Merge pull request #128 from Shoobx/Tomp0801-pfitzinger-xmlformatter-…

…replacing

use_replace tag

.gitignore

@@ -11,4 +11,6 @@ build
 docs/build
 venv
 build
-dist
+dist
+test.ipynb
+.venv/

.pre-commit-config.yaml

@@ -1,18 +1,18 @@
 repos:
 -   repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.4.0
+    rev: v4.5.0
     hooks:
     -   id: check-yaml
     -   id: trailing-whitespace
 -   repo: https://github.com/psf/black
-    rev: 22.12.0
+    rev: 23.12.1
     hooks:
     -   id: black
 -   repo: https://github.com/pycqa/flake8
-    rev: 6.0.0
+    rev: 6.1.0
     hooks:
     -   id: flake8
 -   repo: https://github.com/regebro/pyroma
-    rev: '4.1'
+    rev: '4.2'
     hooks:
     -   id: pyroma

CHANGES.rst

@@ -9,6 +9,11 @@ Changes
 
 - Changed usage of deprecated `pkg_resources` package to `importlib.metadata`.
 
+- A `use_replace` flag was added to the `XMLFormatter` by Thomas Pfitzinger.
+  It changes text replacement from delete and insert tags to a replace tag.
+  It's not currently accessaible thtough the CLI, the question is it is better
+  to add a new formatter name, or an option to pass in formatter flags.
+
 
 2.6.3 (2023-05-21)
 ------------------

CHANGES.txt

@@ -0,0 +1,7 @@
+- Added option to XMLFormatter to use replace tags
+
+- in _make_diff_tags after diffing, neighboring delete/insert diffs are joined to a replace tag
+
+- the deleted text is added as an attribute ("old-text")
+
+- the inserted text is the element's text

README.rst

@@ -90,5 +90,8 @@ Contributors
 
  * Jacek Chałupka, [email protected]
 
-The diff algorithm is based on "`Change Detection in Hierarchically Structured Information <http://infolab.stanford.edu/c3/papers/html/tdiff3-8/tdiff3-8.html>`_",
+ * Thomas Pfitzinger, [email protected]
+
+The diff algorithm is based on
+"`Change Detection in Hierarchically Structured Information <http://infolab.stanford.edu/c3/papers/html/tdiff3-8/tdiff3-8.html>`_",
 and the text diff is using Google's ``diff_match_patch`` algorithm.

tests/test_diff.py

@@ -640,7 +640,6 @@ def test_with_xmlid(self):
         )
 
     def test_change_attribs(self):
-
         left = """<document>
     <story firstPageTemplate="FirstPage">
         <section xml:id="oldfirst" ref="3" single-ref="3">

tests/test_formatting.py

@@ -16,13 +16,13 @@ def test_get_placeholder(self):
         replacer = formatting.PlaceholderMaker()
         # Get a placeholder:
         ph = replacer.get_placeholder(etree.Element("tag"), formatting.T_OPEN, None)
-        self.assertEqual(ph, "\ue005")
+        self.assertEqual(ph, "\ue007")
         # Do it again:
         ph = replacer.get_placeholder(etree.Element("tag"), formatting.T_OPEN, None)
-        self.assertEqual(ph, "\ue005")
+        self.assertEqual(ph, "\ue007")
         # Get another one
         ph = replacer.get_placeholder(etree.Element("tag"), formatting.T_CLOSE, ph)
-        self.assertEqual(ph, "\ue006")
+        self.assertEqual(ph, "\ue008")
 
     def test_do_element(self):
         replacer = formatting.PlaceholderMaker(["p"], ["b"])
@@ -34,7 +34,7 @@ def test_do_element(self):
 
         self.assertEqual(
             etree.tounicode(element),
-            "<p>This is a tag with \ue006formatted\ue005 text.</p>",
+            "<p>This is a tag with \ue008formatted\ue007 text.</p>",
         )
 
         replacer.undo_element(element)
@@ -45,14 +45,14 @@ def test_do_element(self):
         element = etree.fromstring(text)
         replacer.do_element(element)
         result = etree.tounicode(element)
-        self.assertEqual(result, "<p>This is a tag with \ue007 text.</p>")
+        self.assertEqual(result, "<p>This is a tag with \ue009 text.</p>")
 
