HTTP Functions

You use HTTP functions when you want to invoke your function via an HTTP(s) request. To allow for HTTP semantics, HTTP function signatures accept HTTP-specific arguments.

You can configure HTTP functions to control whether they can be triggered by both HTTPS and HTTP, or HTTPS only. See Security levels for details.

Running the samples

To run the samples in this document, make sure that you have set up the environment for your runtime as described in this quickstart. In particular, make sure to clone the sample repository to your local machine to ensure that all the required files are present in your environment.

Sample usage

The example below shows how to process an HTTP POST request containing a name parameter:

Node.js

functions/helloworld/index.js 
const functions = require('@google-cloud/functions-framework');
const escapeHtml = require('escape-html');

/**
 * Responds to an HTTP request using data from the request body parsed according
 * to the "content-type" header.
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
functions.http('helloHttp', (req, res) => {
  res.send(`Hello ${escapeHtml(req.query.name || req.body.name || 'World')}!`);
});

Python

functions/helloworld/main.py 
from flask import escape
import functions_framework

@functions_framework.http
def hello_http(request):
    """HTTP Cloud Function.
    Args:
        request (flask.Request): The request object.
        <https://flask.palletsprojects.com/en/1.1.x/api/#incoming-request-data>
    Returns:
        The response text, or any set of values that can be turned into a
        Response object using `make_response`
        <https://flask.palletsprojects.com/en/1.1.x/api/#flask.make_response>.
    """
    request_json = request.get_json(silent=True)
    request_args = request.args

    if request_json and 'name' in request_json:
        name = request_json['name']
    elif request_args and 'name' in request_args:
        name = request_args['name']
    else:
        name = 'World'
    return 'Hello {}!'.format(escape(name))

Go

functions/helloworld/hello_http.go 

// Package helloworld provides a set of Cloud Functions samples.
package helloworld

import (
	"encoding/json"
	"fmt"
	"html"
	"net/http"
)

// HelloHTTP is an HTTP Cloud Function with a request parameter.
func HelloHTTP(w http.ResponseWriter, r *http.Request) {
	var d struct {
		Name string `json:"name"`
	}
	if err := json.NewDecoder(r.Body).Decode(&d); err != nil {
		fmt.Fprint(w, "Hello, World!")
		return
	}
	if d.Name == "" {
		fmt.Fprint(w, "Hello, World!")
		return
	}
	fmt.Fprintf(w, "Hello, %s!", html.EscapeString(d.Name))
}

Java

functions/helloworld/hello-http/src/main/java/functions/HelloHttp.java 

import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import com.google.gson.Gson;
import com.google.gson.JsonElement;
import com.google.gson.JsonObject;
import com.google.gson.JsonParseException;
import java.io.IOException;
import java.io.PrintWriter;
import java.util.logging.Logger;

public class HelloHttp implements HttpFunction {
  private static final Logger logger = Logger.getLogger(HelloHttp.class.getName());

  private static final Gson gson = new Gson();

  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException {
    // Check URL parameters for "name" field
    // "world" is the default value
    String name = request.getFirstQueryParameter("name").orElse("world");

    // Parse JSON request and check for "name" field
    try {
      JsonElement requestParsed = gson.fromJson(request.getReader(), JsonElement.class);
      JsonObject requestJson = null;

      if (requestParsed != null && requestParsed.isJsonObject()) {
        requestJson = requestParsed.getAsJsonObject();
      }

      if (requestJson != null && requestJson.has("name")) {
        name = requestJson.get("name").getAsString();
      }
    } catch (JsonParseException e) {
      logger.severe("Error parsing JSON: " + e.getMessage());
    }

    var writer = new PrintWriter(response.getWriter());
    writer.printf("Hello %s!", name);
  }
}

C#

functions/helloworld/HelloHttp/Function.cs 
using Google.Cloud.Functions.Framework;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Logging;
using System.IO;
using System.Text.Json;
using System.Threading.Tasks;

namespace HelloHttp
{
    public class Function : IHttpFunction
    {
        private readonly ILogger _logger;

        public Function(ILogger<Function> logger) =>
            _logger = logger;

        public async Task HandleAsync(HttpContext context)
        {
            HttpRequest request = context.Request;
            // Check URL parameters for "name" field
            // "world" is the default value
            string name = ((string) request.Query["name"]) ?? "world";

            // If there's a body, parse it as JSON and check for "name" field.
            using TextReader reader = new StreamReader(request.Body);
            string text = await reader.ReadToEndAsync();
            if (text.Length > 0)
            {
                try
                {
                    JsonElement json = JsonSerializer.Deserialize<JsonElement>(text);
                    if (json.TryGetProperty("name", out JsonElement nameElement) &&
                        nameElement.ValueKind == JsonValueKind.String)
                    {
                        name = nameElement.GetString();
                    }
                }
                catch (JsonException parseException)
                {
                    _logger.LogError(parseException, "Error parsing JSON request");
                }
            }

            await context.Response.WriteAsync($"Hello {name}!");
        }
    }
}

Ruby

functions/helloworld/http/app.rb 
require "functions_framework"
require "cgi"
require "json"

FunctionsFramework.http "hello_http" do |request|
  # The request parameter is a Rack::Request object.
  # See https://www.rubydoc.info/gems/rack/Rack/Request
  name = request.params["name"] ||
         (JSON.parse(request.body.read)["name"] rescue nil) ||
         "World"
  # Return the response body as a string.
  # You can also return a Rack::Response object, a Rack response array, or
  # a hash which will be JSON-encoded into a response.
  "Hello #{CGI.escape_html name}!"
end

PHP

functions/helloworld_http/index.php 
<?php

use Google\CloudFunctions\FunctionsFramework;
use Psr\Http\Message\ServerRequestInterface;

// Register the function with Functions Framework.
// This enables omitting the `FUNCTIONS_SIGNATURE_TYPE=http` environment
// variable when deploying. The `FUNCTION_TARGET` environment variable should
// match the first parameter.
FunctionsFramework::http('helloHttp', 'helloHttp');

function helloHttp(ServerRequestInterface $request): string
{
    $name = 'World';
    $body = $request->getBody()->getContents();
    if (!empty($body)) {
        $json = json_decode($body, true);
        if (json_last_error() != JSON_ERROR_NONE) {
            throw new RuntimeException(sprintf(
                'Could not parse body: %s',
                json_last_error_msg()
            ));
        }
        $name = $json['name'] ?? $name;
    }
    $queryString = $request->getQueryParams();
    $name = $queryString['name'] ?? $name;

    return sprintf('Hello, %s!', htmlspecialchars($name));
}

The following command shows how to call the function and pass it a parameter using curl:

curl -X POST HTTP_TRIGGER_ENDPOINT -H "Content-Type:application/json"  -d '{"name":"Jane"}'

where HTTP_TRIGGER_ENDPOINT is the URL for the function, obtained when the function is deployed. For more information, see HTTP Triggers.

Try it for yourself

If you're new to Google Cloud, create an account to evaluate how Cloud Functions performs in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.

Try Cloud Functions free

Security levels

The security level feature controls whether an HTTP function's URL supports HTTPS only, or both HTTP and HTTPS. Support for HTTPS only is the default.

