hamlab/internal/api/docker.go

package api

// The Docker tab: what is running, grouped the way it was deployed (compose stacks), with
// start / stop / restart, logs and inspect — plus images, volumes and networks to look at.
// It drives the docker CLI (or podman), as the user first and as root if that is refused.
// Creating, removing and pruning are deliberately absent: whatever deployed the stacks
// (Dockhand, compose) stays the one place that changes what exists.

import (
	"context"
	"encoding/json"
	"errors"
	"fmt"
	"io"
	"log"
	"net/http"
	"os"
	"os/exec"
	"regexp"
	"sort"
	"strconv"
	"strings"
	"time"

	"hamlab/internal/priv"
)

type dockerContainer struct {
	ID      string  `json:"id"`
	Name    string  `json:"name"`
	Image   string  `json:"image"`
	State   string  `json:"state"`  // running | exited | paused | restarting | created | dead
	Status  string  `json:"status"` // "Up 3 hours (healthy)"
	Health  string  `json:"health"` // healthy | unhealthy | starting | none
	Ports   string  `json:"ports"`
	Stack   string  `json:"stack"`   // compose project, "" when started by hand
	Service string  `json:"service"` // compose service
	Created string  `json:"created"`
	Mem     float64 `json:"mem"` // bytes, from the container's cgroup; 0 when unknown
}

func dockerTool() string {
	for _, t := range []string{"docker", "podman"} {
		if _, err := exec.LookPath(t); err == nil {
			return t
		}
	}
	return ""
}

// dockerRun runs the tool as this user, and again as root if the socket said no.
func dockerRun(ctx context.Context, args ...string) ([]byte, error) {
	tool := dockerTool()
	if tool == "" {
		return nil, errors.New("neither docker nor podman is installed here")
	}
	out, err := exec.CommandContext(ctx, tool, args...).CombinedOutput()
	if err != nil && priv.Available() && strings.Contains(strings.ToLower(string(out)), "permission denied") {
		full := priv.Wrap(append([]string{tool}, args...)...)
		out, err = exec.CommandContext(ctx, full[0], full[1:]...).CombinedOutput()
	}
	if err != nil {
		msg := strings.TrimSpace(string(out))
		if i := strings.LastIndex(msg, "\n"); i >= 0 {
			msg = msg[i+1:]
		}
		if msg == "" {
			msg = err.Error()
		}
		return out, errors.New(msg)
	}
	return out, nil
}

func jsonLines(out []byte, each func(map[string]any)) {
	for _, line := range strings.Split(strings.TrimSpace(string(out)), "\n") {
		var m map[string]any
		if json.Unmarshal([]byte(line), &m) == nil {
			each(m)
		}
	}
}

func dstr(m map[string]any, k string) string { s, _ := m[k].(string); return s }

func labelOf(labels, key string) string {
	for _, kv := range strings.Split(labels, ",") {
		if v, ok := strings.CutPrefix(kv, key+"="); ok {
			return v
		}
	}
	return ""
}

