src/report.py

"""
Report generation module for RehabWatch.
Creates PDF reports and CSV exports.
"""

from fpdf import FPDF
from datetime import datetime
from typing import Dict, Any, Optional
import io


class RehabWatchPDF(FPDF):
    """Custom PDF class for RehabWatch reports."""

    def header(self):
        """Add header to each page."""
        self.set_font('Helvetica', 'B', 16)
        self.set_text_color(46, 125, 50)  # Green
        self.cell(0, 10, 'RehabWatch', 0, 0, 'L')
        self.set_font('Helvetica', '', 10)
        self.set_text_color(100, 100, 100)
        self.cell(0, 10, 'Mining Rehabilitation Assessment', 0, 1, 'R')
        self.line(10, 25, 200, 25)
        self.ln(10)

    def footer(self):
        """Add footer to each page."""
        self.set_y(-15)
        self.set_font('Helvetica', 'I', 8)
        self.set_text_color(128, 128, 128)
        self.cell(0, 10, f'Page {self.page_no()}/{{nb}}', 0, 0, 'C')

    def chapter_title(self, title: str):
        """Add a section title."""
        self.set_font('Helvetica', 'B', 14)
        self.set_text_color(33, 33, 33)
        self.cell(0, 10, title, 0, 1, 'L')
        self.ln(2)

    def chapter_body(self, body: str):
        """Add body text."""
        self.set_font('Helvetica', '', 11)
        self.set_text_color(66, 66, 66)
        self.multi_cell(0, 6, body)
        self.ln(4)


def generate_pdf_report(
    tenement_id: str,
    stats: Dict[str, float],
    rehab_score: int,
    interpretation: str,
    date_before: str,
    date_after: str,
    mine_name: Optional[str] = None
) -> bytes:
    """
    Generate a PDF report of the rehabilitation assessment.

    Args:
        tenement_id: Mining tenement identifier
        stats: Statistics dictionary from analysis
        rehab_score: Rehabilitation score (0-100)
        interpretation: Plain-language interpretation
        date_before: Analysis start date
        date_after: Analysis end date
        mine_name: Optional name of the mine

    Returns:
        PDF as bytes for download
    """
    pdf = RehabWatchPDF()
    pdf.alias_nb_pages()
    pdf.add_page()

    # Title
    pdf.set_font('Helvetica', 'B', 24)
    pdf.set_text_color(33, 33, 33)
    pdf.cell(0, 15, 'Rehabilitation Assessment Report', 0, 1, 'C')
    pdf.ln(5)

    # Site Information
    pdf.set_font('Helvetica', '', 12)
    pdf.set_text_color(66, 66, 66)

    if mine_name:
        pdf.cell(0, 8, f'Site: {mine_name}', 0, 1, 'C')
    pdf.cell(0, 8, f'Tenement ID: {tenement_id}', 0, 1, 'C')
    pdf.cell(0, 8, f'Analysis Period: {date_before} to {date_after}', 0, 1, 'C')
    pdf.cell(0, 8, f'Report Generated: {datetime.now().strftime("%Y-%m-%d %H:%M")}', 0, 1, 'C')
    pdf.ln(10)

    # Rehabilitation Score Box
    pdf.set_fill_color(240, 240, 240)
    pdf.rect(60, pdf.get_y(), 90, 35, 'F')

    pdf.set_font('Helvetica', 'B', 12)
    pdf.set_text_color(66, 66, 66)
    pdf.cell(0, 8, 'Rehabilitation Score', 0, 1, 'C')

    # Score color based on value
    if rehab_score >= 60:
        pdf.set_text_color(46, 125, 50)  # Green
    elif rehab_score >= 40:
        pdf.set_text_color(255, 152, 0)  # Orange
    else:
        pdf.set_text_color(183, 28, 28)  # Red

    pdf.set_font('Helvetica', 'B', 36)
    pdf.cell(0, 15, f'{rehab_score}/100', 0, 1, 'C')
    pdf.ln(15)

    # Interpretation
    pdf.chapter_title('Summary')
    pdf.chapter_body(interpretation)

    # Statistics Table
    pdf.chapter_title('Detailed Statistics')

    # Table header
    pdf.set_font('Helvetica', 'B', 10)
    pdf.set_fill_color(46, 125, 50)
    pdf.set_text_color(255, 255, 255)
    pdf.cell(90, 8, 'Metric', 1, 0, 'C', True)
    pdf.cell(50, 8, 'Value', 1, 0, 'C', True)
    pdf.cell(50, 8, 'Unit', 1, 1, 'C', True)

