export_report.py

"""
export_report.py — BIRAS Report Export System
==============================================
Generates publication-ready reports from analysis_results in:
  - DOCX (python-docx)  ← primary
  - TXT                 ← always available fallback
  - PDF via DOCX→PDF    ← optional (requires LibreOffice or reportlab)

Usage:
    from export_report import export_report
    pdf_bytes = export_report(analysis_results, fmt='docx')
"""

from __future__ import annotations

import io
import os
import datetime
import logging

logger = logging.getLogger('biras.export')


# ─────────────────────────────────────────────────────────────────────────────
# HELPERS
# ─────────────────────────────────────────────────────────────────────────────

def _get(ar: dict, *keys, default=''):
    """Nested safe-get from analysis_results."""
    for key_path in keys:
        parts = key_path.split('.')
        val = ar
        for p in parts:
            if isinstance(val, dict):
                val = val.get(p, '')
            else:
                val = ''
                break
        if val:
            return val
    return default


def _flatten(ar: dict) -> dict:
    """
    Build a flat context dict from the nested analysis_results structure,
    with backward-compat fallback to flat keys.
    """
    query = (ar.get('meta') or {}).get('query') or ar.get('query', 'Research Topic')
    n     = ((ar.get('meta') or {}).get('article_count') or
             ar.get('total_articles', ar.get('total_results', 0)))

    writing  = ar.get('writing', ar)      # nested or flat fallback
    research = ar.get('research', ar)
    ai_out   = ar.get('ai_outputs', ar)

    rqs = research.get('questions', ar.get('research_questions', []))
    rqs_text = '\n'.join(f"  {i+1}. {q}" for i, q in enumerate(rqs)) if rqs else '  (Not available)'

    return {
        'query':           query,
        'n':               n,
        'timestamp':       datetime.datetime.now().strftime('%Y-%m-%d %H:%M UTC'),
        'llm_source':      (ar.get('meta') or {}).get('llm_source', ar.get('llm_source', 'rule-based')),

        # Writing sections (nested → flat fallback)
        'abstract':          writing.get('abstract',          ar.get('abstract', '')),
        'global_summary':    writing.get('global_summary',    ar.get('global_summary', '')),
        'conclusion':        writing.get('conclusion',         ar.get('conclusion', '')),
        'future_work':       writing.get('future_work',        ar.get('future_work', '')),
        'validation':        writing.get('validation_summary', ar.get('validation_summary', '')),
        'system_position':   writing.get('system_positioning', ar.get('system_positioning', '')),
        'limitations':       writing.get('limitations',        ar.get('limitations', '')),

        # AI outputs
        'insights_text':     ai_out.get('insights_text',        ar.get('insights_text', '')),
        'recommendations':   ai_out.get('recommendations_text', ar.get('recommendations_text', '')),
        'narrative':         ai_out.get('narrative_text',        ar.get('narrative_text', '')),

        # Research
        'rqs_text':          rqs_text,
        'related_work':      research.get('related_work', ar.get('related_work', '')),

        # Evaluation
        'evaluation':        ar.get('evaluation', {}),
        'metrics':           ar.get('metrics', {}),
    }


# ─────────────────────────────────────────────────────────────────────────────
# TXT EXPORT (always-available fallback)
# ─────────────────────────────────────────────────────────────────────────────

def _build_txt(ctx: dict) -> bytes:
    sep = '=' * 70
    lines = [
        sep,
        f"  BIRAS — AI Research Report",
        f"  Query : {ctx['query']}",
        f"  Papers: {ctx['n']}",
        f"  Date  : {ctx['timestamp']}",
        f"  Source: {ctx['llm_source']}",
        sep, '',

        '[ ABSTRACT ]',
        ctx['abstract'] or ctx['global_summary'] or '(not available)', '',

        '[ KEY INSIGHTS ]',
        ctx['insights_text'] or '(not available)', '',

        '[ RESEARCH QUESTIONS ]',
        ctx['rqs_text'], '',

        '[ RELATED WORK ]',
        ctx['related_work'] or '(not available)', '',

        '[ CONCLUSION ]',
        ctx['conclusion'] or '(not available)', '',

        '[ FUTURE WORK ]',
        ctx['future_work'] or '(not available)', '',
    ]

    if ctx.get('limitations'):
        lines += ['[ LIMITATIONS ]', ctx['limitations'], '']

    ev = ctx.get('evaluation', {})
    if ev:
        lines += ['[ EVALUATION ]']
        for k, v in ev.items():
            lines.append(f"  {k}: {v}")
        lines.append('')

    lines += [sep, 'Generated by BIRAS — AI Research Operating System', sep]
    return '\n'.join(lines).encode('utf-8')


