feat: 자재 분류 시스템 대폭 개선
🔧 주요 개선사항: - EXCLUDE 분류기 추가 (WELD GAP 등 제외 대상 처리) - FITTING 분류기 키워드 확장 (ELL, RED 추가) - PIPE 재질 중복 문제 해결 (material_grade 파싱 개선) - NIPPLE 특별 처리 추가 (스케줄 + 길이 정보 포함) - OLET 타입 중복 표시 제거 📊 분류 정확도: - UNKNOWN: 0개 (100% 분류 성공) - EXCLUDE: 1,014개 (제외 대상) - 실제 자재: 1,823개 정확 분류 🎯 해결된 문제: - PIPE 재질 'ASTM A106 ASTM A106' → 'ASTM A106 GR B' - WELD GAP 오분류 → EXCLUDE 카테고리 - FITTING 키워드 인식 실패 → ELL, RED 키워드 추가 - 프론트엔드 중복 표시 제거
This commit is contained in:
Hyungi Ahn
@@ -104,13 +104,19 @@ def parse_dataframe(df):
|
||||
|
||||
material_grade = ""
|
||||
if "ASTM" in description.upper():
|
||||
astm_match = re.search(r'ASTM\s+([A-Z0-9\s]+)', description.upper())
|
||||
# ASTM 표준과 등급만 추출, end_preparation(BOE, POE, BBE 등)은 제외
|
||||
astm_match = re.search(r'ASTM\s+([A-Z0-9]+(?:\s+GR\s+[A-Z0-9]+)?)', description.upper())
|
||||
if astm_match:
|
||||
material_grade = astm_match.group(0).strip()
|
||||
|
||||
main_size = str(row.get(mapped_columns.get('main_size', ''), ''))
|
||||
red_size = str(row.get(mapped_columns.get('red_size', ''), ''))
|
||||
|
||||
# main_nom과 red_nom 별도 저장 (원본 값 유지)
|
||||
main_nom = main_size if main_size != 'nan' and main_size != '' else None
|
||||
red_nom = red_size if red_size != 'nan' and red_size != '' else None
|
||||
|
||||
# 기존 size_spec도 유지 (호환성을 위해)
|
||||
if main_size != 'nan' and red_size != 'nan' and red_size != '':
|
||||
size_spec = f"{main_size} x {red_size}"
|
||||
elif main_size != 'nan' and main_size != '':
|
||||
@@ -133,6 +139,8 @@ def parse_dataframe(df):
|
||||
'quantity': quantity,
|
||||
'unit': "EA",
|
||||
'size_spec': size_spec,
|
||||
'main_nom': main_nom, # 추가
|
||||
'red_nom': red_nom, # 추가
|
||||
'material_grade': material_grade,
|
||||
'length': length_value,
|
||||
'line_number': index + 1,
|
||||
@@ -230,32 +238,60 @@ async def upload_file(
|
||||
except (ValueError, TypeError):
|
||||
length_value = None
|
||||
|
||||
classification_result = classify_pipe("", description, size_spec, length_value)
|
||||
print(f"PIPE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
# main_nom과 red_nom 추출
|
||||
main_nom = material_data.get("main_nom")
|
||||
red_nom = material_data.get("red_nom")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_fitting("", description, size_spec)
|
||||
print(f"FITTING 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_valve("", description, size_spec)
|
||||
print(f"VALVE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_flange("", description, size_spec)
|
||||
print(f"FLANGE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_bolt("", description, size_spec)
|
||||
print(f"BOLT 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_gasket("", description, size_spec)
|
||||
print(f"GASKET 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
classification_result = classify_instrument("", description, size_spec)
|
||||
print(f"INSTRUMENT 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
classification_result = None
|
||||
try:
|
||||
# EXCLUDE 분류기 우선 호출 (제외 대상 먼저 걸러냄)
|
||||
from app.services.exclude_classifier import classify_exclude
|
||||
classification_result = classify_exclude("", description, main_nom or "")
|
||||
print(f"EXCLUDE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 파이프 분류기 호출
|
||||
classification_result = classify_pipe("", description, main_nom or "", length_value)
|
||||
print(f"PIPE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 피팅 분류기 호출 (main_nom, red_nom 개별 전달)
|
||||
classification_result = classify_fitting("", description, main_nom or "", red_nom)
|
||||
print(f"FITTING 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 플랜지 분류기 호출 (main_nom, red_nom 개별 전달)
|
||||
classification_result = classify_flange("", description, main_nom or "", red_nom)
|
||||
print(f"FLANGE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 밸브 분류기 호출
|
||||
classification_result = classify_valve("", description, main_nom or "")
|
||||
