test: browsing agent utils test (generated by Tom, AI test code agent)

tests/unit/test_browsing_agent_utils.py

@@ -0,0 +1,156 @@
+"""Tests for the browsing agent utils."""
+
+import pytest
+
+from openhands.agenthub.browsing_agent.utils import (
+    ParseError,
+    compress_string,
+    extract_html_tags,
+    parse_html_tags,
+    parse_html_tags_raise,
+    yaml_parser,
+)
+
+
+def test_yaml_parser_valid():
+    """Test yaml_parser with valid input."""
+    valid_yaml = """
+    key1: value1
+    key2: value2
+    """
+    value, valid, retry_message = yaml_parser(valid_yaml)
+    assert valid is True
+    assert value == {'key1': 'value1', 'key2': 'value2'}
+    assert retry_message == ''
+
+
+def test_yaml_parser_invalid():
+    """Test yaml_parser with invalid input."""
+    invalid_yaml = """
+    key1: value1
+    key2: : invalid : syntax :
+    """
+    value, valid, retry_message = yaml_parser(invalid_yaml)
+    assert valid is False
+    assert value == {}
+    assert 'valid yaml' in retry_message.lower()
+
+
+def test_compress_string():
+    """Test string compression with redundant paragraphs and lines."""
+    text = """
+First paragraph
+with multiple lines
+that are unique.
+
+Second paragraph
+that repeats.
+
+Third unique
+paragraph here.
+
+Second paragraph
+that repeats.
+"""
+    result = compress_string(text)
+
+    # Check that definitions section exists
+    assert '<definitions>' in result
+    assert '</definitions>' in result
+
+    # Check that repeated content is replaced with identifiers
+    definitions = extract_html_tags(result, ['definitions'])['definitions'][0]
+    assert '§-0' in result  # Paragraph identifier
+    assert 'Second paragraph' in definitions
+
+
+def test_extract_html_tags_single():
+    """Test extracting a single HTML tag."""
+    text = '<test>Content</test>'
+    result = extract_html_tags(text, ['test'])
+    assert result == {'test': ['Content']}
+
+
+def test_extract_html_tags_multiple():
+    """Test extracting multiple HTML tags."""
+    text = '<tag1>First</tag1><tag2>Second</tag2><tag1>Third</tag1>'
+    result = extract_html_tags(text, ['tag1', 'tag2'])
+    assert result == {'tag1': ['First', 'Third'], 'tag2': ['Second']}
+
+
+def test_extract_html_tags_nested():
+    """Test extracting nested HTML tags."""
+    text = '<outer>Outside<inner>Inside</inner>End</outer>'
+    result = extract_html_tags(text, ['outer', 'inner'])
+    assert result == {'outer': ['Outside<inner>Inside</inner>End'], 'inner': ['Inside']}
+
+
+def test_parse_html_tags_basic():
+    """Test basic HTML tag parsing."""
+    text = '<required>Content</required><optional>Extra</optional>'
+    result, valid, message = parse_html_tags(
+        text, keys=('required',), optional_keys=('optional',)
+    )
+    assert valid is True
+    assert result == {'required': 'Content', 'optional': 'Extra'}
+    assert message == ''
+
+
+def test_parse_html_tags_missing_required():
+    """Test parsing with missing required tags."""
+    text = '<optional>Present</optional>'
+    result, valid, message = parse_html_tags(
+        text, keys=('required',), optional_keys=('optional',)
+    )
+    assert valid is False
+    assert 'Missing the key <required>' in message
+
+
+def test_parse_html_tags_multiple_instances():
+    """Test parsing with multiple instances of the same tag."""
+    text = '<tag>First</tag><tag>Second</tag>'
+
+    # Test without merge_multiple
+    result, valid, message = parse_html_tags(text, keys=('tag',))
+    assert valid is False
+    assert 'multiple instances' in message.lower()
+
+    # Test with merge_multiple
+    result, valid, message = parse_html_tags(text, keys=('tag',), merge_multiple=True)
+    assert valid is True
+    assert 'First' in result['tag']
+    assert 'Second' in result['tag']
+
+
+def test_parse_html_tags_raise():
+    """Test parse_html_tags_raise function."""
+    valid_text = '<required>Content</required>'
+    result = parse_html_tags_raise(valid_text, keys=('required',))
+    assert result == {'required': 'Content'}
+
+    invalid_text = '<wrong>Content</wrong>'
+    with pytest.raises(ParseError):
+        parse_html_tags_raise(invalid_text, keys=('required',))
+
+
+def test_parse_html_tags_empty():
+    """Test parsing with empty content."""
+    text = '<tag></tag>'
+    result, valid, message = parse_html_tags(text, keys=('tag',))
+    assert valid is True
+    assert result == {'tag': ''}
+
+
+def test_parse_html_tags_whitespace():
+    """Test parsing with whitespace content."""
+    text = '<tag>  \n  </tag>'
+    result, valid, message = parse_html_tags(text, keys=('tag',))
+    assert valid is True
+    assert result == {'tag': ''}
+
+
+def test_parse_html_tags_case_sensitivity():
+    """Test case sensitivity in tag parsing."""
+    text = '<TAG>Content</TAG>'
+    result = extract_html_tags(text, ['tag'])
+    assert result == {}  # Tags should be case-sensitive