Add greposcope: script for downloading and searching pattern in diffoscope output of unreproducible packages

> Usage: greposcope [OPTIONS] pattern
>
> Searches for 'pattern' in the diffoscope output of every unreproducible packages reported at https://reproducible.archlinux.org.
> This is useful to identify packages that are unreproducible because of a specific issue.
>
> OPTIONS
>     -h, --help            Show this message
>     -i, --ignore-case     Ignore case distinctions in patterns and input data
>
> Examples:
>     $ greposcope "gzip compressed data"
>     $ greposcope -i zipinfo

Diffoscope outputs are downloaded under `${TMPDIR:-/tmp}/greposcope-${UID}"`. Only diffoscope outputs that haven't been downloaded yet (as in, not found in `${TMPDIR:-/tmp}/greposcope-${UID}"`) are downloaded, so only the first run of the script downloads "world" (so we don't make a huge amount of requests to our `rebuilderd` instance at each run).
Tasks are parallelized to reduce execution time.

Dependencies: `rebuilderd-tools` & `parallel`

Author: Antiz96

Makefile

@@ -10,7 +10,8 @@ BASH_SCRIPTS = \
 	package/parse-submodules \
 	package/pkgsearch \
 	package/rebuild-todo \
-	package/pkggrep
+	package/pkggrep \
+	package/greposcope
 
 PYTHON_SCRIPTS = \
 	package/staging2testing \

package/greposcope

@@ -0,0 +1,89 @@
+#!/bin/bash
+
+# SPDX-License-Identifier: GPL-2.0
+
+# Dependencies:
+# rebuilderd-tools
+# parallel
+
+progname="greposcope"
+tmp_dir="${TMPDIR:-/tmp}/${progname}-${UID}"
+rebuilderd_host="https://reproducible.archlinux.org"
+
+help() {
+	cat <<EOF
+Usage: ${progname} [OPTIONS] pattern
+
+Searches for 'pattern' in the diffoscope output of every unreproducible packages reported at ${rebuilderd_host}.
+This is useful to identify packages that are unreproducible because of a specific issue.
+
+OPTIONS
+    -h, --help            Show this message
+    -i, --ignore-case     Ignore case distinctions in patterns and input data
+
+Examples:
+    $ ${progname} "gzip compressed data"
+    $ ${progname} -i zipinfo
+EOF
+}
+
+while ((${#})); do
+	case "${1}" in
+		-h|--help)
+			help
+			exit 0
+		;;
+		-i|--ignore-case)
+			extra_grep_opt="--ignore-case"
+		;;
+		--)
+			shift
+			break
+		;;
+		-*)
+			echo -e >&2 "Invalid argument -- '${1}'\nTry '${progname} --help' for more information"
+			exit 1
+		;;
+		*)
+			pattern="${1}"
+		;;
+	esac
+	shift
+done
+
+if [ -z "${pattern}" ]; then
+    echo -e >&2 "No pattern provided\nTry '${progname} --help' for more information"
+    exit 1
+fi
+
+mkdir -p "${tmp_dir}"
+
+echo -e "Fetching the list of unreproducible packages...\n"
+mapfile -t pkg_list < <(rebuildctl -H "${rebuilderd_host}" pkgs ls --status BAD | awk '{print $3}')
+
+if [ ${#pkg_list[@]} -eq 0 ]; then
+	echo "No unreproducible packages found"
+	exit 0
+fi
+
+# Substract packages that already have their diffoscope output downloaded from the list
+updated_pkg_list=()
+for pkg in "${pkg_list[@]}"; do
+    if [ ! -f "${tmp_dir}/${pkg}.diffoscope" ]; then
+        updated_pkg_list+=("${pkg}")
+    fi
+done
+
+pkg_list=("${updated_pkg_list[@]}")
+
+if [ ${#pkg_list[@]} -eq 0 ]; then
+	echo "All diffoscope outputs already downloaded"
+else
+	echo -e "Downloading diffoscope outputs of ${#pkg_list[@]} unreproducible packages...\nThis may take some time...\n"
+	parallel --bar -j "$(nproc)" \
+		rebuildctl -H "${rebuilderd_host}" pkgs diffoscope --name {} "&>" "${tmp_dir}/{}.diffoscope" ::: "${pkg_list[@]}"
+fi
+
+echo -e "\nSearching for \"${pattern}\" in diffoscope outputs...\n"
+parallel --silent -j "$(nproc)" \
+	"grep --color=always --with-filename --line-number ${extra_grep_opt} '${pattern}' {}" ::: "${tmp_dir}"/*.diffoscope || true