IRhetoric - Convergence in the Simulacrum. A Blog by Karsten Januszewski

Getting Twitter Avatar Via C# With 1.1 API

Wed, 01 May 2013 09:23:26 -0900

If you know someone's Twitter handle and would like to display their twitter avatar on your website, here's some code to get the URL of their avatar. First, you'll need to register at http://dev.twitter.com and acquire a ConsumerKey, ConsumerSecret, Token and TokenSecret -- now that Twitter supports application only authentication, there isn't any handshaking involved; you just need to craft up the right OAuthCredentials for a ProtectedResource, which the Hammock library does for you. Love that library: you can get it here: https://github.com/danielcrenna/hammock or grab it as a NuGet package: http://nuget.org/packages/Hammock

Here’s the code; nothing too fancy:

using System;
using Hammock;
using Hammock.Authentication.OAuth;
using Hammock.Web;
using Newtonsoft.Json.Linq;

namespace ExternalServices
{
    public class TwitterAvatarLookup : ITwitterAvatarLookup
    {
        const string ConsumerKey = "";
        const string ConsumerSecret = "";
        const string Token = "";
        const string TokenSecret = "";

        public string GetTwitterAvatarUrl(string twitterHandle)
        {
            string avatarUrl = string.Empty;
            var request = new RestRequest
            {
                Credentials = new OAuthCredentials
                {
                    Type = OAuthType.ProtectedResource,
                    SignatureMethod = OAuthSignatureMethod.HmacSha1,
                    ParameterHandling = OAuthParameterHandling.HttpAuthorizationHeader,
                    ConsumerKey = ConsumerKey,
                    ConsumerSecret = ConsumerSecret,
                    Token = Token,
                    TokenSecret = TokenSecret,
                }
            };
            request.Path =
                string.Format(
                    "https://api.twitter.com/1.1/users/lookup.json?screen_name={0}&include_entities=0&include_rts=0",
                    twitterHandle);

            request.Method = WebMethod.Get;
            RestClient client = new RestClient();
            try
            {
                RestResponse response = client.Request(request);
                JArray jArray = JArray.Parse(response.Content);
                avatarUrl = (string)jArray[0]["profile_image_url_https"];

            }
            catch (Exception)
            {
                return "default.png";
            }
            return avatarUrl;
        }

    }
}

If you know more than one handle whose avatar you need to get, the API supports passing multiple user handles; see https://dev.twitter.com/docs/api/1.1/get/users/lookup

Saving a PDFSharp PDF File To Azure Blob Storage

Fri, 01 Mar 2013 14:29:39 -0900

Love the PDFSharp library. Here’s how I went about saving a PDF generated with that library to Azure blob storage:

const bool unicode = false;
const PdfFontEmbedding embedding = PdfFontEmbedding.Always;
PdfDocumentRenderer pdfRenderer = new PdfDocumentRenderer(unicode, embedding);
pdfRenderer.Document = document;
pdfRenderer.RenderDocument();
MemoryStream memStream = new MemoryStream();
pdfRenderer.PdfDocument.Save(memStream,false);

var client = new CloudBlobClient(new Uri("http://*.blob.core.windows.net", UriKind.Absolute),
new StorageCredentialsAccountAndKey("*",
"..."));

var container = client.GetContainerReference("temp");
memStream.Seek(0, SeekOrigin.Begin);
string filename = DateTime.Now.ToString().GetHashCode().ToString("x") + ".pdf";
var pdf = container.GetBlobReference(filename);
pdf.Properties.ContentType = "application/pdf";
pdf.UploadFromStream(memStream, new BlobRequestOptions { Timeout = TimeSpan.FromMinutes(10) });
memStream.Close();

The crux of the code is the line where pass false to the Save method of PdfDocument, which keeps the memory stream open. And, then, before giving that stream to the Azure SDK method, you need to rewind the stream to the beginning. Other than that, all pretty boilerplate.