    # Table data
    pdf.set_font('Helvetica', '', 10)
    pdf.set_text_color(66, 66, 66)

    table_data = [
        ('NDVI Before (mean)', f"{stats['ndvi_before_mean']:.4f}", 'index'),
        ('NDVI After (mean)', f"{stats['ndvi_after_mean']:.4f}", 'index'),
        ('NDVI Change (mean)', f"{stats['ndvi_change_mean']:.4f}", 'index'),
        ('Relative Change', f"{stats['percent_change']:.2f}", '%'),
        ('SAVI (mean)', f"{stats.get('savi_after_mean', 0):.4f}", 'index'),
        ('EVI (mean)', f"{stats.get('evi_after_mean', 0):.4f}", 'index'),
        ('NDMI (moisture)', f"{stats.get('ndmi_after_mean', 0):.4f}", 'index'),
        ('BSI (bare soil)', f"{stats.get('bsi_after_mean', 0):.4f}", 'index'),
        ('Water Presence', f"{stats.get('percent_water', 0):.2f}", '%'),
        ('Bare Soil Extent', f"{stats.get('percent_bare_soil', 0):.2f}", '%'),
        ('Moisture Stressed', f"{stats.get('percent_moisture_stressed', 0):.2f}", '%'),
        ('Area Improved', f"{stats['area_improved_ha']:.2f}", 'hectares'),
        ('Area Stable', f"{stats['area_stable_ha']:.2f}", 'hectares'),
        ('Area Degraded', f"{stats['area_degraded_ha']:.2f}", 'hectares'),
        ('Total Area', f"{stats['total_area_ha']:.2f}", 'hectares'),
        ('Percentage Improved', f"{stats['percent_improved']:.2f}", '%'),
        ('Percentage Degraded', f"{stats['percent_degraded']:.2f}", '%'),
    ]

    fill = False
    for row in table_data:
        if fill:
            pdf.set_fill_color(245, 245, 245)
        else:
            pdf.set_fill_color(255, 255, 255)
        pdf.cell(90, 7, row[0], 1, 0, 'L', fill)
        pdf.cell(50, 7, row[1], 1, 0, 'C', fill)
        pdf.cell(50, 7, row[2], 1, 1, 'C', fill)
        fill = not fill

    pdf.ln(10)

    # Methodology
    pdf.chapter_title('Methodology')
    methodology_text = """This assessment uses multiple vegetation and soil indices derived from Sentinel-2 satellite imagery, Copernicus DEM, and IO-LULC land cover data.

Vegetation Indices:
- NDVI (Normalized Difference Vegetation Index): Overall vegetation health
- SAVI (Soil Adjusted Vegetation Index): Better for sparse vegetation
- EVI (Enhanced Vegetation Index): Better for dense vegetation

Soil & Water Indices:
- BSI (Bare Soil Index): Identifies exposed soil areas
- NDWI (Normalized Difference Water Index): Water body detection
- NDMI (Normalized Difference Moisture Index): Vegetation moisture content

Terrain Analysis:
- Slope and aspect from Copernicus DEM GLO-30 (30m resolution)
- Erosion risk combining slope steepness and bare soil exposure

Land Cover:
- IO-LULC annual land cover classification (2017-2023)

The Rehabilitation Score combines vegetation health, improvement trends, soil stability, and moisture status compared to reference conditions."""

    pdf.chapter_body(methodology_text)

    # Data Sources and Disclaimers
    pdf.add_page()
    pdf.chapter_title('Data Sources')
    sources_text = """- Satellite Imagery: Copernicus Sentinel-2 L2A (Surface Reflectance)
- Spatial Resolution: 10 meters
- Temporal Resolution: ~5 days revisit
- Cloud Masking: Applied using Scene Classification Layer (SCL)
- Compositing Method: Median composite over 30-day windows
- Digital Elevation: Copernicus DEM GLO-30 (30m resolution)
- Land Cover: IO-LULC Annual v02 (10m resolution, 2017-2023)
- Data Access: Microsoft Planetary Computer (free, open access)"""

    pdf.chapter_body(sources_text)

    pdf.chapter_title('Disclaimer')
    disclaimer_text = """This report is generated automatically using satellite remote sensing data and should be used for preliminary assessment purposes only. Results may be affected by:

- Cloud cover and atmospheric conditions
- Seasonal vegetation variations
- Sensor calibration differences
- Topographic effects

For regulatory compliance or detailed rehabilitation assessment, ground-based verification is recommended. This analysis does not constitute professional advice and should be interpreted by qualified personnel.

RehabWatch is a demonstration tool and the developers assume no liability for decisions made based on this analysis."""

    pdf.chapter_body(disclaimer_text)

    # Output PDF bytes
    return bytes(pdf.output())


def stats_to_csv(
    stats: Dict[str, float],
    tenement_id: str,
    rehab_score: int,
    date_before: str,
    date_after: str,
    mine_name: Optional[str] = None
) -> str:
    """
    Convert statistics to CSV format.

