances, SearXInstance{
URL: url,
Status: int(status),
SSLGrade: grade,
})
}
}
}
}
log.Printf("Fetched %d SearX instances: %+v", len(searxInstances), searxInstances)
searxInstanceLastFetched = time.Now()
return nil
}
func getRandomSearXInstance() (SearXInstance, error) {
searxInstanceFetchLock.Lock()
defer searxInstanceFetchLock.Unlock()
if searxInstances == nil || time.Since(searxInstanceLastFetched) > 24*time.Hour {
if err := fetchSearXInstances(); err != nil {
log.Printf("Error fetching instances, using backup instances: %v", err)
searxInstances = backupInstances
}
}
if len(searxInstances) == 0 {
return SearXInstance{}, fmt.Errorf("no available SearX instances")
}
for _, instance := range searxInstances {
if isInstanceValid(instance) {
log.Printf("Selected SearX instance: %+v", instance)
return instance, nil
}
}
return SearXInstance{}, fmt.Errorf("no valid SearX instances found")
}
func isInstanceValid(instance SearXInstance) bool {
searchURL := fmt.Sprintf("%s/search?q=test&categories=general&language=en&safe_search=1&page=1&format=json", strings.TrimRight(instance.URL, "/"))
client := &http.Client{
Timeout: 10 * time.Second,
}
req, err := http.NewRequest("GET", searchURL, nil)
if err != nil {
log.Printf("Instance validation failed for URL: %s, Error: %v", searchURL, err)
return false
}
req.Header.Set("User-Agent", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36")
resp, err := client.Do(req)
if err != nil {
log.Printf("Instance validation failed for URL: %s, Error: %v", searchURL, err)
return false
}
defer resp.Body.Close()
if resp.StatusCode == http.StatusOK {
return true
} else {
log.Printf("Instance validation failed for URL: %s, StatusCode: %d", searchURL, resp.StatusCode)
return false
}
}
2024-06-09 10:43:46 +00:00
// func main() {
// instance, err := getRandomSearXInstance()
// if err != nil {
// log.Fatalf("Failed to get a SearX instance: %v", err)
// }
// fmt.Printf("Selected SearX instance: %s\n", instance.URL)
// }