Security Tip: Protecting Session Sensitive Responses with Retrofit 2 and okhttp 3

Problem Statement:

A user could receive and view data that is not theirs, and we must prevent this from happening in a secure application.  Here is a diagram of how this could happen:

Flow Diagram of the Session Mismatch Issue

 

Implementation options with Retrofit 2 and okhttp 3:

Option 1:

Write a custom Retrofit 2 CallAdapter that blocks the response from being processed or passes an exception to the onFailure() method.

  • PROS:
    • Great if you use Retrofit 2 exclusively for session sensitive calls.
    • This can prevent the callback from being invoked, or you could call Retrofit 2’s onFailure() method.
  • CONS:
    • Retrofit 2 CallAdapter is complex and you have to write one for both plain Callbacks as well as for RxJava (if you are using both).
    • We’d have to copy the existing CallAdapter implementation in retrofit2.ExecutorCallAdapterFactory and modify it as it’s a final class.
    • This only works for Retrofit 2 calls, but not all networking calls made through okhttp 3.

Option 2:

When the client session is invalidated, use okhttp 3’s Dispatcher to cancel all running and queued calls.

  • PROS:
    • Very effective. All calls would immediately be cancelled.
  • CONS:
    • If there is a call like “Update Profile” that is still occurring, it would be cancelled, which is not desirable.
    • If there is an unauthenticated call happening, you’d have to ensure that was being executed on non session sensitive Retrofit 2 or okhttp 3 instance to avoid unwanted termination.

Option 3:

Same as the previous option, but instead occurs when a new client session is started, instead of when an old one ends.

  • PROS:
    • Very effective. All calls would immediately be cancelled.
  • CONS:
    • You would have to ensure this code got run before you kicked off any session sensitive calls, otherwise they would be immediately cancelled.
    • If there is an unauthenticated call happening in the background, you’d have to ensure that was being executed on non session sensitive Retrofit 2 or okhttp 3 instance to avoid unwanted termination.

Option 4:

Using an okhttp 3 Network Interceptor to return back an HTTP Response with a custom response code like 999 and removing the response body.

  • PROS:
    • Will work for:
      • All okhttp 3 calls.
      • Retrofit 2 Callback calls.
      • Retrofit 2 RxJava calls.
  • CONS:
    • It will appear like the HTTP response came back from the server so you will have to have custom logic in your response handler to check and see if this response has a HTTP status code of 999.
    • It feels a little “hacky”… but would prevent the security hole.

Option 5: (My Choice)

Using an okhttp 3 Network Interceptor to throw a custom SessionMismatchException when this problem is detected, and handle the Exception appropriately.

  • PROS:
    • Will work for:
      • All okhttp 3 calls
      • Retrofit 2 Callback Calls
      • Retrofit 2 RxJava Calls
    • This is truly an “Exception” scenario, so this makes sense.
    • The request is executed fully, but will not be received by the client. (This could be a “con” depending on your use case)
  • CONS:
    • You must code your onFailure and onError handlers to appropriately handle this SessionMismatchException type.
    • If you just extend IOException, okhttp 3 will by default retry the HTTP call.  However, if you extend java.net.SocketTimeoutException, okhttp 3 will not retry.

My Choice: Option 5, throwing a custom SessionMismatchException.  This seemed to be the most flexible and intuitive since this truly is an Exception/Failure case.  I’ve provided a sample implementation of this below.

Assumption: You have a  “SessionManager” in your Android code which is aware of the current session.

Let me know if this was helpful, or if you’d do something different on Twitter at @HandstandSam

Controlling the Android Emulator and Virtual Devices (AVDs) via the command-line

I Googled all around to try and figure out how to tweak Android Emulator virtual devices (AVDs) via the command-line.  The Android Emulator documentation online has some options for running the emulator, but it didn’t allow me to specify memory at runtime, and many other things I wanted to customize.  There is also documentation specific to the creation of AVDs (vs just the running of them on the emulator), but that kept referring to the UI tools for it. I’ve gotten so used to Googling for help, that I didn’t think of running help locally :-p

emulator -help

That gives us a LOT of options that are not documented on the website including.