func (s *Server) dockerOverview(w http.ResponseWriter, r *http.Request) {
	if dockerTool() == "" {
		jsonOK(w, 200, map[string]any{"available": false, "message": "Docker isn't installed on this machine."})
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 20*time.Second)
	defer cancel()
	out, err := dockerRun(ctx, "ps", "-a", "--no-trunc", "--format", "{{json .}}")
	if err != nil {
		jsonOK(w, 200, map[string]any{"available": false, "message": err.Error()})
		return
	}
	containers := []dockerContainer{}
	jsonLines(out, func(m map[string]any) {
		c := dockerContainer{ID: dstr(m, "ID"), Name: strings.Split(dstr(m, "Names"), ",")[0], Image: dstr(m, "Image"), State: dstr(m, "State"), Status: dstr(m, "Status"),
			Ports: dstr(m, "Ports"), Created: dstr(m, "CreatedAt"), Health: "none", Stack: labelOf(dstr(m, "Labels"), "com.docker.compose.project"), Service: labelOf(dstr(m, "Labels"), "com.docker.compose.service")}
		for _, h := range []string{"unhealthy", "healthy", "starting"} {
			if strings.Contains(c.Status, "("+h+")") || strings.Contains(c.Status, "health: "+h) {
				c.Health = h
				break
			}
		}
		for _, dir := range []string{"/sys/fs/cgroup/system.slice/docker-" + c.ID + ".scope", "/sys/fs/cgroup/docker/" + c.ID, "/sys/fs/cgroup/machine.slice/libpod-" + c.ID + ".scope"} {
			if b, err := os.ReadFile(dir + "/memory.current"); err == nil {
				c.Mem, _ = strconv.ParseFloat(strings.TrimSpace(string(b)), 64)
				break
			}
		}
		containers = append(containers, c)
	})
	sort.Slice(containers, func(i, j int) bool {
		if containers[i].Stack != containers[j].Stack {
			return containers[i].Stack < containers[j].Stack
		}
		return containers[i].Name < containers[j].Name
	})
	list := func(args ...string) []map[string]any {
		rows := []map[string]any{}
		if out, err := dockerRun(ctx, args...); err == nil {
			jsonLines(out, func(m map[string]any) { rows = append(rows, m) })
		}
		return rows
	}
	// Who uses what, from one inspect of every container: an image or a volume nothing uses is
	// the only kind this page offers to remove.
	imageUsed, volumeUsers := map[string]bool{}, map[string][]string{}
	if ids, err := dockerRun(ctx, "ps", "-aq"); err == nil && len(strings.Fields(string(ids))) > 0 {
		const f = `{{.Name}}|{{.Image}}|{{range .Mounts}}{{if eq .Type "volume"}}{{.Name}},{{end}}{{end}}`
		if out, err := dockerRun(ctx, append([]string{"inspect", "-f", f}, strings.Fields(string(ids))...)...); err == nil {
			for _, line := range strings.Split(strings.TrimSpace(string(out)), "\n") {
				p := strings.SplitN(line, "|", 3)
				if len(p) < 3 {
					continue
				}
				if id := strings.TrimPrefix(p[1], "sha256:"); len(id) >= 12 {
					imageUsed[id[:12]] = true
				}
				for _, v := range strings.Split(p[2], ",") {
					if v != "" {
						volumeUsers[v] = append(volumeUsers[v], strings.TrimPrefix(p[0], "/"))
					}
				}
			}
		}
	}
	images := list("images", "--format", "{{json .}}")
	for _, im := range images {
		im["in_use"] = imageUsed[dstr(im, "ID")]
	}
	volumes := list("volume", "ls", "--format", "{{json .}}")
	for _, v := range volumes {
		users := volumeUsers[dstr(v, "Name")]
		if users == nil {
			users = []string{}
		}
		sort.Strings(users)
		v["used_by"] = users
		v["anonymous"] = strings.Contains(dstr(v, "Labels"), "com.docker.volume.anonymous")
	}
	jsonOK(w, 200, map[string]any{"available": true, "tool": dockerTool(), "containers": containers,
		"images": images, "volumes": volumes, "networks": list("network", "ls", "--format", "{{json .}}")})
}

var dockerRef = regexp.MustCompile(`^[A-Za-z0-9][A-Za-z0-9_.-]{0,127}$`)

func (s *Server) dockerAction(w http.ResponseWriter, r *http.Request) {
	id, action := r.PathValue("id"), r.PathValue("action")
	if !dockerRef.MatchString(id) || (action != "start" && action != "stop" && action != "restart" && action != "remove") {
		jsonErr(w, 400, "start, stop, restart or remove a container by name or ID")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 90*time.Second)
	defer cancel()
	verb := action
	if action == "remove" { // never forced: a running container has to be stopped first, on purpose
		verb = "rm"
	}
	if _, err := dockerRun(ctx, verb, id); err != nil {
		code := 500
		if action == "remove" {
			code = 409
		}
		jsonErr(w, code, err.Error())
		return
	}
	log.Printf("docker: %s %s (from %s)", action, id, clientIP(r))
	w.WriteHeader(204)
}

var imageRef = regexp.MustCompile(`^[a-zA-Z0-9][a-zA-Z0-9._/:@-]{0,254}$`)

// imagePull runs `docker pull` as a task the page follows.
func (s *Server) imagePull(w http.ResponseWriter, r *http.Request) {
	var body struct{ Image string }
	if err := json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body); err != nil || !imageRef.MatchString(strings.TrimSpace(body.Image)) {
		jsonErr(w, 400, "an image looks like nginx:alpine or ghcr.io/owner/name:tag")
		return
	}
	ref := strings.TrimSpace(body.Image)
	t, err := s.tasks.Start("docker:pull", "pull "+ref, exec.Command(dockerTool(), "pull", ref))
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	log.Printf("docker: pull %s as task %s (from %s)", ref, t.ID, clientIP(r))
	jsonOK(w, 202, t.Snapshot(false))
}