Capture ClickOnce File Downloads With Event Tracking In Google Analytics

Fri, 01 Mar 2013 14:10:23 -0900

This flummoxed me for a bit, so I figured I post it. If you wire the event up to an onClick handler as the docs suggest:

<a href="app.application" onClick="_gaq.push(['_trackEvent', 'ClickOnce', 
'Download']);">Play</a>

Your event will never fire. The trick is to add the target attribute and set it to _blank which opens a new tab in the browser and immediately closes it:

<a href="app.application" target="_blank" onClick="_gaq.push(['_trackEvent', 'ClickOnce',
'Download']);">Play</a>

Windows Azure Table Storage Emulator UpdateObject Error

Mon, 21 Jan 2013 12:53:00 -0900

Was getting this:

<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<error xmlns="http://schemas.microsoft.com/ado/2007/08/dataservices/metadata">
  <code>InvalidInput</code>
  <message xml:lang="en-US">One of the request inputs is not valid.</message>
</error>

When calling UpdateObject from the Azure SDK when using the table storage emulator. Turns out the emulator doesn’t exactly emulate as per MSDN documentation: http://msdn.microsoft.com/en-us/library/gg433135.aspx “The storage emulator does not support Insert-Or-Replace Entity or Insert-Or-Merge Entity, known as upsert features.”

So, to work around, I changed the code to delete then reinsert. Or not. Turned out I had lots of concurrancy problems with doing a delete/insert which were resolved with the more transactional upsert.

Sorting ListBlobs By LastModifiedUtc

Mon, 03 Dec 2012 09:46:07 -0900

I had a list of JSON objects stored as individual blobs in Azure blob storage. I needed to get them out of blob storage sorted by last modified time and then reconstitute them as an array of JSON objects. Hit a few curiosities in writing this code, which someone else may benefit from. Without further ado, here’s the code:

            CloudBlobDirectory dir = eventContainer.GetDirectoryReference("http://---");
            SortedDictionary<DateTime, string> dictionary = new SortedDictionary<DateTime, string>();
            foreach (CloudBlob blob in dir.ListBlobs())
            {
                if (blob.Name == string.Format("{0}/$$$.$$$", "live")) continue;
                dictionary.Add(blob.Properties.LastModifiedUtc, blob.DownloadText());
            }
            StringBuilder json = new StringBuilder();
            json.Append("Callback([");
            foreach (string s in dictionary.Values)
            {
                json.Append(s);
                json.Append(",");
            }
            json.Remove(json.Length - 1, 1);
            json.Append("])");
            CloudBlob liveJson = dir.GetBlobReference("live.json");
            liveJson.Properties.CacheControl = cacheControl;
            liveJson.UploadText(json.ToString());

Okay, so what is going on here?

I start out with a nifty SortedDictionary, which will sort my items by the key of the dictionary as I add them.

I then call ListBlobs(). There’s this peculiar oddity with Blob Storage when you use CloudXplorer where this ghost file name $$$.$$$ gets created, which is why I have to check the name of the file. I then throw the string and the date into the dictionary.

Once that finishes, I iterate the dictionary, adding the callback and syntax for making a json array. Get rid of the final comma and then throw the whole deal back into blob storage. Hoorah!

At first, I was actually deserializing the strings into objects using JSON.Net and then I realized there was no reason to do that when all I needed to do was manipulate strings.

Converting New Twitter Search Result To Old Format in C# Using JSON.NET

Mon, 05 Nov 2012 14:57:00 -0900

Been playing with the new Twitter API, version 1.1, and want to preserve a bunch of code, so I wrote a routine to convert the new format to the old format. Maybe it’ll save someone time. (Note that I didn’t convert the metadata because I didn’t need it.)

Here’s the code which uses JSON.NET:

