scitex 2.14.0__py3-none-any.whl → 2.15.1__py3-none-any.whl

scitex/canvas/editor/_dearpygui/_rendering.py

@@ -0,0 +1,504 @@
+#!/usr/bin/env python3
+# Timestamp: "2026-01-24 (ywatanabe)"
+# File: /home/ywatanabe/proj/scitex-python/src/scitex/canvas/editor/_dearpygui/_rendering.py
+
+"""
+Figure rendering for DearPyGui editor.
+
+Handles matplotlib figure rendering, element highlights, and hover overlays.
+"""
+
+import io
+from typing import TYPE_CHECKING, List, Tuple
+
+from ._utils import MM_TO_PT, create_checkerboard
+
+if TYPE_CHECKING:
+    from ._state import EditorState
+
+
+def update_preview(state: "EditorState", dpg) -> None:
+    """Update the figure preview (full re-render).
+
+    Parameters
+    ----------
+    state : EditorState
+        Editor state
+    dpg : module
+        DearPyGui module
+    """
+    try:
+        # Mark cache dirty and do full render
+        state.cache_dirty = True
+        img_data, width, height = render_figure(state, dpg)
+
+        # Update texture
+        dpg.set_value("preview_texture", img_data)
+
+        # Update status
+        dpg.set_value("status_text", f"Preview updated ({width}x{height})")
+
+    except Exception as e:
+        dpg.set_value("status_text", f"Error: {str(e)}")
+
+
+def update_hover_overlay(state: "EditorState", dpg) -> None:
+    """Fast hover overlay update using cached base image (no matplotlib re-render).
+
+    Parameters
+    ----------
+    state : EditorState
+        Editor state
+    dpg : module
+        DearPyGui module
+    """
+    import numpy as np
+    from PIL import ImageDraw
+
+    # If no cached base, do full render
+    if state.cached_base_image is None:
+        update_preview(state, dpg)
+        return
+
+    try:
+        # Start with a copy of cached base
+        img = state.cached_base_image.copy()
+        draw = ImageDraw.Draw(img, "RGBA")
+
+        # Get hover element type
+        hovered_type = (
+            state.hovered_element.get("type") if state.hovered_element else None
+        )
+        selected_type = (
+            state.selected_element.get("type") if state.selected_element else None
+        )
+
+        # Draw hover highlight (outline only, no fill) for non-trace elements
+        if hovered_type and hovered_type != "trace" and hovered_type != selected_type:
+            bbox = state.element_bboxes.get(hovered_type)
+            if bbox:
+                x0, y0, x1, y1 = bbox
+                # Transparent outline only - no fill to avoid covering content
+                draw.rectangle(
+                    [x0 - 2, y0 - 2, x1 + 2, y1 + 2],
+                    fill=None,
+                    outline=(100, 180, 255, 100),
+                    width=1,
+                )
+
+        # Draw selection highlight (outline only, no fill) for non-trace elements
+        if selected_type and selected_type != "trace":
+            bbox = state.element_bboxes.get(selected_type)
+            if bbox:
+                x0, y0, x1, y1 = bbox
+                # Transparent outline only - no fill to avoid covering content
+                draw.rectangle(
+                    [x0 - 2, y0 - 2, x1 + 2, y1 + 2],
+                    fill=None,
+                    outline=(255, 200, 80, 150),
+                    width=2,
+                )
+
+        # Convert to DearPyGui texture format
+        img_array = np.array(img).astype(np.float32) / 255.0
+        img_data = img_array.flatten().tolist()
+
+        # Update texture
+        dpg.set_value("preview_texture", img_data)
+
+    except Exception:
+        # Fallback to full render on error
+        update_preview(state, dpg)
+
+
+def render_figure(state: "EditorState", dpg) -> Tuple[List[float], int, int]:
+    """Render figure and return as RGBA data for texture.
