From ca506e9a57d97077fb847d0f2b2eb4aed4b88c53 Mon Sep 17 00:00:00 2001
From: Sarthak Kapoor <sarthakkapoor@hotmail.com>
Date: Wed, 14 Aug 2024 17:12:47 +0200
Subject: [PATCH] Rename and add doc string for utils:
 generate_archive_from_json

---
 src/nomad_polymerization_reactions/utils.py   | 37 ++++++++++++++++++-
 .../processed_reactions/empty.archive.yaml    |  2 +
 tests/data/processed_reactions/empty.json     |  1 +
 tests/test_utils.py                           |  8 +++-
 4 files changed, 45 insertions(+), 3 deletions(-)
 create mode 100644 tests/data/processed_reactions/empty.archive.yaml
 create mode 100644 tests/data/processed_reactions/empty.json

diff --git a/src/nomad_polymerization_reactions/utils.py b/src/nomad_polymerization_reactions/utils.py
index 1ce7c40..764e70a 100644
--- a/src/nomad_polymerization_reactions/utils.py
+++ b/src/nomad_polymerization_reactions/utils.py
@@ -9,7 +9,42 @@
     from structlog.stdlib import BoundLogger
 
 
-def generate_archive_from_llm_output(filepath: str, logger: 'BoundLogger' = None):  # noqa: PLR0912
+def generate_archive_from_json(filepath: str, logger: 'BoundLogger' = None):  # noqa: PLR0912
+    """
+    Generate an archive.yaml file from a JSON file coming from the LLM output.
+    Function expects a JSON of the following format:
+    ```json
+    {
+        "file": "paper_0.json",
+        "monomer1_s": "C=C",
+        "monomer2_s": "C=O",
+        "monomer1": "ethylene",
+        "monomer2": "carbon monoxide",
+        "r_values": {
+            "constant_1": 22.0,
+            "constant_2": 0.0
+        },
+        "conf_intervals": {
+            "constant_conf_1": null,
+            "constant_conf_2": null
+        },
+        "temperature": 20.0,
+        "temperature_unit": "\u00b0C",
+        "solvent": null,
+        "method": "bulk",
+        "r-product": null,
+        "source": "https://doi.org/10.1002/pol.1963.110010415"
+    }
+    ```
+
+    Args:
+        filepath (str): Path to the JSON file.
+        logger (BoundLogger): A structlog logger.
+
+    Returns:
+        dict: The dict used to generate archive.yaml file.
+    """
+
     class OrderedDumper(yaml.Dumper):
         def represent_dict(self, data):
             return self.represent_mapping('tag:yaml.org,2002:map', data.items())
diff --git a/tests/data/processed_reactions/empty.archive.yaml b/tests/data/processed_reactions/empty.archive.yaml
new file mode 100644
index 0000000..5c7228e
--- /dev/null
+++ b/tests/data/processed_reactions/empty.archive.yaml
@@ -0,0 +1,2 @@
+data:
+  m_def: nomad_polymerization_reactions.schema_packages.mypackage.PolymerizationReaction
diff --git a/tests/data/processed_reactions/empty.json b/tests/data/processed_reactions/empty.json
new file mode 100644
index 0000000..9e26dfe
--- /dev/null
+++ b/tests/data/processed_reactions/empty.json
@@ -0,0 +1 @@
+{}
\ No newline at end of file
diff --git a/tests/test_utils.py b/tests/test_utils.py
index 0e0c1f7..e48b2a4 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -2,7 +2,7 @@
 
 import pytest
 import yaml
-from nomad_polymerization_reactions.utils import generate_archive_from_llm_output
+from nomad_polymerization_reactions.utils import generate_archive_from_json
 
 
 @pytest.mark.parametrize(
@@ -20,10 +20,14 @@
                 'tests/data/processed_reactions/paper_5_reaction_1.archive.yaml'
             ),
         },
+        {
+            'filepath': 'tests/data/processed_reactions/empty.json',
+            'reference': ('tests/data/processed_reactions/empty.archive.yaml'),
+        },
     ],
 )
 def test_generate_archive_from_llm_output(params):
-    output = generate_archive_from_llm_output(params['filepath'])
+    output = generate_archive_from_json(params['filepath'])
     with open(params['reference']) as f:
         reference = yaml.load(f, Loader=yaml.FullLoader)
     assert output == reference