Update pubmed25.py

pubmed25.py

@@ -12,8 +12,21 @@
 # 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.
-"""MEDLINE/PubMed data - Modified for full abstract text extraction."""
-
+"""
+MEDLINE/PubMed data loader for Hugging Face Datasets.
+
+This script is based on the original NCBI PubMed dataset script from Hugging Face.
+Modifications by:
+- Hoang Ha (LIG): Updated for 2025 baseline structure and integrated full abstract parsing.
+- Tiziri Terkmani (Research Engineer, LIG, Team SIGMA): Enhanced abstract parsing logic was adapted for the NanoBubble Project: https://nanobubbles.hypotheses.org/
+
+**Important Caution:** The default URL configuration in this script is set up as an
+example and primarily targets files from the **2025 baseline**. To download data
+covering the full intended range (e.g., 2015-2025), you **MUST** modify the `_URLs`
+list below to include the correct file paths for **ALL** desired years from the
+NCBI FTP server (ftp://ftp.ncbi.nlm.nih.gov/pubmed/baseline/). The number of files
+per year varies. Use with caution and verify the downloaded data range.
+"""
 
 import copy
 import gzip