         # Single formatting tags still get two placeholders.
         text = "<p>This is a <b/> with <foo/> text.</p>"
         element = etree.fromstring(text)
         replacer.do_element(element)
         result = etree.tounicode(element)
-        self.assertEqual(result, "<p>This is a \ue009\ue008 with \ue00a text.</p>")
+        self.assertEqual(result, "<p>This is a \ue00b\ue00a with \ue00c text.</p>")
 
     def test_do_undo_element(self):
         replacer = formatting.PlaceholderMaker(["p"], ["b"])
@@ -63,7 +63,7 @@ def test_do_undo_element(self):
         replacer.do_element(element)
 
         self.assertEqual(
-            element.text, "This \ue005 a \ue006 with \ue008formatted" "\ue007 text."
+            element.text, "This \ue007 a \ue008 with \ue00aformatted" "\ue009 text."
         )
 
         replacer.undo_element(element)
@@ -79,7 +79,7 @@ def test_do_undo_element_double_format(self):
         replacer.do_element(element)
 
         self.assertEqual(
-            element.text, "This is \ue006doubly \ue008formatted\ue007" "\ue005 text."
+            element.text, "This is \ue008doubly \ue00aformatted\ue009" "\ue007 text."
         )
 
         replacer.undo_element(element)
@@ -110,17 +110,17 @@ def test_rml_bug(self):
         after_diff = """<document xmlns:diff="http://namespaces.shoobx.com/diff">
   <section>
     <para>
-      <insert>\ue005</insert>.
-      \ue007\ue009At Will Employment\ue008\ue006
+      <insert>\ue007</insert>.
+      \ue009\ue00bAt Will Employment\ue00a\ue008
       .\u201c<insert>New </insert>Text\u201d
     </para>
   </section>
 </document>"""
 
         # The diff formatting will find some text to insert.
         delete_attrib = "{%s}delete-format" % formatting.DIFF_NS
-        replacer.placeholder2tag["\ue006"].element.attrib[delete_attrib] = ""
-        replacer.placeholder2tag["\ue007"].element.attrib[delete_attrib] = ""
+        replacer.placeholder2tag["\ue008"].element.attrib[delete_attrib] = ""
+        replacer.placeholder2tag["\ue009"].element.attrib[delete_attrib] = ""
         tree = etree.fromstring(after_diff)
         replacer.undo_tree(tree)
         result = etree.tounicode(tree)
@@ -153,7 +153,7 @@ def test_placeholder_overflow(self):
 
             #
             self.assertEqual(
-                element.text, "This \uf904 a \uf905 with \uf907some" "\uf906 text."
+                element.text, "This \uf906 a \uf907 with \uf909some" "\uf908 text."
             )
 
             try:
@@ -172,8 +172,8 @@ def test_placeholder_overflow(self):
                 # This should raise an error on a narrow build
                 self.assertEqual(
                     element.text,
-                    "This \U00010004 a \U00010005 with \U00010007some"
-                    "\U00010006 text.",
+                    "This \U00010006 a \U00010007 with \U00010009some"
+                    "\U00010008 text.",
                 )
             except ValueError:
                 if sys.maxunicode > 0x10000:
@@ -186,8 +186,8 @@ def test_placeholder_overflow(self):
 
 
 class XMLFormatTests(unittest.TestCase):
-    def _format_test(self, left, action, expected):
-        formatter = formatting.XMLFormatter(pretty_print=False)
+    def _format_test(self, left, action, expected, use_replace=False):
+        formatter = formatting.XMLFormatter(pretty_print=False, use_replace=use_replace)
         result = formatter.format([action], etree.fromstring(left))
         self.assertEqual(result, expected)
 
@@ -317,6 +317,51 @@ def test_update_text_after_2(self):
 
         self._format_test(left, action, expected)
 
+    def test_replace_text_in(self):
+        left = '<document><node attr="val"/></document>'
+        action = actions.UpdateTextIn("/document/node", "Text")
+        expected = START + ' attr="val"><diff:insert>Text</diff:insert>' + END
+
+        self._format_test(left, action, expected, use_replace=True)
+
+        left = "<document><node>This is a bit of text, right" + END
+        action = actions.UpdateTextIn("/document/node", "Also a bit of text, rick")
+        expected = (
+            START + '><diff:replace old-text="This is">Also</diff:replace>'
+            ' a bit of text, ri<diff:replace old-text="ght">ck'
+            "</diff:replace>" + END
+        )
+
+        self._format_test(left, action, expected, use_replace=True)
+
+    def test_replace_text_after_1(self):
+        left = "<document><node/><node/></document>"
+        action = actions.UpdateTextAfter("/document/node[1]", "Text")
+        expected = START + "/><diff:insert>Text</diff:insert>" "<node/></document>"
+
+        self._format_test(left, action, expected, use_replace=True)
+
+    def test_replace_text_after_2(self):
+        left = "<document><node/>This is a bit of text, right</document>"
+        action = actions.UpdateTextAfter("/document/node", "Also a bit of text, rick")
+        expected = (
+            START + '/><diff:replace old-text="This is">Also</diff:replace>'
+            ' a bit of text, ri<diff:replace old-text="ght">ck'
+            "</diff:replace></document>"
+        )
+
+        self._format_test(left, action, expected, use_replace=True)
+
+    def test_replace_complete_text(self):
+        left = "<document><node>aaaaaaa bbbbbb</node></document>"
+        action = actions.UpdateTextIn("/document/node", "ccccc dddd eee")
+        expected = (
+            START + '><diff:replace old-text="aaaaaaa bbbbbb">ccccc dddd eee'
+            "</diff:replace>" + END
+        )
+
+        self._format_test(left, action, expected, use_replace=True)
+
 
