/**
* Parser for Google Play data safety information
* Improved version with multiple extraction strategies
*/
/**
* Extracts data safety information from Google Play HTML page
* @param {string} html - HTML content from app page
* @returns {Object|null}
*/
export function parseDataSafety(html) {
if (!html || typeof html !== 'string') {
return null;
}
try {
const result = {
dataShared: [],
dataCollected: [],
securityPractices: [],
privacyPolicyUrl: null,
};
// Strategy 1: Extract privacy policy URL with multiple patterns
const privacyPolicyPatterns = [
/<a[^>]*href=["']([^"']*privacy[^"']*)["'][^>]*>/i,
/privacy[^"']*policy["'][^>]*href=["']([^"']+)["']/i,
/<link[^>]*rel=["']privacy-policy["'][^>]*href=["']([^"']+)["']/i,
/privacyPolicyUrl["']:\s*["']([^"']+)["']/i,
/privacy.*?policy.*?url["']:\s*["']([^"']+)["']/i,
];
for (const pattern of privacyPolicyPatterns) {
const match = html.match(pattern);
if (match) {
result.privacyPolicyUrl = match[1];
break;
}
}
// Strategy 2: Look for JSON-LD structured data
const jsonLdMatches = html.matchAll(/<script[^>]*type=["']application\/ld\+json["'][^>]*>([\s\S]*?)<\/script>/gi);
for (const match of jsonLdMatches) {
try {
const jsonLd = JSON.parse(match[1]);
// Look for data safety information in structured data
if (jsonLd.dataSafety || jsonLd['data-safety'] || jsonLd.data_safety) {
const ds = jsonLd.dataSafety || jsonLd['data-safety'] || jsonLd.data_safety;
if (ds.dataShared) {
result.dataShared = Array.isArray(ds.dataShared) ? ds.dataShared : [];
}
if (ds.dataCollected) {
result.dataCollected = Array.isArray(ds.dataCollected) ? ds.dataCollected : [];
}
if (ds.securityPractices) {
result.securityPractices = Array.isArray(ds.securityPractices) ? ds.securityPractices : [];
}
}
} catch (e) {
// Not valid JSON, continue
}
}
// Strategy 3: Extract from script tags with embedded JSON
const scriptMatches = html.matchAll(/<script[^>]*>([\s\S]*?)<\/script>/gi);
for (const match of scriptMatches) {
const scriptContent = match[1];
if (scriptContent.includes('dataSafety') || scriptContent.includes('data-safety') || scriptContent.includes('data_safety')) {
// Try to find data safety JSON structures
const jsonPatterns = [
/dataSafety["']?\s*:\s*\{([\s\S]*?)\}/i,
/"data-safety"["']?\s*:\s*\{([\s\S]*?)\}/i,
/data_safety["']?\s*:\s*\{([\s\S]*?)\}/i,
];
for (const pattern of jsonPatterns) {
const jsonMatch = scriptContent.match(pattern);
if (jsonMatch) {
try {
const jsonStr = '{' + jsonMatch[1] + '}';
const jsonData = JSON.parse(jsonStr);
if (jsonData.dataShared) {
result.dataShared = Array.isArray(jsonData.dataShared) ? jsonData.dataShared : [];
}
if (jsonData.dataCollected) {
result.dataCollected = Array.isArray(jsonData.dataCollected) ? jsonData.dataCollected : [];
}
if (jsonData.securityPractices) {
result.securityPractices = Array.isArray(jsonData.securityPractices) ? jsonData.securityPractices : [];
}
} catch (e) {
// Not valid JSON, continue
}
}
}
}
}
// Strategy 4: Extract from HTML sections with multiple patterns
const dataSafetySectionPatterns = [
/<div[^>]*class=["'][^"']*data-safety["'][^>]*>([\s\S]*?)<\/div>/i,
/<section[^>]*id=["']data-safety["'][^>]*>([\s\S]*?)<\/section>/i,
/<div[^>]*id=["']data-safety["'][^>]*>([\s\S]*?)<\/div>/i,
/<section[^>]*class=["'][^"']*data-safety["'][^>]*>([\s\S]*?)<\/section>/i,
];
let sectionHtml = null;
for (const pattern of dataSafetySectionPatterns) {
const match = html.match(pattern);
if (match) {
sectionHtml = match[1];
break;
}
}
if (sectionHtml) {
// Extract "Data shared" section
const dataSharedPatterns = [
/data.*?shared["']?\s*:\s*\[([\s\S]*?)\]/i,
/<div[^>]*class=["'][^"']*data-shared["'][^>]*>([\s\S]*?)<\/div>/gi,
/<section[^>]*class=["'][^"']*data-shared["'][^>]*>([\s\S]*?)<\/section>/gi,
];
for (const pattern of dataSharedPatterns) {
const matches = sectionHtml.matchAll(pattern);
for (const match of matches) {
const itemHtml = pattern.source.includes('div') || pattern.source.includes('section') ? match[1] : match[0];
// Extract data items
const dataItemPatterns = [
/<span[^>]*>([^<]+)<\/span>/gi,
/<div[^>]*class=["'][^"']*item["'][^>]*>([^<]+)<\/div>/gi,
/"([^"]+)"/g,
];
for (const itemPattern of dataItemPatterns) {
const itemMatches = itemHtml.matchAll(itemPattern);
