11 Commits

Author SHA1 Message Date
Rustem Kamalov
10decd5740 Fix config argument, update deault config. 2026-04-04 19:32:23 +03:00
Rustem Kamalov
78a9f1e264 Merge pull request #22 from karust/dev
Add runtime proxy override, update readme
2026-04-04 18:57:56 +03:00
Rustem Kamalov
9f84a648f4 Update config and readme 2026-04-04 18:54:10 +03:00
Rustem Kamalov
3c07c7789a Runtime proxy override 2026-04-01 17:38:13 +03:00
Rustem Kamalov
3daa48e6a3 Proxy policy wiring and fail-closed execution
- replace proxy wiring with global and per-engine tag policies
- split stats endpoints and lock fail-closed proxy behavior with tests
2026-04-01 00:54:21 +03:00
Rustem Kamalov
8bec5578c0 Implement proxy support with configuration, pooling, and integration tests
- Added ProxyConfig and ProxyPool structures to manage proxy settings and rotation.
- Implemented normalization functions for proxy URLs and configurations.
- Created a new ProxyPool that supports failure tracking and round-robin selection.
- Integrated proxy handling into the ResilientSearcher for search queries.
- Added integration tests for various proxy scenarios including SOCKS5 and HTTP proxies.
- Updated server and resilience stats to include proxy information.
- Refactored search functions to utilize the new proxy client.
- Removed direct proxy handling from search_raw.go and yandex/search_raw.go, using the new core proxy client instead.
- Added unit tests for proxy configuration and pool behavior.
2026-03-30 22:46:03 +03:00
Rustem Kamalov
0146891076 Add bounded TTL cache for dedicated and mega endpoints 2026-03-30 00:31:49 +03:00
Rustem Kamalov
f219c78d84 Add retry/circuit breaker, configurable fallback and CORS.
Based in part on work from PR #21 by @Sai-Prashanth123, adapted and integrated with project-specific fixes.
2026-03-25 23:31:41 +03:00
Rustem Kamalov
0fcfc06baa Add healthcheck endpoint 2026-03-25 04:09:37 +03:00
Rustem Kamalov
a4d459761e Add custom browser path support 2026-03-17 02:18:51 +03:00
Rustem Kamalov
7cdd358cad Add filter param to Google. Add start param for supporting engines 2026-03-17 01:54:25 +03:00
41 changed files with 5381 additions and 691 deletions

3
.gitignore vendored
View File

@@ -25,3 +25,6 @@ logs.txt
.release
core/test/
.aider*
.gocache/
openserp
.gomodcache/

View File

@@ -16,8 +16,13 @@ RUN go build -o /app/openserp .
FROM zenika/alpine-chrome:with-chromedriver
WORKDIR /usr/src/app
COPY --from=builder /app/openserp /usr/local/bin/openserp
ADD config.yaml /usr/src/app
COPY config.yaml ./config.yaml
HEALTHCHECK --interval=30s --timeout=5s --start-period=10s --retries=3 \
CMD wget --no-verbose --tries=1 --spider http://localhost:7000/health || exit 1
ENTRYPOINT ["openserp"]

145
README.md
View File

@@ -1,4 +1,4 @@
# OpenSERP (Search Engine Results Page)
# OpenSERP (Search Engine Results)
![OpenSERP](/logo.svg)
@@ -6,18 +6,18 @@
[![Go Reference](https://pkg.go.dev/badge/github/karust/openserp?style=for-the-badge)](https://pkg.go.dev/github.com/karust/openserp)
[![release](https://img.shields.io/github/release/karust/openserp)](https://github.com/karust/openserp/releases)
<!--[![Docker Pulls](https://img.shields.io/docker/pulls/karust/openserp)](https://hub.docker.com/repository/docker/karust/openserp)-->
<!-- [![Docker Pulls](https://img.shields.io/docker/pulls/karust/openserp)](https://hub.docker.com/repository/docker/karust/openserp) -->
**OpenSERP** provides free API access to multiple search engines including **[Google, Yandex, Baidu, Bing, DuckDuckGo]**. Get comprehensive search results without expensive API subscriptions!
**OpenSERP** provides free API and CLI access to multiple search engines including **Google, Yandex, Baidu, Bing, and DuckDuckGo**. Get comprehensive search results without expensive API subscriptions!
## Features
- 🔍 **Multi-Engine Support**: Google, Yandex, Baidu, Bing, DuckDuckGo...
- 🌐 **Megasearch**: Aggregate results from multiple engines simultaneously
- 🖼 **Images**: Image search is also available!
- 🎯 **Advanced Filtering**: Language, date range, file type, site-specific searches
- 🌍 **Proxy Support**: HTTP/SOCKS5 proxy support
- 🐳 **Docker Ready**: Easy deployment with Docker
- 🔍 **Multi-engine** - search with dedicated endpoints for each engine
- 🌐 **Megasearch** - cross-engine aggregation with deduplication
- 🖼 **Images** - image search is also available
- 🎯 **Advanced filters** - language, date range, file type, and site queries
- 🌍 **Configurable** - proxy, cache, and resilient mode support
- 🐳 **Docker-ready** - local and container deployment
## Quick Start⚡
@@ -31,36 +31,36 @@ docker run -p 127.0.0.1:7000:7000 -it karust/openserp serve -a 0.0.0.0 -p 7000
docker compose up --build
```
### From Source
### From source
```bash
# Clone and build
git clone https://github.com/karust/openserp.git
cd openserp
go build -o openserp .
# Run the server
./openserp serve
```
## 🌐 Megasearch & Megaimage - Search Everything at Once!
## 🌐 Megasearch & Megaimage
**Megasearch** aggregates results from multiple engines simultaneously with automatic deduplication. **Megaimage** does the same for image searches!
### Megasearch (Web Results)
Search all engines at once:
```bash
# Search ALL engines at once
curl "http://localhost:7000/mega/search?text=golang&limit=10"
# Pick specific engines
curl "http://localhost:7000/mega/search?text=golang&engines=duckduckgo,bing&limit=15"
# Advanced filtering
curl "http://localhost:7000/mega/search?text=Donald+Trump&engines=duckduckgo,bing&limit=20&date=20251005..20251005&lang=EN"
curl "http://127.0.0.1:7000/mega/search?text=golang&limit=10"
```
- API response example:
Search only selected engines:
```bash
curl "http://127.0.0.1:7000/mega/search?text=golang&engines=duckduckgo,bing&limit=15"
```
Advanced filtering:
```bash
curl "http://127.0.0.1:7000/mega/search?text=Donald+Trump&engines=duckduckgo,bing&limit=20&date=20251005..20251005&lang=EN"
```
API response example:
```json
[
@@ -68,48 +68,46 @@ curl "http://localhost:7000/mega/search?text=Donald+Trump&engines=duckduckgo,bin
"rank": 1,
"url": "https://en.wikipedia.org/wiki/Golden_Retriever",
"title": "Golden Retriever - Wikipedia",
"description": "The Golden Retriever is a Scottish breed of retriever dog of medium size. It is characterised by a gentle and affectionate nature and a striking golden coat. It is a working dog, and registration is subject to successful completion of a working trial. [2] It is commonly kept as a companion dog and is among the most frequently registered breeds in several Western countries; some may compete in ...",
"description": "The Golden Retriever is a Scottish breed of retriever dog of medium size. It is characterised by a gentle and affectionate nature and a striking golden coat.",
"ad": false,
"engine": "duckduckgo"
},
{
"rank": 2,
"url": "https://www.bing.com/ck/a?!&&p=6f15ac4589858d0a104cd6f55cc8e91e8d8d6da91f905b626921f67f2323a467JmltdHM9MTc1OTE5MDQwMA&ptn=3&ver=2&hsh=4&fclid=2357c2f4-6131-68de-359f-d48c607c691d&u=a1aHR0cHM6Ly93d3cuZ29sZGVucmV0cmlldmVyZm9ydW0uY29tL3RocmVhZHMvdW5kZXJzdGFuZGluZy13aHktZ29sZGVuLXJldHJpZXZlciVFMiU4MCU5OXMtbGlmZXNwYW4taGFsdmVkLWluLXRoZS1sYXN0LTM1LXllYXJzLjM1NzMyMi8&ntb=1",
"title": "Golden Retriever Dog Forums\nhttps://www.goldenretrieverforum.com threads understanding-why-g…",
"description": "Oct 20, 2024 · Back in the 1970s, Golden Retrievers routinely lived until 16 and 17 years old, they are now living until 9 or 10 years old. Golden Retrievers seem to be dying mostly of bone …",
"url": "https://www.bing.com/ck/a?!&&p=6f15ac4589858d0a104cd6f55cc8",
"title": "Golden Retriever Dog Forums",
"description": "Oct 20, 2024 · Back in the 1970s, Golden Retrievers routinely lived until 16 and 17 years old, they are now...",
"ad": false,
"engine": "bing"
},
{
"rank": 3,
"url": "http://www.baidu.com/link?url=2544q3ugc68j0scVxdpWCSX-gl2AmuCy1l7uRR3loIfS1hmJWMiJKW4MDGWoZrLE7X-ybu1L7T8PspoL7iy_dK",
"title": "golden retrievers是什么意思_golden retrievers怎么读_解释_用法...",
"description": "\n\n2025年9月21日golden retrievers 读音:美英 golden retrievers基本解释 金毛猎犬 分词解释 golden金(黄)色的 retrievers寻猎物犬( retriever的名词复数 ) 词组短语 golden retrieversfor sale出售金毛寻回犬 golden retrieversnear me我附近的金毛寻回犬 golden retrieverspuppies金毛寻回犬幼犬...\ndanci.gei6.com/golden...retrievers...",
"url": "http://www.baidu.com/link?url==2544q3ugc68j0scVxdpWCSX-gl2AmuCy1l7uRR3loIfS1",
"title": "golden retrievers是什么意思",
"description": "2025年9月21日golden retrievers 读音:美英 golden retrievers基本解释 金毛猎犬 分词解释 golden金(黄)色的...",
"ad": false,
"engine": "baidu"
}
]
```
### Megaimage (Image Results)
Image search:
```bash
# Search images across ALL engines
curl "http://localhost:7000/mega/image?text=golang logo&limit=20"
curl "http://127.0.0.1:7000/mega/image?text=golang logo&limit=20"
```
### Available Engines
List available engines:
```bash
# Check which engines are available
curl "http://localhost:7000/mega/engines"
curl "http://127.0.0.1:7000/mega/engines"
```
**Available engines:** `google`, `yandex`, `baidu`, `bing`, `duckduckgo`
## 🔍 Individual Engine APIs
### Search Parameters
Common query parameters:
| Parameter | Description | Example |
| --------- | -------------------- | --------------------------------- |
@@ -119,52 +117,63 @@ curl "http://localhost:7000/mega/engines"
| `file` | File extension | `PDF`, `DOC`, `XLS` |
| `site` | Site-specific search | `github.com`, `stackoverflow.com` |
| `limit` | Number of results | `10`, `25`, `50` |
| `answers` | Include Q&A results | `true`, `false` |
### Individual Engine Examples
Engine-specific parameters:
| Parameter | Supported engines | Notes |
| --------- | ----------------------------------- | ------------------------------------------------------------------ |
| `start` | `google`, `bing`, `yandex`, `baidu` | Web search pagination offset. |
| `filter` | `google` | Duplicate filter (`true` hides similar, `false` includes similar). |
| `answers` | `google` | Include Google answer boxes in output with negative ranks. |
Examples:
```bash
# DuckDuckGo search
curl "http://localhost:7000/duck/search?text=golang&limit=7"
# Google search
curl "http://localhost:7000/google/search?text=golang&lang=EN&limit=10"
curl "http://127.0.0.1:7000/duck/search?text=golang&limit=7"
curl "http://127.0.0.1:7000/google/search?text=golang&lang=EN&limit=10"
curl "http://127.0.0.1:7000/bing/search?text=golang&limit=10&start=20"
curl "http://127.0.0.1:7000/yandex/search?text=golang&limit=10&start=10"
curl "http://127.0.0.1:7000/bing/image?text=golang&limit=20"
```
### Image Search
## 🌍 Proxy Support
OpenSERP supports HTTP and SOCKS5 proxies.
Simple global proxy:
```bash
# Bing Images
curl "http://localhost:7000/bing/image?text=golang&limit=20"
# Baidu Images
curl "http://localhost:7000/baidu/image?text=golang&limit=15"
```
## 🌐 Proxy Support
OpenSERP supports HTTP and SOCKS5 proxies with authentication:
```bash
# SOCKS5 proxy
./openserp serve --proxy socks5://127.0.0.1:1080
# HTTP proxy with authentication
./openserp search bing "query" --proxy http://user:pass@127.0.0.1:8080
```
Advanced proxy configuration is available in [config.yaml](./config.yaml).
You can enable tagged proxy pools and per-request override via `X-Use-Proxy: <tag>` or `X-Use-Proxy: direct`.
## Health & Stats
```bash
curl -i "http://127.0.0.1:7000/health"
curl "http://127.0.0.1:7000/stats"
curl "http://127.0.0.1:7000/stats/cache"
curl "http://127.0.0.1:7000/stats/proxy"
curl "http://127.0.0.1:7000/stats/cb"
```
Useful response headers in server mode: `X-Cache`, `X-Fallback-Engine`,`X-Proxy-Mode`, `X-Proxy-Tag`, `X-Proxy-Used`
## License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
This project is licensed under the MIT License. See [LICENSE](LICENSE).
## 🤝 Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
Contributions are welcome. Please feel free to submit a pull request.
## 👾 Issues & Support
If you encounter any issues or have questions:
If you encounter issues or have questions:
- Open an issue on GitHub
- Check existing issues for solutions
- Review the documentation above
- Check existing issues for similar reports
- Review the documentation and example config

View File

@@ -131,7 +131,7 @@ func (baid *Baidu) Search(query core.Query) ([]core.SearchResult, error) {
}
desc = strings.ReplaceAll(desc, title, "")
gR := core.SearchResult{Rank: i + 1, URL: linkText.String(), Title: title, Description: desc}
gR := core.SearchResult{Rank: query.Start + i + 1, URL: linkText.String(), Title: title, Description: desc}
searchResults = append(searchResults, gR)
}

View File

@@ -1,65 +1,20 @@
package baidu
import (
"context"
"crypto/tls"
"fmt"
"net"
"net/http"
"net/url"
"strings"
"time"
"github.com/PuerkitoBio/goquery"
"github.com/corpix/uarand"
"github.com/karust/openserp/core"
"github.com/sirupsen/logrus"
utls "github.com/refraction-networking/utls"
)
func baiduRequest(searchURL string, query core.Query) (*http.Response, error) {
// Create HTTP transport with proxy
transport := &http.Transport{}
if query.ProxyURL != "" {
proxyUrl, err := url.Parse(query.ProxyURL)
if err != nil {
return nil, err
}
transport.Proxy = http.ProxyURL(proxyUrl)
}
// Set insecure TLS
if query.Insecure {
transport.TLSClientConfig = &tls.Config{InsecureSkipVerify: true}
}
transport.DialTLSContext = func(ctx context.Context, network, addr string) (net.Conn, error) {
dialer := &net.Dialer{}
rawConn, err := dialer.DialContext(ctx, network, addr)
if err != nil {
return nil, err
}
hostname := strings.Split(addr, ":")[0]
config := &utls.Config{
ServerName: hostname,
InsecureSkipVerify: query.Insecure,
}
uconn := utls.UClient(rawConn, config, utls.HelloChrome_Auto)
if err := uconn.Handshake(); err != nil {
rawConn.Close()
return nil, err
}
return uconn, nil
}
baseClient := &http.Client{
Transport: transport,
Timeout: time.Second * 10,
baseClient, err := core.NewRawHTTPClient(query)
if err != nil {
return nil, err
}
req, err := http.NewRequest("GET", searchURL, nil)
@@ -138,6 +93,11 @@ func Search(query core.Query) ([]core.SearchResult, error) {
if err != nil {
return nil, err
}
if query.Start > 0 {
for i := range results {
results[i].Rank = query.Start + i + 1
}
}
logrus.Debugf("Baidu Raw results : %v", results)
return core.DeduplicateResults(results), nil

View File

@@ -66,6 +66,13 @@ func BuildURL(q core.Query) (string, error) {
if q.Limit != 0 {
params.Add("rn", strconv.Itoa(q.Limit))
}
if q.Start < 0 {
return "", errors.New("incorrect start provided")
}
if q.Start > 0 {
// Baidu uses "pn" as result offset for pagination.
params.Add("pn", strconv.Itoa(q.Start))
}
if len(params.Get("wd")) == 0 {
return "", errors.New("Empty query built")

View File

@@ -140,7 +140,7 @@ func (bing *Bing) Search(query core.Query) ([]core.SearchResult, error) {
}
bing.logger.Info("Found %d results (%d ads)", totalResults, len(adElements))
rank := 0
rank := query.Start
for _, result := range organicElements {
srchRes := core.SearchResult{}

View File

@@ -43,8 +43,15 @@ func BuildURL(q core.Query) (string, error) {
params.Add("setlang", strings.ToLower(q.LangCode))
}
// Set result offset (pagination) - Bing uses "first" parameter
if q.Limit > 0 {
// Set result offset (pagination) - Bing uses "first" parameter.
// When first is present, Bing may ignore custom count and return default page size.
if q.Start < 0 {
return "", errors.New("incorrect start provided")
}
if q.Start > 0 {
// Bing uses 1-based first-result index for pagination.
params.Add("first", strconv.Itoa(q.Start+1))
} else if q.Limit > 0 {
params.Add("count", strconv.Itoa(q.Limit))
}

30
cmd/proxy_policy.go Normal file
View File

@@ -0,0 +1,30 @@
package cmd
import (
"strings"
"github.com/karust/openserp/core"
)
func buildEngineProxyPolicyMap() map[string]string {
return map[string]string{
"google": config.GoogleConfig.Proxy,
"yandex": config.YandexConfig.Proxy,
"baidu": config.BaiduConfig.Proxy,
"bing": config.BingConfig.Proxy,
"duckduckgo": config.DuckDuckGoConfig.Proxy,
}
}
func buildNormalizedProxyConfig(runtime string) (core.ProxyConfig, error) {
return core.NormalizeProxyConfig(core.ProxyConfig{
Runtime: runtime,
Proxies: config.Proxies,
EnginePolicies: buildEngineProxyPolicyMap(),
})
}
func resolveEngineProxyPolicy(proxyCfg core.ProxyConfig, engineName string) core.ProxyPolicy {
engineKey := strings.ToLower(strings.TrimSpace(engineName))
return core.ResolveEffectiveProxyPolicy(proxyCfg.Proxies.Global, proxyCfg.EnginePolicies[engineKey])
}

View File

@@ -2,6 +2,7 @@ package cmd
import (
"fmt"
"os"
"strconv"
"strings"
@@ -13,48 +14,104 @@ import (
)
const (
version = "0.5.4"
version = "0.6.0"
defaultConfigFilename = "config"
envPrefix = "OPENSERP"
)
type Config struct {
App AppConfig `mapstructure:"app"`
Config2Capcha Config2Captcha `mapstructure:"2captcha"`
GoogleConfig core.SearchEngineOptions `mapstructure:"google"`
YandexConfig core.SearchEngineOptions `mapstructure:"yandex"`
BaiduConfig core.SearchEngineOptions `mapstructure:"baidu"`
BingConfig core.SearchEngineOptions `mapstructure:"bing"`
DuckDuckGoConfig core.SearchEngineOptions `mapstructure:"duckduckgo"`
Server ServerConfig `mapstructure:"server"`
App AppConfig `mapstructure:"app"`
Proxies core.ProxiesConfig `mapstructure:"proxies"`
Cache CacheConfig `mapstructure:"cache"`
Resilience ResilienceConfig `mapstructure:"resilience"`
CircuitBreaker CircuitBreakerConfig `mapstructure:"circuit_breaker"`
CORS CORSConfig `mapstructure:"cors"`
Config2Capcha Config2Captcha `mapstructure:"2captcha"`
GoogleConfig EngineConfig `mapstructure:"google"`
YandexConfig EngineConfig `mapstructure:"yandex"`
BaiduConfig EngineConfig `mapstructure:"baidu"`
BingConfig EngineConfig `mapstructure:"bing"`
DuckDuckGoConfig EngineConfig `mapstructure:"duckduckgo"`
}
type Config2Captcha struct {
ApiKey string `mapstructure:"apikey"`
}
type AppConfig struct {
type ServerConfig struct {
Host string `mapstructure:"host"`
Port int `mapstructure:"port"`
Timeout int `mapstructure:"timeout"`
ConfigPath string `mapstructure:"config_path"`
IsBrowserHead bool `mapstructure:"head"`
IsLeaveHead bool `mapstructure:"leave_head"`
IsLeakless bool `mapstructure:"leakless"`
IsDebug bool `mapstructure:"debug"`
IsVerbose bool `mapstructure:"verbose"`
IsRawRequests bool `mapstructure:"raw_requests"`
ProxyURL string `mapstructure:"proxy"`
Insecure bool `mapstructure:"insecure"`
}
type AppConfig struct {
Timeout int `mapstructure:"timeout"`
BrowserPath string `mapstructure:"browser_path"`
IsBrowserHead bool `mapstructure:"head"`
IsLeaveHead bool `mapstructure:"leave_head"`
IsLeakless bool `mapstructure:"leakless"`
IsStealth bool `mapstructure:"stealth"`
}
type EngineConfig struct {
core.SearchEngineOptions `mapstructure:",squash"`
Proxy string `mapstructure:"proxy"`
}
type CacheConfig struct {
TTLSeconds int `mapstructure:"ttl_seconds"`
MaxSize int `mapstructure:"max_size"`
}
type ResilienceConfig struct {
MaxRetries int `mapstructure:"max_retries"`
AllowEndpointFallback bool `mapstructure:"allow_endpoint_fallback"`
}
type CircuitBreakerConfig struct {
Failures int `mapstructure:"failures"`
RecoverySeconds int `mapstructure:"recovery_seconds"`
Successes int `mapstructure:"successes"`
}
type CORSConfig struct {
Enabled bool `mapstructure:"enabled"`
AllowOrigins string `mapstructure:"allow_origins"`
AllowMethods string `mapstructure:"allow_methods"`
AllowHeaders string `mapstructure:"allow_headers"`
MaxAge int `mapstructure:"max_age"`
}
var config = Config{}
var flagToConfigKey = map[string]string{
"config": "app.config_path",
"leave": "app.leave_head",
"raw": "app.raw_requests",
"2captcha_key": "2captcha.apikey",
"host": "server.host",
"port": "server.port",
"timeout": "app.timeout",
"config": "server.config_path",
"browser-path": "app.browser_path",
"verbose": "server.verbose",
"debug": "server.debug",
"head": "app.head",
"leakless": "app.leakless",
"raw": "server.raw_requests",
"leave": "app.leave_head",
"2captcha_key": "2captcha.apikey",
"proxy": "proxies.global",
"stealth": "app.stealth",
"insecure": "server.insecure",
"cache_ttl": "cache.ttl_seconds",
"cache_max_size": "cache.max_size",
"max_retries": "resilience.max_retries",
"allow_endpoint_fallback": "resilience.allow_endpoint_fallback",
"cb_failures": "circuit_breaker.failures",
"cb_recovery": "circuit_breaker.recovery_seconds",
"cb_successes": "circuit_breaker.successes",
}
var RootCmd = &cobra.Command{
@@ -64,22 +121,15 @@ var RootCmd = &cobra.Command{
Version: version,
SilenceUsage: true,
PersistentPreRunE: func(cmd *cobra.Command, args []string) error {
core.InitLogger(config.App.IsVerbose, config.App.IsDebug)
err := initializeConfig(cmd)
if err != nil {
return err
}
core.InitLogger(config.Server.IsVerbose, config.Server.IsDebug)
logrus.Debugf("Final config: %+v", config)
return nil
},
// Run: func(cmd *cobra.Command, args []string) {
// // Working with OutOrStdout/OutOrStderr allows us to unit test our command easier
// //out := cmd.OutOrStdout()
// logrus.Trace("Config:", config)
// },
}
// Bind each cobra flag to its associated viper configuration (config file and environment variable)
@@ -121,14 +171,27 @@ func parseFlagValue(flg *pflag.Flag) (interface{}, error) {
// Initialize Viper
func initializeConfig(cmd *cobra.Command) error {
v := viper.New()
setConfigDefaults(v)
// Base name of the config file, without the file extension
v.SetConfigName(defaultConfigFilename)
v.AddConfigPath(".")
explicitConfigPath := strings.TrimSpace(cmd.Flag("config").Value.String())
if explicitConfigPath == "" {
explicitConfigPath = strings.TrimSpace(os.Getenv(envPrefix + "_SERVER_CONFIG_PATH"))
}
if explicitConfigPath != "" {
v.SetConfigFile(explicitConfigPath)
} else {
// Base name of the config file, without the file extension
v.SetConfigName(defaultConfigFilename)
v.AddConfigPath(".")
