feat(container): Implement image pulling with authentication and refactor runtime detection

- Add PullImageWithAuth method to DockerRuntime for pulling images with authentication.
- Refactor runtime detection to use DockerRuntime for both Docker and Podman environments.
- Update container manager to utilize authenticated image pulling based on registry configuration.
- Enhance logging for image pulling and container network info retrieval.
- Remove deprecated PodmanRuntime implementation to streamline container management.

commit: 45beea7dda93587970b8c94f6ba59c7ad9a720d0
author: bamen · 16mo ago · 2025-06-07 12:38 UTC
parent: 0c26242
signature: unsigned

12 files · +590 -897

M cmd/start.go +8 -14


@@ -8,17 +8,17 @@ import (
8	8		"syscall"
9	9		"time"
10	10
11		-	"github.com/fsnotify/fsnotify"
12		-	"github.com/google/uuid"
13		-	"github.com/rs/zerolog"
14		-	"github.com/rs/zerolog/log"
15		-	"github.com/spf13/cobra"
16		-	"github.com/spf13/viper"
17	11		"gordon/internal/config"
18	12		"gordon/internal/container"
19	13		"gordon/internal/events"
20	14		"gordon/internal/proxy"
21	15		"gordon/internal/registry"
	16	+
	17	+	"github.com/fsnotify/fsnotify"
	18	+	"github.com/rs/zerolog"
	19	+	"github.com/rs/zerolog/log"
	20	+	"github.com/spf13/cobra"
	21	+	"github.com/spf13/viper"
22	22		)
23	23
24	24		var startCmd = &cobra.Command{

@@ -91,13 +91,7 @@ func runStart(cmd *cobra.Command, args []string) {
91	91		}
92	92
93	93		// Publish config reload event
94		-	configReloadEvent := events.Event{
95		-	ID: uuid.New().String(),
96		-	Type: events.ConfigReload,
97		-	Timestamp: time.Now(),
98		-	}
99		-
100		-	if err := eventBus.Publish(configReloadEvent); err != nil {
	94	+	if err := eventBus.Publish(events.ConfigReload, nil); err != nil {
101	95		log.Error().Err(err).Msg("Failed to publish config reload event")
102	96		} else {
103	97		log.Info().Msg("Configuration reloaded successfully")

@@ -161,4 +155,4 @@ func runStart(cmd *cobra.Command, args []string) {
161	155		case <-time.After(10 * time.Second):
162	156		log.Warn().Msg("Force shutdown after timeout")
163	157		}
164		-	}⏎̸
	158	+	}

 	"syscall"
 	"time"
-	"github.com/fsnotify/fsnotify"
-	"github.com/google/uuid"
-	"github.com/rs/zerolog"
-	"github.com/rs/zerolog/log"
-	"github.com/spf13/cobra"
-	"github.com/spf13/viper"
 	"gordon/internal/config"
 	"gordon/internal/container"
 	"gordon/internal/events"
 	"gordon/internal/proxy"
 	"gordon/internal/registry"
+	"github.com/fsnotify/fsnotify"
+	"github.com/rs/zerolog"
+	"github.com/rs/zerolog/log"
+	"github.com/spf13/cobra"
+	"github.com/spf13/viper"
 )
 var startCmd = &cobra.Command{
 		}
 		// Publish config reload event
-		configReloadEvent := events.Event{
+		if err := eventBus.Publish(events.ConfigReload, nil); err != nil {
-			ID:        uuid.New().String(),
-			Type:      events.ConfigReload,
-			Timestamp: time.Now(),
-		}
-		if err := eventBus.Publish(configReloadEvent); err != nil {
 			log.Error().Err(err).Msg("Failed to publish config reload event")
 		} else {
 			log.Info().Msg("Configuration reloaded successfully")
 	case <-time.After(10 * time.Second):
 		log.Warn().Msg("Force shutdown after timeout")
 	}
-}⏎̸
+}

M internal/container/docker.go +48 -21


@@ -2,6 +2,8 @@ package container
2	2
3	3		import (
4	4		"context"
	5	+	"encoding/base64"
	6	+	"encoding/json"
5	7		"fmt"
6	8		"io"
7	9		"strconv"

@@ -9,6 +11,7 @@ import (
9	11
10	12		"github.com/docker/docker/api/types/container"
11	13		"github.com/docker/docker/api/types/image"
	14	+	"github.com/docker/docker/api/types/registry"
12	15		"github.com/docker/docker/client"
13	16		"github.com/docker/go-connections/nat"
14	17		"github.com/rs/zerolog/log"

@@ -230,6 +233,42 @@ func (d *DockerRuntime) PullImage(ctx context.Context, imageRef string) error {
230	233		return nil
231	234		}
232	235
	236	+	// PullImageWithAuth pulls an image with authentication
	237	+	func (d *DockerRuntime) PullImageWithAuth(ctx context.Context, imageRef, username, password string) error {
	238	+	log.Info().Str("image", imageRef).Str("username", username).Msg("Pulling image with authentication")
	239	+
	240	+	// Create authentication configuration
	241	+	authConfig := registry.AuthConfig{
	242	+	Username: username,
	243	+	Password: password,
	244	+	}
	245	+
	246	+	// Encode authentication to base64 JSON
	247	+	authConfigBytes, err := json.Marshal(authConfig)
	248	+	if err != nil {
	249	+	return fmt.Errorf("failed to marshal auth config: %w", err)
	250	+	}
	251	+	authStr := base64.URLEncoding.EncodeToString(authConfigBytes)
	252	+
	253	+	// Pull with authentication
	254	+	reader, err := d.client.ImagePull(ctx, imageRef, image.PullOptions{
	255	+	RegistryAuth: authStr,
	256	+	})
	257	+	if err != nil {
	258	+	return fmt.Errorf("failed to pull image %s with auth: %w", imageRef, err)
	259	+	}
	260	+	defer reader.Close()
	261	+
	262	+	// Read the response to completion (this is required for the pull to complete)
	263	+	_, err = io.Copy(io.Discard, reader)
	264	+	if err != nil {
	265	+	return fmt.Errorf("failed to read pull response for image %s: %w", imageRef, err)
	266	+	}
	267	+
	268	+	log.Info().Str("image", imageRef).Msg("Image pulled successfully with authentication")
	269	+	return nil
	270	+	}
	271	+
233	272		// RemoveImage removes an image
234	273		func (d *DockerRuntime) RemoveImage(ctx context.Context, imageRef string, force bool) error {
235	274		_, err := d.client.ImageRemove(ctx, imageRef, image.RemoveOptions{Force: force})

@@ -330,10 +369,8 @@ func (d *DockerRuntime) GetImageExposedPorts(ctx context.Context, imageRef strin
330	369		}
331	370		}
332	371
333		-	// If no exposed ports found, return common web ports as fallback
334	372		if len(ports) == 0 {
335		-	log.Warn().Str("image", imageRef).Msg("No EXPOSE directives found in image, using common web ports")
336		-	return []int{80, 8080, 3000}, nil
	373	+	return nil, fmt.Errorf("no EXPOSE directives found in image %s - please add EXPOSE directive to Dockerfile", imageRef)
337	374		}
338	375
339	376		log.Info().Str("image", imageRef).Ints("exposed_ports", ports).Msg("Found exposed ports in image")

@@ -384,7 +421,7 @@ func (d *DockerRuntime) GetContainerNetworkInfo(ctx context.Context, containerID
384	421		return "", 0, fmt.Errorf("no IP address found for container %s", containerID)
385	422		}
386	423
387		-	// Get exposed ports and select the best one (Traefik-style logic)
	424	+	// Get exposed ports from container
388	425		var exposedPorts []int
389	426		if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
390	427		for portSpec := range resp.NetworkSettings.Ports {

@@ -397,29 +434,19 @@ func (d *DockerRuntime) GetContainerNetworkInfo(ctx context.Context, containerID
397	434		}
398	435
399	436		if len(exposedPorts) == 0 {
400		-	return "", 0, fmt.Errorf("no exposed ports found for container %s", containerID)
401		-	}
402		-
403		-	// Sort ports and select the lowest one (Traefik strategy)
404		-	var selectedPort int
405		-	if len(exposedPorts) == 1 {
406		-	selectedPort = exposedPorts[0]
407		-	} else {
408		-	// Find the lowest port number
409		-	selectedPort = exposedPorts[0]
410		-	for _, port := range exposedPorts {
411		-	if port < selectedPort {
412		-	selectedPort = port
413		-	}
414		-	}
	437	+	return "", 0, fmt.Errorf("no EXPOSE directives found for container %s - please add EXPOSE directive to Dockerfile", containerID)
415	438		}
416	439
	440	+	// Use the first exposed port
	441	+	selectedPort := exposedPorts[0]
	442	+
417	443		log.Info().
418	444		Str("container_id", containerID).
419	445		Str("ip", containerIP).
420	446		Int("selected_port", selectedPort).
421		-	Ints("available_ports", exposedPorts).
	447	+	Ints("exposed_ports", exposedPorts).
422	448		Msg("Container network info detected")
423	449
424	450		return containerIP, selectedPort, nil
425		-	}⏎̸
	451	+	}
	452	+

 import (
 	"context"
+	"encoding/base64"
+	"encoding/json"
 	"fmt"
 	"io"
 	"strconv"
 	"github.com/docker/docker/api/types/container"
 	"github.com/docker/docker/api/types/image"
+	"github.com/docker/docker/api/types/registry"
 	"github.com/docker/docker/client"
 	"github.com/docker/go-connections/nat"
 	"github.com/rs/zerolog/log"
 	return nil
 }
+// PullImageWithAuth pulls an image with authentication
+func (d *DockerRuntime) PullImageWithAuth(ctx context.Context, imageRef, username, password string) error {
+	log.Info().Str("image", imageRef).Str("username", username).Msg("Pulling image with authentication")
+	// Create authentication configuration
+	authConfig := registry.AuthConfig{
+		Username: username,
+		Password: password,
+	}
+	// Encode authentication to base64 JSON
+	authConfigBytes, err := json.Marshal(authConfig)
+	if err != nil {
+		return fmt.Errorf("failed to marshal auth config: %w", err)
+	}
+	authStr := base64.URLEncoding.EncodeToString(authConfigBytes)
+	// Pull with authentication
+	reader, err := d.client.ImagePull(ctx, imageRef, image.PullOptions{
+		RegistryAuth: authStr,
+	})
+	if err != nil {
+		return fmt.Errorf("failed to pull image %s with auth: %w", imageRef, err)
+	}
+	defer reader.Close()
+	// Read the response to completion (this is required for the pull to complete)
+	_, err = io.Copy(io.Discard, reader)
+	if err != nil {
+		return fmt.Errorf("failed to read pull response for image %s: %w", imageRef, err)
+	}
+	log.Info().Str("image", imageRef).Msg("Image pulled successfully with authentication")
+	return nil
+}
 // RemoveImage removes an image
 func (d *DockerRuntime) RemoveImage(ctx context.Context, imageRef string, force bool) error {
 	_, err := d.client.ImageRemove(ctx, imageRef, image.RemoveOptions{Force: force})
 		}
 	}
-	// If no exposed ports found, return common web ports as fallback
 	if len(ports) == 0 {
-		log.Warn().Str("image", imageRef).Msg("No EXPOSE directives found in image, using common web ports")
+		return nil, fmt.Errorf("no EXPOSE directives found in image %s - please add EXPOSE directive to Dockerfile", imageRef)
-		return []int{80, 8080, 3000}, nil
 	}
 	log.Info().Str("image", imageRef).Ints("exposed_ports", ports).Msg("Found exposed ports in image")
 		return "", 0, fmt.Errorf("no IP address found for container %s", containerID)
 	}
-	// Get exposed ports and select the best one (Traefik-style logic)
+	// Get exposed ports from container
 	var exposedPorts []int
 	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
 		for portSpec := range resp.NetworkSettings.Ports {
 	}
 	if len(exposedPorts) == 0 {
-		return "", 0, fmt.Errorf("no exposed ports found for container %s", containerID)
+		return "", 0, fmt.Errorf("no EXPOSE directives found for container %s - please add EXPOSE directive to Dockerfile", containerID)
-	}
-	// Sort ports and select the lowest one (Traefik strategy)
-	var selectedPort int
-	if len(exposedPorts) == 1 {
-		selectedPort = exposedPorts[0]
-	} else {
-		// Find the lowest port number
-		selectedPort = exposedPorts[0]
-		for _, port := range exposedPorts {
-			if port < selectedPort {
-				selectedPort = port
-			}
-		}
 	}
+	// Use the first exposed port
+	selectedPort := exposedPorts[0]
 	log.Info().
 		Str("container_id", containerID).
 		Str("ip", containerIP).
 		Int("selected_port", selectedPort).
-		Ints("available_ports", exposedPorts).
+		Ints("exposed_ports", exposedPorts).
 		Msg("Container network info detected")
 	return containerIP, selectedPort, nil
-}⏎̸
+}

M internal/container/factory.go +5 -9


@@ -148,14 +148,10 @@ func tryCreateRuntime(socketPath string) (runtime.Runtime, *RuntimeInfo, error)
148	148		}
149	149
150	150		// Determine runtime type based on socket path and server info
151		-	runtimeType := detectRuntimeType(socketPath, ping.APIVersion)
	151	+	runtimeType := detectRuntimeType(socketPath)
152	152
153		-	var rt runtime.Runtime
154		-	if runtimeType == RuntimeDocker {
155		-	rt = &DockerRuntime{client: cli}
156		-	} else {
157		-	rt = &PodmanRuntime{client: cli}
158		-	}
	153	+	// Since Podman uses Docker-compatible API, we can use DockerRuntime for both
	154	+	rt := &DockerRuntime{client: cli}
159	155
160	156		info := &RuntimeInfo{
161	157		Type: runtimeType,

@@ -167,7 +163,7 @@ func tryCreateRuntime(socketPath string) (runtime.Runtime, *RuntimeInfo, error)
167	163		}
168	164
169	165		// detectRuntimeType determines if the socket is Docker or Podman
170		-	func detectRuntimeType(socketPath, apiVersion string) RuntimeType {
	166	+	func detectRuntimeType(socketPath string) RuntimeType {
171	167		// Check socket path patterns
172	168		if strings.Contains(socketPath, "podman") {
173	169		if strings.Contains(socketPath, "/run/user/") \|\| strings.Contains(socketPath, "$XDG_RUNTIME_DIR") {

@@ -250,7 +246,7 @@ func createPodmanRuntimeWithSocket(socketPath string) (runtime.Runtime, error) {
250	246		return nil, fmt.Errorf("Podman ping failed: %w", err)
251	247		}
252	248
253		-	return &PodmanRuntime{client: cli}, nil
	249	+	return &DockerRuntime{client: cli}, nil
254	250		}
255	251
256	252		// GetAvailableRuntimes returns information about all available runtimes

 	}
 	// Determine runtime type based on socket path and server info
-	runtimeType := detectRuntimeType(socketPath, ping.APIVersion)
+	runtimeType := detectRuntimeType(socketPath)
-	var rt runtime.Runtime
+	// Since Podman uses Docker-compatible API, we can use DockerRuntime for both
-	if runtimeType == RuntimeDocker {
+	rt := &DockerRuntime{client: cli}
-		rt = &DockerRuntime{client: cli}
-	} else {
-		rt = &PodmanRuntime{client: cli}
-	}
 	info := &RuntimeInfo{
 		Type:       runtimeType,
 }
 // detectRuntimeType determines if the socket is Docker or Podman
-func detectRuntimeType(socketPath, apiVersion string) RuntimeType {
+func detectRuntimeType(socketPath string) RuntimeType {
 	// Check socket path patterns
 	if strings.Contains(socketPath, "podman") {
 		if strings.Contains(socketPath, "/run/user/") || strings.Contains(socketPath, "$XDG_RUNTIME_DIR") {
 		return nil, fmt.Errorf("Podman ping failed: %w", err)
 	}
-	return &PodmanRuntime{client: cli}, nil
+	return &DockerRuntime{client: cli}, nil
 }
 // GetAvailableRuntimes returns information about all available runtimes

M internal/container/manager.go +48 -26


@@ -3,6 +3,7 @@ package container
3	3		import (
4	4		"context"
5	5		"fmt"
	6	+	"strings"
6	7		"sync"
7	8
8	9		"github.com/rs/zerolog/log"

@@ -13,10 +14,10 @@ import (
13	14
14	15		// Manager handles container lifecycle and management
15	16		type Manager struct {
16		-	runtime runtime.Runtime
17		-	config *config.Config
	17	+	runtime runtime.Runtime
	18	+	config *config.Config
18	19		containers map[string]*runtime.Container // map[domain] -> container
19		-	mu sync.RWMutex
	20	+	mu sync.RWMutex
20	21		}
21	22
22	23		// NewManager creates a new container manager

@@ -55,42 +56,63 @@ func (m Manager) DeployContainer(ctx context.Context, route config.Route) (run
55	56		// Check if container already exists for this domain
56	57		if existing, exists := m.containers[route.Domain]; exists {
57	58		log.Info().Str("domain", route.Domain).Str("container", existing.ID).Msg("Container already exists, restarting")
58		-
	59	+
59	60		// Stop and remove existing container
60	61		if err := m.runtime.StopContainer(ctx, existing.ID); err != nil {
61	62		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to stop existing container")
62	63		}
63		-
	64	+
64	65		if err := m.runtime.RemoveContainer(ctx, existing.ID, true); err != nil {
65	66		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to remove existing container")
66	67		}
67		-
	68	+
68	69		delete(m.containers, route.Domain)
69	70		}
70	71
	72	+	// Construct the full image reference
	73	+	imageRef := route.Image
	74	+
	75	+	// If registry auth is enabled and image doesn't already contain a registry domain, prepend it
	76	+	if m.config.RegistryAuth.Enabled && m.config.Server.RegistryDomain != "" {
	77	+	// Check if image already contains a registry domain (has a '.' and doesn't start with official Docker Hub libraries)
	78	+	if !strings.Contains(strings.Split(imageRef, ":")[0], ".") {
	79	+	imageRef = fmt.Sprintf("%s/%s", m.config.Server.RegistryDomain, route.Image)
	80	+	}
	81	+	}
	82	+
71	83		// Pull the image first
72		-	log.Info().Str("image", route.Image).Msg("Pulling image")
73		-	if err := m.runtime.PullImage(ctx, route.Image); err != nil {
74		-	return nil, fmt.Errorf("failed to pull image %s: %w", route.Image, err)
	84	+	log.Info().Str("image", imageRef).Msg("Pulling image")
	85	+
	86	+	// Use authenticated pull if registry auth is configured
	87	+	if m.config.RegistryAuth.Enabled {
	88	+	err := m.runtime.PullImageWithAuth(ctx, imageRef, m.config.RegistryAuth.Username, m.config.RegistryAuth.Password)
	89	+	if err != nil {
	90	+	return nil, fmt.Errorf("failed to pull image %s with auth: %w", imageRef, err)
	91	+	}
	92	+	} else {
	93	+	err := m.runtime.PullImage(ctx, imageRef)
	94	+	if err != nil {
	95	+	return nil, fmt.Errorf("failed to pull image %s: %w", imageRef, err)
	96	+	}
75	97		}
76	98
77	99		// Get exposed ports from the image
78		-	exposedPorts, err := m.runtime.GetImageExposedPorts(ctx, route.Image)
	100	+	exposedPorts, err := m.runtime.GetImageExposedPorts(ctx, imageRef)
79	101		if err != nil {
80		-	log.Warn().Err(err).Str("image", route.Image).Msg("Failed to get exposed ports from image, using defaults")
	102	+	log.Warn().Err(err).Str("image", imageRef).Msg("Failed to get exposed ports from image, using defaults")
81	103		exposedPorts = []int{80, 8080, 3000} // Fallback to common web server ports
82	104		}
83	105
84	106		// Create container configuration
85	107		containerConfig := &runtime.ContainerConfig{
86		-	Image: route.Image,
87		-	Name: fmt.Sprintf("gordon-%s", route.Domain),
88		-	Ports: exposedPorts,
	108	+	Image: imageRef,
	109	+	Name: fmt.Sprintf("gordon-%s", route.Domain),
	110	+	Ports: exposedPorts,
89	111		Labels: map[string]string{
90		-	"gordon.domain": route.Domain,
91		-	"gordon.image": route.Image,
	112	+	"gordon.domain": route.Domain,
	113	+	"gordon.image": route.Image,
92	114		"gordon.managed": "true",
93		-	"gordon.route": route.Domain,
	115	+	"gordon.route": route.Domain,
94	116		},
95	117		AutoRemove: false, // Keep containers for inspection
96	118		}

@@ -131,7 +153,7 @@ func (m Manager) DeployContainer(ctx context.Context, route config.Route) (run
131	153		func (m Manager) GetContainer(domain string) (runtime.Container, bool) {
132	154		m.mu.RLock()
133	155		defer m.mu.RUnlock()
134		-
	156	+
135	157		container, exists := m.containers[domain]
136	158		return container, exists
137	159		}

@@ -141,7 +163,7 @@ func (m *Manager) GetContainerPort(ctx context.Context, domain string, internalP
141	163		m.mu.RLock()
142	164		container, exists := m.containers[domain]
143	165		m.mu.RUnlock()
144		-
	166	+
145	167		if !exists {
146	168		return 0, fmt.Errorf("no container found for domain %s", domain)
147	169		}

@@ -153,7 +175,7 @@ func (m *Manager) GetContainerPort(ctx context.Context, domain string, internalP
153	175		func (m Manager) ListContainers() map[string]runtime.Container {
154	176		m.mu.RLock()
155	177		defer m.mu.RUnlock()
156		-
	178	+
157	179		// Return a copy of the map to avoid race conditions
158	180		result := make(map[string]*runtime.Container)
159	181		for domain, container := range m.containers {

@@ -166,7 +188,7 @@ func (m Manager) ListContainers() map[string]runtime.Container {
166	188		func (m *Manager) FindContainerByDomain(domain string) (string, bool) {
167	189		m.mu.RLock()
168	190		defer m.mu.RUnlock()
169		-
	191	+
170	192		if container, exists := m.containers[domain]; exists {
171	193		return container.ID, true
172	194		}

@@ -177,7 +199,7 @@ func (m *Manager) FindContainerByDomain(domain string) (string, bool) {
177	199		func (m *Manager) FindDomainByContainerID(containerID string) (string, bool) {
178	200		m.mu.RLock()
179	201		defer m.mu.RUnlock()
180		-
	202	+
181	203		for domain, container := range m.containers {
182	204		if container.ID == containerID {
183	205		return domain, true

@@ -201,7 +223,7 @@ func (m *Manager) StopContainerByDomain(ctx context.Context, domain string) erro
201	223		m.mu.RLock()
202	224		container, exists := m.containers[domain]
203	225		m.mu.RUnlock()
204		-
	226	+
205	227		if !exists {
206	228		return fmt.Errorf("no container found for domain %s", domain)
207	229		}