print(f"VALVE 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 볼트 분류기 호출
|
||||
classification_result = classify_bolt("", description, main_nom or "")
|
||||
print(f"BOLT 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 가스켓 분류기 호출
|
||||
classification_result = classify_gasket("", description, main_nom or "")
|
||||
print(f"GASKET 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
if classification_result.get("overall_confidence", 0) < 0.5:
|
||||
# 계기 분류기 호출
|
||||
classification_result = classify_instrument("", description, main_nom or "")
|
||||
print(f"INSTRUMENT 분류 결과: {classification_result.get('category', 'UNKNOWN')} (신뢰도: {classification_result.get('overall_confidence', 0)})")
|
||||
|
||||
except Exception as e:
|
||||
print(f"분류기 실행 중 오류 발생: {e}")
|
||||
# 기본 분류 결과 생성
|
||||
classification_result = {
|
||||
"category": "UNKNOWN",
|
||||
"overall_confidence": 0.0,
|
||||
"reason": f"분류기 오류: {str(e)}"
|
||||
}
|
||||
|
||||
print(f"최종 분류 결과: {classification_result.get('category', 'UNKNOWN')}")
|
||||
|
||||
@@ -263,13 +299,13 @@ async def upload_file(
|
||||
material_insert_query = text("""
|
||||
INSERT INTO materials (
|
||||
file_id, original_description, quantity, unit, size_spec,
|
||||
material_grade, line_number, row_number, classified_category,
|
||||
classification_confidence, is_verified, created_at
|
||||
main_nom, red_nom, material_grade, line_number, row_number,
|
||||
classified_category, classification_confidence, is_verified, created_at
|
||||
)
|
||||
VALUES (
|
||||
:file_id, :original_description, :quantity, :unit, :size_spec,
|
||||
:material_grade, :line_number, :row_number, :classified_category,
|
||||
:classification_confidence, :is_verified, :created_at
|
||||
:main_nom, :red_nom, :material_grade, :line_number, :row_number,
|
||||
:classified_category, :classification_confidence, :is_verified, :created_at
|
||||
)
|
||||
RETURNING id
|
||||
""")
|
||||
@@ -287,6 +323,8 @@ async def upload_file(
|
||||
"quantity": material_data["quantity"],
|
||||
"unit": material_data["unit"],
|
||||
"size_spec": material_data["size_spec"],
|
||||
"main_nom": material_data.get("main_nom"), # 추가
|
||||
"red_nom": material_data.get("red_nom"), # 추가
|
||||
"material_grade": material_data["material_grade"],
|
||||
"line_number": material_data["line_number"],
|
||||
"row_number": material_data["row_number"],
|
||||
@@ -309,16 +347,11 @@ async def upload_file(
|
||||
# material_id도 함께 저장하도록 수정
|
||||
pipe_detail_insert_query = text("""
|
||||
INSERT INTO pipe_details (
|
||||
material_id, file_id, material_standard, material_grade, material_type,
|
||||
manufacturing_method, end_preparation, schedule, wall_thickness,
|
||||
nominal_size, length_mm, material_confidence, manufacturing_confidence,
|
||||
end_prep_confidence, schedule_confidence
|
||||
)
|
||||
VALUES (
|
||||
:material_id, :file_id, :material_standard, :material_grade, :material_type,
|
||||
:manufacturing_method, :end_preparation, :schedule, :wall_thickness,
|
||||
:nominal_size, :length_mm, :material_confidence, :manufacturing_confidence,
|
||||
:end_prep_confidence, :schedule_confidence
|
||||
material_id, file_id, outer_diameter, schedule,
|
||||
material_spec, manufacturing_method, end_preparation, length_mm
|
||||
) VALUES (
|
||||
:material_id, :file_id, :outer_diameter, :schedule,
|
||||
:material_spec, :manufacturing_method, :end_preparation, :length_mm
|
||||
)
|
||||
""")
|
||||
|
||||
@@ -329,25 +362,98 @@ async def upload_file(
|
||||
schedule_info = classification_result.get("schedule", {})
|
||||
size_info = classification_result.get("size_info", {})
|
||||
|
||||
# main_nom을 outer_diameter로 활용
|
||||
outer_diameter = material_data.get("main_nom") or material_data.get("size_spec", "")
|
||||
|
||||
# end_preparation 정보 추출 (분류 결과에서)
|
||||
end_prep = ""
|
||||
if isinstance(end_prep_info, dict):
|
||||
end_prep = end_prep_info.get("type", "")
|
||||
else:
|
||||
