From 76d91ea8f65c6ad52efb24ac2c94b22c2908bc5c Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.name>
Date: Mon, 21 Mar 2022 13:51:51 +0000
Subject: Only generate full-size PDF if requested

This avoids the issue that large PDFs require a lot of RAM, so there
are chances of running out of memory. Plus it's a waste of space and
time.
---
 internal/pipeline/pipeline.go | 90 ++++++++++++++++++++++---------------------
 1 file changed, 46 insertions(+), 44 deletions(-)

(limited to 'internal/pipeline')

diff --git a/internal/pipeline/pipeline.go b/internal/pipeline/pipeline.go
index a09a414..d8beeb9 100644
--- a/internal/pipeline/pipeline.go
+++ b/internal/pipeline/pipeline.go
@@ -330,7 +330,7 @@ func Ocr(training string, tesscmd string) func(context.Context, chan string, cha
 	}
 }
 
-func Analyse(conn Downloader) func(context.Context, chan string, chan string, chan error, *log.Logger) {
+func Analyse(conn Downloader, mkfullpdf bool) func(context.Context, chan string, chan string, chan error, *log.Logger) {
 	return func(ctx context.Context, toanalyse chan string, up chan string, errc chan error, logger *log.Logger) {
 		confs := make(map[string][]*bookpipeline.Conf)
 		bestconfs := make(map[string]*bookpipeline.Conf)
@@ -455,12 +455,6 @@ func Analyse(conn Downloader) func(context.Context, chan string, chan string, ch
 			errc <- fmt.Errorf("Failed to set up PDF: %s", err)
 			return
 		}
-		fullsizepdf := new(bookpipeline.Fpdf)
-		err = fullsizepdf.Setup()
-		if err != nil {
-			errc <- fmt.Errorf("Failed to set up PDF: %s", err)
-			return
-		}
 		binhascontent, colourhascontent := false, false
 
 		select {
@@ -583,54 +577,62 @@ func Analyse(conn Downloader) func(context.Context, chan string, chan string, ch
 			up <- fn
 		}
 
-		for _, pg := range colourimgs {
-			select {
-			case <-ctx.Done():
-				errc <- ctx.Err()
+		if mkfullpdf {
+			fullsizepdf := new(bookpipeline.Fpdf)
+			err = fullsizepdf.Setup()
+			if err != nil {
+				errc <- fmt.Errorf("Failed to set up PDF: %s", err)
 				return
-			default:
 			}
+			for _, pg := range colourimgs {
+				select {
+				case <-ctx.Done():
+					errc <- ctx.Err()
+					return
+				default:
+				}
 
-			logger.Println("Downloading colour page to add to PDF", pg.img)
-			colourfn := pg.img
-			err = conn.Download(conn.WIPStorageId(), bookname+"/"+colourfn, filepath.Join(savedir, colourfn))
-			if err != nil {
-				colourfn = strings.Replace(pg.img, ".jpg", ".png", 1)
-				logger.Println("Download failed; trying", colourfn)
+				logger.Println("Downloading colour page to add to PDF", pg.img)
+				colourfn := pg.img
 				err = conn.Download(conn.WIPStorageId(), bookname+"/"+colourfn, filepath.Join(savedir, colourfn))
 				if err != nil {
-					logger.Println("Download failed; skipping page", pg.img)
+					colourfn = strings.Replace(pg.img, ".jpg", ".png", 1)
+					logger.Println("Download failed; trying", colourfn)
+					err = conn.Download(conn.WIPStorageId(), bookname+"/"+colourfn, filepath.Join(savedir, colourfn))
+					if err != nil {
+						logger.Println("Download failed; skipping page", pg.img)
+					}
 				}
-			}
-			if err == nil {
-				err = fullsizepdf.AddPage(filepath.Join(savedir, colourfn), filepath.Join(savedir, pg.hocr), false)
-				if err != nil {
-					errc <- fmt.Errorf("Failed to add page %s to PDF: %s", pg.img, err)
-					return
-				}
-				err = os.Remove(filepath.Join(savedir, colourfn))
-				if err != nil {
-					errc <- err
-					return
+				if err == nil {
+					err = fullsizepdf.AddPage(filepath.Join(savedir, colourfn), filepath.Join(savedir, pg.hocr), false)
+					if err != nil {
+						errc <- fmt.Errorf("Failed to add page %s to PDF: %s", pg.img, err)
+						return
+					}
+					err = os.Remove(filepath.Join(savedir, colourfn))
+					if err != nil {
+						errc <- err
+						return
+					}
 				}
 			}
-		}
-
-		select {
-		case <-ctx.Done():
-			errc <- ctx.Err()
-			return
-		default:
-		}
 
-		if colourhascontent {
-			fn = filepath.Join(savedir, bookname+".original.pdf")
-			err = fullsizepdf.Save(fn)
-			if err != nil {
-				errc <- fmt.Errorf("Failed to save full size pdf: %s", err)
+			select {
+			case <-ctx.Done():
+				errc <- ctx.Err()
 				return
+			default:
+			}
+
+			if colourhascontent {
+				fn = filepath.Join(savedir, bookname+".original.pdf")
+				err = fullsizepdf.Save(fn)
+				if err != nil {
+					errc <- fmt.Errorf("Failed to save full size pdf: %s", err)
+					return
+				}
+				up <- fn
 			}
-			up <- fn
 		}
 
 		select {
-- 
cgit v1.2.1-24-ge1ad