62 lines
9.6 KiB
XML
62 lines
9.6 KiB
XML
|
<svg xmlns="http://www.w3.org/2000/svg" width="2400" height="3600" viewBox="0 0 2400 3600">
|
||
|
|
<defs>
|
||
|
|
<marker id="arrow" markerWidth="10" markerHeight="10" refX="9" refY="3" orient="auto" markerUnits="strokeWidth">
|
||
|
|
<path d="M0,0 L0,6 L9,3 z" fill="#1E40AF"/>
|
||
|
|
</marker>
|
||
|
|
</defs>
|
||
|
|
<rect width="100%" height="100%" fill="#F7F9FC"/>
|
||
|
|
<text x="110" y="120" font-family="Malgun Gothic, Arial" font-size="58" font-weight="700" fill="#172033">크롤 진행 단계와 LLM 사용 판단 구조</text>
|
||
|
|
<text x="112" y="176" font-family="Malgun Gothic, Arial" font-size="28" fill="#5C6B82">API 요청부터 Rule/LLM 추출, 검증, DB 저장까지의 흐름</text>
|
||
|
|
<path d="M 1200 402 L 1200 470" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 602 L 1200 670" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 802 L 1200 870" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 1002 L 1200 1070" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 1222 L 1200 1300" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 1432 L 1200 1525" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 1897 L 1200 1965" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 2097 L 1200 2190" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 2562 L 1200 2635" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 3012 L 1200 3080" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 3240 L 1200 3330" fill="none" stroke="#1E40AF" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 820 1610 L 818 1610 L 818 1809 L 800 1809" fill="none" stroke="#5C6B82" stroke-width="4" marker-end="url(#arrow)"/>\n<rect x="756" y="1779" width="124" height="48" rx="12" fill="#F7F9FC" stroke="#CBD5E1" stroke-width="2"/>\n<text x="818" y="1817" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="25" font-weight="700" fill="#5C6B82">아니오</text>
|
||
|
|
<path d="M 1200 1695 L 1200 1765" fill="none" stroke="#16A34A" stroke-width="4" marker-end="url(#arrow)"/>\n<rect x="1138" y="1735" width="124" height="48" rx="12" fill="#F7F9FC" stroke="#CBD5E1" stroke-width="2"/>\n<text x="1200" y="1773" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="25" font-weight="700" fill="#16A34A">예</text>
|
||
|
|
<path d="M 820 2275 L 758 2275 L 758 2496 L 740 2496" fill="none" stroke="#5C6B82" stroke-width="4" marker-end="url(#arrow)"/>\n<rect x="696" y="2466" width="124" height="48" rx="12" fill="#F7F9FC" stroke="#CBD5E1" stroke-width="2"/>\n<text x="758" y="2504" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="25" font-weight="700" fill="#5C6B82">skip</text>
|
||
|
|
<path d="M 1200 2360 L 1200 2430" fill="none" stroke="#16A34A" stroke-width="4" marker-end="url(#arrow)"/>\n<rect x="1138" y="2400" width="124" height="48" rx="12" fill="#F7F9FC" stroke="#CBD5E1" stroke-width="2"/>\n<text x="1200" y="2438" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="25" font-weight="700" fill="#16A34A">use</text>
|
||
|
|
<path d="M 430 2562 L 430 2946 L 820 2946" fill="none" stroke="#5C6B82" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<path d="M 1200 2785 L 1200 2880" fill="none" stroke="#16A34A" stroke-width="4" marker-end="url(#arrow)"/>
|
||
|
|
<rect x="820" y="270" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="316" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">API 요청</text>\n<text x="1200" y="356" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">/crawl-site</text>
|
||
|
|
<rect x="820" y="470" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="516" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Job 생성 및</text>\n<text x="1200" y="556" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">queue 초기화</text>
|
||
|
|
<rect x="820" y="670" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="716" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">URL depth / domain /</text>\n<text x="1200" y="756" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">robots 검사</text>
|
||
|
|
<rect x="820" y="870" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="916" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Fetch:</text>\n<text x="1200" y="956" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">HTML 수집</text>
|
||
|
|
<rect x="820" y="1070" width="760" height="152" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="1106" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Parse:</text>\n<text x="1200" y="1146" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">본문 / zone /</text>\n<text x="1200" y="1186" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">clean_text 추출</text>
|
||
|
|
<rect x="820" y="1300" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="1346" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Page Classifier:</text>\n<text x="1200" y="1386" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">ProductPage 등 분류</text>
|
||
|
|
<polygon points="1200,1525 1580,1610 1200,1695 820,1610" fill="#FFF7ED" stroke="#D97706" stroke-width="4"/>
|
||
|
|
<text x="1200" y="1590" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">분석 대상</text>\n<text x="1200" y="1630" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">page_type인가?</text>
|
||
|
|
<rect x="120" y="1740" width="680" height="138" rx="24" fill="#F8FAFC" stroke="#5C6B82" stroke-width="4"/>
|
||
|
|
<text x="460" y="1769" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">discovered/skipped</text>\n<text x="460" y="1809" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">저장,</text>\n<text x="460" y="1849" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">링크만 확장</text>
|
||
|
|
<rect x="820" y="1765" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="1831" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Extractor 선택</text>
|
||
|
|
<rect x="820" y="1965" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="2031" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">Rule extraction</text>
|
||
|
|
<polygon points="1200,2190 1580,2275 1200,2360 820,2275" fill="#FFF7ED" stroke="#D97706" stroke-width="4"/>
|
||
|
|
<text x="1200" y="2275" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">LLM 사용 판단</text>
|
||
|
|
<rect x="120" y="2430" width="620" height="132" rx="24" fill="#F8FAFC" stroke="#5C6B82" stroke-width="4"/>
|
||
|
|
<text x="430" y="2496" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">rule_only routed</text>
|
||
|
|
<rect x="820" y="2430" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="2496" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">LLM extraction</text>
|
||
|
|
<rect x="820" y="2635" width="760" height="150" rx="24" fill="#ECFDF5" stroke="#16A34A" stroke-width="4"/>
|
||
|
|
<text x="1200" y="2690" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">rule/LLM merge,</text>\n<text x="1200" y="2730" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">agreement/conflict 계산</text>
|
||
|
|
<rect x="820" y="2880" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="2946" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">validation</text>
|
||
|
|
<rect x="820" y="3080" width="760" height="160" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="3120" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">DB 저장:</text>\n<text x="1200" y="3160" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">entities / claims /</text>\n<text x="1200" y="3200" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">evidence / log</text>
|
||
|
|
<rect x="820" y="3330" width="760" height="132" rx="24" fill="#FFFFFF" stroke="#2563EB" stroke-width="4"/>
|
||
|
|
<text x="1200" y="3396" text-anchor="middle" font-family="Malgun Gothic, Arial" font-size="30" font-weight="700" fill="#172033">progress 업데이트</text>
|
||
|
|
</svg>
|