}
// 1. Config file (lowest priority). Return an error if we cannot parse the config file.
err := v.ReadInConfig()
if err != nil {
if explicitConfigPath != "" {
return fmt.Errorf("cannot read config %q: %w", explicitConfigPath, err)
}
err = fmt.Errorf("cannot read config: %v", err)
logrus.Warn(err)
}
@@ -145,32 +208,143 @@ func initializeConfig(cmd *cobra.Command) error {
// 3. Command flags (highest priority). Bind the current command's flags to viper
bindFlags(cmd, v)
if err := validateRemovedConfigPaths(v); err != nil {
return err
}
// Dump Viper values to config struct
if err := validateEngineProxyTags(v); err != nil {
return err
}
err = v.Unmarshal(&config)
if err != nil {
return fmt.Errorf("cannot unmarshall config: %v", err)
}
if config.App.IsDebug {
config.Proxies, err = core.NormalizeProxiesConfig(config.Proxies)
if err != nil {
return fmt.Errorf("invalid proxies config: %w", err)
}
if config.Server.IsDebug {
logrus.Debug("Viper config:")
v.Debug()
}
return nil
}
func validateEngineProxyTags(v *viper.Viper) error {
for _, engineName := range []string{"google", "yandex", "baidu", "bing", "duckduckgo"} {
key := engineName + ".proxy"
if !v.IsSet(key) {
continue
}
raw := v.Get(key)
tag, ok := raw.(string)
if !ok {
return fmt.Errorf("invalid %s.proxy config: proxy must be a string tag", engineName)
}
if _, err := core.NormalizeProxyTag(tag); err != nil {
return fmt.Errorf("invalid %s.proxy config: %w", engineName, err)
}
}
return nil
}
func validateRemovedConfigPaths(v *viper.Viper) error {
legacyKeys := map[string]string{
"app.proxy": "use proxies.global or proxies.entries with per-engine proxy tags instead",
"proxy_pool": "use proxies.entries and proxies.health.failure_threshold instead",
"proxy_pool.urls": "use proxies.entries instead",
"proxy_pool.failure_threshold": "use proxies.health.failure_threshold instead",
"app.host": "move to server.host",
"app.port": "move to server.port",
"app.debug": "move to server.debug",
"app.verbose": "move to server.verbose",
"app.raw_requests": "move to server.raw_requests",
"app.insecure": "move to server.insecure",
"proxies.defaults": "use proxies.global or per-engine proxy tags instead",
"proxies.defaults.mode": "use proxies.global or per-engine proxy tags instead",
"proxies.defaults.tag": "use per-engine proxy tags on each engine instead",
"google.proxy.mode": "use google.proxy: <tag> or omit it for direct mode",
"google.proxy.tag": "use google.proxy: <tag>",
"yandex.proxy.mode": "use yandex.proxy: <tag> or omit it for direct mode",
"yandex.proxy.tag": "use yandex.proxy: <tag>",
"baidu.proxy.mode": "use baidu.proxy: <tag> or omit it for direct mode",
"baidu.proxy.tag": "use baidu.proxy: <tag>",
"bing.proxy.mode": "use bing.proxy: <tag> or omit it for direct mode",
"bing.proxy.tag": "use bing.proxy: <tag>",
"duckduckgo.proxy.mode": "use duckduckgo.proxy: <tag> or omit it for direct mode",
"duckduckgo.proxy.tag": "use duckduckgo.proxy: <tag>",
}
for key, hint := range legacyKeys {
if v.IsSet(key) {
return fmt.Errorf("config key %q is removed in proxy v2: %s", key, hint)
}
}
return nil
}
func setConfigDefaults(v *viper.Viper) {
v.SetDefault("server.host", "127.0.0.1")
v.SetDefault("server.port", 7070)
v.SetDefault("server.debug", false)
v.SetDefault("server.verbose", false)
v.SetDefault("server.raw_requests", false)
v.SetDefault("server.insecure", false)
v.SetDefault("app.timeout", 30)
v.SetDefault("app.browser_path", "")
v.SetDefault("app.head", false)
v.SetDefault("app.leave_head", false)
v.SetDefault("app.leakless", false)
v.SetDefault("app.stealth", false)
v.SetDefault("proxies.entries", []interface{}{})
v.SetDefault("proxies.global", "")
v.SetDefault("proxies.health.failure_threshold", core.DefaultProxyFailureThreshold)
v.SetDefault("cache.ttl_seconds", 300)
v.SetDefault("cache.max_size", 1000)
// Keep stage2 defaults stable even when config file is absent.
v.SetDefault("resilience.max_retries", 3)
v.SetDefault("resilience.allow_endpoint_fallback", false)
v.SetDefault("circuit_breaker.failures", 5)
v.SetDefault("circuit_breaker.recovery_seconds", 60)
v.SetDefault("circuit_breaker.successes", 2)
v.SetDefault("cors.enabled", true)
v.SetDefault("cors.allow_origins", "*")
v.SetDefault("cors.allow_methods", "GET, POST, OPTIONS")
v.SetDefault("cors.allow_headers", "Origin, Content-Type, Accept, Authorization, X-Use-Proxy")
v.SetDefault("cors.max_age", 86400)
}
func init() {
RootCmd.PersistentFlags().IntVarP(&config.App.Port, "port", "p", 7070, "Port number to run server")
RootCmd.PersistentFlags().StringVarP(&config.App.Host, "host", "a", "127.0.0.1", "Host address to run server")
RootCmd.PersistentFlags().IntVarP(&config.Server.Port, "port", "p", 7070, "Port number to run server")
RootCmd.PersistentFlags().StringVarP(&config.Server.Host, "host", "a", "127.0.0.1", "Host address to run server")
RootCmd.PersistentFlags().IntVarP(&config.App.Timeout, "timeout", "t", 30, "Timeout to fail request")
RootCmd.PersistentFlags().StringVarP(&config.App.ConfigPath, "config", "c", "", "Configuration file path")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsVerbose, "verbose", "v", false, "Use verbose output")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsDebug, "debug", "d", false, "Use debug output. Disable headless browser")
RootCmd.PersistentFlags().StringVarP(&config.Server.ConfigPath, "config", "c", "", "Configuration file path")
RootCmd.PersistentFlags().StringVarP(&config.App.BrowserPath, "browser-path", "", "", "Custom browser binary path (Chrome/Chromium/Edge/Brave..)")
RootCmd.PersistentFlags().BoolVarP(&config.Server.IsVerbose, "verbose", "v", false, "Use verbose output")
RootCmd.PersistentFlags().BoolVarP(&config.Server.IsDebug, "debug", "d", false, "Use debug output. Disable headless browser")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsBrowserHead, "head", "", false, "Enable browser UI")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsLeakless, "leakless", "l", false, "Use leakless mode to insure browser instances are closed after search")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsRawRequests, "raw", "r", false, "Disable browser usage, use HTTP requests")
RootCmd.PersistentFlags().BoolVarP(&config.Server.IsRawRequests, "raw", "r", false, "Disable browser usage, use HTTP requests")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsLeaveHead, "leave", "", false, "Leave browser and tabs opened after search is made")
RootCmd.PersistentFlags().StringVarP(&config.Config2Capcha.ApiKey, "2captcha_key", "", "", "2 captcha api key")
RootCmd.PersistentFlags().StringVarP(&config.App.ProxyURL, "proxy", "x", "", "HTTP or Socks5 proxy URL (e.g. http://user:pass@127.0.0.1:8080)")
RootCmd.PersistentFlags().StringVarP(&config.Proxies.Global, "proxy", "x", "", "Force a single proxy for all engines (same as proxies.global)")
RootCmd.PersistentFlags().BoolVarP(&config.App.IsStealth, "stealth", "s", false, "Use stealth browser plugin")
RootCmd.PersistentFlags().BoolVarP(&config.App.Insecure, "insecure", "k", false, "Allow insecure TLS connections")
RootCmd.PersistentFlags().BoolVarP(&config.Server.Insecure, "insecure", "k", false, "Allow insecure TLS connections")
RootCmd.PersistentFlags().IntVar(&config.Cache.TTLSeconds, "cache_ttl", 300, "Cache TTL in seconds (0 to disable)")
RootCmd.PersistentFlags().IntVar(&config.Cache.MaxSize, "cache_max_size", 1000, "Maximum number of cached responses")
RootCmd.PersistentFlags().IntVar(&config.Resilience.MaxRetries, "max_retries", 3, "Max retry attempts per search engine (0 to disable)")
RootCmd.PersistentFlags().BoolVar(&config.Resilience.AllowEndpointFallback, "allow_endpoint_fallback", false, "Allow dedicated endpoints to fallback to other engines")
RootCmd.PersistentFlags().IntVar(&config.CircuitBreaker.Failures, "cb_failures", 5, "Consecutive failures before circuit breaker opens")
RootCmd.PersistentFlags().IntVar(&config.CircuitBreaker.RecoverySeconds, "cb_recovery", 60, "Seconds before retrying an engine with open circuit")
RootCmd.PersistentFlags().IntVar(&config.CircuitBreaker.Successes, "cb_successes", 2, "Consecutive successful half-open checks needed to close circuit")
}

View File

@@ -25,24 +25,46 @@ var searchCMD = &cobra.Command{
}
func search(cmd *cobra.Command, args []string) {
var err error
engineType := args[0]
engineType := normalizeEngineArg(args[0])
query := core.Query{
Text: args[1],
Limit: 10,
ProxyURL: config.App.ProxyURL,
Insecure: config.App.Insecure,
Filter: true,
Insecure: config.Server.Insecure,
}
proxyRuntime := core.ProxyRuntimeBrowser
if config.Server.IsRawRequests {
proxyRuntime = core.ProxyRuntimeRaw
}
proxyCfg, err := buildNormalizedProxyConfig(proxyRuntime)
if err != nil {
logrus.Errorf("Error validating proxy config: %v", err)
return
}
policy := resolveEngineProxyPolicy(proxyCfg, engineType)
selectedProxy, err := selectCLIProxy(proxyCfg, policy)
if err != nil {
logrus.Errorf("Error selecting proxy for %s: %v", engineType, err)
return
}
if config.Server.IsRawRequests {
query.ProxyURL = selectedProxy
}
logrus.Infof("Starting SERP search request using %s engine for query: %s", engineType, query.Text)
var results []core.SearchResult
if config.App.IsRawRequests {
if config.Server.IsRawRequests {
logrus.Infof("Using raw requests mode for %s search", engineType)
results, err = searchRaw(engineType, query)
} else {
logrus.Infof("Using browser mode for %s search", engineType)
results, err = searchBrowser(engineType, query)
results, err = searchBrowser(engineType, query, selectedProxy)
}
if err != nil {
@@ -60,42 +82,51 @@ func search(cmd *cobra.Command, args []string) {
fmt.Println(string(b))
}
func searchBrowser(engineType string, query core.Query) ([]core.SearchResult, error) {
func searchBrowser(engineType string, query core.Query, browserProxyURL string) ([]core.SearchResult, error) {
var engine core.SearchEngine
if core.IsAuthenticatedSocksProxyURL(browserProxyURL) {
return nil, fmt.Errorf(
"%w: browser runtime does not support authenticated SOCKS proxy %s",
core.ErrProxyUnavailable,
core.MaskProxyURL(browserProxyURL),
)
}
opts := core.BrowserOpts{
IsHeadless: !config.App.IsBrowserHead, // Disable headless if browser head mode is set
IsHeadless: !config.App.IsBrowserHead,
IsLeakless: config.App.IsLeakless,
Timeout: time.Second * time.Duration(config.App.Timeout),
LeavePageOpen: config.App.IsLeaveHead,
CaptchaSolverApiKey: config.Config2Capcha.ApiKey,
ProxyURL: config.App.ProxyURL,
Insecure: config.App.Insecure,
BrowserPath: config.App.BrowserPath,
ProxyURL: browserProxyURL,
Insecure: config.Server.Insecure,
UseStealth: config.App.IsStealth,
}
if config.App.IsDebug {
if config.Server.IsDebug {
opts.IsHeadless = false
}
browser, err := core.NewBrowser(opts)
if err != nil {
logrus.Error(err)
return nil, err
}
switch strings.ToLower(engineType) {
case "yandex":
engine = yandex.New(*browser, config.YandexConfig)
engine = yandex.New(*browser, config.YandexConfig.SearchEngineOptions)
case "google":
engine = google.New(*browser, config.GoogleConfig)
engine = google.New(*browser, config.GoogleConfig.SearchEngineOptions)
case "baidu":
engine = baidu.New(*browser, config.BaiduConfig)
engine = baidu.New(*browser, config.BaiduConfig.SearchEngineOptions)
case "bing":
engine = bing.New(*browser, config.BingConfig)
case "duck":
engine = duckduckgo.New(*browser, config.DuckDuckGoConfig)
engine = bing.New(*browser, config.BingConfig.SearchEngineOptions)
case "duckduckgo":
engine = duckduckgo.New(*browser, config.DuckDuckGoConfig.SearchEngineOptions)
default:
logrus.Infof("No `%s` search engine found", engineType)
return nil, fmt.Errorf("no %q search engine found", engineType)
}
return engine.Search(query)
@@ -114,13 +145,42 @@ func searchRaw(engineType string, query core.Query) ([]core.SearchResult, error)
case "bing":
logrus.Warn("Bing does not support raw HTTP requests mode. Please use browser mode instead.")
return nil, fmt.Errorf("bing does not support raw requests mode")
case "duck":
case "duckduckgo":
logrus.Warn("DuckDuckGo does not support raw HTTP requests mode. Please use browser mode instead.")
return nil, fmt.Errorf("duckduckgo does not support raw requests mode")
default:
logrus.Infof("No `%s` search engine found", engineType)
return nil, fmt.Errorf("no %q search engine found", engineType)
}
}
func selectCLIProxy(proxyCfg core.ProxyConfig, policy core.ProxyPolicy) (string, error) {
if policy.Mode == core.ProxyModeOff {
return "", nil
}
if global := strings.TrimSpace(proxyCfg.Proxies.Global); global != "" {
return global, nil
}
if proxyCfg.Registry == nil {
return "", fmt.Errorf("%w: no proxy registry configured", core.ErrProxyUnavailable)
}
selected := proxyCfg.Registry.NextByTag(policy.Tag)
if selected == "" {
return "", fmt.Errorf("%w: no healthy proxy available for tag %q", core.ErrProxyUnavailable, policy.Tag)
}
return selected, nil
}
func normalizeEngineArg(raw string) string {
switch strings.ToLower(strings.TrimSpace(raw)) {
case "duck":
return "duckduckgo"
default:
return strings.ToLower(strings.TrimSpace(raw))
}
return nil, nil
}
func init() {

View File

@@ -2,6 +2,8 @@ package cmd
import (
"fmt"
"strings"
"sync"
"time"
"github.com/karust/openserp/baidu"
@@ -21,9 +23,7 @@ type rawEngine struct {
}
func (r *rawEngine) Search(q core.Query) ([]core.SearchResult, error) {
// Inject proxy settings from config
q.ProxyURL = config.App.ProxyURL
q.Insecure = config.App.Insecure
q.Insecure = config.Server.Insecure
switch r.name {
case "google":
@@ -63,52 +63,309 @@ var serveCMD = &cobra.Command{
}
func serve(cmd *cobra.Command, args []string) {
if config.App.IsRawRequests {
corsCfg := core.DefaultCORSConfig()
corsCfg.AllowOrigins = config.CORS.AllowOrigins
corsCfg.AllowMethods = config.CORS.AllowMethods
corsCfg.AllowHeaders = config.CORS.AllowHeaders
corsCfg.MaxAge = config.CORS.MaxAge
proxyRuntime := core.ProxyRuntimeBrowser
if config.Server.IsRawRequests {
proxyRuntime = core.ProxyRuntimeRaw
}
proxyCfg, err := buildNormalizedProxyConfig(proxyRuntime)
if err != nil {
logrus.Errorf("invalid proxy configuration: %v", err)
return
}
if config.Server.IsRawRequests {
logrus.Warn("Browserless results are very inconsistent or may not even work!")
serv := core.NewServer(config.App.Host, config.App.Port,
serverOpts := buildServerOptions(corsCfg, proxyCfg)
serv := core.NewServerWithOptions(config.Server.Host, config.Server.Port, serverOpts,
&rawEngine{name: "google"},
&rawEngine{name: "yandex"},
&rawEngine{name: "baidu"},
)
serv.Listen()
if err := serv.Listen(); err != nil {
logrus.Error(err)
}
return
}
opts := core.BrowserOpts{
IsHeadless: !config.App.IsBrowserHead, // Disable headless if browser head mode is set
baseOpts := core.BrowserOpts{
IsHeadless: !config.App.IsBrowserHead,
IsLeakless: config.App.IsLeakless,
Timeout: time.Second * time.Duration(config.App.Timeout),
LeavePageOpen: config.App.IsLeaveHead,
CaptchaSolverApiKey: config.Config2Capcha.ApiKey,
ProxyURL: config.App.ProxyURL,
Insecure: config.App.Insecure,
BrowserPath: config.App.BrowserPath,
Insecure: config.Server.Insecure,
UseStealth: config.App.IsStealth,
}
if config.App.IsDebug {
opts.IsHeadless = false
if config.Server.IsDebug {
baseOpts.IsHeadless = false
}
browser, err := core.NewBrowser(opts)
engines, err := buildBrowserEngines(baseOpts, proxyCfg)
if err != nil {
logrus.Error(err)
return
}
yand := yandex.New(*browser, config.YandexConfig)
gogl := google.New(*browser, config.GoogleConfig)
baidu := baidu.New(*browser, config.BaiduConfig)
bing := bing.New(*browser, config.BingConfig)
ddg := duckduckgo.New(*browser, config.DuckDuckGoConfig)
serv := core.NewServer(config.App.Host, config.App.Port, gogl, yand, baidu, bing, ddg)
err = serv.Listen()
if err != nil {
serverOpts := buildServerOptions(corsCfg, proxyCfg)
serv := core.NewServerWithOptions(config.Server.Host, config.Server.Port, serverOpts, engines...)
if err := serv.Listen(); err != nil {
logrus.Error(err)
}
}
func buildServerOptions(corsCfg core.CORSConfig, proxyCfg core.ProxyConfig) core.ServerOptions {
return core.ServerOptions{
CacheTTL: time.Duration(config.Cache.TTLSeconds) * time.Second,
CacheMaxSize: config.Cache.MaxSize,
EnableCORS: config.CORS.Enabled,
CORS: corsCfg,
AllowEndpointFallback: config.Resilience.AllowEndpointFallback,
Resilience: core.ResilientConfig{
Retry: core.RetryConfig{
MaxRetries: config.Resilience.MaxRetries,
InitialBackoff: 1 * time.Second,
MaxBackoff: 30 * time.Second,
BackoffFactor: 2.0,
},
CircuitBreaker: core.CircuitBreakerConfig{
FailureThreshold: config.CircuitBreaker.Failures,
RecoveryTimeout: time.Duration(config.CircuitBreaker.RecoverySeconds) * time.Second,
SuccessThreshold: config.CircuitBreaker.Successes,
},
Proxy: proxyCfg,
},
}
}
type browserPool struct {
mu sync.Mutex
base core.BrowserOpts
browser map[string]*core.Browser
}
func newBrowserPool(base core.BrowserOpts) *browserPool {
return &browserPool{
base: base,
browser: map[string]*core.Browser{},
}
}
func (p *browserPool) get(proxyURL string) (*core.Browser, error) {
key := strings.TrimSpace(proxyURL)
if key == "" {
key = "direct"
}
p.mu.Lock()
defer p.mu.Unlock()
if b, ok := p.browser[key]; ok {
return b, nil
}
opts := p.base
opts.ProxyURL = proxyURL
b, err := core.NewBrowser(opts)
if err != nil {
return nil, err
}
// Reuse one launched browser per unique effective proxy so startup stays lazy
// and engines with identical proxy policy don't spawn duplicate browser processes.
