patrick
9a24ea29e1
Git-Repository für bestehenden archivdms-Code initialisiert, Branch-/Commit-Konvention (feature/<ticket>-<slug>-Branches, Ticket-Prefix in Commit-Nachricht) etabliert.
65 lines
2.8 KiB
Go
65 lines
2.8 KiB
Go
package storage
|
|
|
|
import (
|
|
"context"
|
|
"fmt"
|
|
)
|
|
|
|
// initMLClassifierSchema creates the Naive-Bayes retraining/classification
|
|
// tables (ml_classifier_tokens/ml_classifier_classes/ml_classifier_runs) plus
|
|
// the assigned_via provenance columns on document_tags/documents that let the
|
|
// classifier tell apart manually-set, rule-engine-set and ML-accepted
|
|
// assignments. Idempotent, called from (*Store).initSchema AFTER
|
|
// initTaxonomySchema (depends on document_types/correspondents/tags/
|
|
// document_tags/documents existing). See migrations/020_ml_classifier.sql.
|
|
func (s *Store) initMLClassifierSchema(ctx context.Context) error {
|
|
_, err := s.db.Exec(ctx, `
|
|
CREATE TABLE IF NOT EXISTS ml_classifier_tokens (
|
|
id BIGSERIAL PRIMARY KEY,
|
|
tenant_id BIGINT NOT NULL,
|
|
kind TEXT NOT NULL CHECK (kind IN ('document_types','correspondents','tags')),
|
|
entity_id BIGINT NOT NULL,
|
|
token TEXT NOT NULL,
|
|
count BIGINT NOT NULL DEFAULT 0,
|
|
UNIQUE (tenant_id, kind, entity_id, token)
|
|
);
|
|
CREATE INDEX IF NOT EXISTS idx_ml_tokens_lookup ON ml_classifier_tokens(tenant_id, kind, token);
|
|
|
|
CREATE TABLE IF NOT EXISTS ml_classifier_classes (
|
|
tenant_id BIGINT NOT NULL,
|
|
kind TEXT NOT NULL,
|
|
entity_id BIGINT NOT NULL,
|
|
doc_count BIGINT NOT NULL DEFAULT 0,
|
|
total_tokens BIGINT NOT NULL DEFAULT 0,
|
|
PRIMARY KEY (tenant_id, kind, entity_id)
|
|
);
|
|
|
|
CREATE TABLE IF NOT EXISTS ml_classifier_runs (
|
|
id BIGSERIAL PRIMARY KEY,
|
|
tenant_id BIGINT NOT NULL,
|
|
started_at TIMESTAMPTZ NOT NULL DEFAULT now(),
|
|
completed_at TIMESTAMPTZ,
|
|
doc_count BIGINT NOT NULL DEFAULT 0,
|
|
status TEXT NOT NULL DEFAULT 'running' CHECK (status IN ('running','completed','failed','skipped_insufficient_data')),
|
|
error TEXT
|
|
);
|
|
CREATE INDEX IF NOT EXISTS idx_ml_classifier_runs_tenant ON ml_classifier_runs(tenant_id);
|
|
`)
|
|
if err != nil {
|
|
return fmt.Errorf("storage: create ml_classifier tables: %w", err)
|
|
}
|
|
|
|
// Provenance columns: default 'manual' so existing rows are NOT retroactively
|
|
// (mis)classified as rule-assigned — conservative default keeps them out of
|
|
// automatic ML-accepted reclassification too.
|
|
_, err = s.db.Exec(ctx, `
|
|
ALTER TABLE document_tags ADD COLUMN IF NOT EXISTS assigned_via TEXT NOT NULL DEFAULT 'manual' CHECK (assigned_via IN ('manual','rule','ml_accepted'));
|
|
ALTER TABLE documents ADD COLUMN IF NOT EXISTS doc_type_assigned_via TEXT NOT NULL DEFAULT 'manual' CHECK (doc_type_assigned_via IN ('manual','rule','ml_accepted'));
|
|
ALTER TABLE documents ADD COLUMN IF NOT EXISTS correspondent_assigned_via TEXT NOT NULL DEFAULT 'manual' CHECK (correspondent_assigned_via IN ('manual','rule','ml_accepted'));
|
|
`)
|
|
if err != nil {
|
|
return fmt.Errorf("storage: alter documents/document_tags for ml_classifier provenance: %w", err)
|
|
}
|
|
return nil
|
|
}
|