Add functions for handling UTF-8 encoding and attachment headers in report_service; enhance chart time parsing in docx_sections; improve report generation logic in docx.py to conditionally include charts for Cloud-to-Ground and Intercloud lightnings based on data availability.

report_service/main.py

@@ -87,6 +87,31 @@ def _build_filename(payload: dict[str, Any]) -> str:
     return "_".join(parts)
 
 
+def _pct_encode_utf8(text: str) -> str:
+    return "".join(f"%{b:02X}" for b in text.encode("utf-8"))
+
+
+def _latin1_header_values(headers: dict[str, str]) -> dict[str, str]:
+    return {
+        k: v.encode("latin-1", errors="replace").decode("latin-1") for k, v in headers.items()
+    }
+
+
+def _ascii_attachment_filename_from_stem(stem: str) -> str:
+    slug = slugify_ascii_underscore(stem)
+    safe = slug.encode("ascii", "ignore").decode("ascii").strip("._-") or "report"
+    return f"{safe}.docx"
+
+
+def _content_disposition_attachment(filename: str) -> str:
+    lower = filename.lower()
+    stem = filename[: -len(".docx")] if lower.endswith(".docx") else filename
+    ascii_fn = _ascii_attachment_filename_from_stem(stem)
+    ascii_fn = "".join(ch for ch in ascii_fn if ord(ch) < 128) or "report.docx"
+    encoded = _pct_encode_utf8(filename)
+    return f'attachment; filename="{ascii_fn}"; filename*=UTF-8\'\'{encoded}'
+
+
 @app.get("/health")
 def health() -> JSONResponse:
     return JSONResponse({"ok": True, "service": "lightning-report", "version": app.version})
@@ -147,13 +172,20 @@ async def generate(request: Request) -> Response:
             pass
 
     logger.info("Generated %s (%d bytes) for %s", filename, len(data), customer_name)
+    stem = filename[: -len(".docx")] if filename.lower().endswith(".docx") else filename
+    ascii_filename = _ascii_attachment_filename_from_stem(stem)
+    ascii_filename = "".join(ch for ch in ascii_filename if ord(ch) < 128) or "report.docx"
+    customer_hdr = _pct_encode_utf8(str(customer_name))
+    hdrs = _latin1_header_values(
+        {
+            "Content-Disposition": _content_disposition_attachment(filename),
+            "X-Report-Filename": ascii_filename,
+            "X-Report-Customer": customer_hdr,
+            "X-Report-Strikes": str(n_strikes),
+        }
+    )
     return Response(
         content=data,
         media_type=DOCX_MIME,
-        headers={
+        headers=hdrs,
-            "Content-Disposition": f'attachment; filename="{filename}"',
-            "X-Report-Filename": filename,
-            "X-Report-Customer": str(customer_name),
-            "X-Report-Strikes": str(n_strikes),
-        },
     )

src/reporting/docx.py

@@ -682,52 +682,58 @@ def create_docx_report(
 		"name": config.wind_farm_name or "Wind Farm",
 	})
 	pre = precompute_distances_and_rings(centroid_lat, centroid_lng, lightning_df, config.distance_rings)
+	type_values = lightning_df.get("p_type", pd.Series(dtype=str)).astype(str)
+	mask_within = pre["mask_within"]
+	has_cg = bool(((type_values == "0") & mask_within).any())
+	has_ic = bool(((type_values != "0") & mask_within).any())
 
-	_add_title(doc, "Cloud-to-Ground Lightnings", size_pt=14)
+	if has_cg:
-	if start_date and end_date:
+		_add_title(doc, "Cloud-to-Ground Lightnings", size_pt=14)
-		_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
+		if start_date and end_date:
-	cg_fig = plot_cloud_to_ground_coordinate_plane(centroid_row, lightning_df, turbine_df, storm_data, precomputed=pre)
+			_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
-	_add_image_from_bytes(doc, cg_fig.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+		cg_fig = plot_cloud_to_ground_coordinate_plane(centroid_row, lightning_df, turbine_df, storm_data, precomputed=pre)
-	doc.add_page_break()
+		_add_image_from_bytes(doc, cg_fig.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+		doc.add_page_break()
 
