From 9bcb65be292bb7aaff5ae2286b80c55aedc35f0b Mon Sep 17 00:00:00 2001
From: Samuel Veiga Rangel <samuelveigarangel@hotmail.com>
Date: Thu, 1 Aug 2024 02:42:38 -0300
Subject: [PATCH 1/5] refactor xml_pmc_aff for aff elements

---
 packtools/sps/formats/pmc.py | 27 +++++++++++++++++++++++----
 1 file changed, 23 insertions(+), 4 deletions(-)

diff --git a/packtools/sps/formats/pmc.py b/packtools/sps/formats/pmc.py
index 7a1374cf1..ccd67b024 100644
--- a/packtools/sps/formats/pmc.py
+++ b/packtools/sps/formats/pmc.py
@@ -43,14 +43,29 @@ def xml_pmc_aff(xml_tree):
     """
     affs = xml_tree.findall(".//aff")
     for aff in affs:
-        aff_institution = aff.find("./institution[@content-type='original']").text
+        original_institution = aff.find("./institution[@content-type='original']")
+        if original_institution is not None:
+            aff_institution = original_institution.text
+        else:
+            aff_with_address = []
+            aff_with_address.append(aff.find("./institution[@content-type='orgname']").text)
+            
+            addr_line = aff.find("./addr-line")
+            if addr_line is not None:
+                named_contents = addr_line.findall(".//named-content")
+                aff_with_address.extend([named_content.text for named_content in named_contents])
+            
+            country = aff.find("./country")
+            if country is not None:
+                aff_with_address.append(country.text)
+            aff_institution = ", ".join(aff_with_address)
+
 
         for institution in aff.findall(".//institution"):
             aff.remove(institution)
 
-        aff.remove(aff.find("./addr-line"))
-
-        aff.remove(aff.find("./country"))
+        for element in [aff.find("./addr-line"), aff.find("./country")]:
+            aff.remove(element)
 
         node_label = aff.find("./label")
 
@@ -154,3 +169,7 @@ def xml_pmc_ref(xml_tree):
     refs = xml_tree.findall(".//ref")
     for ref in refs:
         ref.remove(ref.find("./mixed-citation"))
+
+
+def xml_pmc_refactor_pub_date(xmltree):
+    ...
\ No newline at end of file

From df5ccaa826554ca693f26961fe2282e77ed9e8dd Mon Sep 17 00:00:00 2001
From: Samuel Veiga Rangel <samuelveigarangel@hotmail.com>
Date: Fri, 9 Aug 2024 15:04:01 -0300
Subject: [PATCH 2/5] remove funcao xml_pmc_refactor_pub_date

---
 packtools/sps/formats/pmc.py | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/packtools/sps/formats/pmc.py b/packtools/sps/formats/pmc.py
index ccd67b024..9391789ee 100644
--- a/packtools/sps/formats/pmc.py
+++ b/packtools/sps/formats/pmc.py
@@ -170,6 +170,3 @@ def xml_pmc_ref(xml_tree):
     for ref in refs:
         ref.remove(ref.find("./mixed-citation"))
 
-
-def xml_pmc_refactor_pub_date(xmltree):
-    ...
\ No newline at end of file

From 16152ecba13a9f38537d7bd4c7f73a925eda99d8 Mon Sep 17 00:00:00 2001
From: Samuel Veiga Rangel <samuelveigarangel@hotmail.com>
Date: Fri, 9 Aug 2024 15:04:13 -0300
Subject: [PATCH 3/5] Adiciona teste test_xml_pmc_without_original

---
 tests/sps/formats/test_pmc.py | 71 +++++++++++++++++++++++++++++++++++
 1 file changed, 71 insertions(+)

diff --git a/tests/sps/formats/test_pmc.py b/tests/sps/formats/test_pmc.py
index 2db4c7167..c6e48cd93 100644
--- a/tests/sps/formats/test_pmc.py
+++ b/tests/sps/formats/test_pmc.py
@@ -492,6 +492,77 @@ def test_xml_pmc_ref(self):
 
         self.assertEqual(obtained, expected)
 
