Files
CANveyor/main.py
T

598 lines
21 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# File: main.py
# Copyright (C) 2026 Erick Ahmed
# SPDX-License-Identifier: AGPL-3.0-or-later
"""CANveyor dashboard entry point.
Handles raw log ingestion, J1939 decoding, precomputation of
statistical figures, and exposes a Dash application for browsing the
processed data.
"""
import os
from concurrent.futures import ThreadPoolExecutor
from pathlib import Path
from typing import Dict, List, Tuple
import dash
import dash_bootstrap_components as dbc
import numpy as np
import pandas as pd
import polars as pl
from dash import dcc, html, Input, Output, State
from decoder import decode_j1939_frames
from logs.view import get_logs_table_component, prepare_logs_data
from parser import parse_csv, parse_log
from stats.correlation import calculate_correlation, plot_correlation_heatmap
from stats.entropy import calculate_byte_entropy, plot_entropy_heatmap
from stats.frequency import calculate_frequency, plot_frequency
from stats.id_viewer import _format_can_id_vec, plot_bits
from stats.utils.loader import load_data
from vehicle import get_vehicle_module
RAW_LOG_DIR = "data/logs"
CSV_DIR = "data/csv"
PARQUET_DIR = "data/parquet"
PAGE_SIZE = 25_000
BYTE_COLS = [f"b{i}" for i in range(8)]
BUS_OPTIONS = [
{"label": "Bus 1", "value": "Bus 1"},
{"label": "Bus 2", "value": "Bus 2"},
]
DATA: Dict[str, Dict[str, pd.DataFrame]] = {}
VEHICLE_META: Dict[str, Dict[str, str]] = {}
PRECOMPUTED_FIGURES: Dict[str, object] = {}
DATA_BY_ID: Dict[Tuple[str, str], Dict[str, Tuple[pd.DataFrame, List[str]]]] = {}
CORR_CACHE: Dict[Tuple, object] = {}
PREPARED_LOGS_CACHE: Dict[Tuple[str, str], pd.DataFrame] = {}
def parse_vehicle_from_filename(filename: str) -> Tuple[str, str, str]:
"""Derive (vehicle, brand, model) from a log file name."""
stem = Path(filename).stem
if "-" in stem:
brand, model_part = stem.split("-", 1)
else:
brand, model_part = stem, "Unknown"
model = model_part.replace("_", " ")
vehicle = f"{brand} {model}".strip()
return vehicle, brand, model
def _vehicle_paths(vehicle: str) -> Dict[str, str]:
"""Return all intermediate file paths for a given vehicle."""
return {
"bus1_csv": f"{CSV_DIR}/{vehicle}_bus1.csv",
"bus2_csv": f"{CSV_DIR}/{vehicle}_bus2.csv",
"bus1_parquet": f"{PARQUET_DIR}/{vehicle}_bus1.parquet",
"bus2_parquet": f"{PARQUET_DIR}/{vehicle}_bus2.parquet",
"bus1_decoded": f"{PARQUET_DIR}/{vehicle}_bus1_decoded.parquet",
"bus2_decoded": f"{PARQUET_DIR}/{vehicle}_bus2_decoded.parquet",
}
def run_pipeline() -> None:
"""Parse raw log files, convert to parquet, and decode J1939 frames."""
for directory in (RAW_LOG_DIR, CSV_DIR, PARQUET_DIR):
os.makedirs(directory, exist_ok=True)
for log_file in Path(RAW_LOG_DIR).glob("*.txt"):
vehicle, _, _ = parse_vehicle_from_filename(log_file.name)
paths = _vehicle_paths(vehicle)
if Path(paths["bus1_decoded"]).exists() and Path(paths["bus2_decoded"]).exists():
continue
print(f"Parsing raw log: {log_file.name}...")
parse_log(str(log_file), paths["bus1_csv"], paths["bus2_csv"])
print("Converting to parquet...")
