html2pdf/converter/converter.go

package converter

import (
	"context"
	"errors"
	"fmt"
	"os"
	"path/filepath"
	"runtime"
	"strings"
	"time"

	"gitea.tecamino.com/paadi/html2pdf/models"
	"github.com/chromedp/cdproto/page"
	"github.com/chromedp/chromedp"
)

// html to pdf converter structure for
type Converter struct {
	chromePath  string
	ctx         context.Context
	cancel      context.CancelFunc
	allocCancel context.CancelFunc
	progress    func(progress int)
}

// NewConverter starts a new converter instance with a chrome headless shell executable
func NewConverter(chromePath string) (*Converter, error) {
	var err error

	c := &Converter{chromePath: chromePath}
	chromePath, err = c.getChromePath()
	if err != nil {
		return nil, err
	}

	opts := append(chromedp.DefaultExecAllocatorOptions[:],
		chromedp.ExecPath(chromePath),
		chromedp.NoSandbox,
		chromedp.Headless,
		chromedp.DisableGPU,
	)

	opts = append(opts, platformOptions())

	var allocCtx context.Context
	allocCtx, c.cancel = chromedp.NewExecAllocator(context.Background(), opts...)
	c.ctx, c.allocCancel = chromedp.NewContext(allocCtx)
	return c, nil
}

func (c *Converter) SetProgressCallback(cb func(progress int)) {
	c.progress = cb
}

// Convert converts all given input files
func (c *Converter) Convert(files ...models.File) error {
	for i, f := range files {
		if c.progress != nil {
			c.progress(i + 1)
		}

		if f.Input == "" || filepath.Ext(f.Input) != ".html" {
			return fmt.Errorf("no .html input file path provided: %s", f.Input)
		} else if f.Output == "" || filepath.Ext(f.Output) != ".pdf" {
			return fmt.Errorf("no .pdf output file path provided: %s", f.Output)
		}

		var htmlURL strings.Builder
		htmlURL.WriteString("file://")

		switch runtime.GOOS {
		case "windows":
			htmlURL.WriteString("/")
		}

		// Convert to absolute path
		absPath, err := filepath.Abs(f.Input)
		if err != nil {
			return err
		}

		htmlURL.WriteString(filepath.ToSlash(absPath))

		ctx, cancel := context.WithTimeout(c.ctx, 60*time.Second)
		defer cancel()

		var pdfData []byte
		err = chromedp.Run(ctx,
			chromedp.Navigate(htmlURL.String()),
			chromedp.WaitReady("body", chromedp.ByQuery),
			chromedp.ActionFunc(func(ctx context.Context) error {
				buf, _, err := page.PrintToPDF().
					WithPrintBackground(true).
					WithPaperWidth(8.27).
					WithPaperHeight(11.69).
					Do(ctx)
				if err != nil {
					return err
				}
				pdfData = buf
				return nil
			}),
		)

		if err != nil {
			c.cancel()
			return err
		}

		// Save PDF to file
		if err := os.WriteFile(f.Output, pdfData, 0644); err != nil {
			c.cancel()
			return err
		}
	}
	c.cancel()
	return nil
}

func (c *Converter) Close() {
	if c.cancel != nil {
		c.cancel()
	}
}

// getChromePath checks for system Chrome, else falls back to bundled headless shell
func (c *Converter) getChromePath() (string, error) {

	chromeExec := "chrome-headless-shell"
	if runtime.GOOS == "windows" {
		chromeExec += ".exe"
	}

	path := filepath.Join(c.chromePath, chromeExec)
	if _, err := os.Stat(path); err == nil {
		return path, nil
	}

	// Candidate paths for system Chrome
	candidates := []string{}

	// Fallback:
	switch runtime.GOOS {
	case "windows":
		candidates = []string{
			`C:\Program Files\Google\Chrome\Application\chrome.exe`,
			`C:\Program Files (x86)\Google\Chrome\Application\chrome.exe`,
			`C:\Program Files\Chromium\Application\chrome.exe`,
		}
	case "darwin":
		candidates = []string{
			"/Applications/Google Chrome.app/Contents/MacOS/Google Chrome",
			"/Applications/Chromium.app/Contents/MacOS/Chromium",
		}
	default: // Linux
		candidates = []string{
			"/usr/bin/google-chrome",
			"/usr/bin/chromium-browser",
			"/usr/bin/chromium",
		}
	}

	// Check system paths first
	for _, path := range candidates {
		if _, err := os.Stat(path); err == nil {
			return path, nil
		}
	}

	return "", errors.New("chrome path not found")
}