fix(sqllab): prevent strings with angle brackets from being hidden (#34512)

Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
Evan Rusackas
2025-08-02 22:53:17 -07:00
committed by GitHub
parent 7191ae55c8
commit 4e209e51d0
2 changed files with 94 additions and 1 deletions

View File

@@ -49,6 +49,21 @@ describe('isProbablyHTML', () => {
const trickyText = 'a <= 10 and b > 10';
expect(isProbablyHTML(trickyText)).toBe(false);
});
it('should return false for strings with angle brackets that are not HTML', () => {
// Test case from issue #25561
expect(isProbablyHTML('<abcdef:12345>')).toBe(false);
// Other similar cases
expect(isProbablyHTML('<foo:bar>')).toBe(false);
expect(isProbablyHTML('<123>')).toBe(false);
expect(isProbablyHTML('<test@example.com>')).toBe(false);
expect(isProbablyHTML('<custom-element>')).toBe(false);
// Mathematical expressions
expect(isProbablyHTML('if x < 5 and y > 10')).toBe(false);
expect(isProbablyHTML('price < $100')).toBe(false);
});
});
describe('sanitizeHtmlIfNeeded', () => {

View File

@@ -68,9 +68,87 @@ export function isProbablyHTML(text: string) {
return true;
}
// Check if the string contains common HTML patterns
if (!hasHtmlTagPattern(text)) {
return false;
}
const parser = new DOMParser();
const doc = parser.parseFromString(cleanedStr, 'text/html');
return Array.from(doc.body.childNodes).some(({ nodeType }) => nodeType === 1);
// Check if parsing created actual HTML elements (not just text nodes)
const elements = Array.from(doc.body.childNodes).filter(
node => node.nodeType === 1,
) as Element[];
// If no elements were created, it's not HTML
if (elements.length === 0) {
return false;
}
// Check if the elements are known HTML tags (not custom/unknown tags)
// This prevents strings like "<abcdef:12345>" from being treated as HTML
return elements.some(element => {
const tagName = element.tagName.toLowerCase();
// List of common HTML tags we want to recognize
const knownHtmlTags = [
'div',
'span',
'p',
'a',
'b',
'i',
'u',
'em',
'strong',
'h1',
'h2',
'h3',
'h4',
'h5',
'h6',
'table',
'tr',
'td',
'th',
'tbody',
'thead',
'tfoot',
'ul',
'ol',
'li',
'img',
'br',
'hr',
'pre',
'code',
'blockquote',
'section',
'article',
'nav',
'header',
'footer',
'form',
'input',
'button',
'select',
'option',
'textarea',
'label',
'fieldset',
'legend',
'video',
'audio',
'canvas',
'iframe',
'script',
'style',
'link',
'meta',
'title',
];
return knownHtmlTags.includes(tagName);
});
}
export function sanitizeHtmlIfNeeded(htmlString: string) {