Jez Walter Additional browsers View history page View main browser (Shown) Type instead of Alexa Browser back button Last thing Alexa heard Device that heard request Main screen resizeable  Mobile content tab from built-in web server Web server Additional browser 1 of 3 Window Co-ords Mobile control tab
Getting Started: Simply ‘Run’ the application installer,  register as a user & log in, then say, ‘Alexa … Who sang’ followed by a song name. A built in browser should then open a YouTube list of tracks related to your song choice. Login: You will need to login to your Alexa app so that the application can read the ‘History’ page, to learn what Alexa heard and also which device heard it, The application can not read any part of the login process but if you prefer for whatever reason, you can login to your Alexa app from any browser before launching the Web extractor application but be sure to tick the ‘Keep me logged in’ checkbox otherwise you would need to login through the application every time you started  it. Configuration options: The application accepts ‘Phrases’ or ‘Words’ of your choice as commands which are written to a file named ‘connections.txt’ and found in your installation folder. If the whole ‘Phrase’ or ‘Word’ forms part of what Alexa hears then this is considered a ‘Match’ or ‘Trigger’ to act on. Actions typically mean firing off a number of browser web searches but may also include launching PC applications, batch files, PC to Hub commands and more. Below are just a few ‘Starter’ points on how you might use the Web Extractor. To open the ‘connections.txt file click on the application menu Commands - Open Phrases List You can edit any of the lines and simply save changes, then refresh the application by clicking on menu Commands  - Refresh Phrases List Default values: The application has 4 user browsers built in, these are represented by default values at the top of the connections.txt file Default-0:[SEARCH%20] Default-1:[SEARCH+]&biw=1601&bih=815&source=lnms&tbm=isch Default-2:about:blank Default-3:about:blank The number following the word ‘Default’ is the number of the built in browser where ‘0’ is the main application screen. If no phrases are matched the default values are used instead. if you set any of them to ‘about:blank’ there will be no action taken and no default value for that particular browser. Phrases: These are used to form your own unique commands. Any number can be grouped together meaning that if any one phrase in the group is matched then ‘Actions’ are then triggered. Phrase:Who sings Phrase:Who sang NEWURL-0:[SEARCH] NEWURL-1:[SEARCH] So from the above configuration, if you said ‘Alexa Who sings the look of love’ it would match on the ‘Who sings’ phrase, the phrase is then automatically removed from what Alexa heard and the rest of the text is past to your different URL’s that immediately follow your phrases. In the case above to Browser-0 (Main screen) and to Browser-1. Please note: After making any changes to the connections.txt file and saving them, you can then refresh this in the application by going to Commands - Refresh phrases list without having to restart the application. All URL’s must begin with either ‘http’ or ‘https’ otherwise it will be treated like a PC application. How to configure searches: Go to any website that has a search facility, Google, YouTube, Supermarket etc. Search for something with at least two words e.g. ‘Red Wine’. Copy the URL into the connections.txt file and look for the words ‘Red Wine’. Replace these with [SEARCH]. If the spaces in the URL have been replaced by a character e.g. ‘Red+Wine’ or ‘Red%20Wine’ then put these in as well e.g. [SEARCH+] or [SEARCH%20] so that the website receives the search text how it expects it with correctly configured spaces. Sending to mobile devices or other PC’s, Laptops, Tablets: The Web extractor application has a built in web server allowing anything with a browser to potentially connect to the application through your own home network without actually installing anything. Go to Servers - Server - Console from the main menu and click the start button, then optionally close the server screen. This produces an IP address that is now ready to use by any other devices on your network. You just need to type this into your device browser along with a number of your choice to identify that monitor and device e.g. you need to add /?S=x where ‘x’ is a number of your choice. The same number can be used by multiple devices but then they will all receive content instead of targeting specific devices. To add phrases is simply NEWURL- x : URL Phrase:Cocktails NEWURL-20:[SEARCH-] In the above example, if you say ‘Alexa Cocktails Boston Tea Party’ the phrase part, ‘Cocktails’ is removed and ‘Boston Tea Party’ becomes the search string that launches that particular recipe. This will now go to every device that has been given the number ‘20’ in its web server URL whether that is a single monitor or monitors all over the house. ZONES: Only send content if the question was heard from a specific device: This is also very easy to set up. Go to your ‘Settings’ page in the Alexa app and make a note of the name of each of your devices,  drop these at the bottom of the connections.txt file with a prefix of ZONE: e.g. ZONE:Echo-1 ZONE:Echo-2 Phrase:Main Street NEWURL-62: ZONE:Dot-1 ZONE:Dot-2 ZONE:Dot-3 Whatever phrases you add below each of these device names will only launch the URL’s if the