package parser

import (
	"fmt"
	"os"
	"strings"

	"github.com/ledongthuc/pdf"
)

// PDFParser extracts plain text from PDF files using ledongthuc/pdf (open source).
type PDFParser struct{}

func (p *PDFParser) SupportedExtensions() []string {
	return []string{".pdf"}
}

func (p *PDFParser) ParseToMarkdown(path string) (string, error) {
	f, r, err := pdf.Open(path)
	if err != nil {
		return "", fmt.Errorf("open pdf: %w", err)
	}
	defer f.Close()

	var sb strings.Builder
	totalPages := r.NumPage()

	for i := 1; i <= totalPages; i++ {
		page := r.Page(i)
		if page.V.IsNull() {
			continue
		}
		text, err := page.GetPlainText(nil)
		if err != nil {
			continue
		}
		content := strings.TrimSpace(text)
		if content != "" {
			sb.WriteString(fmt.Sprintf("## Page %d\n\n%s\n\n", i, content))
		}
	}

	if sb.Len() == 0 {
		// Fallback: try reading entire content at once
		content, err := readPDFPlainText(path)
		if err == nil && content != "" {
			return content, nil
		}
		return "", fmt.Errorf("no text content extracted from PDF")
	}

	return sb.String(), nil
}

func readPDFPlainText(path string) (string, error) {
	f, err := os.Open(path)
	if err != nil {
		return "", err
	}
	defer f.Close()
	stat, _ := f.Stat()
	reader, err := pdf.NewReader(f, stat.Size())
	if err != nil {
		return "", err
	}
	var sb strings.Builder
	for i := 1; i <= reader.NumPage(); i++ {
		page := reader.Page(i)
		if page.V.IsNull() {
			continue
		}
		text, _ := page.GetPlainText(nil)
		sb.WriteString(text)
	}
	return sb.String(), nil
}