412821ec56
- jz_bm_material_file_texts caches the pdftotext output per file - the River job jz.bm.extract_material_text extracts and resyncs - the management multipart attach queues it inside its transaction - file_texts in the document and a file snippet with its name
48 lines
1.5 KiB
Go
48 lines
1.5 KiB
Go
package bm
|
|
|
|
import (
|
|
"context"
|
|
"errors"
|
|
"sync"
|
|
|
|
"git.golem15.com/golem15/summercms/modules/diary"
|
|
"git.golem15.com/jakub/sm-bm-plugin/classes"
|
|
"git.golem15.com/jakub/sm-bm-plugin/jobs"
|
|
"gocloud.dev/blob"
|
|
"gorm.io/gorm"
|
|
)
|
|
|
|
// pdfToolWarning is logged at most once per process when pdftotext is
|
|
// missing.
|
|
var pdfToolWarning sync.Once
|
|
|
|
// extractMaterialText runs one jz.bm.extract_material_text job: the text of
|
|
// the material's PDFs that have none yet is extracted and stored, then the
|
|
// material's search document is synced. A missing database or bucket is an
|
|
// error, so River retries; a missing pdftotext is logged once and the
|
|
// document is still synced.
|
|
func (p *Plugin) extractMaterialText(ctx context.Context, args jobs.ExtractMaterialTextArgs) error {
|
|
if p.app == nil {
|
|
return errors.New("jz.bm: app is not booted")
|
|
}
|
|
db, ok := p.app.Lookup[*gorm.DB]()
|
|
if !ok || db == nil {
|
|
return errors.New("jz.bm: database is not configured")
|
|
}
|
|
bucket, ok := p.app.Lookup[*blob.Bucket]()
|
|
if !ok || bucket == nil {
|
|
return errors.New("jz.bm: the uploads bucket is not configured")
|
|
}
|
|
_, err := classes.EnsureFileTexts(ctx, db, bucket, classes.DefaultPDFText(), args.MaterialID)
|
|
switch {
|
|
case errors.Is(err, classes.ErrPDFToolMissing):
|
|
pdfToolWarning.Do(func() {
|
|
diary.For("jz.bm").Warn("pdftotext not found; attached PDFs add no search text until poppler-utils is installed")
|
|
})
|
|
case err != nil:
|
|
return err
|
|
}
|
|
p.materialIndex.Sync(ctx, db, args.MaterialID)
|
|
return nil
|
|
}
|