Initial CivilPlan MCP implementation

2026-04-03 09:08:08 +09:00
commit 544e4e0720
70 changed files with 3364 additions and 0 deletions
@@ -0,0 +1,3 @@
+from civilplan_mcp.updater.scheduler import build_scheduler
+
+__all__ = ["build_scheduler"]
@@ -0,0 +1,93 @@
+from __future__ import annotations
+
+from datetime import datetime
+import json
+from pathlib import Path
+import re
+from typing import Any
+
+import httpx
+
+from civilplan_mcp.config import get_settings
+
+
+LOG_FILE_NAME = "update_log.json"
+
+
+def flag_manual_update_required(update_type: str, message: str, data_dir: Path | None = None) -> Path:
+    target_dir = data_dir or get_settings().data_dir
+    target_dir.mkdir(parents=True, exist_ok=True)
+    flag_path = target_dir / f".update_required_{update_type}"
+    flag_path.write_text(f"{datetime.now():%Y-%m-%d} {message}", encoding="utf-8")
+    return flag_path
+
+
+def clear_update_flag(update_type: str, data_dir: Path | None = None) -> None:
+    target_dir = data_dir or get_settings().data_dir
+    flag_path = target_dir / f".update_required_{update_type}"
+    if flag_path.exists():
+        flag_path.unlink()
+
+
+def check_update_flags(data_dir: Path | None = None) -> list[str]:
+    target_dir = data_dir or get_settings().data_dir
+    warnings: list[str] = []
+    for flag in target_dir.glob(".update_required_*"):
+        warnings.append(flag.read_text(encoding="utf-8"))
+    return warnings
+
+
+def read_update_log(data_dir: Path | None = None) -> list[dict[str, Any]]:
+    target_dir = data_dir or get_settings().data_dir
+    log_path = target_dir / LOG_FILE_NAME
+    if not log_path.exists():
+        return []
+    return json.loads(log_path.read_text(encoding="utf-8"))
+
+
+def write_update_log(entry: dict[str, Any], data_dir: Path | None = None) -> Path:
+    target_dir = data_dir or get_settings().data_dir
+    target_dir.mkdir(parents=True, exist_ok=True)
+    log_path = target_dir / LOG_FILE_NAME
+    log_data = read_update_log(target_dir)
+    log_data.append(entry)
+    log_path.write_text(json.dumps(log_data, ensure_ascii=False, indent=2), encoding="utf-8")
+    return log_path
+
+
+def fetch_source_text(url: str) -> str:
+    response = httpx.get(
+        url,
+        timeout=30,
+        headers={"User-Agent": "CivilPlan-MCP/1.0 (+https://localhost:8765/mcp)"},
+    )
+    response.raise_for_status()
+    return response.text
+
+
+def extract_marker(text: str, patterns: list[str]) -> str | None:
+    for pattern in patterns:
+        match = re.search(pattern, text, flags=re.IGNORECASE)
+        if match:
+            return match.group(0)
+    return None
+
+
+def build_log_entry(
+    *,
+    update_type: str,
+    period: str | None,
+    status: str,
+    source_url: str,
+    detail: str,
+    marker: str | None = None,
+) -> dict[str, Any]:
+    return {
+        "timestamp": datetime.now().isoformat(timespec="seconds"),
+        "update_type": update_type,
+        "period": period,
+        "status": status,
+        "source_url": source_url,
+        "marker": marker,
+        "detail": detail,
+    }
@@ -0,0 +1,20 @@
+from __future__ import annotations
+
+from apscheduler.schedulers.background import BackgroundScheduler
+from apscheduler.triggers.cron import CronTrigger
+
+from civilplan_mcp.updater.standard_updater import update_standard_prices
+from civilplan_mcp.updater.wage_updater import update_wage_rates
+from civilplan_mcp.updater.waste_updater import update_waste_prices
+
+
+def build_scheduler(*, start: bool = False) -> BackgroundScheduler:
+    scheduler = BackgroundScheduler(timezone="Asia/Seoul")
+    scheduler.add_job(update_wage_rates, CronTrigger(month="1", day="2", hour="9"), id="wage_h1", kwargs={"period": "상반기"})
+    scheduler.add_job(update_waste_prices, CronTrigger(month="1", day="2", hour="9"), id="waste_annual")