@@ -225,7 +247,7 @@ func (m *Manager) RemoveContainer(ctx context.Context, containerID string, force
225	247		break
226	248		}
227	249		}
228		-
	250	+
229	251		return nil
230	252		}
231	253

@@ -234,7 +256,7 @@ func (m *Manager) RemoveContainerByDomain(ctx context.Context, domain string, fo
234	256		m.mu.RLock()
235	257		container, exists := m.containers[domain]
236	258		m.mu.RUnlock()
237		-
	259	+
238	260		if !exists {
239	261		return fmt.Errorf("no container found for domain %s", domain)
240	262		}

@@ -292,4 +314,4 @@ func (m *Manager) HealthCheck(ctx context.Context) map[string]bool {
292	314		// Runtime returns the underlying runtime interface
293	315		func (m *Manager) Runtime() runtime.Runtime {
294	316		return m.runtime
295		-	}⏎̸
	317	+	}


@@ -3,6 +3,7 @@ package container
3		import (	3		import (
4		"context"	4		"context"
5		"fmt"	5		"fmt"
			6	+	"strings"
6		"sync"	7		"sync"
7			8
8		"github.com/rs/zerolog/log"	9		"github.com/rs/zerolog/log"

@@ -13,10 +14,10 @@ import (
13			14
14		// Manager handles container lifecycle and management	15		// Manager handles container lifecycle and management
15		type Manager struct {	16		type Manager struct {
16	-	runtime runtime.Runtime	17	+	runtime runtime.Runtime
17	-	config *config.Config	18	+	config *config.Config
18		containers map[string]*runtime.Container // map[domain] -> container	19		containers map[string]*runtime.Container // map[domain] -> container
19	-	mu sync.RWMutex	20	+	mu sync.RWMutex
20		}	21		}
21			22
22		// NewManager creates a new container manager	23		// NewManager creates a new container manager

@@ -55,42 +56,63 @@ func (m Manager) DeployContainer(ctx context.Context, route config.Route) (run
55		// Check if container already exists for this domain	56		// Check if container already exists for this domain
56		if existing, exists := m.containers[route.Domain]; exists {	57		if existing, exists := m.containers[route.Domain]; exists {
57		log.Info().Str("domain", route.Domain).Str("container", existing.ID).Msg("Container already exists, restarting")	58		log.Info().Str("domain", route.Domain).Str("container", existing.ID).Msg("Container already exists, restarting")
58	-		59	+
59		// Stop and remove existing container	60		// Stop and remove existing container
60		if err := m.runtime.StopContainer(ctx, existing.ID); err != nil {	61		if err := m.runtime.StopContainer(ctx, existing.ID); err != nil {
61		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to stop existing container")	62		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to stop existing container")
62		}	63		}
63	-		64	+
64		if err := m.runtime.RemoveContainer(ctx, existing.ID, true); err != nil {	65		if err := m.runtime.RemoveContainer(ctx, existing.ID, true); err != nil {
65		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to remove existing container")	66		log.Warn().Err(err).Str("container", existing.ID).Msg("Failed to remove existing container")
66		}	67		}
67	-		68	+
68		delete(m.containers, route.Domain)	69		delete(m.containers, route.Domain)
69		}	70		}
70			71
			72	+	// Construct the full image reference
			73	+	imageRef := route.Image
			74	+
			75	+	// If registry auth is enabled and image doesn't already contain a registry domain, prepend it
			76	+	if m.config.RegistryAuth.Enabled && m.config.Server.RegistryDomain != "" {
			77	+	// Check if image already contains a registry domain (has a '.' and doesn't start with official Docker Hub libraries)
			78	+	if !strings.Contains(strings.Split(imageRef, ":")[0], ".") {
			79	+	imageRef = fmt.Sprintf("%s/%s", m.config.Server.RegistryDomain, route.Image)
			80	+	}
			81	+	}
			82	+
71		// Pull the image first	83		// Pull the image first
72	-	log.Info().Str("image", route.Image).Msg("Pulling image")	84	+	log.Info().Str("image", imageRef).Msg("Pulling image")
73	-	if err := m.runtime.PullImage(ctx, route.Image); err != nil {	85	+
74	-	return nil, fmt.Errorf("failed to pull image %s: %w", route.Image, err)	86	+	// Use authenticated pull if registry auth is configured
			87	+	if m.config.RegistryAuth.Enabled {
			88	+	err := m.runtime.PullImageWithAuth(ctx, imageRef, m.config.RegistryAuth.Username, m.config.RegistryAuth.Password)
			89	+	if err != nil {
			90	+	return nil, fmt.Errorf("failed to pull image %s with auth: %w", imageRef, err)
			91	+	}
			92	+	} else {
			93	+	err := m.runtime.PullImage(ctx, imageRef)
			94	+	if err != nil {
			95	+	return nil, fmt.Errorf("failed to pull image %s: %w", imageRef, err)
			96	+	}
75		}	97		}
76			98
77		// Get exposed ports from the image	99		// Get exposed ports from the image
78	-	exposedPorts, err := m.runtime.GetImageExposedPorts(ctx, route.Image)	100	+	exposedPorts, err := m.runtime.GetImageExposedPorts(ctx, imageRef)
79		if err != nil {	101		if err != nil {
80	-	log.Warn().Err(err).Str("image", route.Image).Msg("Failed to get exposed ports from image, using defaults")	102	+	log.Warn().Err(err).Str("image", imageRef).Msg("Failed to get exposed ports from image, using defaults")
81		exposedPorts = []int{80, 8080, 3000} // Fallback to common web server ports	103		exposedPorts = []int{80, 8080, 3000} // Fallback to common web server ports
82		}	104		}
83			105
84		// Create container configuration	106		// Create container configuration
85		containerConfig := &runtime.ContainerConfig{	107		containerConfig := &runtime.ContainerConfig{
86	-	Image: route.Image,	108	+	Image: imageRef,
87	-	Name: fmt.Sprintf("gordon-%s", route.Domain),	109	+	Name: fmt.Sprintf("gordon-%s", route.Domain),
88	-	Ports: exposedPorts,	110	+	Ports: exposedPorts,
89		Labels: map[string]string{	111		Labels: map[string]string{
90	-	"gordon.domain": route.Domain,	112	+	"gordon.domain": route.Domain,
91	-	"gordon.image": route.Image,	113	+	"gordon.image": route.Image,
92		"gordon.managed": "true",	114		"gordon.managed": "true",
93	-	"gordon.route": route.Domain,	115	+	"gordon.route": route.Domain,
94		},	116		},
95		AutoRemove: false, // Keep containers for inspection	117		AutoRemove: false, // Keep containers for inspection
96		}	118		}

@@ -131,7 +153,7 @@ func (m Manager) DeployContainer(ctx context.Context, route config.Route) (run
131		func (m Manager) GetContainer(domain string) (runtime.Container, bool) {	153		func (m Manager) GetContainer(domain string) (runtime.Container, bool) {
132		m.mu.RLock()	154		m.mu.RLock()
133		defer m.mu.RUnlock()	155		defer m.mu.RUnlock()
134	-		156	+
135		container, exists := m.containers[domain]	157		container, exists := m.containers[domain]
136		return container, exists	158		return container, exists
137		}	159		}

@@ -141,7 +163,7 @@ func (m *Manager) GetContainerPort(ctx context.Context, domain string, internalP
141		m.mu.RLock()	163		m.mu.RLock()
142		container, exists := m.containers[domain]	164		container, exists := m.containers[domain]
143		m.mu.RUnlock()	165		m.mu.RUnlock()
144	-		166	+
145		if !exists {	167		if !exists {
146		return 0, fmt.Errorf("no container found for domain %s", domain)	168		return 0, fmt.Errorf("no container found for domain %s", domain)
147		}	169		}

@@ -153,7 +175,7 @@ func (m *Manager) GetContainerPort(ctx context.Context, domain string, internalP
153		func (m Manager) ListContainers() map[string]runtime.Container {	175		func (m Manager) ListContainers() map[string]runtime.Container {
154		m.mu.RLock()	176		m.mu.RLock()
155		defer m.mu.RUnlock()	177		defer m.mu.RUnlock()
156	-		178	+
157		// Return a copy of the map to avoid race conditions	179		// Return a copy of the map to avoid race conditions
158		result := make(map[string]*runtime.Container)	180		result := make(map[string]*runtime.Container)
159		for domain, container := range m.containers {	181		for domain, container := range m.containers {

@@ -166,7 +188,7 @@ func (m Manager) ListContainers() map[string]runtime.Container {
166		func (m *Manager) FindContainerByDomain(domain string) (string, bool) {	188		func (m *Manager) FindContainerByDomain(domain string) (string, bool) {
167		m.mu.RLock()	189		m.mu.RLock()
168		defer m.mu.RUnlock()	190		defer m.mu.RUnlock()
169	-		191	+
170		if container, exists := m.containers[domain]; exists {	192		if container, exists := m.containers[domain]; exists {
171		return container.ID, true	193		return container.ID, true
172		}	194		}

@@ -177,7 +199,7 @@ func (m *Manager) FindContainerByDomain(domain string) (string, bool) {
177		func (m *Manager) FindDomainByContainerID(containerID string) (string, bool) {	199		func (m *Manager) FindDomainByContainerID(containerID string) (string, bool) {
178		m.mu.RLock()	200		m.mu.RLock()
179		defer m.mu.RUnlock()	201		defer m.mu.RUnlock()
180	-		202	+
181		for domain, container := range m.containers {	203		for domain, container := range m.containers {
182		if container.ID == containerID {	204		if container.ID == containerID {
183		return domain, true	205		return domain, true

@@ -201,7 +223,7 @@ func (m *Manager) StopContainerByDomain(ctx context.Context, domain string) erro
201		m.mu.RLock()	223		m.mu.RLock()
202		container, exists := m.containers[domain]	224		container, exists := m.containers[domain]
203		m.mu.RUnlock()	225		m.mu.RUnlock()
204	-		226	+
205		if !exists {	227		if !exists {
206		return fmt.Errorf("no container found for domain %s", domain)	228		return fmt.Errorf("no container found for domain %s", domain)
207		}	229		}

@@ -225,7 +247,7 @@ func (m *Manager) RemoveContainer(ctx context.Context, containerID string, force
225		break	247		break
226		}	248		}
227		}	249		}
228	-		250	+
229		return nil	251		return nil
230		}	252		}
231			253

@@ -234,7 +256,7 @@ func (m *Manager) RemoveContainerByDomain(ctx context.Context, domain string, fo
234		m.mu.RLock()	256		m.mu.RLock()
235		container, exists := m.containers[domain]	257		container, exists := m.containers[domain]
236		m.mu.RUnlock()	258		m.mu.RUnlock()
237	-		259	+
238		if !exists {	260		if !exists {
239		return fmt.Errorf("no container found for domain %s", domain)	261		return fmt.Errorf("no container found for domain %s", domain)
240		}	262		}

@@ -292,4 +314,4 @@ func (m *Manager) HealthCheck(ctx context.Context) map[string]bool {
292		// Runtime returns the underlying runtime interface	314		// Runtime returns the underlying runtime interface
293		func (m *Manager) Runtime() runtime.Runtime {	315		func (m *Manager) Runtime() runtime.Runtime {
294		return m.runtime	316		return m.runtime
295	-	}⏎̸	317	+	}

D internal/container/podman.go +0 -431

@@ -1,431 +0,0 @@
1	-	package container
2	-
3	-	import (
4	-	"context"
5	-	"fmt"
6	-	"io"
7	-	"strconv"
8	-	"strings"
9	-
10	-	"github.com/docker/docker/api/types/container"
11	-	"github.com/docker/docker/api/types/image"
12	-	"github.com/docker/docker/client"
13	-	"github.com/docker/go-connections/nat"
14	-	"github.com/rs/zerolog/log"
15	-
16	-	"gordon/pkg/runtime"
17	-	)
18	-
19	-	// PodmanRuntime implements the Runtime interface using Podman API
20	-	// Podman API is compatible with Docker API, so we can reuse the Docker client
21	-	type PodmanRuntime struct {
22	-	client *client.Client
23	-	}
24	-
25	-	// NewPodmanRuntime creates a new Podman runtime instance
26	-	func NewPodmanRuntime() (*PodmanRuntime, error) {
27	-	// Try to create with auto-detection
28	-	socketPath := getDefaultPodmanSocket(false) // Try root first
29	-	rt, err := createPodmanRuntimeWithSocket(socketPath)
30	-	if err != nil {
31	-	// Try rootless
32	-	socketPath = getDefaultPodmanSocket(true)
33	-	rt, err = createPodmanRuntimeWithSocket(socketPath)
34	-	if err != nil {
35	-	return nil, fmt.Errorf("failed to create Podman runtime: %w", err)
36	-	}
37	-	}
38	-
39	-	return rt.(*PodmanRuntime), nil
40	-	}
41	-
42	-	// CreateContainer creates a new container
43	-	func (p PodmanRuntime) CreateContainer(ctx context.Context, config runtime.ContainerConfig) (*runtime.Container, error) {
44	-	// Convert ports to Podman/Docker format
45	-	exposedPorts := make(nat.PortSet)
46	-	portBindings := make(nat.PortMap)
47	-
48	-	for _, port := range config.Ports {
49	-	containerPort := nat.Port(fmt.Sprintf("%d/tcp", port))
50	-	exposedPorts[containerPort] = struct{}{}
51	-
52	-	// Bind to random available port on host
53	-	portBindings[containerPort] = []nat.PortBinding{
54	-	{
55	-	HostIP: "0.0.0.0",
56	-	HostPort: "0", // Podman will assign a random available port
57	-	},
58	-	}
59	-	}
60	-
61	-	// Create container configuration
62	-	containerConfig := &container.Config{
63	-	Image: config.Image,
64	-	Env: config.Env,
65	-	ExposedPorts: exposedPorts,
66	-	WorkingDir: config.WorkingDir,
67	-	Cmd: config.Cmd,
68	-	Labels: config.Labels,
69	-	}
70	-
71	-	hostConfig := &container.HostConfig{
72	-	PortBindings: portBindings,
73	-	AutoRemove: config.AutoRemove,
74	-	}
75	-
76	-	// Create the container
77	-	resp, err := p.client.ContainerCreate(ctx, containerConfig, hostConfig, nil, nil, config.Name)
78	-	if err != nil {
79	-	return nil, fmt.Errorf("failed to create container: %w", err)
80	-	}
81	-
82	-	log.Info().Str("id", resp.ID).Str("name", config.Name).Str("image", config.Image).Msg("Podman container created")
83	-
84	-	// Inspect the created container to get full details
85	-	return p.InspectContainer(ctx, resp.ID)
86	-	}
87	-
88	-	// StartContainer starts a container
89	-	func (p *PodmanRuntime) StartContainer(ctx context.Context, containerID string) error {
90	-	err := p.client.ContainerStart(ctx, containerID, container.StartOptions{})
91	-	if err != nil {
92	-	return fmt.Errorf("failed to start container %s: %w", containerID, err)
93	-	}
94	-
95	-	log.Info().Str("id", containerID).Msg("Podman container started")
96	-	return nil
97	-	}
98	-
99	-	// StopContainer stops a container
100	-	func (p *PodmanRuntime) StopContainer(ctx context.Context, containerID string) error {
101	-	timeout := 30 // 30 seconds timeout
102	-	err := p.client.ContainerStop(ctx, containerID, container.StopOptions{Timeout: &timeout})
103	-	if err != nil {
104	-	return fmt.Errorf("failed to stop container %s: %w", containerID, err)
105	-	}
106	-
107	-	log.Info().Str("id", containerID).Msg("Podman container stopped")
108	-	return nil
109	-	}
110	-
111	-	// RestartContainer restarts a container
112	-	func (p *PodmanRuntime) RestartContainer(ctx context.Context, containerID string) error {
113	-	timeout := 30 // 30 seconds timeout
114	-	err := p.client.ContainerRestart(ctx, containerID, container.StopOptions{Timeout: &timeout})
115	-	if err != nil {
116	-	return fmt.Errorf("failed to restart container %s: %w", containerID, err)
117	-	}
118	-
119	-	log.Info().Str("id", containerID).Msg("Podman container restarted")
120	-	return nil
121	-	}
122	-
123	-	// RemoveContainer removes a container
124	-	func (p *PodmanRuntime) RemoveContainer(ctx context.Context, containerID string, force bool) error {
125	-	err := p.client.ContainerRemove(ctx, containerID, container.RemoveOptions{Force: force})
126	-	if err != nil {
127	-	return fmt.Errorf("failed to remove container %s: %w", containerID, err)
128	-	}
129	-
130	-	log.Info().Str("id", containerID).Bool("force", force).Msg("Podman container removed")
131	-	return nil
132	-	}
133	-
134	-	// ListContainers lists containers
135	-	func (p PodmanRuntime) ListContainers(ctx context.Context, all bool) ([]runtime.Container, error) {
136	-	containers, err := p.client.ContainerList(ctx, container.ListOptions{All: all})
137	-	if err != nil {
138	-	return nil, fmt.Errorf("failed to list containers: %w", err)
139	-	}
140	-
141	-	var result []*runtime.Container
142	-	for _, c := range containers {
143	-	// Extract ports
144	-	var ports []int
145	-	for _, port := range c.Ports {
146	-	if port.PublicPort > 0 {
147	-	ports = append(ports, int(port.PublicPort))
148	-	}
149	-	}
150	-
151	-	// Get the primary name (remove leading slash)
152	-	name := ""
153	-	if len(c.Names) > 0 {
154	-	name = strings.TrimPrefix(c.Names[0], "/")
155	-	}
156	-
157	-	result = append(result, &runtime.Container{
158	-	ID: c.ID,
159	-	Image: c.Image,
160	-	Name: name,
161	-	Status: c.Status,
162	-	Ports: ports,
163	-	Labels: c.Labels,
164	-	})
165	-	}
166	-
167	-	return result, nil
168	-	}
169	-
170	-	// InspectContainer inspects a container
171	-	func (p PodmanRuntime) InspectContainer(ctx context.Context, containerID string) (runtime.Container, error) {
172	-	resp, err := p.client.ContainerInspect(ctx, containerID)
173	-	if err != nil {
174	-	return nil, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
175	-	}
176	-
177	-	// Extract published ports
178	-	var ports []int
179	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
180	-	for _, bindings := range resp.NetworkSettings.Ports {
181	-	for _, binding := range bindings {
182	-	if binding.HostPort != "" {
183	-	if port, err := strconv.Atoi(binding.HostPort); err == nil {
184	-	ports = append(ports, port)
185	-	}
186	-	}
187	-	}
188	-	}
189	-	}
190	-
191	-	// Get container name (remove leading slash)
192	-	name := strings.TrimPrefix(resp.Name, "/")
193	-
194	-	return &runtime.Container{
195	-	ID: resp.ID,
196	-	Image: resp.Config.Image,
197	-	Name: name,
198	-	Status: resp.State.Status,
199	-	Ports: ports,
200	-	Labels: resp.Config.Labels,
201	-	}, nil
202	-	}
203	-
204	-	// GetContainerLogs gets container logs
205	-	func (p *PodmanRuntime) GetContainerLogs(ctx context.Context, containerID string, follow bool) (io.ReadCloser, error) {
206	-	logs, err := p.client.ContainerLogs(ctx, containerID, container.LogsOptions{
207	-	ShowStdout: true,
208	-	ShowStderr: true,
209	-	Follow: follow,
210	-	Timestamps: true,
211	-	})
212	-	if err != nil {
213	-	return nil, fmt.Errorf("failed to get logs for container %s: %w", containerID, err)
214	-	}
215	-
216	-	return logs, nil
217	-	}
218	-
219	-	// PullImage pulls an image
220	-	func (p *PodmanRuntime) PullImage(ctx context.Context, imageRef string) error {
221	-	log.Info().Str("image", imageRef).Msg("Pulling image with Podman")
222	-
223	-	reader, err := p.client.ImagePull(ctx, imageRef, image.PullOptions{})
224	-	if err != nil {
225	-	return fmt.Errorf("failed to pull image %s: %w", imageRef, err)
226	-	}
227	-	defer reader.Close()
228	-
229	-	// Read the response to completion (this is required for the pull to complete)
230	-	_, err = io.Copy(io.Discard, reader)
231	-	if err != nil {
232	-	return fmt.Errorf("failed to read pull response for image %s: %w", imageRef, err)
233	-	}
234	-
235	-	log.Info().Str("image", imageRef).Msg("Image pulled successfully with Podman")
236	-	return nil
237	-	}
238	-
239	-	// RemoveImage removes an image
240	-	func (p *PodmanRuntime) RemoveImage(ctx context.Context, imageRef string, force bool) error {
241	-	_, err := p.client.ImageRemove(ctx, imageRef, image.RemoveOptions{Force: force})
242	-	if err != nil {
243	-	return fmt.Errorf("failed to remove image %s: %w", imageRef, err)
244	-	}
245	-
246	-	log.Info().Str("image", imageRef).Bool("force", force).Msg("Podman image removed")
247	-	return nil
248	-	}
249	-
250	-	// ListImages lists images
251	-	func (p *PodmanRuntime) ListImages(ctx context.Context) ([]string, error) {
252	-	images, err := p.client.ImageList(ctx, image.ListOptions{})
253	-	if err != nil {
254	-	return nil, fmt.Errorf("failed to list images: %w", err)
255	-	}
256	-
257	-	var result []string
258	-	for _, img := range images {
259	-	for _, tag := range img.RepoTags {
260	-	if tag != "<none>:<none>" {
261	-	result = append(result, tag)
262	-	}
263	-	}
264	-	}
265	-
266	-	return result, nil
267	-	}
268	-
269	-	// Ping checks if Podman is responsive
270	-	func (p *PodmanRuntime) Ping(ctx context.Context) error {
271	-	_, err := p.client.Ping(ctx)
272	-	if err != nil {
273	-	return fmt.Errorf("Podman ping failed: %w", err)
274	-	}
275	-	return nil
276	-	}
277	-
278	-	// Version returns Podman version
279	-	func (p *PodmanRuntime) Version(ctx context.Context) (string, error) {
280	-	version, err := p.client.ServerVersion(ctx)
281	-	if err != nil {
282	-	return "", fmt.Errorf("failed to get Podman version: %w", err)
283	-	}
284	-	return version.Version, nil
285	-	}
286	-
287	-	// IsContainerRunning checks if a container is running
288	-	func (p *PodmanRuntime) IsContainerRunning(ctx context.Context, containerID string) (bool, error) {
289	-	container, err := p.InspectContainer(ctx, containerID)
290	-	if err != nil {
291	-	return false, err
292	-	}
293	-	return container.Status == "running", nil
294	-	}
295	-
296	-	// GetContainerPort gets the host port for a container's internal port
297	-	func (p *PodmanRuntime) GetContainerPort(ctx context.Context, containerID string, internalPort int) (int, error) {
298	-	resp, err := p.client.ContainerInspect(ctx, containerID)
299	-	if err != nil {
300	-	return 0, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
301	-	}
302	-
303	-	if resp.NetworkSettings == nil \|\| resp.NetworkSettings.Ports == nil {
304	-	return 0, fmt.Errorf("no port mappings found for container %s", containerID)
305	-	}
306	-
307	-	containerPort := nat.Port(fmt.Sprintf("%d/tcp", internalPort))
308	-	bindings, exists := resp.NetworkSettings.Ports[containerPort]
309	-	if !exists \|\| len(bindings) == 0 {
310	-	return 0, fmt.Errorf("port %d not mapped for container %s", internalPort, containerID)
311	-	}
312	-
313	-	hostPort, err := strconv.Atoi(bindings[0].HostPort)
314	-	if err != nil {
315	-	return 0, fmt.Errorf("invalid host port for container %s: %w", containerID, err)
316	-	}
317	-
318	-	return hostPort, nil
319	-	}
320	-
321	-	// GetImageExposedPorts gets the exposed ports from an image
322	-	func (p *PodmanRuntime) GetImageExposedPorts(ctx context.Context, imageRef string) ([]int, error) {
323	-	imageInspect, err := p.client.ImageInspect(ctx, imageRef)
324	-	if err != nil {
325	-	return nil, fmt.Errorf("failed to inspect image %s: %w", imageRef, err)
326	-	}
327	-
328	-	var ports []int
329	-	if imageInspect.Config != nil && imageInspect.Config.ExposedPorts != nil {
330	-	for portSpec := range imageInspect.Config.ExposedPorts {
331	-	// Parse port from format like "80/tcp"
332	-	portStr := strings.Split(string(portSpec), "/")[0]
333	-	if port, err := strconv.Atoi(portStr); err == nil {
334	-	ports = append(ports, port)
335	-	}
336	-	}
337	-	}
338	-
339	-	// If no exposed ports found, return common web ports as fallback
340	-	if len(ports) == 0 {
341	-	log.Warn().Str("image", imageRef).Msg("No EXPOSE directives found in image, using common web ports")
342	-	return []int{80, 8080, 3000}, nil
343	-	}
344	-
345	-	log.Info().Str("image", imageRef).Ints("exposed_ports", ports).Msg("Found exposed ports in image")
346	-	return ports, nil
347	-	}
348	-
349	-	// GetContainerExposedPorts gets all exposed ports from a running container
350	-	func (p *PodmanRuntime) GetContainerExposedPorts(ctx context.Context, containerID string) ([]int, error) {
351	-	resp, err := p.client.ContainerInspect(ctx, containerID)
352	-	if err != nil {
353	-	return nil, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
354	-	}
355	-
356	-	var ports []int
357	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
358	-	for portSpec := range resp.NetworkSettings.Ports {
359	-	// Parse port from format like "80/tcp"
360	-	portStr := strings.Split(string(portSpec), "/")[0]
361	-	if port, err := strconv.Atoi(portStr); err == nil {
362	-	ports = append(ports, port)
363	-	}
364	-	}
365	-	}
366	-
367	-	return ports, nil
368	-	}
369	-
370	-	// GetContainerNetworkInfo gets container's internal IP and automatically detects the best port to use
371	-	func (p *PodmanRuntime) GetContainerNetworkInfo(ctx context.Context, containerID string) (string, int, error) {
372	-	resp, err := p.client.ContainerInspect(ctx, containerID)
373	-	if err != nil {
374	-	return "", 0, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
375	-	}
376	-
377	-	// Get container's internal IP address
378	-	var containerIP string
379	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Networks != nil {
380	-	// Use the first available network (usually bridge or custom network)
381	-	for _, network := range resp.NetworkSettings.Networks {
382	-	if network.IPAddress != "" {
383	-	containerIP = network.IPAddress
384	-	break
385	-	}
386	-	}
387	-	}
388	-
389	-	if containerIP == "" {
390	-	return "", 0, fmt.Errorf("no IP address found for container %s", containerID)
391	-	}
392	-
393	-	// Get exposed ports and select the best one (Traefik-style logic)
394	-	var exposedPorts []int
395	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
396	-	for portSpec := range resp.NetworkSettings.Ports {
397	-	// Parse port from format like "80/tcp"
398	-	portStr := strings.Split(string(portSpec), "/")[0]
399	-	if port, err := strconv.Atoi(portStr); err == nil {
400	-	exposedPorts = append(exposedPorts, port)
401	-	}
402	-	}
403	-	}
404	-
405	-	if len(exposedPorts) == 0 {
406	-	return "", 0, fmt.Errorf("no exposed ports found for container %s", containerID)
407	-	}
408	-
409	-	// Sort ports and select the lowest one (Traefik strategy)
410	-	var selectedPort int
411	-	if len(exposedPorts) == 1 {
412	-	selectedPort = exposedPorts[0]
413	-	} else {
414	-	// Find the lowest port number
415	-	selectedPort = exposedPorts[0]
416	-	for _, port := range exposedPorts {
417	-	if port < selectedPort {
418	-	selectedPort = port
419	-	}
420	-	}
421	-	}
422	-
423	-	log.Info().
424	-	Str("container_id", containerID).
425	-	Str("ip", containerIP).
426	-	Int("selected_port", selectedPort).
427	-	Ints("available_ports", exposedPorts).
428	-	Msg("Podman container network info detected")
429	-
430	-	return containerIP, selectedPort, nil
431	-	}⏎̸

