Initial commit of Descrybe v2 without local scratch artifacts.
Drop one-shot tmp/axe scripts and agent i18n scratch so the Gitea tree is deployable.
This commit is contained in:
@@ -0,0 +1,65 @@
|
||||
package processing
|
||||
|
||||
import (
|
||||
"crypto/sha256"
|
||||
"encoding/hex"
|
||||
"encoding/json"
|
||||
|
||||
"github.com/descrybe/descrybe-v2/apps/api/internal/company"
|
||||
)
|
||||
|
||||
// FieldEnhanceInputHash is stored on processed_products.field_sources so re-runs
|
||||
// can skip the LLM when enhance inputs are unchanged (mirrors feed content_hash).
|
||||
const FieldEnhanceInputHash = "enhance_input_hash"
|
||||
|
||||
// enhanceInputHashVersion bumps when the enhance prompt/schema changes so prior
|
||||
// hashes are invalidated and products re-enhance once.
|
||||
const enhanceInputHashVersion = "3"
|
||||
|
||||
// HashEnhanceInput returns a stable hex SHA-256 of the inputs that feed the
|
||||
// enhance LLM (same compaction as ProductEnhanceUser / CompactBrandPrompt).
|
||||
// Empty inputs still produce a deterministic hash.
|
||||
func HashEnhanceInput(category, name, description, brandPrompt, language, promptSystem, promptUser string, attrs map[string]any) string {
|
||||
langCode, err := company.ParseLanguage(language, true)
|
||||
if err != nil {
|
||||
langCode = company.DefaultLanguage
|
||||
}
|
||||
payload := map[string]any{
|
||||
"v": enhanceInputHashVersion,
|
||||
"category": SanitizeText(category),
|
||||
"name": SanitizeText(truncateRunes(name, 200)),
|
||||
"description": SanitizeText(truncateRunes(description, MaxProductDescRunes)),
|
||||
"brand_prompt": CompactBrandPrompt(brandPrompt),
|
||||
"language": langCode,
|
||||
"prompt_system": SanitizeText(truncateRunes(promptSystem, 4000)),
|
||||
"prompt_user": SanitizeText(truncateRunes(promptUser, 4000)),
|
||||
"attrs": CompactAttrs(attrs, MaxAttrKeys),
|
||||
}
|
||||
b, err := json.Marshal(payload)
|
||||
if err != nil {
|
||||
// Unreachable for map[string]any of strings/scalars; fall back so callers
|
||||
// never skip LLM on a broken hash.
|
||||
sum := sha256.Sum256([]byte(category + "\x00" + name + "\x00" + description))
|
||||
return hex.EncodeToString(sum[:])
|
||||
}
|
||||
sum := sha256.Sum256(b)
|
||||
return hex.EncodeToString(sum[:])
|
||||
}
|
||||
|
||||
func enhanceHashFromMeta(raw any) string {
|
||||
m, ok := raw.(map[string]any)
|
||||
if !ok {
|
||||
return ""
|
||||
}
|
||||
h, _ := m["input_hash"].(string)
|
||||
return h
|
||||
}
|
||||
|
||||
func enhanceStatusFromMeta(raw any) string {
|
||||
m, ok := raw.(map[string]any)
|
||||
if !ok {
|
||||
return ""
|
||||
}
|
||||
s, _ := m["status"].(string)
|
||||
return s
|
||||
}
|
||||
Reference in New Issue
Block a user