fix(pubmed): trata ausência de PubDate como erro em vez de gerar Journal inválido

packtools/sps/formats/pubmed.py

@@ -16,6 +16,19 @@
 )
 
 
+class MissingRequiredElementError(Exception):
+    """Raised when a SciELO article has no data for an element the PubMed
+    DTD marks as required (no `?`), so a valid <Article> cannot be built
+    for it. `element_path` identifies which element (e.g. "Journal/PubDate")."""
+
+    def __init__(self, element_path):
+        self.element_path = element_path
+        super().__init__(
+            f"{element_path} é obrigatório na DTD do PubMed e não foi "
+            "possível determinar seu valor para este artigo."
+        )
+
+
 def xml_pubmed_article_pipe():
     return ET.Element("Article")
 
@@ -127,20 +140,21 @@ def xml_pubmed_pub_date_pipe(xml_pubmed, xml_tree):
     </PubDate>
     """
     date = get_date(xml_tree)
-    if date is not None:
-        dt = ET.Element("PubDate")
-        dt.set("PubStatus", "epublish")
-        for element in ["year", "month", "day"]:
-            # TODO
-            # Season
-            # The season of publication. e.g.,Winter, Spring, Summer, Fall. Do not use if a Month is available.
-            # There is no example of using this value in the files.
-
-            if date.get(element):
-                el = ET.Element(element.capitalize())
-                el.text = date.get(element)
-                dt.append(el)
-        xml_pubmed.find("Journal").append(dt)
+    if date is None:
+        raise MissingRequiredElementError("Journal/PubDate")
+    dt = ET.Element("PubDate")
+    dt.set("PubStatus", "epublish")
+    for element in ["year", "month", "day"]:
+        # TODO
+        # Season
+        # The season of publication. e.g.,Winter, Spring, Summer, Fall. Do not use if a Month is available.
+        # There is no example of using this value in the files.
+
+        if date.get(element):
+            el = ET.Element(element.capitalize())
+            el.text = date.get(element)
+            dt.append(el)
+    xml_pubmed.find("Journal").append(dt)
 
 
 def xml_pubmed_replaces_pipe(xml_pubmed, xml_tree):

packtools/sps/formats/pubmed_generator.py

@@ -1,4 +1,5 @@
 import argparse
+import sys
 
 from packtools.sps.formats import pubmed
 from packtools.sps.utils import xml_utils
@@ -27,7 +28,12 @@ def main():
     arguments = parser.parse_args()
 
     xml_tree = xml_utils.get_xml_tree(arguments.path_to_read)
-    xml_pubmed = pubmed.pipeline_pubmed(xml_tree)
+    try:
+        xml_pubmed = pubmed.pipeline_pubmed(xml_tree)
+    except pubmed.MissingRequiredElementError as exc:
+        print(f"Erro: {exc}", file=sys.stderr)
+        raise SystemExit(1)
+
     with open(arguments.path_to_write, "w", encoding="utf-8") as file:
         file.write(xml_pubmed)
         print(f"Arquivo criado em: {arguments.path_to_write}")

tests/sps/formats/test_pubmed.py

@@ -25,6 +25,8 @@
     xml_pubmed_citations,
     xml_pubmed_abstract,
     xml_pubmed_other_abstract,
+    pipeline_pubmed,
+    MissingRequiredElementError,
 )
 
 
@@ -506,11 +508,6 @@ def test_xml_pubmed_pub_date_pipe_without_year(self):
         self.assertEqual(obtained, expected)
 
     def test_xml_pubmed_pub_date_pipe_without_date(self):
-        expected = (
-            '<Article>'
-            '<Journal/>'
-            '</Article>'
-        )
         xml_pubmed = ET.fromstring(
             '<Article>'
             '<Journal/>'
@@ -527,11 +524,8 @@ def test_xml_pubmed_pub_date_pipe_without_date(self):
             '</article>'
         )
 
-        xml_pubmed_pub_date_pipe(xml_pubmed, xml_tree)
-
-        obtained = ET.tostring(xml_pubmed, encoding="utf-8").decode("utf-8")
-
-        self.assertEqual(obtained, expected)
+        with self.assertRaises(MissingRequiredElementError):
+            xml_pubmed_pub_date_pipe(xml_pubmed, xml_tree)
 
     def test_xml_pubmed_article_title_pipe(self):
         expected = (
@@ -1828,6 +1822,43 @@ def test_xml_pubmed_other_abstract(self):
         self.assertEqual(obtained, expected)
 
 
+class PipelinePubmedRequiredPubDate(unittest.TestCase):
+    def test_pipeline_pubmed_raises_when_no_pub_date_is_found(self):
+        xml_tree = ET.fromstring(
+            '<article xmlns:mml="http://www.w3.org/1998/Math/MathML" '
+            'xmlns:xlink="http://www.w3.org/1999/xlink" '
+            'article-type="research-article" dtd-version="1.1" specific-use="sps-1.9" xml:lang="en">'
+            '<front>'
+            '<article-meta>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+        )
+
+        with self.assertRaises(MissingRequiredElementError):
+            pipeline_pubmed(xml_tree)
+
+    def test_pipeline_pubmed_does_not_raise_when_pub_date_is_found(self):
+        xml_tree = ET.fromstring(
+            '<article xmlns:mml="http://www.w3.org/1998/Math/MathML" '
+            'xmlns:xlink="http://www.w3.org/1999/xlink" '
+            'article-type="research-article" dtd-version="1.1" specific-use="sps-1.9" xml:lang="en">'
+            '<front>'
+            '<article-meta>'
+            '<pub-date publication-format="electronic" date-type="pub">'
+            '<day>06</day>'
+            '<month>01</month>'
+            '<year>2023</year>'
+            '</pub-date>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+        )
+
+        xml_pubmed = pipeline_pubmed(xml_tree)
+
+        self.assertIn("<PubDate", xml_pubmed)
+
 
 if __name__ == '__main__':
     unittest.main()