// imageRemove is `docker rmi`, unforced: an image something still uses is refused by Docker, and that is the answer.
func (s *Server) imageRemove(w http.ResponseWriter, r *http.Request) {
	var body struct{ Image string }
	if err := json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body); err != nil || !imageRef.MatchString(strings.TrimSpace(body.Image)) {
		jsonErr(w, 400, "which image?")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 90*time.Second)
	defer cancel()
	if _, err := dockerRun(ctx, "rmi", strings.TrimSpace(body.Image)); err != nil {
		jsonErr(w, 409, err.Error())
		return
	}
	log.Printf("docker: rmi %s (from %s)", body.Image, clientIP(r))
	w.WriteHeader(204)
}

// imagePrune removes dangling images — layers no tag points at any more, left behind by pulls —
// or, asked for {"all": true}, every image no container uses. Those can all be pulled again.
func (s *Server) imagePrune(w http.ResponseWriter, r *http.Request) {
	ctx, cancel := context.WithTimeout(r.Context(), 5*time.Minute)
	defer cancel()
	var body struct{ All bool }
	_ = json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body)
	args := []string{"image", "prune", "-f"}
	if body.All {
		args = append(args, "-a")
	}
	out, err := dockerRun(ctx, args...)
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	lines := strings.Split(strings.TrimSpace(string(out)), "\n")
	log.Printf("docker: image prune (from %s)", clientIP(r))
	jsonOK(w, 200, map[string]string{"result": lines[len(lines)-1]})
}

// dockerLogs is the tail of a container's log; ?since=<RFC3339> asks only for what is new,
// which is how the page follows a log without holding a connection open.
func (s *Server) dockerLogs(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	args := []string{"logs", "--timestamps"}
	if since := r.URL.Query().Get("since"); since != "" {
		if _, err := time.Parse(time.RFC3339Nano, since); err != nil {
			jsonErr(w, 400, "since is an RFC 3339 time")
			return
		}
		args = append(args, "--since", since)
	} else {
		n, _ := strconv.Atoi(r.URL.Query().Get("n"))
		if n <= 0 || n > 5000 {
			n = 300
		}
		args = append(args, "--tail", strconv.Itoa(n))
	}
	ctx, cancel := context.WithTimeout(r.Context(), 20*time.Second)
	defer cancel()
	out, err := dockerRun(ctx, append(args, id)...)
	if err != nil {
		code := 500
		if strings.Contains(strings.ToLower(err.Error()), "no such container") {
			code = 404
		}
		jsonErr(w, code, err.Error())
		return
	}
	type line struct {
		T string `json:"t"`
		M string `json:"m"`
	}
	lines := []line{}
	for _, l := range strings.Split(strings.TrimRight(string(out), "\n"), "\n") {
		if l == "" {
			continue
		}
		t, m, ok := strings.Cut(l, " ")
		if _, err := time.Parse(time.RFC3339Nano, t); !ok || err != nil {
			t, m = "", l
		}
		lines = append(lines, line{t, m})
	}
	jsonOK(w, 200, map[string]any{"lines": lines})
}

// dockerInspect is the parts of `docker inspect` worth reading on a phone.
func (s *Server) dockerInspect(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 15*time.Second)
	defer cancel()
	out, err := dockerRun(ctx, "inspect", id)
	var all []map[string]any
	if err != nil || json.Unmarshal(out, &all) != nil || len(all) == 0 {
		jsonErr(w, 500, fmt.Sprint("inspect failed: ", err))
		return
	}
	c := all[0]
	dig := func(m map[string]any, path ...string) any {
		var cur any = m
		for _, p := range path {
			mm, ok := cur.(map[string]any)
			if !ok {
				return nil
			}
			cur = mm[p]
		}
		return cur
	}
	env := []string{}
	if list, ok := dig(c, "Config", "Env").([]any); ok {
		for _, e := range list {
			k, _, _ := strings.Cut(fmt.Sprint(e), "=")
			env = append(env, k) // names only: values are where the secrets live
		}
	}
	sort.Strings(env)
	jsonOK(w, 200, map[string]any{
		"image": dig(c, "Config", "Image"), "command": dig(c, "Config", "Cmd"), "entrypoint": dig(c, "Config", "Entrypoint"),
		"restart": dig(c, "HostConfig", "RestartPolicy", "Name"), "network_mode": dig(c, "HostConfig", "NetworkMode"),
		"memory_limit": dig(c, "HostConfig", "Memory"), "started": dig(c, "State", "StartedAt"), "exit_code": dig(c, "State", "ExitCode"),
		"error": dig(c, "State", "Error"), "restarts": c["RestartCount"], "mounts": c["Mounts"], "ports": dig(c, "NetworkSettings", "Ports"),
		"networks": dig(c, "NetworkSettings", "Networks"), "env_names": env,
		"compose_file": dig(c, "Config", "Labels", "com.docker.compose.project.config_files"),
	})
}

var volumeRef = regexp.MustCompile(`^[a-zA-Z0-9][a-zA-Z0-9_.-]{0,254}$`)

