enable italicisation of portion

indigo/analysis/markup.py

@@ -16,7 +16,7 @@ class TextPatternMarker:
     """ Xpath for candidate text nodes that should be tested for matches. Must be defined by subclasses.
     """
 
-    ancestors = ['coverpage', 'preface', 'preamble', 'body', 'mainBody', 'conclusions']
+    ancestors = ['coverpage', 'preface', 'preamble', 'body', 'mainBody', 'portionBody', 'conclusions']
     """ Tags that the candidate_xpath should be run against.
     """
 

indigo_api/models/documents.py

@@ -548,6 +548,15 @@ def change_date(self, new_date, user, comment=None):
             self.expression_date = new_date
             self.save_with_revision(user, comment=comment)
 
+    def get_provision_xml(self, provision_eid):
+        provision_xml = self.doc.get_portion_element(provision_eid)
+        if not provision_xml:
+            return None
+        portion = StructuredDocument.for_document_type('portion')()
+        portion.frbr_uri = self.frbr_uri
+        portion.main_content.append(provision_xml)
+        return portion.main
+
     def update_provision_xml(self, provision_eid, provision_xml):
         xml = etree.fromstring(provision_xml)
         # portionBody will always have exactly one child

indigo_api/serializers.py

@@ -1,26 +1,24 @@
-import logging
 import os.path
-from itertools import groupby
-from typing import List
-
-from actstream.signals import action
 from collections import OrderedDict
 
-from lxml.etree import LxmlError
-
-from django.contrib.auth.models import User
+import logging
+import reversion
+from actstream.signals import action
+from allauth.account.utils import user_display
 from django.conf import settings
+from django.contrib.auth.models import User
+from lxml import etree
+from lxml.etree import LxmlError
 from rest_framework import serializers
-from rest_framework.reverse import reverse
 from rest_framework.exceptions import ValidationError
+from rest_framework.reverse import reverse
+from typing import List
+
 from cobalt import StructuredDocument, FrbrUri
 from cobalt.akn import AKN_NAMESPACES, DEFAULT_VERSION
-import reversion
-
 from indigo_api.models import Document, Attachment, Annotation, DocumentActivity, Work, Amendment, Language, \
     PublicationDocument, Task, Commencement
 from indigo_api.signals import document_published
-from allauth.account.utils import user_display
 
 log = logging.getLogger(__name__)
 
@@ -438,11 +436,60 @@ class DocumentAPISerializer(serializers.Serializer):
     """
     Helper to handle input documents for general document APIs
     """
-    document = DocumentSerializer(required=True)
+    xml = serializers.CharField()
+    language = serializers.CharField(min_length=3, max_length=3)
+    provision_eid = serializers.CharField(allow_blank=True)
 
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        self.fields['document'].instance = self.instance
+    def validate_xml(self, xml):
+        """ mostly copied from DocumentSerializer.validate()
+        """
+        try:
+            doctype = self.instance.work_uri.doctype if not self.initial_data.get('provision_eid') else 'portion'
+            doc = StructuredDocument.for_document_type(doctype)(xml)
+        except (LxmlError, ValueError) as e:
+            raise ValidationError("Invalid XML: %s" % str(e))
+        # ensure the correct namespace
+        if doc.namespace != AKN_NAMESPACES[DEFAULT_VERSION]:
+            raise ValidationError(
+                f"Document must have namespace {AKN_NAMESPACES[DEFAULT_VERSION]}, but it has {doc.namespace} instead.")
+        return xml
+
+    def update_document(self):
+        document = self.instance
+        # the language could have been changed but not yet saved during editing, which might affect which plugin is used
+        language_code = self.validated_data.get('language')
+        if document.language.code != language_code:
+            document.language = Language.for_code(language_code)
+        # update the content with updated but unsaved changes too
+        self.set_content()
+
+    def set_content(self):
+        document = self.instance
+        xml = self.validated_data.get('xml')
+        provision_eid = self.validated_data.get('provision_eid')
+        if not provision_eid:
+            # update the document's full XML with what's in the editor
+            document.content = xml
+        else:
+            if self.use_full_xml:
+                # we'll need the document's full XML for the analysis,
+                # but update the relevant provision with what's in the editor first
+                document.update_provision_xml(provision_eid, xml)
+            else:
+                # update the document to be a 'portion' and use only what's in the editor as the content
+                document.work.work_uri.doctype = 'portion'
+                document.content = xml
+
+    def updated_xml(self):
+        document = self.instance
+        provision_eid = self.validated_data.get('provision_eid')
+        if not provision_eid:
+            # return the document's full updated XML
+            return document.document_xml
+        # otherwise, return only the provision being edited (NOT including the outer akn tag)
+        # if we used the full XML for the analysis, grab only the appropriate provision as a portion
+        xml = document.get_provision_xml(provision_eid) if self.use_full_xml else document.doc.portion
+        return etree.tostring(xml, encoding='unicode')
 
 
 class NoopSerializer(object):

indigo_api/tests/fixtures.py

@@ -33,6 +33,41 @@
 </akomaNtoso>
 """
 