            JObject newapiresults = null;
            newapiresults = JObject.Parse(response.Content);
            JArray tweets = new JArray();
            foreach (JObject status in tweets)
            {
                JObject oldStatus = new JObject();
                oldStatus["created_at"] = status["created_at"];
                oldStatus["from_user"] = status["user"]["screen_name"];
                oldStatus["from_user_id"] = status["user"]["id"];
                oldStatus["from_user_id_str"] = status["user"]["id_str"];
                oldStatus["from_user_id_name"] = status["user"]["name"];
                oldStatus["geo"] = status["geo"];
                oldStatus["id"] = status["id"];
                oldStatus["id_str"] = status["id_str"];
                oldStatus["iso_language_code"] = status["metadata"]["iso_language_code"];
                oldStatus["metadata"] = status["metadata"];
                oldStatus["profile_image_url"] = status["user"]["profile_image_url"];
                oldStatus["profile_image_url_https"] = status["user"]["profile_image_url_https"];
                oldStatus["source"] = status["source"];
                oldStatus["text"] = status["text"];
                oldStatus["to_user"] = status["in_reply_to_screen_name"];
                oldStatus["to_user_id"] = status["in_reply_to_user_id"];
                oldStatus["to_user_id_str"] = status["in_reply_to_user_id_str"];
                oldStatus["to_user_name"] = status["in_reply_to_screen_name"];
                oldStatus["in_reply_to_status_id"] = status["in_reply_to_status_id"];
                oldStatus["in_reply_to_status_id_str"] = status["in_reply_to_status_id_str"];
                results.Add(oldStatus);


            }
            JObject result = new JObject();
            result["results"] = results;
            //do something with the string
            UploadText(result.ToString());

For what it is worth, here’s the old JSON and then the new JSON – wow the new JSON is a lot more verbose!

{ "completed_in" : 0.029000000000000001,
  "max_id" : 265608574640738304,
  "max_id_str" : "265608574640738304",
  "next_page" : "?page=2&max_id=265608574640738304&q=%40twitterapi%20-via",
  "page" : 1,
  "query" : "%40twitterapi+-via",
  "refresh_url" : "?since_id=265608574640738304&q=%40twitterapi%20-via",
  "results" : [ { "created_at" : "Tue, 06 Nov 2012 00:16:33 +0000",
        "from_user" : "morgules",
        "from_user_id" : 636921780,
        "from_user_id_str" : "636921780",
        "from_user_name" : "Дмитрий Моргулес",
        "geo" : null,
        "id" : 265608574640738304,
        "id_str" : "265608574640738304",
        "iso_language_code" : "ru",
        "metadata" : { "result_type" : "recent" },
        "profile_image_url" : "http://a0.twimg.com/profile_images/2647733681/0daea79d4d5858a39bb1801d64b20e14_normal.jpeg",
        "profile_image_url_https" : "https://si0.twimg.com/profile_images/2647733681/0daea79d4d5858a39bb1801d64b20e14_normal.jpeg",
        "source" : "<a href="http://twitter.com/tweetbutton">Tweet Button</a>",
        "text" : "Руслан Нурисламов: безрукий барабанщик из Златоуста | Общество | Слово http://t.co/ziOSiA78 с помощью @twitterapi",
        "to_user" : null,
        "to_user_id" : 0,
        "to_user_id_str" : "0",
        "to_user_name" : null
      },
      { "created_at" : "Tue, 06 Nov 2012 00:13:12 +0000",
        "from_user" : "oxkarlomejor",
        "from_user_id" : 221827509,
        "from_user_id_str" : "221827509",
        "from_user_name" : "oscar  david ",
        "geo" : null,
        "id" : 265607731417870336,
        "id_str" : "265607731417870336",
        "iso_language_code" : "es",
        "metadata" : { "result_type" : "recent" },
        "profile_image_url" : "http://a0.twimg.com/profile_images/1602975094/Imagen_010_normal.jpg",
        "profile_image_url_https" : "https://si0.twimg.com/profile_images/1602975094/Imagen_010_normal.jpg",
        "source" : "<a href="http://twitter.com/">web</a>",
        "text" : "@twitterapi quiero ser Famoso",
        "to_user" : "twitterapi",
        "to_user_id" : 6253282,
        "to_user_id_str" : "6253282",
        "to_user_name" : "Twitter API"
      }]}

And here’s the new JSON:

