import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from datetime import datetime
import pandas as pd

# Define updated tasks and durations
updated_tasks = [
    ("Refactor BSV modules", "2025-06-01", "2025-07-01"),
    ("Setup C benchmark suite", "2025-06-01", "2025-07-01"),
    ("EuroSys paper revisions (feedback)", "2025-06-01", "2025-07-31"),
    ("Formal progression review", "2025-06-01", "2025-07-01"),
    ("Submit EuroSys for preliminary feedback", "2025-06-01", "2025-07-01"),
    ("Debug Toooba memory pipeline (TLB bypass)", "2025-07-01", "2025-08-31"),
    ("Validate C benchmark suite", "2025-07-01", "2025-08-01"),
    ("Documentation for second paper", "2025-07-15", "2025-08-31"),
    ("Finalise EuroSys revisions", "2025-07-01", "2025-07-31"),
    ("Continue memory debugging", "2025-08-01", "2025-09-30"),
    ("Draft second paper (intro/method)", "2025-08-01", "2025-09-30"),
    ("Generate experimental results", "2025-08-15", "2025-09-30"),
    ("Publish EuroSys paper", "2025-09-01", "2025-10-01"),
    ("Benchmark Toooba design", "2025-09-01", "2025-10-31"),
    ("Draft eval & analysis (second paper)", "2025-09-15", "2025-10-31"),
    ("Third experimental phase", "2025-10-01", "2025-11-30"),
    ("Modify allocators (remove mmap)", "2025-10-15", "2025-11-30"),
    ("Finalise second paper", "2025-11-01", "2025-11-30"),
    ("Strip huge-page optimisations", "2025-11-01", "2025-12-31"),
    ("Analyse instruction-level reductions", "2025-11-01", "2025-12-31"),
    ("Draft third research paper", "2025-11-15", "2025-12-31"),
    ("Evaluate/profile third paper", "2025-12-01", "2026-01-31"),
    ("Write thesis chapters (Exp 1 & 2)", "2025-12-15", "2026-01-31"),
    ("Thesis development & refinement", "2026-01-01", "2026-09-30"),
    ("Submit third paper", "2026-01-01", "2026-09-30"),
    ("Prepare full dissertation", "2026-01-01", "2026-09-30"),
]

# Convert to DataFrame
df = pd.DataFrame(updated_tasks, columns=["Task", "Start", "End"])
df["Start"] = pd.to_datetime(df["Start"])
df["End"] = pd.to_datetime(df["End"])
df["Duration"] = df["End"] - df["Start"]

# Plot
fig, ax = plt.subplots(figsize=(14, 12))
for i, row in df.iterrows():
    ax.barh(i, row["Duration"].days, left=row["Start"], height=0.8)

ax.set_yticks(range(len(df)))
ax.set_yticklabels(df["Task"])
ax.xaxis.set_major_locator(mdates.MonthLocator())
ax.xaxis.set_major_formatter(mdates.DateFormatter("%b %Y"))
plt.xticks(rotation=45)
plt.title("PhD Gantt Chart: June 2025 – September 2026")
plt.tight_layout()
plt.grid(axis='x')
plt.show()