end_prep = str(end_prep_info) if end_prep_info else ""
|
||||
|
||||
# 재질 정보 - 이미 정제된 material_grade 사용
|
||||
material_spec = material_data.get("material_grade", "")
|
||||
|
||||
# 제조방법 추출
|
||||
manufacturing_method = ""
|
||||
if isinstance(manufacturing_info, dict):
|
||||
manufacturing_method = manufacturing_info.get("method", "UNKNOWN")
|
||||
else:
|
||||
manufacturing_method = str(manufacturing_info) if manufacturing_info else "UNKNOWN"
|
||||
|
||||
# 스케줄 정보 추출
|
||||
schedule = ""
|
||||
if isinstance(schedule_info, dict):
|
||||
schedule = schedule_info.get("schedule", "UNKNOWN")
|
||||
else:
|
||||
schedule = str(schedule_info) if schedule_info else "UNKNOWN"
|
||||
|
||||
db.execute(pipe_detail_insert_query, {
|
||||
"material_id": material_id,
|
||||
"file_id": file_id,
|
||||
"material_standard": material_info.get("standard"),
|
||||
"material_grade": material_info.get("grade"),
|
||||
"material_type": material_info.get("material_type"),
|
||||
"manufacturing_method": manufacturing_info.get("method"),
|
||||
"end_preparation": end_prep_info.get("type"),
|
||||
"schedule": schedule_info.get("schedule"),
|
||||
"wall_thickness": schedule_info.get("wall_thickness"),
|
||||
"nominal_size": material_data.get("size_spec", ""), # material_data에서 직접 가져옴
|
||||
"length_mm": length_mm,
|
||||
"material_confidence": material_info.get("confidence", 0.0),
|
||||
"manufacturing_confidence": manufacturing_info.get("confidence", 0.0),
|
||||
"end_prep_confidence": end_prep_info.get("confidence", 0.0),
|
||||
"schedule_confidence": schedule_info.get("confidence", 0.0)
|
||||
"outer_diameter": outer_diameter,
|
||||
"schedule": schedule,
|
||||
"material_spec": material_spec,
|
||||
"manufacturing_method": manufacturing_method,
|
||||
"end_preparation": end_prep,
|
||||
"length_mm": material_data.get("length", 0.0) if material_data.get("length") else 0.0
|
||||
})
|
||||
|
||||
print("PIPE 상세 정보 저장 완료")
|
||||
|
||||
# FITTING 분류 결과인 경우 상세 정보 저장
|
||||
elif classification_result.get("category") == "FITTING":
|
||||
print("FITTING 상세 정보 저장 시작")
|
||||
|
||||
# 피팅 정보 추출
|
||||
fitting_type_info = classification_result.get("fitting_type", {})
|
||||
connection_info = classification_result.get("connection_method", {})
|
||||
pressure_info = classification_result.get("pressure_rating", {})
|
||||
material_info = classification_result.get("material", {})
|
||||
|
||||
# 피팅 타입 및 서브타입
|
||||
fitting_type = fitting_type_info.get("type", "UNKNOWN")
|
||||
fitting_subtype = fitting_type_info.get("subtype", "UNKNOWN")
|
||||
|
||||
# 연결 방식
|
||||
connection_method = connection_info.get("method", "UNKNOWN")
|
||||
|
||||
# 압력 등급
|
||||
pressure_rating = pressure_info.get("rating", "UNKNOWN")
|
||||
|
||||
# 재질 정보
|
||||
material_standard = material_info.get("standard", "")
|
||||
material_grade = material_info.get("grade", "")
|
||||
|
||||
# main_size와 reduced_size
|
||||
main_size = material_data.get("main_nom") or material_data.get("size_spec", "")
|
||||
reduced_size = material_data.get("red_nom", "")
|
||||
|
||||
db.execute(text("""
|
||||
INSERT INTO fitting_details (
|
||||
material_id, file_id, fitting_type, fitting_subtype,
|
||||
connection_method, pressure_rating, material_standard,
|
||||
material_grade, main_size, reduced_size
|
||||
) VALUES (
|
||||
:material_id, :file_id, :fitting_type, :fitting_subtype,
|
||||
:connection_method, :pressure_rating, :material_standard,
|
||||
:material_grade, :main_size, :reduced_size
|
||||
)
|
||||
"""), {
|
||||
"material_id": material_id,
|
||||
"file_id": file_id,
|
||||
"fitting_type": fitting_type,
|
||||
"fitting_subtype": fitting_subtype,
|
||||
"connection_method": connection_method,
|
||||
"pressure_rating": pressure_rating,
|
||||
"material_standard": material_standard,
|
||||
"material_grade": material_grade,
|
||||
"main_size": main_size,
|
||||
"reduced_size": reduced_size
|
||||
})
|
||||
|
||||