 class DiffFormatTests(unittest.TestCase):
     def _format_test(self, action, expected):
@@ -515,7 +560,6 @@ def test_all_actions(self):
 
 
 class FormatterFileTests(unittest.TestCase):
-
     formatter = None  # Override this
     maxDiff = None
 
@@ -524,7 +568,6 @@ def process(self, left, right):
 
 
 class XMLFormatterFileTests(FormatterFileTests):
-
     # The XMLFormatter has no text or formatting tags, so
     formatter = formatting.XMLFormatter(
         pretty_print=False, normalize=formatting.WS_TEXT
@@ -535,7 +578,6 @@ class XMLFormatterFileTests(FormatterFileTests):
 
 
 class HTMLFormatterFileTests(FormatterFileTests):
-
     # We use a few tags for the placeholder tests.
     # <br/> is intentionally left out, to test an edge case
     # with empty non-formatting tags in text.

tests/test_patch.py

@@ -23,7 +23,6 @@
 
 
 class PatcherTests(unittest.TestCase):
-
     patcher = Patcher()
 
     def _test(self, start, action, end):

tests/test_utils.py

@@ -111,7 +111,6 @@ def _diff(self, left, right, result):
         self.assertEqual("".join(res), result)
 
     def test_lcs(self):
-
         self._diff("ABCDEF", "ABCDEF", "ABCDEF")
 
         self._diff("ABCDEF", "GHIJKL", "")

xmldiff/diff_match_patch.py

@@ -74,6 +74,7 @@ def __init__(self):
     DIFF_DELETE = -1
     DIFF_INSERT = 1
     DIFF_EQUAL = 0
+    DIFF_REPLACE = 2
 
     def diff_main(self, text1, text2, checklines=True, deadline=None):
         """Find the differences between two texts.  Simplifies the problem by
@@ -1135,7 +1136,7 @@ def diff_prettyHtml(self, diffs):
           HTML representation.
         """
         html = []
-        for (op, data) in diffs:
+        for op, data in diffs:
             text = (
                 data.replace("&", "&")
                 .replace("<", "&lt;")
@@ -1160,7 +1161,7 @@ def diff_text1(self, diffs):
           Source text.
         """
         text = []
-        for (op, data) in diffs:
+        for op, data in diffs:
             if op != self.DIFF_INSERT:
                 text.append(data)
         return "".join(text)
@@ -1175,7 +1176,7 @@ def diff_text2(self, diffs):
           Destination text.
         """
         text = []
-        for (op, data) in diffs:
+        for op, data in diffs:
             if op != self.DIFF_DELETE:
                 text.append(data)
         return "".join(text)
@@ -1193,7 +1194,7 @@ def diff_levenshtein(self, diffs):
         levenshtein = 0
         insertions = 0
         deletions = 0
-        for (op, data) in diffs:
+        for op, data in diffs:
             if op == self.DIFF_INSERT:
                 insertions += len(data)
             elif op == self.DIFF_DELETE:
@@ -1219,7 +1220,7 @@ def diff_toDelta(self, diffs):
           Delta text.
         """
         text = []
-        for (op, data) in diffs:
+        for op, data in diffs:
             if op == self.DIFF_INSERT:
                 # High ascii will raise UnicodeDecodeError.  Use Unicode instead.
                 data = data.encode("utf-8")
@@ -1707,7 +1708,7 @@ def patch_apply(self, patches, text):
                     else:
                         self.diff_cleanupSemanticLossless(diffs)
                         index1 = 0
-                        for (op, data) in patch.diffs:
+                        for op, data in patch.diffs:
                             if op != self.DIFF_EQUAL:
                                 index2 = self.diff_xIndex(diffs, index1)
                             if op == self.DIFF_INSERT:  # Insertion
@@ -2006,7 +2007,7 @@ def __str__(self):
             coords2 = str(self.start2 + 1) + "," + str(self.length2)
         text = ["@@ -", coords1, " +", coords2, " @@\n"]
         # Escape the body of the patch with %xx notation.
-        for (op, data) in self.diffs:
+        for op, data in self.diffs:
             if op == diff_match_patch.DIFF_INSERT:
                 text.append("+")
             elif op == diff_match_patch.DIFF_DELETE: