Update tests for font-size-based title extraction

Refactor test helpers to use get_text("dict") structure instead of
get_text("blocks"). Add tests for multi-span joining, short span
filtering, regex rejection, and multi-page title detection.

Author: amahuli03

server/api/views/uploadFile/test_title.py

@@ -4,60 +4,75 @@
 from . import title
 
 
+def make_page_dict(blocks):
+    """Helper to build a get_text("dict") return value from a simple list of blocks.
+    Each block is a list of (text, font_size) tuples representing spans.
+    """
+    dict_blocks = []
+    for spans in blocks:
+        dict_blocks.append({
+            "type": 0,
+            "lines": [{
+                "spans": [{"text": text, "size": size} for text, size in spans]
+            }]
+        })
+    return {"blocks": dict_blocks}
+
+
+def make_mock_doc(pages_data, metadata=None):
+    """Build a mock fitz.Document.
+    pages_data: list of block lists, one per page. Each block is a list of (text, size) tuples.
+    """
+    doc = MagicMock()
+    doc.metadata = metadata or {"title": None}
+    doc.__len__ = lambda self: len(pages_data)
+
+    mock_pages = []
+    for page_blocks in pages_data:
+        page = MagicMock()
+        page.get_text.return_value = make_page_dict(page_blocks)
+        mock_pages.append(page)
+
+    doc.__getitem__ = lambda self, idx: mock_pages[idx]
+    return doc
+
+
 class TestGenerateTitle(unittest.TestCase):
     def test_prefers_metadata_title_if_valid(self):
         doc = MagicMock()
         doc.metadata = {"title": "A Study Regarding The Efficacy of Drugs"}
         self.assertEqual(
             "A Study Regarding The Efficacy of Drugs", title.generate_title(doc))
 
-    def test_falls_back_to_first_page_text_if_metadata_title_is_empty(self):
-        doc = MagicMock()
-        doc.metadata = {"title": ""}
-        doc[0].get_text = MagicMock()
-
-        foo_block = [None] * 7
-        foo_block[4] = "foo"
-        foo_block[6] = 0
-
-        title_block = [None] * 7
-        title_block[4] = "Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia"
-        title_block[6] = 0
-
-        bar_block = [None] * 7
-        bar_block[4] = "bar"
-        bar_block[6] = 0
-        doc[0].get_text.return_value = [foo_block, title_block, bar_block]
-
+    def test_falls_back_to_font_size_if_metadata_title_is_empty(self):
+        doc = make_mock_doc(
+            pages_data=[[
+                [("foo", 10.0)],
+                [("Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia", 18.0)],
+                [("bar", 10.0)],
+            ]],
+            metadata={"title": ""},
+        )
         expected_title = "Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia"
         self.assertEqual(expected_title, title.generate_title(doc))
 
-    def test_falls_back_to_first_page_text_if_metadata_title_does_not_match_regex(self):
-        doc = MagicMock()
-        doc.metadata = {"title": "abcd1234"}
-        doc[0].get_text = MagicMock()
-
-        foo_block = [None] * 7
-        foo_block[4] = "foo"
-        foo_block[6] = 0
-
-        title_block = [None] * 7
-        title_block[4] = "Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia"
-        title_block[6] = 0
-
-        bar_block = [None] * 7
-        bar_block[4] = "bar"
-        bar_block[6] = 0
-        doc[0].get_text.return_value = [foo_block, title_block, bar_block]
-
+    def test_falls_back_to_font_size_if_metadata_title_does_not_match_regex(self):
+        doc = make_mock_doc(
+            pages_data=[[
+                [("foo", 10.0)],
+                [("Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia", 18.0)],
+                [("bar", 10.0)],
+            ]],
+            metadata={"title": "abcd1234"},
+        )
         expected_title = "Advances in Mood Disorder Pharmacotherapy: Evaluating New Antipsychotics and Mood Stabilizers for Bipolar Disorder and Schizophrenia"
         self.assertEqual(expected_title, title.generate_title(doc))
 