+
+    Parameters
+    ----------
+    state : EditorState
+        Editor state
+    dpg : module
+        DearPyGui module
+
+    Returns
+    -------
+    tuple
+        (img_data, width, height)
+    """
+    import matplotlib
+
+    matplotlib.use("Agg")
+    import matplotlib.pyplot as plt
+    import numpy as np
+    from matplotlib.ticker import MaxNLocator
+    from PIL import Image
+
+    from ._plotting import plot_from_csv
+
+    o = state.current_overrides
+
+    # Dimensions - use fixed size for preview
+    preview_dpi = 100
+    fig_size = o.get("fig_size", [3.15, 2.68])
+
+    # Create figure with white background for preview
+    fig, ax = plt.subplots(figsize=fig_size, dpi=preview_dpi)
+
+    # For preview, use white background (transparent doesn't show well in GUI)
+    fig.patch.set_facecolor("white")
+    ax.patch.set_facecolor("white")
+
+    # Plot from CSV data (only pass selection, hover is via PIL overlay for speed)
+    if state.csv_data is not None:
+        plot_from_csv(ax, o, state.csv_data, highlight_trace=state.selected_trace_index)
+    else:
+        ax.text(
+            0.5,
+            0.5,
+            "No plot data available\n(CSV not found)",
+            ha="center",
+            va="center",
+            transform=ax.transAxes,
+            fontsize=o.get("axis_fontsize", 7),
+        )
+
+    # Apply labels
+    if o.get("title"):
+        ax.set_title(o["title"], fontsize=o.get("title_fontsize", 8))
+    if o.get("xlabel"):
+        ax.set_xlabel(o["xlabel"], fontsize=o.get("axis_fontsize", 7))
+    if o.get("ylabel"):
+        ax.set_ylabel(o["ylabel"], fontsize=o.get("axis_fontsize", 7))
+
+    # Tick styling
+    ax.tick_params(
+        axis="both",
+        labelsize=o.get("tick_fontsize", 7),
+        length=o.get("tick_length", 0.8) * MM_TO_PT,
+        width=o.get("tick_width", 0.2) * MM_TO_PT,
+        direction=o.get("tick_direction", "out"),
+    )
+
+    # Number of ticks
+    ax.xaxis.set_major_locator(MaxNLocator(nbins=o.get("n_ticks", 4)))
+    ax.yaxis.set_major_locator(MaxNLocator(nbins=o.get("n_ticks", 4)))
+
+    # Grid
+    if o.get("grid"):
+        ax.grid(True, linewidth=o.get("axis_width", 0.2) * MM_TO_PT, alpha=0.3)
+
+    # Axis limits
+    if o.get("xlim"):
+        ax.set_xlim(o["xlim"])
+    if o.get("ylim"):
+        ax.set_ylim(o["ylim"])
+
+    # Spines
+    if o.get("hide_top_spine", True):
+        ax.spines["top"].set_visible(False)
+    if o.get("hide_right_spine", True):
+        ax.spines["right"].set_visible(False)
+
+    for spine in ax.spines.values():
+        spine.set_linewidth(o.get("axis_width", 0.2) * MM_TO_PT)
+
+    # Annotations
+    for annot in o.get("annotations", []):
+        if annot.get("type") == "text":
+            ax.text(
+                annot.get("x", 0.5),
+                annot.get("y", 0.5),
+                annot.get("text", ""),
+                transform=ax.transAxes,
+                fontsize=annot.get("fontsize", o.get("axis_fontsize", 7)),
+            )
+
+    fig.tight_layout()
+
+    # Draw before collecting bboxes so we have accurate positions
+    fig.canvas.draw()
+
+    # Draw hover/selection highlights for non-trace elements
+    draw_element_highlights(state, fig, ax)
+
+    # Store axes transform info for click-to-select
+    fig.canvas.draw()
+    ax_bbox = ax.get_position()
+    fig_width_px = int(fig_size[0] * preview_dpi)
+    fig_height_px = int(fig_size[1] * preview_dpi)
+
+    # Collect element bboxes for click detection
+    _collect_element_bboxes(state, fig, ax)
+
+    # Convert to RGBA data for DearPyGui texture
+    buf = io.BytesIO()
+    fig.savefig(
+        buf,
+        format="png",
+        dpi=preview_dpi,
+        bbox_inches="tight",
+        facecolor="white",
+        edgecolor="none",
+    )
+    buf.seek(0)
+
+    # Load with PIL and convert to normalized RGBA
+    img = Image.open(buf).convert("RGBA")
+    width, height = img.size
+
+    # Resize to fit within max preview size while preserving aspect ratio
+    max_width, max_height = 800, 600
+    ratio = min(max_width / width, max_height / height)
+    new_width = int(width * ratio)
+    new_height = int(height * ratio)
+    img = img.resize((new_width, new_height), Image.LANCZOS)
