Humanities Home page   Office of Digital Humanities
Back     BYU LiveCode Lessons Gateway

Computers & the Humanities 210

Annotated Readings Applications

Language textbooks will often contain culture reading sections: a selection of text about the target country or culture, usually "annotated"; that is, accompanied by supplementary related material, be that additional text, illustrations, or pictures. In a computing environment, using LiveCode, we can duplicate and enhance that function by using animations, sound, and video in addition to text and pictures. This is the essence of hyperlinking. This activity will focus on the mechanics of getting the text to operate in the fashion we need.

Spanish Text

Let's start with a selection of Spanish text in a scrolling text field. Our goal is to have certain words within that text clickable or "hot" so that the user will be taken to a glossary of sorts. The glossary can be either data fields on the card itself, or a separate stack (or any of a half-dozen other methods).

Beginnings

First of all, we have to decide what we want to accomplish and think through the various possibilities and requirements. Working languages is tricky because we have to account for different forms, or inflections, of certain words. We want to make sure we account for those. For the glossary, we would create a card separate from the readings with three fields:

When designed as a separate stack, the Glossary stack would contain one card per key word. Remember that if you want to access another stack, you must either provide the absolute file path to that stack, or you must have previously set the defaultFolder to the folder holding the current stack, as we learned in an earlier lesson.

Ideally we want to allow the student to click on any word in the text and view the corresponding information in the glossary. This would be facilitated by placing a handler in the field's script (don't forget to lock the text of the field):

# This assumes that the defaultFolder has already been set
on mouseUp
  put the clickText into whichWord
  select the clickChunk
  set the cursor to watch
  push card
  lock screen
  go to stack "Spanish Glossary.rev"
  find word whichWord
  if the result is "Not Found" then --word was not found
    answer "The word '" & whichWord & "' was not found in the glossary."
    pop card
    exit mouseUp
  end if
  unlock screen with dissolve
end mouseUp

The use of the clickText function in the first line of the handler gives us the text of the word clicked and puts it into a variable. The user is then taken to the glossary where a search for the text is performed, restricting the search to key fields. A successful search will display the desired information. The text clicked is selected to give some visual feedback to the user to indicate that their click was registered.

Now comes the question of how to get back to the text after a successful search. The push card statement in the mouseUp handler preserves card's ID in memory. Then we need a button with the glossary with this handler in its script:

on mouseUp
  pop card
end mouseUp

This will retrieve the ID of the card with the text, which we "pushed" earlier, and return the user to that card. When set up in this manner, we could have several readings that reference the glossary, and the button will return them to the proper text of origin each time.

We have to account for the inflected forms. So we can set the open the Property Inspector for each field and check the Find command ignores checkbox for each field in the stack that we want the find command to search. (This approach is usually practical only when the glossary is a separate stack).

It is good design to account for all probable possibilities, so we should always plan for the possibility the word will not be found, even if we think they should all be there. The if control statement in the field's handler will trap all unsuccessful searches. A number of LiveCode commands return an error message if unsuccessful. This message is often accessible using a function called the result. In this case, if the search is unsuccessful, the result function returns not found. If the result is empty, that is an indication of a successful search. In this case we don't want to stay in the glossary (looking at the wrong word) so we pop card to return to the text and give an informative word-not-found message.

We also spare the user having to watch while we locate the word in the glossary. We can issue the lock screen command before going to the next card, and then unlock screen after we arrive there. We can also add a an appropriate visual effect to give a clue that the context has changed to the other stack. Notice that you can use the visual effect command in conjunction with a hide/show command.

Link Text

You may have noticed Link option under the Text menu. It's not one of the text styles that you would usually choose. In fact, specifying any text as Link results in it appearing much like the links you see in web browsers. Link style allows a programmer to override LiveCode's definition of a word for the purpose of reporting the clickChunk or the clickText. All contiguous characters that have been linked are considered to be one unit.

In this application we could use it to deal with "bound forms," i.e., multiple words that function as a semantic unit. Suppose zapatos de tenis were an idiomatic expression or slang term (like tennie runners). We can select the entire phrase and link the words. They now will be handled as a unit: they will highlight and be searched for in the glossary.