-	_add_title(doc, "Cloud-to-Ground — Current vs Distance", size_pt=14)
+		_add_title(doc, "Cloud-to-Ground — Current vs Distance", size_pt=14)
-	if start_date and end_date:
+		if start_date and end_date:
-		_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
+			_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
-	cg_chart = _build_current_vs_distance_chart(
+		cg_chart = _build_current_vs_distance_chart(
-		lightning_df,
+			lightning_df,
-		pre["dists_km"],
+			pre["dists_km"],
-		pre["mask_within"],
+			pre["mask_within"],
-		"cg",
+			"cg",
-		"Cloud-to-Ground Lightning — Current vs Distance from Centroid",
+			"Cloud-to-Ground Lightning — Current vs Distance from Centroid",
-		900,
+			900,
-		650,
+			650,
-	)
+		)
-	if cg_chart is not None:
+		if cg_chart is not None:
-		_add_image_from_bytes(doc, cg_chart.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+			_add_image_from_bytes(doc, cg_chart.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
-	doc.add_page_break()
+			doc.add_page_break()
 
-	_add_title(doc, "Intercloud Lightnings", size_pt=14)
+	if has_ic:
-	if start_date and end_date:
+		_add_title(doc, "Intercloud Lightnings", size_pt=14)
-		_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
+		if start_date and end_date:
-	ic_fig = plot_intercloud_coordinate_plane(centroid_row, lightning_df, turbine_df, storm_data, precomputed=pre)
+			_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
-	_add_image_from_bytes(doc, ic_fig.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+		ic_fig = plot_intercloud_coordinate_plane(centroid_row, lightning_df, turbine_df, storm_data, precomputed=pre)
-	doc.add_page_break()
+		_add_image_from_bytes(doc, ic_fig.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+		doc.add_page_break()
 
-	_add_title(doc, "Intercloud — Current vs Distance", size_pt=14)
+		_add_title(doc, "Intercloud — Current vs Distance", size_pt=14)
-	if start_date and end_date:
+		if start_date and end_date:
-		_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
+			_add_paragraph(doc, f"{start_date} - {end_date}", size_pt=10)
-	ic_chart = _build_current_vs_distance_chart(
+		ic_chart = _build_current_vs_distance_chart(
-		lightning_df,
+			lightning_df,
-		pre["dists_km"],
+			pre["dists_km"],
-		pre["mask_within"],
+			pre["mask_within"],
-		"ic",
+			"ic",
-		"Intercloud Lightning — Current vs Distance from Centroid",
+			"Intercloud Lightning — Current vs Distance from Centroid",
-		900,
+			900,
-		650,
+			650,
-	)
+		)
-	if ic_chart is not None:
+		if ic_chart is not None:
-		_add_image_from_bytes(doc, ic_chart.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
+			_add_image_from_bytes(doc, ic_chart.to_image(format="png", width=1200, height=900, scale=2, engine="kaleido"), content_width)
-	doc.add_page_break()
+			doc.add_page_break()
 
 	risk_table_data, risk_row_colors = build_risk_table_data(turbine_df)
 	if risk_table_data and len(risk_table_data) > 1:

src/reporting/docx_sections.py

@@ -8,6 +8,38 @@ from src.config import config
 from src.reporting.precompute import precompute_distances_and_rings
 
