PIC

Code::Blocks

Manual

Version 2.0.1 beta

Thanks to the Code::Blocks team:

Anders F. Bj÷rklund (afb), Biplab Kumar Modak (biplab), Bartomiej wiecki (byo), Paul A. Jimenez (ceniza), Koa Chong Gee (cyberkoa), Daniel Orb (daniel2000), Lieven de Cock (killerbot), Yiannis Mandravellos (mandrav), Mispunt (mispunt), Martin Halle (mortenmacfly), Jens Lody (jens), Jerome Antoine (dje), Damien Moore (dmoore), Pecan Heber (pecan), Ricardo Garcia (rickg22), Thomas Denk (thomasdenk), tiwag (tiwag), stahta01 (stahta01), oBFusCATed (oBFusCATed), BlueHazzard (BlueHazzard)

And many other contributors...

Original manual in English and in German (V1.x) by Mario Cupelli (mariocup)
Translation in French (v1.x), corrections and additions for version 2 by Gerard Durand (gd_on).

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.

updated in october 2020



www.codeblocks.org

1  Code::Blocks Project Management

The instructions in several paragraphs (for example chapter 3 on page 224 or chapter 2 on page 82) are official documentations of the Code::Blocks Wiki site (eventually reviewed and amended) and available in english only. This documentation is an extension of the original version 1.1, compiled and/or written by Mario Cupelli.

The below illustration shows the design of the Code::Blocks user interface.


PIC

Figure 1.1: IDE Code::Blocks

Management
This window contains the interface ’Projects’ which will in the following text be referred to as the project view. This view show all the projects opened in Code::Blocks at a certain time. The ’Symbols’ tab of the Management window shows symbols, variables etc..
Editor
In the above illustration, a source named hello.c is opened with syntax highlighting in the editor.
Open files list
shows a list of all files opened in the editor, in this example: hello.c.
CodeSnippets
can be displayed via the menu ’View’ ’CodeSnippets’ . Here you can manage text modules, links to files and links to urls.
Logs & others
. This window is used for outputting search results, log messages of a compiler etc..

The status bar gives an overview of the following settings:

Code::Blocks offers a very flexible and comprehensive project management. The following text will address only some of the features of the project management.

1.1  Project View

In Code::Blocks, the sources and the settings for the build process are stored in a project file <name>.cbp. C/C++ sources and the corresponding header files are the typical components of a project. The easiest way to create a new project is executing the command ’File’ ’Project’ and selecting a wizard. Then you can add files to the project via the context menu ’Add files’ in the Management window.

Code::Blocks governs the project files in categories according to their file extensions. These are the preset categories:

Sources
includes source files with the extensions *.c;*.cpp;.
ASM Sources
includes source files with the extensions *.s;*.S;*.ss;*.asm.
Headers
includes, among others, files with the extension *.h;.
Resources
includes files for layout descriptions for wxWidgets windows with the extensions *.res;*.xrc;. These file types are shown in the ’Resources’ tab of the Manangement window.

The settings for types and categories of files can be adjusted via the context menu ’Project tree’ ’Edit file types & categories’ . Here you can also define custom categories for file extensions of your own. For example, if you wish to list linker scripts with the *.ld extension in a category called Linkerscript, you only have to create the new category.

Note:
If you deactivate ’Project tree’ ’Categorize by file types’ in the context menu, the category display will be switched off, and the files will be listed as they are stored in the file system.

1.2  Notes for Projects

In Code::Blocks, so-called notes can be stored for a project. These notes should contain short descriptions or hints for the corresponding project. By displaying this information during the opening of a project, other users are provided with a quick survey of the project. The display of notes can be switched on or off in the Notes tab of the Properties of a project.

1.3  Project Templates

Code::Blocks is supplied with a variety of project templates which are displayed when creating a new project. However, it is also possible to store custom templates for collecting your own specifications for compiler switches, the optimisation to be used, machine-specific switches etc. in templates. These templates will be stored in the Documents and Settings\<user>\Application Data\codeblocks\UserTemplates directory in Win 7 (or an equivalent path in the user profile, adapted to each OS). If the templates are to be open to all users, they have to be copied to a corresponding directory of the Code::Blocks installation. These templates will then be displayed at the next startup of Code::Blocks under ’New’ ’ Project’ ’User templates’ .

Note:
The available templates in the Project Wizard can be edited by selection via right-click.

1.4  Create Projects from Build Targets

In projects it is necessary to have different variants of the project available. Variants are called Build Targets. They differ with respect to their compiler options, debug information and/or choice of files. A Build Target can also be outsourced to a separate project. To do so, click ’Project’ ’Properties’ , select the variant from the tab ’Build Targets’ and click the ’Create project from target’ button (see Figure 1.2 on page 18).


PIC

Figure 1.2: Build Targets

1.5  Virtual Targets

Projects can be further structured in Code::Blocks by so-called Virtual Targets. A frequently used project structure consists of two Build Targets, one ’Debug’ Target which contains debug information and one ’Release’ Target without this information. By adding Virtual Targets via ’Project’ ’Properties’ ’Build Targets’ individual Build Targets can be combined. For example, a Virtual Target ’All’ can create the Targets Debug and Release simultaneously. Virtual Targets are shown in the symbol bar of the compiler under Build Targets.

1.6  Pre- and Postbuild steps

Code::Blocks makes it possible to perform additional operations before or after compiling a project. These operations are called Prebuilt or Postbuilt Steps. Typical Postbuilt Steps are:

Example

Creating a Disassembly from an object under Windows. Piping to a file requires calling cmd with the /c option.

cmd /c objdump -D name.elf > name.dis

Archiving a project can be another example for a Postbuilt Step. For this purpose, create a Build Target ’Archive’ and include the following instruction in the Postbuilt Step:

zip -j9 $(PROJECT_NAME)_$(TODAY).zip src h obj $(PROJECT_NAME).cbp

With this command, the active project and its sources, header and objects will be packed as a zip file. In doing so, the Built-in variables $(PROJECT_NAME) and $(TODAY), the project name and the current date will be extracted (see section 3.2 on page 225). After the execution of the Target ’Archive’, the packed file will be stored in the project directory.

In the share/codeblocks/scripts directory you will find some examples for scripts. You can add a script via menu ’Settings’ ’Scripting’ and register in a menu. If you execute e.g. the script make_dist from the menu then all files belonging to a project will be compressed in an archive <project>.tar.gz.

1.7  Adding Scripts in Build Targets

Code::Blocks offers the possibility of using menu actions in scripts. The script represents another degree of freedom for controlling the generation of your project.

Note:
A script can also be included at a Build Target.

1.8  Workspace and Project Dependencies

In Code::Blocks, multiple projects can be open. By saving open projects via ’File’ ’Save workspace’ you can collect them in a single workspace under <name>.workspace. If you open <name>.workspace during the next startup of von Code::Blocks, all projects will show up again.

Complex software systems consist of components which are managed in different Code::Blocks projects. Furthermore, with the generation of such software systems, there are often dependencies between these projects.

Example

A project A contains fundamental functions which are made available to other projects in the form of a library. Now, if the sources of this project are modified, then the library has to be rebuilt. To maintain consistency between a project B which uses the functions and project A which implements the functions, project B has to depend on project A. The necessary information on the dependencies of projects is stored in the relevant workspace, so that each project can be created separately. The usage of dependencies makes it also possible to control the order in which the projects will be generated. The dependencies for projects can be set via the selecting the menu ’Project’ ’Properties’ and then clicking the ’Project’s dependencies’ button.

1.9  Including Assembler files

In the Management window of the Project View, Assembler files are shown in the ASM Sources category. The user can change the listing of files in categories (see section 1.1 on page 14). Right-clicking one of the listed Assembler files will open a context menu. Select ’Properties’ to open a new window. Now select the ’Build’ tab and activate the two fields ’Compile file’ and ’Link file’. Then select the ’Advanced’ tab and execute the following steps:

  1. Set ’Compiler variable’ to CC
  2. Select the compiler under ’For this compiler’
  3. Select ’Use custom command to build this file’
  4. In the window, enter:
    $compiler $options $includes <asopts> -c $file -o $object

The Code::Blocks variables are marked by $ (see section 3.4 on page 236). They are set automatically so that you only have to replace the Assembler option <asopt> by your own settings.

1.10  Editor and Tools

This section describe tools in the Úditor window.

1.10.1  Default Code

The company’s Coding Rules require source files to have a standard design. Code::Blocks makes it possible to include a predefined content at the beginning of a file automatically when creating new C/C++ sources and headers. This predefined content is called default code. This setting can be selected under ’Settings’ ’Editor’ Default Code. If you create a new file then a macro expansion of variables, e.g. defined via menu ’Settings’ ’Global variables’ , is performed. A new file can be created via the menu ’File’ ’New’ ’File’ .