+    def test_xml_pmc_without_original(self):
+        self.maxDiff = None
+        expected = (
+            '<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-id specific-use="scielo-v3" pub-id-type="publisher-id">ZwzqmpTpbhTmtwR9GfDzP7c</article-id>'
+            '<article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0080-62342022000100445</article-id>'
+            '<article-id pub-id-type="doi">10.1590/1980-220X-REEUSP-2021-0569en</article-id>'
+            '<article-id pub-id-type="other">00445</article-id>'
+            '<contrib-group>'
+            '<contrib contrib-type="author">'
+            '<contrib-id contrib-id-type="orcid">0000-0003-0843-6485</contrib-id>'
+            '<name>'
+            '<surname>Boni</surname>'
+            '<given-names>Fernanda Guarilha</given-names>'
+            '</name>'
+            '<xref ref-type="aff" rid="aff1">'
+            '<sup>1</sup>'
+            '</xref>'
+            '</contrib>'
+            '</contrib-group>'
+            '<aff id="aff1">'
+            'Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil'
+            '</aff>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+
+        )
+        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-id specific-use="scielo-v3" pub-id-type="publisher-id">ZwzqmpTpbhTmtwR9GfDzP7c</article-id>'
+            '<article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0080-62342022000100445</article-id>'
+            '<article-id pub-id-type="doi">10.1590/1980-220X-REEUSP-2021-0569en</article-id>'
+            '<article-id pub-id-type="other">00445</article-id>'
+            '<contrib-group>'
+            '<contrib contrib-type="author">'
+            '<contrib-id contrib-id-type="orcid">0000-0003-0843-6485</contrib-id>'
+            '<name>'
+            '<surname>Boni</surname>'
+            '<given-names>Fernanda Guarilha</given-names>'
+            '</name>'
+            '<xref ref-type="aff" rid="aff1">'
+            '<sup>1</sup>'
+            '</xref>'
+            '</contrib>'
+            '</contrib-group>'
+            '<aff id="aff1">'
+            '<institution content-type="orgname">Universidade Federal do Rio Grande do Sul</institution>'
+            '<institution content-type="orgdiv1">Escola de Enfermagem</institution>'
+            '<institution content-type="orgdiv2">Programa de Pós-Graduação em Enfermagem</institution>'
+            '<addr-line>'
+            '<named-content content-type="city">Porto Alegre</named-content>'
+            '<named-content content-type="state">RS</named-content>'
+            '</addr-line>'
+            '<country country="BR">Brazil</country>'
+            '</aff>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+        )        
+
+        xml_pmc_aff(xml_tree)
+
+        obtained = ET.tostring(xml_tree, encoding="utf-8").decode("utf-8")
+
+        self.assertEqual(obtained, expected)
+
 
 if __name__ == '__main__':
     unittest.main()

From 3b70df745c2ea482412cf82a2facbdca0bfae357 Mon Sep 17 00:00:00 2001
From: Samuel Veiga Rangel <samuelveigarangel@hotmail.com>
Date: Sun, 11 Aug 2024 21:06:07 -0300
Subject: [PATCH 4/5] Substitui findall por xpath para named-content, state e
 city

---
 packtools/sps/formats/pmc.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/packtools/sps/formats/pmc.py b/packtools/sps/formats/pmc.py
index 9391789ee..a7fb4701d 100644
--- a/packtools/sps/formats/pmc.py
+++ b/packtools/sps/formats/pmc.py
@@ -52,7 +52,7 @@ def xml_pmc_aff(xml_tree):
             
             addr_line = aff.find("./addr-line")
             if addr_line is not None:
-                named_contents = addr_line.findall(".//named-content")
+                named_contents = addr_line.xpath(".//named-content | .//state | .//city ")
                 aff_with_address.extend([named_content.text for named_content in named_contents])
             
             country = aff.find("./country")