+
+    # Store preview bounds for coordinate conversion (after resize)
+    x_offset = (max_width - new_width) // 2
+    y_offset = (max_height - new_height) // 2
+    state.preview_bounds = (x_offset, y_offset, new_width, new_height)
+
+    # Scale element bboxes to preview coordinates
+    _scale_element_bboxes(state, ratio, x_offset, y_offset, new_height)
+
+    # Store axes transform info (scaled to resized image)
+    ax_x0 = int(ax_bbox.x0 * new_width)
+    ax_y0 = int((1 - ax_bbox.y1) * new_height)  # Flip y (0 at top)
+    ax_width = int(ax_bbox.width * new_width)
+    ax_height = int(ax_bbox.height * new_height)
+    xlim = ax.get_xlim()
+    ylim = ax.get_ylim()
+    state.axes_transform = (ax_x0, ax_y0, ax_width, ax_height, xlim, ylim)
+
+    # Create background - checkerboard for transparent, white otherwise
+    transparent = o.get("transparent", True)
+    if transparent:
+        padded = create_checkerboard(max_width, max_height, square_size=10)
+    else:
+        padded = Image.new("RGBA", (max_width, max_height), (255, 255, 255, 255))
+
+    # Paste figure centered on background
+    padded.paste(img, (x_offset, y_offset), img)
+    img = padded
+    width, height = max_width, max_height
+
+    # Cache the base image (without highlights) for fast hover updates
+    state.cached_base_image = img.copy()
+    state.cache_dirty = False
+
+    # Convert to normalized float array for DearPyGui
+    img_array = np.array(img).astype(np.float32) / 255.0
+    img_data = img_array.flatten().tolist()
+
+    plt.close(fig)
+
+    # Update texture data
+    dpg.set_value("preview_texture", img_data)
+
+    return img_data, width, height
+
+
+def _collect_element_bboxes(state: "EditorState", fig, ax) -> None:
+    """Collect element bboxes for click detection."""
+    renderer = fig.canvas.get_renderer()
+    state.element_bboxes_raw = {}
+
+    # Title bbox
+    if ax.title.get_text():
+        try:
+            title_bbox = ax.title.get_window_extent(renderer)
+            state.element_bboxes_raw["title"] = (
+                title_bbox.x0,
+                title_bbox.y0,
+                title_bbox.x1,
+                title_bbox.y1,
+            )
+        except Exception:
+            pass
+
+    # X label bbox
+    if ax.xaxis.label.get_text():
+        try:
+            xlabel_bbox = ax.xaxis.label.get_window_extent(renderer)
+            state.element_bboxes_raw["xlabel"] = (
+                xlabel_bbox.x0,
+                xlabel_bbox.y0,
+                xlabel_bbox.x1,
+                xlabel_bbox.y1,
+            )
+        except Exception:
+            pass
+
+    # Y label bbox
+    if ax.yaxis.label.get_text():
+        try:
+            ylabel_bbox = ax.yaxis.label.get_window_extent(renderer)
+            state.element_bboxes_raw["ylabel"] = (
+                ylabel_bbox.x0,
+                ylabel_bbox.y0,
+                ylabel_bbox.x1,
+                ylabel_bbox.y1,
+            )
+        except Exception:
+            pass
+
+    # Legend bbox
+    legend = ax.get_legend()
+    if legend:
+        try:
+            legend_bbox = legend.get_window_extent(renderer)
+            state.element_bboxes_raw["legend"] = (
+                legend_bbox.x0,
+                legend_bbox.y0,
+                legend_bbox.x1,
+                legend_bbox.y1,
+            )
+        except Exception:
+            pass
+
+    # X axis (bottom spine area)
+    try:
+        xaxis_bbox = ax.spines["bottom"].get_window_extent(renderer)
+        state.element_bboxes_raw["xaxis"] = (
+            xaxis_bbox.x0,
+            xaxis_bbox.y0 - 20,
+            xaxis_bbox.x1,
+            xaxis_bbox.y1 + 10,
+        )
+    except Exception:
+        pass
+
+    # Y axis (left spine area)
+    try:
+        yaxis_bbox = ax.spines["left"].get_window_extent(renderer)
+        state.element_bboxes_raw["yaxis"] = (
+            yaxis_bbox.x0 - 20,
+            yaxis_bbox.y0,
+            yaxis_bbox.x1 + 10,
+            yaxis_bbox.y1,
+        )
+    except Exception:
+        pass
+
+
+def _scale_element_bboxes(
+    state: "EditorState",
+    ratio: float,
+    x_offset: int,
+    y_offset: int,
+    new_height: int,
+) -> None:
+    """Scale element bboxes to preview coordinates."""
