aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2021-11-03 19:36:09 -0700
committerBryan Newbold <bnewbold@archive.org>2021-11-03 19:36:09 -0700
commit7439ade08f185eaa6ac44b8ab509a8d3c047d574 (patch)
tree75b175b435b661655537cf6d50c2c88a56853696
parent9c1d72ea9dde74ab594795377e47b9b764adf60f (diff)
downloadgrobid_tei_xml-7439ade08f185eaa6ac44b8ab509a8d3c047d574.tar.gz
grobid_tei_xml-7439ade08f185eaa6ac44b8ab509a8d3c047d574.zip
add failing test for single editor tag, multiple persname
-rw-r--r--tests/files/citation/single_editor_multiple_persname.grobid070.tei.xml38
-rw-r--r--tests/test_parse.py20
2 files changed, 58 insertions, 0 deletions
diff --git a/tests/files/citation/single_editor_multiple_persname.grobid070.tei.xml b/tests/files/citation/single_editor_multiple_persname.grobid070.tei.xml
new file mode 100644
index 0000000..8c1aa1b
--- /dev/null
+++ b/tests/files/citation/single_editor_multiple_persname.grobid070.tei.xml
@@ -0,0 +1,38 @@
+<biblStruct >
+ <analytic>
+ <title level="a" type="main">Uterine cancer</title>
+ <author>
+ <persName>
+ <forename type="first">J</forename>
+ <forename type="middle">R</forename>
+ <surname>Lurain</surname>
+ </persName>
+ </author>
+ </analytic>
+ <monogr>
+ <title level="m">Novak’s gynecology</title>
+ <editor>
+ <persName>
+ <forename type="first">J</forename>
+ <forename type="middle">S</forename>
+ <surname>Berek</surname>
+ </persName>
+ <persName>
+ <forename type="first">E</forename>
+ <forename type="middle">Y</forename>
+ <surname>Adashi</surname>
+ </persName>
+ <persName>
+ <forename type="first">P</forename>
+ <forename type="middle">A</forename>
+ <surname>Hillard</surname>
+ </persName>
+ </editor>
+ <imprint>
+ <publisher>Williams and Wilkins</publisher>
+ <date type="published" when="1996">1996</date>
+ <biblScope unit="page" from="1057" to="1110" />
+ </imprint>
+ </monogr>
+ <note>12th ed. Baltimore</note>
+</biblStruct>
diff --git a/tests/test_parse.py b/tests/test_parse.py
index 5f64a48..832ee78 100644
--- a/tests/test_parse.py
+++ b/tests/test_parse.py
@@ -386,3 +386,23 @@ def test_citation_list_utf8() -> None:
assert ref_str.unstructured == unstructured
assert ref_str.first_page == "155"
assert ref_str.pages == "155-172"
+
+def test_citation_multiple_editors() -> None:
+ with open("tests/files/citation/single_editor_multiple_persname.grobid070.tei.xml", "r") as f:
+ tei_xml = f.read()
+
+ ref = parse_citation_xml(tei_xml)
+ assert ref
+ assert ref.title == "Uterine cancer"
+ assert len(ref.authors) == 1
+ assert ref.authors[0].full_name == "J R Lurain"
+ assert ref.authors[0].middle_name == "R"
+ assert ref.authors[0].surname == "Lurain"
+ assert len(ref.editors) == 3
+ assert ref.editors[0].full_name == "J S Berek"
+ assert ref.editors[1].full_name == "E Y Adashi"
+ assert ref.editors[2].full_name == "P A Hillard"
+ assert ref.journal == "Novak’s gynecology"
+ assert ref.publisher == "Williams and Wilkins"
+ assert ref.date == "1996"
+ assert ref.note == "12th ed. Baltimore"