From 4ac04003dc350196a5f67a64174edd507fae497c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rapha=C3=ABl=20Bournhonesque?= Date: Fri, 8 Nov 2024 17:15:07 +0100 Subject: [PATCH] feat: try to export mobile dumps as gzipped TSV --- openfoodfacts_exports/exports/csv.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/openfoodfacts_exports/exports/csv.py b/openfoodfacts_exports/exports/csv.py index e9735df..0e8654d 100644 --- a/openfoodfacts_exports/exports/csv.py +++ b/openfoodfacts_exports/exports/csv.py @@ -10,7 +10,7 @@ logger = logging.getLogger(__name__) MOBILE_APP_DUMP_DATASET_PATH = ( - settings.DATASET_DIR / "en.openfoodfacts.org.products.csv" + settings.DATASET_DIR / "openfoodfacts-mobile-dump-products.tsv.gz" ) MOBILE_APP_DUMP_SQL_QUERY = r""" @@ -37,7 +37,7 @@ def generate_mobile_app_dump(parquet_path: Path, output_path: Path) -> None: raise FileNotFoundError(f"{str(parquet_path)} was not found.") with tempfile.TemporaryDirectory() as tmp_dir: - tmp_file_path = Path(tmp_dir) / "mobile_dump.csv" + tmp_file_path = Path(tmp_dir) / "mobile_dump.csv.gz" query = MOBILE_APP_DUMP_SQL_QUERY.replace( "{dataset_path}", str(parquet_path) ).replace("{output_path}", str(tmp_file_path))