refactor: remove vendored binaries and raw assets from repo

.gitignore

@@ -4,6 +4,11 @@ config.yaml
 # ISO output files
 *.iso
 
+# Vendored/runtime assets fetched outside git
+dcmtk-bin/
+legacy/raw/
+.local/
+
 # Temp directories
 /tmp/dicomdir_*
 
@@ -20,3 +25,10 @@ Thumbs.db
 
 # Local-only docs tell @VArtzy<farrelnikoson@gmail.com> for request
 docs/phases
+
+# Keep placeholder docs for ignored asset directories
+!dcmtk-bin/
+!dcmtk-bin/README.md
+!legacy/
+!legacy/raw/
+!legacy/raw/README.md

README.md

@@ -25,6 +25,9 @@ The service needs:
 - MicroDicom files
 - writable temp storage
 
+Large runtime assets are intentionally not stored in git.
+Use the setup scripts in `scripts/` to stage local copies in ignored directories.
+
 ## Setup
 Before running the service, make sure the VM has:
 - Go installed, if you are building on that machine
@@ -33,6 +36,20 @@ Before running the service, make sure the VM has:
 - network access to PACS, the patient API, and the CD publisher
 - a writable temp directory
 
+Stage local runtime assets if needed:
+
+```bash
+scripts/setup-dcmtk.sh --source-dir /path/to/dcmtk/bin
+scripts/setup-microdicom.sh --source-dir /path/to/microdicom
+```
+
+Or download your hosted release assets directly:
+
+```bash
+scripts/setup-dcmtk.sh --archive-url https://github.com/<owner>/<repo>/releases/download/<tag>/dcmtk-bin.tar.gz
+scripts/setup-microdicom.sh --archive-url https://github.com/<owner>/<repo>/releases/download/<tag>/microdicom.zip
+```
+
 Create a local config file from the template:
 
 ```bash
@@ -40,6 +57,7 @@ cp config.example.yaml config.yaml
 ```
 
 Then adjust the paths, hosts, ports, and tokens for your environment.
+For local staging via the setup scripts, point config at `.local/dcmtk-bin/` and `.local/microdicom/`.
 
 ## Build
 A normal Go build is enough in a friendly environment:

docs/deployment-vm.md

@@ -29,6 +29,13 @@ The service needs:
 - a writable temp directory
 - a free port range for `storescp`
 
+If you do not manage these assets globally on the VM, you can stage them locally first:
+
+```bash
+scripts/setup-dcmtk.sh --source-dir /path/to/dcmtk/bin --install-dir /opt/dicom-iso/dcmtk-bin
+scripts/setup-microdicom.sh --source-dir /path/to/microdicom --install-dir /opt/dicom-iso/microdicom
+```
+
 ## Build note
 The build environment cannot depend on public internet access.
 That means the binary must be built through an approved offline-friendly path.

docs/repo-size-cleanup.md

@@ -0,0 +1,18 @@
+# Repo size cleanup
+
+Implemented in the working tree:
+- removed tracked binaries from `dcmtk-bin/`
+- removed tracked raw assets from `legacy/raw/`
+- added `.gitignore` protections
+- added setup scripts for DCMTK and MicroDicom
+
+## Finish the cleanup in git history
+Rewriting history is still required to shrink the remote repository size.
+
+```bash
+git filter-repo --path dcmtk-bin --path legacy/raw --invert-paths
+git push --force --all
+git push --force --tags
+```
+
+Coordinate this with any collaborators first.

scripts/purge-git-history.sh

@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+cat <<'EOF'
+This rewrites git history to remove large vendored assets.
+Run only after committing the current tree cleanup and coordinating with collaborators.
+
+Commands:
+  git filter-repo --path dcmtk-bin --path legacy/raw --invert-paths
+  git push --force --all
+  git push --force --tags
+
+Afterward, collaborators should reclone or hard-reset to the rewritten history.
+EOF

