From 41c73fecd5955b5150d7b68e6a236ce7e7c4cb75 Mon Sep 17 00:00:00 2001 From: gm Date: Fri, 5 Jun 2026 17:53:25 +0900 Subject: [PATCH] =?UTF-8?q?ui:=20structured=20record=20collection=20?= =?UTF-8?q?=E2=80=94=20collect=20node=20schema=20parsing=20(v0.9)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 수집 노드의 하위 노드를 레코드 필드 스키마로 인식: - buildSchemaMap(): 수집 섹션의 child sections → 파싱 마크 맵 - parseRecordLine(): 한 줄을 inline 마크 기준으로 분리 → {mark: value} 레코드 - parseInstance(): 스키마 있으면 레코드 파싱, 없으면 기존 토큰 분리 - mergedToText(): 레코드 → inline 재구성 (✅김양모⌚12시), 단순 → per_line 토큰 - mergedToJSON(): 레코드 → 첫 필드=부모SECTION, 나머지=자식SECTION 계층 구조 - renderFillerTemplateView(): 스키마 필드를 칩 안에 인라인 배지로 표시 Co-Authored-By: Claude Sonnet 4.6 --- ui/builder.html | 134 ++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 112 insertions(+), 22 deletions(-) diff --git a/ui/builder.html b/ui/builder.html index 5faebfd..6e6835f 100644 --- a/ui/builder.html +++ b/ui/builder.html @@ -286,7 +286,7 @@ input,textarea,button,select{font-family:inherit}
🏗️

CTStruct

- v0.8 + v0.9