Example

/*************************************************************** 
 *  Project: $(project) 
 *  Function: 
 *************************************************************** 
 *  $Author: mario $ 
 *  $Name:  $ 
 *************************************************************** 
 * 
 *  Copyright 2007 by company name 
 * 
 ***************************************************************/

1.10.2  Abbreviation

A lot of typing can be saved in Code::Blocks by defining abbreviation. This is done by selecting ’Settings’ ’Editor’ and defining the abbreviations under the name <name>, which can then be called by the keyboard shortcut Ctrl-J (see Figure 1.3 on page 25).


PIC

Figure 1.3: Defining abbreviations

Parametrisation is also possible by including variables $(NAME) in the abbreviations.

#ifndef $(Guard token) 
#define $(Guard token) 
#endif // $(Guard token)

When performing the abbreviation <name> in the source text and performing Ctrl-J, the content of the variable is requested and included.

1.10.3  Personalities

Code::Blocks settings are saved as application data in a file called <user>.conf in the codeblocks directory. This configuration file contains information such as the last opened projects, settings for the editor, display of symbol bars etc. By default, the ’default’ personality is set so that the configuration is stored in the file default.conf. If Code::Blocks is called from the command line with the parameter --personality=myuser, the settings will be stored in the file myuser.conf. If the profile does not exist already, it will automatically be created. This procedure makes it possible to create the corresponding profiles for different work steps. If you start Code::Blocks from the command line with the additional parameter--personality=ask, a selection box will be displayed for all the available profiles.

Note:
The name of the current profile/personality is displayed in the right corner of the status bar.

1.10.4  Configuration Files

The Code::Blocks settings are stored in the default.conf profile in the codeblocks directory of your Application Data. When using personalities (see subsection 1.10.3 on page 26), the configuration details will be stored in the <personality>.conf file.

The tool cb_share_conf, which can be found in the Code::Blocks installation directory, is used for managing and storing these settings.

If you wish to define standard settings for several users of a computer, the configuration file default.conf has to be stored in the directory \Documents and Settings\Default User\Application Data\codeblocks. During the first startup, Code::Blocks will copy the presettings from ’Default User’ to the application data of the current users.

To create a portable version of Code::Blocks on a USB stick, proceed as follows. Copy the Code::Blocks installation to a USB stick and store the configuration file default.conf in this directory. The configuration will be used as a global setting. Please take care that the file is writeable, otherwise changes of the configuration cannot be stored.

1.10.5  Navigate and Search

In Code::Blocks there are different ways of quick navigation between files and functions. Setting bookmarks is a typical procedure. Via the shortcut Ctrl-B a bookmark is set or deleted in the source file. Via Alt-PgUp you can jump to the previous bookmark, and via Alt-PgDn you can jump to the next bookmark.

If you select the workspace or a project in the workspace in the project view you will be able to search for a file in the project. Just select ’Find file’ from the context menu, then type the name of the file and the file will be selected. If you hit return this file will be opened in the editor (see Figure 1.4 on page 29).


PIC

Figure 1.4: Searching for files

In Code::Blocks you can easily navigate between header/source files like:

  1. Set cursor at the location where a header file is included and open this file via the context menu ’open include file’ (see Figure 1.5 on page 32)
  2. Swap between header and source via the context menu ’Swap header/source’
  3. Select e.g. a define in the editor and choose ’Find declaration’ from the context menu to open the file with its declaration.

PIC

Figure 1.5: Opening of a header file

Code::Blocks offeres several ways of searching within a file or directory. The dialogue box for searching is opened via ’Search’ ’Find’ (Ctrl-F) or ’Find in Files’ (Ctrl-Shift-F).

Alt-G and Ctrl-Alt-G are another useful functions. The dialogue which will open on using this shortcut, lets you select files/functions and then jumps to the implementation of the selected function (see Figure 1.6 on page 35) or opens the selected file in the editor. You may use wildcards like * or ? etc. for an incremental search in the dialog.


PIC

Figure 1.6: Search for functions

Note:
With the Ctrl-PgUp shortcut you can jump to the previous function, and via Ctrl-PgDn you can jump to the next function.

In the editor, you can open a new Open Files dialog via Ctrl-Tab and you can switch between the listed entries. If the Ctrl-key is pressed, then a file can be selected in different ways:

  1. If you select an entry with the left mouse button, then the selected file will be opened.
  2. If you press the Tab-key you will switch between the listed entries. Releasing the Crtl-key will open the selected file.
  3. If you move the mouse over the listed entries, then the current selection will be highlighted. Releasing the Crtl-key will open the selected file.
  4. If the mouse pointer is outside the highlighted selection, then you can use the mouse-wheel to switch between the entries. Releasing the Crtl-key will open the selected file.

A common procedure when developing software is to struggle with a set of functions which are implemented in different files. The Browse Tracker plugin will help you solve this problem by showing you the order in which the files were selected. You can then comfortably navigate the function calls (see section 2.4 on page 110).

The display of line numbers in Code::Blocks can be activated via ’Settings’ ’General Settings’ in the field ’Show line numbers’. The shortcut Ctrl-G or the menu command ’Search’ ’Goto line’ will help you jump to the desired line.

Note:
If you hold the Ctrl key and then select text in the Code::Blocks editor you can perform e.g. a Google search via the context menu.

1.10.6  Symbol view

The Code::Blocks Management window offers a tree view for symbols of C/C++ sources for navigating via functions or variables. As the scope of this view, you can set the current file or project, or the whole workspace.

Note:
Entering a search term or symbol names in the ’Search’ input mask of the Symbol Browser results in a filtered view of the symbols if any hits occurred.

The following categories exist for the symbols:

Global functions
Lists the implementation of global functions.
Global typedefs
Lists the use of typedef definitions.
Global variables
Displays the symbols of global variables.
Preprocessor symbols
Lists the pre-processor directives created by #define.
Global macros
Lists macros of pre-processor directives.

PIC

Figure 1.7: Symbol view

Structures and classes are displayed in the ’bottom tree’ and the sort sequence can be modified via the context menu. If a category is selected by mouse-click, the found symbols will be displayed in the lower part of the window (see Figure 1.7 on page 39). Double-clicking the symbol will open the file in which the symbol is defined or the function implemented, and jumps to the corresponding line. An auto-refresh of the symbol browser without saving a file, can be activated via the menu ’Settings’ ’Editor’ ’Code Completion’ (see Figure 1.8 on page 42). For projects with many symbols the performance within Code::Blocks will be affected.


PIC

Figure 1.8: Enable real-time parsing

Note:
In the editor, a list of the classes can be displayed via the context menus ’Insert Class method declaration implementation’ or ’All class methods without implementation’ .

1.10.7  Including external help files

Code::Blocks comes with just its own help file: normally, developers need much more help and reference for language, libraries, protocols, file formats and so on. The help plugin makes all the needed documentation available from within Code::Blocks itself. Virtually any document can be integrated in the Code::Blocks help system, since the help plugin has the ability to launch external programs, if necessary, to view the added documents.

Once added a new help file or document, there will be a new entry in the ”Help” menu to open it.

The Code::Blocks development environment supports the inclusion of external help files via the menu ’Settings’ ’Environment’ . Include the manual of your choice in the chm format in ’Help Files’ select ’this is the default help file’ (see Figure 1.9 on page 45). The entry $(keyword) is a placeholder for a select item in your editor. Now you can select a function in an opened source file in Code::Blocks by mouse-click, and the corresponding documentation will appear while pressing F1.

If you have included multiple help files, you can select a term in the editor and choose a help file from the context menu ’Locate in’ for Code::Blocks to search in.


PIC

Figure 1.9: Settings for help files

In Code::Blocks you can add even support for man pages. Just add a entry ’man’ and specify the path as follows.

man:/usr/share/man

On Linux, man pages are usually installed anyway, for Windows you might want to download them e.g. from here: http://www.win.tue.nl/~aeb/linux/man

Helpfile options

Code::Blocks provides an ’Embedded HTML Viewer’, which can be used to display simple html file and find keywords within this file. Just configure the path to the html file, which should be parsed and enable the checkbox ’Open this file with embedded help viewer’ via the menu ’Settings’ ’Environment’ ’Help Files’ .


PIC

Figure 1.10: Embedded HTML Viewer

Note:
If you select a html file with a double-click within the file explorer (see section 2.8 on page 122) then the embedded html viewer will be started, as long as no association for html files is made in file extensions handler.

CHM Files

You can find c++ chm help files on the web. Just add them via the dialog.

For Linux you have to install a chm viewer to be able to open chm files. There are severals like gnochm, kchmviewer, xchm and so on.

1.10.8  Including external tools

Including external tools is possible in Code::Blocks via ’Tools’ ’Configure Tools’ ’Add’ . Built-in variables (see section 3.2 on page 225) can also be accessed for tool parameters. Furthermore there are several kinds of launching options for starting external applications. Depending on the option, the externally started applications are stopped when Code::Blocks is quit. If the applications are to remain open after quitting Code::Blocks, the option ’Launch tool visible detached’ must be set.

