Add ratio filtering script for ST

scripts/filter_tsv_based_on_ratio.py

@@ -0,0 +1,83 @@
+# Copyright 2025 FBK
+
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+
+#     http://www.apache.org/licenses/LICENSE-2.0
+
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License
+
+import argparse
+import csv
+import logging
+import os
+
+
+def filter_data(tsv_in, threshold_min, threshold_max, tsv_out):
+    # Setup logger
+    logging.basicConfig(
+        format="%(asctime)s | %(levelname)s | %(name)s | %(message)s",
+        datefmt="%Y-%m-%d %H:%M:%S",
+        level=os.environ.get("LOGLEVEL", "INFO").upper(),
+    )
+    logger = logging.getLogger("filter_on_char_ratio")
+    filtered = 0
+    with open(tsv_in, 'r') as in_f, open(tsv_out, 'w') as out_f:
+        # Copy header
+        reader = csv.DictReader(
+            in_f,
+            delimiter="\t",
+            quotechar=None,
+            doublequote=False,
+            lineterminator="\n",
+            quoting=csv.QUOTE_NONE,
+        )
+        writer = csv.DictWriter(
+            out_f,
+            fieldnames=reader.fieldnames,
+            delimiter="\t",
+            quotechar=None,
+            doublequote=False,
+            lineterminator="\n",
+            quoting=csv.QUOTE_NONE,
+        )
+        writer.writeheader()
+        for sample in reader:
+            filtered += 1
+            if len(sample['src_text']) > 0:
+                ratio = float(len(sample['tgt_text'])) / len(sample['src_text'])
+                if threshold_min < ratio < threshold_max:
+                    writer.writerow(sample)
+                    filtered -= 1
+        logger.info(f"Filtered rows: {filtered}")
+
+
+def process_multiple_files(tsv_files, threshold_min, threshold_max, tsv_out_dir):
+    for tsv_in in tsv_files:
+        # Create output file path
+        filename = os.path.basename(tsv_in)
+        tsv_out = os.path.join(tsv_out_dir, filename)
+        filter_data(tsv_in, threshold_min, threshold_max, tsv_out)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Filter based on character ratio")
+    parser.add_argument("--tsv-datasets", required=True, nargs='+', help="List of TSV files to process")
+    parser.add_argument("--tsv-out-dir", required=True, help="Directory to save the filtered TSV files")
+    parser.add_argument("--threshold-min", required=True, type=float,
+                        help="The minimum ratio admitted. Everything with a lower ratio is filtered.")
+    parser.add_argument("--threshold-max", required=True, type=float,
+                        help="The maximum ratio admitted. Everything with a higher ratio is filtered.")
+
+    args = parser.parse_args()
+
+    # Ensure the output directory exists
+    os.makedirs(args.tsv_out_dir, exist_ok=True)
+
+    # Process multiple TSV files
+    process_multiple_files(args.tsv_datasets, args.threshold_min, args.threshold_max, args.tsv_out_dir)