print(f"FITTING 상세 정보 저장 완료: {fitting_type} - {fitting_subtype}")
|
||||
|
||||
db.commit()
|
||||
print(f"자재 저장 완료: {materials_inserted}개")
|
||||
@@ -457,14 +563,17 @@ async def get_materials(
|
||||
try:
|
||||
query = """
|
||||
SELECT m.id, m.file_id, m.original_description, m.quantity, m.unit,
|
||||
m.size_spec, m.material_grade, m.line_number, m.row_number,
|
||||
m.size_spec, m.main_nom, m.red_nom, m.material_grade, m.line_number, m.row_number,
|
||||
m.created_at, m.classified_category, m.classification_confidence,
|
||||
m.classification_details,
|
||||
f.original_filename, f.project_id, f.job_no, f.revision,
|
||||
p.official_project_code, p.project_name
|
||||
p.official_project_code, p.project_name,
|
||||
pd.outer_diameter, pd.schedule, pd.material_spec, pd.manufacturing_method,
|
||||
pd.end_preparation, pd.length_mm
|
||||
FROM materials m
|
||||
LEFT JOIN files f ON m.file_id = f.id
|
||||
LEFT JOIN projects p ON f.project_id = p.id
|
||||
LEFT JOIN pipe_details pd ON m.id = pd.material_id
|
||||
WHERE 1=1
|
||||
"""
|
||||
params = {}
|
||||
@@ -579,6 +688,8 @@ async def get_materials(
|
||||
"quantity": float(m.quantity) if m.quantity else 0,
|
||||
"unit": m.unit,
|
||||
"size_spec": m.size_spec,
|
||||
"main_nom": m.main_nom, # 추가
|
||||
"red_nom": m.red_nom, # 추가
|
||||
"material_grade": m.material_grade,
|
||||
"line_number": m.line_number,
|
||||
"row_number": m.row_number,
|
||||
@@ -588,22 +699,17 @@ async def get_materials(
|
||||
"created_at": m.created_at
|
||||
}
|
||||
|
||||
# 카테고리별 상세 정보 추가
|
||||
# 카테고리별 상세 정보 추가 (JOIN 결과 사용)
|
||||
if m.classified_category == 'PIPE':
|
||||
pipe_query = text("SELECT * FROM pipe_details WHERE material_id = :material_id")
|
||||
pipe_result = db.execute(pipe_query, {"material_id": m.id})
|
||||
pipe_detail = pipe_result.fetchone()
|
||||
if pipe_detail:
|
||||
# JOIN된 결과에서 pipe_details 정보 가져오기
|
||||
if hasattr(m, 'outer_diameter') and m.outer_diameter is not None:
|
||||
material_dict['pipe_details'] = {
|
||||
"nominal_size": pipe_detail.nominal_size,
|
||||
"schedule": pipe_detail.schedule,
|
||||
"material_standard": pipe_detail.material_standard,
|
||||
"material_grade": pipe_detail.material_grade,
|
||||
"material_type": pipe_detail.material_type,
|
||||
"manufacturing_method": pipe_detail.manufacturing_method,
|
||||
"end_preparation": pipe_detail.end_preparation,
|
||||
"wall_thickness": pipe_detail.wall_thickness,
|
||||
"length_mm": float(pipe_detail.length_mm) if pipe_detail.length_mm else None
|
||||
"outer_diameter": m.outer_diameter,
|
||||
"schedule": m.schedule,
|
||||
"material_spec": m.material_spec,
|
||||
"manufacturing_method": m.manufacturing_method,
|
||||
"end_preparation": m.end_preparation,
|
||||
"length_mm": float(m.length_mm) if m.length_mm else None
|
||||
}
|
||||
elif m.classified_category == 'FITTING':
|
||||
fitting_query = text("SELECT * FROM fitting_details WHERE material_id = :material_id")
|
||||
@@ -906,19 +1012,12 @@ async def get_pipe_details(
|
||||
"original_description": pd.original_description,
|
||||
"quantity": pd.quantity,
|
||||
"unit": pd.unit,
|
||||
"material_standard": pd.material_standard,
|
||||
"material_grade": pd.material_grade,
|
||||
"material_type": pd.material_type,
|
||||
"material_spec": pd.material_spec,
|
||||
"manufacturing_method": pd.manufacturing_method,
|
||||
"end_preparation": pd.end_preparation,
|
||||
"schedule": pd.schedule,
|
||||
"wall_thickness": pd.wall_thickness,
|
||||
"nominal_size": pd.nominal_size,
|
||||
"outer_diameter": pd.outer_diameter,
|
||||
"length_mm": pd.length_mm,
|
||||
"material_confidence": pd.material_confidence,
|
||||
"manufacturing_confidence": pd.manufacturing_confidence,
|
||||
"end_prep_confidence": pd.end_prep_confidence,
|
||||
"schedule_confidence": pd.schedule_confidence,
|
||||
"created_at": pd.created_at,
|
||||
"updated_at": pd.updated_at
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user