1.11  Tips for working with Code::Blocks

In this chapter we will present some useful settings in Code::Blocks.

1.11.1  Tracking of Modifications

Code::Blocks provides a feature to track modifications within a source file and to show a bar in the margin for the changes. Modifications are marked with a yellow changebar and modifications that are already saved will use a green changebar (see Figure 1.11 on page 52). You can navigate between your changes via the menu ’Search’ ’Goto next changed line’ or ’Search’ ’Goto previous changed line’ . The same functionality is also accessible via the shortcuts Ctrl-F3 and Ctrl-Shift-F3.


PIC

Figure 1.11: Tracking of modifications

This feature can be enabled or disabled with the checkbox ’Use Changebar’ in the menu ’Settings’ ’Editor’ ’Margins and caret’ .

Note:
If a modified file is closed, then the changes history like undo/redo and changebars get lost. Via the menu ’Edit’ ’Clear changes history’ or the corresponding context menu you are able to clear the changes history even if the file is kept open.

1.11.2  Data Exchange with other applications

Data can be exchanged between Code::Blocks and other applications. For this interprocess communication DDE (Dynamic Data Exchange) is used for windows and under different operating systems it is a TCP based communication.

With this interface different commands with the following syntax can be sent to a Code::Blocks instance.

[<command>("<parameter>")]

These commands are currently available:

Open

The command

[Open("d:\temp\test.txt")]

uses the parameter, in our case it is a file specified with an absolute path, and opens it in an existing Code::Blocks instance or starts a first instance if required.

OpenLine

This command opens a file at a given line number in a Code::Blocks instance. The line number is specified with :line.

[OpenLine("d:\temp\test.txt:10")]
Raise

Set the focus to the Code::Blocks instance. A parameter must not be passed.

1.11.3  Configuring environmental variables

The configuration for an operating system is specified by so-called environmental variables. The environmental variable PATH for example contains the path to an installed compiler. The operating system will process this environmental variable from beginning to end, i.e. the entries at the end will be searched last. If different versions of a compiler or other applications are installed, the following situations can occur:

So it might be the case that different versions of a compilers or other tools are mandatory for different projects. One possibility in such a case is to change the environmental variables in the system control for every project. However, this procedure is error-prone and not flexible. For this requirement, Code::Blocks offers an elegant solution. Different configurations of environmental variables can be created which are used only internally in Code::Blocks. Additionally, you can switch between these configurations. The Figure 1.12 on page 57 shows the dialogue which you can open via ’Environment Varibales’ under ’Settings’ ’Environment’ . A configuration is created via the ’Create’ button.


PIC

Figure 1.12: Environmental variables

Access and scope of the environmental variables created here, is limited to Code::Blocks. You can expand these environmental variables just like other Code::Blocks variables via $(NAME).

Note:
A configuration for the environmental variable for each project can be selected in the context menu ’Properties’ of the ’EnvVars options’ tab.

Example

You can write the used environment into a postbuild Step (see section 1.6 on page 19) in a file <project>.env and archive it within your project.

cmd /c echo \%PATH\%  > project.env

or under Linux

echo \$PATH > project.env

1.11.4  Switching between perspectives

Depending on the task in hand, it can be useful to have different configurations or views in Code::Blocks and to save these configurations/views. By default, the settings (e. g. show/hide symbol bars, layout, etc.) are stored in the default.conf configuration file. By using the command line option --personality=ask during the start of Code::Blocks, different settings can be selected. Apart from this global setting, a situation might occur where you wish to switch between different views of windows and symbol bars during a session. Editing files and debugging projects are two typical examples for such situations. Code::Blocks offers a mechanism for storing and selecting different perspectives to prevent the user from frequently having to open and close windows and symbol bars manually. To save a perspective, select the menu ’View’ ’Perspectives’ ’Save current’ and enter a name at <name>. The command ’Settings’ ’Editor’ ’Keyboard shortcuts’ ’View’ ’Perspectives’ <name>’ allows a keyboard shortcut to be defined for this process. This mechanism makes it possible to switch between different views by simply using hot keys.

Note:
Another example is editing a file in Full Screen mode without symbol bars. You can create a perspective such as ’Full’ and assign a hot key for this purpose.

1.11.5  Switching between projects

If several projects or files are opened at the same time, the user needs a way to switch quickly between the projects or files. Code::Blocks has a number of shortcuts for such situations.

Alt-F5
Activates the previous project from the project view.
Alt-F6
Activates the next project from the project view.
F11
Switches within the editor between a source file <name>.cpp and the corresponding header file <name>.h

1.11.6  Extended settings for compilers

During the build process of a project, the compiler messages are displayed in the Messages window in the Build Log tab. If you wish to receive detailed information, the display can be extended. For this purpose click ’Settings’ ’Compiler and Debugger’ and select ’Other Settings’ in the drop-down field.


PIC

Figure 1.13: Setting detail information

Take care that the correct compiler is selected. The ’Full command line’ setting in the Compiler Logging field outputs the complete information in the Build Log. In addition, this output can be logged in a HTML file. For this purpose select ’Save build log to HTML file when finished’. Furthermore, Code::Blocks offers a progress bar for the build process in the Build Log window which can be activated via the ’Display build progress bar’ setting.

1.11.7  Zooming within the editor

Code::Blocks offers a very efficient editor. This editor allows you to change the size in which the opened text is displayed. If you use a mouse with a wheel, you only need to press the Ctrl key and scroll via the mouse wheel to zoom in and out of the text.

Note:
With the shortcut Ctrl-Numepad-/ or with the menu ’Edit’ ’Special commands’ ’Zoom’ ’Reset’ the original font size of the active file in the editor is restored.

1.11.8  Wrap Mode

When editing text files, e. g. *.txt, within Code::Blocks, it might be useful to have the text wrapped, meaning long lines will be displayed in several lines on the screen so that they can be properly edited. The ’Word wrap’ function can be activated via ’Settings’ ’Editor’ ’Other Options’ or by setting the checkbox ’Word wrap’ . The Home and End keys position the cursor at the beginning or end of wrapped lines respectively. When setting ’Settings’ ’Editor’ ’Other Options’ and ’Home key always move to caret to first column’ , the cursor will be positioned at the beginning or end of the current line respectively, if the Home or End keys are pressed. If positioning the cursor at the beginning of the first line of the current paragraph is desired, the key combination ’Alt-Home’ is to be used. The same applies analogously for ’Alt-End’ for positioning the cursor at the end of the last line of the current paragraph.

1.11.9  Select modes in editor

Code::Blocks supports different modes for selecting or pasting of strings.

  1. With the left mouse button a text in the active editor can be selected and then the mouse button can be released. With the mouse wheel the user can scroll to a position. If the middle mouse button is pressed then the formerly selected text will be inserted. This feature is available per file and can be seen a clipboard per file.
  2. Pressing the ’ALT’ key will activate the so-called block-select mode and a rectangle selection can be raised with the left mouse button. If the Alt key is released this selection can be copied or pasted. This feature is helpful if you want to select some columns e.g. of an array and copy and paste the content.
  3. In the menu ’Settings’ ’Editor’ ’Margins and Caret’ so-called ’Virtual Spaces’ can be activated. This option enables that a selection in the block select mode can start or end within an empty line.
  4. In the menu ’Settings’ ’Editor’ ’Margins and Caret’ the ’Multiple Selection’ can be activated. While holding the Ctrl-key the user can select different lines in the active editor via the left mouse button. The selections will be appended in the clipboard via the shortcut Ctrl-C or Ctrl-X. Ctrl-V will insert the content at the current cursor position. An additional option called ’Enable typing (and deleting)’ can be activated for multiple selections. This feature is useful if you want to add a pre-processor directive like #ifdef at different source lines or if you want to overwrite or replace a text at several positions.

Note:
Most Linux window managers use ALT-LeftClickDrag to move a window, so you will have to disable this window manager behavior first for block select to work.

1.11.10  Code folding

Code::Blocks supports so called code folding. With this feature you can fold e.g. functions within the Code::Blocks editor. A folding point is marked by minus symbol in the left margin of the editor view. In the margin the beginning and the end of a folding point is visible as vertical line. If you click the minus symbol with the left mouse button the code snippet will be folded or unfolded. Via the menu ’Edit’ ’Folding’ you can select the folding. In the editor you see folded code as continous horizontal line.

Note:
The folding style and the folding depth limit can be configured via menu ’Settings’ ’Editor’ ’Folding’ .

Code::Blocks provides the folding feature also for preprocessor directives. To enable this feature select ’Fold preprocessor commands’ via the menu ’Settings’ ’Editor’ in the folding entry.

Another possibility is to set user defined folding points. The start of folding point is entered as comment with a opening bracket and the end is market with a comment with a closing bracket.

