feat: add embeds metadata

This commit is contained in:
hubert.lenoir
2026-04-03 18:23:27 +02:00
committed by Julien Neuhart
parent eff9444294
commit 3187980ead
24 changed files with 848 additions and 24 deletions

View File

@@ -22,6 +22,7 @@ type multiPdfEngines struct {
writeMetadataEngines []gotenberg.PdfEngine
passwordEngines []gotenberg.PdfEngine
embedEngines []gotenberg.PdfEngine
embedMetadataEngines []gotenberg.PdfEngine
readBookmarksEngines []gotenberg.PdfEngine
writeBookmarksEngines []gotenberg.PdfEngine
watermarkEngines []gotenberg.PdfEngine
@@ -38,6 +39,7 @@ func newMultiPdfEngines(
writeMetadataEngines,
passwordEngines,
embedEngines,
embedMetadataEngines,
readBookmarksEngines,
writeBookmarksEngines,
watermarkEngines,
@@ -53,6 +55,7 @@ func newMultiPdfEngines(
writeMetadataEngines: writeMetadataEngines,
passwordEngines: passwordEngines,
embedEngines: embedEngines,
embedMetadataEngines: embedMetadataEngines,
readBookmarksEngines: readBookmarksEngines,
writeBookmarksEngines: writeBookmarksEngines,
watermarkEngines: watermarkEngines,
@@ -603,6 +606,43 @@ func (multi *multiPdfEngines) Rotate(ctx context.Context, logger *slog.Logger, i
return err
}
// EmbedFilesMetadata sets metadata on embedded files using the first available
// engine that supports it.
//
//nolint:dupl
func (multi *multiPdfEngines) EmbedFilesMetadata(ctx context.Context, logger *slog.Logger, metadata map[string]map[string]string, inputPath string) error {
tracer := gotenberg.Tracer()
ctx, span := tracer.Start(ctx, "pdfengines.EmbedFilesMetadata", trace.WithSpanKind(trace.SpanKindInternal))
defer span.End()
var err error
errChan := make(chan error, 1)
for _, engine := range multi.embedMetadataEngines {
go func(engine gotenberg.PdfEngine) {
errChan <- engine.EmbedFilesMetadata(ctx, logger, metadata, inputPath)
}(engine)
select {
case setErr := <-errChan:
if setErr != nil {
err = errors.Join(err, setErr)
} else {
span.SetStatus(codes.Ok, "")
return nil
}
case <-ctx.Done():
return ctx.Err()
}
}
err = fmt.Errorf("set embeds metadata using multi PDF engines: %w", err)
span.RecordError(err)
span.SetStatus(codes.Error, err.Error())
return err
}
// Interface guards.
var (
_ gotenberg.PdfEngine = (*multiPdfEngines)(nil)

View File

@@ -36,6 +36,7 @@ type PdfEngines struct {
writeMetadataNames []string
encryptNames []string
embedNames []string
embedMetadataNames []string
readBookmarksNames []string
writeBookmarksNames []string
watermarkNames []string
@@ -59,6 +60,7 @@ func (mod *PdfEngines) Descriptor() gotenberg.ModuleDescriptor {
fs.StringSlice("pdfengines-write-metadata-engines", []string{"exiftool"}, "Set the PDF engines and their order for the write metadata feature - empty means all")
fs.StringSlice("pdfengines-encrypt-engines", []string{"qpdf", "pdftk", "pdfcpu"}, "Set the PDF engines and their order for the password protection feature - empty means all")
fs.StringSlice("pdfengines-embed-engines", []string{"pdfcpu"}, "Set the PDF engines and their order for the file embedding feature - empty means all")
fs.StringSlice("pdfengines-embed-metadata-engines", []string{"qpdf"}, "Set the PDF engines and their order for the embed metadata feature - empty means all")
fs.StringSlice("pdfengines-read-bookmarks-engines", []string{"pdfcpu"}, "Set the PDF engines and their order for the read bookmarks feature - empty means all")
fs.StringSlice("pdfengines-write-bookmarks-engines", []string{"pdfcpu"}, "Set the PDF engines and their order for the write bookmarks feature - empty means all")
fs.StringSlice("pdfengines-watermark-engines", []string{"pdfcpu", "pdftk"}, "Set the PDF engines and their order for the watermark feature - empty means all")
@@ -91,6 +93,7 @@ func (mod *PdfEngines) Provision(ctx *gotenberg.Context) error {
writeMetadataNames := flags.MustStringSlice("pdfengines-write-metadata-engines")
encryptNames := flags.MustStringSlice("pdfengines-encrypt-engines")
embedNames := flags.MustStringSlice("pdfengines-embed-engines")
embedMetadataNames := flags.MustStringSlice("pdfengines-embed-metadata-engines")
readBookmarksNames := flags.MustStringSlice("pdfengines-read-bookmarks-engines")
writeBookmarksNames := flags.MustStringSlice("pdfengines-write-bookmarks-engines")
watermarkNames := flags.MustStringSlice("pdfengines-watermark-engines")
@@ -162,6 +165,11 @@ func (mod *PdfEngines) Provision(ctx *gotenberg.Context) error {
mod.embedNames = embedNames
}
mod.embedMetadataNames = defaultNames
if len(embedMetadataNames) > 0 {
mod.embedMetadataNames = embedMetadataNames
}
mod.readBookmarksNames = defaultNames
if len(readBookmarksNames) > 0 {
mod.readBookmarksNames = readBookmarksNames
@@ -236,6 +244,7 @@ func (mod *PdfEngines) Validate() error {
findNonExistingEngines(mod.writeMetadataNames)
findNonExistingEngines(mod.encryptNames)
findNonExistingEngines(mod.embedNames)
findNonExistingEngines(mod.embedMetadataNames)
findNonExistingEngines(mod.readBookmarksNames)
findNonExistingEngines(mod.writeBookmarksNames)
findNonExistingEngines(mod.watermarkNames)
@@ -261,6 +270,7 @@ func (mod *PdfEngines) SystemMessages() []string {
fmt.Sprintf("write metadata engines - %s", strings.Join(mod.writeMetadataNames, " ")),
fmt.Sprintf("encrypt engines - %s", strings.Join(mod.encryptNames, " ")),
fmt.Sprintf("embed engines - %s", strings.Join(mod.embedNames, " ")),
fmt.Sprintf("embed metadata engines - %s", strings.Join(mod.embedMetadataNames, " ")),
fmt.Sprintf("read bookmarks engines - %s", strings.Join(mod.readBookmarksNames, " ")),
fmt.Sprintf("write bookmarks engines - %s", strings.Join(mod.writeBookmarksNames, " ")),
fmt.Sprintf("watermark engines - %s", strings.Join(mod.watermarkNames, " ")),
@@ -294,6 +304,7 @@ func (mod *PdfEngines) PdfEngine() (gotenberg.PdfEngine, error) {
engines(mod.writeMetadataNames),
engines(mod.encryptNames),
engines(mod.embedNames),
engines(mod.embedMetadataNames),
engines(mod.readBookmarksNames),
engines(mod.writeBookmarksNames),
engines(mod.watermarkNames),

