map color to speaker to maintain color consistency across all plotly

utils.py

@@ -338,51 +338,48 @@ def build_fig_pie1(df3, catTypeColors):
     return fig
 
 
-def build_fig_pie2(df4, speakerNames, speakerColors, catColors, get_display_name_fn, currFile):
+def build_fig_pie2(df4, speakerNames, speaker_color_map, catColors, get_display_name_fn, currFile):
     """Speaker / category pie chart."""
     df4 = df4.copy()
-    figColors = [
-        speakerColors[list(speakerNames).index(n)]
-        for n in df4["names"] if n in speakerNames
-    ]
     df4["names"] = df4["names"].apply(lambda s: get_display_name_fn(s, currFile))
+    colors = [speaker_color_map.get(n, _SPEAKER_PALETTE[i % len(_SPEAKER_PALETTE)])
+              for i, n in enumerate(df4["names"])]
     fig = go.Figure()
-    fig.update_layout(
-        title_text="Percentage of Speakers per Role",
-        colorway=_SPEAKER_PALETTE,
-        **TRANSPARENT_BG,
-    )
-    fig.add_trace(go.Pie(values=df4["values"], labels=df4["names"], sort=False))
+    fig.update_layout(title_text="Percentage of Speakers per Role", **TRANSPARENT_BG)
+    fig.add_trace(go.Pie(values=df4["values"], labels=df4["names"],
+                         marker_colors=colors, sort=False))
     return fig
 
 
-def _voice_color_map(df5_labels, speakerColors):
+def _voice_color_map(df5_labels, speaker_color_map):
     """Build the label->color map for sunburst/treemap charts.
 
     Single Voice → _PALETTE[0]  (red shade 0, reserved)
-    Multi Voice  → _PALETTE[3]  (green shade 0, reserved)
+    Multi Voice  → _PALETTE[9]  (green shade 0, reserved)
     No Voice     → _PALETTE[-1] (grey, reserved)
-    Speakers     → _SPEAKER_PALETTE in order (never collides with reserved)
+    Speakers     → looked up from speaker_color_map for cross-chart consistency
     """
     top_labels = ["No Voice", "Single Voice", "Multi Voice"]
     color_map = {
-        "Single Voice": _PALETTE[0],   # red shade 0
-        "Multi Voice":  _PALETTE[9],   # green shade 0
-        "No Voice":     _PALETTE[-1],  # grey
+        "Single Voice": _PALETTE[0],
+        "Multi Voice":  _PALETTE[9],
+        "No Voice":     _PALETTE[-1],
     }
     speaker_labels = [l for l in df5_labels if l not in top_labels]
     for i, lbl in enumerate(speaker_labels):
-        color_map[lbl] = _SPEAKER_PALETTE[i % len(_SPEAKER_PALETTE)]
+        color_map[lbl] = speaker_color_map.get(
+            lbl, _SPEAKER_PALETTE[i % len(_SPEAKER_PALETTE)]
+        )
     return color_map
 
 
-def build_fig_sunburst(df5, catTypeColors, speakerColors, get_display_name_fn, currFile):
+def build_fig_sunburst(df5, catTypeColors, speaker_color_map, get_display_name_fn, currFile):
     """Sunburst voice-category chart."""
     df5 = df5.copy()
     df5["labels"]      = df5["labels"].apply(lambda s: get_display_name_fn(s, currFile))
     df5["parentNames"] = df5["parentNames"].apply(lambda s: get_display_name_fn(s, currFile))
 
-    color_map = _voice_color_map(df5["labels"], speakerColors)
+    color_map = _voice_color_map(df5["labels"], speaker_color_map)
 
     fig = px.sunburst(
         df5,
@@ -405,13 +402,13 @@ def build_fig_sunburst(df5, catTypeColors, speakerColors, get_display_name_fn, c
     return fig
 
 
-def build_fig_treemap(df5, catTypeColors, speakerColors, get_display_name_fn, currFile):
+def build_fig_treemap(df5, catTypeColors, speaker_color_map, get_display_name_fn, currFile):
     """Treemap voice-category chart."""
     df5 = df5.copy()
     df5["labels"]      = df5["labels"].apply(lambda s: get_display_name_fn(s, currFile))
     df5["parentNames"] = df5["parentNames"].apply(lambda s: get_display_name_fn(s, currFile))
 
-    color_map = _voice_color_map(df5["labels"], speakerColors)
+    color_map = _voice_color_map(df5["labels"], speaker_color_map)
 
     fig = px.treemap(
         df5,
@@ -434,7 +431,7 @@ def build_fig_treemap(df5, catTypeColors, speakerColors, get_display_name_fn, cu
     return fig
 
 
-def build_fig_timeline(speakers_dataFrame, currTotalTime, speakerColors, get_display_name_fn, currFile):
+def build_fig_timeline(speakers_dataFrame, currTotalTime, speaker_color_map, get_display_name_fn, currFile):
     """Gantt-style speaker timeline."""
     df = speakers_dataFrame.copy()
     df["Resource"] = df["Resource"].apply(lambda s: get_display_name_fn(s, currFile))
@@ -455,7 +452,7 @@ def build_fig_timeline(speakers_dataFrame, currTotalTime, speakerColors, get_dis
     fig = px.timeline(
         df, x_start="Start", x_end="Finish", y="Resource", color="Resource",
         title="Timeline of Audio with Speakers",
-        color_discrete_sequence=_SPEAKER_PALETTE,
+        color_discrete_map=speaker_color_map,
     )
     fig.update_yaxes(autorange="reversed")
 
@@ -492,7 +489,7 @@ def _seconds_to_hhmmss(seconds):
     return f"{h:02d}:{m:02d}:{s:05.2f}"
 
 
-def build_fig_bar(df2, speakerNames, catColors, speakerColors, get_display_name_fn, currFile):
+def build_fig_bar(df2, speakerNames, catColors, speaker_color_map, get_display_name_fn, currFile):
     """Horizontal bar chart — time spoken per speaker (hh:mm:ss.ss).
     Only individual speakers are shown; role/category rows are excluded.
     """
@@ -505,7 +502,7 @@ def build_fig_bar(df2, speakerNames, catColors, speakerColors, get_display_name_
         df2, x="values", y="names", color="names", orientation="h",
         custom_data=["names", "time_label"],
         title="Time Spoken by each Speaker",
-        color_discrete_sequence=_SPEAKER_PALETTE,
+        color_discrete_map=speaker_color_map,
     )
     # Hide x-axis tick labels — values are crowded with many speakers.
     # The exact time is still visible on hover via the hovertemplate.