for (const itemMatch of itemMatches) {
const dataText = itemMatch[1].trim();
if (dataText && dataText.length > 2 && dataText.length < 100) {
// Check if it's a valid data type (not HTML tags or common words)
if (!dataText.startsWith('<') && !['and', 'or', 'the', 'a', 'an'].includes(dataText.toLowerCase())) {
const existing = result.dataShared.find(d => d.data === dataText);
if (!existing) {
result.dataShared.push({
data: dataText,
optional: itemHtml.includes('optional') || false,
purpose: extractPurpose(itemHtml),
type: extractDataType(dataText),
});
}
}
}
}
}
}
}
// Extract "Data collected" section
const dataCollectedPatterns = [
/data.*?collected["']?\s*:\s*\[([\s\S]*?)\]/i,
/<div[^>]*class=["'][^"']*data-collected["'][^>]*>([\s\S]*?)<\/div>/gi,
/<section[^>]*class=["'][^"']*data-collected["'][^>]*>([\s\S]*?)<\/section>/gi,
];
for (const pattern of dataCollectedPatterns) {
const matches = sectionHtml.matchAll(pattern);
for (const match of matches) {
const itemHtml = pattern.source.includes('div') || pattern.source.includes('section') ? match[1] : match[0];
// Extract data items
const dataItemPatterns = [
/<span[^>]*>([^<]+)<\/span>/gi,
/<div[^>]*class=["'][^"']*item["'][^>]*>([^<]+)<\/div>/gi,
/"([^"]+)"/g,
];
for (const itemPattern of dataItemPatterns) {
const itemMatches = itemHtml.matchAll(itemPattern);
for (const itemMatch of itemMatches) {
const dataText = itemMatch[1].trim();
if (dataText && dataText.length > 2 && dataText.length < 100) {
if (!dataText.startsWith('<') && !['and', 'or', 'the', 'a', 'an'].includes(dataText.toLowerCase())) {
const existing = result.dataCollected.find(d => d.data === dataText);
if (!existing) {
result.dataCollected.push({
data: dataText,
optional: itemHtml.includes('optional') || false,
purpose: extractPurpose(itemHtml),
type: extractDataType(dataText),
});
}
}
}
}
}
}
}
// Extract security practices
const securityPatterns = [
/security.*?practices["']?\s*:\s*\[([\s\S]*?)\]/i,
/<div[^>]*class=["'][^"']*security-practices["'][^>]*>([\s\S]*?)<\/div>/gi,
];
for (const pattern of securityPatterns) {
const matches = sectionHtml.matchAll(pattern);
for (const match of matches) {
const practiceHtml = match[1];
const practiceMatch = practiceHtml.match(/([^<]+)/);
if (practiceMatch) {
const practice = practiceMatch[1].trim();
if (practice && practice.length > 5) {
result.securityPractices.push({
practice: practice,
description: extractDescription(practiceHtml),
});
}
}
}
}
}
// Return result only if we found at least some data
if (result.dataShared.length > 0 || result.dataCollected.length > 0 || result.securityPractices.length > 0 || result.privacyPolicyUrl) {
return result;
}
return null;
} catch (error) {
console.error('Error parsing Google Play data safety:', error);
return null;
}
}
/**
* Extracts purpose from HTML context
* @param {string} html - HTML context
* @returns {string}
*/
function extractPurpose(html) {
const purposePatterns = [
/purpose["']:\s*["']([^"']+)["']/i,
/<span[^>]*class=["'][^"']*purpose["'][^>]*>([^<]+)<\/span>/i,
];
for (const pattern of purposePatterns) {
const match = html.match(pattern);
if (match) {
return match[1].trim();
}
}
return '';
}
/**
* Extracts data type from data name
* @param {string} dataName - Name of the data
* @returns {string}
*/
function extractDataType(dataName) {
const dataTypes = {
'name': 'Personal info',
'email': 'Personal info',
'phone': 'Personal info',
'address': 'Personal info',
'location': 'Location',
'contacts': 'Contacts',
'photos': 'Photos/Media/Files',
'camera': 'Photos/Media/Files',
'microphone': 'Audio',
'calendar': 'Calendar',
'sms': 'Messages',
'call': 'Phone',
};
const lowerName = dataName.toLowerCase();
for (const [key, value] of Object.entries(dataTypes)) {
if (lowerName.includes(key)) {
return value;
}
}
return '';
}
/**
* Extracts description from HTML context
* @param {string} html - HTML context
* @returns {string}
*/
function extractDescription(html) {
const descPatterns = [
/description["']:\s*["']([^"']+)["']/i,
/<p[^>]*>([^<]+)<\/p>/i,
/<span[^>]*>([^<]+)<\/span>/i,
];
for (const pattern of descPatterns) {
const match = html.match(pattern);
if (match) {
return match[1].trim();
}
}
return '';
}
