From 0c913794241cb18ac1eb84b25a17676607c2898a Mon Sep 17 00:00:00 2001 From: Al Date: Sat, 10 Mar 2018 19:03:14 -0500 Subject: [PATCH] [build/fix] using Github for the model releases rather than the Mapzen S3 buckets which are no longer working after the shutdown. It requires a little more effort to get the metadata, but downloads should still be just as fast since Github releases are on S3 as well. Note: still need to implement the upload piece, but this at least provides a model endpoint for users. --- src/libpostal_data | 119 ++++++++++++++++++++++++++------------------- 1 file changed, 68 insertions(+), 51 deletions(-) diff --git a/src/libpostal_data b/src/libpostal_data index 22cedf8d..558b589f 100755 --- a/src/libpostal_data +++ b/src/libpostal_data @@ -3,23 +3,27 @@ set -e if [ "$#" -lt 3 ]; then - echo "Usage: ./libpostal_data [upload|download] [base|geodb] data_dir" + echo "Usage: ./libpostal_data [upload|download] [base|parser|language_classifier|all] data_dir" exit 1 fi LIBPOSTAL_VERSION_STRING="v1" +LIBPOSTAL_RELEASE_VERSION_STRING="v1.0.0" +LIBPOSTAL_REPO_NAME="openvenues/libpostal" LIBPOSTAL_S3_BUCKET_NAME="libpostal" LIBPOSTAL_S3_KEY="s3://$LIBPOSTAL_S3_BUCKET_NAME" -LIBPOSTAL_S3_BUCKET_URL="https://$LIBPOSTAL_S3_BUCKET_NAME.s3.amazonaws.com" -LIBPOSTAL_CLOUDFRONT_URL="https://d1p366rbd94x8u.cloudfront.net" + +GITHUB_API_URL="https://api.github.com" +LIBPOSTAL_RELEASE_API_URL="$GITHUB_API_URL/repos/$LIBPOSTAL_REPO_NAME/releases" + LIBPOSTAL_DATA_FILE="libpostal_data.tar.gz" LIBPOSTAL_PARSER_FILE="parser.tar.gz" LIBPOSTAL_LANG_CLASS_FILE="language_classifier.tar.gz" -LIBPOSTAL_DATA_S3_PREFIX="data/$LIBPOSTAL_VERSION_STRING" -LIBPOSTAL_PARSER_S3_PREFIX="models/address_parser" -LIBPOSTAL_LANG_CLASS_S3_PREFIX="models/language_classifier" +LIBPOSTAL_DATA_S3_PREFIX="$LIBPOSTAL_LATEST_MODEL_VERSION_STRING/parser.tar.gz" +LIBPOSTAL_PARSER_S3_PREFIX="$LIBPOSTAL_LATEST_DATA_VERSION_STRING/libpostal_data.tar.gz" +LIBPOSTAL_LANG_CLASS_S3_PREFIX="$LIBPOSTAL_LATEST_MODEL_VERSION_STRING/language_classifier.tar.gz" COMMAND=$1 FILE=$2 @@ -58,7 +62,7 @@ kill_background_processes() { trap kill_background_processes INT PART_MSG='echo "Downloading part $1: filename=$5, offset=$2, max=$3"' -PART_CURL='curl $4 --silent -H"Range:bytes=$2-$3" --retry 3 --retry-delay 2 -o $5' +PART_CURL='curl -L $4 --silent -H"Range:bytes=$2-$3" --retry 3 --retry-delay 2 -o $5' DOWNLOAD_PART="$PART_MSG;$PART_CURL" @@ -99,10 +103,12 @@ download_multipart() { download_file() { updated_path=$1 data_dir=$2 - prefix=$3 - filename=$4 - name=$5 - shift 5 + metadata_url=$3 + url=$4 + size=$5 + filename=$6 + name=$7 + shift 7 subdirs=$@ local_path=$data_dir/$filename @@ -113,16 +119,13 @@ download_file() { echo "Checking for new libpostal $name..." - url=$LIBPOSTAL_CLOUDFRONT_URL/$prefix/$filename - - if [ $(curl -sI $url -z "$(cat $updated_path)" --remote-time -w %{http_code} -o /dev/null | grep "^200$") ]; then + if [ $(curl -LsI $metadata_url -z "$(cat $updated_path)" --remote-time -w %{http_code} -o /dev/null | grep "^200$") ]; then echo "New libpostal $name