Files
ImaGen/cmd/imagen/generate.go
mAi 237270b204 mAi: #211 - bootstrap ImaGen framework skeleton
First step of the model-agnostic image-generation framework. Lands the
plumbing other components (skill, ComfyUI/Replicate adapters, agents)
will plug into:

- internal/backend: Backend interface (Request/Result), thread-safe
  Registry with init-time Register, plus a Mock reference adapter that
  emits a deterministic gradient PNG for smoke tests.
- internal/config: YAML loader for ~/.config/imagen.yaml. Framework owns
  default_backend + output settings + a per-backend block; each adapter
  owns the schema below its own block via BackendSpec.Raw.
- internal/output: filename templating ({date}/{time}/{slug}/{seed}/
  {backend}/{ext}), JSON metadata sidecar, --output override path.
- internal/prompt: embedded styles.yaml, style-preset suffix application.
- internal/server: 501 stub — HTTP surface lands in a follow-up issue.
- cmd/imagen: generate / backends / config (init|validate|path) / serve
  / version subcommands. Stdlib-only flag parsing with a small helper to
  honour positional prompt args ahead of flags (matches the issue spec).
- Tests for output (slug, naming template, sidecar), backend (mock PNG
  validity + determinism, registry build + duplicate panic), config
  (round-trip + validation), prompt (style apply + unknown-style error).
- CLAUDE.md, README.md, docs/architecture.md, docs/usage.md, Makefile.

Acceptance criteria from #211:
1. go build ./... — clean
2. imagen backends — lists registered backends, exits 0
3. imagen generate "test prompt" --backend mock --output /tmp/x.png —
   writes a 1024x1024 PNG plus an x.png.json sidecar
4. imagen config init | imagen config validate — round-trips cleanly
5. CLAUDE.md "Adding a new adapter" — six-step recipe
2026-05-08 14:37:05 +02:00

