From e9fddc930d67e6784459ad40605c52623d673b9a Mon Sep 17 00:00:00 2001
From: WilmerGaspar <wilmergasparespinoza@gmail.com>
Date: Wed, 17 Jun 2026 21:05:28 -0600
Subject: [PATCH 1/2] fix(metadata): encode multi-word Scopus queries

Encode main and special Scopus query terms after converting spaces to AND for multi-word keyword handling.
---
 .../metadata_extractor/fetch_metadata.py              | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/src/comproscanner/metadata_extractor/fetch_metadata.py b/src/comproscanner/metadata_extractor/fetch_metadata.py
index 6ca8f42b..a4f50ca3 100644
--- a/src/comproscanner/metadata_extractor/fetch_metadata.py
+++ b/src/comproscanner/metadata_extractor/fetch_metadata.py
@@ -115,8 +115,15 @@ def _construct_url(self, cursor, year, query, special_query):
         Returns:
             str: The constructed URL
         """
-        base = f"{self.base_url}PUBYEAR+%3D+{year}+{query}"
-        url = base + (f"+{special_query}" if special_query else "") + "&count=200"
+        encoded_query = urllib.parse.quote(query.replace(" ", " AND "))
+        encoded_special_query = (
+            urllib.parse.quote(special_query.replace(" ", " AND "))
+            if special_query
+            else ""
+        )
+
+        base = f"{self.base_url}PUBYEAR+%3D+{year}+{encoded_query}"
+        url = base + (f"+{encoded_special_query}" if encoded_special_query else "") + "&count=200"
         url += f"&cursor={cursor}"
         return url
 

From dd605f7a02f306e553b8073bfd007a1c46d0bbe3 Mon Sep 17 00:00:00 2001
From: WilmerGaspar <wilmergasparespinoza@gmail.com>
Date: Thu, 18 Jun 2026 22:21:43 -0600
Subject: [PATCH 2/2] fix(metadata): normalize multi-word keyword handling

---
 src/comproscanner/comproscanner.py                 | 14 ++++++++------
 .../metadata_extractor/fetch_metadata.py           | 13 +++++++------
 2 files changed, 15 insertions(+), 12 deletions(-)

diff --git a/src/comproscanner/comproscanner.py b/src/comproscanner/comproscanner.py
index d780e1ed..a2a85c7c 100644
--- a/src/comproscanner/comproscanner.py
+++ b/src/comproscanner/comproscanner.py
@@ -54,11 +54,13 @@
 
 class ComProScanner:
     def __init__(self, main_property_keyword: str = None):
-        self.main_property_keyword = main_property_keyword
-        if self.main_property_keyword is None:
-            raise ValueErrorHandler(
-                "Please provide a main property keyword to proceed."
-            )
+    if main_property_keyword is None:
+        raise ValueErrorHandler(
+            "Please provide a main property keyword to proceed."
+        )
+
+    self.main_property_keyword = main_property_keyword.replace(" ", "_")
+    self.main_property_search_keyword = self.main_property_keyword.replace("_", " ")
 
     def collect_metadata(
         self,
@@ -501,7 +503,7 @@ def extract_composition_property_data(
                 f"results/extracted_data/{self.main_property_keyword}/related_figures"
             )
         if materials_data_identifier_query is None:
-            materials_data_identifier_query = f"Is there any material chemical composition and corresponding {self.main_property_keyword} value mentioned in the paper? Give one word answer. Either yes or no."
+            materials_data_identifier_query = f"Is there any material chemical composition and corresponding {self.main_property_search_keyword} value mentioned in the paper? Give one word answer. Either yes or no."
         preparator = MatPropDataPreparator(
             main_property_keyword=self.main_property_keyword,
             main_extraction_keyword=main_extraction_keyword,
diff --git a/src/comproscanner/metadata_extractor/fetch_metadata.py b/src/comproscanner/metadata_extractor/fetch_metadata.py
index a4f50ca3..2248ba44 100644
--- a/src/comproscanner/metadata_extractor/fetch_metadata.py
+++ b/src/comproscanner/metadata_extractor/fetch_metadata.py
@@ -115,12 +115,13 @@ def _construct_url(self, cursor, year, query, special_query):
         Returns:
             str: The constructed URL
         """
-        encoded_query = urllib.parse.quote(query.replace(" ", " AND "))
-        encoded_special_query = (
-            urllib.parse.quote(special_query.replace(" ", " AND "))
-            if special_query
-            else ""
-        )
+        search_query = query.replace("_", " ")
+encoded_query = urllib.parse.quote(search_query.replace(" ", " AND "))
+encoded_special_query = (
+    urllib.parse.quote(special_query.replace("_", " ").replace(" ", " AND "))
+    if special_query
+    else ""
+)
 
         base = f"{self.base_url}PUBYEAR+%3D+{year}+{encoded_query}"
         url = base + (f"+{encoded_special_query}" if encoded_special_query else "") + "&count=200"