available" - content_length=$(curl -I $url 2> /dev/null | awk 'tolower($0) ~ /^content-length:/ { print $2 }' | tr -d '[[:space:]]') - if [ $content_length -ge $LARGE_FILE_SIZE ]; then - download_multipart $url $local_path $content_length + if [ $size -ge $LARGE_FILE_SIZE ]; then + download_multipart $url $local_path $size else - curl $url --retry 3 --retry-delay 2 -o $local_path + curl -L $url --retry 3 --retry-delay 2 -o $local_path fi if date -ud "@$(date -ur . +%s)" >/dev/null 2>&1; then @@ -163,45 +166,59 @@ if [ $COMMAND = "download" ]; then mkdir -p $LIBPOSTAL_DATA_DIR - if [ $FILE = "base" ] || [ $FILE = "all" ]; then - download_file $LIBPOSTAL_DATA_UPDATED_PATH $LIBPOSTAL_DATA_DIR $LIBPOSTAL_DATA_S3_PREFIX $LIBPOSTAL_DATA_FILE "data file" $BASIC_MODULE_DIRS - fi - if [ $FILE = "parser" ] || [ $FILE = "all" ]; then - latest_parser=$(curl --silent "$LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_PARSER_S3_PREFIX/latest") - parser_s3_prefix="$LIBPOSTAL_PARSER_S3_PREFIX/$latest_parser" - download_file $LIBPOSTAL_PARSER_UPDATED_PATH $LIBPOSTAL_DATA_DIR $parser_s3_prefix $LIBPOSTAL_PARSER_FILE "parser data file" $PARSER_MODULE_DIR - fi - if [ $FILE = "language_classifier" ] || [ $FILE = "all" ]; then - latest_lang_class=$(curl --silent "$LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_LANG_CLASS_S3_PREFIX/latest") - lang_class_s3_prefix="$LIBPOSTAL_LANG_CLASS_S3_PREFIX/$latest_lang_class" - download_file $LIBPOSTAL_LANG_CLASS_UPDATED_PATH $LIBPOSTAL_DATA_DIR $lang_class_s3_prefix $LIBPOSTAL_LANG_CLASS_FILE "language classifier data file" $LANGUAGE_CLASSIFIER_MODULE_DIR - fi + release_id=$(curl -s $LIBPOSTAL_RELEASE_API_URL/tags/$LIBPOSTAL_RELEASE_VERSION_STRING | grep "\"id\"" | head -n1 | grep -o '[0-9][0-9]*') + release_assets="$(curl -s $LIBPOSTAL_RELEASE_API_URL/$release_id/assets)" - if [ "$LIBPOSTAL_DATA_DIR_VERSION" != "$LIBPOSTAL_VERSION_STRING" ]; then - echo $LIBPOSTAL_VERSION_STRING > $LIBPOSTAL_DATA_VERSION_FILE; - fi + asset_names=($(echo "$release_assets" | grep -o '"name": *"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"')) + asset_metadata_urls=($(echo "$release_assets" | grep -o '"url": *"[^"]*/releases/assets/[0-9]*"' | grep -o '"[^"]*"$' | tr -d '"')) + asset_urls=($(echo "$release_assets" | grep -o '"browser_download_url": *"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"')) + + num_assets=${#asset_names[*]} + + asset_sizes=($(echo "$release_assets" | grep -o '"size": *[0-9]*' | grep -o '[0-9]*$')) + + for i in $(seq 0 $((num_assets - 1))); do + asset=${asset_names[i]}; + asset_metadata_url=${asset_metadata_urls[i]}; + asset_url=${asset_urls[i]}; + asset_size=${asset_sizes[i]}; + + if [ $asset = $LIBPOSTAL_DATA_FILE ] && ([ $FILE = "base" ] || [ $FILE == "all" ]); then + download_file $LIBPOSTAL_DATA_UPDATED_PATH $LIBPOSTAL_DATA_DIR $asset_metadata_url $asset_url $asset_size $LIBPOSTAL_DATA_FILE "data file" $BASIC_MODULE_DIRS + fi + if [ $asset = $LIBPOSTAL_PARSER_FILE ] && ([ $FILE = "parser" ] || [ $FILE = "all" ]); then + download_file $LIBPOSTAL_PARSER_UPDATED_PATH $LIBPOSTAL_DATA_DIR $asset_metadata_url $asset_url $asset_size $LIBPOSTAL_PARSER_FILE "parser data file" $PARSER_MODULE_DIR + fi + if [ $asset = $LIBPOSTAL_LANG_CLASS_FILE ] && ([ $FILE = "language_classifier" ] || [ $FILE = "all" ]); then + download_file $LIBPOSTAL_LANG_CLASS_UPDATED_PATH $LIBPOSTAL_DATA_DIR $asset_metadata_url $asset_url $asset_size $LIBPOSTAL_LANG_CLASS_FILE "language classifier data file" $LANGUAGE_CLASSIFIER_MODULE_DIR + fi + + if [ "$LIBPOSTAL_DATA_DIR_VERSION" != "$LIBPOSTAL_VERSION_STRING" ]; then + echo $LIBPOSTAL_VERSION_STRING > $LIBPOSTAL_DATA_VERSION_FILE; + fi + done elif [ $COMMAND = "upload" ]; then + echo "upload not implemented yet" - if [ $FILE = "base" ] || [ $FILE = "all" ]; then - tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_DATA_FILE $BASIC_MODULE_DIRS - aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_DATA_FILE $LIBPOSTAL_S3_KEY/$LIBPOSTAL_DATA_S3_PREFIX/ - fi + #if [ $FILE = "base" ] || [ $FILE = "all" ]; then + # tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_DATA_FILE $BASIC_MODULE_DIRS + # aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_DATA_FILE $LIBPOSTAL_S3_KEY/$LIBPOSTAL_DATA_S3_PREFIX/ + #fi - if [ $FILE = "parser" ] || [ $FILE = "all" ]; then - latest_parser=$(curl --silent $LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_PARSER_S3_PREFIX/latest) - tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_PARSER_FILE $PARSER_MODULE_DIR - parser_s3_dir="$LIBPOSTAL_S3_KEY/$LIBPOSTAL_PARSER_S3_PREFIX/$latest_parser/" - aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_PARSER_FILE $parser_s3_dir - fi - - if [ $FILE = "language_classifier" ] || [ $FILE = "all" ]; then - latest_lang_class=$(curl --silent $LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_LANG_CLASS_S3_PREFIX/latest) - tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_LANG_CLASS_FILE $LANGUAGE_CLASSIFIER_MODULE_DIR - lang_class_s3_dir="$LIBPOSTAL_S3_KEY/$LIBPOSTAL_LANG_CLASS_S3_PREFIX/$latest_lang_class/" - aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_LANG_CLASS_FILE $lang_class_s3_dir - fi + #if [ $FILE = "parser" ] || [ $FILE = "all" ]; then + # latest_parser=$(curl --silent $LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_PARSER_S3_PREFIX/latest) + # tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_PARSER_FILE $PARSER_MODULE_DIR + # parser_s3_dir="$LIBPOSTAL_S3_KEY/$LIBPOSTAL_PARSER_S3_PREFIX/$latest_parser/" + # aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_PARSER_FILE $parser_s3_dir + #fi + #if [ $FILE = "language_classifier" ] || [ $FILE = "all" ]; then + # latest_lang_class=$(curl --silent $LIBPOSTAL_S3_BUCKET_URL/$LIBPOSTAL_LANG_CLASS_S3_PREFIX/latest) + # tar -C $LIBPOSTAL_DATA_DIR -cvzf $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_LANG_CLASS_FILE $LANGUAGE_CLASSIFIER_MODULE_DIR + # lang_class_s3_dir="$LIBPOSTAL_S3_KEY/$LIBPOSTAL_LANG_CLASS_S3_PREFIX/$latest_lang_class/" + # aws s3 cp --acl=public-read $LIBPOSTAL_DATA_DIR/$LIBPOSTAL_LANG_CLASS_FILE $lang_class_s3_dir + #fi else echo "Invalid command: $COMMAND" exit 1