@@ -1,431 +0,0 @@
1	-	package container
2	-
3	-	import (
4	-	"context"
5	-	"fmt"
6	-	"io"
7	-	"strconv"
8	-	"strings"
9	-
10	-	"github.com/docker/docker/api/types/container"
11	-	"github.com/docker/docker/api/types/image"
12	-	"github.com/docker/docker/client"
13	-	"github.com/docker/go-connections/nat"
14	-	"github.com/rs/zerolog/log"
15	-
16	-	"gordon/pkg/runtime"
17	-	)
18	-
19	-	// PodmanRuntime implements the Runtime interface using Podman API
20	-	// Podman API is compatible with Docker API, so we can reuse the Docker client
21	-	type PodmanRuntime struct {
22	-	client *client.Client
23	-	}
24	-
25	-	// NewPodmanRuntime creates a new Podman runtime instance
26	-	func NewPodmanRuntime() (*PodmanRuntime, error) {
27	-	// Try to create with auto-detection
28	-	socketPath := getDefaultPodmanSocket(false) // Try root first
29	-	rt, err := createPodmanRuntimeWithSocket(socketPath)
30	-	if err != nil {
31	-	// Try rootless
32	-	socketPath = getDefaultPodmanSocket(true)
33	-	rt, err = createPodmanRuntimeWithSocket(socketPath)
34	-	if err != nil {
35	-	return nil, fmt.Errorf("failed to create Podman runtime: %w", err)
36	-	}
37	-	}
38	-
39	-	return rt.(*PodmanRuntime), nil
40	-	}
41	-
42	-	// CreateContainer creates a new container
43	-	func (p PodmanRuntime) CreateContainer(ctx context.Context, config runtime.ContainerConfig) (*runtime.Container, error) {
44	-	// Convert ports to Podman/Docker format
45	-	exposedPorts := make(nat.PortSet)
46	-	portBindings := make(nat.PortMap)
47	-
48	-	for _, port := range config.Ports {
49	-	containerPort := nat.Port(fmt.Sprintf("%d/tcp", port))
50	-	exposedPorts[containerPort] = struct{}{}
51	-
52	-	// Bind to random available port on host
53	-	portBindings[containerPort] = []nat.PortBinding{
54	-	{
55	-	HostIP: "0.0.0.0",
56	-	HostPort: "0", // Podman will assign a random available port
57	-	},
58	-	}
59	-	}
60	-
61	-	// Create container configuration
62	-	containerConfig := &container.Config{
63	-	Image: config.Image,
64	-	Env: config.Env,
65	-	ExposedPorts: exposedPorts,
66	-	WorkingDir: config.WorkingDir,
67	-	Cmd: config.Cmd,
68	-	Labels: config.Labels,
69	-	}
70	-
71	-	hostConfig := &container.HostConfig{
72	-	PortBindings: portBindings,
73	-	AutoRemove: config.AutoRemove,
74	-	}
75	-
76	-	// Create the container
77	-	resp, err := p.client.ContainerCreate(ctx, containerConfig, hostConfig, nil, nil, config.Name)
78	-	if err != nil {
79	-	return nil, fmt.Errorf("failed to create container: %w", err)
80	-	}
81	-
82	-	log.Info().Str("id", resp.ID).Str("name", config.Name).Str("image", config.Image).Msg("Podman container created")
83	-
84	-	// Inspect the created container to get full details
85	-	return p.InspectContainer(ctx, resp.ID)
86	-	}
87	-
88	-	// StartContainer starts a container
89	-	func (p *PodmanRuntime) StartContainer(ctx context.Context, containerID string) error {
90	-	err := p.client.ContainerStart(ctx, containerID, container.StartOptions{})
91	-	if err != nil {
92	-	return fmt.Errorf("failed to start container %s: %w", containerID, err)
93	-	}
94	-
95	-	log.Info().Str("id", containerID).Msg("Podman container started")
96	-	return nil
97	-	}
98	-
99	-	// StopContainer stops a container
100	-	func (p *PodmanRuntime) StopContainer(ctx context.Context, containerID string) error {
101	-	timeout := 30 // 30 seconds timeout
102	-	err := p.client.ContainerStop(ctx, containerID, container.StopOptions{Timeout: &timeout})
103	-	if err != nil {
104	-	return fmt.Errorf("failed to stop container %s: %w", containerID, err)
105	-	}
106	-
107	-	log.Info().Str("id", containerID).Msg("Podman container stopped")
108	-	return nil
109	-	}
110	-
111	-	// RestartContainer restarts a container
112	-	func (p *PodmanRuntime) RestartContainer(ctx context.Context, containerID string) error {
113	-	timeout := 30 // 30 seconds timeout
114	-	err := p.client.ContainerRestart(ctx, containerID, container.StopOptions{Timeout: &timeout})
115	-	if err != nil {
116	-	return fmt.Errorf("failed to restart container %s: %w", containerID, err)
117	-	}
118	-
119	-	log.Info().Str("id", containerID).Msg("Podman container restarted")
120	-	return nil
121	-	}
122	-
123	-	// RemoveContainer removes a container
124	-	func (p *PodmanRuntime) RemoveContainer(ctx context.Context, containerID string, force bool) error {
125	-	err := p.client.ContainerRemove(ctx, containerID, container.RemoveOptions{Force: force})
126	-	if err != nil {
127	-	return fmt.Errorf("failed to remove container %s: %w", containerID, err)
128	-	}
129	-
130	-	log.Info().Str("id", containerID).Bool("force", force).Msg("Podman container removed")
131	-	return nil
132	-	}
133	-
134	-	// ListContainers lists containers
135	-	func (p PodmanRuntime) ListContainers(ctx context.Context, all bool) ([]runtime.Container, error) {
136	-	containers, err := p.client.ContainerList(ctx, container.ListOptions{All: all})
137	-	if err != nil {
138	-	return nil, fmt.Errorf("failed to list containers: %w", err)
139	-	}
140	-
141	-	var result []*runtime.Container
142	-	for _, c := range containers {
143	-	// Extract ports
144	-	var ports []int
145	-	for _, port := range c.Ports {
146	-	if port.PublicPort > 0 {
147	-	ports = append(ports, int(port.PublicPort))
148	-	}
149	-	}
150	-
151	-	// Get the primary name (remove leading slash)
152	-	name := ""
153	-	if len(c.Names) > 0 {
154	-	name = strings.TrimPrefix(c.Names[0], "/")
155	-	}
156	-
157	-	result = append(result, &runtime.Container{
158	-	ID: c.ID,
159	-	Image: c.Image,
160	-	Name: name,
161	-	Status: c.Status,
162	-	Ports: ports,
163	-	Labels: c.Labels,
164	-	})
165	-	}
166	-
167	-	return result, nil
168	-	}
169	-
170	-	// InspectContainer inspects a container
171	-	func (p PodmanRuntime) InspectContainer(ctx context.Context, containerID string) (runtime.Container, error) {
172	-	resp, err := p.client.ContainerInspect(ctx, containerID)
173	-	if err != nil {
174	-	return nil, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
175	-	}
176	-
177	-	// Extract published ports
178	-	var ports []int
179	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
180	-	for _, bindings := range resp.NetworkSettings.Ports {
181	-	for _, binding := range bindings {
182	-	if binding.HostPort != "" {
183	-	if port, err := strconv.Atoi(binding.HostPort); err == nil {
184	-	ports = append(ports, port)
185	-	}
186	-	}
187	-	}
188	-	}
189	-	}
190	-
191	-	// Get container name (remove leading slash)
192	-	name := strings.TrimPrefix(resp.Name, "/")
193	-
194	-	return &runtime.Container{
195	-	ID: resp.ID,
196	-	Image: resp.Config.Image,
197	-	Name: name,
198	-	Status: resp.State.Status,
199	-	Ports: ports,
200	-	Labels: resp.Config.Labels,
201	-	}, nil
202	-	}
203	-
204	-	// GetContainerLogs gets container logs
205	-	func (p *PodmanRuntime) GetContainerLogs(ctx context.Context, containerID string, follow bool) (io.ReadCloser, error) {
206	-	logs, err := p.client.ContainerLogs(ctx, containerID, container.LogsOptions{
207	-	ShowStdout: true,
208	-	ShowStderr: true,
209	-	Follow: follow,
210	-	Timestamps: true,
211	-	})
212	-	if err != nil {
213	-	return nil, fmt.Errorf("failed to get logs for container %s: %w", containerID, err)
214	-	}
215	-
216	-	return logs, nil
217	-	}
218	-
219	-	// PullImage pulls an image
220	-	func (p *PodmanRuntime) PullImage(ctx context.Context, imageRef string) error {
221	-	log.Info().Str("image", imageRef).Msg("Pulling image with Podman")
222	-
223	-	reader, err := p.client.ImagePull(ctx, imageRef, image.PullOptions{})
224	-	if err != nil {
225	-	return fmt.Errorf("failed to pull image %s: %w", imageRef, err)
226	-	}
227	-	defer reader.Close()
228	-
229	-	// Read the response to completion (this is required for the pull to complete)
230	-	_, err = io.Copy(io.Discard, reader)
231	-	if err != nil {
232	-	return fmt.Errorf("failed to read pull response for image %s: %w", imageRef, err)
233	-	}
234	-
235	-	log.Info().Str("image", imageRef).Msg("Image pulled successfully with Podman")
236	-	return nil
237	-	}
238	-
239	-	// RemoveImage removes an image
240	-	func (p *PodmanRuntime) RemoveImage(ctx context.Context, imageRef string, force bool) error {
241	-	_, err := p.client.ImageRemove(ctx, imageRef, image.RemoveOptions{Force: force})
242	-	if err != nil {
243	-	return fmt.Errorf("failed to remove image %s: %w", imageRef, err)
244	-	}
245	-
246	-	log.Info().Str("image", imageRef).Bool("force", force).Msg("Podman image removed")
247	-	return nil
248	-	}
249	-
250	-	// ListImages lists images
251	-	func (p *PodmanRuntime) ListImages(ctx context.Context) ([]string, error) {
252	-	images, err := p.client.ImageList(ctx, image.ListOptions{})
253	-	if err != nil {
254	-	return nil, fmt.Errorf("failed to list images: %w", err)
255	-	}
256	-
257	-	var result []string
258	-	for _, img := range images {
259	-	for _, tag := range img.RepoTags {
260	-	if tag != "<none>:<none>" {
261	-	result = append(result, tag)
262	-	}
263	-	}
264	-	}
265	-
266	-	return result, nil
267	-	}
268	-
269	-	// Ping checks if Podman is responsive
270	-	func (p *PodmanRuntime) Ping(ctx context.Context) error {
271	-	_, err := p.client.Ping(ctx)
272	-	if err != nil {
273	-	return fmt.Errorf("Podman ping failed: %w", err)
274	-	}
275	-	return nil
276	-	}
277	-
278	-	// Version returns Podman version
279	-	func (p *PodmanRuntime) Version(ctx context.Context) (string, error) {
280	-	version, err := p.client.ServerVersion(ctx)
281	-	if err != nil {
282	-	return "", fmt.Errorf("failed to get Podman version: %w", err)
283	-	}
284	-	return version.Version, nil
285	-	}
286	-
287	-	// IsContainerRunning checks if a container is running
288	-	func (p *PodmanRuntime) IsContainerRunning(ctx context.Context, containerID string) (bool, error) {
289	-	container, err := p.InspectContainer(ctx, containerID)
290	-	if err != nil {
291	-	return false, err
292	-	}
293	-	return container.Status == "running", nil
294	-	}
295	-
296	-	// GetContainerPort gets the host port for a container's internal port
297	-	func (p *PodmanRuntime) GetContainerPort(ctx context.Context, containerID string, internalPort int) (int, error) {
298	-	resp, err := p.client.ContainerInspect(ctx, containerID)
299	-	if err != nil {
300	-	return 0, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
301	-	}
302	-
303	-	if resp.NetworkSettings == nil \|\| resp.NetworkSettings.Ports == nil {
304	-	return 0, fmt.Errorf("no port mappings found for container %s", containerID)
305	-	}
306	-
307	-	containerPort := nat.Port(fmt.Sprintf("%d/tcp", internalPort))
308	-	bindings, exists := resp.NetworkSettings.Ports[containerPort]
309	-	if !exists \|\| len(bindings) == 0 {
310	-	return 0, fmt.Errorf("port %d not mapped for container %s", internalPort, containerID)
311	-	}
312	-
313	-	hostPort, err := strconv.Atoi(bindings[0].HostPort)
314	-	if err != nil {
315	-	return 0, fmt.Errorf("invalid host port for container %s: %w", containerID, err)
316	-	}
317	-
318	-	return hostPort, nil
319	-	}
320	-
321	-	// GetImageExposedPorts gets the exposed ports from an image
322	-	func (p *PodmanRuntime) GetImageExposedPorts(ctx context.Context, imageRef string) ([]int, error) {
323	-	imageInspect, err := p.client.ImageInspect(ctx, imageRef)
324	-	if err != nil {
325	-	return nil, fmt.Errorf("failed to inspect image %s: %w", imageRef, err)
326	-	}
327	-
328	-	var ports []int
329	-	if imageInspect.Config != nil && imageInspect.Config.ExposedPorts != nil {
330	-	for portSpec := range imageInspect.Config.ExposedPorts {
331	-	// Parse port from format like "80/tcp"
332	-	portStr := strings.Split(string(portSpec), "/")[0]
333	-	if port, err := strconv.Atoi(portStr); err == nil {
334	-	ports = append(ports, port)
335	-	}
336	-	}
337	-	}
338	-
339	-	// If no exposed ports found, return common web ports as fallback
340	-	if len(ports) == 0 {
341	-	log.Warn().Str("image", imageRef).Msg("No EXPOSE directives found in image, using common web ports")
342	-	return []int{80, 8080, 3000}, nil
343	-	}
344	-
345	-	log.Info().Str("image", imageRef).Ints("exposed_ports", ports).Msg("Found exposed ports in image")
346	-	return ports, nil
347	-	}
348	-
349	-	// GetContainerExposedPorts gets all exposed ports from a running container
350	-	func (p *PodmanRuntime) GetContainerExposedPorts(ctx context.Context, containerID string) ([]int, error) {
351	-	resp, err := p.client.ContainerInspect(ctx, containerID)
352	-	if err != nil {
353	-	return nil, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
354	-	}
355	-
356	-	var ports []int
357	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
358	-	for portSpec := range resp.NetworkSettings.Ports {
359	-	// Parse port from format like "80/tcp"
360	-	portStr := strings.Split(string(portSpec), "/")[0]
361	-	if port, err := strconv.Atoi(portStr); err == nil {
362	-	ports = append(ports, port)
363	-	}
364	-	}
365	-	}
366	-
367	-	return ports, nil
368	-	}
369	-
370	-	// GetContainerNetworkInfo gets container's internal IP and automatically detects the best port to use
371	-	func (p *PodmanRuntime) GetContainerNetworkInfo(ctx context.Context, containerID string) (string, int, error) {
372	-	resp, err := p.client.ContainerInspect(ctx, containerID)
373	-	if err != nil {
374	-	return "", 0, fmt.Errorf("failed to inspect container %s: %w", containerID, err)
375	-	}
376	-
377	-	// Get container's internal IP address
378	-	var containerIP string
379	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Networks != nil {
380	-	// Use the first available network (usually bridge or custom network)
381	-	for _, network := range resp.NetworkSettings.Networks {
382	-	if network.IPAddress != "" {
383	-	containerIP = network.IPAddress
384	-	break
385	-	}
386	-	}
387	-	}
388	-
389	-	if containerIP == "" {
390	-	return "", 0, fmt.Errorf("no IP address found for container %s", containerID)
391	-	}
392	-
393	-	// Get exposed ports and select the best one (Traefik-style logic)
394	-	var exposedPorts []int
395	-	if resp.NetworkSettings != nil && resp.NetworkSettings.Ports != nil {
396	-	for portSpec := range resp.NetworkSettings.Ports {
397	-	// Parse port from format like "80/tcp"
398	-	portStr := strings.Split(string(portSpec), "/")[0]
399	-	if port, err := strconv.Atoi(portStr); err == nil {
400	-	exposedPorts = append(exposedPorts, port)
401	-	}
402	-	}
403	-	}
404	-
405	-	if len(exposedPorts) == 0 {
406	-	return "", 0, fmt.Errorf("no exposed ports found for container %s", containerID)
407	-	}
408	-
409	-	// Sort ports and select the lowest one (Traefik strategy)
410	-	var selectedPort int
411	-	if len(exposedPorts) == 1 {
412	-	selectedPort = exposedPorts[0]
413	-	} else {
414	-	// Find the lowest port number
415	-	selectedPort = exposedPorts[0]
416	-	for _, port := range exposedPorts {
417	-	if port < selectedPort {
418	-	selectedPort = port
419	-	}
420	-	}
421	-	}
422	-
423	-	log.Info().
424	-	Str("container_id", containerID).
425	-	Str("ip", containerIP).
426	-	Int("selected_port", selectedPort).
427	-	Ints("available_ports", exposedPorts).
428	-	Msg("Podman container network info detected")
429	-
430	-	return containerIP, selectedPort, nil
431	-	}⏎̸

M internal/events/bus.go +32 -26


@@ -11,22 +11,22 @@ import (
11	11		)
12	12
13	13		type InMemoryEventBus struct {
14		-	handlers []EventHandler
15		-	eventChan chan Event
16		-	done chan struct{}
17		-	mu sync.RWMutex
18		-	ctx context.Context
19		-	cancel context.CancelFunc
20		-	bufferSize int
	14	+	handlers []EventHandler
	15	+	eventChan chan Event
	16	+	done chan struct{}
	17	+	mu sync.RWMutex
	18	+	ctx context.Context
	19	+	cancel context.CancelFunc
	20	+	bufferSize int
21	21		}
22	22
23	23		func NewInMemoryEventBus(bufferSize int) *InMemoryEventBus {
24	24		if bufferSize <= 0 {
25	25		bufferSize = 100
26	26		}
27		-
	27	+
28	28		ctx, cancel := context.WithCancel(context.Background())
29		-
	29	+
30	30		return &InMemoryEventBus{
31	31		handlers: make([]EventHandler, 0),
32	32		eventChan: make(chan Event, bufferSize),

@@ -37,13 +37,19 @@ func NewInMemoryEventBus(bufferSize int) *InMemoryEventBus {
37	37		}
38	38		}
39	39
40		-	func (bus *InMemoryEventBus) Publish(event Event) error {
41		-	if event.ID == "" {
42		-	event.ID = uuid.New().String()
	40	+	func (bus *InMemoryEventBus) Publish(eventType EventType, payload interface{}) error {
	41	+	event := Event{
	42	+	ID: uuid.New().String(),
	43	+	Type: eventType,
	44	+	Timestamp: time.Now(),
	45	+	Data: payload,
43	46		}
44		-
45		-	if event.Timestamp.IsZero() {
46		-	event.Timestamp = time.Now()
	47	+
	48	+	// Add specific fields for known event types
	49	+	switch p := payload.(type) {
	50	+	case ImagePushedPayload:
	51	+	event.ImageName = p.Name
	52	+	event.Tag = p.Reference
47	53		}
48	54
49	55		select {

@@ -65,20 +71,20 @@ func (bus *InMemoryEventBus) Publish(event Event) error {
65	71		func (bus *InMemoryEventBus) Subscribe(handler EventHandler) error {
66	72		bus.mu.Lock()
67	73		defer bus.mu.Unlock()
68		-
	74	+
69	75		bus.handlers = append(bus.handlers, handler)
70	76		log.Debug().
71	77		Str("handler_type", fmt.Sprintf("%T", handler)).
72	78		Int("total_handlers", len(bus.handlers)).
73	79		Msg("Event handler subscribed")
74		-
	80	+
75	81		return nil
76	82		}
77	83
78	84		func (bus *InMemoryEventBus) Unsubscribe(handler EventHandler) error {
79	85		bus.mu.Lock()
80	86		defer bus.mu.Unlock()
81		-
	87	+
82	88		for i, h := range bus.handlers {
83	89		if h == handler {
84	90		bus.handlers = append(bus.handlers[:i], bus.handlers[i+1:]...)