As mentioned before, linked text initially looks like links in a web browser, i.e., underlined and with a blue color. For our reading text, because the assumption is that all words in the text are clickable, it would confusing to have only some words colored and underlined. Fortunately, LiveCode provides a way to change the appearance of linked text without changing its function. The the Property Inspector for the stack, opend the Colors & Patterns panel. There we can choose the color of the three states of linked text. We need to change all of these to black (or to whatever color we decide the text should be). Under the Basic Properties for a stack we find the Underline links check box. By deselecting that, none of the text is underlined. Now any text has a uniform appearance, but still functions as designed.

Shakespeare Text

Now we'll look at a different kind of Annotated Reading where it is assumed the reader already knows most of the words in the text and only needs help with unfamiliar terms. The example here is the famous witches' cauldron scene from Shakespeare's Macbeth. Even native speakers of English need help with Shakespeare!

In a printed edition of Macbeth you would typically see difficult terms marked with superscript numbers that cross-reference to notes at the bottom of the page or along the side. We would like to emulate this with a computer application. As you have probably surmised, this is similar to our previous problem, but in this case we want only certain words to be clickable. To facilitate this, we create a separate fields with a list of the unfamiliar terms which we wish to define (now this could be a separate stack or separate card, like the Spanish glossary). We'll put one word/phrase per line and we'll also create a parallel list of the definitions (again, one definition per line), making sure that they correspond to the wordlist (e.g., line 1 in the wordlist corresponds to line 1 in the definition list). If the Don't wrap property is set on each field, then it makes it easier to ensure that you have a line to line correspondence between the two fields. On the main text card we'll link the unfamiliar terms in the text itself (whether they are a phrase or a single word). We'll create a button named "Show Vocabulary" and script it to alternately show and hide links, to give users a visual cue as to which words have been defined.

We would now want appropriate mouse event handlers in the script of the text field to handler all requests for supplementary information:

 on mouseDown
   if the textStyle of the clickText is link then
	 put the clickText into whichWord
   else
	 exit mouseDown
   end if
   put lineOffset(whichWord, fld "Keys" of card "Shakespeare Wordlist") into defLine
   if defLine is 0 then --word was not found
	 answer "The word '" & whichWord & "' was not found in the glossary."
	 exit mouseDown
   end if
   get line defLine of fld "meanings" of card "Shakespeare Wordlist"
   put it into fld "defBox"
   get the clickLoc
   add 10 to item 2 of it
   set the left of field "defBox" to item 1 of it
   set the top of fld "defBox" to item 2 of it
   show fld "defBox"
 end mouseDown
 
 on mouseUp
   hide fld "defBox"
 end mouseUp
 
 on mouseRelease
   hide fld "defBox"
 end mouseRelease

First we need to check if the clickChunk is link style. If so, we'll put it into a variable; if not, we'll ignore it and abort the handler with the exit command. We can thus effectively eliminate other words. We need to then search the wordlist and find the word in a specific field. If the field is on a different card, as this one is, we would also need to specify the card. Since we only need to search one field, we'll elect to use the lineoffset function, which returns the line number on which the search term was found. The line number is all we need from this in order to get corresponding line from definition field (and consequently the proper definition). Having acquired the proper definition, we can put the definition into a field to display.

Let's take advantage of the power of the computer and have the definition appear next to the word being defined. We employ a mouseDown handler to have the definition appear when the mouse is down and then have corresponding mouseUp and mouseRelease handlers to hide the field when the mouse is released. We need to position the field with the defintion relative to where the user clicked. We can get the clickLoc (it will be in x,y format) and add 10 to the vertical coordinate (the second item, or y, in the clickLoc). We can then use those coordinates to set the top and left of the field with the definition. Since we added 10 the vertical coordinate, the definition will appear below the word/phrase being defined.

Another Approach

As mentioned above, there are many approaches to this kind of word lookup task. Here is one that is more complex, but creates a pleasing word highlight effect. We will not go over this one in class, but if you can study it and adapt it for use in your project if you wish.