question was actually heard from that specific device. Imagine, you have a Dot in your hall with a small monitor and possibly a Raspberry Pi driving it. You ask, ‘Alexa … What’s the traffic like on main street’ the Web extractor could then launch a website to an actual traffic camera onto that particular monitor whilst you’re putting your coat on. Launching PC apps: The easiest way to launch a PC application such as Word, Excel, Access etc is to open the application and save an empty document to a known folder then attach it to a phrase or a number of applications to a phrase Phrase:Banking NEWURL-100:C:\accounts NEWURL-100:C:\accounts\mybank.xlsx In the above example we’ve given the application launcher a number of 100 which doesn’t relate to any monitor, it doesn’t need to . Then we type the path to the empty document, this could also be the actual application path but this can be hard to find for current Microsoft products. Now when you say ‘Alexa Banking’ several key applications that you commonly use can be on your desktop waiting for you to use them. The Web extractor does not close PC applications. Positioning and sizing screens: It is possible to have various PC apps fire up to a specific part of the screen. e.g. Half the screen for Excel and several folders auto opened in the other half along with a calculator. In order for this to work we need an additional attribute to follow the application path. This is your intended application main window text or ‘Form name’. This is usually found at the top of the window or to the left and may include a path to where the document was saved to. You need all of this text. Then we need to get some co-ordinates to use for positioning and sizing. Using one of the built in browsers, open it by clicking 1,2 or 3 on the main screen and drag and size the screen to where you want your application to be placed. With it in position, click File | Get window size. This will show 4 sets of co-ordinates, if you clicked on Browser-1 then click into the respective co-ordinate edit box, right click and copy. Paste this to the end of your URL in the connections.txt file, Please note the structure and number of commas in the URL’s below, NEWURL-(Screen number):Application path,Application form name, Left px, top px, width px, height px Phrase:Banking NEWURL-100:C:\accounts,Accounts,1055,10,533,636 NEWURL-100:C:\accounts\mybank.xlsx,mybank - Excel,14,12,1043,1029 if it’s a built in browser you do not need a form name, just URL, Co-ordinates. NEWURL-1:http://mysite.html,10,20,900,700 Your applications or built in browsers will now auto size and position across a multi monitor PC exactly where you place them. TTS (Text to speech) announcements: You can add text that you would like to be announced or read out loud. You will need a TTS Reader to do this. The link below is for one that has been tested and works well with the Web Extractor & also FREE at the time of writing. Download it, launch it and then configure it ready to receive text from the Web Extractor. Click Tools | Options Tick the ‘Read Clipboard’ check box in the bottom right hand corner, Click OK Now you can add URL’s for your speech, these are formed as, SAY-(Screen number):Text to read out loud Phrase:Say hello to SAY-100:Hello [SEARCH] you are most welcome (You can say anything) Phrase:Please welcome SAY-100:Hi [SEARCH] can I take your coat . looks cheap . but I might get a quid for it on ebay This would result in you saying something like ‘Alexa please welcome Lisa’ and the text spoken out of your PC or HiFi speakers would be, Hi Lisa can I take … You can make some pretty awesome party announcements too. The text to be announced is also passed to remote devices so could potentially be spoken by another TTS set up on that device automatically. You can set up different text strings and drop them into ZONES so that depending on which device heard the question will force the appropriate announcement. If this is then mixed with sensors means you can have automated announcements, doorbells, PIR’s etc Shutting down: You can say ‘Alexa Screensaver’ and this will invoke your default screensaver without having to reach for a keyboard or mouse. You can say ‘Alexa Sleep’ and this will shut down your monitors to standby. You would need to move your mouse or use the keyboard to wake Alexa and your screens up again. You can run the Web extractor on multiple PC’s on the same network simultaneously and they will all respond as though it’s just one big computer. Each PC would still have its own connections.txt file but all fired off from the same questions you put to Alexa. Open source: The built in web server launches the ../webserver/library/index.html file. This is directly editable by the user if you know how to write javascript or jquery scripts.  You could also set up a full php server on your PC and have the Web extractor fire off information or search results directly into a MySQL database. Or link the URL’s to a ‘Hub’ through the PC. USB control is also a possibility with far more intricate switching and sequencing.
Web Extractor Instructions
You can have the Web Extractor up and running within a few seconds of downloading it where it should work instantly and out-of-the-box. This page will show you other features that you can then learn about at your own pace where virtually all of them entails a single line of text to be pasted into a configuration file and it’s then ready to run.
Download Download