When an HTTP function is configured so that it can only be triggered with HTTPS, users who attempt to use the HTTP protocol will be redirected.

You specify an HTTP function's security level during deployment:

  • If you are using the Google Cloud CLI to deploy your function, you can set the function's security level using the --security-level flag. Its possible values are secure-optional or secure-always, which is the default. For example:

    gcloud functions deploy FUNCTION_NAME --trigger-http --security-level=secure-optional...

  • If you are deploying your function from the Google Cloud console, use the Require HTTPS checkbox to toggle whether the function requires HTTPS.

If the function can be triggered with either HTTP or HTTPS, the function code can examine the value of the request header X-Forwarded-Proto to determine which protocol was used. A secure request will have the value https for that header, while other requests will have the value http.

HTTP frameworks

To handle HTTP, Cloud Functions uses a particular HTTP framework version in each runtime:

Runtime HTTP framework
Node.js Express 4.17.1
Python Flask 1.0.2
Go Standard http.HandlerFunc interface
Java Functions Framework Java API
.NET Functions Framework for .NET
Ruby Functions Framework for Ruby
PHP Functions Framework for PHP

Terminating HTTP functions

If a function creates background tasks (such as threads, futures, Node.js Promise objects, callbacks, or system processes), you must terminate or otherwise resolve these tasks before returning an HTTP response. Any tasks not terminated prior to an HTTP response may not be completed, and may also cause undefined behavior.

Parsing HTTP requests

The example below shows how to read HTTP requests in various formats:

Node.js

In Node.js, the body of the request is automatically parsed based on the content-type header and made available via your HTTP function's arguments.

functions/http/index.js 
const escapeHtml = require('escape-html');

/**
 * Responds to an HTTP request using data from the request body parsed according
 * to the "content-type" header.
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
exports.helloContent = (req, res) => {
  let name;

  switch (req.get('content-type')) {
    // '{"name":"John"}'
    case 'application/json':
      ({name} = req.body);
      break;

    // 'John', stored in a Buffer
    case 'application/octet-stream':
      name = req.body.toString(); // Convert buffer to a string
      break;

    // 'John'
    case 'text/plain':
      name = req.body;
      break;

    // 'name=John' in the body of a POST request (not the URL)
    case 'application/x-www-form-urlencoded':
      ({name} = req.body);
      break;
  }

  res.status(200).send(`Hello ${escapeHtml(name || 'World')}!`);
};

Python

functions/helloworld/main.py 
from flask import escape
import functions_framework

@functions_framework.http
def hello_content(request):
    """ Responds to an HTTP request using data from the request body parsed
    according to the "content-type" header.
    Args:
        request (flask.Request): The request object.
        <https://flask.palletsprojects.com/en/1.1.x/api/#incoming-request-data>
    Returns:
        The response text, or any set of values that can be turned into a
        Response object using `make_response`
        <https://flask.palletsprojects.com/en/1.1.x/api/#flask.make_response>.
    """
    content_type = request.headers['content-type']
    if content_type == 'application/json':
        request_json = request.get_json(silent=True)
        if request_json and 'name' in request_json:
            name = request_json['name']
        else:
            raise ValueError("JSON is invalid, or missing a 'name' property")
    elif content_type == 'application/octet-stream':
        name = request.data
    elif content_type == 'text/plain':
        name = request.data
    elif content_type == 'application/x-www-form-urlencoded':
        name = request.form.get('name')
    else:
        raise ValueError("Unknown content type: {}".format(content_type))
    return 'Hello {}!'.format(escape(name))

Go

functions/http/http_content_type.go 

// Package http provides a set of HTTP Cloud Functions samples.
package http

import (
	"encoding/json"
	"fmt"
	"html"
	"io/ioutil"
	"log"
	"net/http"
)

// HelloContentType is an HTTP Cloud function.
// It uses the Content-Type header to identify the request payload format.
func HelloContentType(w http.ResponseWriter, r *http.Request) {
	var name string

	switch r.Header.Get("Content-Type") {
	case "application/json":
		var d struct {
			Name string `json:"name"`
		}
		err := json.NewDecoder(r.Body).Decode(&d)
		if err != nil {
			log.Printf("error parsing application/json: %v", err)
		} else {
			name = d.Name
		}
	case "application/octet-stream":
		body, err := ioutil.ReadAll(r.Body)
		if err != nil {
			log.Printf("error parsing application/octet-stream: %v", err)
		} else {
			name = string(body)
		}
	case "text/plain":
		body, err := ioutil.ReadAll(r.Body)
		if err != nil {
			log.Printf("error parsing text/plain: %v", err)
		} else {
			name = string(body)
		}
	case "application/x-www-form-urlencoded":
		if err := r.ParseForm(); err != nil {
			log.Printf("error parsing application/x-www-form-urlencoded: %v", err)
		} else {
			name = r.FormValue("name")
		}
	}

	if name == "" {
		name = "World"
	}

	fmt.Fprintf(w, "Hello, %s!", html.EscapeString(name))
}

Java

functions/http/parse-content-type/src/main/java/functions/ParseContentType.java 

import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import com.google.gson.Gson;
import com.google.gson.JsonObject;
import java.io.IOException;
import java.io.PrintWriter;
import java.net.HttpURLConnection;
import java.nio.charset.StandardCharsets;
import java.util.Base64;
import java.util.Optional;

public class ParseContentType implements HttpFunction {

  // Use GSON (https://github.com/google/gson) to parse JSON content.
  private static final Gson gson = new Gson();

  // Responds to an HTTP request using data from the request body parsed according to the
  // "content-type" header.
  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException {
    String name = null;

    // Default values avoid null issues (with switch/case) and exceptions from get() (optionals)
    String contentType = request.getContentType().orElse("");

    switch (contentType) {
      case "application/json":
        // '{"name":"John"}'
        JsonObject body = gson.fromJson(request.getReader(), JsonObject.class);
        if (body.has("name")) {
          name = body.get("name").getAsString();
        }
        break;
      case "application/octet-stream":
        // 'John', stored in a Buffer
        name = new String(Base64.getDecoder().decode(request.getInputStream().readAllBytes()),
            StandardCharsets.UTF_8);
        break;
      case "text/plain":
        // 'John'
        name = request.getReader().readLine();
        break;
      case "application/x-www-form-urlencoded":
        // 'name=John' in the body of a POST request (not the URL)
        Optional<String> nameParam = request.getFirstQueryParameter("name");
        if (nameParam.isPresent()) {
          name = nameParam.get();
        }
        break;
      default:
        // Invalid or missing "Content-Type" header
        response.setStatusCode(HttpURLConnection.HTTP_UNSUPPORTED_TYPE);
        return;
    }

    // Verify that a name was provided
    if (name == null) {
      response.setStatusCode(HttpURLConnection.HTTP_BAD_REQUEST);
    }

    // Respond with a name
    var writer = new PrintWriter(response.getWriter());
    writer.printf("Hello %s!", name);
  }
}

