diff --git a/notifications_utils/formatters.py b/notifications_utils/formatters.py
index 0ddd1365..d922fd02 100644
--- a/notifications_utils/formatters.py
+++ b/notifications_utils/formatters.py
@@ -23,7 +23,7 @@
LIST_ITEM_STYLE = 'Margin: 5px 0 5px; padding: 0 0 0 5px; font-size: 16px; line-height: 25px; color: #323A45;'
PARAGRAPH_STYLE = 'Margin: 0 0 20px 0; font-size: 16px; line-height: 25px; color: #323A45;'
THEMATIC_BREAK_STYLE = 'border: 0; height: 1px; background: #BFC1C3; Margin: 30px 0 30px 0;'
-UNORDERED_LIST_STYLE = 'Margin: 0 0 0 20px; padding: 0 0 20px 0; list-style-type: disk; ' \
+UNORDERED_LIST_STYLE = 'Margin: 0 0 0 20px; padding: 0 0 20px 0; list-style-type: disc; ' \
'font-family: Helvetica, Arial, sans-serif;'
OBSCURE_WHITESPACE = (
@@ -346,13 +346,17 @@ def insert_list_spaces(md: str) -> str:
"""
Proper markdown for lists has a space after the number or bullet. This is a preprocessing step to insert
any missing spaces in lists. This preprocessing should take place before any manipulation by Mistune.
+
+ The regular expression for unordered lists replaces the bullet with the minus, which Mistune handles.
+ This is necessary because Utils allows the non-standard literal • in markdown to denote an unordered list.
+ Performing this substitution avoids having to write custom parsing logic for Mistune.
"""
# Ordered lists
- md = re.sub(r'''^(\s*)(\d+\.)''', r'''\1\2 ''', md, flags=re.M)
+ md = re.sub(r'''^(\s*)(\d+\.)(?=\S)''', r'''\1\2 ''', md, flags=re.M)
# Unordered lists
- return re.sub(r'''^(\s*)(\*|-|\+|•)(?!\2)''', r'''\1- ''', md, flags=re.M)
+ return re.sub(r'''^(\s*)(\*|-|\+|•)(?!\2)(\s*)''', r'''\1- ''', md, flags=re.M)
def strip_parentheses_in_link_placeholders(value: str) -> str:
diff --git a/tests/test_formatted_list.py b/tests/test_formatted_list.py
deleted file mode 100644
index e69de29b..00000000
diff --git a/tests/test_formatters.py b/tests/test_formatters.py
index e79219d7..562cf4a4 100644
--- a/tests/test_formatters.py
+++ b/tests/test_formatters.py
@@ -2,23 +2,24 @@
from markupsafe import Markup
from notifications_utils.formatters import (
+ escape_html,
+ formatted_list,
+ insert_list_spaces,
+ make_quotes_smart,
+ nl2li,
+ normalise_whitespace,
notify_html_markdown,
notify_markdown,
- sms_encode,
- formatted_list,
- strip_dvla_markup,
- strip_pipes,
- escape_html,
+ remove_smart_quotes_from_email_addresses,
remove_whitespace_before_punctuation,
- make_quotes_smart,
replace_hyphens_with_en_dashes,
- tweak_dvla_list_markup,
- nl2li,
- strip_whitespace,
+ sms_encode,
strip_and_remove_obscure_whitespace,
- remove_smart_quotes_from_email_addresses,
+ strip_dvla_markup,
+ strip_pipes,
strip_unsupported_characters,
- normalise_whitespace,
+ strip_whitespace,
+ tweak_dvla_list_markup,
)
from notifications_utils.template import (
HTMLEmailTemplate,
@@ -421,7 +422,7 @@ def test_ordered_list():
'\n* three'
),
(
- '\n'
'- '
@@ -430,7 +431,7 @@ def test_ordered_list():
'
\n'
'nested 1
\n'
'nested 2
\n'
- '\n'
'- two
\n'
@@ -469,13 +470,8 @@ def test_paragraph_in_list_has_no_linebreak(test_text, expected):
'+ two\n'
'+ three\n'
),
- pytest.param(( # bullet as bullet - This is non-standard.
