trimming of long strims added to the pipeline

assemble.sh

@@ -148,7 +148,12 @@ java -Xmx4g -jar ${MAP_GENERATOR_DIR}/biohackathon/target/biohackathon-1.0-jar-w
 #java -Xmx4g -jar ${MAP_GENERATOR_DIR}/biohackathon/target/biohackathon-1.0-jar-with-dependencies.jar --enrichr-file ${enriched_paths_out_path} --minerva-file  ${enriched_maps_out_path} --output-file ${map_out_path}
 check_exit_code
 
-log "Pathwaygs assembled into ${map_out_path}"
+log "Pathways assembled into ${map_out_path}"
+
+log "Trimming long strings..."
+map_out_path_trimmed=${map_out_path/.xml/_trim.xml}
+$PYTHON_BIN ${MAP_GENERATOR_DIR}/utils/trim_long_strings.py -i ${map_out_path} > ${map_out_path_trimmed}
+log "Long string trimmed"
 
 log "Combining the map with overlays"
 
@@ -156,13 +161,13 @@ tmp_dir=${RES_DIR}/tmp/
 tmp_dir_layouts=${tmp_dir}/layouts/
 
 mkdir ${tmp_dir}
-cp ${map_out_path} ${tmp_dir}
+cp ${map_out_path_trimmed} ${tmp_dir}
 
 mkdir ${tmp_dir_layouts}
 cp ${minerva_genes_out_path} ${tmp_dir_layouts}
 cp ${minerva_variants_out_path} ${tmp_dir_layouts}
 
-map_zip_out_path=${map_out_path/.xml/.zip}
+map_zip_out_path=${map_out_path_trimmed/.xml/.zip}
 rm ${map_zip_out_path}
 cd ${tmp_dir}
 zip -r tmp.zip .

map_generator/utils/trim_long_strings.py

+import logging
+import argparse
+import xml.etree.ElementTree as ET
+
+
+def trim(path: str) -> str:
+    tree = ET.parse(path)
+    root = tree.getroot()
+
+    for elem in root.iter():
+        atts = elem.attrib
+        # if 'name' in atts:
+        #     print(elem.tag, atts)
+        for k, v in atts.items():
+            if len(v) > 250:
+                logging.warning("Trimming {} (key {} too long)".format(elem, k))
+                if "resource" in k:
+                    #  We want to keep the RDF resource still a valid resource otherwise MINERVA will fail during import
+                    atts[k] = atts[k].split(";")[0]
+                else:
+                    atts[k] = ""
+
+    return ET.tostring(root, encoding='utf8').decode('utf8')
+
+if __name__ == '__main__':
+
+    logging.basicConfig(
+        level=logging.INFO,
+        format='%(asctime)s [%(levelname)s] %(module)s - %(message)s',
+        datefmt='%H:%M:%S')
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("-i", "--input_xml",
+                        required=True,
+                        help="XML file with a map")
+    args = parser.parse_args()
+    print(trim(args.input_xml))
\ No newline at end of file