Naming convention minutiae

[WhiteCube]

What would you do if you had to name something that has no word in the English language?

Lets say my first function takes a string as an argument, and returns a string. What it does is replaces all \ with \\ and all newlines with \n.

My second function does the reverse. The purpose of the functions is to allow me to store multi-line text on a single line within a text file.

I have named them encode_esc and decode_esc, and that is ugly.

One school of thought suggests naming a function according to what it returns. There might be a word for this.

Times like this, I consider making a book of every programming concept I know, and giving them numbers.

 

[WhiteCube]

I don't think the number idea is a good one, that's what astronomers do (M33), and though I like the style, the readability is poor.

Inventing a new word is tempting.

escit = a string with escape sequences encoded in it

escitae : plural

antiescit : opposite

The readability of that isn't great either.

Or perhaps using an old word in a new way, like "percolate up" and "percolate down" in the context of heaps.

 

[WhiteCube]

According to gargle, programmers call strings that contain ASCII control characters, raw.

That doesn't ring a bell, but at least its short.

raw(
cooked(

raw_to_cooked(
cooked_to_raw(

y=cooked(thawed(frozen(x)))

To me, raw just means unprocessed, but I'll take it.

raw_to_escaped(
escaped_to_raw(

Job done, moving on.

 

[c1lonewolf]

create a single function named "convert_str" with the required parameter.
convert_str( option )

inside the function simply place a 'switch statement' to change the option from one to the other and return the string.

 

[WhiteCube]

create a single function named "convert_str" with the required parameter.
convert_str( option )

inside the function simply place a 'switch statement' to change the option from one to the other and return the string.

Not a bad idea. A nice bit of lateral thinking.

I did something similar when I wanted to use a symbol in a name, and the language didn't allow it.

=calc(a,b,"+")
=calc(a,b,"-")
=calc(a,b,"*")
=calc(a,b,"/")

I could have used 4 functions, but this looked better. Calc dispatched to subroutines, and the mainline didn't need to know what I named them.

 

[c1lonewolf]

not sure what language you're using but applying a switch allows multiple options mot just character switches.

convert_str(o, s){
  if(  (o || s)  == ( 'undefined' || undefined) ){ return; }
  switch ( o ){
    case 'encode' : { str = str [operation] }
    case 'decode'  : { str = str [operation] }
   case 'uppercase'  : { str = str.toUpperCase(); }
   case 'lowercase'  : { str = str.toLowerCase(); }
   case 'calc_length' : { str = str.length; }
  }
  return str;
}

 

[WhiteCube]

not sure what language you're using but applying a switch allows multiple options mot just character switches.

convert_str(o, s){
  if(  (o || s)  == ( 'undefined' || undefined) ){ return; }
  switch ( o ){
    case 'encode' : { str = str [operation] }
    case 'decode'  : { str = str [operation] }
   case 'uppercase'  : { str = str.toUpperCase(); }
   case 'lowercase'  : { str = str.toLowerCase(); }
   case 'calc_length' : { str = str.length; }
  }
  return str;
}

That is like the way I implement object oriented style in a language that doesn't support objects.

One argument is the name of the method to apply.

Which just moves the naming problem to the other side of the parenthesis.