@@ -89,7 +95,7 @@ func (bus *InMemoryEventBus) Unsubscribe(handler EventHandler) error {
89	95		return nil
90	96		}
91	97		}
92		-
	98	+
93	99		return fmt.Errorf("handler not found")
94	100		}
95	101

@@ -97,16 +103,16 @@ func (bus *InMemoryEventBus) Start() error {
97	103		log.Info().
98	104		Int("buffer_size", bus.bufferSize).
99	105		Msg("Starting event bus")
100		-
	106	+
101	107		go bus.processEvents()
102	108		return nil
103	109		}
104	110
105	111		func (bus *InMemoryEventBus) Stop() error {
106	112		log.Info().Msg("Stopping event bus")
107		-
	113	+
108	114		bus.cancel()
109		-
	115	+
110	116		select {
111	117		case <-bus.done:
112	118		log.Info().Msg("Event bus stopped")

@@ -119,7 +125,7 @@ func (bus *InMemoryEventBus) Stop() error {
119	125
120	126		func (bus *InMemoryEventBus) processEvents() {
121	127		defer close(bus.done)
122		-
	128	+
123	129		for {
124	130		select {
125	131		case event := <-bus.eventChan:

@@ -136,12 +142,12 @@ func (bus *InMemoryEventBus) handleEvent(event Event) {
136	142		handlers := make([]EventHandler, len(bus.handlers))
137	143		copy(handlers, bus.handlers)
138	144		bus.mu.RUnlock()
139		-
	145	+
140	146		for _, handler := range handlers {
141	147		if handler.CanHandle(event.Type) {
142	148		go func(h EventHandler) {
143	149		start := time.Now()
144		-
	150	+
145	151		if err := h.Handle(event); err != nil {
146	152		log.Error().
147	153		Err(err).

@@ -160,4 +166,4 @@ func (bus *InMemoryEventBus) handleEvent(event Event) {
160	166		}(handler)
161	167		}
162	168		}
163		-	}⏎̸
	169	+	}

 )
 type InMemoryEventBus struct {
-	handlers    []EventHandler
+	handlers   []EventHandler
-	eventChan   chan Event
+	eventChan  chan Event
-	done        chan struct{}
+	done       chan struct{}
-	mu          sync.RWMutex
+	mu         sync.RWMutex
-	ctx         context.Context
+	ctx        context.Context
-	cancel      context.CancelFunc
+	cancel     context.CancelFunc
-	bufferSize  int
+	bufferSize int
 }
 func NewInMemoryEventBus(bufferSize int) *InMemoryEventBus {
 	if bufferSize <= 0 {
 		bufferSize = 100
 	}
 	ctx, cancel := context.WithCancel(context.Background())
 	return &InMemoryEventBus{
 		handlers:   make([]EventHandler, 0),
 		eventChan:  make(chan Event, bufferSize),
 	}
 }
-func (bus *InMemoryEventBus) Publish(event Event) error {
+func (bus *InMemoryEventBus) Publish(eventType EventType, payload interface{}) error {
-	if event.ID == "" {
+	event := Event{
-		event.ID = uuid.New().String()
+		ID:        uuid.New().String(),
+		Type:      eventType,
+		Timestamp: time.Now(),
+		Data:      payload,
 	}
-	if event.Timestamp.IsZero() {
+	// Add specific fields for known event types
-		event.Timestamp = time.Now()
+	switch p := payload.(type) {
+	case ImagePushedPayload:
+		event.ImageName = p.Name
+		event.Tag = p.Reference
 	}
 	select {
 func (bus *InMemoryEventBus) Subscribe(handler EventHandler) error {
 	bus.mu.Lock()
 	defer bus.mu.Unlock()
 	bus.handlers = append(bus.handlers, handler)
 	log.Debug().
 		Str("handler_type", fmt.Sprintf("%T", handler)).
 		Int("total_handlers", len(bus.handlers)).
 		Msg("Event handler subscribed")
 	return nil
 }
 func (bus *InMemoryEventBus) Unsubscribe(handler EventHandler) error {
 	bus.mu.Lock()
 	defer bus.mu.Unlock()
 	for i, h := range bus.handlers {
 		if h == handler {
 			bus.handlers = append(bus.handlers[:i], bus.handlers[i+1:]...)
 			return nil
 		}
 	}
 	return fmt.Errorf("handler not found")
 }
 	log.Info().
 		Int("buffer_size", bus.bufferSize).
 		Msg("Starting event bus")
 	go bus.processEvents()
 	return nil
 }
 func (bus *InMemoryEventBus) Stop() error {
 	log.Info().Msg("Stopping event bus")
 	bus.cancel()
 	select {
 	case <-bus.done:
 		log.Info().Msg("Event bus stopped")
 func (bus *InMemoryEventBus) processEvents() {
 	defer close(bus.done)
 	for {
 		select {
 		case event := <-bus.eventChan:
 	handlers := make([]EventHandler, len(bus.handlers))
 	copy(handlers, bus.handlers)
 	bus.mu.RUnlock()
 	for _, handler := range handlers {
 		if handler.CanHandle(event.Type) {
 			go func(h EventHandler) {
 				start := time.Now()
 				if err := h.Handle(event); err != nil {
 					log.Error().
 						Err(err).
 			}(handler)
 		}
 	}
-}⏎̸
+}

M internal/events/types.go +21 -15


@@ -7,23 +7,29 @@ import (
7	7		type EventType string
8	8
9	9		const (
10		-	ImagePushed EventType = "image.pushed"
11		-	ImageDeleted EventType = "image.deleted"
12		-	ConfigReload EventType = "config.reload"
13		-	ContainerStop EventType = "container.stop"
14		-	ContainerStart EventType = "container.start"
	10	+	ImagePushed EventType = "image.pushed"
	11	+	ImageDeleted EventType = "image.deleted"
	12	+	ConfigReload EventType = "config.reload"
	13	+	ContainerStop EventType = "container.stop"
	14	+	ContainerStart EventType = "container.start"
15	15		ContainerHealthCheck EventType = "container.health_check"
16	16		)
17	17
18	18		type Event struct {
19		-	ID string `json:"id"`
20		-	Type EventType `json:"type"`
21		-	Timestamp time.Time `json:"timestamp"`
22		-	ImageName string `json:"image_name,omitempty"`
23		-	Tag string `json:"tag,omitempty"`
24		-	Route string `json:"route,omitempty"`
25		-	ContainerID string `json:"container_id,omitempty"`
26		-	Data map[string]interface{} `json:"data,omitempty"`
	19	+	ID string `json:"id"`
	20	+	Type EventType `json:"type"`
	21	+	Timestamp time.Time `json:"timestamp"`
	22	+	ImageName string `json:"image_name,omitempty"`
	23	+	Tag string `json:"tag,omitempty"`
	24	+	Route string `json:"route,omitempty"`
	25	+	ContainerID string `json:"container_id,omitempty"`
	26	+	Data interface{} `json:"data,omitempty"`
	27	+	}
	28	+
	29	+	type ImagePushedPayload struct {
	30	+	Name string `json:"name"`
	31	+	Reference string `json:"reference"`
	32	+	Manifest []byte `json:"-"`
27	33		}
28	34
29	35		type EventHandler interface {

@@ -32,9 +38,9 @@ type EventHandler interface {
32	38		}
33	39
34	40		type EventBus interface {
35		-	Publish(event Event) error
	41	+	Publish(eventType EventType, payload interface{}) error
36	42		Subscribe(handler EventHandler) error
37	43		Unsubscribe(handler EventHandler) error
38	44		Start() error
39	45		Stop() error
40		-	}⏎̸
	46	+	}


@@ -7,23 +7,29 @@ import (
7		type EventType string	7		type EventType string
8			8
9		const (	9		const (
10	-	ImagePushed EventType = "image.pushed"	10	+	ImagePushed EventType = "image.pushed"
11	-	ImageDeleted EventType = "image.deleted"	11	+	ImageDeleted EventType = "image.deleted"
12	-	ConfigReload EventType = "config.reload"	12	+	ConfigReload EventType = "config.reload"
13	-	ContainerStop EventType = "container.stop"	13	+	ContainerStop EventType = "container.stop"
14	-	ContainerStart EventType = "container.start"	14	+	ContainerStart EventType = "container.start"
15		ContainerHealthCheck EventType = "container.health_check"	15		ContainerHealthCheck EventType = "container.health_check"
16		)	16		)
17			17
18		type Event struct {	18		type Event struct {
19	-	ID string `json:"id"`	19	+	ID string `json:"id"`
20	-	Type EventType `json:"type"`	20	+	Type EventType `json:"type"`
21	-	Timestamp time.Time `json:"timestamp"`	21	+	Timestamp time.Time `json:"timestamp"`
22	-	ImageName string `json:"image_name,omitempty"`	22	+	ImageName string `json:"image_name,omitempty"`
23	-	Tag string `json:"tag,omitempty"`	23	+	Tag string `json:"tag,omitempty"`
24	-	Route string `json:"route,omitempty"`	24	+	Route string `json:"route,omitempty"`
25	-	ContainerID string `json:"container_id,omitempty"`	25	+	ContainerID string `json:"container_id,omitempty"`
26	-	Data map[string]interface{} `json:"data,omitempty"`	26	+	Data interface{} `json:"data,omitempty"`
			27	+	}
			28	+
			29	+	type ImagePushedPayload struct {
			30	+	Name string `json:"name"`
			31	+	Reference string `json:"reference"`
			32	+	Manifest []byte `json:"-"`
27		}	33		}
28			34
29		type EventHandler interface {	35		type EventHandler interface {

@@ -32,9 +38,9 @@ type EventHandler interface {
32		}	38		}
33			39
34		type EventBus interface {	40		type EventBus interface {
35	-	Publish(event Event) error	41	+	Publish(eventType EventType, payload interface{}) error
36		Subscribe(handler EventHandler) error	42		Subscribe(handler EventHandler) error
37		Unsubscribe(handler EventHandler) error	43		Unsubscribe(handler EventHandler) error
38		Start() error	44		Start() error
39		Stop() error	45		Stop() error
40	-	}⏎̸	46	+	}

M internal/middleware/auth.go +22 -12


@@ -11,18 +11,12 @@ import (
11	11		func RegistryAuth(username, password string) func(http.Handler) http.Handler {
12	12		return func(next http.Handler) http.Handler {
13	13		return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
14		-	// Skip auth for registry info endpoint
15		-	if r.URL.Path == "/v2/" && r.Method == "GET" {
16		-	next.ServeHTTP(w, r)
17		-	return
18		-	}
19		-
20	14		// Require auth for all other registry operations
21	15		if !isAuthenticated(r, username, password) {
22	16		w.Header().Set("WWW-Authenticate", `Basic realm="Gordon Registry"`)
23	17		w.Header().Set("Docker-Distribution-API-Version", "registry/2.0")
24	18		http.Error(w, "Unauthorized", http.StatusUnauthorized)
25		-
	19	+
26	20		log.Warn().
27	21		Str("method", r.Method).
28	22		Str("path", r.URL.Path).

@@ -30,7 +24,7 @@ func RegistryAuth(username, password string) func(http.Handler) http.Handler {
30	24		Msg("Unauthorized registry access attempt")
31	25		return
32	26		}
33		-
	27	+
34	28		next.ServeHTTP(w, r)
35	29		})
36	30		}

@@ -39,19 +33,28 @@ func RegistryAuth(username, password string) func(http.Handler) http.Handler {
39	33		// Helper functions
40	34
41	35		func isAuthenticated(r *http.Request, expectedUsername, expectedPassword string) bool {
	36	+	// Debug: Log raw Authorization header
	37	+	authHeader := r.Header.Get("Authorization")
	38	+	log.Debug().
	39	+	Str("method", r.Method).
	40	+	Str("path", r.URL.Path).
	41	+	Str("authorization_header", authHeader).
	42	+	Msg("Processing authentication request")
	43	+
42	44		username, password, ok := r.BasicAuth()
43	45		if !ok {
44	46		log.Debug().
45	47		Str("method", r.Method).
46	48		Str("path", r.URL.Path).
	49	+	Str("authorization_header", authHeader).
47	50		Msg("No basic auth provided")
48	51		return false
49	52		}
50		-
	53	+
51	54		// Use constant-time comparison to prevent timing attacks
52	55		usernameMatch := subtle.ConstantTimeCompare([]byte(username), []byte(expectedUsername)) == 1
53	56		passwordMatch := subtle.ConstantTimeCompare([]byte(password), []byte(expectedPassword)) == 1
54		-
	57	+
55	58		authenticated := usernameMatch && passwordMatch
56	59		if !authenticated {
57	60		log.Debug().

@@ -59,8 +62,15 @@ func isAuthenticated(r *http.Request, expectedUsername, expectedPassword string)
59	62		Str("expected_username", expectedUsername).
60	63		Str("method", r.Method).
61	64		Str("path", r.URL.Path).
	65	+	Str("authorization_header", authHeader).
62	66		Msg("Authentication failed")
	67	+	} else {
	68	+	log.Debug().
	69	+	Str("provided_username", username).
	70	+	Str("method", r.Method).
	71	+	Str("path", r.URL.Path).
	72	+	Msg("Authentication successful")
63	73		}
64		-
	74	+
65	75		return authenticated
66		-	}⏎̸
	76	+	}

 func RegistryAuth(username, password string) func(http.Handler) http.Handler {
 	return func(next http.Handler) http.Handler {
 		return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-			// Skip auth for registry info endpoint
-			if r.URL.Path == "/v2/" && r.Method == "GET" {
-				next.ServeHTTP(w, r)
-				return
-			}
 			// Require auth for all other registry operations
 			if !isAuthenticated(r, username, password) {
 				w.Header().Set("WWW-Authenticate", `Basic realm="Gordon Registry"`)
 				w.Header().Set("Docker-Distribution-API-Version", "registry/2.0")
 				http.Error(w, "Unauthorized", http.StatusUnauthorized)
 				log.Warn().
 					Str("method", r.Method).
 					Str("path", r.URL.Path).
 					Msg("Unauthorized registry access attempt")
 				return
 			}
 			next.ServeHTTP(w, r)
 		})
 	}
 // Helper functions
 func isAuthenticated(r *http.Request, expectedUsername, expectedPassword string) bool {
+	// Debug: Log raw Authorization header
+	authHeader := r.Header.Get("Authorization")
+	log.Debug().
+		Str("method", r.Method).
+		Str("path", r.URL.Path).
+		Str("authorization_header", authHeader).
+		Msg("Processing authentication request")
 	username, password, ok := r.BasicAuth()
 	if !ok {
 		log.Debug().
 			Str("method", r.Method).
 			Str("path", r.URL.Path).
+			Str("authorization_header", authHeader).
 			Msg("No basic auth provided")
 		return false
 	}
 	// Use constant-time comparison to prevent timing attacks
 	usernameMatch := subtle.ConstantTimeCompare([]byte(username), []byte(expectedUsername)) == 1
 	passwordMatch := subtle.ConstantTimeCompare([]byte(password), []byte(expectedPassword)) == 1
 	authenticated := usernameMatch && passwordMatch
 	if !authenticated {
 		log.Debug().
 			Str("expected_username", expectedUsername).
 			Str("method", r.Method).
 			Str("path", r.URL.Path).
+			Str("authorization_header", authHeader).
 			Msg("Authentication failed")
+	} else {
+		log.Debug().
+			Str("provided_username", username).
+			Str("method", r.Method).
+			Str("path", r.URL.Path).
+			Msg("Authentication successful")
 	}
 	return authenticated
-}⏎̸
+}

M internal/proxy/server.go +79 -13


@@ -6,7 +6,9 @@ import (
6	6		"net/http"
7	7		"net/http/httputil"
8	8		"net/url"
	9	+	"os"
9	10		"strconv"
	11	+	"strings"
10	12
11	13		"github.com/rs/zerolog/log"
12	14		"gordon/internal/config"

@@ -72,19 +74,46 @@ func (s Server) proxyToContainer(w http.ResponseWriter, r http.Request, route
72	74		return
73	75		}
74	76
75		-	// Get container's internal network info (Traefik-style automatic detection)
76		-	containerIP, containerPort, err := s.manager.Runtime().GetContainerNetworkInfo(ctx, container.ID)
77		-	if err != nil {
78		-	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Msg("Failed to get container network info")
79		-	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
80		-	return
81		-	}
82		-
83		-	target, err := url.Parse(fmt.Sprintf("http://%s:%d", containerIP, containerPort))
84		-	if err != nil {
85		-	log.Error().Err(err).Msg("Failed to parse target URL")
86		-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
87		-	return
	77	+	// Detect if Gordon is running in a container or on the host
	78	+	var target *url.URL
	79	+	if s.isRunningInContainer() {
	80	+	// Gordon is in a container - use container network
	81	+	containerIP, containerPort, err := s.manager.Runtime().GetContainerNetworkInfo(ctx, container.ID)
	82	+	if err != nil {
	83	+	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Msg("Failed to get container network info")
	84	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
	85	+	return
	86	+	}
	87	+	target, err = url.Parse(fmt.Sprintf("http://%s:%d", containerIP, containerPort))
	88	+	if err != nil {
	89	+	log.Error().Err(err).Msg("Failed to parse container target URL")
	90	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
	91	+	return
	92	+	}
	93	+	log.Debug().Str("mode", "container").Str("target", target.String()).Msg("Using container-to-container networking")
	94	+	} else {
	95	+	// Gordon is on the host - use host port mapping
	96	+	exposedPorts, err := s.manager.Runtime().GetImageExposedPorts(ctx, route.Image)
	97	+	if err != nil {
	98	+	log.Error().Err(err).Str("domain", route.Domain).Str("image", route.Image).Msg("Failed to get exposed ports from image")
	99	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
	100	+	return
	101	+	}
	102	+
	103	+	hostPort, err := s.manager.Runtime().GetContainerPort(ctx, container.ID, exposedPorts[0])
	104	+	if err != nil {
	105	+	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Int("internal_port", exposedPorts[0]).Msg("Failed to get host port mapping")
	106	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
	107	+	return
	108	+	}
	109	+
	110	+	target, err = url.Parse(fmt.Sprintf("http://localhost:%d", hostPort))
	111	+	if err != nil {
	112	+	log.Error().Err(err).Msg("Failed to parse host target URL")
	113	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
	114	+	return
	115	+	}
	116	+	log.Debug().Str("mode", "host").Str("target", target.String()).Msg("Using host port mapping")
88	117		}
89	118
90	119		proxy := httputil.NewSingleHostReverseProxy(target)

@@ -146,6 +175,43 @@ func (s Server) proxyToRegistry(w http.ResponseWriter, r http.Request) {
146	175		proxy.ServeHTTP(w, r)
147	176		}
148	177
	178	+	// isRunningInContainer detects if Gordon is running inside a container
	179	+	func (s *Server) isRunningInContainer() bool {
	180	+	// Check for common container indicators
	181	+
	182	+	// 1. Check for /.dockerenv file (Docker creates this)
	183	+	if _, err := os.Stat("/.dockerenv"); err == nil {
	184	+	return true
	185	+	}
	186	+
	187	+	// 2. Check cgroup for container indicators
	188	+	if data, err := os.ReadFile("/proc/1/cgroup"); err == nil {
	189	+	content := string(data)
	190	+	if strings.Contains(content, "docker") \|\|
	191	+	strings.Contains(content, "containerd") \|\|
	192	+	strings.Contains(content, "podman") {
	193	+	return true
	194	+	}
	195	+	}
	196	+
	197	+	// 3. Check if hostname matches container ID pattern (64 hex chars)
	198	+	if hostname, err := os.Hostname(); err == nil {
	199	+	if len(hostname) == 12 \|\| len(hostname) == 64 {
	200	+	// Container hostnames are typically 12 or 64 character hex strings
	201	+	return true
	202	+	}
	203	+	}
	204	+
	205	+	// 4. Check environment variables set by container runtimes
	206	+	if os.Getenv("KUBERNETES_SERVICE_HOST") != "" \|\|
	207	+	os.Getenv("DOCKER_CONTAINER") != "" \|\|
	208	+	os.Getenv("container") != "" {
	209	+	return true
	210	+	}
	211	+
	212	+	return false
	213	+	}
	214	+
149	215		func (s *Server) Start(ctx context.Context) error {
150	216		addr := ":" + strconv.Itoa(s.config.Server.Port)
151	217