// volumeRemove deletes one volume and what is in it, which nothing brings back. So it is never
// forced (Docker refuses a volume a container still uses), it is one at a time, and the request
// has to name the volume twice — a slip of a script can't delete one by accident.
func (s *Server) volumeRemove(w http.ResponseWriter, r *http.Request) {
	var body struct{ Name, Confirm string }
	if err := json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body); err != nil || !volumeRef.MatchString(body.Name) {
		jsonErr(w, 400, "which volume?")
		return
	}
	if body.Confirm != body.Name {
		jsonErr(w, 400, "name the volume again in \"confirm\" to delete it")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 90*time.Second)
	defer cancel()
	if _, err := dockerRun(ctx, "volume", "rm", body.Name); err != nil {
		jsonErr(w, 409, err.Error())
		return
	}
	log.Printf("docker: volume rm %s (from %s)", body.Name, clientIP(r))
	w.WriteHeader(204)
}

// volumePrune removes unused *anonymous* volumes only: the nameless ones an image's VOLUME line
// leaves behind each time its container is recreated. Named volumes are somebody's data, and a
// stack that is merely down leaves its volumes "unused" — those go one at a time, by name.
func (s *Server) volumePrune(w http.ResponseWriter, r *http.Request) {
	ctx, cancel := context.WithTimeout(r.Context(), 5*time.Minute)
	defer cancel()
	out, err := dockerRun(ctx, "volume", "prune", "-f")
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	lines := strings.Split(strings.TrimSpace(string(out)), "\n")
	log.Printf("docker: volume prune (from %s)", clientIP(r))
	jsonOK(w, 200, map[string]string{"result": lines[len(lines)-1]})
}

// ---- a container's health, its recent life, its console, and its limits ----

// dockerHealth is the health check's verdict and the last few probe results. Docker keeps
// the last five itself, which is all anyone wants: is it failing, and what did it say.
func (s *Server) dockerHealth(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 15*time.Second)
	defer cancel()
	out, err := dockerRun(ctx, "inspect", "--format", "{{json .State}}", id)
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	var st struct {
		Status string `json:"Status"`
		Health *struct {
			Status        string `json:"Status"`
			FailingStreak int    `json:"FailingStreak"`
			Log           []struct {
				Start    time.Time `json:"Start"`
				End      time.Time `json:"End"`
				ExitCode int       `json:"ExitCode"`
				Output   string    `json:"Output"`
			} `json:"Log"`
		} `json:"Health"`
	}
	if json.Unmarshal(out, &st) != nil {
		jsonErr(w, 500, "could not read the container's state")
		return
	}
	if st.Health == nil {
		jsonOK(w, 200, map[string]any{"configured": false, "state": st.Status})
		return
	}
	jsonOK(w, 200, map[string]any{
		"configured": true, "state": st.Status, "status": st.Health.Status,
		"failing_streak": st.Health.FailingStreak, "log": st.Health.Log,
	})
}

// dockerEvents is what has happened to one container lately — started, died, killed, the
// health checks going bad. Asked of Docker on demand rather than kept: Docker already has
// the history, and a copy of it would only go stale.
func (s *Server) dockerEvents(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	since := r.URL.Query().Get("since")
	if !regexp.MustCompile(`^[0-9]+[mhd]$`).MatchString(since) {
		since = "7d"
	}
	ctx, cancel := context.WithTimeout(r.Context(), 30*time.Second)
	defer cancel()
	out, err := dockerRun(ctx, "events", "--since", since, "--until", "0s",
		"--filter", "container="+id, "--format", "{{json .}}")
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	events := []map[string]any{}
	for _, line := range strings.Split(strings.TrimSpace(string(out)), "\n") {
		if !strings.HasPrefix(line, "{") {
			continue
		}
		var e struct {
			Action   string         `json:"Action"`
			Time     int64          `json:"time"`
			TimeNano int64          `json:"timeNano"`
			Actor    map[string]any `json:"Actor"`
		}
		if json.Unmarshal([]byte(line), &e) != nil {
			continue
		}
		m := map[string]any{"action": e.Action, "t": e.Time * 1000}
		if e.TimeNano > 0 {
			m["t"] = e.TimeNano / 1e6
		}
		if at, ok := e.Actor["Attributes"].(map[string]any); ok {
			// the interesting ones: why it stopped, and what the health check said
			for _, k := range []string{"exitCode", "signal", "name", "image"} {
				if v, ok := at[k]; ok && v != "" {
					m[k] = v
				}
			}
		}
		events = append(events, m)
	}
	// newest first, and bounded: a flapping container can produce thousands
	for i, j := 0, len(events)-1; i < j; i, j = i+1, j-1 {
		events[i], events[j] = events[j], events[i]
	}
	if len(events) > 400 {
		events = events[:400]
	}
	jsonOK(w, 200, map[string]any{"events": events, "since": since})
}