Android Emulator usage: emulator [options] [-qemu args]
  options:
    -list-avds                     list available AVDs
    -sysdir <dir>                  search for system disk images in <dir>
    -system <file>                 read initial system image from <file>
    -writable-system               make system image writable after 'adb remount'
    -datadir <dir>                 write user data into <dir>
    -kernel <file>                 use specific emulated kernel
    -ramdisk <file>                ramdisk image (default <system>/ramdisk.img
    -image <file>                  obsolete, use -system <file> instead
    -initdata <file>               same as '-init-data <file>'
    -data <file>                   data image (default <datadir>/userdata-qemu.img
    -partition-size <size>         system/data partition size in MBs
    -cache <file>                  cache partition image (default is temporary file)
    -cache-size <size>             cache partition size in MBs
    -no-cache                      disable the cache partition
    -nocache                       same as -no-cache
    -sdcard <file>                 SD card image (default <system>/sdcard.img
    -snapstorage <file>            file that contains all state snapshots (default <datadir>/snapshots.img)
    -no-snapstorage                do not mount a snapshot storage file (this disables all snapshot functionality)
    -snapshot <name>               name of snapshot within storage file for auto-start and auto-save (default 'default-boot')
    -no-snapshot                   perform a full boot and do not do not auto-save, but qemu vmload and vmsave operate on snapstorage
    -no-snapshot-save              do not auto-save to snapshot on exit: abandon changed state
    -no-snapshot-load              do not auto-start from snapshot: perform a full boot
    -snapshot-list                 show a list of available snapshots
    -no-snapshot-update-time       do not do try to correct snapshot time on restore
    -wipe-data                     reset the user data image (copy it from initdata)
    -avd <name>                    use a specific android virtual device
    -skindir <dir>                 search skins in <dir> (default <system>/skins)
    -skin <name>                   select a given skin
    -no-skin                       don't use any emulator skin
    -noskin                        same as -no-skin
    -dynamic-skin                  dynamically construct a skin of given size, requires -skin WxH option
    -memory <size>                 physical RAM size in MBs
    -accel <mode>                  Configure emulation acceleration
    -no-accel                      Same as '-accel off'
    -ranchu                        Use new emulator backend instead of the classic one
    -engine <engine>               Select engine. auto|classic|qemu2
    -netspeed <speed>              maximum network download/upload speeds
    -netdelay <delay>              network latency emulation
    -netfast                       disable network shaping
    -code-profile <name>           enable code profiling
    -show-kernel                   display kernel messages
    -shell                         enable root shell on current terminal
    -no-jni                        disable JNI checks in the Dalvik runtime
    -nojni                         same as -no-jni
    -logcat <tags>                 enable logcat output with given tags
    -no-audio                      disable audio support
    -noaudio                       same as -no-audio
    -audio <backend>               use specific audio backend
    -raw-keys                      disable Unicode keyboard reverse-mapping
    -radio <device>                redirect radio modem interface to character device
    -port <port>                   TCP port that will be used for the console
    -ports <consoleport>,<adbport> TCP ports used for the console and adb bridge
    -onion <image>                 use overlay PNG image over screen
    -onion-alpha <%age>            specify onion-skin translucency
    -onion-rotation 0|1|2|3        specify onion-skin rotation
    -dpi-device <dpi>              specify device's resolution in dpi (default 165)
    -scale <scale>                 scale emulator window (deprecated)
    -http-proxy <proxy>            make TCP connections through a HTTP/HTTPS proxy
    -timezone <timezone>           use this timezone instead of the host's default
    -dns-server <servers>          use this DNS server(s) in the emulated system
    -cpu-delay <cpudelay>          throttle CPU emulation
    -no-boot-anim                  disable animation for faster boot
    -no-window                     disable graphical window display
    -version                       display emulator version number
    -report-console <socket>       report console port to remote socket
    -gps <device>                  redirect NMEA GPS to character device
    -keyset <name>                 specify keyset file name
    -shell-serial <device>         specific character device for root shell
    -tcpdump <file>                capture network packets to file
    -bootchart <timeout>           enable bootcharting
    -charmap <file>                use specific key character map
    -prop <name>=<value>           set system property on boot
    -shared-net-id <number>        join the shared network, using IP address 10.1.2.<number>
    -nand-limits <nlimits>         enforce NAND/Flash read/write thresholds
    -gpu <mode>                    set hardware OpenGLES emulation mode
    -camera-back <mode>            set emulation mode for a camera facing back
    -camera-front <mode>           set emulation mode for a camera facing front
    -webcam-list                   lists web cameras available for emulation
    -screen <mode>                 set emulated screen mode
    -force-32bit                   always use 32-bit emulator
    -selinux <disabled|permissive> Set SELinux to either disabled or permissive mode