C#

functions/http/ParseContentType/Function.cs 
using Google.Cloud.Functions.Framework;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Primitives;
using System;
using System.IO;
using System.Linq;
using System.Net;
using System.Net.Mime;
using System.Text;
using System.Text.Json;
using System.Threading.Tasks;

namespace ParseContentType
{
    public class Function : IHttpFunction
    {
        public async Task HandleAsync(HttpContext context)
        {
            HttpRequest request = context.Request;
            HttpResponse response = context.Response;

            string name = null;
            ContentType contentType = new ContentType(request.ContentType);

            switch (contentType.MediaType)
            {
                case "application/json":
                {
                    // '{"name":"John"}'
                    using TextReader reader = new StreamReader(request.Body);
                    string json = await reader.ReadToEndAsync();
                    JsonElement body = JsonSerializer.Deserialize<JsonElement>(json);
                    if (body.TryGetProperty("name", out JsonElement property) && property.ValueKind == JsonValueKind.String)
                    {
                        name = property.GetString();
                    }
                    break;
                }
                case "application/octet-stream":
                {
                    // 'John', encoded to bytes using UTF-8, then encoded as base64
                    using TextReader reader = new StreamReader(request.Body);
                    string base64 = await reader.ReadToEndAsync();
                    byte[] data = Convert.FromBase64String(base64);
                    name = Encoding.UTF8.GetString(data);
                    break;
                }
                case "text/plain":
                {
                    // 'John'
                    using TextReader reader = new StreamReader(request.Body);
                    name = await reader.ReadLineAsync();
                    break;
                }
                case "application/x-www-form-urlencoded":
                {
                    // 'name=John' in the body of a POST request (not the URL)
                    if (request.Form.TryGetValue("name", out StringValues value))
                    {
                        name = value;
                    }
                    break;
                }
            }
            if (name is object)
            {
                await response.WriteAsync($"Hello {name}!");
            }
            else
            {
                // Unrecognized content type, or the name wasn't in the content
                // (e.g. JSON without a "name" property)
                response.StatusCode = (int) HttpStatusCode.BadRequest;
            }
        }
    }
}

Ruby

functions/http/content/app.rb 
require "functions_framework"
require "json"

FunctionsFramework.http "http_content" do |request|
  # The request parameter is a Rack::Request object.
  # See https://www.rubydoc.info/gems/rack/Rack/Request
  content_type = request.content_type
  case content_type
  # '{"name":"John"}'
  when "application/json"
    name = (JSON.parse(request.body.read.to_s)["name"] rescue nil)
  # "John", stored in a Buffer
  when "application/octet-stream"
    name = request.body.read.to_s # Convert buffer to a string
  # "John"
  when "text/plain"
    name = request.body.read.to_s
  # "name=John" in the body of a POST request (not the URL)
  when "application/x-www-form-urlencoded"
    name = (request.params["name"] rescue nil)
  end

  name ||= "World"

  # Return the response body as a string.
  # You can also return a Rack::Response object, a Rack response array, or
  # a hash which will be JSON-encoded into a response.
  "Hello #{name}!"
end

PHP

functions/http_content_type/index.php 

use Psr\Http\Message\ServerRequestInterface;

function helloContent(ServerRequestInterface $request): string
{
    $name = 'World';
    $body = $request->getBody()->getContents();
    switch ($request->getHeaderLine('content-type')) {
        // '{"name":"John"}'
        case 'application/json':
          if (!empty($body)) {
              $json = json_decode($body, true);
              if (json_last_error() != JSON_ERROR_NONE) {
                  throw new RuntimeException(sprintf(
                      'Could not parse body: %s',
                      json_last_error_msg()
                  ));
              }
              $name = $json['name'] ?? $name;
          }
          break;
        // 'John', stored in a stream
        case 'application/octet-stream':
          $name = $body;
          break;
        // 'John'
        case 'text/plain':
          $name = $body;
          break;
        // 'name=John' in the body of a POST request (not the URL)
        case 'application/x-www-form-urlencoded':
          parse_str($body, $data);
          $name = $data['name'] ?? $name;
          break;
    }

    return sprintf('Hello %s!', htmlspecialchars($name));
}

Handling CORS requests

Cross-Origin Resource Sharing (CORS) is a way to let applications running on one domain access content from another domain, for example, letting yourdomain.com make requests to region-project.cloudfunctions.net/yourfunction.

If CORS isn't set up properly, you're likely to get errors that look like this: 

XMLHttpRequest cannot load https://region-project.cloudfunctions.net/function.
No 'Access-Control-Allow-Origin' header is present on the requested resource.
Origin 'http://yourdomain.com' is therefore not allowed access.

Preflight request

CORS consists of two requests:

  • A preflight OPTIONS request.
  • A main request that follows the OPTIONS request.

The preflight request contains headers indicating which method (Access-Control-Request-Method) and which additional headers (Access-Control-Request-Headers) will be sent in the main request, as well as the origin of the main request (Origin).

To handle a preflight request, you must set the appropriate Access-Control-Allow-* headers to match the requests you want to accept:

Node.js

functions/http/index.js 
/**
 * HTTP function that supports CORS requests.
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
exports.corsEnabledFunction = (req, res) => {
  // Set CORS headers for preflight requests
  // Allows GETs from any origin with the Content-Type header
  // and caches preflight response for 3600s

  res.set('Access-Control-Allow-Origin', '*');

  if (req.method === 'OPTIONS') {
    // Send response to OPTIONS requests
    res.set('Access-Control-Allow-Methods', 'GET');
    res.set('Access-Control-Allow-Headers', 'Content-Type');
    res.set('Access-Control-Max-Age', '3600');
    res.status(204).send('');
  } else {
    res.send('Hello World!');
  }
};

Python

functions/http/main.py 
def cors_enabled_function(request):
    # For more information about CORS and CORS preflight requests, see:
    # https://developer.mozilla.org/en-US/docs/Glossary/Preflight_request

    # Set CORS headers for the preflight request
    if request.method == 'OPTIONS':
        # Allows GET requests from any origin with the Content-Type
        # header and caches preflight response for an 3600s
        headers = {
            'Access-Control-Allow-Origin': '*',
            'Access-Control-Allow-Methods': 'GET',
            'Access-Control-Allow-Headers': 'Content-Type',
            'Access-Control-Max-Age': '3600'
        }

        return ('', 204, headers)

    # Set CORS headers for the main request
    headers = {
        'Access-Control-Allow-Origin': '*'
    }

    return ('Hello World!', 200, headers)

Go

functions/http/cors.go 

// Package http provides a set of HTTP Cloud Functions samples.
package http

import (
	"fmt"
	"net/http"
)

// CORSEnabledFunction is an example of setting CORS headers.
// For more information about CORS and CORS preflight requests, see
// https://developer.mozilla.org/en-US/docs/Glossary/Preflight_request.
func CORSEnabledFunction(w http.ResponseWriter, r *http.Request) {
	// Set CORS headers for the preflight request
	if r.Method == http.MethodOptions {
		w.Header().Set("Access-Control-Allow-Origin", "*")
		w.Header().Set("Access-Control-Allow-Methods", "POST")
		w.Header().Set("Access-Control-Allow-Headers", "Content-Type")
		w.Header().Set("Access-Control-Max-Age", "3600")
		w.WriteHeader(http.StatusNoContent)
		return
	}
	// Set CORS headers for the main request.
	w.Header().Set("Access-Control-Allow-Origin", "*")
	fmt.Fprint(w, "Hello, World!")
}

