in dlp/kAnonymityWithEntityIds.js [41:138]
async function kAnonymityWithEntityIds() {
// Specify the BigQuery table to analyze.
const sourceTable = {
projectId: projectId,
datasetId: datasetId,
tableId: sourceTableId,
};
// Specify the unique identifier in the source table for the k-anonymity analysis.
const uniqueIdField = {name: 'Name'};
// These values represent the column names of quasi-identifiers to analyze
const quasiIds = [{name: 'Age'}, {name: 'Mystery'}];
// Configure the privacy metric to compute for re-identification risk analysis.
const privacyMetric = {
kAnonymityConfig: {
entityId: {
field: uniqueIdField,
},
quasiIds: quasiIds,
},
};
// Create action to publish job status notifications to BigQuery table.
const action = [
{
saveFindings: {
outputConfig: {
table: {
projectId: projectId,
datasetId: datasetId,
tableId: outputTableId,
},
},
},
},
];
// Configure the risk analysis job to perform.
const riskAnalysisJob = {
sourceTable: sourceTable,
privacyMetric: privacyMetric,
actions: action,
};
// Combine configurations into a request for the service.
const createDlpJobRequest = {
parent: `projects/${projectId}/locations/global`,
riskJob: riskAnalysisJob,
};
// Send the request and receive response from the service
const [createdDlpJob] = await dlp.createDlpJob(createDlpJobRequest);
const jobName = createdDlpJob.name;
// Waiting for a maximum of 15 minutes for the job to get complete.
let job;
let numOfAttempts = 30;
while (numOfAttempts > 0) {
// Fetch DLP Job status
[job] = await dlp.getDlpJob({name: jobName});
// Check if the job has completed.
if (job.state === 'DONE') {
break;
}
if (job.state === 'FAILED') {
console.log('Job Failed, Please check the configuration.');
return;
}
// Sleep for a short duration before checking the job status again.
await new Promise(resolve => {
setTimeout(() => resolve(), 30000);
});
numOfAttempts -= 1;
}
// Create helper function for unpacking values
const getValue = obj => obj[Object.keys(obj)[0]];
// Print out the results.
const histogramBuckets =
job.riskDetails.kAnonymityResult.equivalenceClassHistogramBuckets;
histogramBuckets.forEach((histogramBucket, histogramBucketIdx) => {
console.log(`Bucket ${histogramBucketIdx}:`);
console.log(
` Bucket size range: [${histogramBucket.equivalenceClassSizeLowerBound}, ${histogramBucket.equivalenceClassSizeUpperBound}]`
);
histogramBucket.bucketValues.forEach(valueBucket => {
const quasiIdValues = valueBucket.quasiIdsValues
.map(getValue)
.join(', ');
console.log(` Quasi-ID values: {${quasiIdValues}}`);
console.log(` Class size: ${valueBucket.equivalenceClassSize}`);
});
});
}