Merge pull request #138 from Capitains/issue-137

Issue 137

CHANGES.md

@@ -1,3 +1,7 @@
+### 2017-08-21 2.0.4 @ponteineptique
+
+- Issue 137 : Fixed a bug where a passage extraction would not work even if the passage was found in the reff extraction. Origin of the issue was replacing too much .// in the xpath and thus breaking it
+
 ### 2017-06-16 2.0.3 @ponteineptique
 
 - Issue 135 : Added support for empty namespace in string expansion for structured metadata"

MyCapytain/__init__.py

@@ -9,4 +9,4 @@
 
 """
 
-__version__ = "2.0.3"
+__version__ = "2.0.4"

MyCapytain/common/utils.py

@@ -236,7 +236,7 @@ def performXpath(parent, xpath):
     loop = False
     if xpath.startswith(".//"):
         result = parent.xpath(
-            xpath.replace(".//", "./"),
+            xpath.replace(".//", "./", 1),
             namespaces=XPATH_NAMESPACES
         )
         if len(result) == 0:

tests/resources/texts/local/commonTests.py

@@ -245,7 +245,7 @@ def test_xml_with_xml_id(self):
         self.assertIn(
             '<l n="1" xml:id="C_l_1"><q><w lemma="ce2"',
             text.getTextualNode(subreference="1.C_w_000001").export(Mimetypes.XML.TEI),
-            "Word chould be there !"
+            "Word should be there !"
         )
         self.assertEqual(
             text.getReffs(level=2), [

tests/test_ad_hoc.py

@@ -0,0 +1,29 @@
+from unittest import TestCase
+from MyCapytain.resources.texts.local.capitains.cts import CapitainsCtsText
+from MyCapytain.common.constants import Mimetypes
+from lxml import etree
+
+
+class TestAdHoc(TestCase):
+    """ This is a series of tests drawn from issues on other repositories and third parties use of the code.
+
+    """
+    def test_passage_extraction_fail_when_reffs_are_found(self):
+        """ This issues is drawn from https://github.com/PerseusDL/canonical-latinLit/issues/226
+        """
+        with open("tests/testing_data/texts/extraction_issue.xml") as text:
+            interactive_text = CapitainsCtsText(resource=etree.parse(text).getroot())
+            reffs = interactive_text.getReffs(level=len(interactive_text.citation))
+            passages = []
+            # The failing passage was 5.1
+            for reff in reffs:
+                try:
+                    passages.append(interactive_text.getTextualNode(reff))
+                except IndexError:
+                    raise Exception("Unable to extract %s " % reff)
+
+            plaintext = [r.export(Mimetypes.PLAINTEXT, exclude=["tei:note"]).strip() for r in passages]
+            self.assertIn(
+                "NUNC et praedictos et regni sorte sequentes", plaintext,
+                "The text of 5.1 should be in plaintext"
+            )