Java

functions/http/cors-enabled/src/main/java/functions/CorsEnabled.java 

import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import java.io.BufferedWriter;
import java.io.IOException;
import java.net.HttpURLConnection;

public class CorsEnabled implements HttpFunction {
  // corsEnabled is an example of setting CORS headers.
  // For more information about CORS and CORS preflight requests, see
  // https://developer.mozilla.org/en-US/docs/Glossary/Preflight_request.
  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException {
    // Set CORS headers
    //   Allows GETs from any origin with the Content-Type
    //   header and caches preflight response for 3600s
    response.appendHeader("Access-Control-Allow-Origin", "*");

    if ("OPTIONS".equals(request.getMethod())) {
      response.appendHeader("Access-Control-Allow-Methods", "GET");
      response.appendHeader("Access-Control-Allow-Headers", "Content-Type");
      response.appendHeader("Access-Control-Max-Age", "3600");
      response.setStatusCode(HttpURLConnection.HTTP_NO_CONTENT);
      return;
    }

    // Handle the main request.
    BufferedWriter writer = response.getWriter();
    writer.write("CORS headers set successfully!");
  }
}

C#

functions/http/Cors/Function.cs 
using Google.Cloud.Functions.Framework;
using Microsoft.AspNetCore.Http;
using System.Net;
using System.Threading.Tasks;

namespace Cors
{
    // For more information about CORS and CORS preflight requests, see
    // https://developer.mozilla.org/en-US/docs/Glossary/Preflight_request.
    public class Function : IHttpFunction
    {
        public async Task HandleAsync(HttpContext context)
        {
            HttpRequest request = context.Request;
            HttpResponse response = context.Response;

            // Set CORS headers
            //   Allows GETs from any origin with the Content-Type
            //   header and caches preflight response for 3600s

            response.Headers.Append("Access-Control-Allow-Origin", "*");
            if (HttpMethods.IsOptions(request.Method))
            {
                response.Headers.Append("Access-Control-Allow-Methods", "GET");
                response.Headers.Append("Access-Control-Allow-Headers", "Content-Type");
                response.Headers.Append("Access-Control-Max-Age", "3600");
                response.StatusCode = (int) HttpStatusCode.NoContent;
                return;
            }

            await response.WriteAsync("CORS headers set successfully!");
        }
    }
}

Ruby

functions/http/cors/app.rb 
FunctionsFramework.http "cors_enabled_function" do |request|
  # For more information about CORS and CORS preflight requests, see
  # https://developer.mozilla.org/en-US/docs/Glossary/Preflight_request
  # for more information.

  # Set CORS headers for the preflight request
  if request.options?
    # Allows GET requests from any origin with the Content-Type
    # header and caches preflight response for an 3600s
    headers = {
      "Access-Control-Allow-Origin"  => "*",
      "Access-Control-Allow-Methods" => "GET",
      "Access-Control-Allow-Headers" => "Content-Type",
      "Access-Control-Max-Age"       => "3600"
    }
    [204, headers, []]
  else
    # Set CORS headers for the main request
    headers = {
      "Access-Control-Allow-Origin" => "*"
    }

    [200, headers, ["Hello World!"]]
  end
end

PHP

functions/http_cors/index.php 

use Psr\Http\Message\ResponseInterface;
use Psr\Http\Message\ServerRequestInterface;
use GuzzleHttp\Psr7\Response;

function corsEnabledFunction(ServerRequestInterface $request): ResponseInterface
{
    // Set CORS headers for preflight requests
    // Allows GETs from any origin with the Content-Type header
    // and caches preflight response for 3600s
    $headers = ['Access-Control-Allow-Origin' => '*'];

    if ($request->getMethod() === 'OPTIONS') {
        // Send response to OPTIONS requests
        $headers = array_merge($headers, [
            'Access-Control-Allow-Methods' => 'GET',
            'Access-Control-Allow-Headers' => 'Content-Type',
            'Access-Control-Max-Age' => '3600'
        ]);
        return new Response(204, $headers, '');
    } else {
        return new Response(200, $headers, 'Hello World!');
    }
}

Alternatively, you can use a third-party library to handle CORS for you.

Limitations

CORS preflight requests are sent without an Authorization header, so they will be rejected on all non-public HTTP Functions. Because the preflight requests fail, the main request will also fail.

HTTP functions require authentication by default, so here are your options for working around this limitation:

Hosting on the same domain

Instead of implementing CORS, you could instead host your website and your functions on the same domain. Since requests would now come from the same origin, CORS won't be enforced. This simplifies your code considerably.

The easiest way to do this is to integrate Firebase Hosting with Google Cloud Functions.

Using Cloud Endpoints to handle CORS

You can deploy a Cloud Endpoints proxy and enable CORS.

If you want authentication capabilities, you can also enable Google ID token validation, which will validate authentication tokens.

Handling HTTP methods

HTTP functions accept all HTTP methods. The following sample shows how to perform different actions based on the HTTP method received (for example, GET and PUT):

Node.js

functions/http/index.js 
/**
 * Responds to a GET request with "Hello World!". Forbids a PUT request.
 *
 * @example
 * gcloud functions call helloHttp
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
exports.helloHttp = (req, res) => {
  switch (req.method) {
    case 'GET':
      res.status(200).send('Hello World!');
      break;
    case 'PUT':
      res.status(403).send('Forbidden!');
      break;
    default:
      res.status(405).send({error: 'Something blew up!'});
      break;
  }
};

Python

functions/helloworld/main.py 
import functions_framework

@functions_framework.http
def hello_method(request):
    """ Responds to a GET request with "Hello world!". Forbids a PUT request.
    Args:
        request (flask.Request): The request object.
        <https://flask.palletsprojects.com/en/1.1.x/api/#incoming-request-data>
    Returns:
        The response text, or any set of values that can be turned into a
        Response object using `make_response`
        <https://flask.palletsprojects.com/en/1.1.x/api/#flask.make_response>.
    """
    from flask import abort

    if request.method == 'GET':
        return 'Hello World!'
    elif request.method == 'PUT':
        return abort(403)
    else:
        return abort(405)

Go

functions/http/http_request_method.go 

// Package http provides a set of HTTP Cloud Functions samples.
package http

import (
	"fmt"
	"net/http"
)

// HelloHTTPMethod is an HTTP Cloud function.
// It uses the request method to differentiate the response.
func HelloHTTPMethod(w http.ResponseWriter, r *http.Request) {
	switch r.Method {
	case http.MethodGet:
		fmt.Fprint(w, "Hello World!")
	case http.MethodPut:
		http.Error(w, "403 - Forbidden", http.StatusForbidden)
	default:
		http.Error(w, "405 - Method Not Allowed", http.StatusMethodNotAllowed)
	}
}

Java

functions/http/http-method/src/main/java/functions/HttpMethod.java 

import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import java.io.BufferedWriter;
import java.io.IOException;
import java.net.HttpURLConnection;

public class HttpMethod implements HttpFunction {
  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException {

    BufferedWriter writer = response.getWriter();

    switch (request.getMethod()) {
      case "GET":
        response.setStatusCode(HttpURLConnection.HTTP_OK);
        writer.write("Hello world!");
        break;
      case "PUT":
        response.setStatusCode(HttpURLConnection.HTTP_FORBIDDEN);
        writer.write("Forbidden!");
        break;
      default:
        response.setStatusCode(HttpURLConnection.HTTP_BAD_METHOD);
        writer.write("Something blew up!");
        break;
    }
  }
}