//{
code with user defined folding
//}

1.11.11  Auto complete

If you open a project in Code::Blocks the ’Search directories’ of your compiler and the project, the sources and headers of your project are parsed. In addition the keywords of the corresponding lexer file are parsed. The parse information is used for the auto complete feature in Code::Blocks. Please check the settings for the editor if this feature is enabled. The auto completion is accessible with the shortcut Ctrl-Space. Via the menu ’Settings’ ’Editor’ ’Syntax highlighting’ you can add user defined keywords to your lexer.

1.11.12  Find broken files

If a file is removed from disk, but is still included in the project file <project>.cbp, then this ’broken file’ will be shown a broken symbol in the project view. You should use the menu ’Remove file from project’ instead of deleting files.

In large projects with a lot of subdirectories the search for broken files can be time consuming. Code::Blocks offers with the plug-in ThreadSearch (see section 2.18 on page 202) a simple solution for this problem. If you enter a search expression in ThreadSearch and select the option ’Project files’ or ’Workspace files’ , then ThreadSearch will parse all files that are included in a project or workspace. If a broken file is found ThreadSerch will issue an error with the missing file.

1.11.13  Including libraries

In the build options of a project, you can add the used libraries via the ’Add’ button in the ’Link libraries’ entry of the ’Linker Settings’. In doing so, you can either use the absolute path to the library or just give the name without the lib prefix and file extension.

Example

For a library called <path>\libs\lib<name>.a, just write <name>. The linker with the corresponding search paths will then include the libraries correctly.

Note:
Another way to include libraries is documented in section 2.12 on page 144.

1.11.14  Object linking order

During compiling, objects name.o are created from the sources name.c/cpp. The linker then binds the individual objects into an application name.exe or for the embedded systems name.elf. In some cases, it might be desirable to predefine the order in which the objects will be linked. In Code::Blocks, this can be achieved by assigning priorities. In the context menu ’Properties’ , you can define the priorities of a file in the Build tab. A low priority will cause the file to be linked earlier.

1.11.15  Autosave

Code::Blocks offers ways of automatically storing projects and source files, or of creating backup copies. This feature can be activated in the menu ’Settings’ ’Environment’ ’Autosave’ . In doing so, ’Save to .save file’ should be specified as the method for creating the backup copy.

1.11.16  Settings for file extensions

In Code::Blocks, you can choose between several ways of treating file extensions. The settings dialogue can be opened via ’Settings’ ’Files extension handling’ . You can either use the applications assigned by Windows for each file extension (open it with the associated application), or change the setting for each extensions in such a way that either a user-defined program will start (launch an external program), or the file will be opened in the Code::Blocks editor (open it inside Code::Blocks editor).

Note:
If a user-defined program is assigned to a certain file extension, the setting ’Disable Code::Blocks while the external program is running’ should be deactivated because otherwise Code::Blocks will be closed whenever a file with this extension is opened.

1.12  Code::Blocks at the command line

IDE Code::Blocks can be executed from the command line without a graphic interface. In such a case, there are several switches available for controlling the build process of a project. Since Code::Blocks is thus scriptable, the creation of executables can be integrated into your own work processes.

codeblocks.exe /na /nd --no-splash-screen --build <name>.cbp --target=Release

<filename>

Specifies the project *.cbp filename or workspace *.workspace filename. For instance, <filename> may be project.cbp. Place this argument at the end of the command line, just before the output redirection if there is any.

--file=<filename>[:line]

Open file in Code::Blocks and optionally jump to a specific line.

/h, --help

Shows a help message regarding the command line arguments.

/na, --no-check-associations

Don’t perform any file association checks (Windows only).

/nd, --no-dde

Don’t start a DDE server (Windows only).

/ni, --no-ipc

Don’t start an IPC server (Linux and Mac only).

/ns, --no-splash-screen

Hides the splash screen while the application is loading.

/d, --debug-log

Display the debug log of the application.

--prefix=<str>

Sets the shared data directory prefix.

/p, --personality=<str>, --profile=<str>

Sets the personality to use. You can use ask as the parameter to list all available personalities.

--rebuild

Clean and build the project or workspace.

--build

Build the project or workspace.

--target=<str>

Sets target for batch build. For example --target=’Release’.

--no-batch-window-close

Keeps the batch log window visible after the batch build is completed.

--batch-build-notify

Shows a message after the batch build is completed.

--safe-mode

All plugins are disabled on startup.

> <build log file>

Placed in the very last position of the command line, this may be used to redirect standard output to log file. This is not a codeblock option as such, but just a standard DOS/*nix shell output redirection.

1.13  Shortcuts

This section describes the shortcuts which are or can be used in Code::Blocks.

1.13.1  Introduction

This plugin can be used to bind one or more key shortcuts to the menu items.

Even if an IDE such as Code::Blocks is mainly handled by mouse, keyboard shortcuts are nevertheless a very helpful way of speeding up and simplifying work processes. In the below table, we have collected some of the available keyboard shortcuts.

1.13.2  Features

Includes a configuation panel and a complete system to view/remove/add/edit command shortcuts.
Supports multiple key profiles and a complete load/save system is present.
Allows the users to customize any menu command they want, and define multiple key-shortcut to each command.

1.13.3  Usage

The configuration page for the plugin can be accessed through the ’Settings’ ’Editor’ menu, and then selecting the Keyboard Shortcuts section.


PIC

Figure 1.14: Keyboard shortcuts configuration dialog

Selecting a command, from the Commands tree, shows the current shortcuts for that command on the right. In the picture Open... is selected and the default shortcut Ctrl-O is displayed.

To add a new shortcut, for the selected command, follow these steps:

  1. Place the focus on the text box under New Shortcut and press the keys, for example F3 or Ctrl-A.
  2. Check Currently assigned to, if another command already has that shortcut you will see it’s name there. If the text says None it’s safe.
  3. Press Add to add the shortcut on the list.
  4. Press OK on the dialog to save changes and return to editor.

1.13.4  Editor

This is a list of shortcuts provided by the Code::Blocks editor component. These shortcuts cannot be rebound.

1.13.5  Files

This is a list of shortcuts provided by the Code::Blocks editor component. These shortcuts cannot be rebound.

1.13.6  View

1.13.7  Search

1.13.8  Build

1.13.9  Debug

2  Plugins

Most of the plugins described ib this chapter are also in the Wiki. Texts and figures have been copied from the Wiki but adapted to be included in Latex documents (Miktex 2.9).

2.1  General

Code::Blocks’ features can be extend by using plugins. There are generally three types of plugins:

Core plugins:
developed and maintained by the core C::B team.
Contrib plugins:
developed and maintained by the community and proven to be very valuable. So they are integrated into the C::B SVN.
3rd party plugins:
developed and maintained by the community but not (yet?) in the C::B repository. Theses plugins often have their own repository or are being posted (including the source code) in the forums.

If you are looking for plugins:

  1. Look in the official release. Notice that the installer / package manager might require you to enable some of the plugins specifically. So READ carefully.
  2. Search the forums for announcements, especially the forums at http://forums.codeblocks.org/index.php/board,14.0.html.
  3. There might be information on the Wiki concerning other plugins on this page and here : http://Wiki.codeblocks.org/index.php/Announcement_for_plugins/patches.

For Windows users, the default behavior of the current installer does not install contrib plugins. You need to manually check the ”contrib plugin” checkbox when asked for selected components to install. There is no way to install them manually afterwards.

If you are developing plugins: Surely you can work with plugin as you like, but here are some suggestions:

  Announce them in the plugin development board in the forums - including the (initial) source code.

OR

  Setup your own webpage (or use a file sharing platform) and post the link to the sources/binaries/svn access in the plugin development board in the forums.

OR

  Setup a repository, probably at BerliOS or SourceForge, post the link to the sources/binaries/svn access in the plugin development board in the forums. Notice: This is very convenient as attachments in our forum might be deleted from time to time. So it is not safe to post source code in the forums.

THEN

  Enter the plugins description on this page.

  Announce the plugin using this template on http://Wiki.codeblocks.org/index.php/Template_for_plugin_announcement

2.2  Astyle

Artistic Style is a source code indenter, source code formatter, and source code beautifier for the C, C++, C# programming languages. It can be used to select different styles of coding rules within Code::Blocks.


PIC

Figure 2.1: Formating your source code

When indenting source code, we as programmers have a tendency to use both spaces and tab characters to create the wanted indentation. Moreover, some editors by default insert spaces instead of tabs when pressing the tab key, and other editors have the ability to prettify lines by automatically setting up the white space before the code on the line, possibly inserting spaces in a code that up to now used only tabs for indentation.

Since the number of space characters shown on screen for each tab character in the source code changes between editors, one of the standard problems programmers are facing when moving from one editor to another is that code containing both spaces and tabs that was up to now perfectly indented, suddenly becomes a mess to look at when changing to another editor. Even if you as a programmer take care to ONLY use spaces or tabs, looking at other people’s source code can still be problematic.

To address this problem, Artistic Style was created - a filter written in C++ that automatically re-indents and re-formats C / C++ / C# source files.

Note:
When copying code, for example from the internet or a manual, this code will automatically be adapted to the coding rules in Code::Blocks.

