import pickle
import time
import logging
import pytest
from cssselect import SelectorError, SelectorSyntaxError
from scrapling import Selector
logging.getLogger("scrapling").setLevel(logging.DEBUG)
@pytest.fixture
def html_content():
return """
Complex Web Page
Products
Product 1
This is product 1
$10.99
In stock: 5
Product 2
This is product 2
$20.99
In stock: 3
Product 3
This is product 3
$15.99
Out of stock
Customer Reviews
Good value for money.
Jane Smith
"""
@pytest.fixture
def page(html_content):
return Selector(html_content, adaptive=False)
# CSS Selector Tests
class TestCSSSelectors:
def test_basic_product_selection(self, page):
"""Test selecting all product elements"""
elements = page.css("main #products .product-list article.product")
assert len(elements) == 3
def test_in_stock_product_selection(self, page):
"""Test selecting in-stock products"""
in_stock_products = page.css(
'main #products .product-list article.product:not(:contains("Out of stock"))'
)
assert len(in_stock_products) == 2
# XPath Selector Tests
class TestXPathSelectors:
def test_high_rating_reviews(self, page):
"""Test selecting reviews with high ratings"""
reviews = page.xpath(
'//section[@id="reviews"]//div[contains(@class, "review") and @data-rating >= 4]'
)
assert len(reviews) == 2
def test_high_priced_products(self, page):
"""Test selecting products above a certain price"""
high_priced_products = page.xpath(
'//article[contains(@class, "product")]'
'[number(translate(substring-after(.//span[@class="price"], "$"), ",", "")) > 15]'
)
assert len(high_priced_products) == 2
# Text Matching Tests
class TestTextMatching:
def test_regex_multiple_matches(self, page):
"""Test finding multiple matches with regex"""
stock_info = page.find_by_regex(r"In stock: \d+", first_match=False)
assert len(stock_info) == 2
def test_regex_first_match(self, page):
"""Test finding the first match with regex"""
stock_info = page.find_by_regex(
r"In stock: \d+", first_match=True, case_sensitive=True
)
assert stock_info.text == "In stock: 5"
def test_partial_text_match(self, page):
"""Test finding elements with partial text match"""
stock_info = page.find_by_text(r"In stock:", partial=True, first_match=False)
assert len(stock_info) == 2
def test_exact_text_match(self, page):
"""Test finding elements with exact text match"""
out_of_stock = page.find_by_text(
"Out of stock", partial=False, first_match=False
)
assert len(out_of_stock) == 1
# Similar Elements Tests
class TestSimilarElements:
def test_finding_similar_products(self, page):
"""Test finding similar product elements"""
first_product = page.css(".product").first
similar_products = first_product.find_similar()
assert len(similar_products) == 2
def test_finding_similar_reviews(self, page):
"""Test finding similar review elements with additional filtering"""
first_review = page.find("div", class_="review")
similar_high_rated_reviews = [
review
for review in first_review.find_similar()
if int(review.attrib.get("data-rating", 0)) >= 4
]
assert len(similar_high_rated_reviews) == 1
# Error Handling Tests
class TestErrorHandling:
def test_invalid_selector_initialization(self):
"""Test various invalid Selector initializations"""
# No arguments
with pytest.raises(ValueError):
_ = Selector(adaptive=False)
with pytest.raises(TypeError):
_ = Selector(content=1, adaptive=False)
def test_invalid_storage(self, page, html_content):
"""Test invalid storage parameter"""
with pytest.raises(ValueError):
_ = Selector(html_content, storage=object, adaptive=True)
def test_bad_selectors(self, page):
"""Test handling of invalid selectors"""
with pytest.raises((SelectorError, SelectorSyntaxError)):
page.css("4 ayo")
with pytest.raises((SelectorError, SelectorSyntaxError)):
page.xpath("4 ayo")
# Pickling and Object Representation Tests
class TestPicklingAndRepresentation:
def test_unpickleable_objects(self, page):
"""Test that Selector objects cannot be pickled"""
table = page.css(".product-list")[0]
with pytest.raises(TypeError):
pickle.dumps(table)
def test_string_representations(self, page):
"""Test custom string representations of objects"""
table = page.css(".product-list")[0]
assert issubclass(type(table.__str__()), str)
assert issubclass(type(table.__repr__()), str)
assert issubclass(type(table.attrib.__str__()), str)
assert issubclass(type(table.attrib.__repr__()), str)
# Navigation and Traversal Tests
class TestElementNavigation:
def test_basic_navigation_properties(self, page):
"""Test basic navigation properties of elements"""
table = page.css(".product-list")[0]
assert table.path is not None
assert table.html_content != ""
assert table.prettify() != ""
def test_parent_and_sibling_navigation(self, page):
"""Test parent and sibling navigation"""
table = page.css(".product-list")[0]
parent = table.parent
assert parent["id"] == "products"
parent_siblings = parent.siblings
assert len(parent_siblings) == 1
def test_child_navigation(self, page):
"""Test child navigation"""
table = page.css(".product-list")[0]
children = table.children
assert len(children) == 3