+    scheduler.add_job(update_standard_prices, CronTrigger(month="1", day="2", hour="9"), id="standard_h1", kwargs={"period": "상반기"})
+    scheduler.add_job(update_standard_prices, CronTrigger(month="7", day="10", hour="9"), id="standard_h2", kwargs={"period": "하반기"})
+    scheduler.add_job(update_wage_rates, CronTrigger(month="9", day="2", hour="9"), id="wage_h2", kwargs={"period": "하반기"})
+    if start:
+        scheduler.start()
+    return scheduler
@@ -0,0 +1,80 @@
+from __future__ import annotations
+
+from typing import Any
+
+from civilplan_mcp.config import get_settings
+from civilplan_mcp.updater.common import (
+    build_log_entry,
+    clear_update_flag,
+    extract_marker,
+    fetch_source_text,
+    flag_manual_update_required,
+    write_update_log,
+)
+
+
+STANDARD_SOURCE_URL = "https://www.molit.go.kr/portal.do"
+STANDARD_MARKER_PATTERNS = [
+    r"표준시장단가",
+    r"제비율",
+    r"20\d{2}년",
+]
+
+
+def update_standard_prices(period: str = "상반기") -> dict[str, Any]:
+    data_dir = get_settings().data_dir
+    try:
+        text = fetch_source_text(STANDARD_SOURCE_URL)
+        marker = extract_marker(text, STANDARD_MARKER_PATTERNS)
+        if not marker:
+            detail = f"Fetched standard-cost source, but no recognizable bulletin marker was found for {period}."
+            flag_manual_update_required("standard", detail, data_dir=data_dir)
+            write_update_log(
+                build_log_entry(
+                    update_type="standard",
+                    period=period,
+                    status="pending_manual_review",
+                    source_url=STANDARD_SOURCE_URL,
+                    detail=detail,
+                ),
+                data_dir=data_dir,
+            )
+            return {"status": "pending_manual_review", "period": period, "source_url": STANDARD_SOURCE_URL}
+
+        clear_update_flag("standard", data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="standard",
+                period=period,
+                status="fetched",
+                source_url=STANDARD_SOURCE_URL,
+                detail="Recognized standard-cost bulletin marker from source page.",
+                marker=marker,
+            ),
+            data_dir=data_dir,
+        )
+        return {
+            "status": "fetched",
+            "period": period,
+            "marker": marker,
+            "source_url": STANDARD_SOURCE_URL,
+        }
+    except Exception as exc:
+        detail = f"Standard update fetch failed for {period}: {exc}"
+        flag_manual_update_required("standard", detail, data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="standard",
+                period=period,
+                status="pending_manual_review",
+                source_url=STANDARD_SOURCE_URL,
+                detail=detail,
+            ),
+            data_dir=data_dir,
+        )
+        return {
+            "status": "pending_manual_review",
+            "period": period,
+            "message": detail,
+            "source_url": STANDARD_SOURCE_URL,
+        }
@@ -0,0 +1 @@
+[]
@@ -0,0 +1,84 @@
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from civilplan_mcp.config import get_settings
+from civilplan_mcp.updater.common import (
+    build_log_entry,
+    check_update_flags,
+    clear_update_flag,
+    extract_marker,
+    fetch_source_text,
+    flag_manual_update_required,
+    write_update_log,
+)
+
+
+WAGE_SOURCE_URL = "https://gwangju.cak.or.kr/lay1/bbs/S1T41C42/A/14/list.do"
+WAGE_MARKER_PATTERNS = [
+    r"20\d{2}년\s*(상반기|하반기)\s*적용\s*건설업\s*임금실태조사",
+    r"20\d{2}년\s*건설업\s*임금실태조사",
+]
+
+
+def update_wage_rates(period: str = "상반기") -> dict[str, Any]:
+    data_dir = get_settings().data_dir
+    try:
+        text = fetch_source_text(WAGE_SOURCE_URL)
+        marker = extract_marker(text, WAGE_MARKER_PATTERNS)
+        if not marker:
+            detail = f"Fetched wage source, but no recognizable wage bulletin marker was found for {period}."