scripts/setup-dcmtk.sh

@@ -0,0 +1,139 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+INSTALL_DIR="${INSTALL_DIR:-$ROOT_DIR/.local/dcmtk-bin}"
+REQUIRED_BINS=(storescp movescu storescu findscu dcmdump dcmodify getscu echoscu dcmj2pnm)
+SOURCE_DIR=""
+ARCHIVE_URL=""
+TMP_DIR=""
+
+usage() {
+  cat <<'EOF'
+Usage:
+  scripts/setup-dcmtk.sh [--source-dir DIR | --archive-url URL] [--install-dir DIR]
+
+Behavior:
+  - Copies required DCMTK binaries into a local ignored directory.
+  - If --source-dir is omitted, binaries are resolved from PATH.
+  - If --archive-url is given, the script downloads and extracts an archive,
+    then searches for the required binaries inside it.
+
+Examples:
+  scripts/setup-dcmtk.sh --source-dir /opt/dcmtk/bin
+  scripts/setup-dcmtk.sh --archive-url https://github.com/<owner>/<repo>/releases/download/<tag>/dcmtk-bin.tar.gz
+  scripts/setup-dcmtk.sh --install-dir /data/dcmtk-bin
+EOF
+}
+
+download() {
+  local url="$1"
+  local out="$2"
+  if command -v curl >/dev/null 2>&1; then
+    curl -fL "$url" -o "$out"
+  elif command -v wget >/dev/null 2>&1; then
+    wget -O "$out" "$url"
+  else
+    echo "need curl or wget to download $url" >&2
+    exit 1
+  fi
+}
+
+extract_archive() {
+  local archive="$1"
+  local dest="$2"
+  case "$archive" in
+    *.tar.gz|*.tgz) tar -xzf "$archive" -C "$dest" ;;
+    *.tar.xz) tar -xJf "$archive" -C "$dest" ;;
+    *.tar) tar -xf "$archive" -C "$dest" ;;
+    *.zip)
+      command -v unzip >/dev/null 2>&1 || { echo "unzip is required for $archive" >&2; exit 1; }
+      unzip -q "$archive" -d "$dest"
+      ;;
+    *)
+      echo "unsupported archive format: $archive" >&2
+      exit 1
+      ;;
+  esac
+}
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --source-dir)
+      SOURCE_DIR="${2:?missing value for --source-dir}"
+      shift 2
+      ;;
+    --archive-url)
+      ARCHIVE_URL="${2:?missing value for --archive-url}"
+      shift 2
+      ;;
+    --install-dir)
+      INSTALL_DIR="${2:?missing value for --install-dir}"
+      shift 2
+      ;;
+    -h|--help)
+      usage
+      exit 0
+      ;;
+    *)
+      echo "Unknown argument: $1" >&2
+      usage >&2
+      exit 1
+      ;;
+  esac
+done
+
+if [[ -n "$SOURCE_DIR" && -n "$ARCHIVE_URL" ]]; then
+  echo "use only one of --source-dir or --archive-url" >&2
+  exit 1
+fi
+
+cleanup() {
+  [[ -n "$TMP_DIR" && -d "$TMP_DIR" ]] && rm -rf "$TMP_DIR"
+}
+trap cleanup EXIT
+
+if [[ -n "$ARCHIVE_URL" ]]; then
+  TMP_DIR="$(mktemp -d)"
+  archive="$TMP_DIR/archive"
+  download "$ARCHIVE_URL" "$archive"
+  mkdir -p "$TMP_DIR/extracted"
+  mv "$archive" "$TMP_DIR/archive$(basename "$ARCHIVE_URL" | sed 's/.*\(\.[^.][^.]*\)$/\1/')" 2>/dev/null || true
+  archive_path="$(find "$TMP_DIR" -maxdepth 1 -type f | head -n 1)"
+  extract_archive "$archive_path" "$TMP_DIR/extracted"
+  SOURCE_DIR="$TMP_DIR/extracted"
+fi
+
+mkdir -p "$INSTALL_DIR"
+
+resolve_bin() {
+  local name="$1"
+  if [[ -n "$SOURCE_DIR" ]]; then
+    local candidate
+    candidate="$(find "$SOURCE_DIR" -type f -name "$name" -perm -u+x | head -n 1 || true)"
+    [[ -n "$candidate" ]] || candidate="$(find "$SOURCE_DIR" -type f -name "$name" | head -n 1 || true)"
+    [[ -n "$candidate" ]] || { echo "missing binary in source: $name" >&2; return 1; }
+    printf '%s\n' "$candidate"
+    return 0
+  fi
+
+  command -v "$name" >/dev/null 2>&1 || { echo "binary not found in PATH: $name" >&2; return 1; }
+  command -v "$name"
+}
+
+for bin in "${REQUIRED_BINS[@]}"; do
+  src="$(resolve_bin "$bin")"
+  install -m 0755 "$src" "$INSTALL_DIR/$bin"
+  echo "installed $bin -> $INSTALL_DIR/$bin"
+done
+
+cat <<EOF
+
+Done.
+Set these config values:
+
+dcmtk:
+  storescp: "$INSTALL_DIR/storescp"
+  movescu:  "$INSTALL_DIR/movescu"
+  storescu: "$INSTALL_DIR/storescu"
+EOF