{ "search_metadata" : { "completed_in" : 0.089999999999999997,
      "count" : 100,
      "max_id" : 265255057749053440,
      "max_id_str" : "265255057749053440",
      "next_results" : "?max_id=265212014639603711&q=nekocase&count=100&include_entities=1",
      "query" : "nekocase",
      "refresh_url" : "?since_id=265255057749053440&q=nekocase&include_entities=1",
      "since_id" : 0,
      "since_id_str" : "0"
    },
  "statuses" : [ { "contributors" : null,
        "coordinates" : null,
        "created_at" : "Mon Nov 05 00:51:49 +0000 2012",
        "entities" : { "hashtags" : [  ],
            "media" : [ { "display_url" : "pic.twitter.com/PEzikMU5",
                  "expanded_url" : "http://twitter.com/jtspicer/status/265255057749053440/photo/1",
                  "id" : 265255057753247745,
                  "id_str" : "265255057753247745",
                  "indices" : [ 115,
                      135
                    ],
                  "media_url" : "http://p.twimg.com/A65gEdRCEAEf0xo.jpg",
                  "media_url_https" : "https://p.twimg.com/A65gEdRCEAEf0xo.jpg",
                  "sizes" : { "large" : { "h" : 766,
                          "resize" : "fit",
                          "w" : 1024
                        },
                      "medium" : { "h" : 449,
                          "resize" : "fit",
                          "w" : 600
                        },
                      "small" : { "h" : 254,
                          "resize" : "fit",
                          "w" : 340
                        },
                      "thumb" : { "h" : 150,
                          "resize" : "crop",
                          "w" : 150
                        }
                    },
                  "type" : "photo",
                  "url" : "http://t.co/PEzikMU5"
                } ],
            "urls" : [  ],
            "user_mentions" : [ { "id" : 126406217,
                  "id_str" : "126406217",
                  "indices" : [ 20,
                      29
                    ],
                  "name" : "Neko Case",
                  "screen_name" : "NekoCase"
                } ]
          },
        "favorited" : false,
        "geo" : null,
        "id" : 265255057749053440,
        "id_str" : "265255057749053440",
        "in_reply_to_screen_name" : null,
        "in_reply_to_status_id" : null,
        "in_reply_to_status_id_str" : null,
        "in_reply_to_user_id" : null,
        "in_reply_to_user_id_str" : null,
        "metadata" : { "iso_language_code" : "en",
            "result_type" : "recent"
          },
        "place" : null,
        "possibly_sensitive" : false,
        "retweet_count" : 0,
        "retweeted" : false,
        "source" : "<a href=\"http://twitter.com/download/iphone\" rel=\"nofollow\">Twitter for iPhone</a>",
        "text" : "The best way to get @NekoCase to RT you is to tweet a context-free photo of an adorable animal. Here goes nothing! http://t.co/PEzikMU5",
        "truncated" : false,
        "user" : { "contributors_enabled" : false,
            "created_at" : "Mon Feb 23 19:57:21 +0000 2009",
            "default_profile" : false,
            "default_profile_image" : false,
            "description" : "Writer of things. Lover of music. Appreciator of film. Unabashed know-it-all. Street walkin' cheetah with a heart full of napalm. And so on. Views=mine.",
            "entities" : { "description" : { "urls" : [  ] } },
            "favourites_count" : 768,
            "follow_request_sent" : null,
            "followers_count" : 360,
            "following" : null,
            "friends_count" : 336,
            "geo_enabled" : false,
            "id" : 21686478,
            "id_str" : "21686478",
            "is_translator" : false,
            "lang" : "en",
            "listed_count" : 8,
            "location" : "Columbus, Ohio (Short North)",
            "name" : "Justin Spicer",
            "notifications" : null,
            "profile_background_color" : "D3D9DB",
            "profile_background_image_url" : "http://a0.twimg.com/profile_background_images/243592391/x3ce2eb796060291bf2a9d40db53a85f.jpg",
            "profile_background_image_url_https" : "https://si0.twimg.com/profile_background_images/243592391/x3ce2eb796060291bf2a9d40db53a85f.jpg",
            "profile_background_tile" : false,
            "profile_banner_url" : "https://si0.twimg.com/profile_banners/21686478/1351622268",
            "profile_image_url" : "http://a0.twimg.com/profile_images/2785722127/42dbe3599093209c9abab988e56f2cdf_normal.jpeg",
