tracker/app.py
Feiko Wielsma ad48295484
Some checks failed
Deploy via SSH / deploy (push) Failing after 1s
Fix NaT bug on empty states and update volume type
2026-03-25 13:11:23 +01:00

295 lines
12 KiB
Python

import streamlit as st
from streamlit_calendar import calendar
import pandas as pd
import database
import datetime
# --- Configuration ---
st.set_page_config(page_title="Work Tracker", layout="wide")
# --- Database Init ---
if "db_initialized" not in st.session_state:
database.init_db()
st.session_state["db_initialized"] = True
# --- UI Layout ---
st.title("📅 Work Hours Tracker")
col1, col2 = st.columns([2, 1])
# --- Sidebar Controls ---
with st.sidebar:
st.header("Settings")
# Global Config for Stats
DAILY_HOURS = st.slider("Work Hours per Day (for Stats)", min_value=6.0, max_value=8.0, value=8.0, step=0.5)
st.divider()
mode = st.radio("Mode", ["Mark Work Day", "Mark Holiday", "Remove Day"])
course_name = ""
if mode == "Mark Work Day":
course_name = st.text_input("Course Name", placeholder="e.g. Python 101")
st.info(f"Current Mode: **{mode}**")
if mode == "Mark Work Day":
if course_name:
st.caption(f"Course: **{course_name}**")
# --- Calendar Section ---
with col1:
st.subheader("Select Days")
# Fetch existing data to populate calendar
df = database.get_all_days()
# Format events for streamlit-calendar
events = []
if not df.empty:
for _, row in df.iterrows():
date_str = row['date']
day_type = row.get('type', 'work')
c_name = row.get('course_name', '')
# Get hours for tooltip/display if needed, though simple title is usually best for calendar
title = "Worked"
color = "#4CAF50" # Green
if day_type == 'holiday':
title = "Holiday"
color = "#FF9800" # Orange
elif c_name:
title = c_name
if row.get('travel_costs'):
title = f"🚗 {title}"
events.append({
"title": title,
"start": date_str,
"allDay": True,
"color": color,
# Store extra props if needed
"extendedProps": {
"type": day_type,
"course": c_name,
"hours": row.get('hours', 8.0),
"travel_costs": bool(row.get('travel_costs'))
}
})
# Calendar options
calendar_options = {
"headerToolbar": {
"left": "today prev,next",
"center": "title",
"right": "dayGridMonth,dayGridWeek"
},
"initialView": "dayGridMonth",
"selectable": True,
"timeZone": "UTC", # Force UTC to prevent off-by-one errors from local timezone
}
# Render Calendar
cal = calendar(events=events, options=calendar_options, key="calendar")
# Handle Click Events
# 1. Date Click (Empty cell)
if cal.get("callback") == "dateClick":
# Check if we already processed this specific event to avoid infinite loops/double toggles
# The component might return the same event object on rerun until interaction changes
current_event_str = str(cal)
if st.session_state.get("last_processed_event") != current_event_str:
st.session_state["last_processed_event"] = current_event_str
# Parse date from ISO string (e.g. "2023-01-01T00:00:00.000Z")
if "date" in cal["dateClick"]:
clicked_date = cal["dateClick"]["date"]
if "T" in clicked_date:
clicked_date = clicked_date.split("T")[0]
# Logic based on Mode
if mode == "Remove Day":
database.remove_day(clicked_date)
elif mode == "Mark Holiday":
# Holidays are standard 8h for now, or could use the slider too if user wanted "half holiday"
database.update_day(clicked_date, day_type="holiday", course_name="Holiday", hours=DAILY_HOURS)
else: # Mark Work Day
database.update_day(clicked_date, day_type="work", course_name=course_name, hours=DAILY_HOURS, travel_costs=True)
st.rerun() # Refresh to update calendar and stats
# 2. Event Click (Existing item)
elif cal.get("callback") == "eventClick":
current_event_str = str(cal)
if st.session_state.get("last_processed_event") != current_event_str:
st.session_state["last_processed_event"] = current_event_str
event_data = cal["eventClick"]["event"]
extended_props = event_data.get("extendedProps", {})
# Only toggle travel costs for "work" days
if extended_props.get("type") == "work":
clicked_date = event_data["start"]
if "T" in clicked_date:
clicked_date = clicked_date.split("T")[0]
# Invert current travel cost status
current_travel_costs = extended_props.get("travel_costs", False)
new_travel_costs = not current_travel_costs
# Update database
database.update_day(
clicked_date,
day_type="work",
course_name=extended_props.get("course"),
hours=extended_props.get("hours", DAILY_HOURS),
travel_costs=new_travel_costs
)
st.rerun()
# --- Statistics Section ---
with col2:
st.subheader("Statistics")
# Configuration: Start Date
START_DATE = pd.Timestamp("2025-09-15")
# Convert to datetime BEFORE filtering
df['date'] = pd.to_datetime(df['date'])
# Filter for data >= Start Date
df = df[df['date'] >= START_DATE]
df = df.sort_values('date')
# Determine "effective" days and hours
# User requested Holidays count as full days for simplicity
def get_effective_days(row):
return 1.0
def get_effective_hours(row):