211 lines
5.4 KiB
Go

package main
import (
"context"
"flag"
"fmt"
"os"
"strconv"
"strings"
"mgit.msbls.de/m/ImaGen/internal/backend"
"mgit.msbls.de/m/ImaGen/internal/config"
"mgit.msbls.de/m/ImaGen/internal/output"
"mgit.msbls.de/m/ImaGen/internal/prompt"
)
func runGenerate(ctx context.Context, args []string) error {
fs := flag.NewFlagSet("generate", flag.ContinueOnError)
var (
backendName string
size string
outPath string
seed int64
steps int
style string
negative string
configPath string
noSidecar bool
)
fs.StringVar(&backendName, "backend", "", "backend instance name (default: config.default_backend)")
fs.StringVar(&size, "size", "1024x1024", "WxH, e.g. 1024x1024")
fs.StringVar(&outPath, "output", "", "explicit output path (overrides config naming template)")
fs.Int64Var(&seed, "seed", 0, "deterministic seed (0 = backend default)")
fs.IntVar(&steps, "steps", 0, "diffusion steps (0 = backend default)")
fs.StringVar(&style, "style", "", "style preset name (see imagen config init for the list)")
fs.StringVar(&negative, "negative", "", "negative prompt (ignored by backends that don't support it)")
fs.StringVar(&configPath, "config", "", "config file path (default: ~/.config/imagen.yaml)")
fs.BoolVar(&noSidecar, "no-sidecar", false, "skip the JSON sidecar even if config enables it")
fs.Usage = func() {
fmt.Fprintln(fs.Output(), `Usage: imagen generate "<prompt>" [flags]`)
fs.PrintDefaults()
}
// stdlib flag stops parsing at the first non-flag arg, so split the
// prompt (leading positional args) from the flags ourselves before parsing.
leadingPositional, flagArgs := splitLeadingPositional(args)
if err := fs.Parse(flagArgs); err != nil {
return err
}
positional := append(leadingPositional, fs.Args()...)
if len(positional) == 0 {
fs.Usage()
return userErr("missing prompt")
}
rawPrompt := strings.Join(positional, " ")
w, h, err := parseSize(size)
if err != nil {
return userErr("bad --size: %v", err)
}
cfg, cfgErr := config.Load(configPath)
if cfgErr != nil && !os.IsNotExist(cfgErr) {
return cfgErr
}
if backendName == "" {
if cfg != nil {
backendName = cfg.DefaultBackend
}
}
if backendName == "" {
return userErr("no --backend given and no default_backend in config")
}
be, err := buildBackend(cfg, backendName)
if err != nil {
return err
}
finalPrompt, err := prompt.Apply(rawPrompt, style)
if err != nil {
return userErr("%v", err)
}
req := backend.Request{
Prompt: finalPrompt,
NegativePrompt: negative,
Width: w,
Height: h,
Steps: steps,
Seed: seed,
Style: style,
}
res, err := be.Generate(ctx, req)
if err != nil {
return fmt.Errorf("backend %q: %w", backendName, err)
}
defer res.ImageReader.Close()
writer := buildWriter(cfg, noSidecar)
in := output.Inputs{
Prompt: rawPrompt,
Backend: be.Name(),
Seed: seedFromMetadata(res.Metadata, seed),
Ext: extFromMime(res.MimeType),
Metadata: res.Metadata,
}
var paths *output.Outputs
if outPath != "" {
paths, err = writer.WriteToPath(res.ImageReader, outPath, in)
} else {
paths, err = writer.Write(res.ImageReader, in)
}
if err != nil {
return err
}
fmt.Println(paths.ImagePath)
if paths.SidecarPath != "" {
fmt.Fprintln(os.Stderr, "sidecar:", paths.SidecarPath)
}
return nil
}
// splitLeadingPositional separates the positional args at the start of args
// from the rest (which begins with the first flag). A literal "--" terminator
// pushes everything after it into the positional list and out of flag parsing.
func splitLeadingPositional(args []string) (positional, flags []string) {
for i, a := range args {
if a == "--" {
return append(positional, args[i+1:]...), flags
}
if strings.HasPrefix(a, "-") {
return positional, args[i:]
}
positional = append(positional, a)
}
return positional, flags
}
func parseSize(s string) (int, int, error) {
parts := strings.SplitN(s, "x", 2)
if len(parts) != 2 {
return 0, 0, fmt.Errorf("expected WxH, got %q", s)
}
w, err := strconv.Atoi(parts[0])
if err != nil {
return 0, 0, err
}
h, err := strconv.Atoi(parts[1])
if err != nil {
return 0, 0, err
}
return w, h, nil
}
func buildBackend(cfg *config.Config, name string) (backend.Backend, error) {
if cfg != nil {
spec, ok := cfg.Backends[name]
if ok {
return backend.Default.Build(spec.Type, name, spec.Raw)
}
}
if backend.Default.Has(name) {
return backend.Default.Build(name, name, nil)
}
return nil, userErr("backend %q not found in config and not a registered type (registered types: %v)",
name, backend.Default.Types())
}
func buildWriter(cfg *config.Config, noSidecar bool) *output.Writer {
w := &output.Writer{}
if cfg != nil {
w.Directory = config.ExpandPath(cfg.Output.Directory)
w.NameTemplate = cfg.Output.Naming
w.WriteSidecar = cfg.Output.WriteMetadataJSON
}
if w.Directory == "" {
w.Directory = "."
}
if noSidecar {
w.WriteSidecar = false
}
return w
}
func seedFromMetadata(meta map[string]any, fallback int64) int64 {
if v, ok := meta["seed"]; ok {
switch n := v.(type) {
case int64:
return n
case int:
return int64(n)
case float64:
return int64(n)
}
}
return fallback
}
func extFromMime(mime string) string {
switch mime {
case "image/png", "":
return "png"
case "image/jpeg":
return "jpg"
case "image/webp":
return "webp"
}
return "bin"
}