refparse.sh

#!/usr/bin/env bash
#
# Companion script for dockerised refparse execution.
#
# Builds the image if it isn't already present, then turns one or more LRG
# (.xml) or GenBank (.gb/.gbk) input files into PDFs (default) or plain-text
# transcriptions (--text).
#
# Usage:
#   ./refparse.sh input/LRG_110.xml
#   ./refparse.sh --text input/GB_TEST.gb input/LRG_1.xml
#
# Output:
#   PDFs            -> output/pdf/
#   generated .tex  -> output/tex/
#   text mode       -> output/txt/
#
# Override the image name with REFPARSE_IMAGE=... if needed.
#
set -euo pipefail

IMAGE="${REFPARSE_IMAGE:-mattwellie/refparse:latest}"
REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"

mode="pdf"
inputs=()
for arg in "$@"; do
    case "$arg" in
        --text) mode="text" ;;
        -h | --help)
            # print the leading comment block (skip the shebang, stop at code)
            awk 'NR==1 {next} /^#/ {sub(/^# ?/, ""); print; next} {exit}' "$0"
            exit 0
            ;;
        -*)
            echo "Unknown option: $arg" >&2
            exit 2
            ;;
        *) inputs+=("$arg") ;;
    esac
done

if [ "${#inputs[@]}" -eq 0 ]; then
    echo "Usage: $0 [--text] FILE [FILE ...]" >&2
    exit 2
fi

# Build the image once if it isn't available locally.
if ! docker image inspect "$IMAGE" >/dev/null 2>&1; then
    echo ">> building $IMAGE"
    docker build -t "$IMAGE" "$REPO_ROOT"
fi

out="$REPO_ROOT/output"
primers="$REPO_ROOT/primers"
mkdir -p "$out" "$primers"

text_flag=""
[ "$mode" = "text" ] && text_flag="--text"

# Generate a .tex (or .txt) per input. The container derives the output name
# from the "input/<file>" path, so each file's directory is mounted as /input.
for f in "${inputs[@]}"; do
    if [ ! -f "$f" ]; then
        echo "!! no such file: $f" >&2
        exit 1
    fi
    dir="$(cd "$(dirname "$f")" && pwd)"
    base="$(basename "$f")"
    echo ">> $base ($mode)"
    docker run --rm \
        -v "$dir:/input" -v "$out:/output" -v "$primers:/primers" \
        "$IMAGE" -i "input/$base" $text_flag
done

if [ "$mode" = "text" ]; then
    mkdir -p "$out/txt"
    mv -f "$out"/*.txt "$out/txt/" 2>/dev/null || true
    echo ">> text transcriptions in output/txt/"
    exit 0
fi

# Compile each freshly generated .tex to PDF using the same image, then tidy up.
shopt -s nullglob
tex_files=("$out"/*.tex)
if [ "${#tex_files[@]}" -eq 0 ]; then
    echo "!! no .tex files were generated" >&2
    exit 1
fi

for tex in "${tex_files[@]}"; do
    docker run --rm --user "$(id -u):$(id -g)" \
        -w /sources -v "$REPO_ROOT:/sources" \
        --entrypoint pdflatex "$IMAGE" \
        -interaction=nonstopmode -output-directory=output "output/$(basename "$tex")"
done

mkdir -p "$out/pdf" "$out/tex"
mv -f "$out"/*.pdf "$out/pdf/" 2>/dev/null || true
mv -f "$out"/*.tex "$out/tex/" 2>/dev/null || true
rm -f "$out"/*.log "$out"/*.aux "$out"/*.out "$out"/*.upa "$out"/*.upb
echo ">> PDFs in output/pdf/, .tex sources in output/tex/"