+    state.element_bboxes = {}
+    for elem_type, raw_bbox in state.element_bboxes_raw.items():
+        if raw_bbox is None:
+            continue
+        rx0, ry0, rx1, ry1 = raw_bbox
+        # Scale to resized image
+        sx0 = int(rx0 * ratio) + x_offset
+        sx1 = int(rx1 * ratio) + x_offset
+        # Flip Y coordinate (matplotlib origin is bottom, preview is top)
+        sy0 = new_height - int(ry1 * ratio) + y_offset
+        sy1 = new_height - int(ry0 * ratio) + y_offset
+        state.element_bboxes[elem_type] = (sx0, sy0, sx1, sy1)
+
+
+def draw_element_highlights(state: "EditorState", fig, ax) -> None:
+    """Draw selection highlights for non-trace elements."""
+    from matplotlib.patches import FancyBboxPatch
+
+    renderer = fig.canvas.get_renderer()
+
+    selected_type = (
+        state.selected_element.get("type") if state.selected_element else None
+    )
+
+    # Skip if selecting traces (handled separately in plot_from_csv)
+    if selected_type == "trace":
+        selected_type = None
+
+    def add_highlight_box(text_obj, color, alpha, linewidth=2):
+        """Add highlight rectangle around a text object (outline only)."""
+        try:
+            bbox = text_obj.get_window_extent(renderer)
+            fig_bbox = bbox.transformed(fig.transFigure.inverted())
+            padding = 0.01
+            rect = FancyBboxPatch(
+                (fig_bbox.x0 - padding, fig_bbox.y0 - padding),
+                fig_bbox.width + 2 * padding,
+                fig_bbox.height + 2 * padding,
+                boxstyle="round,pad=0.02,rounding_size=0.01",
+                facecolor="none",
+                edgecolor=color,
+                alpha=0.7,
+                linewidth=linewidth,
+                transform=fig.transFigure,
+                zorder=100,
+            )
+            fig.patches.append(rect)
+        except Exception:
+            pass
+
+    def add_spine_highlight(spine, color, alpha, linewidth=2):
+        """Add highlight to a spine/axis (outline only)."""
+        try:
+            bbox = spine.get_window_extent(renderer)
+            fig_bbox = bbox.transformed(fig.transFigure.inverted())
+            padding = 0.01
+            rect = FancyBboxPatch(
+                (fig_bbox.x0 - padding, fig_bbox.y0 - padding),
+                fig_bbox.width + 2 * padding,
+                fig_bbox.height + 2 * padding,
+                boxstyle="round,pad=0.01",
+                facecolor="none",
+                edgecolor=color,
+                alpha=0.7,
+                linewidth=linewidth,
+                transform=fig.transFigure,
+                zorder=100,
+            )
+            fig.patches.append(rect)
+        except Exception:
+            pass
+
+    # Map element types to matplotlib objects
+    element_map = {
+        "title": ax.title,
+        "xlabel": ax.xaxis.label,
+        "ylabel": ax.yaxis.label,
+    }
+
+    # Draw selection highlight (outline only, no fill)
+    select_color = "#FFC850"
+    if selected_type in element_map:
+        add_highlight_box(element_map[selected_type], select_color, 0.0, linewidth=2)
+    elif selected_type == "xaxis":
+        add_spine_highlight(ax.spines["bottom"], select_color, 0.0, linewidth=2)
+    elif selected_type == "yaxis":
+        add_spine_highlight(ax.spines["left"], select_color, 0.0, linewidth=2)
+    elif selected_type == "legend":
+        legend = ax.get_legend()
+        if legend:
+            try:
+                bbox = legend.get_window_extent(renderer)
+                fig_bbox = bbox.transformed(fig.transFigure.inverted())
+                padding = 0.01
+                rect = FancyBboxPatch(
+                    (fig_bbox.x0 - padding, fig_bbox.y0 - padding),
+                    fig_bbox.width + 2 * padding,
+                    fig_bbox.height + 2 * padding,
+                    boxstyle="round,pad=0.02",
+                    facecolor="none",
+                    edgecolor=select_color,
+                    alpha=0.7,
+                    linewidth=2,
+                    transform=fig.transFigure,
+                    zorder=100,
+                )
+                fig.patches.append(rect)
+            except Exception:
+                pass
+
+
+# EOF