From ae8cea1bc6362a86fd835c0334d3aace780a8dd7 Mon Sep 17 00:00:00 2001
From: Sven Fischer <sven@leiderfischer.de>
Date: Mon, 3 Feb 2020 18:18:21 +0100
Subject: [PATCH 1/5] Added shell based watcher tool

Improve handling of files due to inotifywait having an event CLOSE_WRITE
after a file has been written.
---
 misc/watcher | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)
 create mode 100755 misc/watcher

diff --git a/misc/watcher b/misc/watcher
new file mode 100755
index 000000000..8b0d2fe21
--- /dev/null
+++ b/misc/watcher
@@ -0,0 +1,46 @@
+#!/bin/sh
+
+usage() {
+	cat <<EOF 1>&2
+Usage: $0 [-i <inputdir>] [-o <outputdir>] [-r] [-- ocroptions]"
+Automatically OCRmyPDF files in input directory and put the result into
+output directory. Specify the options for OCRmyPDF after the double
+hyphen (--).
+
+Arguments:
+	   -i: Input directory
+	   -o: Output directory
+	   -r: Remove input file on successful conversion
+EOF
+		  exit 1;
+		}
+
+INPUTDIR=/input
+OUTPUTDIR=/output
+REMOVEONSUCCESS=true
+
+while getopts ":i:o:hr" o; do
+	case "${o}" in
+		i)
+			INPUTDIR="${OPTARG}"
+			;;
+		o)
+			OUTPUTDIR="${OPTARG}"
+			;;
+		r)
+			REMOVEONSUCCESS=rm
+			;;
+		h)
+			usage
+			;;
+	esac
+done
+shift $((OPTIND-1))
+
+OCRMYPDFOPTS="$*"
+
+while true
+do 
+	F=$(inotifywait -e CLOSE_WRITE -q "$INPUTDIR" --format %f)
+	ocrmypdf $OCRMYPDFOPTS "$INPUTDIR/$F" "$OUTPUTDIR/$F" && $REMOVEONSUCCESS "$INPUTDIR/$F"
+done 

From e64247d365ee1f05722b80d3a0845ef4fccbbdc3 Mon Sep 17 00:00:00 2001
From: Sven Fischer <sven@leiderfischer.de>
Date: Mon, 3 Feb 2020 18:41:02 +0100
Subject: [PATCH 2/5] Handle duplicate output file names

---
 misc/watcher | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/misc/watcher b/misc/watcher
index 8b0d2fe21..8680c87b5 100755
--- a/misc/watcher
+++ b/misc/watcher
@@ -39,8 +39,19 @@ shift $((OPTIND-1))
 
 OCRMYPDFOPTS="$*"
 
+outfilename() {
+	OUTFILE="$1"
+	if [ -e "$OUTFILE" ]; then
+		name="${1%.*}"
+		ext="${1##*.}"
+		OUTFILE="$name".$NUM.$ext
+	fi
+}
+
 while true
 do 
 	F=$(inotifywait -e CLOSE_WRITE -q "$INPUTDIR" --format %f)
-	ocrmypdf $OCRMYPDFOPTS "$INPUTDIR/$F" "$OUTPUTDIR/$F" && $REMOVEONSUCCESS "$INPUTDIR/$F"
-done 
+	outfilename "$OUTPUTDIR/$F"
+
+	ocrmypdf $OCRMYPDFOPTS "$INPUTDIR/$F" "$OUTFILE" && $REMOVEONSUCCESS "$INPUTDIR/$F"
+done

From 700ba06fdeebfc9fc0bf43a0a1bb4dd7ff089dfb Mon Sep 17 00:00:00 2001
From: Sven Fischer <sven@leiderfischer.de>
Date: Tue, 4 Feb 2020 08:11:40 +0100
Subject: [PATCH 3/5] Updated Dockerfile to include watcher script

---
 .docker/Dockerfile | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/.docker/Dockerfile b/.docker/Dockerfile
index 527fed5d5..eaa9f24ca 100644
--- a/.docker/Dockerfile
+++ b/.docker/Dockerfile
@@ -55,6 +55,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   pngquant \
   python3 \
   qpdf \
+  inotify-tools \
   tesseract-ocr \
   tesseract-ocr-chi-sim \
   tesseract-ocr-deu \
@@ -71,6 +72,7 @@ COPY --from=builder /usr/local/bin/ /usr/local/bin/
 
 COPY --from=builder /app/misc/webservice.py /app/
 COPY --from=builder /app/misc/watcher.py /app/
+COPY --from=builder /app/misc/watcher /app/
 
 # Copy minimal project files to get the test suite.
 COPY --from=builder /app/setup.cfg /app/setup.py /app/README.md /app/

From 6c4d916c7b6013d94b0d5c66925d5aa16305f8c2 Mon Sep 17 00:00:00 2001
From: Sven Fischer <sven@leiderfischer.de>
Date: Tue, 4 Feb 2020 08:31:52 +0100
Subject: [PATCH 4/5] Added documentation about watcher

---
 docs/docker.rst | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/docs/docker.rst b/docs/docker.rst
index 392b82d9f..e1e8a0fc6 100644
--- a/docs/docker.rst
+++ b/docs/docker.rst
@@ -168,3 +168,17 @@ also licensed in this way.
 
 In addition to the above, please read our
 :ref:`general remarks on using OCRmyPDF as a service <ocr-service>`.
+
+Using OCRmyPDF for automatic file conversion
+============================================
+
+The OCRmyPDF Docker image includes a script to automatically OCR files put
+into an input directory. The watcher script is started as follows:
+
+.. code-block:: bash
+
+   docker run --name ocrwatcher --entrypoint ./watcher -v $PWD/input:/input -v $PWD/output:/output jbarlow83/ocrmypdf <watcher-options> -- <ocrmypdf-options>
+
+where ``<watcher-options>`` may optionally include ``-r`` to remove correctly
+processed input files from input folder, and after the two hyphens all command
+line options for OCRmyPDF can be included.

From 7a7ae5af829faa473f309f4898428ac8aa7c8960 Mon Sep 17 00:00:00 2001
From: Sven Fischer <sven@leiderfischer.de>
Date: Tue, 4 Feb 2020 22:00:29 +0100
Subject: [PATCH 5/5] Made inotifywait react on all changes

Was only reacting on a single change, if another change occurred during
the conversion the event was missed.
---
 misc/watcher | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/misc/watcher b/misc/watcher
index 8680c87b5..0aceaeeaa 100755
--- a/misc/watcher
+++ b/misc/watcher
@@ -44,14 +44,19 @@ outfilename() {
 	if [ -e "$OUTFILE" ]; then
 		name="${1%.*}"
 		ext="${1##*.}"
+		NUM=$(ls "$name"* | wc -l)
 		OUTFILE="$name".$NUM.$ext
 	fi
 }
 
-while true
-do 
-	F=$(inotifywait -e CLOSE_WRITE -q "$INPUTDIR" --format %f)
-	outfilename "$OUTPUTDIR/$F"
+convert_infile()
+{
+	outfilename "$OUTPUTDIR/$1"
 
 	ocrmypdf $OCRMYPDFOPTS "$INPUTDIR/$F" "$OUTFILE" && $REMOVEONSUCCESS "$INPUTDIR/$F"
-done
+}
+
+inotifywait -m -e CLOSE_WRITE -q "$INPUTDIR" --format %f |
+	while read F; do
+		  convert_infile $F
+	done