Some Notes on the original Article

The IF statement

The BNF shown on this chapter had some bugs. The author said that <program> and <block> are as follows:

<program> ::= <block> END
<block> ::= [ <statement> ]*

That means a <block> can contain zero or more statments. However, there is no way to decide when to terminate a <block>. When dealing with "IF" without "ELSE", the problem is not yet obvious:

IF <condition> <block> ENDIF

Because the author imply that <block> should not be ended with token "END", now we can treat token "ENDIF" which is represent by character 'e' as the termination of block. That means source code like "aibced" is allowed, aka "bc" this two statements are treated as one block.

When it comes to "IF" statement with "ELSE", things just change.

IF <condition> <block> [ ELSE <block>] ENDIF

While we are dealing with the first <block>, there is no way to quit recognizing <block> and match token "ELSE". Especially when the author propose test case "aiblcede". It was expected to recognize "b" as a single block and "l" as token "ELSE". Which however is not possible because there is no way to determine the exit of <block>.

procedure Block;
begin
   while not(Look in ['e']) do begin
      case Look of
       'i': DoIf;
       'o': Other;
      end;
   end;
end;

The source code implies that every block should end with token "END" which is character 'e'. One way to fix this is to explicitly state that a <block> ends with token "END". So the test case should be "aibelceede" and the BNF should explicitly be:

<program> ::= <block>
<block> ::= [ <statement> ]* END

Another way is to tell the block matcher to recognize the "ELSE" token as the author did in the next section.

procedure Block;
begin
   while not(Look in ['e', 'l']) do begin
      case Look of
       'i': DoIf;
       'w': DoWhile;
       else Other;
      end;
   end;
end;

Also note the else branch instead of the original 'o' branch.