Add checking start_line and raw value

sciunto-org · Nov 4, 2023 · e15129b · e15129b
1 parent d1b8d08
commit e15129b
Show file tree

Hide file tree

Showing 2 changed files with 51 additions and 31 deletions.
diff --git a/bibtexparser/splitter.py b/bibtexparser/splitter.py
@@ -395,7 +395,7 @@ def _handle_entry(self, m, m_val) -> Union[Entry, ParsingFailedBlock]:
             entry_type=entry_type,
             key=key,
             fields=fields,
-            raw=self.bibstr[m.start() : end_index + 1],
+            raw=self.bibstr[m.start() : end_index],
         )
 
         # If there were duplicate field keys, we return a DuplicateFieldKeyBlock wrapping

diff --git a/tests/splitter_tests/test_splitter_basic.py b/tests/splitter_tests/test_splitter_basic.py
@@ -321,57 +321,77 @@ def test_handles_duplicate_strings():
     )
 
 
-blocks_not_starting_on_new_lines = """
+blocks_not_starting_on_new_lines = """\
 @article{KUEH2023S98,
 title = {Myocardial Characterisation Using Delayed Dual-Energy Cardiac Computed Tomography},
-journal = {Heart, Lung and Circulation},
-volume = {32},
-pages = {S98},
-year = {2023},
-note = {Abstracts for the Cardiac Society of Australia and New Zealand Annual Scientific Meeting (New Zealand) 2023, 15 - 17 June 2023, Auckland, New Zealand},
-issn = {1443-9506},
-doi = {https://doi.org/10.1016/j.hlc.2023.04.262},
-url = {https://www.sciencedirect.com/science/article/pii/S1443950623004390},
 author = {S.-H. Kueh and J. Benatar and R. Stewart}
 }@INPROCEEDINGS{9837531,
   author={Hassan, Mona Bakri and Saeed, Rashid A. and Khalifa, Othman and Ali, Elmustafa Sayed and Mokhtar, Rania A. and Hashim, Aisha A.},
-  booktitle={2022 IEEE 2nd International Maghreb Meeting of the Conference on Sciences and Techniques of Automatic Control and Computer Engineering (MI-STA)},
   title={Green Machine Learning for Green Cloud Energy Efficiency},
-  year={2022},
-  volume={},
-  number={},
-  pages={288-294},
   doi={10.1109/MI-STA54861.2022.9837531}}@ARTICLE{9372936,
   author={Hu, Ning and Tian, Zhihong and Du, Xiaojiang and Guizani, Nadra and Zhu, Zhihan},
-  journal={IEEE Transactions on Green Communications and Networking},
   title={Deep-Green: A Dispersed Energy-Efficiency Computing Paradigm for Green Industrial IoT},
-  year={2021},
-  volume={5},
-  number={2},
-  pages={750-764},
   doi={10.1109/TGCN.2021.3064683}}@ARTICLE{5445167,
   author={Kumar, Karthik and Lu, Yung-Hsiang},
-  journal={Computer},
   title={Cloud Computing for Mobile Users: Can Offloading Computation Save Energy?},
-  year={2010},
-  volume={43},
-  number={4},
-  pages={51-56},
   doi={10.1109/MC.2010.98}}
 """
 
 
-def test_blocks_not_starting_on_new_lines():
+@pytest.mark.parametrize(
+    "expected",
+    [
+        {
+            "key": "KUEH2023S98",
+            "type": "article",
+            "raw": "@article{KUEH2023S98,\n"
+            "title = {Myocardial Characterisation Using Delayed Dual-Energy Cardiac Computed Tomography},\n"
+            "author = {S.-H. Kueh and J. Benatar and R. Stewart}\n}",
+            "start_line": 0,
+        },
+        {
+            "key": "9837531",
+            "type": "inproceedings",
+            "raw": "@INPROCEEDINGS{9837531,\n"
+            "  author={Hassan, Mona Bakri and Saeed, Rashid A. and Khalifa, Othman and Ali, Elmustafa Sayed and Mokhtar, Rania A. and Hashim, Aisha A.},\n"
+            "  title={Green Machine Learning for Green Cloud Energy Efficiency},\n"
+            "  doi={10.1109/MI-STA54861.2022.9837531}}",
+            "start_line": 3,
+        },
+        {
+            "key": "9372936",
+            "type": "article",
+            "raw": "@ARTICLE{9372936,\n"
+            "  author={Hu, Ning and Tian, Zhihong and Du, Xiaojiang and Guizani, Nadra and Zhu, Zhihan},\n"
+            "  title={Deep-Green: A Dispersed Energy-Efficiency Computing Paradigm for Green Industrial IoT},\n"
+            "  doi={10.1109/TGCN.2021.3064683}}",
+            "start_line": 6,
+        },
+        {
+            "key": "5445167",
+            "type": "article",
+            "raw": "@ARTICLE{5445167,\n"
+            "  author={Kumar, Karthik and Lu, Yung-Hsiang},\n"
+            "  title={Cloud Computing for Mobile Users: Can Offloading Computation Save Energy?},\n"
+            "  doi={10.1109/MC.2010.98}}",
+            "start_line": 9,
+        },
+    ],
+)
+def test_blocks_not_starting_on_new_lines(expected: Dict[str, Any]):
     """Test the new blocks that are not on new lines.
 
     Discussed at https://github.com/sciunto-org/python-bibtexparser/issues/411.
     """
     import bibtexparser
 
     lib = bibtexparser.parse_string(blocks_not_starting_on_new_lines)
-    assert len(lib.blocks) == 4
     assert len(lib.entries) == 4
-    assert lib.entries[0].entry_type == "article"
-    assert lib.entries[1].entry_type == "inproceedings"
-    assert lib.entries[2].entry_type == "article"
-    assert lib.entries[3].entry_type == "article"
+    entry_by_key = lib.entries_dict
+    assert len(entry_by_key) == 4
+
+    entry = entry_by_key[expected["key"]]
+    assert entry.key == expected["key"]
+    assert entry.entry_type == expected["type"]
+    assert entry.raw == expected["raw"]
+    assert entry.start_line == expected["start_line"]