From 34993e36eeeec02e55132d42fed608394e7405cb Mon Sep 17 00:00:00 2001
From: celogeek <65178+celogeek@users.noreply.github.com>
Date: Fri, 30 Dec 2022 14:10:19 +0100
Subject: [PATCH] convert pdf

---
 go.mod                            |  1 +
 go.sum                            |  2 ++
 internal/epub/image_processing.go | 38 ++++++++++++++++++++++++++++++-
 3 files changed, 40 insertions(+), 1 deletion(-)

diff --git a/go.mod b/go.mod
index bf1a2f6..8322649 100644
--- a/go.mod
+++ b/go.mod
@@ -5,6 +5,7 @@ go 1.19
 require (
 	github.com/gofrs/uuid v3.1.0+incompatible
 	github.com/nwaples/rardecode v1.1.3
+	github.com/raff/pdfreader v0.0.0-20220308062436-033e8ac577f0
 	github.com/schollz/progressbar/v3 v3.12.2
 	golang.org/x/image v0.2.0
 )
diff --git a/go.sum b/go.sum
index 2fc520b..4e605f3 100644
--- a/go.sum
+++ b/go.sum
@@ -13,6 +13,8 @@ github.com/nwaples/rardecode v1.1.3 h1:cWCaZwfM5H7nAD6PyEdcVnczzV8i/JtotnyW/dD9l
 github.com/nwaples/rardecode v1.1.3/go.mod h1:5DzqNKiOdpKKBH87u8VlvAnPZMXcGRhxWkRpHbbfGS0=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
+github.com/raff/pdfreader v0.0.0-20220308062436-033e8ac577f0 h1:fuFvfwIc+cpySYurvDNTs5LIHXP9Cj3reVRplj9Whv4=
+github.com/raff/pdfreader v0.0.0-20220308062436-033e8ac577f0/go.mod h1:Ql3QqeGiYGlPOtYz+F/L7J27spqDcdH9LhDHOrrdsD4=
 github.com/rivo/uniseg v0.2.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
 github.com/rivo/uniseg v0.4.3 h1:utMvzDsuh3suAEnhH0RdHmoPbU648o6CvXxTx4SBMOw=
 github.com/rivo/uniseg v0.4.3/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
diff --git a/internal/epub/image_processing.go b/internal/epub/image_processing.go
index 5f83246..12948e7 100644
--- a/internal/epub/image_processing.go
+++ b/internal/epub/image_processing.go
@@ -15,6 +15,9 @@ import (
 	"sync"
 
 	"github.com/nwaples/rardecode"
+	pdfimage "github.com/raff/pdfreader/image"
+	"github.com/raff/pdfreader/pdfread"
+	"golang.org/x/image/tiff"
 )
 
 type Image struct {
@@ -51,7 +54,7 @@ func LoadImages(path string, options *ImageOptions) ([]*Image, error) {
 		case ".cbr", "rar":
 			imageCount, imageInput, err = loadCbr(path)
 		case ".pdf":
-			err = fmt.Errorf("not implemented")
+			imageCount, imageInput, err = loadPdf(path)
 		default:
 			err = fmt.Errorf("unknown file format (%s): support .cbz, .cbr, .pdf", ext)
 		}
@@ -276,3 +279,36 @@ func loadCbr(input string) (int, chan *imageTask, error) {
 
 	return len(names), output, nil
 }
+
+func loadPdf(input string) (int, chan *imageTask, error) {
+	pdf := pdfread.Load(input)
+	if pdf == nil {
+		return 0, nil, fmt.Errorf("can't read pdf")
+	}
+
+	nbPages := len(pdf.Pages())
+	output := make(chan *imageTask)
+	go func() {
+		defer close(output)
+		defer pdf.Close()
+		for i := 0; i < nbPages; i++ {
+			img, err := pdfimage.Extract(pdf, i+1)
+			if err != nil {
+				panic(err)
+			}
+
+			b := bytes.NewBuffer([]byte{})
+			err = tiff.Encode(b, img, nil)
+			if err != nil {
+				panic(err)
+			}
+
+			output <- &imageTask{
+				Id:     i,
+				Reader: io.NopCloser(b),
+			}
+		}
+	}()
+
+	return nbPages, output, nil
+}