parse_csv(paths["bus1_csv"]).sink_parquet(paths["bus1_parquet"])
parse_csv(paths["bus2_csv"]).sink_parquet(paths["bus2_parquet"])
print("Decoding J1939...")
df1 = pl.read_parquet(paths["bus1_parquet"])
df2 = pl.read_parquet(paths["bus2_parquet"])
decode_j1939_frames(df1).write_parquet(paths["bus1_decoded"])
decode_j1939_frames(df2).write_parquet(paths["bus2_decoded"])
def load_vehicle_data() -> None:
"""Load all decoded parquet files into the in-memory DATA store."""
for log_file in Path(RAW_LOG_DIR).glob("*.txt"):
vehicle, brand, model = parse_vehicle_from_filename(log_file.name)
VEHICLE_META[vehicle] = {"brand": brand, "model": model}
paths = _vehicle_paths(vehicle)
if Path(paths["bus1_decoded"]).exists() and Path(paths["bus2_decoded"]).exists():
DATA[vehicle] = {
"Bus 1": load_data(paths["bus1_decoded"]),
"Bus 2": load_data(paths["bus2_decoded"]),
}
def _can_id_column(df: pd.DataFrame) -> str:
return "ID" if "ID" in df.columns else "Identifier"
def _downsample_unchanged(group: pd.DataFrame, byte_cols: List[str]) -> pd.DataFrame:
"""Keep only rows where at least one byte changed vs. the previous row."""
if group.empty or not byte_cols:
return group
arr = group[byte_cols].to_numpy(dtype=np.float32, copy=False)
if len(arr) <= 1:
return group
changed = np.any(arr[1:] != arr[:-1], axis=1)
keep = np.concatenate(([True], changed))
return group.iloc[keep]
def process_bus_data(
vehicle: str, bus: str, df: pd.DataFrame
) -> Tuple[str, str, Dict[str, object], Dict[str, Tuple[pd.DataFrame, List[str]]]]:
"""Compute per-bus figures and ID-grouped, downsampled frames."""
precomp = {
f"{vehicle}_{bus}_freq": plot_frequency(
calculate_frequency(df), title=f"{vehicle} {bus} Frequency"
),
f"{vehicle}_{bus}_entropy": plot_entropy_heatmap(
calculate_byte_entropy(df), title=f"{vehicle} {bus} Byte-Level Entropy"
),
}
df = df.assign(Formatted_ID=_format_can_id_vec(df[_can_id_column(df)]))
df = df.sort_values(["Formatted_ID", "Timestamp"], kind="stable")
grouped: Dict[str, Tuple[pd.DataFrame, List[str]]] = {}
for can_id, group in df.groupby(by="Formatted_ID"):
byte_cols = [c for c in BYTE_COLS if c in group.columns]
group = _downsample_unchanged(group, byte_cols)
grouped[can_id] = (group, byte_cols)
return vehicle, bus, precomp, grouped
def precompute_all() -> None:
"""Run :func:`process_bus_data` across every vehicle/bus pair in parallel."""
with ThreadPoolExecutor() as executor:
futures = [
executor.submit(process_bus_data, vehicle, bus, df)
for vehicle, buses in DATA.items()
for bus, df in buses.items()
]
for future in futures:
v, b, precomp, grouped = future.result()
PRECOMPUTED_FIGURES.update(precomp)
DATA_BY_ID[(v, b)] = grouped
run_pipeline()
print("Loading data into memory...")