C#

functions/http/HttpRequestMethod/Function.cs 
using Google.Cloud.Functions.Framework;
using Microsoft.AspNetCore.Http;
using System.Net;
using System.Threading.Tasks;

namespace HttpRequestMethod
{
    public class Function : IHttpFunction
    {
        public async Task HandleAsync(HttpContext context)
        {
            HttpResponse response = context.Response;
            switch (context.Request.Method)
            {
                case "GET":
                    response.StatusCode = (int) HttpStatusCode.OK;
                    await response.WriteAsync("Hello world!");
                    break;
                case "PUT":
                    response.StatusCode = (int) HttpStatusCode.Forbidden;
                    await response.WriteAsync("Forbidden!");
                    break;
                default:
                    response.StatusCode = (int) HttpStatusCode.MethodNotAllowed;
                    await response.WriteAsync("Something blew up!");
                    break;
            }
        }
    }
}

Ruby

functions/http/method/app.rb 
require "functions_framework"
require "json"

FunctionsFramework.http "http_method" do |request|
  # The request parameter is a Rack::Request object.
  # See https://www.rubydoc.info/gems/rack/Rack/Request
  case request.request_method
  when "GET"
    status = 200
    body = "Hello World!"
  when "PUT"
    status = 403
    body = "Forbidden!"
  else
    status = 405
    body = '{"error":"Something blew up!"}'
  end

  # Return the response body as a Rack::Response object.
  ::Rack::Response.new body, status
end

PHP

functions/http_method/index.php 

use Psr\Http\Message\ResponseInterface;
use Psr\Http\Message\ServerRequestInterface;
use GuzzleHttp\Psr7\Response;

function httpMethod(ServerRequestInterface $request): ResponseInterface
{
    switch ($request->getMethod()) {
        case 'GET':
            // Example: read request
            return new Response(
                200, // OK
                [],
                'Hello, World!' . PHP_EOL
            );
            break;
        case 'PUT':
            // Example: write request to a read-only resource
            return new Response(
                403, // Permission denied
                [],
                'Forbidden!' . PHP_EOL
            );
            break;
        default:
            // Example: request type not supported by the application
            $json_payload = json_encode([
                'error' => 'something blew up!'
            ]);
            return new Response(
                405, // Method not allowed
                ['Content-Type' => 'application/json'],
                $json_payload
            );
            break;
    }
}

Handling Content Types

For Node.js, Cloud Functions parses request body content types of application/json and application/x-www-form-urlencoded as shown above. Plain text content types (text/plain) are passed through as strings using UTF-8 as a default encoding (or a custom encoding provided in the content-type header).

Other content types can be accessed by inspecting your HTTP function's argument. Methods for doing this vary by language.

The example below handles a request with a content type of text/xml:

Node.js

The rawBody property contains the unparsed bytes of the request body.

functions/http/index.js 
/**
 * Parses a document of type 'text/xml'
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
exports.parseXML = (req, res) => {
  // Convert the request to a Buffer and a string
  // Use whichever one is accepted by your XML parser
  const data = req.rawBody;
  const xmlData = data.toString();

  const {parseString} = require('xml2js');

  parseString(xmlData, (err, result) => {
    if (err) {
      console.error(err);
      res.status(500).end();
      return;
    }
    res.send(result);
  });
};

Python

functions/http/main.py 
import json
import xmltodict

def parse_xml(request):
    """ Parses a document of type 'text/xml'
    Args:
        request (flask.Request): The request object.
    Returns:
        The response text, or any set of values that can be turned into a
         Response object using `make_response`
        <http://flask.pocoo.org/docs/1.0/api/#flask.Flask.make_response>.
    """
    data = xmltodict.parse(request.data)
    return json.dumps(data, indent=2)

Go

functions/http/xml.go 

// Package http provides a set of HTTP Cloud Functions samples.
package http

import (
	"encoding/xml"
	"fmt"
	"html"
	"io/ioutil"
	"net/http"
)

// ParseXML is an example of parsing a text/xml request.
func ParseXML(w http.ResponseWriter, r *http.Request) {
	var d struct {
		Name string
	}
	b, err := ioutil.ReadAll(r.Body)
	if err != nil {
		http.Error(w, "Could not read request", http.StatusBadRequest)
	}
	if err := xml.Unmarshal(b, &d); err != nil {
		http.Error(w, "Could not parse request", http.StatusBadRequest)
	}
	if d.Name == "" {
		d.Name = "World"
	}
	fmt.Fprintf(w, "Hello, %v!", html.EscapeString(d.Name))
}

Java

functions/http/parse-xml/src/main/java/functions/ParseXml.java 
import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import java.io.ByteArrayInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.PrintWriter;
import java.net.HttpURLConnection;
import javax.xml.parsers.DocumentBuilder;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.parsers.ParserConfigurationException;
import org.w3c.dom.Document;
import org.xml.sax.SAXException;

public class ParseXml implements HttpFunction {
  private static DocumentBuilderFactory dbFactory = DocumentBuilderFactory.newInstance();

  // Parses a HTTP request in XML format
  // (Responds with a 400 error if the HTTP request isn't valid XML.)
  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException, ParserConfigurationException {

    try {
      DocumentBuilder docBuilder = dbFactory.newDocumentBuilder();
      var writer = new PrintWriter(response.getWriter());

      // Get request body
      InputStream bodyStream = new ByteArrayInputStream(
          request.getInputStream().readAllBytes());

      // Parse + process XML
      Document doc = docBuilder.parse(bodyStream);
      writer.printf("Root element: %s", doc.getDocumentElement().getNodeName());
    } catch (SAXException e) {
      response.setStatusCode(HttpURLConnection.HTTP_BAD_REQUEST);
      return;
    }
  }
}

Multipart Data

The following example shows how to process data with a multipart/form-data content type. Depending on your chosen language, you may need to use a parsing library.

Node.js

functions/http/index.js 
/**
 * Parses a 'multipart/form-data' upload request
 *
 * @param {Object} req Cloud Function request context.
 * @param {Object} res Cloud Function response context.
 */
const path = require('path');
const os = require('os');
const fs = require('fs');

// Node.js doesn't have a built-in multipart/form-data parsing library.
// Instead, we can use the 'busboy' library from NPM to parse these requests.
const Busboy = require('busboy');

exports.uploadFile = (req, res) => {
  if (req.method !== 'POST') {
    // Return a "method not allowed" error
    return res.status(405).end();
  }
  const busboy = Busboy({headers: req.headers});
  const tmpdir = os.tmpdir();

  // This object will accumulate all the fields, keyed by their name
  const fields = {};

  // This object will accumulate all the uploaded files, keyed by their name.
  const uploads = {};

  // This code will process each non-file field in the form.
  busboy.on('field', (fieldname, val) => {
    /**
     *  TODO(developer): Process submitted field values here
     */
    console.log(`Processed field ${fieldname}: ${val}.`);
    fields[fieldname] = val;
  });

  const fileWrites = [];