    Args:
        stats: Statistics dictionary
        tenement_id: Mining tenement identifier
        rehab_score: Rehabilitation score
        date_before: Analysis start date
        date_after: Analysis end date
        mine_name: Optional mine name

    Returns:
        CSV string for download
    """
    lines = []

    # Header
    lines.append("RehabWatch Rehabilitation Assessment Export")
    lines.append(f"Generated,{datetime.now().strftime('%Y-%m-%d %H:%M')}")
    if mine_name:
        lines.append(f"Site Name,{mine_name}")
    lines.append(f"Tenement ID,{tenement_id}")
    lines.append(f"Analysis Start Date,{date_before}")
    lines.append(f"Analysis End Date,{date_after}")
    lines.append("")

    # Statistics
    lines.append("Metric,Value,Unit")
    lines.append(f"Rehabilitation Score,{rehab_score},/100")
    lines.append(f"NDVI Before (mean),{stats['ndvi_before_mean']:.4f},index")
    lines.append(f"NDVI After (mean),{stats['ndvi_after_mean']:.4f},index")
    lines.append(f"NDVI Change (mean),{stats['ndvi_change_mean']:.4f},index")
    lines.append(f"NDVI Change (std dev),{stats['ndvi_change_std']:.4f},index")
    lines.append(f"Relative Change,{stats['percent_change']:.2f},%")
    lines.append(f"SAVI Before,{stats.get('savi_before_mean', 0):.4f},index")
    lines.append(f"SAVI After,{stats.get('savi_after_mean', 0):.4f},index")
    lines.append(f"EVI Before,{stats.get('evi_before_mean', 0):.4f},index")
    lines.append(f"EVI After,{stats.get('evi_after_mean', 0):.4f},index")
    lines.append(f"NDWI After,{stats.get('ndwi_after_mean', 0):.4f},index")
    lines.append(f"NDMI After,{stats.get('ndmi_after_mean', 0):.4f},index")
    lines.append(f"BSI After,{stats.get('bsi_after_mean', 0):.4f},index")
    lines.append(f"Water Presence,{stats.get('percent_water', 0):.2f},%")
    lines.append(f"Bare Soil Extent,{stats.get('percent_bare_soil', 0):.2f},%")
    lines.append(f"Moisture Stressed,{stats.get('percent_moisture_stressed', 0):.2f},%")
    lines.append(f"Sparse Vegetation,{stats.get('percent_sparse_veg', 0):.2f},%")
    lines.append(f"Low Vegetation,{stats.get('percent_low_veg', 0):.2f},%")
    lines.append(f"Moderate Vegetation,{stats.get('percent_moderate_veg', 0):.2f},%")
    lines.append(f"Dense Vegetation,{stats.get('percent_dense_veg', 0):.2f},%")
    lines.append(f"Area Improved,{stats['area_improved_ha']:.2f},hectares")
    lines.append(f"Area Stable,{stats['area_stable_ha']:.2f},hectares")
    lines.append(f"Area Degraded,{stats['area_degraded_ha']:.2f},hectares")
    lines.append(f"Total Area,{stats['total_area_ha']:.2f},hectares")
    lines.append(f"Percentage Improved,{stats['percent_improved']:.2f},%")
    lines.append(f"Percentage Stable,{stats['percent_stable']:.2f},%")
    lines.append(f"Percentage Degraded,{stats['percent_degraded']:.2f},%")

    return "\n".join(lines)


def generate_summary_text(
    tenement_id: str,
    stats: Dict[str, float],
    rehab_score: int,
    interpretation: str,
    date_before: str,
    date_after: str
) -> str:
    """
    Generate a plain text summary of the assessment.

    Args:
        tenement_id: Mining tenement identifier
        stats: Statistics dictionary
        rehab_score: Rehabilitation score
        interpretation: Interpretation text
        date_before: Analysis start date
        date_after: Analysis end date

    Returns:
        Formatted text summary
    """
    summary = f"""
================================================================================
                    REHABWATCH REHABILITATION ASSESSMENT
================================================================================

Tenement: {tenement_id}
Analysis Period: {date_before} to {date_after}
Report Generated: {datetime.now().strftime('%Y-%m-%d %H:%M')}

--------------------------------------------------------------------------------
                           REHABILITATION SCORE
--------------------------------------------------------------------------------

                              {rehab_score} / 100

--------------------------------------------------------------------------------
                              KEY FINDINGS
--------------------------------------------------------------------------------

{interpretation}

--------------------------------------------------------------------------------
                           DETAILED STATISTICS
--------------------------------------------------------------------------------

Vegetation Index (NDVI):
  - Before:  {stats['ndvi_before_mean']:.4f}
  - After:   {stats['ndvi_after_mean']:.4f}
  - Change:  {stats['ndvi_change_mean']:.4f} ({stats['percent_change']:.2f}%)

Area Analysis:
  - Total Area:      {stats['total_area_ha']:.2f} ha
  - Area Improved:   {stats['area_improved_ha']:.2f} ha ({stats['percent_improved']:.2f}%)
  - Area Stable:     {stats['area_stable_ha']:.2f} ha ({stats['percent_stable']:.2f}%)
  - Area Degraded:   {stats['area_degraded_ha']:.2f} ha ({stats['percent_degraded']:.2f}%)

================================================================================
                              DATA SOURCES
================================================================================

Satellite: Copernicus Sentinel-2 L2A
Resolution: 10 meters
Analysis: NDVI vegetation index

================================================================================
"""
    return summary