load_vehicle_data()
precompute_all()
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
app.config.suppress_callback_exceptions = True
def _vehicle_dropdown(dropdown_id: str) -> dcc.Dropdown:
return dcc.Dropdown(
id=dropdown_id,
options=[{"label": v, "value": v} for v in DATA.keys()],
value=list(DATA.keys())[0] if DATA else None,
clearable=False,
)
def _bus_dropdown(dropdown_id: str) -> dcc.Dropdown:
return dcc.Dropdown(
id=dropdown_id,
options=BUS_OPTIONS,
value="Bus 1",
clearable=False,
)
def _label(text: str) -> html.Label:
return html.Label(text, className="mt-2")
app.layout = dbc.Container(
[
html.H1("CANveyor", className="my-4"),
dbc.Tabs(
[
dbc.Tab(
label="Overview",
tab_id="overview",
children=[html.Div(id="overview-content")],
),
dbc.Tab(
label="Vehicles",
tab_id="vehicles",
children=[
dbc.Row(
[
dbc.Col(_label("Vehicle:"), width="auto"),
dbc.Col(
_vehicle_dropdown("vehicles-vehicle-selector"),
width=3, className="me-4",
),
],
className="mb-3 mt-3", align="end",
),
html.Div(id="vehicles-content", className="mt-3"),
],
),
dbc.Tab(
label="Logs",
tab_id="logs",
children=[
dbc.Row(
[
dbc.Col(_label("Vehicle:"), width="auto"),
dbc.Col(
_vehicle_dropdown("logs-vehicle-selector"),
width=3, className="me-4",
),
dbc.Col(_label("Bus:"), width="auto"),
dbc.Col(
_bus_dropdown("logs-bus-selector"),
width=2,
),
],
className="mb-3 mt-3", align="end",
),
get_logs_table_component(),
],
),
dbc.Tab(
label="Statistics",
tab_id="statistics",
children=[
dbc.Row(
[
dbc.Col(_label("Vehicle:"), width="auto"),
dbc.Col(
_vehicle_dropdown("vehicle-selector"),
width=3, className="me-4",
),
dbc.Col(_label("Bus:"), width="auto"),
dbc.Col(
_bus_dropdown("bus-selector"),
width=2,
),
],
className="mb-3 mt-3", align="end",
),
dbc.Tabs(
[
dbc.Tab(label="Frequency", tab_id="freq"),
dbc.Tab(label="ID Viewer", tab_id="id_viewer"),
dbc.Tab(label="Correlation", tab_id="corr"),
dbc.Tab(label="Entropy", tab_id="entropy"),
],
id="tabs",
active_tab="freq",
),
html.Div(id="tab-content", className="mt-3"),
],
),
],
id="main-tabs",
active_tab="statistics",
),
],
fluid=True,
)
def get_prepared_logs(vehicle: str, bus: str) -> pd.DataFrame:
"""Lazily prepare and cache log table data for a vehicle/bus pair."""
cache_key = (vehicle, bus)
if cache_key not in PREPARED_LOGS_CACHE:
PREPARED_LOGS_CACHE[cache_key] = prepare_logs_data(DATA[vehicle][bus])
return PREPARED_LOGS_CACHE[cache_key]
def build_page_buttons(
current_page: int, total_pages: int, max_buttons: int = 15
) -> List:
"""Build the pagination button list with ellipses where appropriate."""