     -qemu args...                 pass arguments to qemu
     -qemu -h                      display qemu help

     -verbose                      same as '-debug-init'
     -debug <tags>                 enable/disable debug messages
     -debug-<tag>                  enable specific debug messages
     -debug-no-<tag>               disable specific debug messages

     -help                         print this help
     -help-<option>                print option-specific help

     -help-disk-images             about disk images
     -help-keys                    supported key bindings
     -help-debug-tags              debug tags for -debug <tags>
     -help-char-devices            character <device> specification
     -help-environment             environment variables
     -help-keyset-file             key bindings configuration file
     -help-virtual-device          virtual device management
     -help-sdk-images              about disk images when using the SDK
     -help-build-images            about disk images when building Android
     -help-all                     prints all help content

I hope you found this from YOUR Google search about this, and it helped 🙂

Running Multiple Android Emulators Simultaneously on OSX

I had a long standing issue where I could only launch two emulators on my Macbook Pro simultaneously.  Any other emulators would launch and get a black screen and wouldn’t load.  I accepted this as fact, and carried on with life.

screen-shot-2016-09-07-at-7-11-45-pm

Finally today I found the need to figure out “why” this was the case as I wanted to start sharding my Android UI tests to speed them up.  I was given the suggestion that it could be the # of CPUs or the amount of RAM.  If you look in the logs you may see something like this (if you launch the emulator from android avd):

screen-shot-2016-09-07-at-7-15-46-pm

or you may see:

“HAXM does not have enough memory remaining to load this AVD.”

In order to change your memory allocation for HAXM, you need to re-install it, so run the installer again which is a DMG at: [$ANDROID_HOME/sdk/extras/intel/Hardware_Accelerated_Execution_Manager/ ] and this time select more memory. [HAXM Technical Docs]screen-shot-2016-09-07-at-8-27-05-pm

But… before you install, make sure you have the latest version. The latest version as of September 7th, 2016 is 6.0.3 (Which was released on June 21, 2016).
screen-shot-2016-09-07-at-7-33-01-pm

On my Macbook, I was given 2048MB of memory by default which allowed me to run 2 emulators simultaneously (as seen above).screen-shot-2016-09-07-at-6-58-46-pm

I have 16GB total memory though, so lets make it so I can run a bunch of emulators at once by allocating 8GB 🙂
8192
confirm

Here is my Macbook Pro running 6 Android emulators simultaneously after the changes:
screen-shot-2016-09-07-at-8-17-17-pm

Success!!!  May the sharding begin!

Related Links:

Continuous Integration for Android Using Jenkins

Dan Jarvis and I co-presented a talk entitled “Continuous Integration for Android Using Jenkins” at the Android Summit 2016 on August 26, 2016 in McLean, Virginia.  Here are our slides.  Video should be available in the coming weeks and will be posted here and on Twitter.

Some Highlights:

  • Automate everything (If you can script it, do it)
  • “There’s an app for that” on a smartphone, “There’s a plugin for that” on Jenkins
  • The Android Emulator is great for scalability in CI, it’s fast and you can configure it in many ways
  • Shard for speed

Lots of Upcoming Speaking Engagements!

I’m excited to announce I’ve been accepted to talk at a bunch of upcoming events.