+            flag_manual_update_required("wage", detail, data_dir=data_dir)
+            write_update_log(
+                build_log_entry(
+                    update_type="wage",
+                    period=period,
+                    status="pending_manual_review",
+                    source_url=WAGE_SOURCE_URL,
+                    detail=detail,
+                ),
+                data_dir=data_dir,
+            )
+            return {"status": "pending_manual_review", "period": period, "source_url": WAGE_SOURCE_URL}
+
+        clear_update_flag("wage", data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="wage",
+                period=period,
+                status="fetched",
+                source_url=WAGE_SOURCE_URL,
+                detail="Recognized wage bulletin marker from source page.",
+                marker=marker,
+            ),
+            data_dir=data_dir,
+        )
+        return {
+            "status": "fetched",
+            "period": period,
+            "marker": marker,
+            "source_url": WAGE_SOURCE_URL,
+        }
+    except Exception as exc:
+        detail = f"Wage update fetch failed for {period}: {exc}"
+        flag_manual_update_required("wage", detail, data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="wage",
+                period=period,
+                status="pending_manual_review",
+                source_url=WAGE_SOURCE_URL,
+                detail=detail,
+            ),
+            data_dir=data_dir,
+        )
+        return {
+            "status": "pending_manual_review",
+            "period": period,
+            "message": detail,
+            "source_url": WAGE_SOURCE_URL,
+        }
+
+
+__all__ = ["check_update_flags", "flag_manual_update_required", "update_wage_rates"]
@@ -0,0 +1,73 @@
+from __future__ import annotations
+
+from typing import Any
+
+from civilplan_mcp.config import get_settings
+from civilplan_mcp.updater.common import (
+    build_log_entry,
+    clear_update_flag,
+    extract_marker,
+    fetch_source_text,
+    flag_manual_update_required,
+    write_update_log,
+)
+
+
+WASTE_SOURCE_URL = "https://www.data.go.kr/data/15052266/fileData.do"
+WASTE_MARKER_PATTERNS = [
+    r"20\d{2}[-./]\d{2}[-./]\d{2}",
+    r"20\d{2}년",
+]
+
+
+def update_waste_prices() -> dict[str, Any]:
+    data_dir = get_settings().data_dir
+    try:
+        text = fetch_source_text(WASTE_SOURCE_URL)
+        marker = extract_marker(text, WASTE_MARKER_PATTERNS)
+        if not marker:
+            detail = "Fetched waste source, but no recognizable release marker was found."
+            flag_manual_update_required("waste", detail, data_dir=data_dir)
+            write_update_log(
+                build_log_entry(
+                    update_type="waste",
+                    period=None,
+                    status="pending_manual_review",
+                    source_url=WASTE_SOURCE_URL,
+                    detail=detail,
+                ),
+                data_dir=data_dir,
+            )
+            return {"status": "pending_manual_review", "source_url": WASTE_SOURCE_URL}
+
+        clear_update_flag("waste", data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="waste",
+                period=None,
+                status="fetched",
+                source_url=WASTE_SOURCE_URL,
+                detail="Recognized release marker from waste source page.",
+                marker=marker,
+            ),
+            data_dir=data_dir,
+        )
+        return {"status": "fetched", "marker": marker, "source_url": WASTE_SOURCE_URL}
+    except Exception as exc:
+        detail = f"Waste update fetch failed: {exc}"
+        flag_manual_update_required("waste", detail, data_dir=data_dir)
+        write_update_log(
+            build_log_entry(
+                update_type="waste",
+                period=None,
+                status="pending_manual_review",
+                source_url=WASTE_SOURCE_URL,
+                detail=detail,
+            ),
+            data_dir=data_dir,
+        )
+        return {
+            "status": "pending_manual_review",
+            "message": detail,
+            "source_url": WASTE_SOURCE_URL,
+        }