 
+def _parse_chart_times(values: pd.Series) -> pd.Series:
+	def _to_local_naive(ts: pd.Series) -> pd.Series:
+		if config.timezone:
+			try:
+				ts = ts.dt.tz_convert(config.timezone)
+			except Exception:
+				pass
+		try:
+			return ts.dt.tz_localize(None)
+		except Exception:
+			return ts
+
+	numeric = pd.to_numeric(values, errors="coerce")
+	numeric_valid = numeric.dropna()
+
+	if len(numeric_valid) > 0 and len(numeric_valid) >= max(1, int(len(values) * 0.7)):
+		abs_max = float(numeric_valid.abs().max())
+		if abs_max >= 1e17:
+			unit = "ns"
+		elif abs_max >= 1e14:
+			unit = "us"
+		elif abs_max >= 1e11:
+			unit = "ms"
+		else:
+			unit = "s"
+		parsed = pd.to_datetime(numeric, errors="coerce", unit=unit, utc=True)
+		return _to_local_naive(parsed)
+
+	parsed = pd.to_datetime(values, errors="coerce", utc=True)
+	return _to_local_naive(parsed)
+
+
 def _build_current_vs_distance_chart(
 	lightning_df: pd.DataFrame,
 	dists_km: np.ndarray,
@@ -30,8 +62,19 @@ def _build_current_vs_distance_chart(
 	distances = dists_km[combined_mask]
 	currents = subset["current"].values.astype(float)
 
-	time_series = pd.to_datetime(subset["local_time"])
+	time_series = _parse_chart_times(subset["local_time"])
-	time_dt = time_series.sort_values()
+	valid_mask = ~time_series.isna()
+	if valid_mask.sum() == 0:
+		return None
+
+	time_series = time_series.loc[valid_mask]
+	distances = distances[valid_mask.values]
+	currents = currents[valid_mask.values]
+
+	sort_idx = np.argsort(time_series.values.astype("datetime64[ns]"))
+	time_values = time_series.values[sort_idx]
+	distances = distances[sort_idx]
+	currents = currents[sort_idx]
 
 	rings_km = np.array(config.distance_rings, dtype=float) / 1000.0
 	ring_colors_cfg = getattr(config, "ring_colors", None) or []
@@ -45,21 +88,18 @@ def _build_current_vs_distance_chart(
 		else:
 			ring_names.append(f"{rings_km[i - 1]:.1f}-{rings_km[i]:.1f} km")
 
-	t_min = time_dt.min()
-	t_max = time_dt.max()
-	tick_vals = pd.date_range(t_min, t_max, periods=4)
-	tick_text = [t.strftime("%d-%m-%Y %H:%M") for t in tick_vals]
-
 	fig = go.Figure()
 	for i in range(len(rings_km)):
 		mask_ring = ring_indices == i
 		if mask_ring.sum() == 0:
 			continue
 		color = ring_colors_cfg[i] if i < len(ring_colors_cfg) else "gray"
-		r_times = time_series.values[mask_ring]
+		r_times = time_values[mask_ring]
 		r_currents = currents[mask_ring]
 		r_dists = distances[mask_ring]
+		tz_suffix = f" ({config.timezone})" if config.timezone else ""
 		r_time_labels = pd.to_datetime(r_times).strftime("%d-%m-%Y %H:%M").values
+		r_time_labels = np.array([f"{t}{tz_suffix}" for t in r_time_labels])
 		fig.add_trace(
 			go.Scatter(
 				x=r_times,
@@ -78,19 +118,22 @@ def _build_current_vs_distance_chart(
 			)
 		)
 
+	timezone_label = config.timezone or "UTC"
+
 	fig.update_layout(
 		font=dict(size=16),
 		title=dict(text=title, x=0.5, font=dict(size=22)),
-		xaxis_title="Time",
+		xaxis_title=f"Time ({timezone_label})",
 		yaxis_title="Current (A)",
 		plot_bgcolor="white",
 		paper_bgcolor="white",
 		xaxis=dict(
+			type="date",
 			showgrid=True,
 			gridcolor="lightgray",
 			zeroline=False,
-			tickvals=tick_vals,
+			tickformat="%d-%m-%Y %H:%M",
-			ticktext=tick_text,
+			nticks=6,
 			tickangle=-25,
 			tickfont=dict(size=22),
 			title_font=dict(size=28),