View File

@@ -443,6 +443,30 @@ func FormDataPdfEmbeds(form *api.FormData) []string {
return embedPaths
}
// FormDataPdfEmbedsMetadata extracts embeds metadata from form data.
// The "embedsMetadata" field is a JSON string keyed by filename.
func FormDataPdfEmbedsMetadata(form *api.FormData) map[string]map[string]string {
var metadata map[string]map[string]string
form.EmbedsMetadata(&metadata)
return metadata
}
// EmbedFilesMetadataStub sets metadata on embedded files in PDFs.
func EmbedFilesMetadataStub(ctx *api.Context, engine gotenberg.PdfEngine, metadata map[string]map[string]string, inputPaths []string) error {
if len(metadata) == 0 {
return nil
}
for _, inputPath := range inputPaths {
err := engine.EmbedFilesMetadata(ctx, ctx.Log(), metadata, inputPath)
if err != nil {
return fmt.Errorf("set embeds metadata on PDF '%s': %w", inputPath, err)
}
}
return nil
}
// FormDataPdfEncrypt extracts encryption parameters from form data.
func FormDataPdfEncrypt(form *api.FormData) (userPassword, ownerPassword string) {
form.String("userPassword", &userPassword, "")
@@ -638,6 +662,7 @@ func mergeRoute(engine gotenberg.PdfEngine) api.Route {
stamp := FormDataPdfStamp(form, false)
stampFile := FormDataPdfStampFile(form)
angle, rotatePages := FormDataPdfRotate(form, false)
embedsMetadata := FormDataPdfEmbedsMetadata(form)
var inputPaths []string
var flatten bool
@@ -754,6 +779,11 @@ func mergeRoute(engine gotenberg.PdfEngine) api.Route {
return fmt.Errorf("embed files into PDFs: %w", err)
}
err = EmbedFilesMetadataStub(ctx, engine, embedsMetadata, outputPaths)
if err != nil {
return fmt.Errorf("set embeds metadata: %w", err)
}
err = EncryptPdfStub(ctx, engine, userPassword, ownerPassword, outputPaths)
if err != nil {
return fmt.Errorf("encrypt PDFs: %w", err)
@@ -789,6 +819,7 @@ func splitRoute(engine gotenberg.PdfEngine) api.Route {
stamp := FormDataPdfStamp(form, false)
stampFile := FormDataPdfStampFile(form)
angle, rotatePages := FormDataPdfRotate(form, false)
embedsMetadata := FormDataPdfEmbedsMetadata(form)
var inputPaths []string
var flatten bool
@@ -856,6 +887,11 @@ func splitRoute(engine gotenberg.PdfEngine) api.Route {
return fmt.Errorf("embed files into PDFs: %w", err)
}
err = EmbedFilesMetadataStub(ctx, engine, embedsMetadata, convertOutputPaths)
if err != nil {
return fmt.Errorf("set embeds metadata: %w", err)
}
err = EncryptPdfStub(ctx, engine, userPassword, ownerPassword, convertOutputPaths)
if err != nil {
return fmt.Errorf("encrypt PDFs: %w", err)
@@ -1180,6 +1216,7 @@ func embedRoute(engine gotenberg.PdfEngine) api.Route {
form := ctx.FormData()
embedPaths := FormDataPdfEmbeds(form)
embedsMetadata := FormDataPdfEmbedsMetadata(form)
var inputPaths []string
err := form.
@@ -1193,6 +1230,11 @@ func embedRoute(engine gotenberg.PdfEngine) api.Route {
return fmt.Errorf("embed files into PDFs: %w", err)
}
err = EmbedFilesMetadataStub(ctx, engine, embedsMetadata, inputPaths)
if err != nil {
return fmt.Errorf("set embeds metadata: %w", err)
}
err = ctx.AddOutputPaths(inputPaths...)
if err != nil {
return fmt.Errorf("add output paths: %w", err)