     @patch("api.views.uploadFile.title.openAIServices.openAI")
     def test_falls_back_to_chatgpt_if_no_title_found(self, mock_openAI):
-        doc = MagicMock()
-        doc.metadata = {"title": None}
-        doc[0].get_text.return_value = []
+        doc = make_mock_doc(
+            pages_data=[[]]  # no blocks at all
+        )
 
         mock_response = MagicMock()
         mock_response.choices = [MagicMock()]
@@ -70,9 +85,7 @@ def test_falls_back_to_chatgpt_if_no_title_found(self, mock_openAI):
 
     @patch("api.views.uploadFile.title.openAIServices.openAI")
     def test_strips_quotes_from_openai_title(self, mock_openAI):
-        doc = MagicMock()
-        doc.metadata = {"title": None}
-        doc[0].get_text.return_value = []
+        doc = make_mock_doc(pages_data=[[]])
 
         mock_response = MagicMock()
         mock_response.choices = [MagicMock()]
@@ -85,9 +98,7 @@ def test_strips_quotes_from_openai_title(self, mock_openAI):
 
     @patch("api.views.uploadFile.title.openAIServices.openAI")
     def test_truncates_long_openai_title(self, mock_openAI):
-        doc = MagicMock()
-        doc.metadata = {"title": None}
-        doc[0].get_text.return_value = []
+        doc = make_mock_doc(pages_data=[[]])
 
         mock_response = MagicMock()
         mock_response.choices = [MagicMock()]
@@ -98,3 +109,55 @@ def test_truncates_long_openai_title(self, mock_openAI):
 
         # Ensure the title is truncated to fit the UploadFile model's title field (max_length=255), since OpenAI responses may exceed this limit
         self.assertLessEqual(len(result), 255)
+
+    def test_font_size_joins_adjacent_spans_in_same_block(self):
+        """A title split across multiple spans in the same block should be joined."""
+        doc = make_mock_doc(
+            pages_data=[[
+                [("Author Name", 10.0)],
+                [("Advances in Mood Disorder", 18.0), ("Pharmacotherapy", 18.0)],
+                [("Some journal info", 10.0)],
+            ]],
+        )
+        result = title.extract_title_by_font_size(doc)
+        self.assertEqual(result, "Advances in Mood Disorder Pharmacotherapy")
+
+    def test_font_size_ignores_short_spans(self):
+        """Superscript markers and other tiny spans should be filtered out."""
+        doc = make_mock_doc(
+            pages_data=[[
+                [("Advances in Mood Disorder Pharmacotherapy", 18.0), ("*", 18.0)],
+                [("Author Name et al.", 10.0)],
+            ]],
+        )
+        # The "*" span is < 2 chars, so it should be ignored; title is just the real text
+        result = title.extract_title_by_font_size(doc)
+        self.assertEqual(result, "Advances in Mood Disorder Pharmacotherapy")
+
+    def test_font_size_returns_none_when_no_regex_match(self):
+        """If the largest-font text doesn't match the title regex, return None."""
+        doc = make_mock_doc(
+            pages_data=[[
+                # Only 2 words — regex requires at least 3
+                [("Psychiatry Research", 18.0)],
+                [("Author Name et al.", 10.0)],
+            ]],
+        )
+        result = title.extract_title_by_font_size(doc)
+        self.assertIsNone(result)
+
+    def test_font_size_finds_title_on_later_page(self):
+        """Title on page 2 should still be found if it has the largest font."""
+        doc = make_mock_doc(
+            pages_data=[
+                [  # page 1: cover page with smaller text
+                    [("Some preamble text here", 12.0)],
+                ],
+                [  # page 2: actual title in larger font
+                    [("Advances in Mood Disorder Pharmacotherapy", 18.0)],
+                    [("Author Name et al.", 10.0)],
+                ],
+            ],
+        )
+        result = title.extract_title_by_font_size(doc)
+        self.assertEqual(result, "Advances in Mood Disorder Pharmacotherapy")