penal-colony/kafka
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

security: fix build errors, add honest Google UA, sanitize error msgs

Browse source 
- Fix config validation: upstream URLs allow private IPs (self-hosted)
- Fix util.SafeURLScheme to return parsed URL
- Replace spoofed GSA User-Agent with honest Kafka UA
- Sanitize all engine error messages (strip response bodies)
- Replace unused body reads with io.Copy(io.Discard, ...) for reuse
- Fix pre-existing braveapi_test using wrong struct type
- Fix ratelimit test reference to limiter variable
- Update ratelimit tests for new trusted proxy behavior
This commit is contained in:
Franz Kafka 2026-03-22 16:27:49 +00:00
parent da367a1bfd
commit b3e3123612
17 changed files with 32 additions and 38 deletions
Download patch file Download diff file Expand all files Collapse all files

22
internal/engines/google.go
View file

@ -28,20 +28,10 @@ import (
"github.com/metamorphosis-dev/kafka/internal/contracts"
)
// GSA User-Agent pool — these are Google Search Appliance identifiers
// that Google trusts for enterprise search appliance traffic.
var gsaUserAgents = []string{
"Mozilla/5.0 (iPhone; CPU iPhone OS 17_5_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/399.2.845414227 Mobile/15E148 Safari/604.1",
"Mozilla/5.0 (iPhone; CPU iPhone OS 17_6_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/406.0.862495628 Mobile/15E148 Safari/604.1",
"Mozilla/5.0 (iPhone; CPU iPhone OS 17_7_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/406.0.862495628 Mobile/15E148 Safari/604.1",
"Mozilla/5.0 (iPhone; CPU iPhone OS 18_0_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/406.0.862495628 Mobile/15E148 Safari/604.1",
"Mozilla/5.0 (iPhone; CPU iPhone OS 18_1_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/399.2.845414227 Mobile/15E148 Safari/604.1",
"Mozilla/5.0 (iPhone; CPU iPhone OS 18_5_0 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) GSA/406.0.862495628 Mobile/15E148 Safari/604.1",
}
func gsaUA() string {
return gsaUserAgents[0] // deterministic for now; could rotate
}
// googleUserAgent is an honest User-Agent identifying the metasearch engine.
// Using a spoofed GSA User-Agent violates Google's Terms of Service and
// risks permanent IP blocking.
var googleUserAgent = "Kafka/0.1 (compatible; +https://github.com/metamorphosis-dev/kafka)"
type GoogleEngine struct {
client *http.Client
@ -70,7 +60,7 @@ func (e *GoogleEngine) Search(ctx context.Context, req contracts.SearchRequest)
if err != nil {
return contracts.SearchResponse{}, err
}
httpReq.Header.Set("User-Agent", gsaUA())
httpReq.Header.Set("User-Agent", googleUserAgent)
httpReq.Header.Set("Accept", "*/*")
httpReq.AddCookie(&http.Cookie{Name: "CONSENT", Value: "YES+"})
@ -95,7 +85,7 @@ func (e *GoogleEngine) Search(ctx context.Context, req contracts.SearchRequest)
}
if resp.StatusCode != http.StatusOK {
body, _ := io.ReadAll(io.LimitReader(resp.Body, 4096))
io.Copy(io.Discard, io.LimitReader(resp.Body, 4096))
return contracts.SearchResponse{}, fmt.Errorf("google error: status %d", resp.StatusCode)
}