p.browser[key] = b
return b, nil
}
type pooledBrowserEngine struct {
name string
limiter *rate.Limiter
opts core.SearchEngineOptions
factory func(core.Browser, core.SearchEngineOptions) core.SearchEngine
pool *browserPool
mu sync.Mutex
engines map[string]core.SearchEngine
}
func (e *pooledBrowserEngine) Search(q core.Query) ([]core.SearchResult, error) {
engine, err := e.getOrCreate(q.ProxyURL)
if err != nil {
return nil, err
}
return engine.Search(q)
}
func (e *pooledBrowserEngine) SearchImage(q core.Query) ([]core.SearchResult, error) {
engine, err := e.getOrCreate(q.ProxyURL)
if err != nil {
return nil, err
}
return engine.SearchImage(q)
}
func (e *pooledBrowserEngine) IsInitialized() bool {
return true
}
func (e *pooledBrowserEngine) Name() string {
return e.name
}
func (e *pooledBrowserEngine) GetRateLimiter() *rate.Limiter {
return e.limiter
}
func (e *pooledBrowserEngine) getOrCreate(proxyURL string) (core.SearchEngine, error) {
key := strings.TrimSpace(proxyURL)
if key == "" {
key = "direct"
}
e.mu.Lock()
defer e.mu.Unlock()
if engine, ok := e.engines[key]; ok {
return engine, nil
}
browser, err := e.pool.get(proxyURL)
if err != nil {
return nil, err
}
engine := e.factory(*browser, e.opts)
e.engines[key] = engine
return engine, nil
}
type browserEngineSpec struct {
name string
opts core.SearchEngineOptions
factory func(core.Browser, core.SearchEngineOptions) core.SearchEngine
}
func browserEngineSpecs() []browserEngineSpec {
return []browserEngineSpec{
{
name: "google",
opts: config.GoogleConfig.SearchEngineOptions,
factory: func(browser core.Browser, opts core.SearchEngineOptions) core.SearchEngine {
return google.New(browser, opts)
},
},
{
name: "yandex",
opts: config.YandexConfig.SearchEngineOptions,
factory: func(browser core.Browser, opts core.SearchEngineOptions) core.SearchEngine {
return yandex.New(browser, opts)
},
},
{
name: "baidu",
opts: config.BaiduConfig.SearchEngineOptions,
factory: func(browser core.Browser, opts core.SearchEngineOptions) core.SearchEngine {
return baidu.New(browser, opts)
},
},
{
name: "bing",
opts: config.BingConfig.SearchEngineOptions,
factory: func(browser core.Browser, opts core.SearchEngineOptions) core.SearchEngine {
return bing.New(browser, opts)
},
},
{
name: "duckduckgo",
opts: config.DuckDuckGoConfig.SearchEngineOptions,
factory: func(browser core.Browser, opts core.SearchEngineOptions) core.SearchEngine {
return duckduckgo.New(browser, opts)
},
},
}
}
func buildBrowserEngines(baseOpts core.BrowserOpts, proxyCfg core.ProxyConfig) ([]core.SearchEngine, error) {
pool := newBrowserPool(baseOpts)
specs := browserEngineSpecs()
engines := make([]core.SearchEngine, 0, len(specs))
for _, spec := range specs {
policy := resolveEngineProxyPolicy(proxyCfg, spec.name)
if err := validateBrowserProxyPolicy(proxyCfg, policy); err != nil {
return nil, fmt.Errorf("browser proxy validation failed for engine %s: %w", spec.name, err)
}
opts := spec.opts
opts.Init()
engines = append(engines, &pooledBrowserEngine{
name: spec.name,
limiter: rate.NewLimiter(rate.Every(opts.GetRatelimit()), opts.RateBurst),
opts: opts,
factory: spec.factory,
pool: pool,
engines: map[string]core.SearchEngine{},
})
}
return engines, nil
}
func validateBrowserProxyPolicy(proxyCfg core.ProxyConfig, policy core.ProxyPolicy) error {
if policy.Mode != core.ProxyModeTagPool {
return nil
}
proxyURL := strings.TrimSpace(proxyCfg.Proxies.Global)
if proxyURL != "" {
return validateBrowserProxyURL(proxyURL)
}
for _, entry := range proxyCfg.Proxies.Entries {
if !entryHasTag(entry, policy.Tag) {
continue
}
if err := validateBrowserProxyURL(entry.URL); err != nil {
return err
}
}
return nil
}
func validateBrowserProxyURL(proxyURL string) error {
// Browser startup must stop immediately on authenticated SOCKS because Chrome
// cannot use that proxy shape reliably and retrying a different proxy hides the misconfiguration.
if core.IsAuthenticatedSocksProxyURL(proxyURL) {
return fmt.Errorf(
"%w: browser runtime does not support authenticated SOCKS proxy %s",
core.ErrProxyUnavailable,
core.MaskProxyURL(proxyURL),
)
}
return nil
}
func entryHasTag(entry core.ProxyEntryConfig, tag string) bool {
tag = strings.TrimSpace(strings.ToLower(tag))
if tag == "" {
return false
}
for _, entryTag := range entry.Tags {
if strings.TrimSpace(strings.ToLower(entryTag)) == tag {
return true
}
}
return false
}
func init() {
RootCmd.AddCommand(serveCMD)
}

105
cmd/serve_test.go Normal file
View File

@@ -0,0 +1,105 @@
package cmd
import (
"strings"
"testing"
"github.com/karust/openserp/core"
)
func TestValidateBrowserProxyPolicyRejectsAuthenticatedSocks(t *testing.T) {
tests := []struct {
name string
proxyCfg core.ProxyConfig
policy core.ProxyPolicy
}{
{
name: "global authenticated socks",
proxyCfg: core.ProxyConfig{
Proxies: core.ProxiesConfig{
Global: "socks5h://user:pass@127.0.0.1:1080",
},
},
policy: core.ProxyPolicy{Mode: core.ProxyModeTagPool},
},
{
name: "tag pool authenticated socks",
proxyCfg: core.ProxyConfig{
Proxies: core.ProxiesConfig{
Entries: []core.ProxyEntryConfig{
{URL: "socks5://user:pass@127.0.0.1:1080", Tags: []string{"us"}},
},
},
},
policy: core.ProxyPolicy{Mode: core.ProxyModeTagPool, Tag: "us"},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
err := validateBrowserProxyPolicy(tt.proxyCfg, tt.policy)
if err == nil {
t.Fatal("expected browser proxy validation to fail")
}
if !strings.Contains(err.Error(), "authenticated SOCKS proxy") {
t.Fatalf("expected explicit authenticated SOCKS error, got %v", err)
}
})
}
}
func TestValidateBrowserProxyPolicyAllowsHTTPAuthAndPlainSocks(t *testing.T) {
tests := []struct {
name string
proxyCfg core.ProxyConfig
policy core.ProxyPolicy
}{
{
name: "global http auth",
proxyCfg: core.ProxyConfig{
Proxies: core.ProxiesConfig{
Global: "http://user:pass@127.0.0.1:8080",
},
},
policy: core.ProxyPolicy{Mode: core.ProxyModeTagPool},
},
{
name: "tag pool plain socks",
proxyCfg: core.ProxyConfig{
Proxies: core.ProxiesConfig{
Entries: []core.ProxyEntryConfig{
{URL: "socks5://127.0.0.1:1080", Tags: []string{"eu"}},
},
},
},
policy: core.ProxyPolicy{Mode: core.ProxyModeTagPool, Tag: "eu"},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if err := validateBrowserProxyPolicy(tt.proxyCfg, tt.policy); err != nil {
t.Fatalf("expected browser proxy validation to succeed, got %v", err)
}
})
}
}
func TestValidateBrowserProxyPolicyRejectsTaggedAuthenticatedSocksInPool(t *testing.T) {
proxyCfg := core.ProxyConfig{
Proxies: core.ProxiesConfig{
Entries: []core.ProxyEntryConfig{
{URL: "http://127.0.0.1:8080", Tags: []string{"default"}},
{URL: "socks5://user:pass@127.0.0.1:1080", Tags: []string{"default"}},
},
},
}
err := validateBrowserProxyPolicy(proxyCfg, core.ProxyPolicy{Mode: core.ProxyModeTagPool, Tag: "default"})
if err == nil {
t.Fatal("expected browser proxy validation to fail for tag pool")
}
if !strings.Contains(err.Error(), "authenticated SOCKS proxy") {
t.Fatalf("expected explicit authenticated SOCKS error, got %v", err)
}
}

View File

@@ -1,22 +1,60 @@
app:
host: 0.0.0.0
port: 7000
debug: false
verbose: true
timeout: 15
head: false
leakless: false
leave_head: false
stealth: false
insecure: true
server:
host: 0.0.0.0 # API host to bind
port: 7000 # API port to bind
debug: false # Enable debug logs and force browser UI mode
verbose: true # Enable info-level request logs
raw_requests: false # true = raw HTTP mode, false = browser mode
insecure: true # Allow insecure TLS connections
2captcha:
apikey: "123123123123123"
app:
timeout: 15 # Browser/search timeout in seconds
browser_path: "" # Custom browser binary path (chrome/chromium/edge..)
head: false # Show browser UI (headful mode)
leakless: false # Force browser process cleanup after request
leave_head: false # Keep tabs open after request for debugging
stealth: false # Enable stealth browser plugin
proxies:
# Force a single proxy for all engines.
# Same behavior as passing --proxy on the CLI.
#global: http://127.0.0.1:8080
# Advanced mode: define tagged proxy pools and opt engines in with `proxy: <tag>`.
#entries:
# - url: http://127.0.0.1:8080
# tags: [default, us]
# - url: socks5h://127.0.0.1:1080
# tags: [eu]
health:
failure_threshold: 3 # Disable proxy after this many consecutive failures
cache:
ttl_seconds: 60 # Dedicated endpoint cache TTL in seconds (0 disables cache)
max_size: 1000 # Maximum cached dedicated responses before oldest-entry eviction
resilience:
max_retries: 2 # Retry attempts per engine request (0 disables retries)
allow_endpoint_fallback: false # Keep dedicated endpoints engine-pure by default
# circuit_breaker:
# failures: 5 # Consecutive failures required to open circuit
# recovery_seconds: 60 # Wait time before moving open circuit to half-open
# successes: 2 # Consecutive half-open successes required to close circuit
cors:
enabled: true
allow_origins: "*"
allow_methods: "GET, POST, OPTIONS"
allow_headers: "Origin, Content-Type, Accept, Authorization, X-Use-Proxy"
max_age: 86400
# 2captcha:
# apikey: "123123123123123"
google:
rate_requests: 4 # Number of requests per Minute
rate_burst: 2 # Number of non-ratelimited requests per Minute
captcha: true
rate_requests: 4 # Allowed average requests per minute
rate_burst: 2 # Burst requests before limiter applies
captcha: true # Enable captcha solver path
yandex:
rate_requests: 4
@@ -25,10 +63,12 @@ yandex:
baidu:
rate_requests: 4
rate_burst: 2
# No proxy tag means direct traffic
bing:
rate_requests: 4
rate_burst: 2
# No proxy tag means direct traffic
duckduckgo:
rate_requests: 4

View File

@@ -1,8 +1,11 @@
package core
import (
"context"
"errors"
"fmt"
"net/url"
"os"
"strings"
"time"
@@ -23,6 +26,7 @@ type BrowserOpts struct {
LeavePageOpen bool // Leave pages and browser open
WaitLoadTime time.Duration // Time to wait till page loads
CaptchaSolverApiKey string // 2Captcha api key
BrowserPath string // Explicit browser executable path
ProxyURL string // Proxy URL
Insecure bool // Allow insecure TLS connections
UseStealth bool // Use go-rod stealth plugin
@@ -51,35 +55,45 @@ func NewBrowser(opts BrowserOpts) (*Browser, error) {
opts.Check()
logrus.Debugf("Browser options: %+v", opts)
path, has := launcher.LookPath()
logrus.Debug("Browser found: ", has)
path, err := resolveBrowserBinaryPath(opts.BrowserPath, launcher.LookPath)
if err != nil {
return nil, err
}
// Create launcher
l := launcher.New().Bin(path).Leakless(opts.IsLeakless).Headless(opts.IsHeadless).Set("disable-blink-features", "AutomationControlled").
l := launcher.New().Leakless(opts.IsLeakless).Headless(opts.IsHeadless).Set("disable-blink-features", "AutomationControlled").
Delete("enable-automation")
if path != "" {
logrus.Debugf("Using browser binary: %s", path)
l = l.Bin(path)
}
// Configure proxy if specified
if opts.ProxyURL != "" {
normalizedProxyURL, err := NormalizeProxyURL(opts.ProxyURL)
if err != nil {
return nil, fmt.Errorf("invalid proxy URL: %v", err)
}
opts.ProxyURL = normalizedProxyURL
proxyUrl, err := url.Parse(opts.ProxyURL)
if err != nil {
return nil, fmt.Errorf("invalid proxy URL: %v", err)
}
// Make sure the proxy URL includes the scheme when passed to launcher
// This ensures proper handling of SOCKS5 proxies
proxyStr := proxyUrl.String()
logrus.Debugf("Setting up proxy: %s", proxyStr)
// Chrome's proxy-server flag must not contain credentials.
// Auth (if needed) is handled separately via DevTools auth callbacks.
proxyStr := proxyURLForBrowserLaunch(proxyUrl)
logrus.Debugf("Setting up proxy: %s", MaskProxyURL(proxyStr))
l = l.Proxy(proxyStr)
// Check if proxy has auth credentials
if proxyUrl.User != nil {
username := proxyUrl.User.Username()
logrus.Debugf("Using proxy authentication: %s:****", username)
// We'll handle auth in the Navigate method
logrus.Debugf("Proxy credentials configured for %s proxy: %s:****", proxyUrl.Scheme, username)
}
}
var err error
b := Browser{BrowserOpts: opts}
b.browserAddr, err = l.Launch()
@@ -91,6 +105,52 @@ func NewBrowser(opts BrowserOpts) (*Browser, error) {
return &b, err
}
func proxyURLForBrowserLaunch(u *url.URL) string {
if u == nil {
return ""
}
clone := *u
// Chrome expects socks5 scheme in --proxy-server; socks5h is not accepted.
if clone.Scheme == "socks5h" {
clone.Scheme = "socks5"
}
clone.User = nil
clone.Path = ""
clone.RawPath = ""
clone.RawQuery = ""
clone.Fragment = ""
return clone.String()
}
func validateBrowserBinaryPath(path string) error {
info, err := os.Stat(path)
if err != nil {
return err
}
if info.IsDir() {
return fmt.Errorf("path points to a directory")
}
return nil
}
// resolveBrowserBinaryPath prefers an explicit browser path. If no explicit path is provided,
// it falls back to launcher autodiscovery and lets Rod handle auto-download when no binary is found.
func resolveBrowserBinaryPath(browserPath string, lookPath func() (string, bool)) (string, error) {
if browserPath != "" {
if err := validateBrowserBinaryPath(browserPath); err != nil {
return "", fmt.Errorf("invalid browser_path %q: %w", browserPath, err)
}
return browserPath, nil
}
path, has := lookPath()
if has {
return path, nil
}
return "", nil
}
// Check whether browser instance is already created
func (b *Browser) IsInitialized() bool {
if b.browserAddr != "" {
@@ -104,9 +164,14 @@ func (b *Browser) IsInitialized() bool {
func (b *Browser) Navigate(URL string) (*rod.Page, error) {
logrus.Debug("Navigate to: ", URL)
b.browser = rod.New().ControlURL(b.browserAddr)
b.browser.MustConnect()
b.browser.SetCookies(nil)
browser := rod.New().ControlURL(b.browserAddr).Timeout(b.Timeout)
if err := browser.Connect(); err != nil {
return nil, fmt.Errorf("browser connect failed: %w", err)
}
b.browser = browser
if err := b.browser.SetCookies(nil); err != nil {
return nil, fmt.Errorf("browser cookie reset failed: %w", err)
}
// Handle proxy authentication before any navigations
if b.ProxyURL != "" {
@@ -114,39 +179,66 @@ func (b *Browser) Navigate(URL string) (*rod.Page, error) {
// Always ignore certificate errors when using proxies
// This fixes the ERR_CERT_AUTHORITY_INVALID error for SOCKS5 proxies
b.browser.MustIgnoreCertErrors(true)
if err := b.browser.IgnoreCertErrors(true); err != nil {
return nil, fmt.Errorf("configure proxy cert handling failed: %w", err)
}
if proxyUrl.User != nil {
if proxyUrl.User != nil && (proxyUrl.Scheme == "http" || proxyUrl.Scheme == "https") {
username := proxyUrl.User.Username()
password, _ := proxyUrl.User.Password()
// Launch auth handler before any navigation occurs
go b.browser.MustHandleAuth(username, password)()
go func() {
if err := b.browser.HandleAuth(username, password)(); err != nil {
logrus.Debugf("Proxy auth handler stopped: %v", err)
}
}()
} else if proxyUrl.User != nil && (proxyUrl.Scheme == "socks5" || proxyUrl.Scheme == "socks5h") {
// This callback handles HTTP proxy auth challenges; it doesn't authenticate SOCKS proxies.
logrus.Debug("SOCKS proxy credentials are not handled by browser auth callback")
}
} else if b.Insecure {
// Still respect the insecure flag if no proxy is used
b.browser.MustIgnoreCertErrors(true)
if err := b.browser.IgnoreCertErrors(true); err != nil {
return nil, fmt.Errorf("configure insecure mode failed: %w", err)
}
}
ua := strings.ReplaceAll(b.browser.MustVersion().UserAgent, "HeadlessChrome/", "Chrome/")
version, err := b.browser.Version()
if err != nil {
return nil, fmt.Errorf("read browser version failed: %w", err)
}
ua := strings.ReplaceAll(version.UserAgent, "HeadlessChrome/", "Chrome/")
var page *rod.Page
if b.UseStealth {
page = stealth.MustPage(b.browser)
page.MustEmulate(devices.Device{
page, err = stealth.Page(b.browser)
if err != nil {
return nil, fmt.Errorf("create stealth page failed: %w", err)
}
err = page.Emulate(devices.Device{
AcceptLanguage: b.LanguageCode,
UserAgent: ua,
})
if err != nil {
return nil, fmt.Errorf("emulate stealth page failed: %w", err)
}
} else {
page = b.browser.MustPage("about:blank")
page, err = b.browser.Page(proto.TargetCreateTarget{URL: "about:blank"})
if err != nil {
return nil, fmt.Errorf("create page failed: %w", err)
}
page.MustEmulate(devices.Device{
err = page.Emulate(devices.Device{
AcceptLanguage: b.LanguageCode,
UserAgent: ua,
})
if err != nil {
return nil, fmt.Errorf("emulate page failed: %w", err)
}
proto.EmulationSetDeviceMetricsOverride{
err = proto.EmulationSetDeviceMetricsOverride{
Width: 1920,
Height: 1080,
DeviceScaleFactor: 1,
@@ -154,25 +246,37 @@ func (b *Browser) Navigate(URL string) (*rod.Page, error) {
ScreenWidth: &[]int{1920}[0],
ScreenHeight: &[]int{1080}[0],
}.Call(page)
if err != nil {
return nil, fmt.Errorf("set device metrics failed: %w", err)
}
}
//EnableCustomStealth(page)
err := page.Navigate(URL)
timedPage := page.Timeout(b.Timeout)
err = timedPage.Navigate(URL)
if err != nil {
return nil, err
}
// Avoid panics from MustWaitLoad when the target navigates/closes mid-wait
if werr := page.WaitLoad(); werr != nil {
logrus.Debugf("WaitLoad returned early: %v", werr)
if werr := timedPage.WaitLoad(); werr != nil {
if errors.Is(werr, context.DeadlineExceeded) {
// Some engines keep loading background resources while the DOM is already usable.
// Treat load timeout as non-fatal and let engine-specific selector timeouts decide.
logrus.Debugf("WaitLoad timed out after %s; continuing with partial page state", b.Timeout)
} else {
logrus.Debugf("WaitLoad returned early: %v", werr)
}
}
wait := page.MustWaitRequestIdle()
// may cause bugs with google
if b.WaitRequests {
wait := timedPage.WaitRequestIdle(300*time.Millisecond, nil, nil, nil)
wait()
}
time.Sleep(2 * time.Second)
time.Sleep(b.WaitLoadTime)
return page, nil
}

140
core/cache.go Normal file
View File

@@ -0,0 +1,140 @@
package core
import (
"crypto/sha256"
"encoding/hex"
"fmt"
"sync"
"time"
)
type CacheEntry struct {
Data []byte
CreatedAt time.Time
ExpiresAt time.Time
}
// ResponseCache is a bounded in-memory TTL cache for dedicated endpoint responses.