# ─────────────────────────────────────────────────────────────────────────────
# DOCX EXPORT
# ─────────────────────────────────────────────────────────────────────────────

def _build_docx(ctx: dict) -> bytes:
    try:
        from docx import Document
        from docx.shared import Pt, RGBColor, Inches
        from docx.enum.text import WD_ALIGN_PARAGRAPH
        from docx.oxml.ns import qn
    except ImportError as e:
        logger.warning("python-docx not available: %s — falling back to TXT", e)
        raise

    doc = Document()

    # ── Page margins ──────────────────────────────────────────────────────────
    for section in doc.sections:
        section.top_margin    = Inches(1)
        section.bottom_margin = Inches(1)
        section.left_margin   = Inches(1.25)
        section.right_margin  = Inches(1.25)

    def h(text: str, level: int = 1):
        doc.add_heading(text, level=level)

    def body(text: str, italic: bool = False):
        if not text:
            return
        p = doc.add_paragraph(text)
        run = p.runs[0] if p.runs else p.add_run(text)
        run.font.size = Pt(11)
        if italic:
            run.italic = True

    def muted(text: str):
        p = doc.add_paragraph()
        run = p.add_run(text)
        run.font.size = Pt(9)
        run.font.color.rgb = RGBColor(0x88, 0x88, 0x88)
        p.alignment = WD_ALIGN_PARAGRAPH.CENTER

    # ── Cover ─────────────────────────────────────────────────────────────────
    title = doc.add_heading(
        f"AI-Assisted Systematic Literature Review\non: {ctx['query'].title()}", level=0
    )
    title.alignment = WD_ALIGN_PARAGRAPH.CENTER
    muted(f"Generated by BIRAS  |  {ctx['timestamp']}  |  {ctx['n']} articles")
    doc.add_paragraph()

    # ── Sections ──────────────────────────────────────────────────────────────
    _SECTIONS = [
        ('Abstract',                    'abstract'),
        ('1. Global Research Summary',  'global_summary'),
        ('2. Key Insights',             'insights_text'),
        ('3. Research Questions',       'rqs_text'),
        ('4. Related Work',             'related_work'),
        ('5. Conclusion',               'conclusion'),
        ('6. Future Work',              'future_work'),
    ]

    for heading_txt, key in _SECTIONS:
        content = ctx.get(key, '').strip()
        if not content:
            continue
        h(heading_txt)
        if key == 'rqs_text':
            for line in content.split('\n'):
                if line.strip():
                    doc.add_paragraph(line.strip(), style='List Number')
        else:
            body(content)
        doc.add_paragraph()

    if ctx.get('limitations'):
        h('7. Limitations')
        body(ctx['limitations'])
        doc.add_paragraph()

    ev = ctx.get('evaluation', {})
    if ev:
        h('8. Evaluation Metrics')
        for k, v in ev.items():
            doc.add_paragraph(f"{k.replace('_', ' ').title()}: {v}", style='List Bullet')
        doc.add_paragraph()

    m = ctx.get('metrics', {})
    if m:
        h('Appendix — Performance Metrics')
        for k, v in m.items():
            doc.add_paragraph(f"{k}: {v}", style='List Bullet')

    # Footer note
    body(
        f"\nThis report was auto-generated by BIRAS (Bibliometric Intelligent Research "
        f"Assistant System). AI source: {ctx['llm_source']}.", italic=True
    )

    buf = io.BytesIO()
    doc.save(buf)
    buf.seek(0)
    return buf.read()