# Use the global setting, ignoring DB value if necessary
# (Or could prioritize DB if users want override? User asked to recalculate based on entered days)
# "I don't want it to change the actual hours worked. I just want to... recalculate the averages"
# This implies ignoring stored hour values and reapplying the global constant.
return DAILY_HOURS
# Ensure 'type' exists
if 'type' not in df.columns:
df['type'] = 'work'
df['effective_days'] = df.apply(get_effective_days, axis=1)
df['effective_hours'] = df.apply(get_effective_hours, axis=1)
# --- Calculate Weekly Stats ---
# Resample to weekly frequency (W-MON)
weekly_data = df.resample('W-MON', on='date')[['effective_days', 'effective_hours']].sum()
# Reindex to ensure we cover from START_DATE to Today (or Max Date)
# This ensures weeks with 0 work are counted in the average
max_date = pd.Timestamp.now() if df.empty or pd.isna(df['date'].max()) else max(df['date'].max(), pd.Timestamp.now())
full_weeks = pd.date_range(start=START_DATE, end=max_date, freq='W-MON')
weekly_data = weekly_data.reindex(full_weeks, fill_value=0)
weekly_data.columns = ['Days', 'Hours']
if not weekly_data.empty:
avg_days = weekly_data['Days'].mean()
avg_hours = weekly_data['Hours'].mean()
m1, m2 = st.columns(2)
m1.metric("Avg Days / Week", f"{avg_days:.2f}")
m2.metric("Avg Hours / Week", f"{avg_hours:.1f}")
# Plotting
st.caption("Days per Week (Effective)")
# Calculate running average (expanding mean) to show trend over time
weekly_data['Running Average'] = weekly_data['Days'].expanding().mean()
st.line_chart(weekly_data[['Days', 'Running Average']], use_container_width=True)
# --- Monthly Breakdown ---
monthly_data = df.resample('M', on='date')[['effective_days', 'effective_hours']].sum()
# Reindex for full months too
# (Use 'MS' for Month Start to align better, or 'M' for End)
# But reindexing with 'M' is tricky if today is mid-month.
# Let's stick to simple reindexing of what we have + fillna(0) for displayed months
# Actually, standard resample handles missing months WITHIN the range.
# We just need to make sure the range starts at START_DATE.
full_months = pd.date_range(start=START_DATE, end=max_date, freq='M')
monthly_data = monthly_data.reindex(full_months, fill_value=0)
monthly_data.columns = ['Days Worked', 'Hours Worked']
# Calculate Avg Hours/Week for the month
# (Total Hours / (Days in Month / 7))
monthly_data['Avg Hours/Week'] = monthly_data['Hours Worked'] / (monthly_data.index.days_in_month / 7)
# Format index to be more readable (e.g. "January 2023")
monthly_data.index = monthly_data.index.strftime('%B %Y')
st.write("### Monthly Breakdown")
st.dataframe(monthly_data.style.format({
"Avg Hours/Week": "{:.1f}",
"Hours Worked": "{:.1f}",
"Days Worked": "{:.1f}"
}))
st.divider()
st.write("### 🚗 Export Travel Costs")
# Select month for export
available_months = df.resample('M', on='date').size().index.strftime('%B %Y').tolist()
if available_months:
selected_export_month = st.selectbox("Select Month to Export", reversed(available_months))
if st.button("Generate Export Text"):
# Filter data for selected month
month_dt = pd.to_datetime(selected_export_month, format='%B %Y')
month_df = df[(df['date'].dt.month == month_dt.month) & (df['date'].dt.year == month_dt.year)]
if not month_df.empty:
export_lines = ["Hierbij ook nog even het lijstje aan werkdagen met reiskosten:"]
# Group by course
courses = month_df['course_name'].unique()
total_travel_days = 0
for course in courses:
if not course or course == 'Holiday':
continue
course_df = month_df[month_df['course_name'] == course].sort_values('date')
travel_days = course_df[course_df['travel_costs'] == 1]
no_travel_days = course_df[course_df['travel_costs'] == 0]
if not travel_days.empty:
dates_str = ", ".join(travel_days['date'].dt.strftime('%d %b'))
line = f"{course}: {dates_str}"
# Add "skipped" days if they exist within the range
if not no_travel_days.empty:
skip_dates = ", ".join(no_travel_days['date'].dt.strftime('%d %b'))
line += f" (dus niet {skip_dates} want toen was het Online)"
export_lines.append(line)
total_travel_days += len(travel_days)
export_lines.append("")
export_lines.append(f"Totaal lijkt het dus {total_travel_days} dagen geweest te zijn. Zoals eerder bekeken leek het zo'n 30km per dag.")
export_text = "\n".join(export_lines)
st.code(export_text, language="text")
st.success("You can copy the text above!")
else:
st.warning("No data found for this month.")
else:
st.info("No work days logged yet. Click dates on the calendar.")
# --- Debugging (Optional) ---
# with st.expander("Raw Data"):
# st.dataframe(df)