scraper2-hj3415 2.1.0__tar.gz → 2.1.2__tar.gz

{scraper2_hj3415-2.1.0 → scraper2_hj3415-2.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scraper2-hj3415
-Version: 2.1.0
+Version: 2.1.2
 Summary: Naver WiseReport scraper
 Keywords: example,demo
 Author-email: Hyungjin Kim <hj3415@gmail.com>
@@ -16,6 +16,7 @@ Requires-Dist: pandas-stubs>=2.3.3
 Requires-Dist: lxml>=6.0.2
 Requires-Dist: typer>=0.21.0
 Requires-Dist: db2-hj3415
+Requires-Dist: contracts-hj3415
 
 # scraper2
 

{scraper2_hj3415-2.1.0 → scraper2_hj3415-2.1.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "scraper2-hj3415"                          # PyPI 이름 (하이픈 허용)
-version = "2.1.0"
+version = "2.1.2"
 description = "Naver WiseReport scraper"
 readme = "README.md"
 requires-python = ">=3.11"
@@ -24,6 +24,7 @@ dependencies = [
   "lxml>=6.0.2",
   "typer>=0.21.0",
   "db2-hj3415",
+  "contracts-hj3415",
 ]
 
 [tool.flit.module]

{scraper2_hj3415-2.1.0 → scraper2_hj3415-2.1.2}/src/scraper2/main.py

@@ -92,6 +92,7 @@ async def _run_ingest_with_progress(
     sleep_sec: float,
     show: bool,
     show_n: int,
+    asof: datetime,
     chunk_size: int = 10,
     progress_every: int = 1,
 ) -> None:
@@ -101,7 +102,7 @@ async def _run_ingest_with_progress(
         return
 
     t0 = time.perf_counter()   # ✅ 시작 시각
-    run_asof = datetime.now(timezone.utc)
+    run_asof = asof
 
     def _chunks(xs: list[str], n: int):
         for i in range(0, len(xs), n):
@@ -159,6 +160,39 @@ def _format_elapsed(sec: float) -> str:
     h, m = divmod(m, 60)
     return f"{h}h {m}m {s}s"
 
+def _parse_asof(asof: str | None) -> datetime:
+    """
+    Parse ISO8601 string to timezone-aware UTC datetime.
+    Accepts:
+      - 2026-01-09T05:00:00+00:00
+      - 2026-01-09T05:00:00Z
+      - 2026-01-09T05:00:00  (treated as UTC)
+    """
+    if not asof:
+        return datetime.now(timezone.utc)
+
+    s = asof.strip()
+    if not s:
+        return datetime.now(timezone.utc)
+
+    # allow trailing 'Z'
+    if s.endswith("Z"):
+        s = s[:-1] + "+00:00"
+
+    try:
+        dt = datetime.fromisoformat(s)
+    except ValueError as e:
+        raise typer.BadParameter(
+            f"--asof must be ISO8601 datetime (e.g. 2026-01-09T05:00:00Z). got={asof!r}"
+        ) from e
+
+    # if naive, assume UTC
+    if dt.tzinfo is None:
+        dt = dt.replace(tzinfo=timezone.utc)
+
+    # normalize to UTC
+    return dt.astimezone(timezone.utc)
+
 # -------------------------
 # nfs subcommands: one / all
 # -------------------------
@@ -170,6 +204,7 @@ def nfs_one(
     sleep_sec: float = typer.Option(2.0, "--sleep"),
     sink: Sink = typer.Option("memory", "--sink"),
     show: bool = typer.Option(True, "--show/--no-show", help="결과 DTO 출력"),
+    asof: str | None = typer.Option(None, "--asof", help="배치 기준시각(ISO8601, UTC 권장). 예: 2026-01-09T05:00:00Z"),
 ):
     code = code.strip()
     if not code:
@@ -184,7 +219,7 @@ def nfs_one(
             await _mongo_bootstrap(ucs.db)
 
         try:
-            run_asof = datetime.now(timezone.utc)
+            run_asof = _parse_asof(asof)
             for ep in _endpoint_list(endpoint):
                 ingest_uc = cast(IngestPort, getattr(ucs.ingest, ep))
                 results = await ingest_uc.execute_many([code], sleep_sec=sleep_sec, asof=run_asof)
@@ -214,6 +249,7 @@ def nfs_all(
     chunk_size: int = typer.Option(5, "--chunk", help="진행률 표시용 배치 크기"),
     show: bool = typer.Option(False, "--show/--no-show", help="일부 코드만 출력"),
     show_n: int = typer.Option(3, "--show-n"),
+    asof: str | None = typer.Option(None, "--asof", help="배치 기준시각(ISO8601). 예: 2026-01-09T05:00:00Z"),
 ):
     async def _run():
         ucs = build_usecases(sink_kind=sink)
@@ -228,6 +264,8 @@ def nfs_all(
         if limit and limit > 0:
             codes = codes[:limit]
 
+        run_asof = _parse_asof(asof)
+
         typer.echo(f"\n=== NFS ALL === universe={universe}, endpoint={endpoint}, codes={len(codes)}, sink={sink}")
 
         try:
@@ -238,6 +276,7 @@ def nfs_all(
                 sleep_sec=sleep_sec,
                 show=show,
                 show_n=show_n,
+                asof=run_asof,
                 chunk_size=chunk_size,
                 progress_every=1,   # chunk마다
             )