type ResponseCache struct {
mu sync.Mutex
entries map[string]CacheEntry
ttl time.Duration
maxSize int
hits int
misses int
bypasses int
evictions int
}
func NewResponseCache(ttl time.Duration, maxSize int) *ResponseCache {
return &ResponseCache{
entries: make(map[string]CacheEntry),
ttl: ttl,
maxSize: maxSize,
}
}
func BuildCacheKey(engine string, action string, q Query) string {
raw := fmt.Sprintf(
"%s|%s|%s|%s|%s|%s|%s|%d|%d|%t|%t|%s",
engine,
action,
q.Text,
q.LangCode,
q.DateInterval,
q.Filetype,
q.Site,
q.Limit,
q.Start,
q.Filter,
q.Answers,
q.ProxyOverride,
)
hash := sha256.Sum256([]byte(raw))
return hex.EncodeToString(hash[:])
}
func (c *ResponseCache) Get(key string) ([]byte, bool) {
c.mu.Lock()
defer c.mu.Unlock()
c.pruneExpiredLocked(time.Now())
entry, ok := c.entries[key]
if !ok {
c.misses++
return nil, false
}
c.hits++
return entry.Data, true
}
func (c *ResponseCache) Set(key string, data []byte) {
c.mu.Lock()
defer c.mu.Unlock()
now := time.Now()
c.pruneExpiredLocked(now)
if _, exists := c.entries[key]; !exists && len(c.entries) >= c.maxSize {
c.evictOldestLocked()
}
c.entries[key] = CacheEntry{
Data: data,
CreatedAt: now,
ExpiresAt: now.Add(c.ttl),
}
}
func (c *ResponseCache) RecordBypass() {
c.mu.Lock()
defer c.mu.Unlock()
c.bypasses++
}
func (c *ResponseCache) Stats() map[string]interface{} {
c.mu.Lock()
defer c.mu.Unlock()
c.pruneExpiredLocked(time.Now())
return map[string]interface{}{
"status": true,
"entries": len(c.entries),
"hits": c.hits,
"misses": c.misses,
"bypasses": c.bypasses,
"evictions": c.evictions,
"ttl_seconds": int(c.ttl / time.Second),
"max_size": c.maxSize,
}
}
func (c *ResponseCache) pruneExpiredLocked(now time.Time) {
for key, entry := range c.entries {
if !now.Before(entry.ExpiresAt) {
delete(c.entries, key)
}
}
}
func (c *ResponseCache) evictOldestLocked() {
var (
oldestKey string
oldestCreated time.Time
)
for key, entry := range c.entries {
if oldestKey == "" || entry.CreatedAt.Before(oldestCreated) {
oldestKey = key
oldestCreated = entry.CreatedAt
}
}
if oldestKey != "" {
delete(c.entries, oldestKey)
c.evictions++
}
}

155
core/cache_test.go Normal file
View File

@@ -0,0 +1,155 @@
package core
import (
"testing"
"time"
)
func TestResponseCacheSetAndGet(t *testing.T) {
cache := NewResponseCache(5*time.Second, 10)
key := BuildCacheKey("google", "search", Query{Text: "golang", Limit: 10})
if _, ok := cache.Get(key); ok {
t.Fatal("expected initial cache miss")
}
data := []byte(`[{"rank":1}]`)
cache.Set(key, data)
got, ok := cache.Get(key)
if !ok {
t.Fatal("expected cache hit")
}
if string(got) != string(data) {
t.Fatalf("unexpected cached value: got %s want %s", got, data)
}
}
func TestResponseCacheExpiration(t *testing.T) {
cache := NewResponseCache(40*time.Millisecond, 10)
key := BuildCacheKey("google", "search", Query{Text: "expire"})
cache.Set(key, []byte(`[]`))
if _, ok := cache.Get(key); !ok {
t.Fatal("expected cache hit before expiration")
}
time.Sleep(60 * time.Millisecond)
if _, ok := cache.Get(key); ok {
t.Fatal("expected cache miss after expiration")
}
}
func TestResponseCacheEvictsOldestEntry(t *testing.T) {
cache := NewResponseCache(time.Minute, 2)
firstKey := BuildCacheKey("google", "search", Query{Text: "first"})
secondKey := BuildCacheKey("google", "search", Query{Text: "second"})
thirdKey := BuildCacheKey("google", "search", Query{Text: "third"})
cache.Set(firstKey, []byte(`["first"]`))
time.Sleep(10 * time.Millisecond)
cache.Set(secondKey, []byte(`["second"]`))
time.Sleep(10 * time.Millisecond)
cache.Set(thirdKey, []byte(`["third"]`))
if _, ok := cache.Get(firstKey); ok {
t.Fatal("expected oldest entry to be evicted")
}
if _, ok := cache.Get(secondKey); !ok {
t.Fatal("expected newer entry to remain cached")
}
if _, ok := cache.Get(thirdKey); !ok {
t.Fatal("expected newest entry to remain cached")
}
}
func TestResponseCacheStats(t *testing.T) {
cache := NewResponseCache(time.Minute, 2)
key := BuildCacheKey("google", "search", Query{Text: "stats"})
if _, ok := cache.Get(key); ok {
t.Fatal("expected miss for empty cache")
}
cache.Set(key, []byte(`[]`))
if _, ok := cache.Get(key); !ok {
t.Fatal("expected cache hit")
}
cache.RecordBypass()
stats := cache.Stats()
if got := stats["status"]; got != true {
t.Fatalf("expected enabled status, got %v", got)
}
if got := stats["entries"].(int); got != 1 {
t.Fatalf("expected 1 entry, got %d", got)
}
if got := stats["hits"].(int); got != 1 {
t.Fatalf("expected 1 hit, got %d", got)
}
if got := stats["misses"].(int); got != 1 {
t.Fatalf("expected 1 miss, got %d", got)
}
if got := stats["bypasses"].(int); got != 1 {
t.Fatalf("expected 1 bypass, got %d", got)
}
}
func TestBuildCacheKeyChangesWithPaginationAndFlags(t *testing.T) {
base := Query{
Text: "golang",
LangCode: "EN",
Limit: 10,
Start: 0,
Filter: true,
Answers: false,
}
baseKey := BuildCacheKey("google", "search", base)
if same := BuildCacheKey("google", "search", base); same != baseKey {
t.Fatal("expected deterministic key for same query")
}
if changed := BuildCacheKey("google", "search", Query{
Text: "golang",
LangCode: "EN",
Limit: 20,
Start: 0,
Filter: true,
Answers: false,
}); changed == baseKey {
t.Fatal("expected limit to affect cache key")
}
if changed := BuildCacheKey("google", "search", Query{
Text: "golang",
LangCode: "EN",
Limit: 10,
Start: 10,
Filter: true,
Answers: false,
}); changed == baseKey {
t.Fatal("expected start to affect cache key")
}
if changed := BuildCacheKey("google", "search", Query{
Text: "golang",
LangCode: "EN",
Limit: 10,
Start: 0,
Filter: false,
Answers: false,
}); changed == baseKey {
t.Fatal("expected filter to affect cache key")
}
if changed := BuildCacheKey("google", "search", Query{
Text: "golang",
LangCode: "EN",
Limit: 10,
Start: 0,
Filter: true,
Answers: true,
}); changed == baseKey {
t.Fatal("expected answers to affect cache key")
}
}

210
core/circuit_breaker.go Normal file
View File

@@ -0,0 +1,210 @@
package core
import (
"fmt"
"sync"
"time"
"github.com/sirupsen/logrus"
)
type CircuitState int
const (
CircuitClosed CircuitState = iota
CircuitOpen
CircuitHalfOpen
)
func (s CircuitState) String() string {
switch s {
case CircuitClosed:
return "closed"
case CircuitOpen:
return "open"
case CircuitHalfOpen:
return "half-open"
default:
return "unknown"
}
}
type CircuitBreakerConfig struct {
FailureThreshold int
RecoveryTimeout time.Duration
SuccessThreshold int
}
func DefaultCircuitBreakerConfig() CircuitBreakerConfig {
return CircuitBreakerConfig{
FailureThreshold: 5,
RecoveryTimeout: 60 * time.Second,
SuccessThreshold: 2,
}
}
// CircuitBreaker tracks failure state for one engine.
type CircuitBreaker struct {
mu sync.RWMutex
name string
state CircuitState
config CircuitBreakerConfig
failureCount int
successCount int
lastFailureTime time.Time
lastStateChange time.Time
}
func NewCircuitBreaker(name string, cfg CircuitBreakerConfig) *CircuitBreaker {
return &CircuitBreaker{
name: name,
state: CircuitClosed,
config: cfg,
lastStateChange: time.Now(),
}
}
func (cb *CircuitBreaker) AllowRequest() bool {
cb.mu.Lock()
defer cb.mu.Unlock()
switch cb.state {
case CircuitClosed:
return true
case CircuitOpen:
if time.Since(cb.lastFailureTime) >= cb.config.RecoveryTimeout {
cb.setState(CircuitHalfOpen)
logrus.Infof("[CircuitBreaker][%s] Recovery timeout elapsed, moving to half-open", cb.name)
return true
}
return false
case CircuitHalfOpen:
return true
default:
return true
}
}
func (cb *CircuitBreaker) RecordSuccess() {
cb.mu.Lock()
defer cb.mu.Unlock()
switch cb.state {
case CircuitHalfOpen:
cb.successCount++
if cb.successCount >= cb.config.SuccessThreshold {
cb.setState(CircuitClosed)
cb.failureCount = 0
cb.successCount = 0
logrus.Infof("[CircuitBreaker][%s] Recovered, circuit closed", cb.name)
}
case CircuitClosed:
cb.failureCount = 0
}
}
func (cb *CircuitBreaker) RecordFailure() {
cb.mu.Lock()
defer cb.mu.Unlock()
cb.lastFailureTime = time.Now()
switch cb.state {
case CircuitClosed:
cb.failureCount++
if cb.failureCount >= cb.config.FailureThreshold {
cb.setState(CircuitOpen)
logrus.Warnf("[CircuitBreaker][%s] Circuit OPENED after %d consecutive failures (will retry in %s)",
cb.name, cb.failureCount, cb.config.RecoveryTimeout)
}
case CircuitHalfOpen:
cb.setState(CircuitOpen)
cb.successCount = 0
logrus.Warnf("[CircuitBreaker][%s] Failed during half-open, circuit re-opened", cb.name)
}
}
func (cb *CircuitBreaker) State() CircuitState {
cb.mu.RLock()
defer cb.mu.RUnlock()
return cb.state
}
func (cb *CircuitBreaker) Stats() map[string]interface{} {
cb.mu.RLock()
defer cb.mu.RUnlock()
stats := map[string]interface{}{
"engine": cb.name,
"state": cb.state.String(),
"failure_count": cb.failureCount,
"last_changed": cb.lastStateChange.Format(time.RFC3339),
}
if cb.state == CircuitOpen {
remaining := cb.config.RecoveryTimeout - time.Since(cb.lastFailureTime)
if remaining < 0 {
remaining = 0
}
// Expose retry_in as integer seconds for easier client-side processing.
retryInSeconds := int64(0)
if remaining > 0 {
retryInSeconds = int64((remaining + time.Second - time.Nanosecond) / time.Second)
}
stats["retry_in"] = retryInSeconds
}
return stats
}
func (cb *CircuitBreaker) setState(state CircuitState) {
cb.state = state
cb.lastStateChange = time.Now()
}
type CircuitBreakerManager struct {
mu sync.RWMutex
breakers map[string]*CircuitBreaker
config CircuitBreakerConfig
}
func NewCircuitBreakerManager(cfg CircuitBreakerConfig) *CircuitBreakerManager {
return &CircuitBreakerManager{
breakers: make(map[string]*CircuitBreaker),
config: cfg,
}
}
func (m *CircuitBreakerManager) Get(engineName string) *CircuitBreaker {
m.mu.RLock()
if cb, ok := m.breakers[engineName]; ok {
m.mu.RUnlock()
return cb
}
m.mu.RUnlock()
m.mu.Lock()
defer m.mu.Unlock()
if cb, ok := m.breakers[engineName]; ok {
return cb
}
cb := NewCircuitBreaker(engineName, m.config)
m.breakers[engineName] = cb
return cb
}
func (m *CircuitBreakerManager) AllStats() []map[string]interface{} {
m.mu.RLock()
defer m.mu.RUnlock()
stats := make([]map[string]interface{}, 0, len(m.breakers))
for _, cb := range m.breakers {
stats = append(stats, cb.Stats())
}
return stats
}
var ErrCircuitOpen = fmt.Errorf("circuit breaker is open - engine temporarily disabled")

View File

@@ -0,0 +1,165 @@
package core
import (
"testing"
"time"
)
func newTestCircuitBreaker(t *testing.T, cfg CircuitBreakerConfig) *CircuitBreaker {
t.Helper()
return NewCircuitBreaker("test-engine", cfg)
}
// TestCircuitBreaker_OpensAfterThreshold verifies that consecutive failures in closed state
// move the breaker to open exactly on configured threshold and block new requests.
func TestCircuitBreaker_OpensAfterThreshold(t *testing.T) {
cfg := CircuitBreakerConfig{
FailureThreshold: 3,
RecoveryTimeout: time.Second,
SuccessThreshold: 1,
}
cb := newTestCircuitBreaker(t, cfg)
cb.RecordFailure()
cb.RecordFailure()
if cb.State() != CircuitClosed {
t.Fatalf("expected closed after 2 failures, got: %s", cb.State())
}
cb.RecordFailure()
if cb.State() != CircuitOpen {
t.Fatalf("expected open after %d failures, got: %s", cfg.FailureThreshold, cb.State())
}
if cb.AllowRequest() {
t.Error("expected request blocked in open state")
}
}
// TestCircuitBreaker_RecoveryToHalfOpen verifies timed recovery from open to half-open
// when recovery timeout elapses and a new request is attempted.
func TestCircuitBreaker_RecoveryToHalfOpen(t *testing.T) {
cfg := CircuitBreakerConfig{
FailureThreshold: 2,
RecoveryTimeout: 50 * time.Millisecond,
SuccessThreshold: 1,
}
cb := newTestCircuitBreaker(t, cfg)
cb.RecordFailure()
cb.RecordFailure()
if cb.State() != CircuitOpen {
t.Fatal("expected open")
}
time.Sleep(60 * time.Millisecond)
if !cb.AllowRequest() {
t.Error("should allow request after recovery timeout")
}
if cb.State() != CircuitHalfOpen {
t.Errorf("expected half-open, got: %s", cb.State())
}
}
// TestCircuitBreaker_HalfOpenSuccessClosesCircuit verifies that half-open state closes
// only after configured number of successful probes.
func TestCircuitBreaker_HalfOpenSuccessClosesCircuit(t *testing.T) {
cfg := CircuitBreakerConfig{
FailureThreshold: 1,
RecoveryTimeout: 20 * time.Millisecond,
SuccessThreshold: 2,
}
cb := newTestCircuitBreaker(t, cfg)
cb.RecordFailure()
if cb.State() != CircuitOpen {
t.Fatalf("expected open, got: %s", cb.State())
}
time.Sleep(30 * time.Millisecond)
if !cb.AllowRequest() {
t.Fatal("expected request to pass in recovery window")
}
if cb.State() != CircuitHalfOpen {
t.Fatalf("expected half-open after recovery timeout, got: %s", cb.State())
}
cb.RecordSuccess()
if cb.State() != CircuitHalfOpen {
t.Fatalf("expected to stay half-open until success threshold reached, got: %s", cb.State())
}
cb.RecordSuccess()
if cb.State() != CircuitClosed {
t.Fatalf("expected closed after success threshold reached, got: %s", cb.State())
}
}
// TestCircuitBreaker_HalfOpenFailureReopens verifies that a failed probe in half-open
// immediately re-opens the circuit.
func TestCircuitBreaker_HalfOpenFailureReopens(t *testing.T) {
cfg := CircuitBreakerConfig{
FailureThreshold: 1,
RecoveryTimeout: 20 * time.Millisecond,
SuccessThreshold: 1,
}
cb := newTestCircuitBreaker(t, cfg)
cb.RecordFailure()
time.Sleep(30 * time.Millisecond)
if !cb.AllowRequest() {
t.Fatal("expected probe request in half-open")
}
if cb.State() != CircuitHalfOpen {
t.Fatalf("expected half-open, got: %s", cb.State())
}
cb.RecordFailure()
if cb.State() != CircuitOpen {
t.Fatalf("expected open after failed half-open probe, got: %s", cb.State())
}
}
// TestCircuitBreaker_Stats verifies stats payload fields and that retry_in is exposed
// only when breaker is open.
func TestCircuitBreaker_Stats(t *testing.T) {
cb := NewCircuitBreaker("test-engine", DefaultCircuitBreakerConfig())
cb.RecordFailure()
stats := cb.Stats()
if stats["engine"] != "test-engine" {
t.Fatalf("expected engine=test-engine, got: %v", stats["engine"])
}
if stats["state"] != "closed" {
t.Fatalf("expected state=closed, got: %v", stats["state"])
}
if stats["failure_count"].(int) != 1 {
t.Fatalf("expected failure_count=1, got: %v", stats["failure_count"])
}
if _, ok := stats["retry_in"]; ok {
t.Fatalf("did not expect retry_in in closed state, got: %v", stats["retry_in"])
}
openCfg := CircuitBreakerConfig{FailureThreshold: 1, RecoveryTimeout: time.Second, SuccessThreshold: 1}
openCB := NewCircuitBreaker("open-engine", openCfg)
openCB.RecordFailure()
openStats := openCB.Stats()
retryIn, ok := openStats["retry_in"].(int64)
if !ok {
t.Fatalf("expected retry_in int64 in open state, got: %T", openStats["retry_in"])
}
if retryIn <= 0 {
t.Fatalf("expected retry_in > 0 in open state, got: %d", retryIn)
}
}
// TestCircuitBreakerManager_AllStats verifies manager creates and reports per-engine breakers.
func TestCircuitBreakerManager_AllStats(t *testing.T) {
mgr := NewCircuitBreakerManager(DefaultCircuitBreakerConfig())
mgr.Get("google")
mgr.Get("yandex")
stats := mgr.AllStats()
if len(stats) != 2 {
t.Errorf("expected 2 entries, got: %d", len(stats))
}
}

View File

@@ -54,15 +54,28 @@ func ConvertSearchResultsMap(searchResultsMap map[string]SearchResult) *[]Search
}
type Query struct {
Text string
LangCode string // eg. EN, ES, RU...
DateInterval string // format: YYYYMMDD..YYYMMDD - 20181010..20231010
Filetype string // File extension to search.
Site string // Search site
Limit int // Limit the number of results
Answers bool // Include question and answers from SERP page to results with negative indexes
ProxyURL string // Proxy URL for raw requests
Insecure bool // Allow insecure TLS connections
Text string
LangCode string // eg. EN, ES, RU...
DateInterval string // format: YYYYMMDD..YYYMMDD - 20181010..20231010
Filetype string // File extension to search.
Site string // Search site
Limit int // Limit the number of results
Start int // Search offset for pagination (Google uses 0, 10, 20...)
Filter bool // Filter duplicates (google) (false: include similar, true: hide similar)
Answers bool // Include question and answers from SERP page to results with negative indexes
ProxyURL string // Proxy URL for raw requests
ProxyOverride string // Request-scoped proxy override: tag or direct
Insecure bool // Allow insecure TLS connections
}
func ComputePagination(start int, pageSize int) (int, int, error) {
if pageSize <= 0 {
return 0, 0, errors.New("pageSize must be > 0")
}
if start < 0 {
return 0, 0, errors.New("start must be >= 0")
}
return start / pageSize, start % pageSize, nil
}
func (q Query) IsEmpty() bool {
@@ -85,11 +98,30 @@ func (searchQuery *Query) InitFromContext(reqCtx *fiber.Ctx) error {
}
searchQuery.Limit = limit
start, err := strconv.Atoi(reqCtx.Query("start", "0"))
if err != nil {
return err
}
if start < 0 {
return errors.New("start must be >= 0")
}
searchQuery.Start = start
searchQuery.Filter, err = strconv.ParseBool(reqCtx.Query("filter", "1"))
if err != nil {
return err
}
searchQuery.Answers, err = strconv.ParseBool(reqCtx.Query("answers", "0"))
if err != nil {
return err
}
searchQuery.ProxyOverride, err = NormalizeProxyRequestOverride(reqCtx.Get("X-Use-Proxy"))
if err != nil {
return err
}
if searchQuery.IsEmpty() {
return errors.New("Query cannot be empty")
}

74
core/http_client.go Normal file
View File

@@ -0,0 +1,74 @@
package core
import (
"context"
"crypto/tls"
"net"
"net/http"
"net/url"
"strings"
"time"
utls "github.com/refraction-networking/utls"
)
const rawHTTPTimeout = 10 * time.Second
func NewRawHTTPClient(query Query) (*http.Client, error) {
transport, err := newRawTransport(query)
if err != nil {
return nil, err
}
return &http.Client{
Transport: transport,
Timeout: rawHTTPTimeout,
}, nil
}
func newRawTransport(query Query) (*http.Transport, error) {
transport := &http.Transport{}
if query.Insecure {
transport.TLSClientConfig = &tls.Config{InsecureSkipVerify: true}
}
proxyURL, err := NormalizeProxyURL(query.ProxyURL)
if err != nil {
return nil, err
}
if proxyURL != "" {
parsed, err := url.Parse(proxyURL)
if err != nil {
return nil, err
}
// Keep proxied requests on the standard transport path so SOCKS5/SOCKS5H
// resolution and routing are handled by the configured proxy correctly.
transport.Proxy = http.ProxyURL(parsed)
return transport, nil
}
transport.DialTLSContext = func(ctx context.Context, network, addr string) (net.Conn, error) {
dialer := &net.Dialer{}
rawConn, err := dialer.DialContext(ctx, network, addr)
if err != nil {
return nil, err
}
hostname := strings.Split(addr, ":")[0]
config := &utls.Config{
ServerName: hostname,
InsecureSkipVerify: query.Insecure,
}
uconn := utls.UClient(rawConn, config, utls.HelloChrome_Auto)
if err := uconn.Handshake(); err != nil {
rawConn.Close()
return nil, err
}
return uconn, nil
}
return transport, nil
}

144
core/middleware.go Normal file
View File

@@ -0,0 +1,144 @@
package core
import (
"fmt"
"strings"
"time"
"github.com/gofiber/fiber/v2"
"github.com/sirupsen/logrus"
)
type JSONErrorResponse struct {
Error string `json:"error"`
Code int `json:"code"`
Message string `json:"message,omitempty"`
}
type CORSConfig struct {
AllowOrigins string
AllowMethods string
AllowHeaders string
MaxAge int
}
func DefaultCORSConfig() CORSConfig {
return CORSConfig{
AllowOrigins: "*",
AllowMethods: "GET, POST, OPTIONS",
AllowHeaders: "Origin, Content-Type, Accept, Authorization, X-Use-Proxy",
MaxAge: 86400,
}
}
func CORSMiddleware(cfg CORSConfig) fiber.Handler {
cfg = normalizeCORSConfig(cfg)
return func(c *fiber.Ctx) error {
c.Set("Access-Control-Allow-Origin", cfg.AllowOrigins)
c.Set("Access-Control-Allow-Methods", cfg.AllowMethods)
c.Set("Access-Control-Allow-Headers", cfg.AllowHeaders)
c.Set("Access-Control-Max-Age", fmt.Sprintf("%d", cfg.MaxAge))
if c.Method() == "OPTIONS" {
return c.SendStatus(fiber.StatusNoContent)
}
return c.Next()
}
}
// normalizeCORSConfig keeps CORS behavior predictable when config provides partial values.
func normalizeCORSConfig(cfg CORSConfig) CORSConfig {
defaults := DefaultCORSConfig()
if strings.TrimSpace(cfg.AllowOrigins) == "" {
cfg.AllowOrigins = defaults.AllowOrigins
}
if strings.TrimSpace(cfg.AllowMethods) == "" {
cfg.AllowMethods = defaults.AllowMethods
}
if strings.TrimSpace(cfg.AllowHeaders) == "" {
cfg.AllowHeaders = defaults.AllowHeaders
}
if cfg.MaxAge <= 0 {
cfg.MaxAge = defaults.MaxAge
}
return cfg
}
func RequestLoggerMiddleware() fiber.Handler {
return func(c *fiber.Ctx) error {
start := time.Now()
err := c.Next()
latency := time.Since(start)
status := c.Response().StatusCode()
if err != nil {
if e, ok := err.(*fiber.Error); ok {
status = e.Code
} else {
status = fiber.StatusInternalServerError
}
}
logFields := logrus.Fields{
"method": c.Method(),
"path": c.Path(),
"status": status,
"latency": latency.String(),
"ip": c.IP(),
}
if query := c.Query("text"); query != "" {
logFields["query"] = query
}
entry := logrus.WithFields(logFields)
if status >= 500 {
entry.Errorf("%s - request failed", c.Path())
} else if status >= 400 {
entry.Warnf("%s - request error", c.Path())
} else {
entry.Infof("%s - request completed", c.Path())
}
return err
}
}
func JSONErrorMiddleware() fiber.ErrorHandler {
return func(c *fiber.Ctx, err error) error {
code := fiber.StatusInternalServerError
if e, ok := err.(*fiber.Error); ok {
code = e.Code
}
resp := JSONErrorResponse{
Error: statusText(code),
Code: code,
Message: err.Error(),
}
c.Set("Content-Type", "application/json")
return c.Status(code).JSON(resp)
}
}
func statusText(code int) string {
switch {
case code == 400:
return "bad_request"
case code == 404:
return "not_found"
case code == 429:
return "rate_limited"
case code == 503:
return "service_unavailable"
case code >= 400 && code < 500:
return "client_error"
case code >= 500:
return "server_error"
default:
return "error"
}
}

109
core/middleware_test.go Normal file
View File

@@ -0,0 +1,109 @@
package core
import (
"net/http"
"net/http/httptest"
"strings"
"testing"
"github.com/gofiber/fiber/v2"
)
func TestStatusText(t *testing.T) {
tests := []struct {
code int
expected string
}{
{400, "bad_request"},
{404, "not_found"},
{429, "rate_limited"},
{503, "service_unavailable"},
{401, "client_error"},
{500, "server_error"},
{200, "error"},
}
for _, tt := range tests {
result := statusText(tt.code)
if result != tt.expected {
t.Errorf("statusText(%d) = %s, want %s", tt.code, result, tt.expected)
}
}
}
// Middleware unit tests validate CORS behavior itself (header composition and preflight semantics).