2.3  AutoVersioning

An application versioning plug in that increments the version and build number of your application every time a change has been made and stores it in version.h with easy to use variable declarations. Also have a feature for committing changes a la SVN style, a version scheme editor, a change log generator and more

2.3.1  Introduction

The idea of the AutoVersioning plugin was made during the development of a pre-alpha software that required the version info and status. Been to busy coding, without time to maintain the version number, just decided to develop a plugin that could do the job with little intervention as possible.

2.3.2  Features

Here is the list of features the plugin covers summarized:

2.3.3  Usage

Just go to ’Project’ ’Autoversioning’ menu. A pop up window like this will appear:


PIC

Figure 2.2: Configure project for Autoversioning

When hitting yes on the ask to configure message box, the main auto versioning configuration dialog will open, to let you configure the version info of your project.

After configuring your project for auto versioning, the settings that you entered on the configuration dialog will be stored on the project file, and a version.h file will be created. For now, every time that you hit the ’Project’ ’Autoversioning’ menu the configuration dialog will popup to let you edit your project version and versioning related settings, unless you don’t save the new changes made by the plugin to the project file.

2.3.4  Dialog notebook tabs

Version Values

Here you just enter the corresponding version values or let the auto versioning plugin increment them for you (see Figure 2.3 on page 93).

Major
Increments by 1 when the minor version reaches its maximum
Minor
Increments by 1 when the build number pass the barrier of build times, the value is reset to 0 when it reach its maximum value.
Build Number
(also equivalent to Release) - Increments by 1 every time that the revision number is incremented.
Revision
Increments randomly when the project has been modified and then compiled.

PIC

Figure 2.3: Set Version Values

Status

Some fields to keep track of your software status with a list of predefined values for convenience(see Figure 2.4 on page 96).

Software Status
The typical example should be v1.0 Alpha
Abbreviation
Same as software status but like this: v1.0a

PIC

Figure 2.4: Set Status of Autoversioning

Scheme

Lets you edit how the plugin will increment the version values (see Figure 2.5 on page 99).


PIC

Figure 2.5: Scheme of autoversioning

Minor maximum
The maximum number that the Minor value can reach, after this value is reached the Major is incremented by 1 and next time project is compiled the Minor is set to 0.
Build Number maximum
When the value is reached, the next time the project is compiled is set to 0. Put a 0 for unlimited.
Revision maximum
Same as Build Number maximum. Put a 0 for unlimited
Revision random maximum
The revision increments by random numbers that you decide, if you put here 1, the revision obviously will increment by 1.
Build times before incrementing Minor
After successful changes to code and compilation the build history will increment, and when it reaches this value the Minor will increment.

Settings

Here you can set some settings of the auto versioning behavior (see Figure 2.6 on page 102).


PIC

Figure 2.6: Settings of Autoversioning

Autoincrement Major and Minor
Lets the plugin increments this values by you using the scheme. If not marked only the Build Number and Revision will increment.
Create date declarations
Create entries in the version.h file with dates and ubuntu style version.
Do Auto Increment
This tells the plugin to automatically increment the changes when a modification is made, this incrementation will occur before compilation.
Header language
Select the language output of version.h
Ask to increment
If marked, Do Auto Increment, it ask you before compilation (if changes has been made) to increment the version values.
svn enabled
Search for the svn revision and date in the current folder and generates the correct entries in version.h

Changes Log

This lets you enter every change made to the project to generate a ChangesLog.txt file (see Figure 2.7 on page 105).


PIC

Figure 2.7: Changelog of Autoversioning

Show changes editor when incrementing version
Will pop up the changes log editor when incrementing the version.
Title Format
A format able title with a list of predefined values.

2.3.5  Including in your code

To use the variables generated by the plugin just #include <version.h>. An example code would be like the following:

#include <iostream> 
#include "version.h" 
 
void main(){ 
   std::cout<<AutoVersion::Major<<endl; 
}

Output of version.h

The generated header file. Here is a sample content of the file on c++ mode:

#ifndef VERSION_H 
#define VERSION_H 
 
namespace AutoVersion{ 
 
   //Date Version Types 
   static const char DATE[] = "15"; 
   static const char MONTH[] = "09"; 
   static const char YEAR[] = "2007"; 
   static const double UBUNTU_VERSION_STYLE = 7.09; 
 
   //Software Status 
   static const char STATUS[] = "Pre-alpha"; 
   static const char STATUS_SHORT[] = "pa"; 
 
   //Standard Version Type 
   static const long MAJOR = 0; 
   static const long MINOR = 10; 
   static const long BUILD = 1086; 
   static const long REVISION = 6349; 
 
   //Miscellaneous Version Types 
   static const long BUILDS_COUNT = 1984; 
   #define RC_FILEVERSION 0,10,1086,6349 
   #define RC_FILEVERSION_STRING "0, 10, 1086, 6349\0" 
   static const char FULLVERSION_STRING[] = "0.10.1086.6349"; 
 
} 
#endif //VERSION_h

On C mode is the same as C++ but without the namespace:

#ifndef VERSION_H 
#define VERSION_H 
 
   //Date Version Types 
   static const char DATE[] = "15"; 
   static const char MONTH[] = "09"; 
   static const char YEAR[] = "2007"; 
   static const double UBUNTU_VERSION_STYLE = 7.09; 
 
   //Software Status 
   static const char STATUS[] = "Pre-alpha"; 
   static const char STATUS_SHORT[] = "pa"; 
 
   //Standard Version Type 
   static const long MAJOR = 0; 
   static const long MINOR = 10; 
   static const long BUILD = 1086; 
   static const long REVISION = 6349; 
 
   //Miscellaneous Version Types 
   static const long BUILDS_COUNT = 1984; 
   #define RC_FILEVERSION 0,10,1086,6349 
   #define RC_FILEVERSION_STRING "0, 10, 1086, 6349\0" 
   static const char FULLVERSION_STRING[] = "0.10.1086.6349"; 
 
#endif //VERSION_h

2.3.6  Change log generator

This dialog is accessible from the menu ’Project’ ’Changes Log’ . Also if checked Show changes editor when incrementing version on the changes log settings, the window will open to let you enter the list of changes after a modification to the project sources or an incrementation event (see Figure 2.8 on page 109).


PIC

Figure 2.8: Changes for a project

Buttons Summary

Add
Appends a row in to the data grid
Edit
Enables the modification of the selected cell
Delete
Removes the current row from the data grid
Save
Stores into a temporary file (changes.tmp) the actual data for later processing into the changes log file
Write
Process the data grid data to the changes log file
Cancel
Just closes the dialog without taking any action

Here is an example of the output generated by the plugin to the ChangesLog.txt file:

03 September 2007 
   released version 0.7.34 of AutoVersioning-Linux 
 
    Change log: 
       -Fixed: pointer declaration 
       -Bug: blah blah 
 
02 September 2007 
   released version 0.7.32 of AutoVersioning-Linux 
 
    Change log: 
       -Documented some areas of the code 
       -Reorganized the code for readability 
 
01 September 2007 
   released version 0.7.30 of AutoVersioning-Linux 
 
    Change log: 
       -Edited the change log window 
       -If the change log windows is leave blank no changes.txt is modified

2.4  Browse Tracker

Browse Tracker is a plug-in that helps navigating between recently opened files in Code::Blocks. The list of recent files is saved in a history. With the menu ’View’ ’Browse Tracker’ ’Clear All’ the history is cleared.

With the window ’Browsed Tabs’ you can navigate between the items of the recently opened files using the menu entry ’View’ ’Browse Tracker’ ’Backward Ed/Forward Ed’ or the shortcut Alt-Left/Alt-Right. The Browse Tracker menu is also accessible as context menu. The markers are saved in the layout file <projectName>.bmarks

A common procedure when developing software is to struggle with a set of functions which are implemented in different files. The BrowseTracks plug-in will help you solve this problem by showing you the order in which the files were selected. You can then comfortably navigate the function calls.

The plug-in allows even browse markers within each file in the Code::Blocks editor. The cursor position is memorized for every file. You can set this markers using the menu item ’View’ ’ Browse Tracker’ ’ Set BrowseMarks’ or with selecting a line with the left mouse button. A marker with is shown in the left margin. With the menu ’View’ ’Browse Tracker’ ’Prev Mark/Next Mark’ or the shortcut Alt-up/Alt-down you can navigate through the markers within a file. If you want to navigate in a file between markers sorted by line numbers then just select the menu ’View’ ’Browse Tracker’ ’Sort BrowseMark’ .

With the ’Clear BrowseMark’ the marker in a selected line is removed. If a marker is set for a line, holding left-mouse button down for 1/4 second while pressing the Ctrl key will delete the marker for this line. Via the menu ’Clear All BrowseMarks’ or with a Ctrl-left click on any unmarked line will reset the markers within a file.

The settings of the plug-in can be configure via the menu ’Settings’ ’Editor’ ’Browse Tracker’ .