buttons: List = []
if total_pages <= 1:
return buttons
half = max_buttons // 2
start = max(0, current_page - half)
end = min(total_pages, start + max_buttons)
if end - start < max_buttons:
start = max(0, end - max_buttons)
if start > 0:
buttons.append(
dbc.Button(
"1",
id={"type": "page-btn", "index": 0},
color="secondary", outline=True, size="sm", className="me-1",
)
)
if start > 1:
buttons.append(html.Span("…", className="mx-1 align-middle"))
for i in range(start, end):
is_current = i == current_page
buttons.append(
dbc.Button(
str(i + 1),
id={"type": "page-btn", "index": i},
size="sm",
color="primary" if is_current else "secondary",
outline=not is_current,
className="me-1",
disabled=is_current,
)
)
if end < total_pages:
if end < total_pages - 1:
buttons.append(html.Span("…", className="mx-1 align-middle"))
buttons.append(
dbc.Button(
str(total_pages),
id={"type": "page-btn", "index": total_pages - 1},
color="secondary", outline=True, size="sm", className="me-1",
)
)
return buttons
@app.callback(
Output("logs-table", "data"),
Output("logs-table", "columns"),
Output("logs-info-text", "children"),
Output("logs-page-nav", "children"),
Output("logs-current-page", "data"),
Input("logs-vehicle-selector", "value"),
Input("logs-bus-selector", "value"),
Input("logs-prev-btn", "n_clicks"),
Input("logs-next-btn", "n_clicks"),
Input({"type": "page-btn", "index": dash.ALL}, "n_clicks"),
State("logs-current-page", "data"),
)
def update_logs_table(vehicle, bus, prev_clicks, next_clicks, page_btn_clicks, current_page):
if not vehicle or not bus or vehicle not in DATA or bus not in DATA[vehicle]:
return [], [], "No data available", [], 0
prepared_df = get_prepared_logs(vehicle, bus)
total_rows = len(prepared_df)
if total_rows == 0:
return [], [], "No data available", [], 0
total_pages = max(1, (total_rows + PAGE_SIZE - 1) // PAGE_SIZE)
ctx = dash.callback_context
triggered_id = ctx.triggered_id
current_page = current_page if current_page is not None else 0
if triggered_id in ("logs-vehicle-selector", "logs-bus-selector"):
current_page = 0
elif triggered_id == "logs-prev-btn":
current_page = max(0, current_page - 1)
elif triggered_id == "logs-next-btn":
current_page = current_page + 1
elif isinstance(triggered_id, dict) and triggered_id.get("type") == "page-btn":
if ctx.triggered and ctx.triggered[0]["value"]:
current_page = triggered_id["index"]
current_page = max(0, min(current_page, total_pages - 1))
start_idx = current_page * PAGE_SIZE
end_idx = min(start_idx + PAGE_SIZE, total_rows)
page_data = prepared_df.iloc[start_idx:end_idx].to_dict("records")
columns = [{"name": c, "id": c} for c in prepared_df.columns]
info_text = (
f"Page {current_page + 1} of {total_pages} | "
f"Showing rows {start_idx + 1:,}{end_idx:,} "
f"of {total_rows:,} total frames"
)
page_buttons = build_page_buttons(current_page, total_pages)
return page_data, columns, info_text, page_buttons, current_page
@app.callback(
Output("tab-content", "children"),
Input("tabs", "active_tab"),
Input("vehicle-selector", "value"),
Input("bus-selector", "value"),
)
def render_content(tab, vehicle, bus):
if not vehicle or not bus or vehicle not in DATA or bus not in DATA[vehicle]:
return html.Div("No data available")
df = DATA[vehicle][bus]
if tab == "freq":
return dcc.Graph(
figure=PRECOMPUTED_FIGURES[f"{vehicle}_{bus}_freq"],
style={"height": "80vh"},
)
if tab == "id_viewer":
ids = sorted(DATA_BY_ID.get((vehicle, bus), {}).keys())
return html.Div(
[
html.Label("CAN ID:"),
dcc.Dropdown(
id="id-selector",
options=[{"label": i, "value": i} for i in ids],
value=ids[0] if ids else None,
clearable=False,
style={"width": "50%", "marginBottom": "10px"},
),
dcc.Graph(id="id-viewer-graph", style={"height": "70vh"}),
]
)
if tab == "corr":