func TestCORSMiddleware_UsesConfiguredHeaders(t *testing.T) {
app := fiber.New()
app.Use(CORSMiddleware(CORSConfig{
AllowOrigins: "https://example.com",
AllowMethods: "GET,OPTIONS",
AllowHeaders: "Authorization,Content-Type",
MaxAge: 1200,
}))
app.Get("/ping", func(c *fiber.Ctx) error {
return c.SendStatus(fiber.StatusOK)
})
req := httptest.NewRequest(http.MethodGet, "/ping", nil)
resp, err := app.Test(req, -1)
if err != nil {
t.Fatalf("request failed: %v", err)
}
if got := resp.Header.Get("Access-Control-Allow-Origin"); got != "https://example.com" {
t.Fatalf("unexpected allow-origin: %q", got)
}
if got := resp.Header.Get("Access-Control-Allow-Methods"); got != "GET,OPTIONS" {
t.Fatalf("unexpected allow-methods: %q", got)
}
if got := resp.Header.Get("Access-Control-Allow-Headers"); got != "Authorization,Content-Type" {
t.Fatalf("unexpected allow-headers: %q", got)
}
if got := resp.Header.Get("Access-Control-Max-Age"); got != "1200" {
t.Fatalf("unexpected max-age: %q", got)
}
}
func TestCORSMiddleware_OPTIONSReturnsNoContent(t *testing.T) {
app := fiber.New()
app.Use(CORSMiddleware(DefaultCORSConfig()))
app.Get("/ping", func(c *fiber.Ctx) error {
return c.SendStatus(fiber.StatusOK)
})
req := httptest.NewRequest(http.MethodOptions, "/ping", nil)
resp, err := app.Test(req, -1)
if err != nil {
t.Fatalf("request failed: %v", err)
}
if resp.StatusCode != fiber.StatusNoContent {
t.Fatalf("expected 204 for OPTIONS, got %d", resp.StatusCode)
}
}
func TestNormalizeCORSConfig_FillsMissingValues(t *testing.T) {
cfg := normalizeCORSConfig(CORSConfig{
AllowOrigins: "https://example.com",
})
def := DefaultCORSConfig()
if cfg.AllowOrigins != "https://example.com" {
t.Fatalf("expected custom allow_origins preserved, got %q", cfg.AllowOrigins)
}
if cfg.AllowMethods != def.AllowMethods {
t.Fatalf("expected default allow_methods, got %q", cfg.AllowMethods)
}
if cfg.AllowHeaders != def.AllowHeaders {
t.Fatalf("expected default allow_headers, got %q", cfg.AllowHeaders)
}
if cfg.MaxAge != def.MaxAge {
t.Fatalf("expected default max_age, got %d", cfg.MaxAge)
}
}
func TestDefaultCORSConfig_IncludesProxyOverrideHeader(t *testing.T) {
cfg := DefaultCORSConfig()
if got := cfg.AllowHeaders; !strings.Contains(got, "X-Use-Proxy") {
t.Fatalf("expected allow_headers to include X-Use-Proxy, got %q", got)
}
}

561
core/proxy.go Normal file
View File

@@ -0,0 +1,561 @@
package core
import (
"errors"
"fmt"
"net/url"
"sort"
"strings"
"sync"
"github.com/sirupsen/logrus"
)
const (
ProxyRuntimeBrowser = "browser"
ProxyRuntimeRaw = "raw"
ProxyModeOff = "off"
ProxyModeTagPool = "tag_pool"
DefaultProxyFailureThreshold = 3
ProxyOverrideDirect = "direct"
)
var supportedProxySchemes = map[string]struct{}{
"http": {},
"https": {},
"socks5": {},
"socks5h": {},
}
var ErrProxyUnavailable = errors.New("proxy unavailable")
type ProxyPolicy struct {
Mode string `json:"mode" mapstructure:"mode"`
Tag string `json:"tag,omitempty" mapstructure:"tag"`
}
type ProxyEntryConfig struct {
URL string `json:"url" mapstructure:"url"`
Tags []string `json:"tags" mapstructure:"tags"`
}
type ProxiesHealthConfig struct {
FailureThreshold int `json:"failure_threshold" mapstructure:"failure_threshold"`
}
type ProxiesConfig struct {
Global string `json:"global,omitempty" mapstructure:"global"`
Entries []ProxyEntryConfig `json:"entries" mapstructure:"entries"`
Health ProxiesHealthConfig `json:"health" mapstructure:"health"`
}
type ProxyConfig struct {
Runtime string // raw or browser runtime behavior
Proxies ProxiesConfig // canonical proxy inventory
EnginePolicies map[string]string // engine-specific proxy tags
Registry *ProxyRegistry // optional shared registry from caller
}
type ProxyTagSummary struct {
Configured int `json:"configured"`
Healthy int `json:"healthy"`
}
type ProxyStatsEntry struct {
Proxy string `json:"proxy"`
Tags []string `json:"tags"`
Healthy bool `json:"healthy"`
Failures int `json:"failures"`
Disabled bool `json:"disabled"`
}
type ProxyEngineStats struct {
Tag string `json:"tag,omitempty"`
SelectedProxy string `json:"selected_proxy"`
}
type ProxyStats struct {
ConfiguredCount int `json:"configured_count"`
HealthyCount int `json:"healthy_count"`
UnhealthyCount int `json:"unhealthy_count"`
Tags map[string]ProxyTagSummary `json:"tags"`
Entries []ProxyStatsEntry `json:"entries"`
Engines map[string]ProxyEngineStats `json:"engines,omitempty"`
}
type proxyState struct {
url string
tags []string
failures int
disabled bool
}
type ProxyRegistry struct {
mu sync.Mutex
states map[string]*proxyState
order []string
tagIndex map[string][]string
nextByTag map[string]int
failureThreshold int
}
func DefaultProxiesConfig() ProxiesConfig {
return ProxiesConfig{
Global: "",
Entries: []ProxyEntryConfig{},
Health: ProxiesHealthConfig{FailureThreshold: DefaultProxyFailureThreshold},
}
}
func DefaultProxyConfig() ProxyConfig {
return ProxyConfig{
Runtime: ProxyRuntimeBrowser,
Proxies: DefaultProxiesConfig(),
EnginePolicies: map[string]string{},
}
}
func NormalizeProxyConfig(cfg ProxyConfig) (ProxyConfig, error) {
cfg.Runtime = normalizeProxyRuntime(cfg.Runtime)
var err error
cfg.Proxies, err = NormalizeProxiesConfig(cfg.Proxies)
if err != nil {
return cfg, err
}
if cfg.EnginePolicies == nil {
cfg.EnginePolicies = map[string]string{}
}
normalizedEnginePolicies := make(map[string]string, len(cfg.EnginePolicies))
for rawEngine, rawTag := range cfg.EnginePolicies {
engine := normalizeEngineName(rawEngine)
if engine == "" {
continue
}
tag := normalizeTag(rawTag)
if tag == "" {
continue
}
normalizedEnginePolicies[engine] = tag
}
cfg.EnginePolicies = normalizedEnginePolicies
if cfg.Registry == nil {
if len(cfg.Proxies.Entries) > 0 {
registry, err := NewProxyRegistry(cfg.Proxies.Entries, cfg.Proxies.Health.FailureThreshold)
if err != nil {
return cfg, err
}
cfg.Registry = registry
}
}
return cfg, nil
}
func NormalizeProxiesConfig(cfg ProxiesConfig) (ProxiesConfig, error) {
global, err := NormalizeProxyURL(cfg.Global)
if err != nil {
return cfg, fmt.Errorf("invalid proxies.global: %w", err)
}
cfg.Global = global
failureThreshold := cfg.Health.FailureThreshold
if failureThreshold <= 0 {
failureThreshold = DefaultProxyFailureThreshold
}
normalizedEntries := make([]ProxyEntryConfig, 0, len(cfg.Entries))
entryByURL := make(map[string]int, len(cfg.Entries))
for i, rawEntry := range cfg.Entries {
proxyURL, err := NormalizeProxyURL(rawEntry.URL)
if err != nil {
return cfg, fmt.Errorf("invalid proxies.entries[%d].url: %w", i, err)
}
if proxyURL == "" {
return cfg, fmt.Errorf("invalid proxies.entries[%d].url: value is required", i)
}
tags, err := normalizeProxyTags(rawEntry.Tags)
if err != nil {
return cfg, fmt.Errorf("invalid proxies.entries[%d].tags: %w", i, err)
}
if idx, ok := entryByURL[proxyURL]; ok {
normalizedEntries[idx].Tags = mergeTags(normalizedEntries[idx].Tags, tags)
continue
}
normalizedEntries = append(normalizedEntries, ProxyEntryConfig{
URL: proxyURL,
Tags: tags,
})
entryByURL[proxyURL] = len(normalizedEntries) - 1
}
cfg.Entries = normalizedEntries
cfg.Health = ProxiesHealthConfig{FailureThreshold: failureThreshold}
return cfg, nil
}
func NormalizeProxyURL(raw string) (string, error) {
raw = strings.TrimSpace(raw)
if raw == "" {
return "", nil
}
parsed, err := url.Parse(raw)
if err != nil {
return "", err
}
if parsed.Scheme == "" {
return "", fmt.Errorf("proxy URL must include a scheme")
}
if parsed.Host == "" {
return "", fmt.Errorf("proxy URL must include a host")
}
parsed.Scheme = strings.ToLower(parsed.Scheme)
if _, ok := supportedProxySchemes[parsed.Scheme]; !ok {
return "", fmt.Errorf("unsupported proxy scheme %q", parsed.Scheme)
}
return parsed.String(), nil
}
func NormalizeProxyURLs(rawURLs []string) ([]string, error) {
normalized := make([]string, 0, len(rawURLs))
seen := make(map[string]struct{}, len(rawURLs))
for _, raw := range rawURLs {
proxyURL, err := NormalizeProxyURL(raw)
if err != nil {
return nil, err
}
if proxyURL == "" {
continue
}
if _, ok := seen[proxyURL]; ok {
continue
}
seen[proxyURL] = struct{}{}
normalized = append(normalized, proxyURL)
}
return normalized, nil
}
func MaskProxyURL(raw string) string {
proxyURL, err := NormalizeProxyURL(raw)
if err != nil || proxyURL == "" {
return "invalid-proxy"
}
parsed, err := url.Parse(proxyURL)
if err != nil {
return "invalid-proxy"
}
return fmt.Sprintf("%s://%s", parsed.Scheme, parsed.Host)
}
func ResolveEffectiveProxyPolicy(globalProxyURL string, engineTag string) ProxyPolicy {
if strings.TrimSpace(globalProxyURL) != "" {
return ProxyPolicy{Mode: ProxyModeTagPool}
}
tag := normalizeTag(engineTag)
if tag == "" {
return ProxyPolicy{Mode: ProxyModeOff}
}
return ProxyPolicy{Mode: ProxyModeTagPool, Tag: tag}
}
func NormalizeProxyTag(raw string) (string, error) {
tag := normalizeTag(raw)
if tag == "" {
return "", fmt.Errorf("value is required")
}
return tag, nil
}
func NormalizeProxyRequestOverride(raw string) (string, error) {
override := normalizeTag(raw)
if override == "" {
return "", nil
}
if override == ProxyOverrideDirect {
return ProxyOverrideDirect, nil
}
return NormalizeProxyTag(override)
}
func IsAuthenticatedSocksProxyURL(raw string) bool {
normalized, err := NormalizeProxyURL(raw)
if err != nil || normalized == "" {
return false
}
parsed, err := url.Parse(normalized)
if err != nil {
return false
}
if (parsed.Scheme == "socks5" || parsed.Scheme == "socks5h") && parsed.User != nil {
return true
}
return false
}
func NewProxyRegistry(entries []ProxyEntryConfig, failureThreshold int) (*ProxyRegistry, error) {
if failureThreshold <= 0 {
failureThreshold = DefaultProxyFailureThreshold
}
states := make(map[string]*proxyState, len(entries))
order := make([]string, 0, len(entries))
tagIndex := make(map[string][]string)
for idx, entry := range entries {
proxyURL, err := NormalizeProxyURL(entry.URL)
if err != nil {
return nil, fmt.Errorf("invalid proxy registry entry[%d] url: %w", idx, err)
}
if proxyURL == "" {
return nil, fmt.Errorf("invalid proxy registry entry[%d] url: value is required", idx)
}
tags, err := normalizeProxyTags(entry.Tags)
if err != nil {
return nil, fmt.Errorf("invalid proxy registry entry[%d] tags: %w", idx, err)
}
states[proxyURL] = &proxyState{url: proxyURL, tags: tags}
order = append(order, proxyURL)
for _, tag := range tags {
tagIndex[tag] = append(tagIndex[tag], proxyURL)
}
}
return &ProxyRegistry{
states: states,
order: order,
tagIndex: tagIndex,
nextByTag: make(map[string]int, len(tagIndex)),
failureThreshold: failureThreshold,
}, nil
}
func (r *ProxyRegistry) NextByTag(tag string) string {
tag = normalizeTag(tag)
if tag == "" {
return ""
}
r.mu.Lock()
defer r.mu.Unlock()
urls := r.tagIndex[tag]
if len(urls) == 0 {
return ""
}
if r.allDisabledLocked(urls) {
logrus.Warnf("Proxy tag pool exhausted for %q, re-enabling tagged proxies", tag)
for _, proxyURL := range urls {
state := r.states[proxyURL]
state.disabled = false
state.failures = 0
}
}
start := r.nextByTag[tag]
for i := 0; i < len(urls); i++ {
idx := (start + i) % len(urls)
proxyURL := urls[idx]
state := r.states[proxyURL]
if state.disabled {
continue
}
r.nextByTag[tag] = (idx + 1) % len(urls)
logrus.Debugf("Selected proxy for tag=%s: %s", tag, MaskProxyURL(proxyURL))
return proxyURL
}
return ""
}
func (r *ProxyRegistry) ReportFailure(proxyURL string) {
proxyURL, err := NormalizeProxyURL(proxyURL)
if err != nil || proxyURL == "" {
return
}
r.mu.Lock()
defer r.mu.Unlock()
state, ok := r.states[proxyURL]
if !ok {
return
}
state.failures++
if state.failures >= r.failureThreshold {
state.disabled = true
logrus.Warnf("Disabled proxy after %d failures: %s", state.failures, MaskProxyURL(proxyURL))
}
}
func (r *ProxyRegistry) ReportSuccess(proxyURL string) {
proxyURL, err := NormalizeProxyURL(proxyURL)
if err != nil || proxyURL == "" {
return
}
r.mu.Lock()
defer r.mu.Unlock()
state, ok := r.states[proxyURL]
if !ok {
return
}
state.failures = 0
state.disabled = false
}
func (r *ProxyRegistry) HasHealthyProxyForTag(tag string) bool {
tag = normalizeTag(tag)
if tag == "" {
return false
}
r.mu.Lock()
defer r.mu.Unlock()
for _, proxyURL := range r.tagIndex[tag] {
if state, ok := r.states[proxyURL]; ok && !state.disabled {
return true
}
}
return false
}
func (r *ProxyRegistry) BuildStats() ProxyStats {
r.mu.Lock()
defer r.mu.Unlock()
stats := ProxyStats{
Tags: map[string]ProxyTagSummary{},
Entries: make([]ProxyStatsEntry, 0, len(r.order)),
}
for _, proxyURL := range r.order {
state := r.states[proxyURL]
healthy := !state.disabled
if healthy {
stats.HealthyCount++
} else {
stats.UnhealthyCount++
}
stats.ConfiguredCount++
stats.Entries = append(stats.Entries, ProxyStatsEntry{
Proxy: MaskProxyURL(state.url),
Tags: append([]string(nil), state.tags...),
Healthy: healthy,
Failures: state.failures,
Disabled: state.disabled,
})
for _, tag := range state.tags {
summary := stats.Tags[tag]
summary.Configured++
if healthy {
summary.Healthy++
}
stats.Tags[tag] = summary
}
}
return stats
}
func (r *ProxyRegistry) allDisabledLocked(urls []string) bool {
if len(urls) == 0 {
return false
}
for _, proxyURL := range urls {
if state, ok := r.states[proxyURL]; ok && !state.disabled {
return false
}
}
return true
}
func normalizeProxyRuntime(runtime string) string {
switch strings.ToLower(strings.TrimSpace(runtime)) {
case ProxyRuntimeRaw:
return ProxyRuntimeRaw
default:
return ProxyRuntimeBrowser
}
}
func normalizeProxyTags(tags []string) ([]string, error) {
if len(tags) == 0 {
return nil, fmt.Errorf("at least one tag is required")
}
seen := make(map[string]struct{}, len(tags))
normalized := make([]string, 0, len(tags))
for _, rawTag := range tags {
tag := normalizeTag(rawTag)
if tag == "" {
continue
}
if _, ok := seen[tag]; ok {
continue
}
seen[tag] = struct{}{}
normalized = append(normalized, tag)
}
if len(normalized) == 0 {
return nil, fmt.Errorf("at least one non-empty tag is required")
}
sort.Strings(normalized)
return normalized, nil
}
func normalizeTag(raw string) string {
return strings.ToLower(strings.TrimSpace(raw))
}
func mergeTags(base []string, additional []string) []string {
combined := make(map[string]struct{}, len(base)+len(additional))
for _, tag := range base {
combined[tag] = struct{}{}
}
for _, tag := range additional {
combined[tag] = struct{}{}
}
merged := make([]string, 0, len(combined))
for tag := range combined {
merged = append(merged, tag)
}
sort.Strings(merged)
return merged
}
func normalizeEngineName(raw string) string {
return strings.ToLower(strings.TrimSpace(raw))
}

View File

@@ -0,0 +1,234 @@
//go:build integration
// +build integration
package core
import (
"context"
"encoding/json"
"io"
"net"
"net/http"
"net/url"
"os"
"strings"
"testing"
"github.com/karust/openserp/testutil"
"golang.org/x/time/rate"
)
const proxyIntegrationEnabledEnv = "OPENSERP_PROXY_TESTS"
type proxyIntegrationURLs struct {
targetURL string
socks5hAuthURL string
socks5hPlainURL string
httpAuthURL string
httpPlainURL string
badSocks5URL string
badHTTPURL string
}
func TestIntegrationSocks5hAuthProxyDNS(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyFetchesTarget(t, cfg.targetURL, cfg.socks5hAuthURL)
}
func TestIntegrationSocks5hPlainProxyDNS(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyFetchesTarget(t, cfg.targetURL, cfg.socks5hPlainURL)
}
func TestIntegrationHTTPAuthProxy(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyFetchesTarget(t, cfg.targetURL, cfg.httpAuthURL)
}
func TestIntegrationHTTPPlainProxy(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyFetchesTarget(t, cfg.targetURL, cfg.httpPlainURL)
}
func TestIntegrationRawSOCKSProxyPoolRotation(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyPoolRotation(t, cfg.targetURL, []string{cfg.badSocks5URL, cfg.socks5hAuthURL})
}
func TestIntegrationRawHTTPProxyPoolRotation(t *testing.T) {
cfg := proxyIntegrationConfig(t)
assertProxyPoolRotation(t, cfg.targetURL, []string{cfg.badHTTPURL, cfg.httpAuthURL})
}
type proxyIntegrationEngine struct {
targetURL string
limiter *rate.Limiter
proxies []string
}
func (e *proxyIntegrationEngine) Name() string {
return "google"
}
func (e *proxyIntegrationEngine) IsInitialized() bool {
return true
}
func (e *proxyIntegrationEngine) GetRateLimiter() *rate.Limiter {
return e.limiter
}
func (e *proxyIntegrationEngine) Search(q Query) ([]SearchResult, error) {
e.proxies = append(e.proxies, q.ProxyURL)
body, err := fetchViaRawProxy(q.ProxyURL, q.Insecure, e.targetURL)
if err != nil {
return nil, err
}
return []SearchResult{{
Rank: 1,
URL: e.targetURL,
Title: "proxy-ok",
Description: body,
}}, nil
}
func (e *proxyIntegrationEngine) SearchImage(q Query) ([]SearchResult, error) {
return nil, ErrSearchTimeout
}
func proxyIntegrationConfig(t *testing.T) proxyIntegrationURLs {
t.Helper()
testutil.RequireIntegration(t)
if os.Getenv(proxyIntegrationEnabledEnv) != "1" {
t.Skipf("set %s=1 to run proxy integration tests", proxyIntegrationEnabledEnv)
}
return proxyIntegrationURLs{
targetURL: envOrDefault("OPENSERP_PROXY_TEST_TARGET_URL", "http://proxy-target:8080/"),
socks5hAuthURL: envOrDefault("OPENSERP_PROXY_TEST_SOCKS5H_AUTH_URL", "socks5h://test:test@127.0.0.1:19080"),
socks5hPlainURL: envOrDefault("OPENSERP_PROXY_TEST_SOCKS5H_PLAIN_URL", "socks5h://127.0.0.1:19082"),
httpAuthURL: envOrDefault("OPENSERP_PROXY_TEST_HTTP_AUTH_URL", "http://test:test@127.0.0.1:18888"),
httpPlainURL: envOrDefault("OPENSERP_PROXY_TEST_HTTP_PLAIN_URL", "http://127.0.0.1:18889"),
badSocks5URL: envOrDefault("OPENSERP_PROXY_TEST_BAD_SOCKS5_URL", "socks5://127.0.0.1:19081"),
badHTTPURL: envOrDefault("OPENSERP_PROXY_TEST_BAD_HTTP_URL", "http://127.0.0.1:18890"),
}
}
func assertProxyFetchesTarget(t *testing.T, targetURL, proxyURL string) {
t.Helper()
target := mustParseURL(t, targetURL)
assertHostCannotResolveTarget(t, target.Hostname())
body, err := fetchViaRawProxy(proxyURL, false, targetURL)
if err != nil {
t.Fatalf("expected proxied request via %s to succeed, got %v", proxyURL, err)
}
if !strings.Contains(body, "proxy-ok") {
t.Fatalf("expected proxy target response, got %q", body)
}
}
func assertProxyPoolRotation(t *testing.T, targetURL string, pool []string) {
t.Helper()
engine := &proxyIntegrationEngine{
targetURL: targetURL,
limiter: rate.NewLimiter(rate.Inf, 1),
}