            "profile_image_url_https" : "https://si0.twimg.com/profile_images/2785722127/42dbe3599093209c9abab988e56f2cdf_normal.jpeg",
            "profile_link_color" : "DB6995",
            "profile_sidebar_border_color" : "5AC3E9",
            "profile_sidebar_fill_color" : "2D1E29",
            "profile_text_color" : "A177AB",
            "profile_use_background_image" : true,
            "protected" : false,
            "screen_name" : "jtspicer",
            "show_all_inline_media" : true,
            "statuses_count" : 11964,
            "time_zone" : "Eastern Time (US & Canada)",
            "url" : null,
            "utc_offset" : -18000,
            "verified" : false
          }
      },
      { "contributors" : null,
        "coordinates" : null,
        "created_at" : "Mon Nov 05 00:42:28 +0000 2012",
        "entities" : { "hashtags" : [  ],
            "urls" : [  ],
            "user_mentions" : [ { "id" : 126406217,
                  "id_str" : "126406217",
                  "indices" : [ 0,
                      9
                    ],
                  "name" : "Neko Case",
                  "screen_name" : "NekoCase"
                },
                { "id" : 134073162,
                  "id_str" : "134073162",
                  "indices" : [ 10,
                      21
                    ],
                  "name" : "brigid ",
                  "screen_name" : "MBbyBrigid"
                }
              ]
          },
        "favorited" : false,
        "geo" : null,
        "id" : 265252709555376128,
        "id_str" : "265252709555376128",
        "in_reply_to_screen_name" : "NekoCase",
        "in_reply_to_status_id" : 265237602242818049,
        "in_reply_to_status_id_str" : "265237602242818049",
        "in_reply_to_user_id" : 126406217,
        "in_reply_to_user_id_str" : "126406217",
        "metadata" : { "iso_language_code" : "en",
            "result_type" : "recent"
          },
        "place" : null,
        "retweet_count" : 0,
        "retweeted" : false,
        "source" : "<a href=\"http://twitter.com/download/iphone\" rel=\"nofollow\">Twitter for iPhone</a>",
        "text" : "@NekoCase @mbbybrigid thank you so much for sharing the fundraiser!",
        "truncated" : false,
        "user" : { "contributors_enabled" : false,
            "created_at" : "Thu Dec 16 20:06:39 +0000 2010",
            "default_profile" : false,
            "default_profile_image" : false,
            "description" : "The owner/crafter/dreamer behind Cambridge's first stitch lounge and craft studio. Avid gardener. Opinionated stitcher. And I want you to make something...",
            "entities" : { "description" : { "urls" : [  ] },
                "url" : { "urls" : [ { "expanded_url" : null,
                          "indices" : [ 0,
                              27
                            ],
                          "url" : "http://gatherhereonline.com"
                        } ] }
              },
            "favourites_count" : 520,
            "follow_request_sent" : null,
            "followers_count" : 1024,
            "following" : null,
            "friends_count" : 437,
            "geo_enabled" : false,
            "id" : 227419407,
            "id_str" : "227419407",
            "is_translator" : false,
            "lang" : "en",
            "listed_count" : 47,
            "location" : "Cambridge, MA",
            "name" : "virginia b. johnson",
            "notifications" : null,
            "profile_background_color" : "C0DEED",
            "profile_background_image_url" : "http://a0.twimg.com/profile_background_images/293836111/spools_of_thread.jpg",
            "profile_background_image_url_https" : "https://si0.twimg.com/profile_background_images/293836111/spools_of_thread.jpg",
            "profile_background_tile" : true,
            "profile_image_url" : "http://a0.twimg.com/profile_images/1192172838/etsy_gather_here_avatar_normal.jpg",
            "profile_image_url_https" : "https://si0.twimg.com/profile_images/1192172838/etsy_gather_here_avatar_normal.jpg",
            "profile_link_color" : "0084B4",
            "profile_sidebar_border_color" : "C0DEED",
            "profile_sidebar_fill_color" : "DDEEF6",
            "profile_text_color" : "333333",
            "profile_use_background_image" : true,
            "protected" : false,
            "screen_name" : "gather_here",
            "show_all_inline_media" : false,
            "statuses_count" : 7655,
            "time_zone" : "Central Time (US & Canada)",
            "url" : "http://gatherhereonline.com",
            "utc_offset" : -21600,
            "verified" : false
          }
      }]}

