feat: Add code samples for DLP text redaction (#61)

Author: jmorrise

dlp/redact.js

@@ -15,6 +15,52 @@
 
 'use strict';
 
+function redactText(callingProjectId, string, minLikelihood, infoTypes) {
+  // [START dlp_redact_text]
+  // Imports the Google Cloud Data Loss Prevention library
+  const DLP = require('@google-cloud/dlp');
+
+  // Instantiates a client
+  const dlp = new DLP.DlpServiceClient();
+
+  // Construct transformation config which replaces sensitive info with its info type.
+  // E.g., "Her email is [email protected]" => "Her email is [EMAIL_ADDRESS]"
+  const replaceWithInfoTypeTransformation = {
+    primitiveTransformation: {
+      replaceWithInfoTypeConfig: {},
+    },
+  };
+
+  // Construct redaction request
+  const request = {
+    parent: dlp.projectPath(callingProjectId),
+    item: {
+      value: string,
+    },
+    deidentifyConfig: {
+      infoTypeTransformations: {
+        transformations: [replaceWithInfoTypeTransformation],
+      },
+    },
+    inspectConfig: {
+      minLikelihood: minLikelihood,
+      infoTypes: infoTypes,
+    },
+  };
+
+  // Run string redaction
+  dlp
+    .deidentifyContent(request)
+    .then(response => {
+      const resultString = response[0].item.value;
+      console.log(`Redacted text: ${resultString}`);
+    })
+    .catch(err => {
+      console.log(`Error in deidentifyContent: ${err.message || err}`);
+    });
+  // [END dlp_redact_text]
+}
+
 function redactImage(
   callingProjectId,
   filepath,
@@ -89,6 +135,18 @@ function redactImage(
 
 const cli = require(`yargs`)
   .demand(1)
+  .command(
+    `string <string>`,
+    `Redact a string using the Data Loss Prevention API.`,
+    {},
+    opts =>
+      redactText(
+        opts.callingProject,
+        opts.string,
+        opts.minLikelihood,
+        opts.infoTypes
+      )
+  )
   .command(
     `image <filepath> <outputPath>`,
     `Redact sensitive data from an image using the Data Loss Prevention API.`,

dlp/system-test/redact.test.js

@@ -28,6 +28,30 @@ const testResourcePath = 'system-test/resources';
 
 test.before(tools.checkCredentials);
 
+test(`should redact a single sensitive data type from a string`, async t => {
+  const output = await tools.runAsync(
+    `${cmd} string "My email is [email protected]" -t EMAIL_ADDRESS`,
+    cwd
+  );
+  t.regex(output, /My email is \[EMAIL_ADDRESS\]/);
+});
+
+test(`should redact multiple sensitive data types from a string`, async t => {
+  const output = await tools.runAsync(
+    `${cmd} string "I am 29 years old and my email is [email protected]" -t EMAIL_ADDRESS AGE`,
+    cwd
+  );
+  t.regex(output, /I am \[AGE\] and my email is \[EMAIL_ADDRESS\]/);
+});
+
+test(`should handle string with no sensitive data`, async t => {
+  const output = await tools.runAsync(
+    `${cmd} string "No sensitive data to redact here" -t EMAIL_ADDRESS AGE`,
+    cwd
+  );
+  t.regex(output, /No sensitive data to redact here/);
+});
+
 // redact_image
 test(`should redact a single sensitive data type from an image`, async t => {
   const testName = `redact-single-type`;
@@ -61,6 +85,14 @@ test(`should redact multiple sensitive data types from an image`, async t => {
   t.deepEqual(correct, result);
 });
 
+test(`should report info type errors`, async t => {
+  const output = await tools.runAsync(
+    `${cmd} string "My email is [email protected]" -t NONEXISTENT`,
+    cwd
+  );
+  t.regex(output, /Error in deidentifyContent/);
+});
+
 test(`should report image redaction handling errors`, async t => {
   const output = await tools.runAsync(
     `${cmd} image ${testImage} nonexistent.result.png -t BAD_TYPE`,