Mark Style
Browse Marks are displayed per default as within the margin. With the setting ’Book_Marks’ they will be displayed like Bookmarks as blue arrow in the margin. With hide the display of Browse Marks is suppressed.
Toggle Browse Mark key
Markers can be set or removed either by a click with the left mouse button or with a click while holding the crtl key.
Toggle Delay
The duration of holding the left mouse button to enter the Browse Marker mode.
Clear All BrowseMarks
while holding Ctrl key either by a simple or a double click with the left mouse button.

The configuration of the plug-in is stored in your application data directory in the file default.conf. If you use the personality feature of Code::Blocks the configuration is read from the file <personality>.conf.

2.5  CodeSnippets

The CodeSnippets plug-in makes it possible to structure text modules and links to files according to categories in a tree view. The modules are used for storing often used files and constructs in text modules and managing them in a central place. Imagine the following situation: A number of frequently used source files are stored in different directories of the file system. The CodeSnippets window provides the opportunity to create categories, and below the categories, links to the required files. With these features, you can control the access to the files independently from where they are stored within the file system, and you can navigate quickly between the files without the need to search the whole system.

Note:
You can use Code::Blocks variables or environment variables in file links e.g. $(VARNAME)/name.pdf to parametrise a link in the CodeSnippets browser.

The list of text modules and links can be stored in the CodeSnippets window by right-clicking and selecting ’Save Index’ from the context menu. The file codesnippets.xml which will be created by this procedure, can then be found in the codeblocks subdirectory of your Documents and Settings\Application data directory. Under Linux, this information is stored in the .codeblocks subdirectory of your HOME directory. The Code::Blocks configuration files will be loaded during the next start-up. If you wish to save the content of CodeSnippets at a different location, select the ’Save Index As’ entry. To load this file, select ’Load Index File’ during the next start-up of Code::Blocks or include the directory in the ’Settings’ context menu under ’Snippet Folder’. The settings are saved in the corresponding file codesnippets.ini in your application data.

For including a category, use the ’Add SubCategory’ menu. A category can contain Snippets (text modules) or File Links. A text module is created via the ’Add Snippet’ command in the context menu. The content is integrated into the text module as ’New snippet’ by selecting the text passage in the Code::Blocks editor and dragging and dropping it onto the module and the properties dialog pops up. Double-clicking the newly included entry or selecting ’Edit Text’ will open an editor for the content.


PIC

Figure 2.9: Editing a text module

Output of a text module is handled in Code::Blocks via the context menu command ’Apply’ or by dragging and dropping into the editor. Under Windows, the contents of a Snippet can also be dragged and dropped into other applications. In the CodeSnippets Browser you can copy a selected item with drag and drop to a different category.

Beyond this, text modules can be parametrised by <name> variables which can be accessed via $(name) (see Figure 2.9 on page 114). The values of the variables can be retrieved in an entry field if the text module is called via the context menu command ’Apply’.

Besides the text modules, links to files can also be created. If, after having created a text module, you click the context menu command ’Properties’, then you can select the link target by clicking the ’Link target’ button. This procedure will automatically convert the text module into a link to a file. In CodeSnippets, all text modules will be marked by a T symbol, links to a file by an F symbol and urls by an U symbol. If you want to open a selected file (link) in the codesnippets view just select the context menu ’Open File’ or hold the ’Alt’ key and make a double click on the file.

Note:
You can add even url (e.g. http://www.codeblocks.org) in text modules. The url can be opened using the context menu ’Open Url’ or using drag and drop to your favorite web browser.

With this setting, if open a link to a pdf file from the codesnippets view a pdf viewer will be started automatically. This method makes it possible for a user to access files which are spread over the whole network, such as cad data, layouts, documentations etc., with the common applications, simply via the link. The content of the codesnippets is stored in the file codesnippets.xml, the configuration is stored in the file codesnippets.ini in your application data directory. This ini file will, for example, contain the path of the file codesnippets.xml.

Code::Blocks supports the usage of different profiles. These profiles are called personalities. Starting Code::Blocks with the command line option --personality=<profile> will create a new or use an existing profile. Then the settings will not be stored in the file default.conf, but in <personality>.conf in your application data directory instead. The Codesnippets plugin will then store its settings in a specific file named <personality>.codesnippets.ini. Now, if you load a new content <name.xml> in the Codesnippets settings via ’Load Index File’, this content will be stored in the corresponding ini file. The advantage of this method lies in the fact that in case of different profiles, different configurations for text modules and links can be managed.

The plug-in offers an additional search function for navigating between the categories and Snippets. The scope for searching Snippets, categories or Snippets and categories can be adjusted. By entering the required search expression, the corresponding entry is automatically selected in the view. Figure 2.10 on page 118 shows a typical display in the CodeSnippets window.


PIC

Figure 2.10: CodeSnippets View

Note:
When using voluminous text modules, the content of these modules should be saved in files via ’Convert to File Link’ in order to reduce memory usage within the system. If you delete a codesnippet or file link it will be moved to the category .trash; if you hold the Shift key the item will be deleted.

2.6  Doxyblocks

DoxyBlocks is a plugin for Code::Blocks that integrates doxygen into the IDE. It allows you to create documentation, insert comment blocks and run HTML or CHM documents. It also provides configuration of some of the more commonly used settings and access to doxywizard for more detailed configuration.

The settings in the DoxyBlocks toolbar have the following meanings:

PIC
Run doxywizard. Ctrl-Alt-D
PIC
Extract documentation for the current project. Ctrl-Alt-E
PIC
Insert a comment block at the current line. Additionally, DoxyBlocks will try to intelligently read if a method exists on the line for which a comment is being added. Ctrl-Alt-B
/** \brief 
 * 
 * \param bar bool 
 * \return void 
 * 
 */ 
void MyClass::Foo(bool bar) 
{ 
   fooBar(bar); 
}
PIC
Insert a line comment at the current cursor position. Ctrl-Alt-L
void MyClass::Foo(bool bar) 
{ 
   fooBar(bar); /**<  */ 
}
PIC
View generated HTML documentation. Ctrl-Alt-H
PIC
View generated HTML Help documentation. Ctrl-Alt-C
PIC
Open DoxyBlocks’ preferences. Ctrl-Alt-P

Doxyblocks works only if doxygen is installed on your system. You need at least the executables doxygen and doxywizard (available in official doxygen distribution at http://www.doxygen.nl/). Optionally you can have the executable ”dot” from the graphviz package (see https://graphviz.gitlab.io/. On Windows, the help compiler (hhc) may be used to generate chm files.

Notes

In the preferences you have a check box to allow or not allow DoxyBlocks to overwrite the doxyfile. By default, if a doxyfile already exists it is not overwritten to protect any changes that have been made outside DoxyBlocks however this behaviour prevents changes made within DoxyBlocks being written to an existing doxyfile.
If a text field is blank in ”Preferences”, DoxyBlocks will assume that the corresponding executable is available somewhere in your environment’s path. You can use macros such as $(CODEBLOCKS) in your path and they will be expanded automatically.
OUTPUT_DIRECTORY
Used to specify the (relative or absolute) base path where the generated documentation will be put. If a relative path is entered, it will be relative to the location where doxygen was started. If left blank the current directory will be used. DoxyBlocks will use the path name entered here to create a directory relative to <project dir>. This allows you to create separate doxygen directories for projects that reside in the same directory, or just use a different directory name. If this field is blank, documents will be created in <project dir>/doxygen. Enter directory names without dots, leading separators, volume names, etc. DoxyBlocks does validation on the path name and will strip extraneous characters.
Examples: 
[blank]          -> <project dir>/doxygen. 
"docs"           -> <project dir>/docs. 
"docs/sub1/sub2"  -> <project dir>/docs/sub1/sub2. 
"doxygen/docs"    -> <project dir>/doxygen/docs.
OUTPUT_LANGUAGE
Used to specify the language in which all documentation generated by doxygen is written. Doxygen will use this information to generate all constant output in the proper language. The default language is English. Other languages are supported.
More information in doxygen help files

2.7  Editor Tweaks plugin

The EditorTweaks plugin contains several different features. On a per-file basis, it controls :

From the merge with the Aligner plugin, it has the ability to make sections of code more readable by aligning a specific character. For example, aligning the ”=” in

int var = 1; 
int longVarName = 2; 
int foobar = 3;

will result in:

int var        = 1; 
int longVarName = 2; 
int foobar     = 3;

2.8  FileManager and PowerShell Plugin

The File Explorer Figure 2.11 on page 124 is included in the FileManager plugin, and can be found in the ’Files’ tab. The composition of the File Explorer is shown in Figure 2.11 on page 124.

On top you will find a field for entering the path. By clicking the button at the end of this field, the drop-down field will list a history of the past entries which can be navigated via a scroll bar. The up arrow key on the right-hand side of the field moves up the directory structure one directory.