+PORTION_FIXTURE = """<akomaNtoso xmlns="http://docs.oasis-open.org/legaldocml/ns/akn/3.0">
+  <portion name="portion">
+    <meta>
+      <identification source="">
+        <FRBRWork>
+          <FRBRthis value="/akn/za/act/1900/1/!main"/>
+          <FRBRuri value="/akn/za/act/1900/1"/>
+          <FRBRalias value="Untitled"/>
+          <FRBRdate date="1900-01-01" name="Generation"/>
+          <FRBRauthor href="#council" as="#author"/>
+          <FRBRcountry value="za"/>
+        </FRBRWork>
+        <FRBRExpression>
+          <FRBRthis value="/akn/za/act/1900/1/eng@/!main"/>
+          <FRBRuri value="/akn/za/act/1900/1/eng@"/>
+          <FRBRdate date="1900-01-01" name="Generation"/>
+          <FRBRauthor href="#council" as="#author"/>
+          <FRBRlanguage language="eng"/>
+        </FRBRExpression>
+        <FRBRManifestation>
+          <FRBRthis value="/akn/za/act/1900/1/eng@/!main"/>
+          <FRBRuri value="/akn/za/act/1900/1/eng@"/>
+          <FRBRdate date="1900-01-01" name="Generation"/>
+          <FRBRauthor href="#council" as="#author"/>
+        </FRBRManifestation>
+      </identification>
+      <publication date="2005-07-24" name="Province of Western Cape: σπαθιοῦ Gazette" number="6277" showAs="Province of Western Cape: 😀 Provincial Gazette"/>
+    </meta>
+    <portionBody>
+      %s
+    </portionBody>
+  </portion>
+</akomaNtoso>
+"""
+
 BODY_FIXTURE = """
 <body>
   <section eId="sec_1">
@@ -130,6 +165,13 @@ def document_fixture(text=None, xml=None):
     return DOCUMENT_FIXTURE % xml
 
 
+def portion_fixture(text=None, xml=None):
+    if text:
+        xml = """<section eId="sec_1"><content><p>%s</p></content></section>""" % text
+
+    return PORTION_FIXTURE % xml
+
+
 def component_fixture(text=None, xml=None):
     if text:
         xml = """<section eId="sec_1"><content><p>%s</p></content></section>""" % text

indigo_api/tests/test_analysis_api.py

@@ -1,5 +1,3 @@
-# -*- coding: utf-8 -*-
-
 from nose.tools import *  # noqa
 from rest_framework.test import APITestCase
 
@@ -19,17 +17,12 @@
 
     def test_link_terms_no_input(self):
         response = self.client.post('/api/documents/1/analysis/link-terms', {
-            'document': {},
         })
         assert_equal(response.status_code, 400)
 
     def test_link_terms(self):
         response = self.client.post('/api/documents/1/analysis/link-terms', {
-            'document': {
-                'frbr_uri': '/za/act/1992/1',
-                'expression_date': '2001-01-01',
-                'language': 'eng',
-                'content': document_fixture(xml="""
+            'xml': document_fixture(xml="""
 <section id="section-1">
   <num>1.</num>
   <heading>Definitions and interpretation</heading>
@@ -64,37 +57,84 @@
     </content>
   </subsection>
 </section>
-""")
-            }
+"""),
+            'language': 'eng',
+            'provision_eid': ""
         })
         assert_equal(response.status_code, 200)
 
-        content = response.data['document']['content']
+        content = response.data['xml']
 
         assert_true(content.startswith('<akomaNtoso'))
         assert_in('<def ', content)
         assert_in('<TLCTerm ', content)
 
+    def test_link_terms_portion(self):
+        response = self.client.post('/api/documents/1/analysis/link-terms', {
+            'xml': portion_fixture(xml="""
+<section id="section-1">
+  <num>1.</num>
+  <heading>Definitions and interpretation</heading>
+  <subsection id="section-1.subsection-0">
+    <content>
+      <p>In these By-laws, any word or expression that has been defined in the National Road Traffic Act, 1996 (Act No. 93 of 1996) has that meaning and, unless the context otherwise indicates –</p>
+    </content>
+  </subsection>
+  <subsection id="section-1.subsection-1">
+    <content>
+      <blockList id="section-1.subsection-1.list1">
+        <listIntroduction>"authorised official" means –</listIntroduction>
+        <item id="section-1.subsection-1.list1.a">
+          <num>(a)</num>
+          <p>a member of the Johannesburg Metropolitan Police established in terms of section 64A of the South African Police Service Act, 1995 (Act No. 68 of 1995); or</p>
+        </item>
+        <item id="section-1.subsection-1.list1.b">
+          <num>(b)</num>
+          <p>any person or official authorised as such, in writing, by the Council;</p>
+        </item>
+      </blockList>
+    </content>
+  </subsection>
+  <subsection id="section-1.subsection-2">
+    <content>
+      <p>"backfill" means to replace the structural layers, including the base, sub-base, sudgrade and subgrade but excluding the surfacing, in a trench dug in, or other excavation of, a road reserve, and “backfilling” is construed accordingly;</p>
+    </content>
+  </subsection>
+  <subsection id="section-1.subsection-3">
+    <content>
+      <p>"these By-Laws" includes the schedules;</p>
+    </content>
+  </subsection>
+</section>
+"""),
+            'language': 'eng',
+            'provision_eid': "section-1"
+        })
+        assert_equal(response.status_code, 200)
+
+        content = response.data['xml']
+
+        assert_true(content.startswith('<portion '))
+        assert_in('<def ', content)
+        assert_in('<TLCTerm ', content)
+
     def test_link_terms_no_perms(self):
         self.client.logout()
         self.assertTrue(self.client.login(username='no-perms@example.com', password='password'))
         user = User.objects.get(username='no-perms@example.com')
 
         data = {
-            'document': {
-                'frbr_uri': '/za/act/1992/1',
-                'expression_date': '2001-01-01',
-                'language': 'eng',
-                'content': document_fixture(xml="""
+            'xml': document_fixture(xml="""
 <section id="section-1">
   <num>1.</num>
   <heading>Definitions and interpretation</heading>
   <content>
     <p>test</p>
   </content>
 </section>
-""")
-            }
+"""),
+            'language': 'eng',
+            'provision_eid': ""
         }
 
         # user doesn't have perms