diff --git a/README.md b/README.md index 69240758..570ee0e1 100644 --- a/README.md +++ b/README.md @@ -440,12 +440,12 @@ Training data In machine learning, large amounts of training data are often essential for getting good results. Many open-source machine learning projects either release only the model code (results reproducible if and only if you're Google), or a pre-baked model where the training conditions are unknown. -Libpostal is a bit different because it's trained on open data that's available to everyone, so we've released the entire training pipeline (the [geodata](https://github.com/openvenues/libpostal/tree/master/scripts/geodata) package in this repo), as well as the resulting training data itself on S3. It's over 100GB unzipped. +Libpostal is a bit different because it's trained on open data that's available to everyone, so we've released the entire training pipeline (the [geodata](https://github.com/openvenues/libpostal/tree/master/scripts/geodata) package in this repo), as well as the resulting training data itself on the Internet Archive. It's over 100GB unzipped. -Training data are stored on S3 by the date they were created. There's also a file stored on S3 to point to the most recent training data. To always point to the latest data, use something like: ```latest=$(curl https://s3.amazonaws.com/libpostal/training_data/latest)``` and use that variable in place of the date. +Training data are stored on archive.org by the date they were created. There's also a file stored in the main directory of this repo called `current_parser_training_set` which stores the date of the most recently created training set. To always point to the latest data, try something like: ```latest=$(cat current_parser_training_set)``` and use that variable in place of the date. ### Parser training sets ### -All files can be found at https://d1p366rbd94x8u.cloudfront.net/training_data/$YYYY-MM-DD/parser/$FILE as gzip'd tab-separated values (TSV) files formatted like:```language\tcountry\taddress```. +All files can be found at https://archive.org/download/libpostal-parser-training-data-YYYYMMDD/$FILE as gzip'd tab-separated values (TSV) files formatted like:```language\tcountry\taddress```. - **formatted_addresses_tagged.random.tsv.gz** (ODBL): OSM addresses. Apartments, PO boxes, categories, etc. are added primarily to these examples - **formatted_places_tagged.random.tsv.gz** (ODBL): every toponym in OSM (even cities represented as points, etc.), reverse-geocoded to its parent admins, possibly including postal codes if they're listed on the point/polygon. Every place gets a base level of representation and places with higher populations get proportionally more. diff --git a/current_parser_training_set b/current_parser_training_set new file mode 100644 index 00000000..df9876a0 --- /dev/null +++ b/current_parser_training_set @@ -0,0 +1 @@ +20170304 diff --git a/src/libpostal_data b/src/libpostal_data index 22cedf8d..77c2dd0b 100755 --- a/src/libpostal_data +++ b/src/libpostal_data @@ -3,23 +3,27 @@ set -e if [ "$#" -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 +166,66 @@ 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 > $LIBPOSTAL_DATA_VERSION_FILE; - fi + asset_names_tempfile="$LIBPOSTAL_DATA_DIR/asset_names.tmp" + echo "$release_assets" | grep -o '"name": *"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"' > $asset_names_tempfile + asset_metadata_tempfile="$LIBPOSTAL_DATA_DIR/asset_metadata.tmp" + echo "$release_assets" | grep -o '"url": *"[^"]*/releases/assets/[0-9]*"' | grep -o '"[^"]*"$' | tr -d '"' > $asset_metadata_tempfile + asset_urls_tempfile="$LIBPOSTAL_DATA_DIR/asset_urls.tmp" + echo "$release_assets" | grep -o '"browser_download_url": *"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"' > $asset_urls_tempfile + asset_sizes_tempfile="$LIBPOSTAL_DATA_DIR/asset_sizes.tmp" + echo "$release_assets" | grep -o '"size": *[0-9]*' | grep -o '[0-9]*$' > $asset_sizes_tempfile + + assets_tempfile="$LIBPOSTAL_DATA_DIR/assets.tmp" + paste -d' ' $asset_names_tempfile $asset_metadata_tempfile $asset_urls_tempfile $asset_sizes_tempfile > $assets_tempfile + + rm $asset_names_tempfile $asset_metadata_tempfile $asset_urls_tempfile $asset_sizes_tempfile + + while read -r line; do + asset=$(echo "$line" | cut -f1 -d' ') + asset_metadata_url=$(echo "$line" | cut -f2 -d' ') + asset_url=$(echo "$line" | cut -f3 -d' ') + asset_size=$(echo "$line" | cut -f4 -d' ') + + 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 < $assets_tempfile; + rm $assets_tempfile 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