In the ’Wildcard’ field you can enter a filter term for the file display. Leaving the field empty or entering * results in all files being displayed. Entering *.c;*.h, for example will result in solely C sources and header files being displayed. Opening the pull-down field will, again, list a history of the last entries.


PIC

Figure 2.11: The file manager

Pressing the Shift key and clicking selects a group of files or directories, pressing the Ctrl key and clicking selects multiple separate files or directories.

The following operations can be started via the context menu if one or multiple directories are selected in the File Explorer:

Make Root
defines the current directory as the root directory.
Add to Favorites
sets a marker for the directory and stores it as a favourite. This function allows you to navigate quickly between frequently used directories, also on different network drives.
New File
creates a new file in the selected directory.
New Directory
creates a new subdirectory in the selected directory.

The following operations can be started via the context menu if one or multiple files or directories are selected in the File Explorer:

Duplicate
copies a file/directory and renames it.
Copy To
opens a dialogue for entering the target directory in which the copied file/directory is to be stored.
Move To
moves the selection to the target location.
Delete
deletes the selected files/directories.
Show Hidden Files
activates/deactivates the display of hidden system files. When activated, this menu entry is checkmarked.
Refresh
update the display of the directory tree.

The following operations can be started via the context menu if one or multiple files are selected in the File Explorer:

Open in CB Editor
opens the selected file in the Code::Blocks editor.
Rename
renames the selected file.
Add to active project
adds the file(s) to the active project.

Note:
The files/directories selected in the File Explorer can be accessed in the PowerShell plugin via the mpaths variable.

User-defined functions can be specified via the menu command ’Settings’ ’Environment’ ’PowerShell’ . In the PowerShell mask, a new function which can be named at random, is created via the ’New’ button. In the ’ShellCommand Executable’ field, the executable program is stated, and in the field at the bottom of the window, additional parameters can be passed to the program. By clicking the function in the context menu or the PowerShell menu, the function is started and will then process the selected files/directories. The output is redirected to a separate shell window.

For example a menu entry in ’PowerShell’ ’SVN’ and in the context menu is created for ’SVN’. $file in this context means the file selected in the File Explorer, $mpath the selected files or directories (see section 3.2 on page 225).

 Add;$interpreter add $mpaths;;;

This and every subsequent command will create a submenu, in this case called ’Extensions’ ’SVN’ ’Add’ . The context menu is extended accordingly. Clicking the command in the context menu will make the SVN command add process the selected files/directories.

TortoiseSVN is a widespread SVN program with integration in the explorer. The program TortoiseProc.exe of TortoiseSVN can be started in the command line and dispalys a dialogue to collect user input. So you can perform the commands, that are available as context menu in the explorer also in the command line. Therefore you can integrate it also a shell extension in Code::Blocks. For example the command

TortoiseProc.exe /command:diff /path:$file

will diff a selected file in the Code::Blocks file explorer with the SVN base. See Figure 2.12 on page 129 how to integrate this command.

Note:
For files that are under SVN control the file explorer shows overlay icons if they are actived via menu ’View’ ’SVN Decorators’ .

PIC

Figure 2.12: Add a shell extension to the context menu

Example

You can use the file explorer to diff files or directories. Follow these steps:

  1. Add the name via menu ’Settings’ ’Environment’ ’PowerShell’ . This is shown as entry in the interpreter menu and the context menu.
  2. Select the absolute path of Diff executable (e.g. kdiff3). The program is accessed with the variable $interpreter.
  3. Add parameters of the interpreter
    Diff;$interpreter $mpaths;;;

This command will be executed using the selected files or directories as parameter. The selection is accessed via the variable $mpaths. This is an easy way to diff files or directories.

Note:
The plug-in supports the use of Code::Blocks variables within the shell extension.
$interpreter

Call this executable.

$fname

Name of the file without extension.

$fext

Extension of the selected file.

$file

Name of the file.

$relfile

Name of the file without path info.

$dir

Name of the selected directory.

$reldir

Name of directory without path info.

$path

Absolute path.

$relpath

Relative path of file or directory.

$mpaths

List of current selected files or directories.

$inputstr{<msg>}

String that is entered in a message window.

$parentdir

Parent directory (../).

Note:
The entries of shell extension are also available as context menu in the Code::Blocks editor.

2.9  HexEditor

How a file can be opened in HexEditor within Code::Blocks.

  1. ’File’ ’ Open with HexEditor’
  2. Project Navigator context menu (’Open with’ ’ Hex editor’
  3. Select the Tab Files in the Management Panel. By selecting a file in the FileManager and executing the context menu ’Open With Hex editor’ this file is opened in HexEditor.

Division of windows:

left is HexEditor view and right is the display as strings

Upper line: Current position (value in decimal/hex) and percentage (ratio of current cursor position to whole file).

Buttons:

Search functions

Goto Button: Jump to an absolute position. Format in decimal or hex. Relative jump forward or backward by specifying a sign.

Search: Search for hex patterns in the HexEditor view or for strings in the file preview.

Configuration of the number of columns: Exactly, Multiple of, Power of

Display mode: Hex, Binary

Bytes: Select how many bytes should be displayed per column.

Choice of Endianess: BE: Big Endian LE: Little Endian

Value Preview: Adds an additional view in HexEditor. For a selected value in HexEditor, the value is also displayed as Word, Dword, Float, Double.

Expression Input: Allows you to perform an arithmetic operation on a value in HexEditor. Result of the operation is displayed at the right margin.

Calc: Expression Tester

Editing a file in the HexEditor:

Includes Undo and Redo History.

For example, move the cursor to the string view: Insert spaces with the Insert key. Delete characters by pressing the Del key.

By entering a text, the existing content is overwritten as a string.

By entering numbers in the HexEditor view the values are overwritten and the preview is updated.

2.10  Incremental Search

For an efficient search in open files, Code::Blocks provides the so-called Incremental Search. This search method is initiated for an open file via the menu ’Search’ ’Incremental Search’ or by the keyboard shortcut Ctrl-I. The focus is then automatically set to the search mask of the corresponding toolbar. As soon as you begin entering the search term, the background of the search mask will be adjusted in accordance with the occurrence of the term. If a hit is found in the active editor, the respective position in the text is marked in colour. By default the current hit will be highlighted in green. This setting can be changed via ’Settings’ ’ Editor’ ’ Incremental Search’ (see Figure 2.13 on page 135). Pressing the Return key induces the search to proceed to the next occurrence of the search string within the file. With Shift-Return the previous occurrence can be selected. This functionality is not supported by Scintilla if the incremental search uses regular expressions.

PIC

If the search string cannot be found within the active file, this fact is highlighted by the background of the search mask being displayed in red.


PIC

Figure 2.13: Settings for Incremental Search

ESC
Leave the Incremental Search modus.
ALT-DELETE
Clear the input of the incremental search field.

The icons in the Incremental Search toolbar have the following meanings:

PIC
Deleting the text within the search mask of the Incremental Search toolbar.
PIC,PIC
Navigating between the occurrences of a search string.
PIC
Clicking this button results in all the occurrences of the search string within the editor being highlighted in colour, instead of only the initial occurrence.
PIC
Activating this option restricts the search to the text passage marked within the editor.
PIC
This option means a case sensitive search is performed.
PIC
Regular expression can be used in the input field of incremental search.

Note:
The standard settings of this toolbar can be configured in ’Settings’ ’Editor’ ’Incremental Search’ .

2.11  NassiShneiderman plugin

NassiShneiderman plugin allows the creation of Nassi Shneiderman diagrams within Code::Blocks ([url :NassiShneiderman]).

2.11.1  Create a diagram

There are two possibilities to create a diagram.

  1. To create an empty diagram select the menu options ’File’ ’New’ ’Nassi Shneiderman diagram’ .
  2. The second option is to creates a diagram from C/C++ source code.

In an editor window highlight some code to create the diagram from. For example the body of a function/method from the opening braces till the closing brace. Then right click the selection and choose ’Nassi Shneiderman’ ’Create diagram’ (see Figure 2.14 on page 139).


PIC

Figure 2.14: NassiShneiderman Create

You should get a new diagram (see Figure 2.15 on page 142).


PIC

Figure 2.15: NassiShneiderman Diagram Example

The parser has some limitations:

2.11.2  Editing structograms

2.11.2.1 What to do with a diagram?

You can do a lot of things with a structogram:

  1. Store for later usage. The diagram can be stored ’File’ ’Save file’ or ’File’ ’Save file as...’ .
  2. It is possible to export to different formats ’File’ ’Export’
  3. Directly insert the code into the editor: Open or create a diagram. Back in an editor window right click and choose ’Nassi Shneiderman’ ’insert from xy’ (You get a list with all open diagrams here).
  4. Drag’n’Drop the diagram (or parts of it) to other tools. For example to OpenOffice Writer to get an image in your documentation.

If the chosen diagram has a selection, the export or code-generation takes only this part of the diagram.