Espresso: A Screenshot is Worth 1,000 Words

Do your product owners, designers and the people that pay you understand what in the world your Espresso tests are doing and why they are valuable? You’ve spent so much time and effort writing these tests and your whole team deserves to get the most benefit out of them. In this talk you’ll learn how to setup your Espresso tests to take programatic screenshots, and leverage the Robot pattern of testing for clean, readable, and maintainable tests. You’ll also learn guidelines on when it’s appropriate to write Espresso tests instead of Robolectric or Unit tests, and how to leverage mock data to make your Espresso tests run with Tesla-like speed.

I have architected the Espresso test setup for our Capital One Wallet Android team and helped execute our ongoing continuous integration efforts. We’ve seen a 4x+ speed improvement over Appium, have more maintainable tests, and now have visibility for anyone to look into our test coverage.


“Continuous Integration Tips & Tricks for Android”
This is a co-talk with Dan Jarvis, the tech lead for Capital One Wallet for Android.

DNS on the Android Emulator

The last few days I have been struggling to get the Android Emulator to connect to certain domains through a proxy we have at my job.  Because of this, I started messing with my OS X DNS configuration and got myself in a DNS mess.  I finally figured my way out, so I wanted to share with you how DNS works with the Android Emulator, and hopefully avoid the troubles I ran into.

Android Documentation:

Screen Shot 2016-05-03 at 7.44.53 PM
http://developer.android.com/tools/help/emulator.html#dns

The Android Emulator uses the DNS settings of your OS X machine from/etc/resolv.conf as mentioned above.  The file looks something like:

#
# Mac OS X Notice
#
# This file is not used by the host name and address resolution
# or the DNS query routing mechanisms used by most processes on
# this Mac OS X system.
#
# This file is automatically generated.
#
search fios-router.home
nameserver 192.168.1.1

This file is re-generated by OS X as you connect to different networks (different WiFi hotspots or ethernet connections) to properly resolve DNS entries while connected to that network.

The thing I didn’t know is that /etc/resolv.conf is by default a symbolic link to /var/run/resolv.conf.  So, if you ever mess with /etc/resolv.conf and want to put back the original, you’ll have to re-create the symbolic link on your system, so that your global overrides are no longer in place.  Create the symbolic link using the following command:

sudo ln -s /var/run/resolv.conf /etc/resolv.conf

Anyways, you can override your DNS entries for the Android Emulator via the command line using the “-dns-server” option, or you can edit /etc/resolv.conf and override it globally.  Just realize that you need to create a symbolic link to /var/run/resolv.conf in order to “revert” changes if you mess with /etc/resolv.conf.

Happy DNS’ing with the Android Emulator and I hope for your case you don’t ever have to deal with it!

Using the Android Emulator for Continuous Integration

Screen Shot 2016-04-23 at 9.06.54 PM

I spent the last few weeks figuring out how to get our Android Tests running on our Jenkins CI (Continuous Integration) server using the latest Android Emulator. I am sharing what I learned so that I can hopefully help someone else and so I can refer back to it in the future.

Result of this tutorial/example:

You will be able to programmatically create an Android Emulator AVD (Android Virtual Device), start it, wait for it to load, run your tests, then kill the emulator, generate a Spoon html test report.

Screen Shot 2016-04-23 at 8.55.17 PM

Assumptions for this example:

  • You have the Android SDK configured on this machine with:
    • SDK components for API 23
    • The latest build tools
    • Google apis support
    • API 23, x86_64 with Google APIs Image
    • HAXM support installed
  • You have an Android build flavor named “qa”.
  • You are running your tests on a “debug” build.
  • You have built the APK and test APK with:
    • ./gradlew clean app:assembleQaDebug and app:assembleQaDebugAndroidTest
  • You have Android tests in a Java package called “integration_tests”
  • spoon-runner-1.2.0-jar-with-dependencies.jar exists in your working directory.
  • You want to use an API 23, x86_64, Google APIs enabled AVD (Note, you could easily change these, but this is what this example uses)

To download all the necessary SDK components referenced above, you can run the following command, or just download the components using the UI version of the Android SDK Manager.  NOTE: These commands aren’t scripted for a CI server.  You have to manually accept licensing terms:

android update sdk --no-ui --all --filter 'platform-tools,extra-android-support,build-tools-21.1.1,extra-android-m2repository,extra-google-m2repository,android-23,addon-google_apis-google-23,sys-img-x86_64-google_apis-23'

This Android test execution script will:

  • Forcefully create an AVD for this test (so that it is a fresh device every time)
  • Use port 6000 for the emulator (This is arbitrary, but by default, it is 5554 and assigned by Android).
  • Start the emulator for the AVD on a specific port (so we can target the device by serial with “emulator-{PORT}”
  • Ensure the emulator started by polling for the emulator to be booted
  • Use Spoon to run the tests in a specific package
  • Kill this emulator instance after our tests complete
#CONFIG
FLAVOR="qa"
PACKAGE_NAME="com.example.myapp"
AVD_NAME="integration-tests"
PORT=6000

##############################################################################

#Ensure the Test APK is built already.
TEST_APK_FILE="app/build/outputs/apk/app-${FLAVOR}-debug-androidTest-unaligned.apk"
if [ ! -f "${TEST_APK_FILE}" ]
then
	echo "androidTest APK doesn't exist, exiting.  Make sure you run ./gradlew app:assembleQaDebug app:assembleQaDebugAndroidTest"
exit
else
	echo "androidTest APK Exists, continuing"
fi

#Calculate the Serial Number of the emulator instance
SERIAL=emulator-${PORT}

echo "Creating (forceful) AVD with name ${AVD_NAME}"
# We have to echo "no" because it will ask us if we want to use a custom hardware profile, and we don't.
echo "no" | android create avd \
    --name "${AVD_NAME}" \
    --target "android-23" \
    --abi "google_apis/x86_64" \
    --skin "WXGA720" \
	--force
echo "AVD ${AVD_NAME} created."

#Start the Android Emulator
#"2>&1" combines stderr and stdout into the stdout stream
START_EMULATOR="emulator \
	-avd ${AVD_NAME} \
	-netspeed full \
	-gpu on \
	-netdelay none \
	-port ${PORT}"

echo $START_EMULATOR
$START_EMULATOR 2>&1 &

#Ensure Android Emulator has booted successfully before continuing
EMU_BOOTED='unknown'
while [[ ${EMU_BOOTED} != *"stopped"* ]]; do
    sleep 5
    EMU_BOOTED=`adb -s ${SERIAL} shell getprop init.svc.bootanim || echo unknown`
done

duration=$(( SECONDS - start ))
echo "Android Emulator started after $duration seconds."

# Get the file name of the QA APK (we need to cope with the version number changing)
APK_LIST_AS_TEXT="$(ls -t app/build/outputs/apk/*-android-debug-*-${FLAVOR}.apk)"
APK_FILES=( $APK_LIST_AS_TEXT )

# Use the Spoon utility as a test runner
SPOON_COMMAND="java -jar spoon-runner-1.2.0-jar-with-dependencies.jar \
	-serial ${SERIAL} \
	--apk ${APK_FILES[0]} \
	--test-apk ${TEST_APK_FILE} \
	--no-animations \
	--fail-on-failure \
	--debug
echo "Running: ${SPOON_COMMAND}"
${SPOON_COMMAND}

#Stop the Android Emulator
echo "Killing the Android Emulator with serial: ${SERIAL}"
adb -s ${SERIAL} emu kill

Final notes & tips:

  • Because the Android Emulator’s serial number is in the format of “emulator-{SERIAL}”, it is very important to start the emulator on a specific port (using the “-port {PORT}” param) so that you can reference the image by serial number.  This allows you to launch more than one emulator on the same build machine.
  • Command to list available Android platforms
    android list targets
  • Command to list available AVDs
    android list avd
  • Command to list all available SDK components that you can –filter for in the “update sdk” command:
    android list sdk --all --extended

If you have any questions or comments, let me know on Twitter at @HandstandSam.

Exercise Your Mobile App’s API with AngularJS and a Simple HTTP Proxy

Capturing your mobile app’s API responses typically very arduous:

  • Run your app, and run through a scenario to generate the calls you want while you grab the logs.
  • or… run a bunch of API request with Postman or curl, but having to run the calls in order and copy/paste Authentication headers, etc to make it work.