  // This code will process each file uploaded.
  busboy.on('file', (fieldname, file, {filename}) => {
    // Note: os.tmpdir() points to an in-memory file system on GCF
    // Thus, any files in it must fit in the instance's memory.
    console.log(`Processed file ${filename}`);
    const filepath = path.join(tmpdir, filename);
    uploads[fieldname] = filepath;

    const writeStream = fs.createWriteStream(filepath);
    file.pipe(writeStream);

    // File was processed by Busboy; wait for it to be written.
    // Note: GCF may not persist saved files across invocations.
    // Persistent files must be kept in other locations
    // (such as Cloud Storage buckets).
    const promise = new Promise((resolve, reject) => {
      file.on('end', () => {
        writeStream.end();
      });
      writeStream.on('finish', resolve);
      writeStream.on('error', reject);
    });
    fileWrites.push(promise);
  });

  // Triggered once all uploaded files are processed by Busboy.
  // We still need to wait for the disk writes (saves) to complete.
  busboy.on('finish', async () => {
    await Promise.all(fileWrites);

    /**
     * TODO(developer): Process saved files here
     */
    for (const file in uploads) {
      fs.unlinkSync(uploads[file]);
    }
    res.send();
  });

  busboy.end(req.rawBody);
};

Python

functions/http/main.py 
import os
import tempfile
from werkzeug.utils import secure_filename

# Helper function that computes the filepath to save files to
def get_file_path(filename):
    # Note: tempfile.gettempdir() points to an in-memory file system
    # on GCF. Thus, any files in it must fit in the instance's memory.
    file_name = secure_filename(filename)
    return os.path.join(tempfile.gettempdir(), file_name)


def parse_multipart(request):
    """ Parses a 'multipart/form-data' upload request
    Args:
        request (flask.Request): The request object.
    Returns:
        The response text, or any set of values that can be turned into a
         Response object using `make_response`
        <http://flask.pocoo.org/docs/1.0/api/#flask.Flask.make_response>.
    """

    # This code will process each non-file field in the form
    fields = {}
    data = request.form.to_dict()
    for field in data:
        fields[field] = data[field]
        print('Processed field: %s' % field)

    # This code will process each file uploaded
    files = request.files.to_dict()
    for file_name, file in files.items():
        # Note: GCF may not keep files saved locally between invocations.
        # If you want to preserve the uploaded files, you should save them
        # to another location (such as a Cloud Storage bucket).
        file.save(get_file_path(file_name))
        print('Processed file: %s' % file_name)

    # Clear temporary directory
    for file_name in files:
        file_path = get_file_path(file_name)
        os.remove(file_path)

    return "Done!"

Go

functions/http/form.go 

// Package http provides a set of HTTP Cloud Functions samples.
package http

import (
	"fmt"
	"log"
	"net/http"
)

// UploadFile processes a 'multipart/form-data' upload request.
func UploadFile(w http.ResponseWriter, r *http.Request) {
	const maxMemory = 2 * 1024 * 1024 // 2 megabytes.

	// ParseMultipartForm parses a request body as multipart/form-data.
	// The whole request body is parsed and up to a total of maxMemory bytes of
	// its file parts are stored in memory, with the remainder stored on
	// disk in temporary files.

	// Note that any files saved during a particular invocation may not
	// persist after the current invocation completes; persistent files
	// should be stored elsewhere, such as in a Cloud Storage bucket.
	if err := r.ParseMultipartForm(maxMemory); err != nil {
		http.Error(w, "Unable to parse form", http.StatusBadRequest)
		log.Printf("Error parsing form: %v", err)
		return
	}

	// Be sure to remove all temporary files after your function is finished.
	defer func() {
		if err := r.MultipartForm.RemoveAll(); err != nil {
			http.Error(w, "Error cleaning up form files", http.StatusInternalServerError)
			log.Printf("Error cleaning up form files: %v", err)
		}
	}()

	// r.MultipartForm.File contains *multipart.FileHeader values for every
	// file in the form. You can access the file contents using
	// *multipart.FileHeader's Open method.
	for _, headers := range r.MultipartForm.File {
		for _, h := range headers {
			fmt.Fprintf(w, "File uploaded: %q (%v bytes)", h.Filename, h.Size)
			// Use h.Open() to read the contents of the file.
		}
	}

}

Java

functions/http/http-form-data/src/main/java/functions/HttpFormData.java 

import com.google.cloud.functions.HttpFunction;
import com.google.cloud.functions.HttpRequest;
import com.google.cloud.functions.HttpResponse;
import java.io.IOException;
import java.net.HttpURLConnection;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.nio.file.StandardCopyOption;
import java.util.logging.Logger;

public class HttpFormData implements HttpFunction {
  private static final Logger logger = Logger.getLogger(HttpFormData.class.getName());

  @Override
  public void service(HttpRequest request, HttpResponse response)
      throws IOException {

    if (!"POST".equals(request.getMethod())) {
      response.setStatusCode(HttpURLConnection.HTTP_BAD_METHOD);
      return;
    }

    // This code will process each file uploaded.
    String tempDirectory = System.getProperty("java.io.tmpdir");
    for (HttpRequest.HttpPart httpPart : request.getParts().values()) {
      String filename = httpPart.getFileName().orElse(null);
      if (filename == null) {
        continue;
      }

      logger.info("Processed file: " + filename);

      // Note: GCF's temp directory is an in-memory file system
      // Thus, any files in it must fit in the instance's memory.
      Path filePath = Paths.get(tempDirectory, filename).toAbsolutePath();

      // Note: files saved to a GCF instance itself may not persist across executions.
      // Persistent files should be stored elsewhere, e.g. a Cloud Storage bucket.
      Files.copy(httpPart.getInputStream(), filePath, StandardCopyOption.REPLACE_EXISTING);

      // TODO(developer): process saved files here
      Files.delete(filePath);
    }

    // This code will process other form fields.
    request.getQueryParameters().forEach(
        (fieldName, fieldValues) -> {
          String firstFieldValue = fieldValues.get(0);

          // TODO(developer): process field values here
          logger.info(String.format(
              "Processed field: %s (value: %s)", fieldName, firstFieldValue));
        });
  }
}

C#

functions/http/HttpFormData/Function.cs 
using Google.Cloud.Functions.Framework;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Logging;
using Microsoft.Extensions.Primitives;
using System.Collections.Generic;
using System.IO;
using System.Net;
using System.Threading.Tasks;

namespace HttpFormData
{
    public class Function : IHttpFunction
    {
        private readonly ILogger _logger;

        public Function(ILogger<Function> logger) =>
            _logger = logger;

        public async Task HandleAsync(HttpContext context)
        {
            HttpResponse response = context.Response;
            HttpRequest request = context.Request;

            if (request.Method != "POST")
            {
                response.StatusCode = (int) HttpStatusCode.MethodNotAllowed;
                return;
            }

            // This code will process each file uploaded.
            string tempDirectory = Path.GetTempPath();
            foreach (IFormFile file in request.Form.Files)
            {
                if (string.IsNullOrEmpty(file.FileName))
                {
                    continue;
                }
                _logger.LogInformation("Processed file: {file}", file.FileName);

                // Note: GCF's temp directory is an in-memory file system
                // Thus, any files in it must fit in the instance's memory.
                string outputPath = Path.Combine(tempDirectory, file.FileName);