scripts/setup-microdicom.sh

@@ -0,0 +1,118 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+INSTALL_DIR="${INSTALL_DIR:-$ROOT_DIR/.local/microdicom}"
+SOURCE_DIR=""
+ARCHIVE_URL=""
+TMP_DIR=""
+
+usage() {
+  cat <<'EOF'
+Usage:
+  scripts/setup-microdicom.sh [--source-dir DIR | --archive-url URL] [--install-dir DIR]
+
+Behavior:
+  - Copies a prepared MicroDicom directory into a local ignored directory.
+  - With --archive-url, downloads and extracts an archive, then searches for
+    a directory containing AUTORUN.INF and MICROD/.
+
+Example:
+  scripts/setup-microdicom.sh --archive-url https://github.com/<owner>/<repo>/releases/download/<tag>/microdicom.zip
+EOF
+}
+
+download() {
+  local url="$1"
+  local out="$2"
+  if command -v curl >/dev/null 2>&1; then
+    curl -fL "$url" -o "$out"
+  elif command -v wget >/dev/null 2>&1; then
+    wget -O "$out" "$url"
+  else
+    echo "need curl or wget to download $url" >&2
+    exit 1
+  fi
+}
+
+extract_archive() {
+  local archive="$1"
+  local dest="$2"
+  case "$archive" in
+    *.tar.gz|*.tgz) tar -xzf "$archive" -C "$dest" ;;
+    *.tar.xz) tar -xJf "$archive" -C "$dest" ;;
+    *.tar) tar -xf "$archive" -C "$dest" ;;
+    *.zip)
+      command -v unzip >/dev/null 2>&1 || { echo "unzip is required for $archive" >&2; exit 1; }
+      unzip -q "$archive" -d "$dest"
+      ;;
+    *)
+      echo "unsupported archive format: $archive" >&2
+      exit 1
+      ;;
+  esac
+}
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --source-dir)
+      SOURCE_DIR="${2:?missing value for --source-dir}"
+      shift 2
+      ;;
+    --archive-url)
+      ARCHIVE_URL="${2:?missing value for --archive-url}"
+      shift 2
+      ;;
+    --install-dir)
+      INSTALL_DIR="${2:?missing value for --install-dir}"
+      shift 2
+      ;;
+    -h|--help)
+      usage
+      exit 0
+      ;;
+    *)
+      echo "Unknown argument: $1" >&2
+      usage >&2
+      exit 1
+      ;;
+  esac
+done
+
+if [[ -n "$SOURCE_DIR" && -n "$ARCHIVE_URL" ]]; then
+  echo "use only one of --source-dir or --archive-url" >&2
+  exit 1
+fi
+
+cleanup() {
+  [[ -n "$TMP_DIR" && -d "$TMP_DIR" ]] && rm -rf "$TMP_DIR"
+}
+trap cleanup EXIT
+
+if [[ -n "$ARCHIVE_URL" ]]; then
+  TMP_DIR="$(mktemp -d)"
+  archive="$TMP_DIR/$(basename "$ARCHIVE_URL")"
+  download "$ARCHIVE_URL" "$archive"
+  mkdir -p "$TMP_DIR/extracted"
+  extract_archive "$archive" "$TMP_DIR/extracted"
+  SOURCE_DIR="$(find "$TMP_DIR/extracted" -type f -name AUTORUN.INF -printf '%h\n' | head -n 1 || true)"
+fi
+
+[[ -n "$SOURCE_DIR" ]] || { echo "one of --source-dir or --archive-url is required" >&2; exit 1; }
+[[ -d "$SOURCE_DIR" ]] || { echo "source directory does not exist: $SOURCE_DIR" >&2; exit 1; }
+[[ -f "$SOURCE_DIR/AUTORUN.INF" ]] || { echo "missing AUTORUN.INF in source directory" >&2; exit 1; }
+[[ -d "$SOURCE_DIR/MICROD" ]] || { echo "missing MICROD/ in source directory" >&2; exit 1; }
+
+rm -rf "$INSTALL_DIR"
+mkdir -p "$INSTALL_DIR"
+cp -a "$SOURCE_DIR"/. "$INSTALL_DIR"/
+
+echo "installed MicroDicom assets -> $INSTALL_DIR"
+cat <<EOF
+
+Done.
+Set this config value:
+
+iso:
+  microdicom_path: "$INSTALL_DIR"
+EOF