add some new tools

iiif/dl_welcomecollection.sh

@@ -0,0 +1,99 @@
+#!/bin/bash
+
+usage() {
+	echo "Usage: $0 --method [size|full] --output-dir DIR --url URL"
+	echo "Options:"
+	echo "  --method         Method to use (allowed values: size, full)"
+	echo "  --output-dir     Output directory (will be created if it doesn't exist)"
+	echo "  --url            URL to process"
+	echo "  --help           Display this help message"
+	exit 1
+}
+
+
+[ $# -eq 0 ] && usage
+METHOD=
+OUTPUT_DIR=
+URL=
+
+while [[ $# -gt 0 ]]; do
+    key="$1"
+
+    case $key in
+        --method)
+			METHOD="$2"
+			if [[ "$METHOD" != "size" && "$METHOD" != "full" ]]; then
+				echo "Error: invalid value for --method. Allowed values are 'size' or 'full'."
+				exit 1
+			fi
+			shift
+			;;
+        --output-dir)
+			OUTPUT_DIR="$2"
+			shift
+			;;
+        --url)
+			URL="$2"
+			shift
+			;;
+        --help)
+			usage
+			;;
+    *)
+        echo "Error: unknown option $1"
+        usage
+        ;;
+    esac
+	shift
+done
+
+if [ -z "$METHOD" ] || [ -z "$OUTPUT_DIR" ] || [ -z "$URL" ]; then
+    echo "Error: missing required arguments."
+    usage
+fi
+
+if [ ! -d "$OUTPUT_DIR" ]; then
+    mkdir -p "$OUTPUT_DIR"
+    if [ $? -ne 0 ]; then
+        echo "Error: failed to create directory '$OUTPUT_DIR'."
+        exit 1
+    fi
+fi
+
+cd "$OUTPUT_DIR"
+if [ ! -f data.json ]; then
+	echo "> fetching json data..."
+	curl -s "$URL" | sed -n 's|.*<script id="__NEXT_DATA__" type="application/json">\([^<]*\)</script>.*|\1|p' \
+		| jq '{width: .props.pageProps.compressedTransformedManifest.compressedCanvases.width, height: .props.pageProps.compressedTransformedManifest.compressedCanvases.height, prefix: .props.pageProps.compressedTransformedManifest.compressedCanvases.imageServiceId.prefix, suffix: .props.pageProps.compressedTransformedManifest.compressedCanvases.imageServiceId.suffix, remainders: .props.pageProps.compressedTransformedManifest.compressedCanvases.imageServiceId.remainders}' > data.json
+fi
+
+echo "> deleting junk files..."
+find . -type f -size 111c -exec rm {} +
+
+prefix=$(cat data.json | jq -r '.prefix')
+suffix=$(cat data.json | jq -r '.suffix')
+
+echo "> downloading missing pages..."
+cat data.json | jq -r '.remainders[]' | while read -r page; do
+	[ -e "$page.jpg" ] && continue
+	url="${prefix}${page}${suffix}"
+	case $METHOD in
+		full)
+			url="${url}/full/full/0/default.jpg"
+			;;
+		size)
+			jpage=$(echo $page | sed 's/^0*//')
+			jpage=$(( jpage-1 ))
+			w=$(cat sizes.json | jq ".width[$jpage]")
+			h=$(cat sizes.json | jq ".height[$jpage]")
+			url="${url}/0,0,${w},${h}/${w},${h}/0/default.jpg"
+			;;
+	esac
+	output="${page}.jpg"
+	curl --silent "$url" -H 'User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:127.0) Gecko/20100101 Firefox/127.0' -H 'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8' -H 'Accept-Language: en-US,en;q=0.5' -H 'Accept-Encoding: gzip, deflate, br, zstd' -H 'DNT: 1' -H 'Connection: keep-alive' -H 'Upgrade-Insecure-Requests: 1' -H 'Sec-Fetch-Dest: document' -H 'Sec-Fetch-Mode: navigate' -H 'Sec-Fetch-Site: same-site' -H 'Priority: u=1' -H 'Pragma: no-cache' -H 'Cache-Control: no-cache' -H 'TE: trailers' -o "$output"
+	if [ $(stat -c%s "$output") -eq 111 ]; then
+		echo "> $page FAIL"
+	else
+		echo "> $page OK"
+	fi
+done

iiif/folger.sh

@@ -0,0 +1,69 @@
+#!/bin/sh
+
+set -e
+#set -x
+
+usage() {
+	echo "usage: $0 --output-dir|-o <output-directory> [--format|-f jpg|tif] <IIIF-manifest-URL>"
+	exit 1
+}
+
+output_dir=
+url=
+format=jpg
+
+while [ "$#" -gt 0 ]; do
+	case $1 in
+		--output-dir|-o)
+			output_dir="$2"
+			shift
+			;;
+		--format|-f)
+			format="$2"
+			shift
+			;;
+		*)
+			url="$1"
+			;;
+	esac
+	shift
+done
+
+if [ -z "$output_dir" ] || [ -z "$url" ]; then
+	usage
+fi
+
+if [ -d "$output_dir" ]; then
+	echo -n "The directory '$output_dir' already exists. Do you want to delete/recreate it (d) or skip and go on (s)? "
+	read response
+	case $response in
+		[dD])
+			rm -rf "$output_dir"
+			mkdir -p "$output_dir"
+			;;
+		[sS])
+			;;
+		*)
+			2>&1 echo "error: invalid option"
+			exit 1
+			;;
+	esac
+else
+	mkdir -p "$output_dir"
+fi
+
+image_urls=$(curl -s "$url" | jq -r ".sequences[].canvases[].images[].resource.\"@id\"")
+counter=1
+for image_url in $image_urls; do
+	image_name=$(basename "$image_url")
+	filename=$(printf "$output_dir/%04d.$format" $counter)
+	if [ "$format" = "tif" ]; then
+		base="${image_url%.jpg}"
+		iamge_url="${image_url}.tif"
+	fi
+	curl -sSL "$image_url" -o "$filename"
+	counter=$((counter + 1))
+	echo "$(basename "$filename") saved"
+done
+
+cleanup

validate_tiff.sh

@@ -0,0 +1,43 @@
+#!/bin/sh
+# NOTE: tifftopnm is a part of netpbm package
+
+usage() {
+	echo "Usage: $0 [--delete]"
+	echo
+	echo "This script checks if TIFF files in the current directory are valid and prints invalid ones."
+	echo
+	echo "Options:"
+	echo "  --delete   Delete invalid TIFF files."
+	echo "  --help     Display this help message."
+}
+
+DELETE_FILES=false
+for arg in "$@"; do
+	case $arg in
+		--delete)
+			DELETE_FILES=true
+			;;
+		--help)
+			usage
+			exit 0
+			;;
+		*)
+			echo "Unknown argument: $arg" >&2
+			usage >&2
+			exit 1
+			;;
+	esac
+done
+
+for f in *.tif; do
+	echo -n "$f... "
+	tifftopnm "$f" 2>&1 >/dev/null | grep -i error >/dev/null && {
+		if [ "$DELETE_FILES" = true ]; then
+			rm "$f"
+			echo -n "DELETED"
+		else
+			echo -n "ERROR"
+		fi
+	}
+	echo 
+done

web/get-refs.sh

@@ -1,4 +1,4 @@
-#!/bin/bash
+#!/usr/local/bin/bash
 
 source ~/.local/my/bash/bashrc