Let's create a hyperlinked set of scriptures by employing a method of bringing the user's attention to linked items within the text. This can be accomplished by highlighting the linked text as the user moves the mouse over it. We can also link other actions to this event, such as displaying the footnotes attached to that particular word or phrase. The controlling handler would be something similar to this:

local prevHilite
on mouseMove
  if prevHilite is not empty and prevHilite is not the mouseChunk then
    set the foregroundColor of prevHilite to "black"
    set the backgroundColor of prevHilite to empty
    put empty into prevHilite
    put empty into field "footnote"
  end if
  if the mouseChunk is not empty and "link" is in the textStyle of the mouseChunk then
    set the foregroundColor of the mouseChunk to "green"
    set the backgroundColor of the mouseChunk to "dark green"
    put the mouseChunk into prevHilite
    put the mouseText into thisRef
    put word 2 of the mouseLine into thisLine
    repeat forever
	find whole thisRef in field "wordList"
	if the result is "Not found" then
	  exit mouseMove
	end if
	put word 2 of the foundLine into possLine
	put line possLine of field "wordList" into footnote
	set the itemDelimiter to "•"
	if item 1 of footnote = thisLine then 
	  exit repeat
	end if
    end repeat
    put line possLine of field "footnotes" into field "footnote"
    find empty
  end if
  pass mouseMove
end mouseMove

on mouseLeave
  if prevHilite is not empty then
    set the foregroundColor of prevHilite to "black"
    set the backgroundColor of prevHilite to empty
    put empty into prevHilite
    put empty into field "footnote"
  end if
  pass mouseLeave
end mouseLeave

As with previous methods, this employs two data fields, one with the list of linked text and the other with the footnote data. We also need to distinguish between identical phrases with different associated footnotes. So each word/phrase in the word list field has a number associated with it identifying it with the verse to which it is attached. Again, the word list and the footnote fields correspond line to line.

The handler begins with a script local variable, meaning it is accessible to each handler within the script. This will contain the chunk data for the highlighted word or phrase. We intercept the mouseMove message to highlight the appropriate text with the user's movement of the mouse over the passage. First we check the contents of the variable to see if it contains chunk data and compare it to the present location of the mouse. If different, then the color attributes of the text are set back to normal, the variable is emptied, and the footnote field is cleared. This is the portion of the handler that un-highlights the text upon leaving it.

The next control structure is the poriton that highlights the text that are footnoted. It first checks to ensure that the user is indeed over a word and that the word is linked. Once those conditions are met, the color attributes of the text are changed to indicate its special status. We also store the chunk expression of that text in the special variable (to remove the highlight later as described above). We obtain the actual text and line number and start our search. It is set up such that the text searched for (thisRef) must also be associated with the particular verse (thisLine). The repeat forever loop will continue until both conditions are met (use with caution!). Once found, the correct footnote information is displayed. The find empty is necessary for the next footnote to be found correctly.

Technically the mouseLeave handler set up to remove the highlight is overkill and not necessary, as the first control structure in the previous handler should remove the highlight in all cases. However, you should always account for the possibility that the user may be quick and that the processor may be slow. This handler will "mop up" in the off chance the previous handler misses it.

Conclusion

Once again, these methods are not necessarily the most polished approach to the task at hand. They are definitely not the only way to approach these tasks. For example, we have been dealing exclusively with text. There is the possibility of linking to pictures, sound, and video if one wishes. You have at your disposal many tools to achieve these same ends, and often in a more elegant and resourceful manner. Trust yourself and use your creativity.


Now It's Your Turn

Your assignment, then, is to create an annotated reading with hyperlinked text. (If you want your final project to have an activity like this in it, you might want to consider using this assignment as a "building block" for your project, and use textual material that relates to your project topic.)

  1. Create a stack and entitle it YourName-AnnotatedReading. Give the stack an appropriate label (based on the language/subject with which you are working).
  2. Choose one of approaches outlined in these notes or one of your own liking where at least ten words are hyperlinked to supplementary information. If you decide to create a glossary, make sure that it accompanies your stack by putting them both within a folder (appropriately named, of course).
  3. As in all assignments, remember to observe LiveCode best practices.
  4. Put a copy of your finished stack (or folder with stacks) in the Assignments folder.

Back     BYU LiveCode Lessons Gateway
Maintained by Devin Asay.
Copyright © 2005 Brigham Young University