package parser

import (
	"bytes"
	"context"
	"fmt"
	"os/exec"
	"strings"
	"time"
)

// pdfToTextTimeout bounds extraction so a malformed PDF cannot wedge an import.
const pdfToTextTimeout = 2 * time.Minute

// pdfToText renders a PDF as text with the original column positions
// preserved, which is what the statement parsers key off.
//
// This shells out to poppler's pdftotext rather than decoding the PDF in Go:
// the layout reconstruction it does is the whole reason the column-based
// parsers work, and no Go library matches it.
func pdfToText(path string) (string, error) {
	ctx, cancel := context.WithTimeout(context.Background(), pdfToTextTimeout)
	defer cancel()

	cmd := exec.CommandContext(ctx, "pdftotext", "-layout", path, "-")
	var stdout, stderr bytes.Buffer
	cmd.Stdout = &stdout
	cmd.Stderr = &stderr

	if err := cmd.Run(); err != nil {
		if ctx.Err() == context.DeadlineExceeded {
			return "", fmt.Errorf("pdftotext timed out after %s on %s", pdfToTextTimeout, path)
		}
		if errors := strings.TrimSpace(stderr.String()); errors != "" {
			return "", fmt.Errorf("pdftotext %s: %w: %s", path, err, errors)
		}
		if _, lookErr := exec.LookPath("pdftotext"); lookErr != nil {
			return "", fmt.Errorf("pdftotext is not installed (it ships with poppler-utils): %w", lookErr)
		}
		return "", fmt.Errorf("pdftotext %s: %w", path, err)
	}
	return stdout.String(), nil
}

// pages splits pdftotext output on form feeds.
func pages(text string) []string {
	return strings.Split(text, "\f")
}

// indent counts the leading spaces of a line.
func indent(line string) int {
	return len(line) - len(strings.TrimLeft(line, " "))
}