@@ -6,7 +6,9 @@ import (
6		"net/http"	6		"net/http"
7		"net/http/httputil"	7		"net/http/httputil"
8		"net/url"	8		"net/url"
			9	+	"os"
9		"strconv"	10		"strconv"
			11	+	"strings"
10			12
11		"github.com/rs/zerolog/log"	13		"github.com/rs/zerolog/log"
12		"gordon/internal/config"	14		"gordon/internal/config"

@@ -72,19 +74,46 @@ func (s Server) proxyToContainer(w http.ResponseWriter, r http.Request, route
72		return	74		return
73		}	75		}
74			76
75	-	// Get container's internal network info (Traefik-style automatic detection)	77	+	// Detect if Gordon is running in a container or on the host
76	-	containerIP, containerPort, err := s.manager.Runtime().GetContainerNetworkInfo(ctx, container.ID)	78	+	var target *url.URL
77	-	if err != nil {	79	+	if s.isRunningInContainer() {
78	-	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Msg("Failed to get container network info")	80	+	// Gordon is in a container - use container network
79	-	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)	81	+	containerIP, containerPort, err := s.manager.Runtime().GetContainerNetworkInfo(ctx, container.ID)
80	-	return	82	+	if err != nil {
81	-	}	83	+	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Msg("Failed to get container network info")
82	-		84	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
83	-	target, err := url.Parse(fmt.Sprintf("http://%s:%d", containerIP, containerPort))	85	+	return
84	-	if err != nil {	86	+	}
85	-	log.Error().Err(err).Msg("Failed to parse target URL")	87	+	target, err = url.Parse(fmt.Sprintf("http://%s:%d", containerIP, containerPort))
86	-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)	88	+	if err != nil {
87	-	return	89	+	log.Error().Err(err).Msg("Failed to parse container target URL")
			90	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
			91	+	return
			92	+	}
			93	+	log.Debug().Str("mode", "container").Str("target", target.String()).Msg("Using container-to-container networking")
			94	+	} else {
			95	+	// Gordon is on the host - use host port mapping
			96	+	exposedPorts, err := s.manager.Runtime().GetImageExposedPorts(ctx, route.Image)
			97	+	if err != nil {
			98	+	log.Error().Err(err).Str("domain", route.Domain).Str("image", route.Image).Msg("Failed to get exposed ports from image")
			99	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
			100	+	return
			101	+	}
			102	+
			103	+	hostPort, err := s.manager.Runtime().GetContainerPort(ctx, container.ID, exposedPorts[0])
			104	+	if err != nil {
			105	+	log.Error().Err(err).Str("domain", route.Domain).Str("container_id", container.ID).Int("internal_port", exposedPorts[0]).Msg("Failed to get host port mapping")
			106	+	http.Error(w, "Service Unavailable", http.StatusServiceUnavailable)
			107	+	return
			108	+	}
			109	+
			110	+	target, err = url.Parse(fmt.Sprintf("http://localhost:%d", hostPort))
			111	+	if err != nil {
			112	+	log.Error().Err(err).Msg("Failed to parse host target URL")
			113	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
			114	+	return
			115	+	}
			116	+	log.Debug().Str("mode", "host").Str("target", target.String()).Msg("Using host port mapping")
88		}	117		}
89			118
90		proxy := httputil.NewSingleHostReverseProxy(target)	119		proxy := httputil.NewSingleHostReverseProxy(target)

@@ -146,6 +175,43 @@ func (s Server) proxyToRegistry(w http.ResponseWriter, r http.Request) {
146		proxy.ServeHTTP(w, r)	175		proxy.ServeHTTP(w, r)
147		}	176		}
148			177
			178	+	// isRunningInContainer detects if Gordon is running inside a container
			179	+	func (s *Server) isRunningInContainer() bool {
			180	+	// Check for common container indicators
			181	+
			182	+	// 1. Check for /.dockerenv file (Docker creates this)
			183	+	if _, err := os.Stat("/.dockerenv"); err == nil {
			184	+	return true
			185	+	}
			186	+
			187	+	// 2. Check cgroup for container indicators
			188	+	if data, err := os.ReadFile("/proc/1/cgroup"); err == nil {
			189	+	content := string(data)
			190	+	if strings.Contains(content, "docker") \|\|
			191	+	strings.Contains(content, "containerd") \|\|
			192	+	strings.Contains(content, "podman") {
			193	+	return true
			194	+	}
			195	+	}
			196	+
			197	+	// 3. Check if hostname matches container ID pattern (64 hex chars)
			198	+	if hostname, err := os.Hostname(); err == nil {
			199	+	if len(hostname) == 12 \|\| len(hostname) == 64 {
			200	+	// Container hostnames are typically 12 or 64 character hex strings
			201	+	return true
			202	+	}
			203	+	}
			204	+
			205	+	// 4. Check environment variables set by container runtimes
			206	+	if os.Getenv("KUBERNETES_SERVICE_HOST") != "" \|\|
			207	+	os.Getenv("DOCKER_CONTAINER") != "" \|\|
			208	+	os.Getenv("container") != "" {
			209	+	return true
			210	+	}
			211	+
			212	+	return false
			213	+	}
			214	+
149		func (s *Server) Start(ctx context.Context) error {	215		func (s *Server) Start(ctx context.Context) error {
150		addr := ":" + strconv.Itoa(s.config.Server.Port)	216		addr := ":" + strconv.Itoa(s.config.Server.Port)
151			217

M internal/registry/server.go +171 -265


@@ -7,18 +7,18 @@ import (
7	7		"io"
8	8		"net/http"
9	9		"strconv"
10		-	"strings"
11	10
12		-	"github.com/rs/zerolog/log"
13	11		"gordon/internal/config"
14	12		"gordon/internal/events"
15	13		"gordon/internal/middleware"
	14	+
	15	+	"github.com/rs/zerolog/log"
16	16		)
17	17
18	18		type Server struct {
19		-	config *config.Config
20		-	mux *http.ServeMux
21		-	storage Storage
	19	+	config *config.Config
	20	+	mux *http.ServeMux
	21	+	storage Storage
22	22		eventBus events.EventBus
23	23		}
24	24

@@ -26,7 +26,7 @@ func NewServer(cfg config.Config, eventBus events.EventBus) (Server, error) {
26	26		// Initialize storage
27	27		registryPath := cfg.Server.DataDir + "/registry"
28	28		log.Debug().Str("data_dir", cfg.Server.DataDir).Str("registry_path", registryPath).Msg("Initializing registry storage")
29		-
	29	+
30	30		storage, err := NewFilesystemStorage(registryPath)
31	31		if err != nil {
32	32		return nil, fmt.Errorf("failed to initialize storage: %w", err)

@@ -44,188 +44,71 @@ func NewServer(cfg config.Config, eventBus events.EventBus) (Server, error) {
44	44
45	45		func (s *Server) setupRoutes() {
46	46		// Docker Registry API v2 endpoints using Go 1.22+ ServeMux patterns
47		-
48		-	// Registry info endpoint
49		-	s.mux.HandleFunc("GET /v2/", s.handleRegistryInfo)
50		-
51		-	// Since Go's ServeMux requires wildcards to be at the end, we need a different approach
52		-	// Use a catch-all handler for /v2/ paths and route internally
53		-	s.mux.HandleFunc("/v2/", s.handleV2Routes)
	47	+	s.mux.HandleFunc("GET /v2/", s.handleBase)
	48	+	s.mux.HandleFunc("HEAD /v2/{name}/manifests/{reference}", s.handleGetManifest)
	49	+	s.mux.HandleFunc("GET /v2/{name}/manifests/{reference}", s.handleGetManifest)
	50	+	s.mux.HandleFunc("PUT /v2/{name}/manifests/{reference}", s.handlePutManifest)
	51	+	s.mux.HandleFunc("HEAD /v2/{name}/blobs/{digest}", s.handleGetBlob)
	52	+	s.mux.HandleFunc("GET /v2/{name}/blobs/{digest}", s.handleGetBlob)
	53	+	s.mux.HandleFunc("POST /v2/{name}/blobs/uploads/", s.handleStartBlobUpload)
	54	+	s.mux.HandleFunc("PATCH /v2/{name}/blobs/uploads/{uuid}", s.handleBlobUpload)
	55	+	s.mux.HandleFunc("PUT /v2/{name}/blobs/uploads/{uuid}", s.handleBlobUpload)
	56	+	s.mux.HandleFunc("GET /v2/{name}/tags/list", s.handleListTags)
54	57		}
55	58
56		-	func (s Server) handleV2Routes(w http.ResponseWriter, r http.Request) {
57		-	path := r.URL.Path
58		-	method := r.Method
59		-
60		-	// Route based on path pattern and method
61		-	switch {
62		-	case path == "/v2/" && method == "GET":
63		-	s.handleRegistryInfo(w, r)
64		-	case method == "GET" && matches(path, "/v2/./manifests/."):
65		-	s.handleGetManifest(w, r)
66		-	case method == "PUT" && matches(path, "/v2/./manifests/."):
67		-	s.handlePutManifest(w, r)
68		-	case (method == "GET" \|\| method == "HEAD") && matches(path, "/v2/./blobs/."):
69		-	s.handleGetBlob(w, r)
70		-	case method == "POST" && matches(path, "/v2/.*/blobs/uploads/"):
71		-	s.handleStartBlobUpload(w, r)
72		-	case (method == "PUT" \|\| method == "PATCH") && matches(path, "/v2/./blobs/uploads/."):
73		-	s.handleBlobUpload(w, r)
74		-	case method == "GET" && matches(path, "/v2/.*/tags/list"):
75		-	s.handleListTags(w, r)
76		-	default:
77		-	http.NotFound(w, r)
78		-	}
	59	+	func (s Server) handleBase(w http.ResponseWriter, r http.Request) {
	60	+	// For Docker clients, we need to respond with a 200 OK to this endpoint
	61	+	// even if authentication is required. The client will then check the
	62	+	// WWW-Authenticate header.
	63	+	// see: https://docs.docker.com/registry/spec/api/#base
	64	+	//
	65	+	// Podman, as of v4.4.1, will not send credentials in the initial /v2/
	66	+	// request. It expects a 401 Unauthorized to then send credentials.
	67	+	//
	68	+	// To support both, we'll check for the "Docker-Client" user agent.
	69	+	// For now, we will just return a 200 OK to all clients.
	70	+	w.Header().Set("Docker-Distribution-API-Version", "registry/2.0")
	71	+	w.WriteHeader(http.StatusOK)
79	72		}
80	73
81		-	// Simple pattern matching helper
82		-	func matches(path, pattern string) bool {
83		-	// Convert pattern to regex-like matching
84		-	// For now, just check if the pattern structure matches
85		-	if pattern == "/v2/./manifests/." {
86		-	return len(path) > 4 && path[:4] == "/v2/" &&
87		-	len(path) > 14 && path[len(path)-10:] != "/manifests/" &&
88		-	contains(path, "/manifests/")
89		-	}
90		-	if pattern == "/v2/./blobs/." {
91		-	return len(path) > 4 && path[:4] == "/v2/" && contains(path, "/blobs/")
92		-	}
93		-	if pattern == "/v2/.*/blobs/uploads/" {
94		-	return len(path) > 4 && path[:4] == "/v2/" &&
95		-	contains(path, "/blobs/uploads/") && path[len(path)-1:] == "/"
96		-	}
97		-	if pattern == "/v2/./blobs/uploads/." {
98		-	return len(path) > 4 && path[:4] == "/v2/" && contains(path, "/blobs/uploads/")
99		-	}
100		-	if pattern == "/v2/.*/tags/list" {
101		-	return len(path) > 4 && path[:4] == "/v2/" &&
102		-	len(path) > 10 && path[len(path)-10:] == "/tags/list"
103		-	}
104		-	return false
105		-	}
	74	+	func (s Server) handleGetManifest(w http.ResponseWriter, r http.Request) {
	75	+	name := r.PathValue("name")
	76	+	reference := r.PathValue("reference")
106	77
107		-	func contains(s, substr string) bool {
108		-	for i := 0; i <= len(s)-len(substr); i++ {
109		-	if s[i:i+len(substr)] == substr {
110		-	return true
111		-	}
112		-	}
113		-	return false
114		-	}
	78	+	log.Debug().Str("name", name).Str("reference", reference).Msg("GET manifest")
115	79
116		-	// Path parsing helpers
117		-	func (s *Server) parseManifestPath(path string) (name, reference string) {
118		-	// Expected format: /v2/{name}/manifests/{reference}
119		-	parts := strings.Split(path, "/")
120		-	if len(parts) >= 5 && parts[3] == "manifests" {
121		-	name = parts[2]
122		-	reference = parts[4]
	80	+	manifest, contentType, err := s.storage.GetManifest(name, reference)
	81	+	if err != nil {
	82	+	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Manifest not found")
	83	+	http.NotFound(w, r)
	84	+	return
123	85		}
124		-	return
125		-	}
126	86
127		-	func (s *Server) parseBlobPath(path string) (name, digest string) {
128		-	// Expected format: /v2/{name}/blobs/{digest}
129		-	parts := strings.Split(path, "/")
130		-	if len(parts) >= 5 && parts[3] == "blobs" {
131		-	name = parts[2]
132		-	digest = parts[4]
133		-	}
134		-	return
135		-	}
	87	+	log.Debug().Str("name", name).Str("reference", reference).Str("content_type", contentType).Msg("GET manifest - serving")
136	88
137		-	func (s *Server) parseUploadPath(path string) (name, uuid string) {
138		-	// Expected format: /v2/{name}/blobs/uploads/{uuid}
139		-	parts := strings.Split(path, "/")
140		-	if len(parts) >= 6 && parts[3] == "blobs" && parts[4] == "uploads" {
141		-	name = parts[2]
142		-	if len(parts) > 5 {
143		-	uuid = parts[5]
144		-	}
145		-	}
146		-	return
147		-	}
	89	+	// Return the appropriate content type based on what was stored
	90	+	w.Header().Set("Content-Type", contentType)
	91	+	w.Header().Set("Content-Length", strconv.Itoa(len(manifest)))
	92	+	w.WriteHeader(http.StatusOK)
148	93
149		-	func (s *Server) parseTagsPath(path string) string {
150		-	// Expected format: /v2/{name}/tags/list
151		-	parts := strings.Split(path, "/")
152		-	if len(parts) >= 4 && parts[3] == "tags" {
153		-	return parts[2]
	94	+	if r.Method == "GET" {
	95	+	_, _ = w.Write(manifest)
154	96		}
155		-	return ""
156	97		}
157	98
158		-	func (s *Server) Start(ctx context.Context) error {
159		-	addr := ":" + strconv.Itoa(s.config.Server.RegistryPort)
160		-
161		-	// Build middleware chain
162		-	middlewares := []func(http.Handler) http.Handler{
163		-	middleware.PanicRecovery,
164		-	middleware.RequestLogger,
165		-	}
166		-
167		-	// Add registry authentication if enabled
168		-	if s.config.RegistryAuth.Enabled {
169		-	middlewares = append(middlewares, middleware.RegistryAuth(s.config.RegistryAuth.Username, s.config.RegistryAuth.Password))
170		-	}
171		-
172		-	// Wrap the mux with middleware
173		-	handler := middleware.Chain(middlewares...)(s.mux)
174		-
175		-	server := &http.Server{
176		-	Addr: addr,
177		-	Handler: handler,
178		-	}
179		-
180		-	log.Info().Str("address", addr).Msg("Registry server starting")
181		-
182		-	// Start server in a goroutine
183		-	errChan := make(chan error, 1)
184		-	go func() {
185		-	if err := server.ListenAndServe(); err != nil && err != http.ErrServerClosed {
186		-	errChan <- err
187		-	}
188		-	}()
189		-
190		-	// Wait for context cancellation or server error
191		-	select {
192		-	case err := <-errChan:
193		-	return err
194		-	case <-ctx.Done():
195		-	log.Info().Msg("Registry server shutting down...")
196		-	return server.Shutdown(context.Background())
197		-	}
198		-	}
	99	+	func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
	100	+	name := r.PathValue("name")
	101	+	reference := r.PathValue("reference")
199	102
200		-	func (s Server) handleRegistryInfo(w http.ResponseWriter, r http.Request) {
201		-	w.Header().Set("Content-Type", "application/json")
202		-	w.WriteHeader(http.StatusOK)
203		-	fmt.Fprint(w, `{"name":"gordon-registry"}`)
204		-	}
	103	+	contentType := r.Header.Get("Content-Type")
	104	+	log.Debug().Str("name", name).Str("reference", reference).Str("content_type", contentType).Msg("PUT manifest")
205	105
206		-	func (s Server) handleGetManifest(w http.ResponseWriter, r http.Request) {
207		-	name, reference := s.parseManifestPath(r.URL.Path)
208		-
209		-	log.Debug().Str("name", name).Str("reference", reference).Msg("GET manifest")
210		-
211		-	manifest, err := s.storage.GetManifest(name, reference)
212		-	if err != nil {
213		-	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Manifest not found")
214		-	w.WriteHeader(http.StatusNotFound)
	106	+	if contentType == "" {
	107	+	log.Warn().Str("name", name).Str("reference", reference).Msg("Content-Type header missing for PUT manifest")
	108	+	http.Error(w, "Content-Type header required", http.StatusBadRequest)
215	109		return
216	110		}
217		-
218		-	w.Header().Set("Content-Type", "application/vnd.docker.distribution.manifest.v2+json")
219		-	w.Header().Set("Content-Length", strconv.Itoa(len(manifest)))
220		-	w.WriteHeader(http.StatusOK)
221		-	w.Write(manifest)
222		-	}
223	111
224		-	func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
225		-	name, reference := s.parseManifestPath(r.URL.Path)
226		-
227		-	log.Debug().Str("name", name).Str("reference", reference).Msg("PUT manifest")
228		-
229	112		// Read manifest data
230	113		data, err := io.ReadAll(r.Body)
231	114		if err != nil {