You would think building a AngularJS web interface for your API to do the tedious parts for you would be easy… which it technically is… but not in a way that you can trick the browser into thinking it’s secure.  That’s because a web browser uses security policies that typical block CORS (Cross-Origin Resource Sharing).  That means that if you are on http://myui.com and want to make a request to your API (which is on http://myapi.com) and you API doesn’t have JSONP (JSON with Padding) support, you are in a hard spot and need something server-side to make those calls for you, and then return the results.

A trivial example of the pain CORS will cause is shown here with my GitHub profile.  If I want to pull this data, and add it to my website, I can’t do it because of CORS blocking.

curl https://api.github.com/users/handstandsam

{
  "login": "handstandsam",
  "id": 264948,
  "avatar_url": "https://avatars.githubusercontent.com/u/264948?v=3",
  "gravatar_id": "",
  "url": "https://api.github.com/users/handstandsam",
  "html_url": "https://github.com/handstandsam",
  "followers_url": "https://api.github.com/users/handstandsam/followers",
  "following_url": "https://api.github.com/users/handstandsam/following{/other_user}",
  "gists_url": "https://api.github.com/users/handstandsam/gists{/gist_id}",
  "starred_url": "https://api.github.com/users/handstandsam/starred{/owner}{/repo}",
  "subscriptions_url": "https://api.github.com/users/handstandsam/subscriptions",
  "organizations_url": "https://api.github.com/users/handstandsam/orgs",
  "repos_url": "https://api.github.com/users/handstandsam/repos",
  "events_url": "https://api.github.com/users/handstandsam/events{/privacy}",
  "received_events_url": "https://api.github.com/users/handstandsam/received_events",
  "type": "User",
  "site_admin": false,
  "name": "Sam Edwards",
  "company": "http://github.com/handstandtech",
  "blog": "http://handstandsam.com",
  "location": null,
  "email": null,
  "hireable": null,
  "bio": null,
  "public_repos": 7,
  "public_gists": 2,
  "followers": 10,
  "following": 58,
  "created_at": "2010-05-04T22:15:11Z",
  "updated_at": "2016-02-26T23:08:36Z"
}

GitHub is smart enough to provide the ability to use JSONP (invokes a local javascript method with the result of your API request), which allows us to actually do this.

curl https://api.github.com/users/handstandsam?callback=mycallback

/**/mycallback({
  "meta": {
    "Content-Type": "application/javascript; charset=utf-8",
    "X-RateLimit-Limit": "60",
    "X-RateLimit-Remaining": "55",
    "X-RateLimit-Reset": "1457061648",
    "Cache-Control": "public, max-age=60, s-maxage=60",
    "Vary": "Accept",
    "ETag": "\"5dd9556a9a8f8067c294a028b1f1a21d\"",
    "Last-Modified": "Fri, 26 Feb 2016 23:08:36 GMT",
    "X-GitHub-Media-Type": "github.v3",
    "status": 200
  },
  "data": {
    "login": "handstandsam",
    "id": 264948,
    "avatar_url": "https://avatars.githubusercontent.com/u/264948?v=3",
    "gravatar_id": "",
    "url": "https://api.github.com/users/handstandsam",
    "html_url": "https://github.com/handstandsam",
    "followers_url": "https://api.github.com/users/handstandsam/followers",
    "following_url": "https://api.github.com/users/handstandsam/following{/other_user}",
    "gists_url": "https://api.github.com/users/handstandsam/gists{/gist_id}",
    "starred_url": "https://api.github.com/users/handstandsam/starred{/owner}{/repo}",
    "subscriptions_url": "https://api.github.com/users/handstandsam/subscriptions",
    "organizations_url": "https://api.github.com/users/handstandsam/orgs",
    "repos_url": "https://api.github.com/users/handstandsam/repos",
    "events_url": "https://api.github.com/users/handstandsam/events{/privacy}",
    "received_events_url": "https://api.github.com/users/handstandsam/received_events",
    "type": "User",
    "site_admin": false,
    "name": "Sam Edwards",
    "company": "http://github.com/handstandtech",
    "blog": "http://handstandsam.com",
    "location": null,
    "email": null,
    "hireable": null,
    "bio": null,
    "public_repos": 7,
    "public_gists": 2,
    "followers": 10,
    "following": 58,
    "created_at": "2010-05-04T22:15:11Z",
    "updated_at": "2016-02-26T23:08:36Z"
  }
})

Most of us aren’t that lucky however, and JSONP is not supported by the API we are using.  And while this is great in many cases, it still doesn’t help us with anything other than GET requests.  POST, PUT and DELETE commands are all blocked from CORS so we are very limited.