From 1b8a4a673422b7c5b84688393daea6bbff037fd1 Mon Sep 17 00:00:00 2001
From: Samuel Veiga Rangel <samuelveigarangel@hotmail.com>
Date: Sun, 11 Aug 2024 21:06:43 -0300
Subject: [PATCH 5/5] Adiciona teste para xml com city e satate

---
 tests/sps/formats/test_pmc.py | 71 +++++++++++++++++++++++++++++++++++
 1 file changed, 71 insertions(+)

diff --git a/tests/sps/formats/test_pmc.py b/tests/sps/formats/test_pmc.py
index c6e48cd93..8c86b182c 100644
--- a/tests/sps/formats/test_pmc.py
+++ b/tests/sps/formats/test_pmc.py
@@ -563,6 +563,77 @@ def test_xml_pmc_without_original(self):
 
         self.assertEqual(obtained, expected)
 
+    def test_xml_pmc_with_state_and_city(self):
+        self.maxDiff = None
+        expected = (
+            '<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-id specific-use="scielo-v3" pub-id-type="publisher-id">ZwzqmpTpbhTmtwR9GfDzP7c</article-id>'
+            '<article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0080-62342022000100445</article-id>'
+            '<article-id pub-id-type="doi">10.1590/1980-220X-REEUSP-2021-0569en</article-id>'
+            '<article-id pub-id-type="other">00445</article-id>'
+            '<contrib-group>'
+            '<contrib contrib-type="author">'
+            '<contrib-id contrib-id-type="orcid">0000-0003-0843-6485</contrib-id>'
+            '<name>'
+            '<surname>Boni</surname>'
+            '<given-names>Fernanda Guarilha</given-names>'
+            '</name>'
+            '<xref ref-type="aff" rid="aff1">'
+            '<sup>1</sup>'
+            '</xref>'
+            '</contrib>'
+            '</contrib-group>'
+            '<aff id="aff1">'
+            'Fundação Oswaldo Cruz, Manguinhos, RJ, Brasil'
+            '</aff>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+
+        )
+        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-id specific-use="scielo-v3" pub-id-type="publisher-id">ZwzqmpTpbhTmtwR9GfDzP7c</article-id>'
+            '<article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0080-62342022000100445</article-id>'
+            '<article-id pub-id-type="doi">10.1590/1980-220X-REEUSP-2021-0569en</article-id>'
+            '<article-id pub-id-type="other">00445</article-id>'
+            '<contrib-group>'
+            '<contrib contrib-type="author">'
+            '<contrib-id contrib-id-type="orcid">0000-0003-0843-6485</contrib-id>'
+            '<name>'
+            '<surname>Boni</surname>'
+            '<given-names>Fernanda Guarilha</given-names>'
+            '</name>'
+            '<xref ref-type="aff" rid="aff1">'
+            '<sup>1</sup>'
+            '</xref>'
+            '</contrib>'
+            '</contrib-group>'
+            '<aff id="aff1">'
+            '<institution content-type="orgname">Fundação Oswaldo Cruz</institution>'
+            '<institution content-type="orgdiv1">Escola Nacional de Saúde Pública Sérgio Arouca</institution>'
+            '<institution content-type="orgdiv2">Centro de Estudos da Saúde do Trabalhador e Ecologia Humana</institution>'
+            '<addr-line>'
+            '<city>Manguinhos</city>'
+            '<state>RJ</state>'
+            '</addr-line>'
+            '<country country="BR">Brasil</country>'
+            '</aff>'
+            '</article-meta>'
+            '</front>'
+            '</article>'
+        )        
+
+        xml_pmc_aff(xml_tree)
+
+        obtained = ET.tostring(xml_tree, encoding="utf-8").decode("utf-8")
+
+        self.assertEqual(obtained, expected)
+
 
 if __name__ == '__main__':
     unittest.main()