Merge pull request #29 from YerevaNN/minor_changes_for_run

add try except for whole text formatting
YerevaNN · May 19, 2024 · 79bbd5d · 79bbd5d
2 parents 595e029 + b6ed503
commit 79bbd5d
Showing 1 changed file with 34 additions and 31 deletions.
diff --git a/chemlactica/utils/text_format_utils.py b/chemlactica/utils/text_format_utils.py
@@ -105,38 +105,41 @@ def generate_formatted_string(compound_json, rng, model_config):
 
 
 def format_key_value(key, value, rng):
-    if key == "CID":
+    try:
+        if key == "CID":
+            return ""
+        formatted_string = ""
+        if key == "related":
+            if len(value) > 10:
+                # value = random.sample(value, 5)
+                value = rng.choice(value, size=10, replace=False, shuffle=False)
+            for pair in value:
+                rounded_sim = "{:.2f}".format(float(pair["similarity"]))
+                formatted_string += f"{SPECIAL_TAGS['similarity']['start']}{pair['SMILES']} {rounded_sim}{SPECIAL_TAGS['similarity']['end']}"  # noqa
+        elif key == "experimental":
+            for pair in value:
+                formatted_string += f"[PROPERTY]{pair['PROPERTY_NAME']} {pair['PROPERTY_VALUE']}[/PROPERTY]"  # noqa
+        elif key == "synonyms":
+            for val in value:
+                formatted_string += f"{SPECIAL_TAGS['synonym']['start']}{val['name']}{SPECIAL_TAGS['synonym']['end']}"  # noqa
+        else:
+            try:
+                if SPECIAL_TAGS[key].get("type") is float:
+                    value = "{:.2f}".format(float(value))
+                    assert len(value.split(".")[-1]) == 2
+                start = SPECIAL_TAGS[key]["start"]
+                end = SPECIAL_TAGS[key]["end"]
+            except Exception as e:
+                print(e)
+                print("Failed to parse: ", key, value)
+                start = value = end = ""
+            return f"{start}{value}{end}"
+
+        return formatted_string
+    except Exception as e:
+        print(e)
+        print("Failed to parse: ", key, value)
         return ""
-    formatted_string = ""
-    if key == "related":
-        if len(value) > 10:
-            # value = random.sample(value, 5)
-            value = rng.choice(value, size=10, replace=False, shuffle=False)
-        for pair in value:
-            rounded_sim = "{:.2f}".format(float(pair["similarity"]))
-            formatted_string += f"{SPECIAL_TAGS['similarity']['start']}{pair['SMILES']} {rounded_sim}{SPECIAL_TAGS['similarity']['end']}"  # noqa
-    elif key == "experimental":
-        for pair in value:
-            formatted_string += (
-                f"[PROPERTY]{pair['PROPERTY_NAME']} {pair['PROPERTY_VALUE']}[/PROPERTY]"
-            )
-    elif key == "synonyms":
-        for val in value:
-            formatted_string += f"{SPECIAL_TAGS['synonym']['start']}{val['name']}{SPECIAL_TAGS['synonym']['end']}"  # noqa
-    else:
-        try:
-            if SPECIAL_TAGS[key].get("type") is float:
-                value = "{:.2f}".format(float(value))
-                assert len(value.split(".")[-1]) == 2
-            start = SPECIAL_TAGS[key]["start"]
-            end = SPECIAL_TAGS[key]["end"]
-        except Exception as e:
-            print(e)
-            print("Failed to parse: ", key, value)
-            start = value = end = ""
-        return f"{start}{value}{end}"
-
-    return formatted_string
 
 
 def main():