def test_next_and_previous_navigation(self, page):
"""Test next and previous element navigation"""
child = page.css(".product-list")[0].find({"data-id": "1"})
next_element = child.next
assert next_element.attrib["data-id"] == "2"
prev_element = next_element.previous
assert prev_element.tag == child.tag
def test_ancestor_finding(self, page):
"""Test finding ancestors of elements"""
all_prices = page.css(".price")
products_with_prices = [
price.find_ancestor(lambda p: p.has_class("product"))
for price in all_prices
]
assert len(products_with_prices) == 3
# JSON and Attribute Tests
class TestJSONAndAttributes:
def test_json_conversion(self, page):
"""Test converting content to JSON"""
script_content = page.css("#page-data::text")[0].get()
assert issubclass(type(script_content.sort()), str)
page_data = script_content.json()
assert page_data["totalProducts"] == 3
assert "lastUpdated" in page_data
def test_attribute_operations(self, page):
"""Test various attribute-related operations"""
# Product ID extraction
products = page.css(".product")
product_ids = [product.attrib["data-id"] for product in products]
assert product_ids == ["1", "2", "3"]
assert "data-id" in products[0]
# Review rating calculations
reviews = page.css(".review")
review_ratings = [int(review.attrib["data-rating"]) for review in reviews]
assert sum(review_ratings) / len(review_ratings) == 4.5
# Attribute searching
key_value = list(products[0].attrib.search_values("1", partial=False))
assert list(key_value[0].keys()) == ["data-id"]
key_value = list(products[0].attrib.search_values("1", partial=True))
assert list(key_value[0].keys()) == ["data-id"]
# JSON attribute conversion
attr_json = page.css("#products").first.attrib["schema"].json()
assert attr_json == {"jsonable": "data"}
assert isinstance(page.css("#products")[0].attrib.json_string, bytes)
# Performance Test
def test_large_html_parsing_performance():
"""Test parsing and selecting performance on large HTML"""
large_html = (
""
+ '' * 5000
+ "
" * 5000
+ ""
)
start_time = time.time()
parsed = Selector(large_html, adaptive=False)
elements = parsed.css(".item")
end_time = time.time()
# assert len(elements) == 5000 # GitHub actions don't like this line
# Converting 5000 elements to a class and doing operations on them will take time
# Based on my tests with 100 runs, 1 loop each Scrapling (given the extra work/features) takes 10.4ms on average
assert (
end_time - start_time < 0.5
) # Locally I test on 0.1 but on GitHub actions with browsers and threading sometimes closing adds fractions of seconds
# Selector Generation Test
def test_selectors_generation(page):
"""Try to create selectors for all elements in the page"""
def _traverse(element: Selector):
assert isinstance(element.generate_css_selector, str)
assert isinstance(element.generate_full_css_selector, str)
assert isinstance(element.generate_xpath_selector, str)
assert isinstance(element.generate_full_xpath_selector, str)
for branch in element.children:
_traverse(branch)
_traverse(page)
def test_full_path_selector_no_duplicate_ids():
"""Test that full path selectors don't duplicate id segments (regression test)"""
html = ''
page = Selector(html)
target = page.css("#target").first
# CSS full path should not duplicate id selectors
css_full = target.generate_full_css_selector
assert css_full.count("#target") == 1, f"Duplicate #target in CSS full path: {css_full}"
assert css_full.count("#main") == 1, f"Duplicate #main in CSS full path: {css_full}"
# XPath full path should not duplicate id selectors
xpath_full = target.generate_full_xpath_selector
assert xpath_full.count("@id='target'") == 1, f"Duplicate @id='target' in XPath full path: {xpath_full}"
assert xpath_full.count("@id='main'") == 1, f"Duplicate @id='main' in XPath full path: {xpath_full}"
# The generated CSS selector should actually select the correct element
result = page.css(css_full)
assert len(result) == 1
assert result.first.text == "Hello"
# The generated XPath selector should also select the correct element
result = page.xpath(xpath_full)
assert len(result) == 1, f"XPath '{xpath_full}' selected {len(result)} elements, expected 1"
assert result.first.text == "Hello"
def test_full_path_selector_mixed_id_and_no_id():
"""Test full path selectors with a mix of elements with and without ids"""
html = ''
page = Selector(html)
target = page.css("p").first
css_full = target.generate_full_css_selector
# p has no id, so it should appear as a tag name; div has id
assert "#wrapper" in css_full
assert css_full.count("#wrapper") == 1
# Verify the selector works
result = page.css(css_full)
assert len(result) == 1
assert result.first.text == "Text"
# Miscellaneous Tests
def test_getting_all_text(page):
"""Test getting all text from the page"""
assert page.get_all_text() != ""
def test_regex_on_text(page):
"""Test regex operations on text"""
element = page.css('[data-id="1"] .price')[0]
match = element.re_first(r"[\.\d]+")
assert match == "10.99"
match = element.text.re(r"(\d+)", replace_entities=False)
assert len(match) == 2