If you are just interested in GET requests, and JSONP is not a possibility, you could instead modify the server code to return the header “Access-Control-Allow-Origin: *”. Most modern browsers would allow you to do a CORS GET request. Unfortunately in a lot of cases, GET requests are not enough or we don’t have access to the server code.

Well, how do I get around this to make REST calls to our server, like POST?

Well, write some server code that can do the requests for you since it won’t have security restrictions like the browser does.  But, most of us don’t have time or that.  The solution I came up with on the latest project I’m learning the API for is “php-guzzle-http-proxy“. This is as dirt simple as I could get and navigate around all these CORS problems I have been mentioning without having to write ANY server-side code.

Just plop this repo in a directory of your PHP server, and start making requests (/proxy in this case).  You’ll just add in a special header to your request I’ve named “X-Proxy-Url”.  This is the URL you REALLY want to call, but CORS won’t let you.  You can do any type of request you want: GET, POST, PUT or DELETE.

In curl it would look like:

curl -H "X-Proxy-Url: https://myapi.com/login" -H "Accept: application/json" -X POST -d '{"username": "sam", "password": "notsecure"}' http://myui.com/proxy

In AngularJS it would look like:

$http({
  method: 'POST',
  url: '/proxy,
  headers: {
    'Accept': 'application/json',
    'X-Proxy-Url': 'https://myapi.com/login'
  },
  data: {"username": "sam", "password": "notsecure"}
}).success(function(response){
  console.log(response.data);
});

That’s it. No server-side code required, just this one tiny PHP proxy script to allow you to make all types of REST calls including GET, POST, etc.

The tool I wrote (that unfortunately I can’t share) used Angular and Bootstrap.  When using the two in combo, when you get your JSON response back, display it with the following tag to make it formatted and easy to read.

<pre>{{data | json}}</pre>

Let me know if you found this helpful on Twitter at @HandstandSam

———————–

Related Links:

Continuous Integration at LesFurets.com – Richmond Java Users Group (RJUG)

“Our highest priority is to satisfy the customer through early and continuous delivery of valuable software.” – 1st principal of the Agile Manifesto

This presentation by Raphaël Brugier (@rbrugier) at the Richmond Java Users Group last night (Feb 17th) was really good and highlighted:

  • Continuous Integration.  A case study of LesFurets.com. 40,000 unit tests running in 3 minutes.  Selenium tests running in 15 minutes.  Passing tests == shippable code.
  • They went from shipping a feature in a month, to as soon as ready (possibly only a few days).
  • A different kind of Git Flow using git octopus. – https://github.com/lesfurets/git-octopus
  • They have new developers hit the deploy button their first day.  Share deployment responsibilities within the dev team.
  • “If it hurts, do it more often.” – In regards to doing a deployment/release.
  • Work on feature branches.  Ship features as ready, not based on sprints. (git octopus facilitates this)
  • Detect feature branch merge conflicts after every push and make sure it’s resolved ASAP.
  • QA is owned by developers and sign off is given by product when dev demos it.
  • and more…

I would highly recommend watching the recording of the event.  Heads up: The speaker had a French accent and the video is from my MacBook’s camera, however, you get to watch it and otherwise you wouldn’t be able to :-p.

YouTube Link: https://www.youtube.com/watch?v=uTOPvC3lV1Q

Slides: http://www.slideshare.net/raphaelbrugier/continuous-delivery-journey-at-lesfuretscom

Meetup Info: http://www.meetup.com/Richmond-Java-Users-Group/events/226909145/