                // Note: files saved to a GCF instance itself may not persist across executions.
                // Persistent files should be stored elsewhere, e.g. a Cloud Storage bucket.
                using (FileStream output = File.Create(outputPath))
                {
                    await file.CopyToAsync(output);
                }

                // TODO(developer): process saved files here
                File.Delete(outputPath);
            }

            // This code will process other form fields.
            foreach (KeyValuePair<string, StringValues> parameter in request.Form)
            {
                // TODO(developer): process field values here
                _logger.LogInformation("Processed field '{key}' (value: '{value}')",
                    parameter.Key, (string) parameter.Value);
            }
        }
    }
}

Ruby

functions/http/form_data/app.rb 
require "functions_framework"

FunctionsFramework.http "http_form_data" do |request|
  # The request parameter is a Rack::Request object.
  # See https://www.rubydoc.info/gems/rack/Rack/Request

  # This Rack call parses multipart form data, returning the params as a hash.
  # The returned params hash includes an entry for each field. File uploads
  # are written to Tempfiles and represented by hashes, while plain fields are
  # represented by strings.
  params = request.POST

  begin
    params.each do |name, part|
      if part.is_a? Hash
        # Handle a file upload part by logging the md5 hash.
        md5 = Digest::MD5.hexdigest part[:tempfile].read
        file_name = part[:filename]
        logger.info "Processed file=#{file_name} md5=#{md5}"
      else
        # Handle a non-file part by logging the value.
        logger.info "Processed field=#{name} value=#{part}"
      end
    end
  ensure
    # Ensure that all Tempfile objects are closed and deleted. The Cloud
    # Functions runtime keeps temporary files in an in-memory file system,
    # so to lower memory usage it is good practice to clean up Tempfiles
    # explicitly rather than wait for object finalization.
    params.each_value do |part|
      part[:tempfile].close! if part.is_a? Hash
    end
  end

  # The HTTP response body.
  "OK"
end

PHP

functions/http_form_data/index.php 

use Psr\Http\Message\ServerRequestInterface;
use Psr\Http\Message\ResponseInterface;
use GuzzleHttp\Psr7\Response;

function uploadFile(ServerRequestInterface $request): ResponseInterface
{
    if ($request->getMethod() != 'POST') {
        return new Response(405, [], 'Method Not Allowed: expected POST, found ' . $request->getMethod());
    }

    $contentType = $request->getHeader('Content-Type')[0];
    if (strpos($contentType, 'multipart/form-data') !== 0) {
        return new Response(400, [], 'Bad Request: content of type "multipart/form-data" not provided, found ' . $contentType);
    }

    $fileList = [];
    /** @var $file Psr\Http\Message\UploadedFileInterface */
    foreach ($request->getUploadedFiles() as $name => $file) {
        // Use caution when trusting the client-provided filename:
        // https://owasp.org/www-community/vulnerabilities/Unrestricted_File_Upload
        $fileList[] = $file->getClientFilename();

        infoLog('Processing ' . $file->getClientFilename());
        $filename = tempnam(sys_get_temp_dir(), $name . '.') . '-' . $file->getClientFilename();

        // Use $file->getStream() to process the file contents in ways other than a direct "file save".
        infoLog('Saving to ' . $filename);
        $file->moveTo($filename);
    }

    if (empty($fileList)) {
        $msg = 'Bad Request: no files sent for upload';
        errorLog($msg);
        return new Response(400, [], $msg);
    }

    return new Response(201, [], 'Saved ' . join(', ', $fileList));
}

function errorLog($msg): void
{
    $stream = fopen('php://stderr', 'wb');
    $entry = json_encode(['msg' => $msg, 'severity' => 'error'], JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES);
    fwrite($stream, $entry . PHP_EOL);
}

function infoLog($msg): void
{
    $stream = fopen('php://stderr', 'wb');
    $entry = json_encode(['message' => $msg, 'severity' => 'info'], JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES);
    fwrite($stream, $entry . PHP_EOL);
}

Uploading Files via Cloud Storage

A common use case for Cloud Functions is file processing. For larger files or those that require persistent storage beyond the scope of a single request, you can use Cloud Storage as an entry point for your file uploads. To accomplish this, you must generate a Signed URL, which provides temporary write access to a Cloud Storage bucket.

If you're using Cloud Functions directly, generate a signed URL using the appropriate Cloud Storage client library.

Uploading files to a Cloud Function using Cloud Storage is a three step process:

  1. Clients call a Cloud Function directly to retrieve a signed URL.

  2. Clients then send file data to the signed URL via an HTTP PUT request.

  3. A second Cloud Function is triggered by the mutation in the storage bucket to further process the file.

You can see an example below of using the Cloud Storage client library to generate a signed URL.

Cloud Functions have a "default application credential" which typically does not include the iam.serviceAccounts.signBlob permission. To allow this, you'll need to first make sure that your function's service account has the appropriate role. You can achieve this using either the Google Cloud console or the gcloud command-line tool:

console

To make sure that your function's service account has the appropriate role, you can directly modify the IAM roles for an account:

  1. Go to the Google Cloud console:

    Go to Google Cloud console

  2. Select the appropriate account, and choose Editor > Service Accounts > Service Account Token Creator.

gcloud

To make sure that your function's service account has the appropriate role, run the following command. The pre-defined serviceAccountTokenCreator role has the iam.serviceAccounts.signBlob permission you need:

gcloud projects add-iam-policy-binding YOUR_PROJECT \
--member serviceAccount:YOUR_SERVICE_ACCOUNT --role roles/iam.serviceAccountTokenCreator

You can determine the service account used by your functions using either the Google Cloud console or the gcloud command-line tool:

console

To determine the service account used by your functions using Google Cloud console:

  1. Go to the Google Cloud console:

    Go to Google Cloud console

  2. Select the function you want to inspect from the list.

You can see the service account on the details page for the function.

gcloud

To determine the service account used by your functions, run the following command and look for the serviceAccountEmail property:

gcloud beta functions describe YOUR_FUNCTION_NAME

Here's an example of generating a signed URL:

Node.js

samples/generateV4UploadSignedUrl.js 
/**
 * TODO(developer): Uncomment the following lines before running the sample.
 */
// The ID of your GCS bucket
// const bucketName = 'your-unique-bucket-name';

// The full path of your file inside the GCS bucket, e.g. 'yourFile.jpg' or 'folder1/folder2/yourFile.jpg'
// const fileName = 'your-file-name';

// Imports the Google Cloud client library
const {Storage} = require('@google-cloud/storage');

// Creates a client
const storage = new Storage();

async function generateV4UploadSignedUrl() {
  // These options will allow temporary uploading of the file with outgoing
  // Content-Type: application/octet-stream header.
  const options = {
    version: 'v4',
    action: 'write',
    expires: Date.now() + 15 * 60 * 1000, // 15 minutes
    contentType: 'application/octet-stream',
  };

  // Get a v4 signed URL for uploading file
  const [url] = await storage
    .bucket(bucketName)
    .file(fileName)
    .getSignedUrl(options);

  console.log('Generated PUT signed URL:');
  console.log(url);
  console.log('You can use this URL with any user agent, for example:');
  console.log(
    "curl -X PUT -H 'Content-Type: application/octet-stream' " +
      `--upload-file my-file '${url}'`
  );
}

generateV4UploadSignedUrl().catch(console.error);

Python

samples/snippets/storage_generate_upload_signed_url_v4.py 
import datetime

from google.cloud import storage


def generate_upload_signed_url_v4(bucket_name, blob_name):
    """Generates a v4 signed URL for uploading a blob using HTTP PUT.

