wehub-resource-sync
103 lines
3.3 KiB
Bash
103 lines
3.3 KiB
Bash
#!/bin/bash
|
|
# This script initializes environment variables and paths,
|
|
# prepares Tesseract data directories, and then runs the main init script.
|
|
|
|
set -euo pipefail
|
|
|
|
append_env_path() {
|
|
local target="$1" current="$2" separator=":"
|
|
if [ -d "$target" ] && [[ ":${current}:" != *":${target}:"* ]]; then
|
|
if [ -n "$current" ]; then
|
|
printf '%s' "${target}${separator}${current}"
|
|
else
|
|
printf '%s' "${target}"
|
|
fi
|
|
else
|
|
printf '%s' "$current"
|
|
fi
|
|
}
|
|
|
|
python_site_dir() {
|
|
local venv_dir="$1"
|
|
local python_bin="$venv_dir/bin/python"
|
|
if [ -x "$python_bin" ]; then
|
|
local py_tag
|
|
if py_tag="$("$python_bin" -c 'import sys; print(f"python{sys.version_info.major}.{sys.version_info.minor}")' 2>/dev/null)" \
|
|
&& [ -n "$py_tag" ] \
|
|
&& [ -d "$venv_dir/lib/$py_tag/site-packages" ]; then
|
|
printf '%s' "$venv_dir/lib/$py_tag/site-packages"
|
|
fi
|
|
fi
|
|
}
|
|
|
|
# === LD_LIBRARY_PATH ===
|
|
# Adjust the library path depending on CPU architecture.
|
|
ARCH=$(uname -m)
|
|
case "$ARCH" in
|
|
x86_64)
|
|
[ -d /usr/lib/x86_64-linux-gnu ] && export LD_LIBRARY_PATH="/usr/lib/x86_64-linux-gnu${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
|
|
;;
|
|
aarch64)
|
|
[ -d /usr/lib/aarch64-linux-gnu ] && export LD_LIBRARY_PATH="/usr/lib/aarch64-linux-gnu${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
|
|
;;
|
|
esac
|
|
|
|
# Add LibreOffice program directory to library path if available.
|
|
if [ -d /usr/lib/libreoffice/program ]; then
|
|
export LD_LIBRARY_PATH="/usr/lib/libreoffice/program${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
|
|
fi
|
|
|
|
# === Python PATH ===
|
|
# Add virtual environments to PATH and PYTHONPATH.
|
|
for dir in /opt/venv/bin /opt/unoserver-venv/bin; do
|
|
PATH="$(append_env_path "$dir" "$PATH")"
|
|
done
|
|
export PATH
|
|
|
|
PYTHON_PATH_ENTRIES=()
|
|
for venv in /opt/venv /opt/unoserver-venv; do
|
|
if [ -d "$venv" ]; then
|
|
site_dir="$(python_site_dir "$venv")"
|
|
[ -n "${site_dir:-}" ] && PYTHON_PATH_ENTRIES+=("$site_dir")
|
|
fi
|
|
done
|
|
if [ ${#PYTHON_PATH_ENTRIES[@]} -gt 0 ]; then
|
|
PYTHONPATH="$(IFS=:; printf '%s' "${PYTHON_PATH_ENTRIES[*]}")${PYTHONPATH:+:$PYTHONPATH}"
|
|
export PYTHONPATH
|
|
fi
|
|
|
|
# # === tessdata ===
|
|
# # Prepare Tesseract OCR data directory.
|
|
# In Debian, tesseract looks in /usr/share/tesseract-ocr/5/tessdata
|
|
# For backwards compatibility, copy any user-mounted files from /usr/share/tessdata
|
|
TESSDATA_SYSTEM="/usr/share/tesseract-ocr/5/tessdata"
|
|
TESSDATA_MOUNT="/usr/share/tessdata"
|
|
|
|
log_warn() {
|
|
echo "[init][warn] $*" >&2
|
|
}
|
|
|
|
# Ensure system tessdata directory exists
|
|
mkdir -p "$TESSDATA_SYSTEM" 2>/dev/null || true
|
|
|
|
# For backwards compatibility: if user mounted custom languages to /usr/share/tessdata,
|
|
# copy them to the system location where Tesseract actually looks
|
|
if [ -d "$TESSDATA_MOUNT" ] && [ "$(ls -A "$TESSDATA_MOUNT" 2>/dev/null)" ]; then
|
|
log_warn "Found user-mounted tessdata in $TESSDATA_MOUNT, copying to system location $TESSDATA_SYSTEM"
|
|
cp -rn "$TESSDATA_MOUNT"/* "$TESSDATA_SYSTEM"/ 2>/dev/null || true
|
|
fi
|
|
|
|
# Set TESSDATA_PREFIX to system location
|
|
export TESSDATA_PREFIX="$TESSDATA_SYSTEM"
|
|
log_warn "Using TESSDATA_PREFIX=$TESSDATA_PREFIX"
|
|
|
|
# === Temp dir ===
|
|
# Ensure the temporary directory exists and has proper permissions.
|
|
mkdir -p /tmp/stirling-pdf
|
|
chown -R stirlingpdfuser:stirlingpdfgroup /tmp/stirling-pdf || true
|
|
chmod -R 755 /tmp/stirling-pdf || true
|
|
|
|
# === Start application ===
|
|
# Run the main init script that handles the full startup logic.
|
|
exec /scripts/init-without-ocr.sh
|