- '• one\n'
- '• two\n'
- '• three\n'
- ), marks=pytest.mark.xfail(strict=False)),
),
- ids=['two_spaces', 'tab', 'dash_as_bullet', 'plus_as_bullet', 'bullet_as_bullet']
+ ids=['two_spaces', 'tab', 'dash_as_bullet', 'plus_as_bullet']
)
@pytest.mark.parametrize(
'markdown_function, expected',
@@ -484,7 +480,7 @@ def test_paragraph_in_list_has_no_linebreak(test_text, expected):
notify_html_markdown,
(
'\n'
'- '
@@ -993,3 +989,63 @@ def test_strip_unsupported_characters():
def test_normalise_whitespace():
assert normalise_whitespace('\u200C Your tax is\ndue\n\n') == 'Your tax is due'
+
+
+@pytest.mark.parametrize(
+ 'actual, expected',
+ [
+ (
+ '1.one\n2.two\n3.three',
+ '1. one\n2. two\n3. three',
+ ),
+ (
+ '-one\n -two\n-three',
+ '- one\n - two\n- three',
+ ),
+ (
+ '+one\n +two\n+three',
+ '- one\n - two\n- three',
+ ),
+ (
+ '*one\n *two\n*three',
+ '- one\n - two\n- three',
+ ),
+ (
+ '•one\n •two\n•three',
+ '- one\n - two\n- three',
+ ),
+ # Next 2 tests: Shouldn't misinterpret a thematic break as a list item
+ (
+ '***one\n *two\n*three',
+ '***one\n - two\n- three',
+ ),
+ (
+ '-one\n ---two\n-three',
+ '- one\n ---two\n- three',
+ ),
+ (
+ '# This is Heading 1\n'
+ '## This is heading 2\n'
+ '### This is heading 3\n'
+ '\n'
+ '__This has been emboldened__\n'
+ '\n'
+ '- this is a bullet list\n'
+ '- list list list\n'
+ '\n'
+ 'Testing personalisation, ((name)).\n',
+ '# This is Heading 1\n'
+ '## This is heading 2\n'
+ '### This is heading 3\n'
+ '\n'
+ '__This has been emboldened__\n'
+ '\n'
+ '- this is a bullet list\n'
+ '- list list list\n'
+ '\n'
+ 'Testing personalisation, ((name)).\n',
+ ),
+ ]
+)
+def test_insert_list_spaces(actual, expected):
+ assert insert_list_spaces(actual) == expected
diff --git a/tests/test_template_types.py b/tests/test_template_types.py
index ddc2be00..eea2a643 100644
--- a/tests/test_template_types.py
+++ b/tests/test_template_types.py
@@ -1584,6 +1584,7 @@ def test_ordered_list_without_spaces(template_type, expected):
assert expected in str(template_type({'content': content, 'subject': ''}))
+@pytest.mark.parametrize('with_spaces', [True, False])
@pytest.mark.parametrize(
'template_type, expected',
[
@@ -1599,12 +1600,13 @@ def test_ordered_list_without_spaces(template_type, expected):
]
)
@pytest.mark.parametrize('bullet', ['*', '-', '+', '•'])
-def test_unordered_list_without_spaces(bullet, template_type, expected):
+def test_unordered_list(bullet, template_type, expected, with_spaces):
"""
Proper markdown for unordered lists has a space after the bullet.
"""
- content = f'{bullet}one\n{bullet}two\n{bullet}three\n'
+ space = ' ' if with_spaces else ''
+ content = f'{bullet}{space}one\n{bullet}{space}two\n{bullet}{space}three\n'
if isinstance(template_type, PlainTextEmailTemplate):
assert str(template_type({'content': content, 'subject': ''})) == expected