@@ -233,151 +116,174 @@ func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
233	116		http.Error(w, "Bad Request", http.StatusBadRequest)
234	117		return
235	118		}
236		-
	119	+
237	120		// Store manifest
238		-	if err := s.storage.PutManifest(name, reference, data); err != nil {
	121	+	if err := s.storage.PutManifest(name, reference, contentType, data); err != nil {
239	122		log.Error().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to store manifest")
240	123		http.Error(w, "Internal Server Error", http.StatusInternalServerError)
241	124		return
242	125		}
243		-
	126	+
244	127		// Calculate digest for Docker-Content-Digest header
245	128		digest := fmt.Sprintf("sha256:%x", sha256.Sum256(data))
246	129		w.Header().Set("Docker-Content-Digest", digest)
247	130		w.Header().Set("Location", fmt.Sprintf("/v2/%s/manifests/%s", name, reference))
248	131		w.WriteHeader(http.StatusCreated)
249		-
	132	+
250	133		// Emit image pushed event
251		-	event := events.Event{
252		-	Type: events.ImagePushed,
253		-	ImageName: name,
254		-	Tag: reference,
255		-	Data: map[string]interface{}{
256		-	"digest": digest,
257		-	"size": len(data),
258		-	},
259		-	}
260		-
261		-	if err := s.eventBus.Publish(event); err != nil {
262		-	log.Error().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to publish image pushed event")
263		-	} else {
264		-	log.Info().Str("name", name).Str("reference", reference).Msg("Image pushed event published")
265		-	}
	134	+	s.eventBus.Publish(events.ImagePushed, events.ImagePushedPayload{
	135	+	Name: name,
	136	+	Reference: reference,
	137	+	Manifest: data,
	138	+	})
266	139		}
267	140
268	141		func (s Server) handleGetBlob(w http.ResponseWriter, r http.Request) {
269		-	name, digest := s.parseBlobPath(r.URL.Path)
270		-
271		-	log.Debug().Str("name", name).Str("digest", digest).Str("method", r.Method).Msg("Handle blob request")
272		-
273		-	reader, err := s.storage.GetBlob(digest)
	142	+	name := r.PathValue("name")
	143	+	digest := r.PathValue("digest")
	144	+
	145	+	log.Debug().Str("name", name).Str("digest", digest).Msg("GET blob")
	146	+
	147	+	path, err := s.storage.GetBlobPath(digest)
274	148		if err != nil {
275		-	log.Warn().Err(err).Str("digest", digest).Msg("Blob not found")
276		-	w.WriteHeader(http.StatusNotFound)
277		-	return
278		-	}
279		-	defer reader.Close()
280		-
281		-	w.Header().Set("Content-Type", "application/octet-stream")
282		-	w.WriteHeader(http.StatusOK)
283		-
284		-	// For HEAD requests, don't send the body
285		-	if r.Method == "HEAD" {
	149	+	log.Warn().Err(err).Str("name", name).Str("digest", digest).Msg("Blob not found")
	150	+	http.NotFound(w, r)
286	151		return
287	152		}
288		-
289		-	_, err = io.Copy(w, reader)
290		-	if err != nil {
291		-	log.Error().Err(err).Str("digest", digest).Msg("Failed to serve blob")
292		-	}
	153	+
	154	+	// Serve the file directly
	155	+	http.ServeFile(w, r, path)
293	156		}
294	157
295	158		func (s Server) handleStartBlobUpload(w http.ResponseWriter, r http.Request) {
296		-	name, _ := s.parseUploadPath(r.URL.Path)
297		-
298		-	log.Debug().Str("name", name).Msg("Start blob upload")
299		-
	159	+	name := r.PathValue("name")
	160	+
	161	+	log.Debug().Str("name", name).Msg("Starting blob upload")
	162	+
	163	+	// Generate a unique upload ID
300	164		uuid, err := s.storage.StartBlobUpload(name)
301	165		if err != nil {
302		-	log.Error().Err(err).Str("name", name).Msg("Failed to start blob upload")
	166	+	log.Error().Err(err).Msg("Failed to start blob upload")
303	167		http.Error(w, "Internal Server Error", http.StatusInternalServerError)
304	168		return
305	169		}
306		-
	170	+
	171	+	// Set headers for the response
307	172		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))
308		-	w.Header().Set("Range", "0-0")
	173	+	w.Header().Set("Range", "0-0") // Indicate that no bytes have been received yet
309	174		w.WriteHeader(http.StatusAccepted)
310	175		}
311	176
312	177		func (s Server) handleBlobUpload(w http.ResponseWriter, r http.Request) {
313		-	name, uuid := s.parseUploadPath(r.URL.Path)
314		-
315		-	log.Debug().Str("name", name).Str("uuid", uuid).Msg("Blob upload")
316		-
317		-	// Check if this is a completion request (has digest parameter)
	178	+	name := r.PathValue("name")
	179	+	uuid := r.PathValue("uuid")
318	180		digest := r.URL.Query().Get("digest")
319		-
	181	+
	182	+	log.Debug().
	183	+	Str("name", name).
	184	+	Str("uuid", uuid).
	185	+	Str("digest", digest).
	186	+	Str("method", r.Method).
	187	+	Msg("Handling blob upload chunk")
	188	+
	189	+	// Read the chunk from the request body
	190	+	chunk, err := io.ReadAll(r.Body)
	191	+	if err != nil {
	192	+	log.Error().Err(err).Msg("Failed to read blob chunk")
	193	+	http.Error(w, "Bad Request", http.StatusBadRequest)
	194	+	return
	195	+	}
	196	+
	197	+	// Append the chunk to the upload
	198	+	length, err := s.storage.AppendBlobChunk(name, uuid, chunk)
	199	+	if err != nil {
	200	+	log.Error().Err(err).Msg("Failed to append blob chunk")
	201	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
	202	+	return
	203	+	}
	204	+
	205	+	// If this is the final chunk (PUT request with digest), finalize the upload
320	206		if r.Method == "PUT" && digest != "" {
321		-	// This is a completion request
322	207		if err := s.storage.FinishBlobUpload(uuid, digest); err != nil {
323		-	log.Error().Err(err).Str("uuid", uuid).Str("digest", digest).Msg("Failed to finish blob upload")
324		-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
	208	+	log.Error().Err(err).Str("digest", digest).Msg("Failed to finalize blob upload")
	209	+	// The upload is invalid, delete it
	210	+	_ = s.storage.CancelBlobUpload(uuid)
	211	+	http.Error(w, "Bad Request: digest mismatch", http.StatusBadRequest)
325	212		return
326	213		}
327		-
	214	+
	215	+	// Respond with 201 Created
328	216		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/%s", name, digest))
329	217		w.Header().Set("Docker-Content-Digest", digest)
330	218		w.WriteHeader(http.StatusCreated)
331		-
332		-	log.Info().Str("name", name).Str("uuid", uuid).Str("digest", digest).Msg("Blob upload completed")
333	219		return
334	220		}
335		-
336		-	// This is a chunk upload (PATCH)
337		-	writer, err := s.storage.GetBlobUpload(uuid)
338		-	if err != nil {
339		-	log.Error().Err(err).Str("uuid", uuid).Msg("Failed to get blob upload")
340		-	http.Error(w, "Not Found", http.StatusNotFound)
341		-	return
342		-	}
343		-	defer writer.Close()
344		-
345		-	written, err := io.Copy(writer, r.Body)
346		-	if err != nil {
347		-	log.Error().Err(err).Str("uuid", uuid).Msg("Failed to write blob chunk")
348		-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
349		-	return
350		-	}
351		-
	221	+
	222	+	// For PATCH requests, respond with 202 Accepted
352	223		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))
353		-	w.Header().Set("Range", fmt.Sprintf("0-%d", written-1))
	224	+	w.Header().Set("Range", fmt.Sprintf("0-%d", length-1))
354	225		w.WriteHeader(http.StatusAccepted)
355		-
356		-	log.Debug().Str("uuid", uuid).Int64("bytes", written).Msg("Blob chunk uploaded")
357	226		}
358	227
359	228		func (s Server) handleListTags(w http.ResponseWriter, r http.Request) {
360		-	name := s.parseTagsPath(r.URL.Path)
361		-
362		-	log.Debug().Str("name", name).Msg("List tags")
363		-
	229	+	name := r.PathValue("name")
	230	+
	231	+	log.Debug().Str("name", name).Msg("Listing tags")
	232	+
364	233		tags, err := s.storage.ListTags(name)
365	234		if err != nil {
366		-	log.Error().Err(err).Str("name", name).Msg("Failed to list tags")
367		-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
	235	+	log.Warn().Err(err).Str("name", name).Msg("Tag not found")
	236	+	w.WriteHeader(http.StatusNotFound)
368	237		return
369	238		}
370		-
	239	+
	240	+	// Construct the response JSON
	241	+	response := fmt.Sprintf(`{"name":"%s","tags":%s}`, name, tags)
	242	+
371	243		w.Header().Set("Content-Type", "application/json")
	244	+	w.Header().Set("Content-Length", strconv.Itoa(len(response)))
372	245		w.WriteHeader(http.StatusOK)
373		-
374		-	// Build JSON response
375		-	fmt.Fprintf(w, `{"name":"%s","tags":[`, name)
376		-	for i, tag := range tags {
377		-	if i > 0 {
378		-	fmt.Fprint(w, ",")
	246	+	w.Write([]byte(response))
	247	+	}
	248	+
	249	+	func (s *Server) Start(ctx context.Context) error {
	250	+	addr := ":" + strconv.Itoa(s.config.Server.RegistryPort)
	251	+
	252	+	// Build middleware chain
	253	+	middlewares := []func(http.Handler) http.Handler{
	254	+	middleware.PanicRecovery,
	255	+	middleware.RequestLogger,
	256	+	}
	257	+
	258	+	// Add registry authentication if enabled
	259	+	if s.config.RegistryAuth.Enabled {
	260	+	middlewares = append(middlewares, middleware.RegistryAuth(s.config.RegistryAuth.Username, s.config.RegistryAuth.Password))
	261	+	}
	262	+
	263	+	// Wrap the mux with middleware
	264	+	handler := middleware.Chain(middlewares...)(s.mux)
	265	+
	266	+	server := &http.Server{
	267	+	Addr: addr,
	268	+	Handler: handler,
	269	+	}
	270	+
	271	+	log.Info().Str("address", addr).Msg("Registry server starting")
	272	+
	273	+	// Start server in a goroutine
	274	+	errChan := make(chan error, 1)
	275	+	go func() {
	276	+	if err := server.ListenAndServe(); err != nil && err != http.ErrServerClosed {
	277	+	errChan <- err
379	278		}
380		-	fmt.Fprintf(w, `"%s"`, tag)
	279	+	}()
	280	+
	281	+	// Wait for context cancellation or server error
	282	+	select {
	283	+	case err := <-errChan:
	284	+	return err
	285	+	case <-ctx.Done():
	286	+	log.Info().Msg("Registry server shutting down...")
	287	+	return server.Shutdown(context.Background())
381	288		}
382		-	fmt.Fprint(w, "]}")
383		-	}⏎̸
	289	+	}


@@ -7,18 +7,18 @@ import (
7		"io"	7		"io"
8		"net/http"	8		"net/http"
9		"strconv"	9		"strconv"
10	-	"strings"
11			10
12	-	"github.com/rs/zerolog/log"
13		"gordon/internal/config"	11		"gordon/internal/config"
14		"gordon/internal/events"	12		"gordon/internal/events"
15		"gordon/internal/middleware"	13		"gordon/internal/middleware"
			14	+
			15	+	"github.com/rs/zerolog/log"
16		)	16		)
17			17
18		type Server struct {	18		type Server struct {
19	-	config *config.Config	19	+	config *config.Config
20	-	mux *http.ServeMux	20	+	mux *http.ServeMux
21	-	storage Storage	21	+	storage Storage
22		eventBus events.EventBus	22		eventBus events.EventBus
23		}	23		}
24			24

@@ -26,7 +26,7 @@ func NewServer(cfg config.Config, eventBus events.EventBus) (Server, error) {
26		// Initialize storage	26		// Initialize storage
27		registryPath := cfg.Server.DataDir + "/registry"	27		registryPath := cfg.Server.DataDir + "/registry"
28		log.Debug().Str("data_dir", cfg.Server.DataDir).Str("registry_path", registryPath).Msg("Initializing registry storage")	28		log.Debug().Str("data_dir", cfg.Server.DataDir).Str("registry_path", registryPath).Msg("Initializing registry storage")
29	-		29	+
30		storage, err := NewFilesystemStorage(registryPath)	30		storage, err := NewFilesystemStorage(registryPath)
31		if err != nil {	31		if err != nil {
32		return nil, fmt.Errorf("failed to initialize storage: %w", err)	32		return nil, fmt.Errorf("failed to initialize storage: %w", err)

@@ -44,188 +44,71 @@ func NewServer(cfg config.Config, eventBus events.EventBus) (Server, error) {
44			44
45		func (s *Server) setupRoutes() {	45		func (s *Server) setupRoutes() {
46		// Docker Registry API v2 endpoints using Go 1.22+ ServeMux patterns	46		// Docker Registry API v2 endpoints using Go 1.22+ ServeMux patterns
47	-		47	+	s.mux.HandleFunc("GET /v2/", s.handleBase)
48	-	// Registry info endpoint	48	+	s.mux.HandleFunc("HEAD /v2/{name}/manifests/{reference}", s.handleGetManifest)
49	-	s.mux.HandleFunc("GET /v2/", s.handleRegistryInfo)	49	+	s.mux.HandleFunc("GET /v2/{name}/manifests/{reference}", s.handleGetManifest)
50	-		50	+	s.mux.HandleFunc("PUT /v2/{name}/manifests/{reference}", s.handlePutManifest)
51	-	// Since Go's ServeMux requires wildcards to be at the end, we need a different approach	51	+	s.mux.HandleFunc("HEAD /v2/{name}/blobs/{digest}", s.handleGetBlob)
52	-	// Use a catch-all handler for /v2/ paths and route internally	52	+	s.mux.HandleFunc("GET /v2/{name}/blobs/{digest}", s.handleGetBlob)
53	-	s.mux.HandleFunc("/v2/", s.handleV2Routes)	53	+	s.mux.HandleFunc("POST /v2/{name}/blobs/uploads/", s.handleStartBlobUpload)
			54	+	s.mux.HandleFunc("PATCH /v2/{name}/blobs/uploads/{uuid}", s.handleBlobUpload)
			55	+	s.mux.HandleFunc("PUT /v2/{name}/blobs/uploads/{uuid}", s.handleBlobUpload)
			56	+	s.mux.HandleFunc("GET /v2/{name}/tags/list", s.handleListTags)
54		}	57		}
55			58
56	-	func (s Server) handleV2Routes(w http.ResponseWriter, r http.Request) {	59	+	func (s Server) handleBase(w http.ResponseWriter, r http.Request) {
57	-	path := r.URL.Path	60	+	// For Docker clients, we need to respond with a 200 OK to this endpoint
58	-	method := r.Method	61	+	// even if authentication is required. The client will then check the
59	-		62	+	// WWW-Authenticate header.
60	-	// Route based on path pattern and method	63	+	// see: https://docs.docker.com/registry/spec/api/#base
61	-	switch {	64	+	//
62	-	case path == "/v2/" && method == "GET":	65	+	// Podman, as of v4.4.1, will not send credentials in the initial /v2/
63	-	s.handleRegistryInfo(w, r)	66	+	// request. It expects a 401 Unauthorized to then send credentials.
64	-	case method == "GET" && matches(path, "/v2/./manifests/."):	67	+	//
65	-	s.handleGetManifest(w, r)	68	+	// To support both, we'll check for the "Docker-Client" user agent.
66	-	case method == "PUT" && matches(path, "/v2/./manifests/."):	69	+	// For now, we will just return a 200 OK to all clients.
67	-	s.handlePutManifest(w, r)	70	+	w.Header().Set("Docker-Distribution-API-Version", "registry/2.0")
68	-	case (method == "GET" \|\| method == "HEAD") && matches(path, "/v2/./blobs/."):	71	+	w.WriteHeader(http.StatusOK)
69	-	s.handleGetBlob(w, r)
70	-	case method == "POST" && matches(path, "/v2/.*/blobs/uploads/"):
71	-	s.handleStartBlobUpload(w, r)
72	-	case (method == "PUT" \|\| method == "PATCH") && matches(path, "/v2/./blobs/uploads/."):
73	-	s.handleBlobUpload(w, r)
74	-	case method == "GET" && matches(path, "/v2/.*/tags/list"):
75	-	s.handleListTags(w, r)
76	-	default:
77	-	http.NotFound(w, r)
78	-	}
79		}	72		}
80			73
81	-	// Simple pattern matching helper	74	+	func (s Server) handleGetManifest(w http.ResponseWriter, r http.Request) {
82	-	func matches(path, pattern string) bool {	75	+	name := r.PathValue("name")
83	-	// Convert pattern to regex-like matching	76	+	reference := r.PathValue("reference")
84	-	// For now, just check if the pattern structure matches
85	-	if pattern == "/v2/./manifests/." {
86	-	return len(path) > 4 && path[:4] == "/v2/" &&
87	-	len(path) > 14 && path[len(path)-10:] != "/manifests/" &&
88	-	contains(path, "/manifests/")
89	-	}
90	-	if pattern == "/v2/./blobs/." {
91	-	return len(path) > 4 && path[:4] == "/v2/" && contains(path, "/blobs/")
92	-	}
93	-	if pattern == "/v2/.*/blobs/uploads/" {
94	-	return len(path) > 4 && path[:4] == "/v2/" &&
95	-	contains(path, "/blobs/uploads/") && path[len(path)-1:] == "/"
96	-	}
97	-	if pattern == "/v2/./blobs/uploads/." {
98	-	return len(path) > 4 && path[:4] == "/v2/" && contains(path, "/blobs/uploads/")
99	-	}
100	-	if pattern == "/v2/.*/tags/list" {
101	-	return len(path) > 4 && path[:4] == "/v2/" &&
102	-	len(path) > 10 && path[len(path)-10:] == "/tags/list"
103	-	}
104	-	return false
105	-	}
106			77
107	-	func contains(s, substr string) bool {	78	+	log.Debug().Str("name", name).Str("reference", reference).Msg("GET manifest")
108	-	for i := 0; i <= len(s)-len(substr); i++ {
109	-	if s[i:i+len(substr)] == substr {
110	-	return true
111	-	}
112	-	}
113	-	return false
114	-	}
115			79
116	-	// Path parsing helpers	80	+	manifest, contentType, err := s.storage.GetManifest(name, reference)
117	-	func (s *Server) parseManifestPath(path string) (name, reference string) {	81	+	if err != nil {
118	-	// Expected format: /v2/{name}/manifests/{reference}	82	+	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Manifest not found")
119	-	parts := strings.Split(path, "/")	83	+	http.NotFound(w, r)
120	-	if len(parts) >= 5 && parts[3] == "manifests" {	84	+	return
121	-	name = parts[2]
122	-	reference = parts[4]
123		}	85		}
124	-	return
125	-	}
126			86
127	-	func (s *Server) parseBlobPath(path string) (name, digest string) {	87	+	log.Debug().Str("name", name).Str("reference", reference).Str("content_type", contentType).Msg("GET manifest - serving")
128	-	// Expected format: /v2/{name}/blobs/{digest}
129	-	parts := strings.Split(path, "/")
130	-	if len(parts) >= 5 && parts[3] == "blobs" {
131	-	name = parts[2]
132	-	digest = parts[4]
133	-	}
134	-	return
135	-	}
136			88
137	-	func (s *Server) parseUploadPath(path string) (name, uuid string) {	89	+	// Return the appropriate content type based on what was stored
138	-	// Expected format: /v2/{name}/blobs/uploads/{uuid}	90	+	w.Header().Set("Content-Type", contentType)
139	-	parts := strings.Split(path, "/")	91	+	w.Header().Set("Content-Length", strconv.Itoa(len(manifest)))
140	-	if len(parts) >= 6 && parts[3] == "blobs" && parts[4] == "uploads" {	92	+	w.WriteHeader(http.StatusOK)
141	-	name = parts[2]
142	-	if len(parts) > 5 {
143	-	uuid = parts[5]
144	-	}
145	-	}
146	-	return
147	-	}
148			93
149	-	func (s *Server) parseTagsPath(path string) string {	94	+	if r.Method == "GET" {
150	-	// Expected format: /v2/{name}/tags/list	95	+	_, _ = w.Write(manifest)
151	-	parts := strings.Split(path, "/")
152	-	if len(parts) >= 4 && parts[3] == "tags" {
153	-	return parts[2]
154		}	96		}
155	-	return ""
156		}	97		}
157			98
158	-	func (s *Server) Start(ctx context.Context) error {	99	+	func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
159	-	addr := ":" + strconv.Itoa(s.config.Server.RegistryPort)	100	+	name := r.PathValue("name")
160	-		101	+	reference := r.PathValue("reference")
161	-	// Build middleware chain
162	-	middlewares := []func(http.Handler) http.Handler{
163	-	middleware.PanicRecovery,
164	-	middleware.RequestLogger,
165	-	}
166	-
167	-	// Add registry authentication if enabled
168	-	if s.config.RegistryAuth.Enabled {
169	-	middlewares = append(middlewares, middleware.RegistryAuth(s.config.RegistryAuth.Username, s.config.RegistryAuth.Password))
170	-	}
171	-
172	-	// Wrap the mux with middleware
173	-	handler := middleware.Chain(middlewares...)(s.mux)
174	-
175	-	server := &http.Server{
176	-	Addr: addr,
177	-	Handler: handler,
178	-	}
179	-
180	-	log.Info().Str("address", addr).Msg("Registry server starting")
181	-
182	-	// Start server in a goroutine
183	-	errChan := make(chan error, 1)
184	-	go func() {
185	-	if err := server.ListenAndServe(); err != nil && err != http.ErrServerClosed {
186	-	errChan <- err
187	-	}
188	-	}()
189	-
190	-	// Wait for context cancellation or server error
191	-	select {
192	-	case err := <-errChan:
193	-	return err
194	-	case <-ctx.Done():
195	-	log.Info().Msg("Registry server shutting down...")
196	-	return server.Shutdown(context.Background())
197	-	}
198	-	}
199			102
200	-	func (s Server) handleRegistryInfo(w http.ResponseWriter, r http.Request) {	103	+	contentType := r.Header.Get("Content-Type")
201	-	w.Header().Set("Content-Type", "application/json")	104	+	log.Debug().Str("name", name).Str("reference", reference).Str("content_type", contentType).Msg("PUT manifest")
202	-	w.WriteHeader(http.StatusOK)
203	-	fmt.Fprint(w, `{"name":"gordon-registry"}`)
204	-	}
205			105
206	-	func (s Server) handleGetManifest(w http.ResponseWriter, r http.Request) {	106	+	if contentType == "" {
207	-	name, reference := s.parseManifestPath(r.URL.Path)	107	+	log.Warn().Str("name", name).Str("reference", reference).Msg("Content-Type header missing for PUT manifest")
208	-		108	+	http.Error(w, "Content-Type header required", http.StatusBadRequest)
209	-	log.Debug().Str("name", name).Str("reference", reference).Msg("GET manifest")
210	-
211	-	manifest, err := s.storage.GetManifest(name, reference)
212	-	if err != nil {
213	-	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Manifest not found")
214	-	w.WriteHeader(http.StatusNotFound)
215		return	109		return
216		}	110		}
217	-
218	-	w.Header().Set("Content-Type", "application/vnd.docker.distribution.manifest.v2+json")
219	-	w.Header().Set("Content-Length", strconv.Itoa(len(manifest)))
220	-	w.WriteHeader(http.StatusOK)
221	-	w.Write(manifest)
222	-	}
223			111
224	-	func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
225	-	name, reference := s.parseManifestPath(r.URL.Path)
226	-
227	-	log.Debug().Str("name", name).Str("reference", reference).Msg("PUT manifest")
228	-
229		// Read manifest data	112		// Read manifest data
230		data, err := io.ReadAll(r.Body)	113		data, err := io.ReadAll(r.Body)
231		if err != nil {	114		if err != nil {