    Note that this method requires a service account key file. You can not use
    this if you are using Application Default Credentials from Google Compute
    Engine or from the Google Cloud SDK.
    """
    # bucket_name = 'your-bucket-name'
    # blob_name = 'your-object-name'

    storage_client = storage.Client()
    bucket = storage_client.bucket(bucket_name)
    blob = bucket.blob(blob_name)

    url = blob.generate_signed_url(
        version="v4",
        # This URL is valid for 15 minutes
        expiration=datetime.timedelta(minutes=15),
        # Allow PUT requests using this URL.
        method="PUT",
        content_type="application/octet-stream",
    )

    print("Generated PUT signed URL:")
    print(url)
    print("You can use this URL with any user agent, for example:")
    print(
        "curl -X PUT -H 'Content-Type: application/octet-stream' "
        "--upload-file my-file '{}'".format(url)
    )
    return url

Go

storage/objects/generate_v4_put_object_signed_url.go 
import (
	"context"
	"fmt"
	"io"
	"time"

	"cloud.google.com/go/storage"
)

// generateV4GetObjectSignedURL generates object signed URL with PUT method.
func generateV4PutObjectSignedURL(w io.Writer, bucket, object string) (string, error) {
	// bucket := "bucket-name"
	// object := "object-name"

	ctx := context.Background()
	client, err := storage.NewClient(ctx)
	if err != nil {
		return "", fmt.Errorf("storage.NewClient: %v", err)
	}
	defer client.Close()

	// Signing a URL requires credentials authorized to sign a URL. You can pass
	// these in through SignedURLOptions with one of the following options:
	//    a. a Google service account private key, obtainable from the Google Developers Console
	//    b. a Google Access ID with iam.serviceAccounts.signBlob permissions
	//    c. a SignBytes function implementing custom signing.
	// In this example, none of these options are used, which means the SignedURL
	// function attempts to use the same authentication that was used to instantiate
	// the Storage client. This authentication must include a private key or have
	// iam.serviceAccounts.signBlob permissions.
	opts := &storage.SignedURLOptions{
		Scheme: storage.SigningSchemeV4,
		Method: "PUT",
		Headers: []string{
			"Content-Type:application/octet-stream",
		},
		Expires: time.Now().Add(15 * time.Minute),
	}

	u, err := client.Bucket(bucket).SignedURL(object, opts)
	if err != nil {
		return "", fmt.Errorf("Bucket(%q).SignedURL: %v", bucket, err)
	}

	fmt.Fprintln(w, "Generated PUT signed URL:")
	fmt.Fprintf(w, "%q\n", u)
	fmt.Fprintln(w, "You can use this URL with any user agent, for example:")
	fmt.Fprintf(w, "curl -X PUT -H 'Content-Type: application/octet-stream' --upload-file my-file %q\n", u)
	return u, nil
}

Java

samples/snippets/src/main/java/com/example/storage/object/GenerateV4PutObjectSignedUrl.java 

import com.google.cloud.storage.BlobId;
import com.google.cloud.storage.BlobInfo;
import com.google.cloud.storage.HttpMethod;
import com.google.cloud.storage.Storage;
import com.google.cloud.storage.StorageException;
import com.google.cloud.storage.StorageOptions;
import java.net.URL;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.TimeUnit;

public class GenerateV4PutObjectSignedUrl {
  /**
   * Signing a URL requires Credentials which implement ServiceAccountSigner. These can be set
   * explicitly using the Storage.SignUrlOption.signWith(ServiceAccountSigner) option. If you don't,
   * you could also pass a service account signer to StorageOptions, i.e.
   * StorageOptions().newBuilder().setCredentials(ServiceAccountSignerCredentials). In this example,
   * neither of these options are used, which means the following code only works when the
   * credentials are defined via the environment variable GOOGLE_APPLICATION_CREDENTIALS, and those
   * credentials are authorized to sign a URL. See the documentation for Storage.signUrl for more
   * details.
   */
  public static void generateV4PutObjectSignedUrl(
      String projectId, String bucketName, String objectName) throws StorageException {
    // String projectId = "my-project-id";
    // String bucketName = "my-bucket";
    // String objectName = "my-object";

    Storage storage = StorageOptions.newBuilder().setProjectId(projectId).build().getService();

    // Define Resource
    BlobInfo blobInfo = BlobInfo.newBuilder(BlobId.of(bucketName, objectName)).build();

    // Generate Signed URL
    Map<String, String> extensionHeaders = new HashMap<>();
    extensionHeaders.put("Content-Type", "application/octet-stream");

    URL url =
        storage.signUrl(
            blobInfo,
            15,
            TimeUnit.MINUTES,
            Storage.SignUrlOption.httpMethod(HttpMethod.PUT),
            Storage.SignUrlOption.withExtHeaders(extensionHeaders),
            Storage.SignUrlOption.withV4Signature());

    System.out.println("Generated PUT signed URL:");
    System.out.println(url);
    System.out.println("You can use this URL with any user agent, for example:");
    System.out.println(
        "curl -X PUT -H 'Content-Type: application/octet-stream' --upload-file my-file '"
            + url
            + "'");
  }
}

C#

storage/api/Storage.Samples/GenerateV4UploadSignedUrl.cs 

using Google.Cloud.Storage.V1;
using System;
using System.Collections.Generic;
using System.Net.Http;

public class GenerateV4UploadSignedUrlSample
{
    public string GenerateV4UploadSignedUrl(
        string bucketName = "your-unique-bucket-name",
        string objectName = "your-object-name",
        string credentialFilePath = "my-local-path/my-credential-file-name")
    {
        UrlSigner urlSigner = UrlSigner.FromServiceAccountPath(credentialFilePath);

        var contentHeaders = new Dictionary<string, IEnumerable<string>>
        {
            { "Content-Type", new[] { "text/plain" } }
        };

        // V4 is the default signing version.
        UrlSigner.Options options = UrlSigner.Options.FromDuration(TimeSpan.FromHours(1));

        UrlSigner.RequestTemplate template = UrlSigner.RequestTemplate
            .FromBucket(bucketName)
            .WithObjectName(objectName)
            .WithHttpMethod(HttpMethod.Put)
            .WithContentHeaders(contentHeaders);

        string url = urlSigner.Sign(template, options);
        Console.WriteLine("Generated PUT signed URL:");
        Console.WriteLine(url);
        Console.WriteLine("You can use this URL with any user agent, for example:");
        Console.WriteLine($"curl -X PUT -H 'Content-Type: text/plain' --upload-file my-file '{url}'");
        return url;
    }
}

When the client uploads a file to the signed URL, you can trigger a second function from this mutation if you want to take further action on the upload. See the Cloud Storage Tutorial for more information on triggering a Cloud Function upon changes to a Cloud Storage bucket.

Next steps