Getting Twitter Avatar Via C#

Fri, 05 Oct 2012 10:42:00 -0900

[See http://rhizohm.net/irhetoric/post/2013/05/01/Getting-Twitter-Avatar-Via-C-With-11-API.aspx for how to do this with the Twitter 1.1 API]

Until Twitter deprecates their version 1 api, getting an avatar from Twitter anonymously serverside with C# is a snap. A little goofy since it returns a 302 so you have to set AllowAutoRedirect to false and then look in the Header location to get the URI:

private string GetTwitterAvatarUrl(string twitterHandle)
        {
            string avatarUrl = string.Empty;
            HttpWebRequest request = (HttpWebRequest)WebRequest.Create(string.Format("https://api.twitter.com/1/users/profile_image?screen_name={0}", twitterHandle));
            request.AllowAutoRedirect = false;
            request.Method = "GET";
            try
            {
                WebResponse response = request.GetResponse();
                if (response.Headers["Location"] != null)
                    avatarUrl = response.Headers["Location"];
            }
            catch (WebException) {}
            
            return avatarUrl;
        }

WordPress Audio Player Now Falls Back To HTML5

Wed, 15 Aug 2012 09:17:41 -0900

I realized people on iOS and other devices that don’t support Flash couldn’t use the embedded Flash player for listening to audio when using the WordPress Audio Player plugin. I wanted to customize it so that sites would fall back to HTML5 if Flash wasn’t present. I found a post in the WordPress forums of someone who did this here.

However, his script didn’t work as it was for two reasons. First, he was Base64 encoding the url to the mp3 — not sure why. So I remmed out that code. Second, he was replacing the parent element with the <audio> tag, which was blowing away the entire post! By replacing the element witht the id of the audio tag itself, walla! Everything works!

You can download the modified file here and minify if you’d like.

Skinless jPlayer Implementation

Wed, 15 Aug 2012 09:13:01 -0900

I recently implemented a skinless audio player using jPlayer that plays m4a files that it retrieves from the iTunes Search API.

I am using jPlayer 2.1.0 and the main gotchas I hit were problems with not having the latest version of Flash on target computers. jPlayer requires Flash 10 or higher for fallback, which it seems to do for m4a files in Firefox.

Anyway, here’s the code if you are interested. Here’s where I instantiate the jPlayer. The only parameters I pass are the supplied parameter. I got away with the defaults for everything else.

$("#jplayer").jPlayer({
ready: function() {
$("#jp_container .track-default").click();
},
supplied: "m4a, m4v"
});

Then, here’s how I wire up the click handler, toggling between the play and pause icons:

var current;
 $(".play").click(function(e) {
 $("#jplayer").jPlayer("setMedia", {
 m4a: $(this).attr("href")
 });

 if (current == this) {
 $(current).empty();
 $(current).append("<img style='vertical-align:middle;' 
   border='0' src='/images/ic_menu_play.png'/>");
 $("#jplayer").jPlayer("pause");
 current = null;
 return false;

 }
 $("#jplayer").jPlayer("play");

 if (current != null) {
 $(current).empty();
 $(current).append("<img style='vertical-align:middle;' 
   border='0' src='/images/ic_menu_play.png'/>");
 }
 $(this).empty();
 $(this).append("<img style='vertical-align:middle;' 
   border='0' src='/images/ic_menu_pause.png'/>");
 current = this;
 return false;
 });

Works great!

Technorati Tags: Jplayer

Having jPlayer dynamically switch between m4a and mp3

Wed, 15 Aug 2012 09:07:52 -0900

According to the docs of jPlayer http://www.jplayer.org/latest/developer-guide/#jPlayer-essential-formats, you are only supposed to use either mp3 or m4a. But in my case, I’m not sure which gets passed to the player from the database. So, I had to handle that in the script. Here’s what it looks like:

$("#jplayer").jPlayer({
 ready: function() {
 $("#jp_container .track-default").click();
 },

 supplied: "m4a, mp3, mv4"
 });

When I wire up the handlers, I check the file extension before passing the URL to the setMedia method, like this:

$(".play").click(function(e) {
 var filename = $(this).attr("href");
 var ext = filename.split('.').pop();

Technorati Tags: jPlayer

 if (ext == "m4a") {
 $("#jplayer").jPlayer("setMedia", {
 m4a: $(this).attr("href")
 });
 }
 else {
 $("#jplayer").jPlayer("setMedia", {
 mp3: $(this).attr("href")
 });

 }

A little hacky, but it works!