@@ -233,151 +116,174 @@ func (s Server) handlePutManifest(w http.ResponseWriter, r http.Request) {
233		http.Error(w, "Bad Request", http.StatusBadRequest)	116		http.Error(w, "Bad Request", http.StatusBadRequest)
234		return	117		return
235		}	118		}
236	-		119	+
237		// Store manifest	120		// Store manifest
238	-	if err := s.storage.PutManifest(name, reference, data); err != nil {	121	+	if err := s.storage.PutManifest(name, reference, contentType, data); err != nil {
239		log.Error().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to store manifest")	122		log.Error().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to store manifest")
240		http.Error(w, "Internal Server Error", http.StatusInternalServerError)	123		http.Error(w, "Internal Server Error", http.StatusInternalServerError)
241		return	124		return
242		}	125		}
243	-		126	+
244		// Calculate digest for Docker-Content-Digest header	127		// Calculate digest for Docker-Content-Digest header
245		digest := fmt.Sprintf("sha256:%x", sha256.Sum256(data))	128		digest := fmt.Sprintf("sha256:%x", sha256.Sum256(data))
246		w.Header().Set("Docker-Content-Digest", digest)	129		w.Header().Set("Docker-Content-Digest", digest)
247		w.Header().Set("Location", fmt.Sprintf("/v2/%s/manifests/%s", name, reference))	130		w.Header().Set("Location", fmt.Sprintf("/v2/%s/manifests/%s", name, reference))
248		w.WriteHeader(http.StatusCreated)	131		w.WriteHeader(http.StatusCreated)
249	-		132	+
250		// Emit image pushed event	133		// Emit image pushed event
251	-	event := events.Event{	134	+	s.eventBus.Publish(events.ImagePushed, events.ImagePushedPayload{
252	-	Type: events.ImagePushed,	135	+	Name: name,
253	-	ImageName: name,	136	+	Reference: reference,
254	-	Tag: reference,	137	+	Manifest: data,
255	-	Data: map[string]interface{}{	138	+	})
256	-	"digest": digest,
257	-	"size": len(data),
258	-	},
259	-	}
260	-
261	-	if err := s.eventBus.Publish(event); err != nil {
262	-	log.Error().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to publish image pushed event")
263	-	} else {
264	-	log.Info().Str("name", name).Str("reference", reference).Msg("Image pushed event published")
265	-	}
266		}	139		}
267			140
268		func (s Server) handleGetBlob(w http.ResponseWriter, r http.Request) {	141		func (s Server) handleGetBlob(w http.ResponseWriter, r http.Request) {
269	-	name, digest := s.parseBlobPath(r.URL.Path)	142	+	name := r.PathValue("name")
270	-		143	+	digest := r.PathValue("digest")
271	-	log.Debug().Str("name", name).Str("digest", digest).Str("method", r.Method).Msg("Handle blob request")	144	+
272	-		145	+	log.Debug().Str("name", name).Str("digest", digest).Msg("GET blob")
273	-	reader, err := s.storage.GetBlob(digest)	146	+
			147	+	path, err := s.storage.GetBlobPath(digest)
274		if err != nil {	148		if err != nil {
275	-	log.Warn().Err(err).Str("digest", digest).Msg("Blob not found")	149	+	log.Warn().Err(err).Str("name", name).Str("digest", digest).Msg("Blob not found")
276	-	w.WriteHeader(http.StatusNotFound)	150	+	http.NotFound(w, r)
277	-	return
278	-	}
279	-	defer reader.Close()
280	-
281	-	w.Header().Set("Content-Type", "application/octet-stream")
282	-	w.WriteHeader(http.StatusOK)
283	-
284	-	// For HEAD requests, don't send the body
285	-	if r.Method == "HEAD" {
286		return	151		return
287		}	152		}
288	-		153	+
289	-	_, err = io.Copy(w, reader)	154	+	// Serve the file directly
290	-	if err != nil {	155	+	http.ServeFile(w, r, path)
291	-	log.Error().Err(err).Str("digest", digest).Msg("Failed to serve blob")
292	-	}
293		}	156		}
294			157
295		func (s Server) handleStartBlobUpload(w http.ResponseWriter, r http.Request) {	158		func (s Server) handleStartBlobUpload(w http.ResponseWriter, r http.Request) {
296	-	name, _ := s.parseUploadPath(r.URL.Path)	159	+	name := r.PathValue("name")
297	-		160	+
298	-	log.Debug().Str("name", name).Msg("Start blob upload")	161	+	log.Debug().Str("name", name).Msg("Starting blob upload")
299	-		162	+
			163	+	// Generate a unique upload ID
300		uuid, err := s.storage.StartBlobUpload(name)	164		uuid, err := s.storage.StartBlobUpload(name)
301		if err != nil {	165		if err != nil {
302	-	log.Error().Err(err).Str("name", name).Msg("Failed to start blob upload")	166	+	log.Error().Err(err).Msg("Failed to start blob upload")
303		http.Error(w, "Internal Server Error", http.StatusInternalServerError)	167		http.Error(w, "Internal Server Error", http.StatusInternalServerError)
304		return	168		return
305		}	169		}
306	-		170	+
			171	+	// Set headers for the response
307		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))	172		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))
308	-	w.Header().Set("Range", "0-0")	173	+	w.Header().Set("Range", "0-0") // Indicate that no bytes have been received yet
309		w.WriteHeader(http.StatusAccepted)	174		w.WriteHeader(http.StatusAccepted)
310		}	175		}
311			176
312		func (s Server) handleBlobUpload(w http.ResponseWriter, r http.Request) {	177		func (s Server) handleBlobUpload(w http.ResponseWriter, r http.Request) {
313	-	name, uuid := s.parseUploadPath(r.URL.Path)	178	+	name := r.PathValue("name")
314	-		179	+	uuid := r.PathValue("uuid")
315	-	log.Debug().Str("name", name).Str("uuid", uuid).Msg("Blob upload")
316	-
317	-	// Check if this is a completion request (has digest parameter)
318		digest := r.URL.Query().Get("digest")	180		digest := r.URL.Query().Get("digest")
319	-		181	+
			182	+	log.Debug().
			183	+	Str("name", name).
			184	+	Str("uuid", uuid).
			185	+	Str("digest", digest).
			186	+	Str("method", r.Method).
			187	+	Msg("Handling blob upload chunk")
			188	+
			189	+	// Read the chunk from the request body
			190	+	chunk, err := io.ReadAll(r.Body)
			191	+	if err != nil {
			192	+	log.Error().Err(err).Msg("Failed to read blob chunk")
			193	+	http.Error(w, "Bad Request", http.StatusBadRequest)
			194	+	return
			195	+	}
			196	+
			197	+	// Append the chunk to the upload
			198	+	length, err := s.storage.AppendBlobChunk(name, uuid, chunk)
			199	+	if err != nil {
			200	+	log.Error().Err(err).Msg("Failed to append blob chunk")
			201	+	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
			202	+	return
			203	+	}
			204	+
			205	+	// If this is the final chunk (PUT request with digest), finalize the upload
320		if r.Method == "PUT" && digest != "" {	206		if r.Method == "PUT" && digest != "" {
321	-	// This is a completion request
322		if err := s.storage.FinishBlobUpload(uuid, digest); err != nil {	207		if err := s.storage.FinishBlobUpload(uuid, digest); err != nil {
323	-	log.Error().Err(err).Str("uuid", uuid).Str("digest", digest).Msg("Failed to finish blob upload")	208	+	log.Error().Err(err).Str("digest", digest).Msg("Failed to finalize blob upload")
324	-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)	209	+	// The upload is invalid, delete it
			210	+	_ = s.storage.CancelBlobUpload(uuid)
			211	+	http.Error(w, "Bad Request: digest mismatch", http.StatusBadRequest)
325		return	212		return
326		}	213		}
327	-		214	+
			215	+	// Respond with 201 Created
328		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/%s", name, digest))	216		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/%s", name, digest))
329		w.Header().Set("Docker-Content-Digest", digest)	217		w.Header().Set("Docker-Content-Digest", digest)
330		w.WriteHeader(http.StatusCreated)	218		w.WriteHeader(http.StatusCreated)
331	-
332	-	log.Info().Str("name", name).Str("uuid", uuid).Str("digest", digest).Msg("Blob upload completed")
333		return	219		return
334		}	220		}
335	-		221	+
336	-	// This is a chunk upload (PATCH)	222	+	// For PATCH requests, respond with 202 Accepted
337	-	writer, err := s.storage.GetBlobUpload(uuid)
338	-	if err != nil {
339	-	log.Error().Err(err).Str("uuid", uuid).Msg("Failed to get blob upload")
340	-	http.Error(w, "Not Found", http.StatusNotFound)
341	-	return
342	-	}
343	-	defer writer.Close()
344	-
345	-	written, err := io.Copy(writer, r.Body)
346	-	if err != nil {
347	-	log.Error().Err(err).Str("uuid", uuid).Msg("Failed to write blob chunk")
348	-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)
349	-	return
350	-	}
351	-
352		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))	223		w.Header().Set("Location", fmt.Sprintf("/v2/%s/blobs/uploads/%s", name, uuid))
353	-	w.Header().Set("Range", fmt.Sprintf("0-%d", written-1))	224	+	w.Header().Set("Range", fmt.Sprintf("0-%d", length-1))
354		w.WriteHeader(http.StatusAccepted)	225		w.WriteHeader(http.StatusAccepted)
355	-
356	-	log.Debug().Str("uuid", uuid).Int64("bytes", written).Msg("Blob chunk uploaded")
357		}	226		}
358			227
359		func (s Server) handleListTags(w http.ResponseWriter, r http.Request) {	228		func (s Server) handleListTags(w http.ResponseWriter, r http.Request) {
360	-	name := s.parseTagsPath(r.URL.Path)	229	+	name := r.PathValue("name")
361	-		230	+
362	-	log.Debug().Str("name", name).Msg("List tags")	231	+	log.Debug().Str("name", name).Msg("Listing tags")
363	-		232	+
364		tags, err := s.storage.ListTags(name)	233		tags, err := s.storage.ListTags(name)
365		if err != nil {	234		if err != nil {
366	-	log.Error().Err(err).Str("name", name).Msg("Failed to list tags")	235	+	log.Warn().Err(err).Str("name", name).Msg("Tag not found")
367	-	http.Error(w, "Internal Server Error", http.StatusInternalServerError)	236	+	w.WriteHeader(http.StatusNotFound)
368		return	237		return
369		}	238		}
370	-		239	+
			240	+	// Construct the response JSON
			241	+	response := fmt.Sprintf(`{"name":"%s","tags":%s}`, name, tags)
			242	+
371		w.Header().Set("Content-Type", "application/json")	243		w.Header().Set("Content-Type", "application/json")
			244	+	w.Header().Set("Content-Length", strconv.Itoa(len(response)))
372		w.WriteHeader(http.StatusOK)	245		w.WriteHeader(http.StatusOK)
373	-		246	+	w.Write([]byte(response))
374	-	// Build JSON response	247	+	}
375	-	fmt.Fprintf(w, `{"name":"%s","tags":[`, name)	248	+
376	-	for i, tag := range tags {	249	+	func (s *Server) Start(ctx context.Context) error {
377	-	if i > 0 {	250	+	addr := ":" + strconv.Itoa(s.config.Server.RegistryPort)
378	-	fmt.Fprint(w, ",")	251	+
			252	+	// Build middleware chain
			253	+	middlewares := []func(http.Handler) http.Handler{
			254	+	middleware.PanicRecovery,
			255	+	middleware.RequestLogger,
			256	+	}
			257	+
			258	+	// Add registry authentication if enabled
			259	+	if s.config.RegistryAuth.Enabled {
			260	+	middlewares = append(middlewares, middleware.RegistryAuth(s.config.RegistryAuth.Username, s.config.RegistryAuth.Password))
			261	+	}
			262	+
			263	+	// Wrap the mux with middleware
			264	+	handler := middleware.Chain(middlewares...)(s.mux)
			265	+
			266	+	server := &http.Server{
			267	+	Addr: addr,
			268	+	Handler: handler,
			269	+	}
			270	+
			271	+	log.Info().Str("address", addr).Msg("Registry server starting")
			272	+
			273	+	// Start server in a goroutine
			274	+	errChan := make(chan error, 1)
			275	+	go func() {
			276	+	if err := server.ListenAndServe(); err != nil && err != http.ErrServerClosed {
			277	+	errChan <- err
379		}	278		}
380	-	fmt.Fprintf(w, `"%s"`, tag)	279	+	}()
			280	+
			281	+	// Wait for context cancellation or server error
			282	+	select {
			283	+	case err := <-errChan:
			284	+	return err
			285	+	case <-ctx.Done():
			286	+	log.Info().Msg("Registry server shutting down...")
			287	+	return server.Shutdown(context.Background())
381		}	288		}
382	-	fmt.Fprint(w, "]}")	289	+	}
383	-	}⏎̸

M internal/registry/storage.go +155 -65


@@ -15,25 +15,27 @@ import (
15	15		// Storage interface defines the contract for registry storage backends
16	16		type Storage interface {
17	17		// Manifest operations
18		-	GetManifest(name, reference string) ([]byte, error)
19		-	PutManifest(name, reference string, data []byte) error
	18	+	GetManifest(name, reference string) ([]byte, string, error)
	19	+	PutManifest(name, reference, contentType string, data []byte) error
20	20		DeleteManifest(name, reference string) error
21		-
	21	+
22	22		// Blob operations
23	23		GetBlob(digest string) (io.ReadCloser, error)
	24	+	GetBlobPath(digest string) (string, error)
24	25		PutBlob(digest string, data io.Reader, size int64) error
25	26		DeleteBlob(digest string) error
26	27		BlobExists(digest string) bool
27		-
	28	+
28	29		// Upload operations
29	30		StartBlobUpload(name string) (string, error) // returns upload UUID
	31	+	AppendBlobChunk(name, uuid string, chunk []byte) (int64, error)
30	32		GetBlobUpload(uuid string) (io.WriteCloser, error)
31	33		FinishBlobUpload(uuid, digest string) error
32	34		CancelBlobUpload(uuid string) error
33		-
	35	+
34	36		// Tag operations
35	37		ListTags(name string) ([]string, error)
36		-
	38	+
37	39		// Repository operations
38	40		ListRepositories() ([]string, error)
39	41		}

@@ -51,15 +53,15 @@ func NewFilesystemStorage(rootDir string) (*FilesystemStorage, error) {
51	53		filepath.Join(rootDir, "blobs"),
52	54		filepath.Join(rootDir, "uploads"),
53	55		}
54		-
	56	+
55	57		for _, dir := range dirs {
56	58		if err := os.MkdirAll(dir, 0755); err != nil {
57	59		return nil, fmt.Errorf("failed to create directory %s: %w", dir, err)
58	60		}
59	61		}
60		-
	62	+
61	63		log.Info().Str("root_dir", rootDir).Msg("Filesystem storage initialized")
62		-
	64	+
63	65		return &FilesystemStorage{
64	66		rootDir: rootDir,
65	67		}, nil

@@ -67,57 +69,75 @@ func NewFilesystemStorage(rootDir string) (*FilesystemStorage, error) {
67	69
68	70		// Manifest operations
69	71
70		-	func (fs *FilesystemStorage) GetManifest(name, reference string) ([]byte, error) {
	72	+	func (fs *FilesystemStorage) GetManifest(name, reference string) ([]byte, string, error) {
71	73		manifestPath := fs.getManifestPath(name, reference)
72		-
	74	+
73	75		data, err := os.ReadFile(manifestPath)
74	76		if err != nil {
75	77		if os.IsNotExist(err) {
76		-	return nil, fmt.Errorf("manifest not found: %s/%s", name, reference)
	78	+	return nil, "", fmt.Errorf("manifest not found: %s/%s", name, reference)
77	79		}
78		-	return nil, fmt.Errorf("failed to read manifest: %w", err)
	80	+	return nil, "", fmt.Errorf("failed to read manifest: %w", err)
	81	+	}
	82	+
	83	+	contentType, err := fs.getManifestContentType(name, reference)
	84	+	if err != nil {
	85	+	// Fallback for manifests stored before this change
	86	+	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Could not get manifest content type, falling back to default")
	87	+	contentType = "application/vnd.docker.distribution.manifest.v2+json"
79	88		}
80		-
81		-	return data, nil
	89	+
	90	+	return data, contentType, nil
82	91		}
83	92
84		-	func (fs *FilesystemStorage) PutManifest(name, reference string, data []byte) error {
	93	+	func (fs *FilesystemStorage) PutManifest(name, reference, contentType string, data []byte) error {
85	94		manifestPath := fs.getManifestPath(name, reference)
86		-
	95	+
87	96		// Create directory if it doesn't exist
88	97		if err := os.MkdirAll(filepath.Dir(manifestPath), 0755); err != nil {
89	98		return fmt.Errorf("failed to create manifest directory: %w", err)
90	99		}
91		-
	100	+
92	101		// Write manifest file
93	102		if err := os.WriteFile(manifestPath, data, 0644); err != nil {
94	103		return fmt.Errorf("failed to write manifest: %w", err)
95	104		}
96		-
	105	+
	106	+	// Store content type
	107	+	if err := fs.putManifestContentType(name, reference, contentType); err != nil {
	108	+	// Don't fail the whole operation, but log a warning
	109	+	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to store manifest content type")
	110	+	}
	111	+
97	112		// Update tags list
98	113		if err := fs.updateTagsList(name, reference); err != nil {
99	114		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to update tags list")
100	115		}
101		-
	116	+
102	117		log.Info().Str("name", name).Str("reference", reference).Msg("Manifest stored")
103	118		return nil
104	119		}
105	120
106	121		func (fs *FilesystemStorage) DeleteManifest(name, reference string) error {
107	122		manifestPath := fs.getManifestPath(name, reference)
108		-
	123	+
109	124		if err := os.Remove(manifestPath); err != nil {
110	125		if os.IsNotExist(err) {
111	126		return fmt.Errorf("manifest not found: %s/%s", name, reference)
112	127		}
113	128		return fmt.Errorf("failed to delete manifest: %w", err)
114	129		}
115		-
	130	+
	131	+	// Delete content type file
	132	+	if err := fs.deleteManifestContentType(name, reference); err != nil {
	133	+	log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to delete manifest content type")
	134	+	}
	135	+
116	136		// Remove from tags list
117	137		if err := fs.removeFromTagsList(name, reference); err != nil {
118	138		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to update tags list")
119	139		}
120		-
	140	+
121	141		log.Info().Str("name", name).Str("reference", reference).Msg("Manifest deleted")
122	142		return nil
123	143		}

@@ -126,7 +146,7 @@ func (fs *FilesystemStorage) DeleteManifest(name, reference string) error {
126	146
127	147		func (fs *FilesystemStorage) GetBlob(digest string) (io.ReadCloser, error) {
128	148		blobPath := fs.getBlobPath(digest)
129		-
	149	+
130	150		file, err := os.Open(blobPath)
131	151		if err != nil {
132	152		if os.IsNotExist(err) {

@@ -134,18 +154,29 @@ func (fs *FilesystemStorage) GetBlob(digest string) (io.ReadCloser, error) {
134	154		}
135	155		return nil, fmt.Errorf("failed to open blob: %w", err)
136	156		}
137		-
	157	+
138	158		return file, nil
139	159		}
140	160
	161	+	func (fs *FilesystemStorage) GetBlobPath(digest string) (string, error) {
	162	+	path := fs.getBlobPath(digest)
	163	+	if _, err := os.Stat(path); err != nil {
	164	+	if os.IsNotExist(err) {
	165	+	return "", fmt.Errorf("blob not found: %s", digest)
	166	+	}
	167	+	return "", err
	168	+	}
	169	+	return path, nil
	170	+	}
	171	+
141	172		func (fs *FilesystemStorage) PutBlob(digest string, data io.Reader, size int64) error {
142	173		blobPath := fs.getBlobPath(digest)
143		-
	174	+
144	175		// Create directory if it doesn't exist
145	176		if err := os.MkdirAll(filepath.Dir(blobPath), 0755); err != nil {
146	177		return fmt.Errorf("failed to create blob directory: %w", err)
147	178		}
148		-
	179	+
149	180		// Create temporary file first
150	181		tmpPath := blobPath + ".tmp"
151	182		file, err := os.Create(tmpPath)

@@ -153,40 +184,40 @@ func (fs *FilesystemStorage) PutBlob(digest string, data io.Reader, size int64)
153	184		return fmt.Errorf("failed to create temporary blob file: %w", err)
154	185		}
155	186		defer file.Close()
156		-
	187	+
157	188		// Copy data to file
158	189		written, err := io.Copy(file, data)
159	190		if err != nil {
160	191		os.Remove(tmpPath)
161	192		return fmt.Errorf("failed to write blob data: %w", err)
162	193		}
163		-
	194	+
164	195		// Verify size if provided
165	196		if size > 0 && written != size {
166	197		os.Remove(tmpPath)
167	198		return fmt.Errorf("blob size mismatch: expected %d, got %d", size, written)
168	199		}
169		-
	200	+
170	201		// Move to final location
171	202		if err := os.Rename(tmpPath, blobPath); err != nil {
172	203		os.Remove(tmpPath)
173	204		return fmt.Errorf("failed to move blob to final location: %w", err)
174	205		}
175		-
	206	+
176	207		log.Info().Str("digest", digest).Int64("size", written).Msg("Blob stored")
177	208		return nil
178	209		}
179	210
180	211		func (fs *FilesystemStorage) DeleteBlob(digest string) error {
181	212		blobPath := fs.getBlobPath(digest)
182		-
	213	+
183	214		if err := os.Remove(blobPath); err != nil {
184	215		if os.IsNotExist(err) {
185	216		return fmt.Errorf("blob not found: %s", digest)
186	217		}
187	218		return fmt.Errorf("failed to delete blob: %w", err)
188	219		}
189		-
	220	+
190	221		log.Info().Str("digest", digest).Msg("Blob deleted")
191	222		return nil
192	223		}

@@ -202,28 +233,60 @@ func (fs *FilesystemStorage) BlobExists(digest string) bool {
202	233		func (fs *FilesystemStorage) StartBlobUpload(name string) (string, error) {
203	234		// Generate UUID-like upload ID using timestamp
204	235		uuid := fmt.Sprintf("%d-%s", time.Now().UnixNano(), strings.ReplaceAll(name, "/", "_"))
205		-
	236	+
206	237		uploadPath := fs.getUploadPath(uuid)
207		-
	238	+
208	239		// Create directory if it doesn't exist
209	240		if err := os.MkdirAll(filepath.Dir(uploadPath), 0755); err != nil {
210	241		return "", fmt.Errorf("failed to create upload directory: %w", err)
211	242		}
212		-
	243	+
213	244		// Create empty upload file
214	245		file, err := os.Create(uploadPath)
215	246		if err != nil {
216	247		return "", fmt.Errorf("failed to create upload file: %w", err)
217	248		}
218	249		file.Close()
219		-
	250	+
220	251		log.Info().Str("uuid", uuid).Str("name", name).Msg("Blob upload started")
221	252		return uuid, nil
222	253		}
223	254
	255	+	func (fs *FilesystemStorage) AppendBlobChunk(name, uuid string, chunk []byte) (int64, error) {
	256	+	uploadPath := fs.getUploadPath(uuid)
	257	+
	258	+	file, err := os.OpenFile(uploadPath, os.O_WRONLY\|os.O_APPEND, 0644)
	259	+	if err != nil {
	260	+	if os.IsNotExist(err) {
	261	+	return 0, fmt.Errorf("upload not found: %s", uuid)
	262	+	}
	263	+	return 0, fmt.Errorf("failed to open upload file: %w", err)
	264	+	}
	265	+	defer file.Close()
	266	+
	267	+	written, err := file.Write(chunk)
	268	+	if err != nil {
	269	+	return 0, fmt.Errorf("failed to write chunk to upload file: %w", err)
	270	+	}
	271	+
	272	+	fi, err := file.Stat()
	273	+	if err != nil {
	274	+	return 0, fmt.Errorf("failed to get file info: %w", err)
	275	+	}
	276	+
	277	+	log.Debug().
	278	+	Str("uuid", uuid).
	279	+	Str("name", name).
	280	+	Int("chunk_size", written).
	281	+	Int64("total_size", fi.Size()).
	282	+	Msg("Appended chunk to blob upload")
	283	+
	284	+	return fi.Size(), nil
	285	+	}
	286	+
224	287		func (fs *FilesystemStorage) GetBlobUpload(uuid string) (io.WriteCloser, error) {
225	288		uploadPath := fs.getUploadPath(uuid)
226		-
	289	+
227	290		file, err := os.OpenFile(uploadPath, os.O_WRONLY\|os.O_APPEND, 0644)
228	291		if err != nil {
229	292		if os.IsNotExist(err) {

@@ -231,38 +294,38 @@ func (fs *FilesystemStorage) GetBlobUpload(uuid string) (io.WriteCloser, error)
231	294		}
232	295		return nil, fmt.Errorf("failed to open upload file: %w", err)
233	296		}
234		-
	297	+
235	298		return file, nil
236	299		}
237	300
238	301		func (fs *FilesystemStorage) FinishBlobUpload(uuid, digest string) error {
239	302		uploadPath := fs.getUploadPath(uuid)
240	303		blobPath := fs.getBlobPath(digest)
241		-
	304	+
242	305		// Create blob directory if it doesn't exist
243	306		if err := os.MkdirAll(filepath.Dir(blobPath), 0755); err != nil {
244	307		return fmt.Errorf("failed to create blob directory: %w", err)
245	308		}
246		-
	309	+
247	310		// Move upload file to blob location
248	311		if err := os.Rename(uploadPath, blobPath); err != nil {
249	312		return fmt.Errorf("failed to move upload to blob location: %w", err)
250	313		}
251		-
	314	+
252	315		log.Info().Str("uuid", uuid).Str("digest", digest).Msg("Blob upload finished")
253	316		return nil
254	317		}
255	318
256	319		func (fs *FilesystemStorage) CancelBlobUpload(uuid string) error {
257	320		uploadPath := fs.getUploadPath(uuid)
258		-
	321	+
259	322		if err := os.Remove(uploadPath); err != nil {
260	323		if os.IsNotExist(err) {
261	324		return fmt.Errorf("upload not found: %s", uuid)
262	325		}
263	326		return fmt.Errorf("failed to cancel upload: %w", err)
264	327		}
265		-
	328	+
266	329		log.Info().Str("uuid", uuid).Msg("Blob upload cancelled")
267	330		return nil
268	331		}

@@ -271,7 +334,7 @@ func (fs *FilesystemStorage) CancelBlobUpload(uuid string) error {
271	334
272	335		func (fs *FilesystemStorage) ListTags(name string) ([]string, error) {
273	336		tagsPath := fs.getTagsPath(name)
274		-
	337	+
275	338		data, err := os.ReadFile(tagsPath)
276	339		if err != nil {
277	340		if os.IsNotExist(err) {

@@ -279,12 +342,12 @@ func (fs *FilesystemStorage) ListTags(name string) ([]string, error) {
279	342		}
280	343		return nil, fmt.Errorf("failed to read tags file: %w", err)
281	344		}
282		-
	345	+
283	346		var tags []string
284	347		if err := json.Unmarshal(data, &tags); err != nil {
285	348		return nil, fmt.Errorf("failed to parse tags file: %w", err)
286	349		}
287		-
	350	+
288	351		return tags, nil
289	352		}
290	353

@@ -292,34 +355,34 @@ func (fs *FilesystemStorage) ListTags(name string) ([]string, error) {
292	355
293	356		func (fs *FilesystemStorage) ListRepositories() ([]string, error) {
294	357		reposDir := filepath.Join(fs.rootDir, "repositories")
295		-
	358	+
296	359		var repositories []string
297	360		err := filepath.Walk(reposDir, func(path string, info os.FileInfo, err error) error {
298	361		if err != nil {
299	362		return err
300	363		}
301		-
	364	+
302	365		if info.IsDir() && path != reposDir {
303	366		// Get relative path from repositories directory
304	367		relPath, err := filepath.Rel(reposDir, path)
305	368		if err != nil {
306	369		return err
307	370		}
308		-
	371	+
309	372		// Check if this directory contains manifests
310	373		manifestsDir := filepath.Join(path, "manifests")
311	374		if _, err := os.Stat(manifestsDir); err == nil {
312	375		repositories = append(repositories, relPath)
313	376		}
314	377		}
315		-
	378	+
316	379		return nil
317	380		})
318		-
	381	+
319	382		if err != nil {
320	383		return nil, fmt.Errorf("failed to list repositories: %w", err)
321	384		}
322		-
	385	+
323	386		return repositories, nil
324	387		}
325	388

@@ -329,6 +392,33 @@ func (fs *FilesystemStorage) getManifestPath(name, reference string) string {
329	392		return filepath.Join(fs.rootDir, "repositories", name, "manifests", reference)
330	393		}
331	394
	395	+	func (fs *FilesystemStorage) getManifestContentTypePath(name, reference string) string {
	396	+	return fs.getManifestPath(name, reference) + ".contenttype"
	397	+	}
	398	+
	399	+	func (fs *FilesystemStorage) getManifestContentType(name, reference string) (string, error) {
	400	+	contentTypePath := fs.getManifestContentTypePath(name, reference)
	401	+	data, err := os.ReadFile(contentTypePath)
	402	+	if err != nil {
	403	+	return "", err
	404	+	}
	405	+	return string(data), nil
	406	+	}
	407	+
	408	+	func (fs *FilesystemStorage) putManifestContentType(name, reference, contentType string) error {
	409	+	contentTypePath := fs.getManifestContentTypePath(name, reference)
	410	+	return os.WriteFile(contentTypePath, []byte(contentType), 0644)
	411	+	}
	412	+
	413	+	func (fs *FilesystemStorage) deleteManifestContentType(name, reference string) error {
	414	+	contentTypePath := fs.getManifestContentTypePath(name, reference)
	415	+	err := os.Remove(contentTypePath)
	416	+	if err != nil && !os.IsNotExist(err) {
	417	+	return err
	418	+	}
	419	+	return nil
	420	+	}
	421	+
332	422		func (fs *FilesystemStorage) getBlobPath(digest string) string {
333	423		// Split digest into directory structure (e.g., sha256:abc123... -> sha256/ab/abc123...)
334	424		parts := strings.SplitN(digest, ":", 2)

@@ -336,15 +426,15 @@ func (fs *FilesystemStorage) getBlobPath(digest string) string {
336	426		// Fallback if digest format is unexpected
337	427		return filepath.Join(fs.rootDir, "blobs", digest)
338	428		}
339		-
	429	+
340	430		algorithm := parts[0]
341	431		hash := parts[1]
342		-
	432	+
343	433		// Create two-level directory structure for better performance
344	434		if len(hash) >= 2 {
345	435		return filepath.Join(fs.rootDir, "blobs", algorithm, hash[:2], hash)
346	436		}
347		-
	437	+
348	438		return filepath.Join(fs.rootDir, "blobs", algorithm, hash)
349	439		}
350	440

@@ -363,14 +453,14 @@ func (fs *FilesystemStorage) updateTagsList(name, reference string) error {
363	453		if err != nil {
364	454		return err
365	455		}
366		-
	456	+
367	457		// Add tag if not already present
368	458		for _, tag := range tags {
369	459		if tag == reference {
370	460		return nil // Already exists
371	461		}
372	462		}
373		-
	463	+
374	464		tags = append(tags, reference)
375	465		return fs.saveTagsList(name, tags)
376	466		}

@@ -380,7 +470,7 @@ func (fs *FilesystemStorage) removeFromTagsList(name, reference string) error {
380	470		if err != nil {
381	471		return err
382	472		}
383		-
	473	+
384	474		// Remove tag if present
385	475		var newTags []string
386	476		for _, tag := range tags {

@@ -388,26 +478,26 @@ func (fs *FilesystemStorage) removeFromTagsList(name, reference string) error {
388	478		newTags = append(newTags, tag)
389	479		}
390	480		}
391		-
	481	+
392	482		return fs.saveTagsList(name, newTags)
393	483		}
394	484
395	485		func (fs *FilesystemStorage) saveTagsList(name string, tags []string) error {
396	486		tagsPath := fs.getTagsPath(name)
397		-
	487	+
398	488		// Create directory if it doesn't exist
399	489		if err := os.MkdirAll(filepath.Dir(tagsPath), 0755); err != nil {
400	490		return fmt.Errorf("failed to create tags directory: %w", err)
401	491		}
402		-
	492	+
403	493		data, err := json.Marshal(tags)
404	494		if err != nil {
405	495		return fmt.Errorf("failed to marshal tags: %w", err)
406	496		}
407		-
	497	+
408	498		if err := os.WriteFile(tagsPath, data, 0644); err != nil {
409	499		return fmt.Errorf("failed to write tags file: %w", err)
410	500		}
411		-
	501	+
412	502		return nil
413		-	}⏎̸
	503	+	}

 // Storage interface defines the contract for registry storage backends
 type Storage interface {
 	// Manifest operations
-	GetManifest(name, reference string) ([]byte, error)
+	GetManifest(name, reference string) ([]byte, string, error)
-	PutManifest(name, reference string, data []byte) error
+	PutManifest(name, reference, contentType string, data []byte) error
 	DeleteManifest(name, reference string) error
 	// Blob operations
 	GetBlob(digest string) (io.ReadCloser, error)
+	GetBlobPath(digest string) (string, error)
 	PutBlob(digest string, data io.Reader, size int64) error
 	DeleteBlob(digest string) error
 	BlobExists(digest string) bool
 	// Upload operations
 	StartBlobUpload(name string) (string, error) // returns upload UUID
+	AppendBlobChunk(name, uuid string, chunk []byte) (int64, error)
 	GetBlobUpload(uuid string) (io.WriteCloser, error)
 	FinishBlobUpload(uuid, digest string) error
 	CancelBlobUpload(uuid string) error
 	// Tag operations
 	ListTags(name string) ([]string, error)
 	// Repository operations
 	ListRepositories() ([]string, error)
 }
 		filepath.Join(rootDir, "blobs"),
 		filepath.Join(rootDir, "uploads"),
 	}
 	for _, dir := range dirs {
 		if err := os.MkdirAll(dir, 0755); err != nil {
 			return nil, fmt.Errorf("failed to create directory %s: %w", dir, err)
 		}
 	}
 	log.Info().Str("root_dir", rootDir).Msg("Filesystem storage initialized")
 	return &FilesystemStorage{
 		rootDir: rootDir,
 	}, nil
 // Manifest operations
-func (fs *FilesystemStorage) GetManifest(name, reference string) ([]byte, error) {
+func (fs *FilesystemStorage) GetManifest(name, reference string) ([]byte, string, error) {
 	manifestPath := fs.getManifestPath(name, reference)
 	data, err := os.ReadFile(manifestPath)
 	if err != nil {
 		if os.IsNotExist(err) {
-			return nil, fmt.Errorf("manifest not found: %s/%s", name, reference)
+			return nil, "", fmt.Errorf("manifest not found: %s/%s", name, reference)
 		}
-		return nil, fmt.Errorf("failed to read manifest: %w", err)
+		return nil, "", fmt.Errorf("failed to read manifest: %w", err)
+	}
+	contentType, err := fs.getManifestContentType(name, reference)
+	if err != nil {
+		// Fallback for manifests stored before this change
+		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Could not get manifest content type, falling back to default")
+		contentType = "application/vnd.docker.distribution.manifest.v2+json"
 	}
-	return data, nil
+	return data, contentType, nil
 }
-func (fs *FilesystemStorage) PutManifest(name, reference string, data []byte) error {
+func (fs *FilesystemStorage) PutManifest(name, reference, contentType string, data []byte) error {
 	manifestPath := fs.getManifestPath(name, reference)
 	// Create directory if it doesn't exist
 	if err := os.MkdirAll(filepath.Dir(manifestPath), 0755); err != nil {
 		return fmt.Errorf("failed to create manifest directory: %w", err)
 	}
 	// Write manifest file
 	if err := os.WriteFile(manifestPath, data, 0644); err != nil {
 		return fmt.Errorf("failed to write manifest: %w", err)
 	}
+	// Store content type
+	if err := fs.putManifestContentType(name, reference, contentType); err != nil {
+		// Don't fail the whole operation, but log a warning
+		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to store manifest content type")
+	}
 	// Update tags list
 	if err := fs.updateTagsList(name, reference); err != nil {
 		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to update tags list")
 	}
 	log.Info().Str("name", name).Str("reference", reference).Msg("Manifest stored")
 	return nil
 }
 func (fs *FilesystemStorage) DeleteManifest(name, reference string) error {
 	manifestPath := fs.getManifestPath(name, reference)
 	if err := os.Remove(manifestPath); err != nil {
 		if os.IsNotExist(err) {
 			return fmt.Errorf("manifest not found: %s/%s", name, reference)
 		}
 		return fmt.Errorf("failed to delete manifest: %w", err)
 	}
+	// Delete content type file
+	if err := fs.deleteManifestContentType(name, reference); err != nil {
+		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to delete manifest content type")
+	}
 	// Remove from tags list
 	if err := fs.removeFromTagsList(name, reference); err != nil {
 		log.Warn().Err(err).Str("name", name).Str("reference", reference).Msg("Failed to update tags list")
 	}
 	log.Info().Str("name", name).Str("reference", reference).Msg("Manifest deleted")
 	return nil
 }
 func (fs *FilesystemStorage) GetBlob(digest string) (io.ReadCloser, error) {
 	blobPath := fs.getBlobPath(digest)
 	file, err := os.Open(blobPath)
 	if err != nil {
 		if os.IsNotExist(err) {
 		}
 		return nil, fmt.Errorf("failed to open blob: %w", err)
 	}
 	return file, nil
 }
+func (fs *FilesystemStorage) GetBlobPath(digest string) (string, error) {
+	path := fs.getBlobPath(digest)
+	if _, err := os.Stat(path); err != nil {
+		if os.IsNotExist(err) {
+			return "", fmt.Errorf("blob not found: %s", digest)
+		}
+		return "", err
+	}
+	return path, nil
+}
 func (fs *FilesystemStorage) PutBlob(digest string, data io.Reader, size int64) error {
 	blobPath := fs.getBlobPath(digest)
 	// Create directory if it doesn't exist
 	if err := os.MkdirAll(filepath.Dir(blobPath), 0755); err != nil {
 		return fmt.Errorf("failed to create blob directory: %w", err)
 	}
 	// Create temporary file first
 	tmpPath := blobPath + ".tmp"
 	file, err := os.Create(tmpPath)
 		return fmt.Errorf("failed to create temporary blob file: %w", err)
 	}
 	defer file.Close()
 	// Copy data to file
 	written, err := io.Copy(file, data)
 	if err != nil {
 		os.Remove(tmpPath)
 		return fmt.Errorf("failed to write blob data: %w", err)
 	}
 	// Verify size if provided
 	if size > 0 && written != size {
 		os.Remove(tmpPath)
 		return fmt.Errorf("blob size mismatch: expected %d, got %d", size, written)
 	}
 	// Move to final location
 	if err := os.Rename(tmpPath, blobPath); err != nil {
 		os.Remove(tmpPath)
 		return fmt.Errorf("failed to move blob to final location: %w", err)
 	}
 	log.Info().Str("digest", digest).Int64("size", written).Msg("Blob stored")
 	return nil
 }
 func (fs *FilesystemStorage) DeleteBlob(digest string) error {
 	blobPath := fs.getBlobPath(digest)
 	if err := os.Remove(blobPath); err != nil {
 		if os.IsNotExist(err) {
 			return fmt.Errorf("blob not found: %s", digest)
 		}
 		return fmt.Errorf("failed to delete blob: %w", err)
 	}
 	log.Info().Str("digest", digest).Msg("Blob deleted")
 	return nil
 }
 func (fs *FilesystemStorage) StartBlobUpload(name string) (string, error) {
 	// Generate UUID-like upload ID using timestamp
 	uuid := fmt.Sprintf("%d-%s", time.Now().UnixNano(), strings.ReplaceAll(name, "/", "_"))
 	uploadPath := fs.getUploadPath(uuid)
 	// Create directory if it doesn't exist
 	if err := os.MkdirAll(filepath.Dir(uploadPath), 0755); err != nil {
 		return "", fmt.Errorf("failed to create upload directory: %w", err)
 	}
 	// Create empty upload file
 	file, err := os.Create(uploadPath)
 	if err != nil {
 		return "", fmt.Errorf("failed to create upload file: %w", err)
 	}
 	file.Close()
 	log.Info().Str("uuid", uuid).Str("name", name).Msg("Blob upload started")
 	return uuid, nil
 }
+func (fs *FilesystemStorage) AppendBlobChunk(name, uuid string, chunk []byte) (int64, error) {
+	uploadPath := fs.getUploadPath(uuid)
+	file, err := os.OpenFile(uploadPath, os.O_WRONLY|os.O_APPEND, 0644)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return 0, fmt.Errorf("upload not found: %s", uuid)
+		}
+		return 0, fmt.Errorf("failed to open upload file: %w", err)
+	}
+	defer file.Close()
+	written, err := file.Write(chunk)
+	if err != nil {
+		return 0, fmt.Errorf("failed to write chunk to upload file: %w", err)
+	}
+	fi, err := file.Stat()
+	if err != nil {
+		return 0, fmt.Errorf("failed to get file info: %w", err)
+	}
+	log.Debug().
+		Str("uuid", uuid).
+		Str("name", name).
+		Int("chunk_size", written).
+		Int64("total_size", fi.Size()).
+		Msg("Appended chunk to blob upload")
+	return fi.Size(), nil
+}
 func (fs *FilesystemStorage) GetBlobUpload(uuid string) (io.WriteCloser, error) {
 	uploadPath := fs.getUploadPath(uuid)
 	file, err := os.OpenFile(uploadPath, os.O_WRONLY|os.O_APPEND, 0644)
 	if err != nil {
 		if os.IsNotExist(err) {
 		}
 		return nil, fmt.Errorf("failed to open upload file: %w", err)
 	}
 	return file, nil
 }
 func (fs *FilesystemStorage) FinishBlobUpload(uuid, digest string) error {
 	uploadPath := fs.getUploadPath(uuid)
 	blobPath := fs.getBlobPath(digest)
 	// Create blob directory if it doesn't exist
 	if err := os.MkdirAll(filepath.Dir(blobPath), 0755); err != nil {
 		return fmt.Errorf("failed to create blob directory: %w", err)
 	}
 	// Move upload file to blob location
 	if err := os.Rename(uploadPath, blobPath); err != nil {
 		return fmt.Errorf("failed to move upload to blob location: %w", err)
 	}
 	log.Info().Str("uuid", uuid).Str("digest", digest).Msg("Blob upload finished")
 	return nil
 }
 func (fs *FilesystemStorage) CancelBlobUpload(uuid string) error {
 	uploadPath := fs.getUploadPath(uuid)
 	if err := os.Remove(uploadPath); err != nil {
 		if os.IsNotExist(err) {
 			return fmt.Errorf("upload not found: %s", uuid)
 		}
 		return fmt.Errorf("failed to cancel upload: %w", err)
 	}
 	log.Info().Str("uuid", uuid).Msg("Blob upload cancelled")
 	return nil
 }
 func (fs *FilesystemStorage) ListTags(name string) ([]string, error) {
 	tagsPath := fs.getTagsPath(name)
 	data, err := os.ReadFile(tagsPath)
 	if err != nil {
 		if os.IsNotExist(err) {
 		}
 		return nil, fmt.Errorf("failed to read tags file: %w", err)
 	}
 	var tags []string
 	if err := json.Unmarshal(data, &tags); err != nil {
 		return nil, fmt.Errorf("failed to parse tags file: %w", err)
 	}
 	return tags, nil
 }
 func (fs *FilesystemStorage) ListRepositories() ([]string, error) {
 	reposDir := filepath.Join(fs.rootDir, "repositories")
 	var repositories []string
 	err := filepath.Walk(reposDir, func(path string, info os.FileInfo, err error) error {
 		if err != nil {
 			return err
 		}
 		if info.IsDir() && path != reposDir {
 			// Get relative path from repositories directory
 			relPath, err := filepath.Rel(reposDir, path)
 			if err != nil {
 				return err
 			}
 			// Check if this directory contains manifests
 			manifestsDir := filepath.Join(path, "manifests")
 			if _, err := os.Stat(manifestsDir); err == nil {
 				repositories = append(repositories, relPath)
 			}
 		}
 		return nil
 	})
 	if err != nil {
 		return nil, fmt.Errorf("failed to list repositories: %w", err)
 	}
 	return repositories, nil
 }
 	return filepath.Join(fs.rootDir, "repositories", name, "manifests", reference)
 }
+func (fs *FilesystemStorage) getManifestContentTypePath(name, reference string) string {
+	return fs.getManifestPath(name, reference) + ".contenttype"
+}
+func (fs *FilesystemStorage) getManifestContentType(name, reference string) (string, error) {
+	contentTypePath := fs.getManifestContentTypePath(name, reference)
+	data, err := os.ReadFile(contentTypePath)
+	if err != nil {
+		return "", err
+	}
+	return string(data), nil
+}
+func (fs *FilesystemStorage) putManifestContentType(name, reference, contentType string) error {
+	contentTypePath := fs.getManifestContentTypePath(name, reference)
+	return os.WriteFile(contentTypePath, []byte(contentType), 0644)
+}
+func (fs *FilesystemStorage) deleteManifestContentType(name, reference string) error {
+	contentTypePath := fs.getManifestContentTypePath(name, reference)
+	err := os.Remove(contentTypePath)
+	if err != nil && !os.IsNotExist(err) {
+		return err
+	}
+	return nil
+}
 func (fs *FilesystemStorage) getBlobPath(digest string) string {
 	// Split digest into directory structure (e.g., sha256:abc123... -> sha256/ab/abc123...)
 	parts := strings.SplitN(digest, ":", 2)
 		// Fallback if digest format is unexpected
 		return filepath.Join(fs.rootDir, "blobs", digest)
 	}
 	algorithm := parts[0]
 	hash := parts[1]
 	// Create two-level directory structure for better performance
 	if len(hash) >= 2 {
 		return filepath.Join(fs.rootDir, "blobs", algorithm, hash[:2], hash)
 	}
 	return filepath.Join(fs.rootDir, "blobs", algorithm, hash)
 }
 	if err != nil {
 		return err
 	}
 	// Add tag if not already present
 	for _, tag := range tags {
 		if tag == reference {
 			return nil // Already exists
 		}
 	}
 	tags = append(tags, reference)
 	return fs.saveTagsList(name, tags)
 }
 	if err != nil {
 		return err
 	}
 	// Remove tag if present
 	var newTags []string
 	for _, tag := range tags {
 			newTags = append(newTags, tag)
 		}
 	}
 	return fs.saveTagsList(name, newTags)
 }
 func (fs *FilesystemStorage) saveTagsList(name string, tags []string) error {
 	tagsPath := fs.getTagsPath(name)
 	// Create directory if it doesn't exist
 	if err := os.MkdirAll(filepath.Dir(tagsPath), 0755); err != nil {
 		return fmt.Errorf("failed to create tags directory: %w", err)
 	}
 	data, err := json.Marshal(tags)
 	if err != nil {
 		return fmt.Errorf("failed to marshal tags: %w", err)
 	}
 	if err := os.WriteFile(tagsPath, data, 0644); err != nil {
 		return fmt.Errorf("failed to write tags file: %w", err)
 	}
 	return nil
-}⏎̸
+}

M pkg/runtime/interface.go +1 -0


@@ -43,6 +43,7 @@ type Runtime interface {
43	43
44	44		// Image operations
45	45		PullImage(ctx context.Context, image string) error
	46	+	PullImageWithAuth(ctx context.Context, image, username, password string) error
46	47		RemoveImage(ctx context.Context, image string, force bool) error
47	48		ListImages(ctx context.Context) ([]string, error)
48	49


@@ -43,6 +43,7 @@ type Runtime interface {
43		43
44	// Image operations	44		// Image operations
45	PullImage(ctx context.Context, image string) error	45		PullImage(ctx context.Context, image string) error
		46	+	PullImageWithAuth(ctx context.Context, image, username, password string) error
46	RemoveImage(ctx context.Context, image string, force bool) error	47		RemoveImage(ctx context.Context, image string, force bool) error
47	ListImages(ctx context.Context) ([]string, error)	48		ListImages(ctx context.Context) ([]string, error)
48		49