// dockerExec opens a shell inside a container, as a terminal session like any other — so
// it reattaches, resizes and scrolls back the same way, and works on a managed machine
// through the same proxy.
func (s *Server) dockerExec(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	tool := dockerTool()
	if tool == "" {
		jsonErr(w, 500, "neither docker nor podman is installed here")
		return
	}
	var body struct {
		Shell string `json:"shell"`
	}
	_ = json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body)
	shell := strings.TrimSpace(body.Shell)
	if shell == "" {
		// most images have sh; bash is nicer where it exists, so try it first and fall back
		shell = "if command -v bash >/dev/null 2>&1; then exec bash; else exec sh; fi"
	} else if strings.ContainsAny(shell, "\n\r") {
		jsonErr(w, 400, "bad shell")
		return
	}
	argv := []string{tool, "exec", "-it", id, "sh", "-c", shell}
	if priv.Available() && !dockerReachable() {
		argv = priv.Wrap(argv...)
	}
	sess, err := s.mgr.CreateCmd("", "sh "+shortName(id), argv)
	if err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	log.Printf("docker exec into %s as terminal %s by %s", id, sess.ID, clientIP(r))
	jsonOK(w, 200, map[string]string{"id": sess.ID, "name": "sh " + shortName(id)})
}

func shortName(id string) string {
	if len(id) > 20 {
		return id[:20]
	}
	return id
}

// dockerReachable reports whether docker answers without help.
func dockerReachable() bool {
	tool := dockerTool()
	if tool == "" {
		return false
	}
	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
	defer cancel()
	return exec.CommandContext(ctx, tool, "version", "--format", "{{.Server.Version}}").Run() == nil
}

// dockerUpdate changes a running container's limits and restart policy. `docker update`
// does this live, without recreating it and without touching the compose file — which is
// also the catch, and the caller is told: a deploy from compose puts back whatever compose
// says.
func (s *Server) dockerUpdate(w http.ResponseWriter, r *http.Request) {
	id := r.PathValue("id")
	if !dockerRef.MatchString(id) {
		jsonErr(w, 400, "bad container")
		return
	}
	var body struct {
		Memory  string `json:"memory"`  // "512m", "2g", or "" to leave alone; "0" removes the limit
		CPUs    string `json:"cpus"`    // "1.5", or "0" for no limit
		Restart string `json:"restart"` // no | on-failure | unless-stopped | always
	}
	if err := json.NewDecoder(io.LimitReader(r.Body, 4096)).Decode(&body); err != nil {
		jsonErr(w, 400, "bad request")
		return
	}
	args := []string{"update"}
	if m := strings.TrimSpace(body.Memory); m != "" {
		if !regexp.MustCompile(`^[0-9]+([bkmgBKMG])?$`).MatchString(m) {
			jsonErr(w, 400, "memory: a number, optionally with b, k, m or g — 0 removes the limit")
			return
		}
		args = append(args, "--memory", m, "--memory-swap", m)
	}
	if c := strings.TrimSpace(body.CPUs); c != "" {
		if !regexp.MustCompile(`^[0-9]+(\.[0-9]+)?$`).MatchString(c) {
			jsonErr(w, 400, "cpus: a number like 1.5 — 0 removes the limit")
			return
		}
		args = append(args, "--cpus", c)
	}
	if p := strings.TrimSpace(body.Restart); p != "" {
		switch p {
		case "no", "on-failure", "unless-stopped", "always":
		default:
			jsonErr(w, 400, "restart: no, on-failure, unless-stopped or always")
			return
		}
		args = append(args, "--restart", p)
	}
	if len(args) == 1 {
		jsonErr(w, 400, "nothing to change")
		return
	}
	ctx, cancel := context.WithTimeout(r.Context(), 30*time.Second)
	defer cancel()
	if _, err := dockerRun(ctx, append(args, id)...); err != nil {
		jsonErr(w, 500, err.Error())
		return
	}
	log.Printf("docker update %s %v by %s", id, args[1:], clientIP(r))
	w.WriteHeader(204)
}