3 Programming with Eloquence

The Integrated Development Environment (IDE)

Program Development

Major benefits are:

• You can have any number of text or program files active at one time, resticted only by memory limitations.

• The new Browse Toolwindow provides an entry for each editor window. For Eloquence programs it also provides the list of active segments. Double clicking on an entry brings the associated window on top and positions the cursor.

• Splittable editor windows and additional views make it convinient to operate at several places of the same document at one time without having to scroll around.

• Language sensitive editor.

• Integrated context sensitive online help.

• Includes its own client/server file sharing capabilities. This will make you independend of the availability of specific network file systems (NFS/SMB) and overcomes network topology and filesystem limitations.

• Consistent authorization scheme in a heterogeneous environment

• Supports the HP Roman-8 and ISO 8859-1 character sets.

• The editor is sensitive about line termination sequences.

• Integrated debugger

The Browser

Double clicking on the folder symbol will bring the associated client window on top. Double clicking on a segment name will additionally position the cursor at the beginning of that section. (This will even work if a program has been edited). Recompiling a program will re-createe its section list.

This is a real live saver if you ever had to jump between different segments (in even different files) while editing a program.

Compiling a program

Please Note: The compiler will reject to compile a program containing references to line numbers. Since the Development Environment does not deal with line numbers those programs would become useless.

Storing and loading program files

Loading a text file and converting it into a program document is easy. Just load the text file, open the Document Properties dialog and change the file type.

Debugging

Eloquence makes it possible to debug your program in different environ
ments:

• Execute your program on your local system

• Execute your program in a remote system

When initiating a remote debugging session, the Eloquence contacts the eloqd server at the remote system. The remote eloqd process will check your authorization, start the debug session and connect it to the Development Environment.

You normally don't want to execute the program in the editor window directly. It is usually part of a larger project and not self contained. Instead you create a Debug configuration which tells Eloquence how to invoke the start program of your project.

• The system where the debug process should be executed on

• The arguments to pass to eloqcore

• The environment to provide to your program

NOTE: You can only have one debugging session active at a time. Starting a debug session requires configuration of the Debug settings in the Application Properties.

Initiating a debug session

This way, the debug process will be executed. If any breakpoints are defined, it will be halted at the first breakpoint. If not, it will run until finished.

Alternatively, the debug process can be started by selecting one of the Step commands (Step into, Step over, Step out). In this case, execution of the debug process will be halted before executing the first line.

The "Run to cursor" command can also be used to start the debug process. In this case, a temporary breakpoint is set at the cursor location and then the debug process will be executed.

Whenever the debug process is suspended, the source file containing the next line to be executed will be automatically be loaded, unless it's already opened in a window. The current line is marked by a yellow arrow.

Stepping through a program

• Continue. This will continue execution until the program exists or is halted (e.g. if a breakpoint is reached).

• Step into. This will continue the execution for one line. (This is similar to the former Halt key.)

• Step over. This will continue the execution until the current line has been finished. (This is similar to the former Advanced Step.) Please note, that recursive calls are taken care of. Execution will not be suspended in a different recursion level.

• Step out. This will continue execution until the end of the current segment has been reached.

• Run to cursor. This will set a temporary breakpoint at the cursor location and continue execution until the program exits or is halted. If another breakpoint is reached before the cursor line is executed, the debug process will stop at that breakpoint and remove the temporary breakpoint from the cursor location.

Breakpoint

Breakpoints are associated with the relative line in a program segment (when the file is loaded or stored), so they may be garbaged, when you edit a particular program file outside the Development Environment. You cannot set a breakpoint unless the program has been previously compiled.

A breakpoint becomes submitted to the debug process, when the program execution has been suspended. If you add or delete a breakpoint, while the debug process is currently executing, you need to suspend execution (for example, by selecting Halt) to activate the changes.

The Breakpoints dialog (available in the Program menu pane or by pressing the accelerator key Shift-F9) can be used to review or remove breakpints without having the associated file in an editor window. When double-clicking the first column, the corresponding file is loaded automatically and the cursor is positioned in the corresponding line.

The Output Toolwindow

The Call Stack Toolwindow

The Variables Toolwindow

Debugging a character oriented program

In addition to a server, you need either a terminal or a terminal emulator (such as Reflection). After the debug process is started, the input and output is redirected to the terminal.

In order to debug a character oriented program, you must configure the TTY device in the Application Properties accordingly. To obtain the tty devicefile name, you should follow the procedure below:

• If you use a terminal device connected to a serial device, simply enter the tty device file (for example /dev/tty0p1). You should not run a getty on this port and you must configure it properly before using it (setting baud rate, data format etc.).

• - or -

2 Obtain the tty device using the tty command. The device file is typically similar to /dev/ttyp4.

3 Disable the shell input (for example by executing a sleep 10000)

4 Enter the device file in the configuration.

Please be aware, that you must define a TERM environment variable, describing the type of terminal used. After that, start the debug process as usual.

(Yes, I know - this is complicated. We are looking for a way to make this easier and even automatic.)