ids = sorted(DATA_BY_ID.get((vehicle, bus), {}).keys())
return html.Div(
[
dbc.Row(
[
dbc.Col(html.Label("Method:"), width=1, className="mt-2"),
dbc.Col(
dcc.Dropdown(
id="corr-method",
options=[
{"label": "Pearson", "value": "pearson"},
{"label": "Spearman", "value": "spearman"},
],
value="pearson",
clearable=False,
),
width=2,
),
dbc.Col(html.Label("Target ID:"), width=1, className="mt-2"),
dbc.Col(
dcc.Dropdown(
id="corr-target",
options=[{"label": "All IDs (Max Corr)", "value": "all"}]
+ [{"label": i, "value": i} for i in ids],
value="all",
clearable=True,
),
width=4,
),
],
className="mb-3",
),
dcc.Graph(id="corr-graph", style={"height": "80vh"}),
]
)
if tab == "entropy":
return dcc.Graph(
figure=PRECOMPUTED_FIGURES[f"{vehicle}_{bus}_entropy"],
style={"height": "80vh"},
)
return html.Div("Tab not found")
@app.callback(
Output("id-viewer-graph", "figure"),
Input("id-selector", "value"),
Input("vehicle-selector", "value"),
Input("bus-selector", "value"),
Input("tabs", "active_tab"),
)
def update_id_viewer(selected_id, vehicle, bus, tab):
if tab != "id_viewer" or not selected_id or not vehicle or not bus:
return dash.no_update
grouped_data = DATA_BY_ID.get((vehicle, bus), {})
if selected_id not in grouped_data:
return dash.no_update
filtered_df, byte_cols = grouped_data[selected_id]
return plot_bits(
filtered_df, byte_cols, selected_id,
title=f"{vehicle} {bus} Byte Visualization",
)
@app.callback(
Output("corr-graph", "figure"),
Input("corr-method", "value"),
Input("corr-target", "value"),
Input("vehicle-selector", "value"),
Input("bus-selector", "value"),
Input("tabs", "active_tab"),
)
def update_corr(method, target, vehicle, bus, tab):
if tab != "corr" or not vehicle or not bus:
return dash.no_update
target_id = None if target == "all" or not target else target
cache_key = (vehicle, bus, method, target_id)
if cache_key not in CORR_CACHE:
df = DATA[vehicle][bus]
CORR_CACHE[cache_key] = calculate_correlation(
df, method=method, target_id=target_id
)
corr_df = CORR_CACHE[cache_key]
title = f"{vehicle} {bus} Correlation"
if target_id:
title += f" ({target_id})"
return plot_correlation_heatmap(corr_df, target_id=target_id, title=title)
@app.callback(
Output("vehicles-content", "children"),
Input("vehicles-vehicle-selector", "value"),
)
def render_vehicles(vehicle):
if not vehicle or vehicle not in DATA:
return html.Div("No data available", className="text-muted")
brand = VEHICLE_META.get(vehicle, {}).get("brand", "")
vehicle_module = get_vehicle_module(brand)
dfs = list(DATA[vehicle].values())
if not dfs:
return html.Div("No data available", className="text-muted")
df = pd.concat(dfs, ignore_index=True)
if "Timestamp" in df.columns:
df = df.sort_values("Timestamp", kind="stable").reset_index(drop=True)
cards = []
for nid, frame_def in vehicle_module.DECODER_RULES.items():
decoded = vehicle_module.decode_dataframe(df, frame_def.can_id)
for item in frame_def.signals:
if hasattr(item, "plot_func") and callable(item.plot_func):
title = f"{frame_def.can_id} - {item.name}"
fig = item.plot_func(decoded, frame_def.color)
else:
sig = item
if getattr(sig, "skip_plot", False):
continue
unit_str = f" ({sig.unit})" if sig.unit else ""
title = f"{sig.name}{unit_str}"
fig = vehicle_module.plot_signal(
decoded, sig.name, title=title, color=frame_def.color
)
cards.append(
dbc.Col(
dbc.Card(
[
dbc.CardBody(
[
dcc.Graph(
figure=fig,
config={"displayModeBar": False},
style={"height": "280px"},
)
],
className="p-2",
),
],
className="shadow-sm border-0 h-100",
),
xs=12, sm=6, md=4, lg=3, className="mb-3",
)
)
if not cards:
return html.Div(
"No decoded signals available. Add rules in the vehicle module.",
className="text-muted",
)
return dbc.Row(cards)
if __name__ == "__main__":
app.run(debug=False)