Files
netbird/proxy/internal/middleware/builtin/llm_response_parser/gzip_test.go
Maycon Santos b416063bcc [management,proxy] Agent network: per-account LLM gateway (policy, metering, multi-provider) (#6555)
* [agent-network] Shared proto, OpenAPI schema, and generated types

* [agent-network] Management: store, manager, synthesizer, policy engine, provider catalog, HTTP/gRPC API

Adds the account-scoped agent-network module: provider/policy/budget CRUD and
store, the reverse-proxy service synthesizer, policy selection + limit
enforcement, the provider catalog (incl. Vertex AI and AWS Bedrock entries),
and the management HTTP + proxy gRPC surfaces.

* [management] Fix agent-network proxy-peer fan-out on affected-peer recompute

The affected-peers resolver loaded only persisted reverse-proxy services, but
agent-network services are synthesized on demand and never persisted. As a
result the embedded proxy peer was never folded into the affected set when a
client's group changed, so the proxy received no network-map update for a newly
authorised client and rejected its handshake until a full resync (restart).

loadProxyServices now merges the synthesized agent-network services (injected
via a registration hook to avoid an import cycle), so proxy peers learn newly
authorised clients immediately.

* [proxy] Reverse-proxy middleware framework, chain, and request plumbing

The per-target middleware chain (slots, dispatcher, mutation gate, metadata
merger), body capture, access-log terminal sink, and the proxy wiring that
builds + runs chains for synthesized agent-network services.

* [proxy] LLM parsers, pricing, and builtin middlewares (OpenAI, Anthropic, Vertex AI, AWS Bedrock)

Request/response parsers and SSE/event-stream metering, the embedded pricing
table, and the builtin middleware set: request parser, router, policy
limit-check/record, cost meter, guardrail, identity inject, response parser.
Includes the path-routed providers — Google Vertex AI (keyfile:: service-account
OAuth minting) and AWS Bedrock (bearer auth, invoke/converse/streaming, optional
/bedrock prefix) — plus the Models allowlist and unmeterable-publisher deny.

* [proxy] IPv6 in-place apply and TCP accept-loop hardening on netstack listeners

* [agent-network] End-to-end test suite, module docs, and deployment preset

* [agent-network] Fix codespell typos and exclude false positives

- labelgen word pool: vermillion -> vermilion, racoon -> raccoon.
- codespell ignore list: add flate (Go compress/flate package), recordin
  (a test-local identifier), and unparseable (a valid alternative spelling used
  consistently across identifiers + a metadata-value constant).

* [management] Set LastSeen on injected proxy peer in realstack test (MySQL strict-mode)

The injected embedded proxy peer had a PeerStatus with a zero LastSeen, which
serializes to '0000-00-00' and is rejected by MySQL in strict mode (SQLite
tolerates it). Set LastSeen to a valid time so SaveAccount succeeds on both
engines.

* [agent-network] Remove e2e shell-script suite from this branch

The end-to-end shell scripts under scripts/e2e/ are maintained in a separate
testing suite and are not part of this change set.

* [agent-network] Polish module docs: remove internal review scaffolding, fix links, verify diagrams

Strip PR-review framing, commit references, absolute paths, and stale internal
references from the agent-network module docs; fix broken relative links; verify
all diagrams against the current architecture. Remove the internal AI-reviewer
prompt file.

* [management] Refine session expiration handling to support 3-state encoding for SSO deadlines

* [agent-network] Relocate agentnetwork package to internals/modules

Move management/server/agentnetwork (and its catalog/, labelgen/, types/
subpackages) to management/internals/modules/agentnetwork, alongside the
reverse-proxy module, and rewrite all importers. Pure relocation: package names,
the synthesizer + affectedpeers registration hook, and store access (shared
store.Store) are unchanged, so no import cycle is introduced (affectedpeers
still depends only on the agentnetwork/types leaf).

* [agent-network] Co-locate HTTP handlers in the module (RegisterEndpoints)

Move the agent-network HTTP handlers from server/http/handlers/agentnetwork into
the module at internals/modules/agentnetwork/handlers (package handlers) and
rename the entrypoint AddEndpoints -> RegisterEndpoints, matching the
reverse-proxy module convention. Wiring in http/handler.go updated accordingly.
2026-06-27 13:41:00 +02:00

