| Prev | Next | Start of Chapter | Next Chapter | Contents | Glossary | Index | Comments | (13 out of 13)

The syntax of the G2 4.0 error handling statements is described under Superseded On Error Statement Syntax and Superseded Signal Statement Syntax.

G2 distinguishes between the G2 4.0 and 5.0 on error and signal statements by the number of arguments in the statement. To allow code to mix the two techniques without requiring special action, G2 automatically interconverts between them. Such interconversion is possible because:

• The error-name argument to a G2 4.0 on error or signal statement is analogous to the class name of the error-class argument to the corresponding G2 5.0 statement.

• The error-text argument to a G2 4.0 on error or signal statement is analogous to the Error-description attribute of the error-class argument to the corresponding G2 5.0 statement.

• The rows of the table show the three ways in which a search for an error handler can originate.

• The columns show the two types of error handler on which the search can terminate.

  	A one-argument on-error statement	A two-argument on-error statement
  An execution error	No conversion needed.	G2 supplies the on error statement with an error-symbol and an error-text.
  A one-argument signal statement	No conversion needed.	G2 converts the error object's class name to an error-symbol and the error object's error-description to an error-text, and supplies the two arguments to the on error statement.
  A two-argument signal statement	G2 supplies the on error statement with an error object whose Error-description is error-text, and whose class name is: g2-error if error-name is error. g2-rpc-error if error-name is rpc-error. default-error otherwise.	No conversion needed.

| Prev | Next | Start of Chapter | Next Chapter | Contents | Glossary | Index | Comments | (13 out of 13)