# ─────────────────────────────────────────────────────────────────────────────
# PDF EXPORT (via reportlab — graceful fallback)
# ─────────────────────────────────────────────────────────────────────────────

def _build_pdf(ctx: dict) -> bytes:
    try:
        from reportlab.lib.pagesizes import A4
        from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
        from reportlab.lib.units import cm
        from reportlab.lib import colors
        from reportlab.platypus import (
            SimpleDocTemplate, Paragraph, Spacer, HRFlowable
        )

        buf = io.BytesIO()
        doc = SimpleDocTemplate(
            buf, pagesize=A4,
            leftMargin=3*cm, rightMargin=3*cm,
            topMargin=2.5*cm, bottomMargin=2.5*cm,
        )
        styles = getSampleStyleSheet()
        story  = []

        title_style = ParagraphStyle(
            'BIRASTitle', parent=styles['Title'],
            fontSize=16, leading=22, spaceAfter=6,
        )
        h1_style = ParagraphStyle(
            'BIRASH1', parent=styles['Heading1'],
            fontSize=13, textColor=colors.HexColor('#1e3a5f'), spaceAfter=4,
        )
        body_style = ParagraphStyle(
            'BIRASBody', parent=styles['Normal'],
            fontSize=10.5, leading=15, spaceAfter=8,
        )
        meta_style = ParagraphStyle(
            'BIRASmeta', parent=styles['Normal'],
            fontSize=9, textColor=colors.grey, alignment=1,
        )

        story.append(Paragraph(
            f"AI-Assisted Systematic Literature Review<br/><i>{ctx['query'].title()}</i>",
            title_style
        ))
        story.append(Paragraph(
            f"Generated by BIRAS  |  {ctx['timestamp']}  |  {ctx['n']} articles  |  {ctx['llm_source']}",
            meta_style
        ))
        story.append(HRFlowable(width='100%', thickness=1, color=colors.lightgrey))
        story.append(Spacer(1, 0.4*cm))

        _SECTIONS = [
            ('Abstract',               'abstract'),
            ('Global Research Summary','global_summary'),
            ('Key Insights',           'insights_text'),
            ('Research Questions',     'rqs_text'),
            ('Related Work',           'related_work'),
            ('Conclusion',             'conclusion'),
            ('Future Work',            'future_work'),
        ]

        for heading_txt, key in _SECTIONS:
            content = ctx.get(key, '').strip()
            if not content:
                continue
            story.append(Paragraph(heading_txt, h1_style))
            # Escape & for reportlab
            safe = content.replace('&', '&amp;').replace('<', '&lt;').replace('>', '&gt;')
            for para in safe.split('\n'):
                if para.strip():
                    story.append(Paragraph(para, body_style))
            story.append(Spacer(1, 0.3*cm))

        doc.build(story)
        buf.seek(0)
        return buf.read()

    except ImportError:
        logger.warning("reportlab not installed — falling back to DOCX bytes")
        try:
            return _build_docx(ctx)
        except Exception:
            return _build_txt(ctx)


# ─────────────────────────────────────────────────────────────────────────────
# PUBLIC API
# ─────────────────────────────────────────────────────────────────────────────

def export_report(analysis_results: dict, fmt: str = 'docx') -> tuple[bytes, str, str]:
    """
    Export analysis_results to a report file.

    Args:
        analysis_results: The full pipeline output dict.
        fmt: 'docx' | 'pdf' | 'txt'

    Returns:
        (file_bytes, mimetype, filename)
    """
    fmt = fmt.lower().strip()
    ctx = _flatten(analysis_results)
    query_slug = ctx['query'][:30].replace(' ', '_').replace('/', '-')
    date_slug  = datetime.datetime.now().strftime('%Y%m%d')

    try:
        if fmt == 'pdf':
            data     = _build_pdf(ctx)
            mime     = 'application/pdf'
            filename = f"BIRAS_Report_{query_slug}_{date_slug}.pdf"
        elif fmt == 'docx':
            data     = _build_docx(ctx)
            mime     = 'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
            filename = f"BIRAS_Report_{query_slug}_{date_slug}.docx"
        else:  # txt
            data     = _build_txt(ctx)
            mime     = 'text/plain; charset=utf-8'
            filename = f"BIRAS_Report_{query_slug}_{date_slug}.txt"

        logger.info("Report exported: fmt=%s size=%d bytes", fmt, len(data))
        return data, mime, filename

    except Exception as exc:
        logger.error("Export failed (%s): %s — falling back to TXT", fmt, exc)
        data     = _build_txt(ctx)
        filename = f"BIRAS_Report_{query_slug}_{date_slug}.txt"
        return data, 'text/plain; charset=utf-8', filename