134 lines
4.4 KiB
Go

package llm_response_parser
import (
"bytes"
"compress/flate"
"compress/gzip"
"compress/zlib"
"context"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"github.com/netbirdio/netbird/proxy/internal/middleware"
)
// gzipBytes returns data gzip-compressed — the wire shape Anthropic
// returns when the client (Claude Code) negotiated Accept-Encoding: gzip.
func gzipBytes(t *testing.T, data []byte) []byte {
t.Helper()
var buf bytes.Buffer
w := gzip.NewWriter(&buf)
_, err := w.Write(data)
require.NoError(t, err, "gzip write must succeed")
require.NoError(t, w.Close(), "gzip close must succeed")
return buf.Bytes()
}
// TestInvoke_AnthropicStreaming_Gzip is the regression guard for the live
// bug: Claude Code negotiates gzip, Anthropic gzips the SSE stream, the
// proxy captures the compressed bytes, and the parser must decompress
// before accumulating — otherwise token usage is silently dropped and
// cost_meter skips with missing_tokens.
func TestInvoke_AnthropicStreaming_Gzip(t *testing.T) {
m := newTestMiddleware(t)
body := gzipBytes(t, loadFixture(t, "anthropic_stream.txt"))
in := &middleware.Input{
Slot: middleware.SlotOnResponse,
Status: 200,
RespHeaders: []middleware.KV{
{Key: "Content-Type", Value: "text/event-stream; charset=utf-8"},
{Key: "Content-Encoding", Value: "gzip"},
},
RespBody: body,
Metadata: []middleware.KV{
{Key: middleware.KeyLLMProvider, Value: "anthropic"},
{Key: middleware.KeyLLMModel, Value: "claude-opus-4-8"},
},
}
out, err := m.Invoke(context.Background(), in)
require.NoError(t, err, "Invoke must not error on a gzip-encoded streaming body")
in123, ok := metaValue(out.Metadata, middleware.KeyLLMInputTokens)
require.True(t, ok, "input tokens must be emitted from a gzip SSE stream")
assert.Equal(t, "123", in123, "input tokens must survive gzip decompression")
outTok, _ := metaValue(out.Metadata, middleware.KeyLLMOutputTokens)
assert.Equal(t, "45", outTok, "output tokens must survive gzip decompression")
totTok, _ := metaValue(out.Metadata, middleware.KeyLLMTotalTokens)
assert.Equal(t, "168", totTok, "total tokens must survive gzip decompression")
}
// TestInvoke_AnthropicBuffered_Gzip covers the non-streaming JSON path
// under gzip — the same decode must happen before ParseResponse.
func TestInvoke_AnthropicBuffered_Gzip(t *testing.T) {
m := newTestMiddleware(t)
body := gzipBytes(t, loadFixture(t, "anthropic_messages.json"))
in := &middleware.Input{
Slot: middleware.SlotOnResponse,
Status: 200,
RespHeaders: []middleware.KV{
{Key: "Content-Type", Value: "application/json"},
{Key: "Content-Encoding", Value: "gzip"},
},
RespBody: body,
Metadata: []middleware.KV{
{Key: middleware.KeyLLMProvider, Value: "anthropic"},
{Key: middleware.KeyLLMModel, Value: "claude-opus-4-8"},
},
}
out, err := m.Invoke(context.Background(), in)
require.NoError(t, err, "Invoke must not error on a gzip-encoded buffered body")
_, ok := metaValue(out.Metadata, middleware.KeyLLMInputTokens)
require.True(t, ok, "input tokens must be emitted from a gzip JSON body")
}
// TestDecodeResponseBody covers the encoding matrix directly.
func TestDecodeResponseBody(t *testing.T) {
plain := []byte(`{"hello":"world"}`)
t.Run("identity passthrough", func(t *testing.T) {
assert.Equal(t, plain, decodeResponseBody(plain, ""))
assert.Equal(t, plain, decodeResponseBody(plain, "identity"))
})
t.Run("gzip", func(t *testing.T) {
assert.Equal(t, plain, decodeResponseBody(gzipBytes(t, plain), "gzip"))
})
t.Run("gzip with multi-coding header takes outermost", func(t *testing.T) {
assert.Equal(t, plain, decodeResponseBody(gzipBytes(t, plain), "identity, gzip"))
})
t.Run("deflate zlib-wrapped", func(t *testing.T) {
var buf bytes.Buffer
zw := zlib.NewWriter(&buf)
_, _ = zw.Write(plain)
_ = zw.Close()
assert.Equal(t, plain, decodeResponseBody(buf.Bytes(), "deflate"))
})
t.Run("deflate raw flate fallback", func(t *testing.T) {
var buf bytes.Buffer
fw, _ := flate.NewWriter(&buf, flate.DefaultCompression)
_, _ = fw.Write(plain)
_ = fw.Close()
assert.Equal(t, plain, decodeResponseBody(buf.Bytes(), "deflate"))
})
t.Run("gzip header but not actually gzip falls back to raw", func(t *testing.T) {
assert.Equal(t, plain, decodeResponseBody(plain, "gzip"))
})
t.Run("unknown encoding (br) returns raw", func(t *testing.T) {
assert.Equal(t, plain, decodeResponseBody(plain, "br"))
})
}