From d1222208081c6608d7601297db0cbffc28ae5dee Mon Sep 17 00:00:00 2001
From: JensDiemer <git@jensdiemer.de>
Date: Sat, 9 Oct 2021 15:39:37 +0200
Subject: [PATCH] Better tracebacks on html validation errors

See also: https://github.com/boxine/bx_py_utils/pull/97
---
 src/inventory_project/tests/temp_utils.py | 157 ++++++++++++++++++++--
 1 file changed, 146 insertions(+), 11 deletions(-)

diff --git a/src/inventory_project/tests/temp_utils.py b/src/inventory_project/tests/temp_utils.py
index ce4cb59..c9965eb 100644
--- a/src/inventory_project/tests/temp_utils.py
+++ b/src/inventory_project/tests/temp_utils.py
@@ -1,31 +1,166 @@
 """
     Remove this if https://github.com/boxine/bx_py_utils/pull/95 merged!
 """
+import inspect
 from pathlib import Path
 
 import pytest
 from bs4 import BeautifulSoup
 from bx_py_utils.test_utils.snapshot import assert_text_snapshot
 from django.http import HttpResponse
-from lxml import html
+from lxml import etree
+from lxml.etree import XMLSyntaxError
 
 
-def validate_html(data):
-    parser = html.HTMLParser(
-        recover=False,  # Crash faster on broken HTML
-    )
-    parser.feed(data)
-    parser.close()
+def cutout(text, line_no, column, extra_lines=2):
+    assert isinstance(text, str)
+    assert line_no >= 0
+    assert column >= 0
+    assert extra_lines >= 0
+
+    lines = text.splitlines()
+    line_count = len(lines)
+
+    assert line_no <= line_count
+
+    from_line = line_no - extra_lines - 1
+    if from_line < 0:
+        from_line = 0
+
+    to_line = line_no + extra_lines
+    if to_line > line_count:
+        to_line = line_count
+
+    line_no_width = len(str(from_line)) + 1
+
+    lines = lines[from_line: to_line]
+    result = []
+    for no, line in enumerate(lines, from_line + 1):
+        result.append(
+            f'{no:0{line_no_width}} {line}'
+        )
+        if no == line_no:
+            result.append(
+                f'{"-"*(line_no_width+column+1)}^'
+            )
+
+    return '\n'.join(result)
+
+
+def test_cutout():
+    text = inspect.cleandoc('''
+        line 1
+        line 2
+        01234567890 line 3
+        line 4
+        line 5
+    ''')
+
+    output = cutout(text, line_no=3, column=5, extra_lines=1)
+    assert output == inspect.cleandoc('''
+        02 line 2
+        03 01234567890 line 3
+        --------^
+        04 line 4
+    ''')
+
+    output = cutout(text, line_no=3, column=1, extra_lines=0)
+    assert output == inspect.cleandoc('''
+        03 01234567890 line 3
+        ----^
+    ''')
+
+    output = cutout(text, line_no=3, column=10, extra_lines=2)
+    assert output == inspect.cleandoc('''
+        01 line 1
+        02 line 2
+        03 01234567890 line 3
+        -------------^
+        04 line 4
+        05 line 5
+    ''')
+
+    text = '\n'.join(f'The Line {no}' for no in range(20))
+    output = cutout(text, line_no=18, column=9, extra_lines=2)
+    assert output == inspect.cleandoc('''
+        016 The Line 15
+        017 The Line 16
+        018 The Line 17
+        -------------^
+        019 The Line 18
+        020 The Line 19
+    ''')
+
+
+class InvalidHtml(AssertionError):
+    """
+    XMLSyntaxError with better error messages: used in validate_html()
+    """
+
+    def __init__(self, *args):
+        self.args = args
+
+        data, origin_err = args
+        assert isinstance(data, str)
+        assert isinstance(origin_err, XMLSyntaxError)
+
+        self.origin_msg = origin_err.msg
+
+        line_no, column = origin_err.position
+        self.cutoput_text = cutout(data, line_no, column, extra_lines=3)
+
+    def __str__(self):
+        return (
+            f'{self.origin_msg}\n'
+            f'{"-"*80}\n'
+            f'{self.cutoput_text}\n'
+            f'{"-"*80}'
+        )
+
+
+def validate_html(data, **parser_kwargs):
+    parser = etree.XMLParser(**parser_kwargs)
+    try:
+        parser.feed(data)
+        parser.close()
+    except XMLSyntaxError as err:
+        raise InvalidHtml(data, err)
 
 
 def test_validate_html():
     validate_html('<p>Test</p>')
     validate_html('<a><b/></a>')
 
-    from lxml.etree import XMLSyntaxError
-    with pytest.raises(XMLSyntaxError) as exc_info:
-        validate_html('<p> >broken< </p>')
-    assert exc_info.value.args[0] == 'htmlParseStartTag: invalid element name, line 1, column 13'
+    validate_html('<foo></foo>')
+    validate_html('<nav></nav>')
+    validate_html('<nav class="sticky" id="nav-sidebar"></nav>')
+
+    with pytest.raises(InvalidHtml) as exc_info:
+        validate_html(inspect.cleandoc('''
+            <no-html>
+                <foo>
+                    <bar>
+                        <h1>Test</h1>
+                        <p> >broken< </p>
+                        <p>the end</p>
+                    <bar>
+                </foo>
+            </no-html>
+        '''))
+    error_message = str(exc_info.value)
+    assert error_message == inspect.cleandoc('''
+        StartTag: invalid element name, line 5, column 25
+        --------------------------------------------------------------------------------
+        02     <foo>
+        03         <bar>
+        04             <h1>Test</h1>
+        05             <p> >broken< </p>
+        ----------------------------^
+        06             <p>the end</p>
+        07         <bar>
+        08     </foo>
+        --------------------------------------------------------------------------------
+    ''')
 
 
 def pretty_format_html(data):