opts := DefaultServerOptions()
opts.Resilience.Retry.MaxRetries = 1
opts.Resilience.Retry.InitialBackoff = 0
opts.Resilience.Retry.MaxBackoff = 0
opts.Resilience.Retry.BackoffFactor = 1
opts.Resilience.Proxy = ProxyConfig{
Runtime: ProxyRuntimeRaw,
Proxies: ProxiesConfig{
Health: ProxiesHealthConfig{FailureThreshold: 1},
},
EnginePolicies: map[string]string{"google": "default"},
}
for _, proxyURL := range pool {
opts.Resilience.Proxy.Proxies.Entries = append(opts.Resilience.Proxy.Proxies.Entries, ProxyEntryConfig{
URL: proxyURL,
Tags: []string{"default"},
})
}
srv := NewServerWithOptions("127.0.0.1", 7190, opts, engine)
resp := request(t, srv, "/google/search?text=proxy")
if resp.StatusCode != http.StatusOK {
t.Fatalf("expected rotated proxy request to succeed, got %d", resp.StatusCode)
}
if len(engine.proxies) != 2 {
t.Fatalf("expected 2 proxy attempts, got %d", len(engine.proxies))
}
if engine.proxies[0] != pool[0] || engine.proxies[1] != pool[1] {
t.Fatalf("unexpected proxy rotation order: %#v", engine.proxies)
}
statsResp := request(t, srv, "/stats/proxy")
var stats map[string]interface{}
if err := json.NewDecoder(statsResp.Body).Decode(&stats); err != nil {
t.Fatalf("decode stats: %v", err)
}
if got := stats["healthy_count"].(float64); got != 1 {
t.Fatalf("expected healthy_count=1, got %v", got)
}
if got := stats["unhealthy_count"].(float64); got != 1 {
t.Fatalf("expected unhealthy_count=1, got %v", got)
}
}
func fetchViaRawProxy(proxyURL string, insecure bool, targetURL string) (string, error) {
client, err := NewRawHTTPClient(Query{
ProxyURL: proxyURL,
Insecure: insecure,
})
if err != nil {
return "", err
}
resp, err := client.Get(targetURL)
if err != nil {
return "", err
}
defer resp.Body.Close()
bodyBytes, err := io.ReadAll(resp.Body)
if err != nil {
return "", err
}
return string(bodyBytes), nil
}
func assertHostCannotResolveTarget(t *testing.T, hostname string) {
t.Helper()
if _, err := net.DefaultResolver.LookupHost(context.Background(), hostname); err == nil {
t.Fatalf("expected direct host-side DNS lookup for %q to fail", hostname)
}
}
func mustParseURL(t *testing.T, raw string) *url.URL {
t.Helper()
parsed, err := url.Parse(raw)
if err != nil {
t.Fatalf("parse target URL %q: %v", raw, err)
}
return parsed
}
func envOrDefault(key, fallback string) string {
if value := strings.TrimSpace(os.Getenv(key)); value != "" {
return value
}
return fallback
}

345
core/proxy_test.go Normal file
View File

@@ -0,0 +1,345 @@
package core
import (
"fmt"
"io"
"log"
"net"
"net/http"
"net/http/httptest"
"net/url"
"testing"
"time"
socks5 "github.com/armon/go-socks5"
xcontext "golang.org/x/net/context"
)
func TestNormalizeProxyURL(t *testing.T) {
tests := []struct {
name string
raw string
want string
wantErr bool
}{
{name: "empty", raw: "", want: ""},
{name: "http", raw: "http://127.0.0.1:8080", want: "http://127.0.0.1:8080"},
{name: "https", raw: "https://127.0.0.1:8443", want: "https://127.0.0.1:8443"},
{name: "socks5", raw: "socks5://127.0.0.1:1080", want: "socks5://127.0.0.1:1080"},
{name: "socks5h upper", raw: "SOCKS5H://127.0.0.1:1080", want: "socks5h://127.0.0.1:1080"},
{name: "missing scheme", raw: "127.0.0.1:8080", wantErr: true},
{name: "missing host", raw: "http://", wantErr: true},
{name: "unsupported scheme", raw: "ftp://127.0.0.1:21", wantErr: true},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
got, err := NormalizeProxyURL(tt.raw)
if tt.wantErr {
if err == nil {
t.Fatalf("expected error for %q", tt.raw)
}
return
}
if err != nil {
t.Fatalf("unexpected error: %v", err)
}
if got != tt.want {
t.Fatalf("expected %q, got %q", tt.want, got)
}
})
}
}
func TestNormalizeProxiesConfigDefaultsAndDeduplicates(t *testing.T) {
cfg, err := NormalizeProxiesConfig(ProxiesConfig{
Global: " HTTP://proxy-global:8080 ",
Entries: []ProxyEntryConfig{
{URL: " http://proxy-one:8080 ", Tags: []string{"default", "us"}},
{URL: "http://proxy-one:8080", Tags: []string{"de", "us"}},
{URL: "socks5://proxy-two:1080", Tags: []string{"default"}},
},
})
if err != nil {
t.Fatalf("normalize proxies config: %v", err)
}
if cfg.Global != "http://proxy-global:8080" {
t.Fatalf("expected normalized global proxy, got %q", cfg.Global)
}
if cfg.Health.FailureThreshold != DefaultProxyFailureThreshold {
t.Fatalf("expected default failure threshold %d, got %d", DefaultProxyFailureThreshold, cfg.Health.FailureThreshold)
}
if len(cfg.Entries) != 2 {
t.Fatalf("expected 2 deduplicated entries, got %d", len(cfg.Entries))
}
if cfg.Entries[0].URL != "http://proxy-one:8080" {
t.Fatalf("unexpected normalized URL for first entry: %s", cfg.Entries[0].URL)
}
if len(cfg.Entries[0].Tags) != 3 {
t.Fatalf("expected merged tags in first entry, got %#v", cfg.Entries[0].Tags)
}
}
func TestNormalizeProxiesConfigRejectsInvalidEntries(t *testing.T) {
_, err := NormalizeProxiesConfig(ProxiesConfig{
Entries: []ProxyEntryConfig{{URL: "ftp://proxy:21", Tags: []string{"default"}}},
})
if err == nil {
t.Fatal("expected invalid scheme error")
}
_, err = NormalizeProxiesConfig(ProxiesConfig{
Entries: []ProxyEntryConfig{{URL: "http://proxy:8080", Tags: []string{" "}}},
})
if err == nil {
t.Fatal("expected empty tags error")
}
_, err = NormalizeProxiesConfig(ProxiesConfig{
Global: "ftp://proxy:21",
Entries: []ProxyEntryConfig{{URL: "http://proxy:8080", Tags: []string{"default"}}},
})
if err == nil {
t.Fatal("expected invalid global proxy scheme error")
}
}
func TestResolveEffectiveProxyPolicy(t *testing.T) {
offPolicy := ResolveEffectiveProxyPolicy("", "")
if offPolicy.Mode != ProxyModeOff {
t.Fatalf("expected mode off, got %s", offPolicy.Mode)
}
if offPolicy.Tag != "" {
t.Fatalf("expected empty tag for off mode, got %q", offPolicy.Tag)
}
tagOnlyPolicy := ResolveEffectiveProxyPolicy("", "US")
if tagOnlyPolicy.Mode != ProxyModeTagPool || tagOnlyPolicy.Tag != "us" {
t.Fatalf("unexpected effective policy with tag override: %#v", tagOnlyPolicy)
}
globalPolicy := ResolveEffectiveProxyPolicy("http://proxy-global:8080", "eu")
if globalPolicy.Mode != ProxyModeTagPool || globalPolicy.Tag != "" {
t.Fatalf("expected global proxy to ignore engine tags, got %#v", globalPolicy)
}
}
func TestProxyRegistryRoundRobinAndFailureRecovery(t *testing.T) {
registry, err := NewProxyRegistry([]ProxyEntryConfig{
{URL: "http://proxy1:8080", Tags: []string{"default"}},
{URL: "http://proxy2:8080", Tags: []string{"default"}},
}, 2)
if err != nil {
t.Fatalf("new proxy registry: %v", err)
}
if got := registry.NextByTag("default"); got != "http://proxy1:8080" {
t.Fatalf("expected first proxy1, got %s", got)
}
if got := registry.NextByTag("default"); got != "http://proxy2:8080" {
t.Fatalf("expected second proxy2, got %s", got)
}
registry.ReportFailure("http://proxy1:8080")
registry.ReportFailure("http://proxy1:8080")
if got := registry.NextByTag("default"); got != "http://proxy2:8080" {
t.Fatalf("expected proxy2 while proxy1 disabled, got %s", got)
}
registry.ReportFailure("http://proxy2:8080")
registry.ReportFailure("http://proxy2:8080")
if got := registry.NextByTag("default"); got != "http://proxy1:8080" {
t.Fatalf("expected tag pool reset to proxy1 after exhaustion, got %s", got)
}
registry.ReportFailure("http://proxy1:8080")
registry.ReportSuccess("http://proxy1:8080")
stats := registry.BuildStats()
if stats.UnhealthyCount != 0 {
t.Fatalf("expected no unhealthy proxies after success recovery, got %d", stats.UnhealthyCount)
}
if stats.HealthyCount != 2 {
t.Fatalf("expected two healthy proxies, got %d", stats.HealthyCount)
}
}
func TestMaskProxyURLRedactsCredentials(t *testing.T) {
if got := MaskProxyURL("http://user:pass@127.0.0.1:8080"); got != "http://127.0.0.1:8080" {
t.Fatalf("unexpected masked proxy value: %s", got)
}
}
func TestProxyURLForBrowserLaunchStripsCredentials(t *testing.T) {
u, err := url.Parse("http://user:pass@127.0.0.1:18888")
if err != nil {
t.Fatalf("parse proxy URL: %v", err)
}
got := proxyURLForBrowserLaunch(u)
want := "http://127.0.0.1:18888"
if got != want {
t.Fatalf("expected %q, got %q", want, got)
}
}
func TestProxyURLForBrowserLaunchNormalizesSocks5h(t *testing.T) {
u, err := url.Parse("socks5h://test:test@127.0.0.1:19080")
if err != nil {
t.Fatalf("parse proxy URL: %v", err)
}
got := proxyURLForBrowserLaunch(u)
want := "socks5://127.0.0.1:19080"
if got != want {
t.Fatalf("expected %q, got %q", want, got)
}
}
func TestProxyStatsMaskCredentials(t *testing.T) {
registry, err := NewProxyRegistry([]ProxyEntryConfig{
{URL: "http://user:pass@proxy.example:8080", Tags: []string{"default"}},
}, 1)
if err != nil {
t.Fatalf("new proxy registry: %v", err)
}
stats := registry.BuildStats()
if len(stats.Entries) != 1 {
t.Fatalf("expected one proxy stats entry, got %d", len(stats.Entries))
}
if got := stats.Entries[0].Proxy; got != "http://proxy.example:8080" {
t.Fatalf("expected masked proxy in stats, got %q", got)
}
}
func TestNormalizeProxyTag(t *testing.T) {
tag, err := NormalizeProxyTag(" US ")
if err != nil {
t.Fatalf("normalize proxy tag: %v", err)
}
if tag != "us" {
t.Fatalf("expected normalized tag us, got %q", tag)
}
if _, err := NormalizeProxyTag(" "); err == nil {
t.Fatal("expected empty proxy tag validation error")
}
}
func TestNormalizeProxyRequestOverride(t *testing.T) {
override, err := NormalizeProxyRequestOverride(" direct ")
if err != nil {
t.Fatalf("normalize direct proxy override: %v", err)
}
if override != ProxyOverrideDirect {
t.Fatalf("expected direct override, got %q", override)
}
override, err = NormalizeProxyRequestOverride(" US ")
if err != nil {
t.Fatalf("normalize tagged proxy override: %v", err)
}
if override != "us" {
t.Fatalf("expected normalized tag override us, got %q", override)
}
override, err = NormalizeProxyRequestOverride("")
if err != nil {
t.Fatalf("normalize empty override: %v", err)
}
if override != "" {
t.Fatalf("expected empty override, got %q", override)
}
}
func TestIsAuthenticatedSocksProxyURL(t *testing.T) {
if !IsAuthenticatedSocksProxyURL("socks5h://user:pass@127.0.0.1:1080") {
t.Fatal("expected authenticated socks proxy to be detected")
}
if IsAuthenticatedSocksProxyURL("socks5://127.0.0.1:1080") {
t.Fatal("expected plain socks proxy to remain browser-compatible")
}
if IsAuthenticatedSocksProxyURL("http://user:pass@127.0.0.1:8080") {
t.Fatal("expected HTTP auth proxy to remain browser-compatible")
}
}
func TestNewRawHTTPClientSocks5hUsesProxyDNS(t *testing.T) {
target := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
w.WriteHeader(http.StatusOK)
_, _ = w.Write([]byte("proxied"))
}))
defer target.Close()
targetAddr, err := net.ResolveTCPAddr("tcp", target.Listener.Addr().String())
if err != nil {
t.Fatalf("resolve target listener: %v", err)
}
const proxyOnlyHost = "proxy-target.invalid"
proxyAddr := startSOCKS5TestServer(t, proxyOnlyHost, targetAddr.IP)
directClient := &http.Client{Timeout: 500 * time.Millisecond}
targetURL := fmt.Sprintf("http://%s:%d/", proxyOnlyHost, targetAddr.Port)
if _, err := directClient.Get(targetURL); err == nil {
t.Fatal("expected direct request to fail without proxy DNS")
}
client, err := NewRawHTTPClient(Query{ProxyURL: "socks5h://" + proxyAddr})
if err != nil {
t.Fatalf("new raw http client: %v", err)
}
resp, err := client.Get(targetURL)
if err != nil {
t.Fatalf("expected proxied request to succeed, got %v", err)
}
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
if err != nil {
t.Fatalf("read proxied body: %v", err)
}
if string(body) != "proxied" {
t.Fatalf("unexpected proxied body: %q", string(body))
}
}
type staticResolver struct {
host string
ip net.IP
}
func (r staticResolver) Resolve(ctx xcontext.Context, name string) (xcontext.Context, net.IP, error) {
if name == r.host {
return ctx, r.ip, nil
}
return ctx, nil, net.UnknownNetworkError(name)
}
func startSOCKS5TestServer(t *testing.T, host string, ip net.IP) string {
t.Helper()
server, err := socks5.New(&socks5.Config{
Resolver: staticResolver{host: host, ip: ip},
Logger: log.New(io.Discard, "", 0),
})
if err != nil {
t.Fatalf("create socks5 server: %v", err)
}
listener, err := net.Listen("tcp", "127.0.0.1:0")
if err != nil {
t.Fatalf("listen socks5: %v", err)
}
t.Cleanup(func() {
_ = listener.Close()
})
go func() {
_ = server.Serve(listener)
}()
return listener.Addr().String()
}

436
core/resilient.go Normal file
View File

@@ -0,0 +1,436 @@
package core
import (
"context"
"errors"
"fmt"
"strings"
"sync"
"github.com/sirupsen/logrus"
)
// ResilientSearcher wraps engines with retry and circuit breaker protection.
type ResilientSearcher struct {
engines []SearchEngine
cbManager *CircuitBreakerManager
retryCfg RetryConfig
proxyRuntime string
proxyCfg ProxyConfig
proxyRegistry *ProxyRegistry
proxyDefaults ProxyPolicy
effectivePolicies map[string]ProxyPolicy
}
type ProxyExecutionMeta struct {
Mode string `json:"mode"`
Tag string `json:"tag,omitempty"`
Used string `json:"used"`
}
type ResilientConfig struct {
Retry RetryConfig
CircuitBreaker CircuitBreakerConfig
Proxy ProxyConfig
}
func DefaultResilientConfig() ResilientConfig {
return ResilientConfig{
Retry: DefaultRetryConfig(),
CircuitBreaker: DefaultCircuitBreakerConfig(),
Proxy: DefaultProxyConfig(),
}
}
func NewResilientSearcher(engines []SearchEngine, cfg ResilientConfig) *ResilientSearcher {
proxyCfg, err := NormalizeProxyConfig(cfg.Proxy)
if err != nil {
logrus.Errorf("Invalid proxy config, using defaults: %v", err)
proxyCfg = DefaultProxyConfig()
proxyCfg, _ = NormalizeProxyConfig(proxyCfg)
}
rs := &ResilientSearcher{
engines: engines,
cbManager: NewCircuitBreakerManager(cfg.CircuitBreaker),
retryCfg: cfg.Retry,
proxyRuntime: proxyCfg.Runtime,
proxyCfg: proxyCfg,
proxyRegistry: proxyCfg.Registry,
proxyDefaults: ResolveEffectiveProxyPolicy(proxyCfg.Proxies.Global, ""),
effectivePolicies: make(map[string]ProxyPolicy, len(engines)),
}
for _, engine := range engines {
engineName := normalizeEngineName(engine.Name())
override := proxyCfg.EnginePolicies[engineName]
effective := ResolveEffectiveProxyPolicy(proxyCfg.Proxies.Global, override)
rs.effectivePolicies[engineName] = effective
}
return rs
}
// SearchPrimary keeps dedicated endpoints engine-pure (no fallback).
func (rs *ResilientSearcher) SearchPrimary(primaryEngine SearchEngine, q Query) ([]SearchResult, string, ProxyExecutionMeta, error) {
results, proxyMeta, err := rs.searchWithProtection(primaryEngine, q, false)
if err != nil {
return nil, primaryEngine.Name(), proxyMeta, err
}
return results, primaryEngine.Name(), proxyMeta, nil
}
// SearchWithFallback retries primary and then tries other initialized engines.
func (rs *ResilientSearcher) SearchWithFallback(primaryEngine SearchEngine, q Query) ([]SearchResult, string, ProxyExecutionMeta, error) {
return rs.searchWithFallback(primaryEngine, q, false)
}
func (rs *ResilientSearcher) SearchImagePrimary(primaryEngine SearchEngine, q Query) ([]SearchResult, string, ProxyExecutionMeta, error) {
results, proxyMeta, err := rs.searchWithProtection(primaryEngine, q, true)
if err != nil {
return nil, primaryEngine.Name(), proxyMeta, err
}
return results, primaryEngine.Name(), proxyMeta, nil
}
func (rs *ResilientSearcher) SearchImageWithFallback(primaryEngine SearchEngine, q Query) ([]SearchResult, string, ProxyExecutionMeta, error) {
return rs.searchWithFallback(primaryEngine, q, true)
}
func (rs *ResilientSearcher) searchWithFallback(primaryEngine SearchEngine, q Query, isImage bool) ([]SearchResult, string, ProxyExecutionMeta, error) {
results, proxyMeta, err := rs.searchWithProtection(primaryEngine, q, isImage)
if err == nil {
return results, primaryEngine.Name(), proxyMeta, nil
}
if errors.Is(err, ErrProxyUnavailable) {
logrus.Warnf("[Resilient] Primary engine %s proxy policy failed closed: %s", primaryEngine.Name(), err)
return nil, primaryEngine.Name(), proxyMeta, err
}
action := "failed"
successMessage := "Fallback to %s succeeded with %d results"
if isImage {
action = "image search failed"
successMessage = "Image fallback to %s succeeded with %d results"
}
logrus.Warnf("[Resilient] Primary engine %s %s: %s. Trying fallback engines...", primaryEngine.Name(), action, err)
for _, fallbackEngine := range rs.engines {
if fallbackEngine.Name() == primaryEngine.Name() || !fallbackEngine.IsInitialized() {
continue
}
results, fallbackMeta, fallbackErr := rs.searchWithProtection(fallbackEngine, q, isImage)
if fallbackErr == nil {
logrus.Infof("[Resilient] "+successMessage, fallbackEngine.Name(), len(results))
return results, fallbackEngine.Name(), fallbackMeta, nil
}
logrus.Warnf("[Resilient] Fallback engine %s also failed: %s", fallbackEngine.Name(), fallbackErr)
}
return nil, primaryEngine.Name(), proxyMeta, ErrAllEnginesFailed
}
func (rs *ResilientSearcher) searchWithProtection(engine SearchEngine, q Query, isImage bool) ([]SearchResult, ProxyExecutionMeta, error) {
cb := rs.cbManager.Get(engine.Name())
if !cb.AllowRequest() {
return nil, ProxyExecutionMeta{}, ErrCircuitOpen
}
policy := rs.effectivePolicyForQuery(engine.Name(), q)
attemptMeta := rs.baseProxyMeta(policy)
result := RetryableSearch(rs.retryCfg, engine.Name(), func() ([]SearchResult, error) {
limiter := engine.GetRateLimiter()
if limiter != nil {
if err := limiter.Wait(context.Background()); err != nil {
return nil, err
}
}
attemptQuery := q
proxyURL := ""
reportToRegistry := false
attemptMeta = rs.baseProxyMeta(policy)
switch policy.Mode {
case ProxyModeOff:
attemptQuery.ProxyURL = ""
attemptMeta.Used = "direct"
case ProxyModeTagPool:
proxyURL = rs.selectProxyForQuery(policy, q)
if proxyURL == "" {
return nil, fmt.Errorf("%w: no healthy proxy available for tag %q", ErrProxyUnavailable, policy.Tag)
}
attemptQuery.ProxyURL = proxyURL
reportToRegistry = policy.Tag != ""
attemptMeta.Used = MaskProxyURL(proxyURL)
}
var (
results []SearchResult
err error
)
if isImage {
results, err = engine.SearchImage(attemptQuery)
} else {
results, err = engine.Search(attemptQuery)
}
if reportToRegistry {
rs.reportProxyAttempt(proxyURL, err)
}
return results, err
})
if result.Err != nil {
if !errors.Is(result.Err, ErrProxyUnavailable) {
cb.RecordFailure()
}
return nil, attemptMeta, result.Err
}
cb.RecordSuccess()
return result.Results, attemptMeta, nil
}
// SearchAllParallel applies retry/circuit protections per engine for mega search.
func (rs *ResilientSearcher) SearchAllParallel(q Query, engines []SearchEngine) []MegaSearchResult {
var wg sync.WaitGroup
var mu sync.Mutex
var allResults []MegaSearchResult
for _, engine := range engines {
if !engine.IsInitialized() {
continue
}
if !rs.cbManager.Get(engine.Name()).AllowRequest() {
logrus.Infof("[Resilient] Skipping %s in megasearch (circuit open)", engine.Name())
continue
}
wg.Add(1)
go func(eng SearchEngine) {
defer wg.Done()
results, _, err := rs.searchWithProtection(eng, q, false)
if err != nil {
return
}
mu.Lock()
for _, r := range results {
allResults = append(allResults, MegaSearchResult{
SearchResult: r,
Engine: eng.Name(),
})
}
mu.Unlock()
}(engine)
}
wg.Wait()
return allResults
}
func (rs *ResilientSearcher) SearchAllImageParallel(q Query, engines []SearchEngine) []MegaSearchResult {
var wg sync.WaitGroup
var mu sync.Mutex
var allResults []MegaSearchResult
for _, engine := range engines {
if !engine.IsInitialized() {
continue
}
if !rs.cbManager.Get(engine.Name()).AllowRequest() {
logrus.Infof("[Resilient] Skipping %s in megaimage (circuit open)", engine.Name())
continue
}
wg.Add(1)
go func(eng SearchEngine) {
defer wg.Done()
results, _, err := rs.searchWithProtection(eng, q, true)
if err != nil {
return
}
mu.Lock()
for _, r := range results {
allResults = append(allResults, MegaSearchResult{
SearchResult: r,
Engine: eng.Name(),
})
}
mu.Unlock()
}(engine)
}
wg.Wait()
return allResults
}
func (rs *ResilientSearcher) GetCircuitBreakerStats() []map[string]interface{} {
return rs.cbManager.AllStats()
}
func (rs *ResilientSearcher) GetProxyStats() ProxyStats {
stats := ProxyStats{
ConfiguredCount: 0,
HealthyCount: 0,
UnhealthyCount: 0,
Tags: map[string]ProxyTagSummary{},
Entries: []ProxyStatsEntry{},
}
if rs.proxyRegistry != nil {
stats = rs.proxyRegistry.BuildStats()
}
engines := map[string]ProxyEngineStats{}
for _, engine := range rs.engines {
engineName := normalizeEngineName(engine.Name())
policy := rs.effectivePolicyForEngine(engineName)
engineStats := ProxyEngineStats{}
switch policy.Mode {
case ProxyModeOff:
engineStats.SelectedProxy = "direct"
case ProxyModeTagPool:
engineStats.Tag = policy.Tag
if global := strings.TrimSpace(rs.proxyCfg.Proxies.Global); global != "" {
engineStats.SelectedProxy = MaskProxyURL(global)
} else {
engineStats.SelectedProxy = "pooled"
}
}
engines[engineName] = engineStats
}
if len(engines) > 0 {
stats.Engines = engines
}
return stats
}
func (rs *ResilientSearcher) ResolveMegaProxyMeta(q Query, engines []SearchEngine) ProxyExecutionMeta {
if len(engines) == 0 {
return ProxyExecutionMeta{Mode: ProxyModeOff, Used: "direct"}
}
allOff := true
proxiedTags := map[string]struct{}{}
hasOff := false
for _, engine := range engines {
policy := rs.effectivePolicyForQuery(engine.Name(), q)
if policy.Mode == ProxyModeOff {
hasOff = true
continue
}
allOff = false
if policy.Tag != "" {
proxiedTags[policy.Tag] = struct{}{}
}
}
if allOff {
return ProxyExecutionMeta{Mode: ProxyModeOff, Used: "direct"}
}
meta := ProxyExecutionMeta{Mode: ProxyModeTagPool}
if len(proxiedTags) == 1 {
for tag := range proxiedTags {
meta.Tag = tag
}
}
if q.ProxyOverride == "" {
if global := strings.TrimSpace(rs.proxyCfg.Proxies.Global); global != "" && !hasOff {
meta.Used = MaskProxyURL(global)
return meta
}
}
if rs.proxyRuntime == ProxyRuntimeRaw {
meta.Used = "multiple"
if hasOff {
meta.Used = "mixed"
}
return meta
}
meta.Used = "multiple"
if hasOff {
meta.Used = "mixed"
}
return meta
}
func (rs *ResilientSearcher) baseProxyMeta(policy ProxyPolicy) ProxyExecutionMeta {
meta := ProxyExecutionMeta{Mode: policy.Mode}
if policy.Mode == ProxyModeTagPool {
meta.Tag = policy.Tag
return meta
}
meta.Used = "direct"
return meta
}
func (rs *ResilientSearcher) effectivePolicyForEngine(engineName string) ProxyPolicy {
engineName = normalizeEngineName(engineName)
if policy, ok := rs.effectivePolicies[engineName]; ok {
return policy
}
return rs.proxyDefaults
}
func (rs *ResilientSearcher) effectivePolicyForQuery(engineName string, q Query) ProxyPolicy {
switch q.ProxyOverride {
case "":
return rs.effectivePolicyForEngine(engineName)
case ProxyOverrideDirect:
return ProxyPolicy{Mode: ProxyModeOff}
default:
return ProxyPolicy{Mode: ProxyModeTagPool, Tag: q.ProxyOverride}
}
}
func (rs *ResilientSearcher) selectProxyForTag(tag string) string {
if rs.proxyRegistry == nil {
return ""
}
return rs.proxyRegistry.NextByTag(tag)
}
func (rs *ResilientSearcher) reportProxyAttempt(proxyURL string, err error) {
if rs.proxyRegistry == nil || proxyURL == "" {
return
}
if err != nil {
rs.proxyRegistry.ReportFailure(proxyURL)
return
}
rs.proxyRegistry.ReportSuccess(proxyURL)
}
func (rs *ResilientSearcher) selectProxyForQuery(policy ProxyPolicy, q Query) string {
if policy.Mode != ProxyModeTagPool {
return ""
}
if q.ProxyOverride == "" {
if global := strings.TrimSpace(rs.proxyCfg.Proxies.Global); global != "" {
return global
}
}
return rs.selectProxyForTag(policy.Tag)
}
var ErrAllEnginesFailed = fmt.Errorf("all search engines failed")

100
core/retry.go Normal file
View File

@@ -0,0 +1,100 @@
package core
import (
"errors"
"fmt"
"math"
"time"
"github.com/sirupsen/logrus"
)
// RetryConfig controls retry behavior.
type RetryConfig struct {
MaxRetries int
InitialBackoff time.Duration
MaxBackoff time.Duration
BackoffFactor float64
}
func DefaultRetryConfig() RetryConfig {
return RetryConfig{
MaxRetries: 3,
InitialBackoff: time.Second,
MaxBackoff: 30 * time.Second,
BackoffFactor: 2.0,
}
}
type RetryResult struct {
Results []SearchResult
Err error
Attempts int
Engine string
}
// RetryableSearch executes searchFn with exponential backoff retries.
// CAPTCHA and proxy-unavailable errors are not retried.
func RetryableSearch(cfg RetryConfig, engineName string, searchFn func() ([]SearchResult, error)) RetryResult {
if cfg.BackoffFactor <= 0 {
cfg.BackoffFactor = 2.0
}
var lastErr error
for attempt := 0; attempt <= cfg.MaxRetries; attempt++ {
if attempt > 0 {
backoff := calculateBackoff(cfg, attempt)
logrus.Warnf("[%s] Retry attempt %d/%d after %s", engineName, attempt, cfg.MaxRetries, backoff)
time.Sleep(backoff)
}
results, err := searchFn()
if err == nil {
if attempt > 0 {
logrus.Infof("[%s] Succeeded on retry attempt %d", engineName, attempt)
}
return RetryResult{
Results: results,
Attempts: attempt + 1,
Engine: engineName,
}
}
lastErr = err
if errors.Is(err, ErrCaptcha) {
logrus.Warnf("[%s] CAPTCHA detected, skipping retries", engineName)
return RetryResult{
Err: err,
Attempts: attempt + 1,
Engine: engineName,
}
}
if errors.Is(err, ErrProxyUnavailable) {
logrus.Warnf("[%s] Proxy unavailable, skipping retries", engineName)
return RetryResult{
Err: err,
Attempts: attempt + 1,
Engine: engineName,
}
}
logrus.Warnf("[%s] Attempt %d failed: %s", engineName, attempt+1, err)
}
return RetryResult{
Err: fmt.Errorf("all %d attempts failed for %s: %w", cfg.MaxRetries+1, engineName, lastErr),
Attempts: cfg.MaxRetries + 1,
Engine: engineName,
}
}
func calculateBackoff(cfg RetryConfig, attempt int) time.Duration {
backoff := float64(cfg.InitialBackoff) * math.Pow(cfg.BackoffFactor, float64(attempt-1))
if backoff > float64(cfg.MaxBackoff) {
backoff = float64(cfg.MaxBackoff)
}
if backoff < 0 {
backoff = 0
}
return time.Duration(backoff)
}

86
core/retry_test.go Normal file
View File

@@ -0,0 +1,86 @@
package core
import (
"errors"
"testing"
"time"
)
func TestRetryableSearch_SuccessOnFirstAttempt(t *testing.T) {
cfg := RetryConfig{MaxRetries: 3, InitialBackoff: 10 * time.Millisecond, MaxBackoff: 100 * time.Millisecond, BackoffFactor: 2.0}
calls := 0
result := RetryableSearch(cfg, "test", func() ([]SearchResult, error) {
calls++
return []SearchResult{{Title: "result1"}}, nil
})
if result.Err != nil {
t.Fatalf("expected no error, got: %v", result.Err)
}
if result.Attempts != 1 {
t.Errorf("expected 1 attempt, got: %d", result.Attempts)
}
if calls != 1 {
t.Errorf("expected 1 call, got: %d", calls)
}
}
func TestRetryableSearch_AllAttemptsFail(t *testing.T) {
cfg := RetryConfig{MaxRetries: 2, InitialBackoff: 10 * time.Millisecond, MaxBackoff: 50 * time.Millisecond, BackoffFactor: 2.0}
calls := 0
result := RetryableSearch(cfg, "test", func() ([]SearchResult, error) {
calls++
return nil, errors.New("persistent failure")
})
if result.Err == nil {
t.Fatal("expected error, got nil")
}
if calls != 3 {
t.Errorf("expected 3 calls (1 + 2 retries), got: %d", calls)
}
if result.Attempts != 3 {
t.Errorf("expected 3 attempts, got: %d", result.Attempts)
}
}
func TestRetryableSearch_CaptchaNotRetried(t *testing.T) {
cfg := RetryConfig{MaxRetries: 3, InitialBackoff: 10 * time.Millisecond, MaxBackoff: 100 * time.Millisecond, BackoffFactor: 2.0}
calls := 0
result := RetryableSearch(cfg, "test", func() ([]SearchResult, error) {
calls++
return nil, ErrCaptcha
})
if !errors.Is(result.Err, ErrCaptcha) {
t.Fatalf("expected ErrCaptcha, got: %v", result.Err)
}
if calls != 1 {
t.Errorf("expected 1 call, got: %d", calls)
}
}
func TestCalculateBackoff(t *testing.T) {
cfg := RetryConfig{InitialBackoff: 1 * time.Second, MaxBackoff: 10 * time.Second, BackoffFactor: 2.0}
tests := []struct {
attempt int
expected time.Duration
}{
{1, 1 * time.Second},
{2, 2 * time.Second},
{3, 4 * time.Second},
{4, 8 * time.Second},
{5, 10 * time.Second},
}
for _, tt := range tests {
got := calculateBackoff(cfg, tt.attempt)
if got != tt.expected {
t.Errorf("attempt %d: expected %s, got %s", tt.attempt, tt.expected, got)
}
}
}

View File

@@ -1,11 +1,13 @@
package core
import (
"context"
"encoding/json"
"errors"
"fmt"
"runtime"
"sort"
"strings"
"sync"
"time"
"github.com/gofiber/fiber/v2"
"github.com/sirupsen/logrus"
@@ -24,227 +26,363 @@ type Server struct {
app *fiber.App
addr string
searchEngines []SearchEngine
cache *ResponseCache
resilient *ResilientSearcher
startTime time.Time
opts ServerOptions
}
type ServerOptions struct {
CacheTTL time.Duration
CacheMaxSize int
EnableCORS bool
CORS CORSConfig
AllowEndpointFallback bool
Resilience ResilientConfig
}
func DefaultServerOptions() ServerOptions {
return ServerOptions{
CacheTTL: 5 * time.Minute,
CacheMaxSize: 1000,
EnableCORS: true,
CORS: DefaultCORSConfig(),
AllowEndpointFallback: false,
Resilience: DefaultResilientConfig(),
}
}
func NewServer(host string, port int, searchEngines ...SearchEngine) *Server {
return NewServerWithOptions(host, port, DefaultServerOptions(), searchEngines...)
}
func NewServerWithOptions(host string, port int, opts ServerOptions, searchEngines ...SearchEngine) *Server {
addr := fmt.Sprintf("%s:%d", host, port)
app := fiber.New(fiber.Config{
ErrorHandler: JSONErrorMiddleware(),
})
serv := Server{
app: fiber.New(),
app: app,
addr: addr,
searchEngines: searchEngines,
resilient: NewResilientSearcher(searchEngines, opts.Resilience),
startTime: time.Now(),
opts: opts,
}
logrus.Info("Resilient search enabled: retry + circuit breaker")
if opts.AllowEndpointFallback {
logrus.Warn("Dedicated endpoint fallback is enabled")
}
if opts.CacheTTL > 0 && opts.CacheMaxSize > 0 {
serv.cache = NewResponseCache(opts.CacheTTL, opts.CacheMaxSize)
logrus.Infof("Response cache enabled: TTL=%s, MaxSize=%d", opts.CacheTTL, opts.CacheMaxSize)
}
if opts.EnableCORS {
app.Use(CORSMiddleware(opts.CORS))
}
app.Use(RequestLoggerMiddleware())
app.Get("/health", serv.handleHealthCheck)
app.Get("/stats", serv.handleStats)
app.Get("/stats/cache", serv.handleCacheStats)
app.Get("/stats/proxy", serv.handleProxyStats)
app.Get("/stats/cb", serv.handleCircuitBreakerStats)
for _, engine := range searchEngines {
locEngine := engine
limiter := engine.GetRateLimiter()
// Custom endpoint mapping for DuckDuckGo
endpointName := strings.ToLower(locEngine.Name())
if endpointName == "duckduckgo" {
endpointName = "duck"
}
serv.app.Get(fmt.Sprintf("/%s/search", endpointName), func(c *fiber.Ctx) error {
q := Query{}
err := q.InitFromContext(c)
if err != nil {
logrus.Errorf("Error while setting %s query: %s", locEngine.Name(), err)
return err
}
logrus.Infof("Starting SERP search request using %s engine for query: %s", locEngine.Name(), q.Text)
err = limiter.Wait(context.Background())
if err != nil {
logrus.Errorf("Ratelimiter error during %s query: %s", locEngine.Name(), err)
}
res, err := locEngine.Search(q)
if err != nil {
switch err {
case ErrCaptcha:
err = fmt.Errorf("captcha found, please stop sending requests for a while\n%s", err)
case ErrSearchTimeout:
err = fmt.Errorf("%s", err)
}
logrus.Errorf("Error during %s search: %s", locEngine.Name(), err)
return fiber.NewError(fiber.StatusServiceUnavailable, err.Error())
}
logrus.Infof("Successfully completed SERP search using %s engine, returned %d results", locEngine.Name(), len(res))
return c.JSON(res)
return serv.handleDedicatedEndpoint(c, locEngine, false)
})
serv.app.Get(fmt.Sprintf("/%s/image", endpointName), func(c *fiber.Ctx) error {
q := Query{}
err := q.InitFromContext(c)
if err != nil {
logrus.Errorf("Error while setting %s query: %s", locEngine.Name(), err)
return err
}
logrus.Infof("Starting SERP image search request using %s engine for query: %s", locEngine.Name(), q.Text)
err = limiter.Wait(context.Background())
if err != nil {
logrus.Errorf("Ratelimiter error during %s query: %s", locEngine.Name(), err)
}
res, err := locEngine.SearchImage(q)
if err != nil && len(res) > 0 {
logrus.Warnf("Partial results returned from %s image search despite error: %s", locEngine.Name(), err)
c.Status(503)
return c.JSON(res)
}
if err != nil {
switch err {
case ErrCaptcha:
err = fmt.Errorf("captcha found, please stop sending requests for a while: %s", err)
case ErrSearchTimeout:
err = fmt.Errorf("%s", err)
}
logrus.Errorf("Error during %s image search: %s", locEngine.Name(), err)
return fiber.NewError(fiber.StatusServiceUnavailable, err.Error())
}
logrus.Infof("Successfully completed SERP image search using [%s], returned %d results", locEngine.Name(), len(res))
return c.JSON(res)
return serv.handleDedicatedEndpoint(c, locEngine, true)
})
}
// Add megasearch endpoint
serv.app.Get("/mega/search", serv.handleMegaSearch)
// Add megasearch image endpoint
serv.app.Get("/mega/image", serv.handleMegaImage)
// Add endpoint to list available engines
serv.app.Get("/mega/engines", serv.handleListEngines)
return &serv
}
// MegaSearchResult represents a search result with engine information
func (s *Server) handleDedicatedEndpoint(c *fiber.Ctx, engine SearchEngine, isImage bool) error {
q := Query{}
if err := q.InitFromContext(c); err != nil {
logrus.Errorf("Error while setting %s query: %s", engine.Name(), err)
return err
}
action := "search"
if isImage {
action = "image"
}
logrus.Infof("Starting SERP %s request using %s engine for query: %s", action, engine.Name(), q.Text)
if hit, err := s.tryServeCacheHit(
c,
cacheHitCandidate{
key: BuildCacheKey(engine.Name(), action, q),
logMessage: fmt.Sprintf("Cache hit for %s %s: %s", engine.Name(), action, q.Text),
},
); hit || err != nil {
return err
}
var (
res []SearchResult
usedEngine string
proxyMeta ProxyExecutionMeta
searchErr error
)
if isImage {
if s.opts.AllowEndpointFallback {
res, usedEngine, proxyMeta, searchErr = s.resilient.SearchImageWithFallback(engine, q)
} else {
res, usedEngine, proxyMeta, searchErr = s.resilient.SearchImagePrimary(engine, q)
}
} else {
if s.opts.AllowEndpointFallback {
res, usedEngine, proxyMeta, searchErr = s.resilient.SearchWithFallback(engine, q)
} else {
res, usedEngine, proxyMeta, searchErr = s.resilient.SearchPrimary(engine, q)
}
}
s.applyProxyHeaders(c, proxyMeta)
if searchErr != nil {
errToReturn := searchErr
switch searchErr {
case ErrCaptcha:
errToReturn = fmt.Errorf("captcha found, please stop sending requests for a while: %w", searchErr)
case ErrSearchTimeout:
errToReturn = fmt.Errorf("%s", searchErr)
default:
if errors.Is(searchErr, ErrProxyUnavailable) {
errToReturn = fmt.Errorf("%s", searchErr)
}
}
logrus.Errorf("Error during resilient %s %s: %s", engine.Name(), action, searchErr)
return fiber.NewError(fiber.StatusServiceUnavailable, errToReturn.Error())
}
cacheStatus := ""
// Avoid caching fallback-served responses so the requested engine can recover
// without the endpoint continuing to serve another engine until TTL expiry.
if s.cache != nil {
cacheStatus = "BYPASS"
switch {
case usedEngine != engine.Name():
s.cache.RecordBypass()
case len(res) == 0:
s.cache.RecordBypass()
default:
cacheKey := BuildCacheKey(engine.Name(), action, q)
if s.cacheJSON(cacheKey, res) {
cacheStatus = "MISS"
}
}
c.Set("X-Cache", cacheStatus)
}
if usedEngine != "" && usedEngine != engine.Name() {
c.Set("X-Fallback-Engine", usedEngine)
}
logrus.Infof("Successfully completed SERP %s using %s, returned %d results", action, usedEngine, len(res))
return c.JSON(res)
}
type HealthStatus struct {
Status string `json:"status"`
Uptime string `json:"uptime"`
Engines []EngineHealth `json:"engines"`
System map[string]interface{} `json:"system"`
}
type EngineHealth struct {
Name string `json:"name"`
Initialized bool `json:"initialized"`
Status string `json:"status"`
}
// handleHealthCheck returns current service and engine status.
// Degraded state stays HTTP 200 to avoid unnecessary restarts in orchestrators.
func (s *Server) handleHealthCheck(c *fiber.Ctx) error {
engines := make([]EngineHealth, 0, len(s.searchEngines))
availableEngines := 0
for _, engine := range s.searchEngines {
status := "ready"
isAvailable := true
if !engine.IsInitialized() {
status = "not_initialized"
isAvailable = false
}
for _, cbStat := range s.resilient.GetCircuitBreakerStats() {
engineName, _ := cbStat["engine"].(string)
if engineName != engine.Name() {
continue
}
circuitState, _ := cbStat["state"].(string)
if circuitState == "open" {
status = "circuit_open"
isAvailable = false
}
break
}
if isAvailable {
availableEngines++
}
engines = append(engines, EngineHealth{
Name: engine.Name(),
Initialized: engine.IsInitialized(),
Status: status,
})
}
overallStatus := "healthy"
totalEngines := len(s.searchEngines)
switch {
case totalEngines == 0 || availableEngines == 0:
overallStatus = "unhealthy"
case availableEngines < totalEngines:
overallStatus = "degraded"
}
var memStats runtime.MemStats
runtime.ReadMemStats(&memStats)
health := HealthStatus{
Status: overallStatus,
Uptime: time.Since(s.startTime).Round(time.Second).String(),
Engines: engines,
System: map[string]interface{}{
"goroutines": runtime.NumGoroutine(),
"memory_mb": memStats.Alloc / 1024 / 1024,
"go_version": runtime.Version(),
},
}
if overallStatus == "unhealthy" {
c.Status(fiber.StatusServiceUnavailable)
}
return c.JSON(health)
}
func (s *Server) handleStats(c *fiber.Ctx) error {
return c.JSON(map[string]interface{}{
"cache": s.cacheStatsPayload(),
"proxy": s.resilient.GetProxyStats(),
"circuit_breakers": s.resilient.GetCircuitBreakerStats(),
})
}
func (s *Server) handleCacheStats(c *fiber.Ctx) error {
return c.JSON(s.cacheStatsPayload())
}
func (s *Server) handleProxyStats(c *fiber.Ctx) error {
return c.JSON(s.resilient.GetProxyStats())
}
func (s *Server) handleCircuitBreakerStats(c *fiber.Ctx) error {
return c.JSON(map[string]interface{}{
"circuit_breakers": s.resilient.GetCircuitBreakerStats(),
})
}
type MegaSearchResult struct {
SearchResult
Engine string `json:"engine"`
}
// handleMegaSearch handles the /megasearch endpoint
func (s *Server) handleMegaSearch(c *fiber.Ctx) error {
q := Query{}
err := q.InitFromContext(c)
if err != nil {
logrus.Errorf("Error while setting megasearch query: %s", err)
return err
}
// Get engines parameter to filter which engines to use
enginesParam := c.Query("engines", "")
var enginesToUse []SearchEngine
if enginesParam != "" {
// Parse comma-separated list of engines
engineNames := strings.Split(enginesParam, ",")
for _, engineName := range engineNames {
engineName = strings.TrimSpace(strings.ToLower(engineName))
for _, engine := range s.searchEngines {
if strings.ToLower(engine.Name()) == engineName {
enginesToUse = append(enginesToUse, engine)
break
}
}
}
} else {
// Use all engines if no specific engines specified
enginesToUse = s.searchEngines
}
if len(enginesToUse) == 0 {
return fiber.NewError(fiber.StatusBadRequest, "No valid search engines specified")
}
// Log which engines will be used
engineNames := make([]string, len(enginesToUse))
for i, engine := range enginesToUse {
engineNames[i] = engine.Name()
}
logrus.Infof("Starting SERP megasearch request using engines: %s for query: %s", strings.Join(engineNames, ", "), q.Text)
// Execute searches in parallel across selected engines
results := s.searchSelectedEngines(q, enginesToUse)
// Deduplicate results while preserving engine information
dedupedResults := s.deduplicateMegaResults(results)
logrus.Infof("Successfully completed SERP megasearch using %d engines, returned %d deduplicated results", len(enginesToUse), len(dedupedResults))
return c.JSON(dedupedResults)
return s.handleMegaEndpoint(c, "search", s.resilient.SearchAllParallel)
}
// handleMegaImage handles the /mega/image endpoint
func (s *Server) handleMegaImage(c *fiber.Ctx) error {
return s.handleMegaEndpoint(c, "image", s.resilient.SearchAllImageParallel)
}
func (s *Server) handleMegaEndpoint(c *fiber.Ctx, action string, run func(Query, []SearchEngine) []MegaSearchResult) error {
q := Query{}
err := q.InitFromContext(c)
if err != nil {
logrus.Errorf("Error while setting megasearch image query: %s", err)
if err := q.InitFromContext(c); err != nil {
logrus.Errorf("Error while setting mega %s query: %s", action, err)
return err
}
// Get engines parameter to filter which engines to use
enginesParam := c.Query("engines", "")
var enginesToUse []SearchEngine
if enginesParam != "" {
// Parse comma-separated list of engines
engineNames := strings.Split(enginesParam, ",")
for _, engineName := range engineNames {
engineName = strings.TrimSpace(strings.ToLower(engineName))
for _, engine := range s.searchEngines {
if strings.ToLower(engine.Name()) == engineName {
enginesToUse = append(enginesToUse, engine)
break
}
}
}
} else {
// Use all engines if no specific engines specified
enginesToUse = s.searchEngines
}
enginesToUse := s.resolveEngines(c.Query("engines", ""))
if len(enginesToUse) == 0 {
return fiber.NewError(fiber.StatusBadRequest, "No valid search engines specified")
}
// Log which engines will be used
engineNames := make([]string, len(enginesToUse))
for i, engine := range enginesToUse {
engineNames[i] = engine.Name()
}
logrus.Infof("Starting SERP megasearch image request using engines: %s for query: %s", strings.Join(engineNames, ", "), q.Text)
engineNamesJoined := strings.Join(engineNames, ",")
s.applyProxyHeaders(c, s.resilient.ResolveMegaProxyMeta(q, enginesToUse))
logrus.Infof("Starting SERP mega %s request using engines: %s for query: %s", action, engineNamesJoined, q.Text)
// Execute image searches in parallel across selected engines
results := s.searchSelectedEnginesImage(q, enginesToUse)
cacheHitCandidates := []cacheHitCandidate{
{
key: s.buildMegaCacheKey(action, enginesToUse, q),
logMessage: fmt.Sprintf("Cache hit for mega %s: engines=%s query=%s", action, engineNamesJoined, q.Text),
},
}
cacheableEngines := s.megaCacheableEngines(enginesToUse)
if len(cacheableEngines) > 0 && len(cacheableEngines) < len(enginesToUse) {
cacheHitCandidates = append(cacheHitCandidates, cacheHitCandidate{
key: s.buildMegaCacheKey(action, cacheableEngines, q),
logMessage: fmt.Sprintf("Cache hit for mega %s partial set: engines=%s query=%s", action, engineNamesJoined, q.Text),
})
}
if hit, err := s.tryServeCacheHit(c, cacheHitCandidates...); hit || err != nil {
return err
}
// Deduplicate results while preserving engine information
results := run(q, enginesToUse)
dedupedResults := s.deduplicateMegaResults(results)
logrus.Infof("Successfully completed SERP megasearch image using %d engines, returned %d deduplicated results", len(enginesToUse), len(dedupedResults))
if s.cache != nil {
c.Set("X-Cache", s.cacheMegaResults(action, enginesToUse, q, dedupedResults))
}
logrus.Infof("Successfully completed SERP mega %s using %d engines, returned %d deduplicated results", action, len(enginesToUse), len(dedupedResults))
return c.JSON(dedupedResults)
}
// handleListEngines lists all available search engines
func (s *Server) handleListEngines(c *fiber.Ctx) error {
var engines []map[string]interface{}
for _, engine := range s.searchEngines {
engines = append(engines, map[string]interface{}{
engineInfo := map[string]interface{}{
"name": engine.Name(),
"initialized": engine.IsInitialized(),
})
}
for _, cbStat := range s.resilient.GetCircuitBreakerStats() {
engineName, _ := cbStat["engine"].(string)
if engineName == engine.Name() {
engineInfo["circuit_state"] = cbStat["state"]
break
}
}
engines = append(engines, engineInfo)
}
return c.JSON(map[string]interface{}{
@@ -253,123 +391,178 @@ func (s *Server) handleListEngines(c *fiber.Ctx) error {
})
}
// searchSelectedEngines performs parallel searches across selected engines
func (s *Server) searchSelectedEngines(q Query, engines []SearchEngine) []MegaSearchResult {
var wg sync.WaitGroup
var mu sync.Mutex
var allResults []MegaSearchResult
for _, engine := range engines {
wg.Add(1)
go func(eng SearchEngine) {
defer wg.Done()
// Apply rate limiting
limiter := eng.GetRateLimiter()
if limiter != nil {
err := limiter.Wait(context.Background())
if err != nil {
logrus.Errorf("Ratelimiter error during %s megasearch: %s", eng.Name(), err)
}
}
// Perform search
results, err := eng.Search(q)
if err != nil {
logrus.Errorf("Error during %s megasearch: %s", eng.Name(), err)
return
}
// Convert to MegaSearchResult with engine info
mu.Lock()
for _, result := range results {
megaResult := MegaSearchResult{
SearchResult: result,
Engine: eng.Name(),
}
allResults = append(allResults, megaResult)
}
mu.Unlock()
}(engine)
func (s *Server) resolveEngines(enginesParam string) []SearchEngine {
if enginesParam == "" {
return s.searchEngines
}
wg.Wait()
return allResults
}
// searchSelectedEnginesImage performs parallel image searches across selected engines
func (s *Server) searchSelectedEnginesImage(q Query, engines []SearchEngine) []MegaSearchResult {
var wg sync.WaitGroup
var mu sync.Mutex
var allResults []MegaSearchResult
for _, engine := range engines {
wg.Add(1)
go func(eng SearchEngine) {
defer wg.Done()
// Apply rate limiting
limiter := eng.GetRateLimiter()
if limiter != nil {
err := limiter.Wait(context.Background())
if err != nil {
logrus.Errorf("Ratelimiter error during %s megasearch image: %s", eng.Name(), err)
}
var enginesToUse []SearchEngine
seen := make(map[string]bool)
engineNames := strings.Split(enginesParam, ",")
for _, engineName := range engineNames {
engineName = strings.TrimSpace(strings.ToLower(engineName))
if engineName == "" || seen[engineName] {
continue
}
for _, engine := range s.searchEngines {
if strings.ToLower(engine.Name()) == engineName {
enginesToUse = append(enginesToUse, engine)
seen[engineName] = true
break
}
// Perform image search
results, err := eng.SearchImage(q)
if err != nil {
logrus.Errorf("Error during %s megasearch image: %s", eng.Name(), err)
return
}
// Convert to MegaSearchResult with engine info
mu.Lock()
for _, result := range results {
megaResult := MegaSearchResult{
SearchResult: result,
Engine: eng.Name(),
}
allResults = append(allResults, megaResult)
}
mu.Unlock()
}(engine)
}
}
wg.Wait()
return allResults
return enginesToUse
}
// deduplicateMegaResults deduplicates results while preserving engine information
func (s *Server) deduplicateMegaResults(results []MegaSearchResult) []MegaSearchResult {
urlMap := make(map[string]MegaSearchResult)
// Process results and keep the first occurrence of each URL
for _, result := range results {
if result.URL == "" {
continue
}
if _, exists := urlMap[result.URL]; !exists {
urlMap[result.URL] = result
}
}
// Convert map back to slice and sort by rank
var deduped []MegaSearchResult
for _, result := range urlMap {
deduped = append(deduped, result)
}
// Sort by rank
sort.Slice(deduped, func(i, j int) bool {
return deduped[i].Rank < deduped[j].Rank
})
return deduped
}
type cacheHitCandidate struct {
key string
logMessage string
}
func (s *Server) tryServeCacheHit(c *fiber.Ctx, candidates ...cacheHitCandidate) (bool, error) {
if s.cache == nil {
return false, nil
}
for _, candidate := range candidates {
cached, ok := s.cache.Get(candidate.key)
if !ok {
continue
}
c.Set("Content-Type", "application/json")
c.Set("X-Cache", "HIT")
logrus.Info(candidate.logMessage)
return true, c.Send(cached)
}
return false, nil
}
func (s *Server) cacheJSON(cacheKey string, payload interface{}) bool {
if s.cache == nil {
return false
}
data, err := json.Marshal(payload)
if err != nil {
s.cache.RecordBypass()
return false
}
s.cache.Set(cacheKey, data)
return true
}
func (s *Server) cacheMegaResults(action string, enginesToUse []SearchEngine, q Query, dedupedResults []MegaSearchResult) string {
cacheStatus := "BYPASS"
if s.cache == nil {
return cacheStatus
}
if len(dedupedResults) == 0 {
s.cache.RecordBypass()
return cacheStatus
}
cacheEngines := s.megaCacheableEngines(enginesToUse)
if len(cacheEngines) == 0 {
s.cache.RecordBypass()
return cacheStatus
}
if s.cacheJSON(s.buildMegaCacheKey(action, cacheEngines, q), dedupedResults) {
return "MISS"
}
return cacheStatus
}
func (s *Server) buildMegaCacheKey(action string, engines []SearchEngine, q Query) string {
names := make([]string, 0, len(engines))
for _, eng := range engines {
names = append(names, strings.ToLower(strings.TrimSpace(eng.Name())))
}
sort.Strings(names)
// Deduplicate engine names in key to keep cache stable when order differs
// or repeated names are passed in the engines query parameter.
uniq := names[:0]
last := ""
for _, name := range names {
if name == last {
continue
}
uniq = append(uniq, name)
last = name
}
return BuildCacheKey("mega:"+strings.Join(uniq, ","), action, q)
}
func (s *Server) megaCacheableEngines(engines []SearchEngine) []SearchEngine {
open := make(map[string]bool)
for _, stat := range s.resilient.GetCircuitBreakerStats() {
name, _ := stat["engine"].(string)
state, _ := stat["state"].(string)
if strings.EqualFold(state, "open") {
open[strings.ToLower(strings.TrimSpace(name))] = true
}
}
cacheable := make([]SearchEngine, 0, len(engines))
for _, eng := range engines {
if open[strings.ToLower(strings.TrimSpace(eng.Name()))] {
continue
}
cacheable = append(cacheable, eng)
}
return cacheable
}
func (s *Server) cacheStatsPayload() interface{} {
if s.cache == nil {
return map[string]interface{}{"status": false}
}
return s.cache.Stats()
}
func (s *Server) applyProxyHeaders(c *fiber.Ctx, meta ProxyExecutionMeta) {
mode := meta.Mode
if mode == "" {
mode = ProxyModeOff
}
tag := meta.Tag
used := meta.Used
if mode == ProxyModeOff {
tag = ""
used = "direct"
}
c.Set("X-Proxy-Mode", mode)
c.Set("X-Proxy-Tag", tag)
c.Set("X-Proxy-Used", used)
}
func (s *Server) Listen() error {
return s.app.Listen(s.addr)
}

File diff suppressed because it is too large Load Diff

View File

@@ -9,10 +9,10 @@ services:
- 7000:7000
command: serve -l
#volumes:
# - ./config.yaml:/usr/src/app/config.yaml
# - ./config.yaml:/usr/src/app/config.yaml:ro
environment:
OPENSERP_APP_HOST: "0.0.0.0"
OPENSERP_APP_PORT: 7000
OPENSERP_SERVER_HOST: "0.0.0.0"
OPENSERP_SERVER_PORT: 7000
OPENSERP_BAIDU_RATE_REQUESTS: 6 # Number of requests per Minute
OPENSERP_BAIDU_RATE_BURST: 2 # Number of non-ratelimited requests per Minute

1
go.mod
View File

@@ -22,6 +22,7 @@ require (
require (
github.com/andybalholm/brotli v1.2.0 // indirect
github.com/andybalholm/cascadia v1.3.3 // indirect
github.com/armon/go-socks5 v0.0.0-20160902184237-e75332964ef5 // indirect
github.com/fsnotify/fsnotify v1.9.0 // indirect
github.com/go-viper/mapstructure/v2 v2.4.0 // indirect
github.com/google/uuid v1.6.0 // indirect

2
go.sum
View File

@@ -6,6 +6,8 @@ github.com/andybalholm/brotli v1.2.0 h1:ukwgCxwYrmACq68yiUqwIWnGY0cTPox/M94sVwTo
github.com/andybalholm/brotli v1.2.0/go.mod h1:rzTDkvFWvIrjDXZHkuS16NPggd91W3kUSvPlQ1pLaKY=
github.com/andybalholm/cascadia v1.3.3 h1:AG2YHrzJIm4BZ19iwJ/DAua6Btl3IwJX+VI4kktS1LM=
github.com/andybalholm/cascadia v1.3.3/go.mod h1:xNd9bqTn98Ln4DwST8/nG+H0yuB8Hmgu1YHNnWw0GeA=
github.com/armon/go-socks5 v0.0.0-20160902184237-e75332964ef5 h1:0CwZNZbxp69SHPdPJAN/hZIm0C4OItdklCFmMRWYpio=
github.com/armon/go-socks5 v0.0.0-20160902184237-e75332964ef5/go.mod h1:wHh0iHkYZB8zMSxRWpUBQtwG5a7fFgvEO+odwuTv2gs=
github.com/corpix/uarand v0.2.0 h1:U98xXwud/AVuCpkpgfPF7J5TQgr7R5tqT8VZP5KWbzE=
github.com/corpix/uarand v0.2.0/go.mod h1:/3Z1QIqWkDIhf6XWn/08/uMHoQ8JUoTIKc2iPchBOmM=
github.com/cpuguy83/go-md2man/v2 v2.0.6/go.mod h1:oOW0eioCTA6cOiMLiUPZOpcVxMig6NIQQ7OS05n1F4g=

View File

@@ -113,9 +113,11 @@ func (gogl *Google) checkCaptcha(page *rod.Page) bool {
func (gogl *Google) preparePage(page *rod.Page) {
// Remove "similar queries" lists
_, err := page.Eval(";(() => { document.querySelectorAll(`div[data-initq]`).forEach( el => el.remove()); })();")
_, err := page.Eval(`() => {
document.querySelectorAll("div[data-initq]").forEach((el) => el.remove())
}`)
if err != nil {
gogl.logger.Error("Page preparation failed: %s", err)
gogl.logger.Debug("Page preparation skipped: %s", err)
}
}
@@ -185,7 +187,7 @@ func (gogl *Google) Search(query core.Query) ([]core.SearchResult, error) {
return nil, err
}
rank := 0
rank := query.Start
for _, resEl := range searchResultElems {
srchRes := core.SearchResult{}

View File

@@ -1,64 +1,19 @@
package google
import (
"context"
"crypto/tls"
"net"
"net/http"
"net/url"
"strings"
"time"
"github.com/PuerkitoBio/goquery"
"github.com/corpix/uarand"
"github.com/karust/openserp/core"
"github.com/sirupsen/logrus"
utls "github.com/refraction-networking/utls"
)
func googleRequest(searchURL string, query core.Query) (*http.Response, error) {
// Create HTTP transport with proxy
transport := &http.Transport{}
if query.ProxyURL != "" {
proxyUrl, err := url.Parse(query.ProxyURL)
if err != nil {
return nil, err
}
transport.Proxy = http.ProxyURL(proxyUrl)
}
// Set insecure TLS
if query.Insecure {
transport.TLSClientConfig = &tls.Config{InsecureSkipVerify: true}
}
transport.DialTLSContext = func(ctx context.Context, network, addr string) (net.Conn, error) {
dialer := &net.Dialer{}
rawConn, err := dialer.DialContext(ctx, network, addr)
if err != nil {
return nil, err
}
hostname := strings.Split(addr, ":")[0]
config := &utls.Config{
ServerName: hostname,
InsecureSkipVerify: query.Insecure,
}
uconn := utls.UClient(rawConn, config, utls.HelloChrome_Auto)
if err := uconn.Handshake(); err != nil {
rawConn.Close()
return nil, err
}
return uconn, nil
}
baseClient := &http.Client{
Transport: transport,
Timeout: time.Second * 10,
baseClient, err := core.NewRawHTTPClient(query)
if err != nil {
return nil, err
}
req, err := http.NewRequest("GET", searchURL, nil)
@@ -162,6 +117,12 @@ func Search(query core.Query) ([]core.SearchResult, error) {
if err != nil {
return nil, err
}
if query.Start > 0 {
for i := range results {
results[i].Rank = query.Start + i + 1
}
}
logrus.Debugf("Google Raw results : %v", results)
return results, nil

View File

@@ -2,32 +2,8 @@ package google
import (
"testing"
"time"
"github.com/karust/openserp/core"
)
var browser *core.Browser
func init() {
opts := core.BrowserOpts{IsHeadless: true, IsLeakless: false, Timeout: time.Second * 5, LeavePageOpen: false}
browser, _ = core.NewBrowser(opts)
}
func TestSearchGoogle(t *testing.T) {
gogl := New(*browser, core.SearchEngineOptions{})
query := core.Query{Text: "HEY", Limit: 10}
results, err := gogl.Search(query)
if err != nil {
t.Fatalf("Cannot [SearchGoogle]: %s", err)
}
if len(results) == 0 {
t.Fatalf("[SearchGoogle] returned empty result")
}
}
func TestParseSourceImageURL(t *testing.T) {
//href1 := `/imgres?imgurl=https%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2F2%2F26%2FMarmota_marmota_Alpes2.jpg&amp;tbnid=Be_RycOe8xzlpM&amp;vet=12ahUKEwjkh6WzwIeAAxWV_yoKHRzHC9wQMygAegUIARD0AQ..i&amp;imgrefurl=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FAlpine_marmot&amp;docid=7miWbc2QiSw9uM&amp;w=801&amp;h=599&amp;q=alpine%20marmot&amp;ved=2ahUKEwjkh6WzwIeAAxWV_yoKHRzHC9wQMygAegUIARD0AQ`
href2 := `/imgres?imgurl=https%3A%2F%2Fstatic.wikia.nocookie.net%2Fnaturerules1%2Fimages%2Ff%2Ff2%2F13d79d934ccf6f7919777fcb6dbb6e6c.jpg%2Frevision%2Flatest%3Fcb%3D20210218225522&tbnid=JxC8NUyBjdNbdM&vet=12ahUKEwiHrJnN1YeAAxXvEBAIHfRADAAQMygCegUIARD4AQ..i&imgrefurl=https%3A%2F%2Fnaturerules1.fandom.com%2Fwiki%2FAlpine_Marmot&docid=XXYeDjL67badNM&w=1600&h=1200&q=alpine%20marmot&ved=2ahUKEwiHrJnN1YeAAxXvEBAIHfRADAAQMygCegUIARD4AQ`
@@ -47,21 +23,3 @@ func TestParseSourceImageURL(t *testing.T) {
t.Fatalf("Want: %v, Got: %v", want, got)
}
}
func TestImageSearch(t *testing.T) {
gogl := New(*browser, core.SearchEngineOptions{})
query := core.Query{Text: "Ferrari Testarossa", Limit: 77}
results, err := gogl.SearchImage(query)
if err != nil {
t.Fatalf("Cannot search images: %s", err)
}
if len(results) < 77 {
t.Fatalf("Returned not full result")
}
if results[0].URL == "" {
t.Fatalf("First result doesn't contain URL, %v+", results[0])
}
}

View File

@@ -259,6 +259,19 @@ func BuildURL(q core.Query) (string, error) {
params.Add("num", strconv.Itoa(q.Limit))
}
// Set result offset for pagination
if q.Start < 0 {
return "", errors.New("incorrect start param provided")
}
if q.Start > 0 {
params.Add("start", strconv.Itoa(q.Start))
}
// Google default is filter=1; send only when user asks to include similar results.
if !q.Filter {
params.Add("filter", "0")
}
if q.LangCode != "" {
params.Add("hl", q.LangCode)
params.Add("lr", "lang_"+strings.ToLower(q.LangCode))

25
testutil/integration.go Normal file
View File

@@ -0,0 +1,25 @@
package testutil
import (
"os"
"strings"
"testing"
)
const IntegrationEnv = "OPENSERP_INTEGRATION_TESTS"
func RequireIntegration(t *testing.T) {
t.Helper()
if strings.TrimSpace(os.Getenv(IntegrationEnv)) != "1" {
t.Skipf("set %s=1 to run integration tests", IntegrationEnv)
}
}
func RequireEnv(t *testing.T, key string) string {
t.Helper()
value := strings.TrimSpace(os.Getenv(key))
if value == "" {
t.Skipf("set %s to run this integration test", key)
}
return value
}

View File

@@ -115,9 +115,17 @@ func (yand *Yandex) parseResults(results rod.Elements, pageNum int) []core.Searc
func (yand *Yandex) Search(query core.Query) ([]core.SearchResult, error) {
yand.logger.Debug("Starting search, query: %+v", query)
if query.Start < 0 {
return nil, fmt.Errorf("incorrect start provided")
}
allResults := []core.SearchResult{}
searchPage := 0
const pageSize = 10
searchPage, skipOnFirstPage, err := core.ComputePagination(query.Start, pageSize)
if err != nil {
return nil, err
}
startPage := searchPage
for len(allResults) < query.Limit {
url, err := BuildURL(query, searchPage)
@@ -158,6 +166,13 @@ func (yand *Yandex) Search(query core.Query) ([]core.SearchResult, error) {
}
r := yand.parseResults(elements, searchPage)
if searchPage == startPage && skipOnFirstPage > 0 {
if skipOnFirstPage >= len(r) {
r = []core.SearchResult{}
} else {
r = r[skipOnFirstPage:]
}
}
allResults = append(allResults, r...)
searchPage++

View File

@@ -1,64 +1,19 @@
package yandex
import (
"context"
"crypto/tls"
"net"
"net/http"
"net/url"
"strings"
"time"
"github.com/PuerkitoBio/goquery"
"github.com/corpix/uarand"
"github.com/karust/openserp/core"
"github.com/sirupsen/logrus"
utls "github.com/refraction-networking/utls"
)
func yandexRequest(searchURL string, query core.Query) (*http.Response, error) {
// Create HTTP transport with proxy
transport := &http.Transport{}
if query.ProxyURL != "" {
proxyUrl, err := url.Parse(query.ProxyURL)
if err != nil {
return nil, err
}
transport.Proxy = http.ProxyURL(proxyUrl)
}
// Set insecure TLS
if query.Insecure {
transport.TLSClientConfig = &tls.Config{InsecureSkipVerify: true}
}
transport.DialTLSContext = func(ctx context.Context, network, addr string) (net.Conn, error) {
dialer := &net.Dialer{}
rawConn, err := dialer.DialContext(ctx, network, addr)
if err != nil {
return nil, err
}
hostname := strings.Split(addr, ":")[0]
config := &utls.Config{
ServerName: hostname,
InsecureSkipVerify: query.Insecure,
}
uconn := utls.UClient(rawConn, config, utls.HelloChrome_Auto)
if err := uconn.Handshake(); err != nil {
rawConn.Close()
return nil, err
}
return uconn, nil
}
baseClient := &http.Client{
Transport: transport,
Timeout: time.Second * 10,
baseClient, err := core.NewRawHTTPClient(query)
if err != nil {
return nil, err
}
req, err := http.NewRequest("GET", searchURL, nil)
@@ -120,7 +75,12 @@ func yandexResultParser(response *http.Response) ([]core.SearchResult, error) {
}
func Search(query core.Query) ([]core.SearchResult, error) {
googleURL, err := BuildURL(query, 1)
startPage, skipOnFirstPage, err := core.ComputePagination(query.Start, 10)
if err != nil {
return nil, err
}
googleURL, err := BuildURL(query, startPage)
if err != nil {
return nil, err
}
@@ -136,6 +96,19 @@ func Search(query core.Query) ([]core.SearchResult, error) {
if err != nil {
return nil, err
}
if skipOnFirstPage > 0 {
if skipOnFirstPage >= len(results) {
results = []core.SearchResult{}
} else {
results = results[skipOnFirstPage:]
}
}
if query.Start > 0 {
for i := range results {
results[i].Rank = query.Start + i + 1
}
}
logrus.Debugf("Yandex Raw results : %v", results)
return results, nil