2.11.2.2 Extensions

The NassiShneiderman plugin supports some extensions of Nassi-Shneiderman diagrams:

2.12  LibFinder

If you want to use some libraries in your application, you have to configure your project to use them. Such configuration process may be hard and annoying because each library can use custom options scheme. Another problem is that configuration differs on platforms which result in incompatibility between unix and windows projects.

LibFinder provides two major functionalities:

2.12.1  Searching for libraries

Searching for libraries is available under ’Plugins’ ’Library finder’ menu. It’s purpose is to detect libraries installed on your system and store the results inside LibFinder’s database (note that these results are not written into Code::Blocks project files). Searching starts with dialogue where you can provide set of directories with installed libraries. LibFinder will scan them recursively so if you’re not sure you may select some generic directories. You may even enter whole disks – in such case searching process will take more time but it may detect more libraries (see Figure 2.16 on page 147).


PIC

Figure 2.16: List of directories

When LibFinder scans for libraries, it uses special rules to detect presence of library. Each set of rules is located in xml file. Currently LibFinder can search for wxWidgets 2.6/2.8, Code::Blocks SDK and GLFW – the list will be extended in future.

Note:
To get more details on how to add library support into LibFinder, read src/plugins/contrib/lib_finder/lib_finder/readme.txt in Code::Blocks sources.

After completing the scan, LibFinder shows the results (see Figure 2.17 on page 150).


PIC

Figure 2.17: Search results

In the list you check libraries which should be stored into LibFinder’s database. Note that each library may have more than one valid configuration and settings added ealier are more likely to be used while building projects.

Below the list you can select what to do with results of previous scans:

Do not clear previous results
This option works like an update to existing results – it adds new ones and updates those which already exist. This option is not recommended.
Second option (Clear previous results for selected libraries)
will clear all results for libraries which are selected before adding new results. This is the recommended option.
Clear all previous library settings
when you select this option, LibFinder’s database will be cleared before adding new results. It’s useful when you want to clean some invalid LibFinder’s database.

Another option in this dialogue is ’Set up Global Variables’ . When you check this option, LibFinder will try automatically configure Global Variables which are also used to help dealing with libraries.

If you have pkg-config installed on your system (it’s installed automatically on most linux versions) LibFinder will also provide libraries from this tool. There is no need to perform scanning for them – they are automatically loaded when Code::Blocks starts.

2.12.2  Including libraries in projects

LibFinder adds extra tab in Project Properties ’Libraries’ – this tab shows libs used in project and libs known in LibFinder. To add library into your project, select it in right pane and click < button. To remove library from project, select it on the left pane and click > button (see Figure 2.18 on page 153).


PIC

Figure 2.18: Project configuration

You can filter libraries known to LibFinder by providing search filter. The ’Show as Tree’ checkbox allows to switch between categorized and uncategorized view.

If you want to add library which is not available in LibFinder’s database, you may use ’Unknown Library’ field. Note that you should enter library’s shortcode (which usually matches global variable name) or name of library in pkg-config. List of suggested shortcodes can be found at Global Variables. Using this option is recommended only when preparing project to be built on other machines where such library exists and is properly detected by LibFinder. You can access a global variable within Code::Blocks like:

$(#GLOBAL_VAR_NAME.include)

Checking the ’Don’t setup automatically’ option will notify LibFinder that it should not add libraries automatically while compiling this project. In such case, LibFinder can be invoked from build script. Example of such script is generated and added to project by pressing ’Add manual build script’ .

2.12.3  Using LibFinder and projects generated from wizards

Wizards will create projects that don’t use LibFinder. To integrate them with this plugin, you will have to manually update project build options. This can be easily achieved by removing all library-specific settings and adding library through ’Libraries’ tab in project properties.

Such project becomes cross-platform. As long as used libs are defined in LibFinder’s database, project’s build options will be automatically updated to match platform-specific library settings.

2.13  SpellChecker plugin

A plugin to check the spelling of strings and comments.

2.13.1  Introduction

A plugin to check the spelling of strings and comments. The spelling gets checked during typing. Additionally a thesaurus is provided. Both may be accessed on-demand by selecting the word in question, then choosing either Spelling... or Thesaurus... from the Edit menu (the operation can be bound to a hot-key via the Keyboard Shortcuts plugin). The context menu (right click the word) provides spelling suggestions.

2.13.2  Configuration

Configuration is in the menu ’Settings’ ’Editor’ . The spell check option are about half way down the list on the left.


PIC

Figure 2.19: SpellChecker Configuration

The meaning of the controls are:

Enable online spell checker
Enable or disable the spell checker.
Language
The language used for spell checking and the thesaurus is selected by choosing a dictionary. This can also be changed in the status bar.
Path settings, Dictionaries
The plugin is looking in this path for the dictionary files.
Path settings, Thesauri
The plugin is looking in this path for the files needed by the thesaurus.
Path settings, Bitmaps
(Optional) The plugin is looking in this path for the flags to show in the status bar.

Note:
You can use Macros in the above three path settings, such as $(CODEBLOCKS)/share/codeblocks/SpellChecker. See Variable expansion for more details. This is quite convenient if you use portable Code::Blocks.

2.13.3  Dictionaries

SpellChecker uses a library called hunspell. Hunspell is the spell checker of OpenOffice.org, Mozilla Firefox and other projects. Dictionaries available for those applications are compatible with this plugin.

Open Office provides a collection of dictionaries for several languages and dialects to download. The OOo 3.x extensions (*.oxt) are compressed archives which can be opened with your favourite archiver (for example 7-Zip or File Roller). Copy the .aff file and the .dic file to the directory configured in ’Path settings, Dictionaries’ (see above).

If you’re running Linux you’ve probably already got compatible dictionaries installed. Look in /usr/share/hunspell or my choice is /usr/share/myspell/dicts. The reason I like the myspell files is they already include thesaurus files which are named correctly to work with the thesaurus, and everything is all in one location. Don’t copy these files. Just point the spell checker to where the files are already located.

I understand on Windows, Firefox and Thunderbird also install compatible dictionary files. These can be found in... C:\Program Files\Mozilla Firefox\dictionaries or C:\Program Files\Mozilla Thunderbird\dictionaries. In addition, both OpenOffice.org and LibreOffice install dictionary files to
C:\Program Files\(Open/Libre)Office\share\extensions\dict-*.

The Google Chrome browser also installs dictionaries, but they are .bdic format and the Code::Blocks spell checker plugin will not work with them.

2.13.4  Thesaurus files

The files for the thesaurus are also available from OOo, like the dictionaries. Copy the thesaurus files (th_*.dat and th_*.idx) to the directory configured in ’Path settings, Thesauri’ (see above) and rename them to match the name of the dictionary but prepend ”th_” and let the extension as is.

Example: If the dictionary files (for one language) are ”en_GB.aff” and ”en_GB.dic” the files used for the thesaurus are ”th_en_GB.idx” and ”th_en_GB.dat”.

On my Linux system I found thesaurus files already installed in /usr/share/myspell/dicts and /usr/share/mythes. Again, don’t move the files. Set the spell checker to use the files from their current location.

On Windows, if either OpenOffice.org or LibreOffice is installed, they often include thesaurus files in C:\Program Files\(Open/Libre)Office\share\extensions\dict-*.

2.13.5  Bitmaps (flags)

The bitmap of the actually selected language is shown in the status bar. If no bitmap is found, the name of the language is shown. The bitmap must be a PNG image. Choose a flag from the famfamfam_flag_icons and copy it to the directory configured in ’Path settings, Bitmaps’ (see above) and rename it to match the name of the dictionary but let the extension png.

2.13.6  Styles to check

Only text with specific styles gets checked (for example only comments and strings). Styles are automatically set by Scintilla (CodeBlocks editing component).

The file OnlineSpellChecking.xml contains a list with indices of the styles to check. The indices differ for different programming-languages so the file contains a list for every programming-language. To add styles, look for the name of the programming-language and the indices in the corresponding lexer_*.xml file and add this information to the file OnlineSpellChecking.xml.

For example, to check the spelling in bash shell scripts (*.sh files), add the line:

<Language name="Bash"index="2,5,6"/>

2.14  Source Code Exporter

The necessity occurs frequently of transferring source code to other applications or to e-mails. If the text is simply copied, formatting will be lost, thus rendering the text very unclear. The Code::Blocks export function serves as a remedy for such situations. The required format for the export file can be selected via ’File’ ’Export’ . The program will then adopt the file name and target directory from the opened source file and propose these for saving the export file. The appropriate file extension in each case will be determined by the export format. The following formats are available.

html
A text-based format which can be displayed in a web browser or in word processing applications.
rtf
The Rich Text format is a text-based format which can be opened in word processing applications such as Word or OpenOffice.
odt
Open Document Text format is a standardised format which was specified by Sun and O’Reilly. This format can be processed by Word, OpenOffice and other word processing applications.
pdf
The Portable Document Format can be opened by applications such as the Acrobat Reader.

2.15  SVN Support

Note:
This extension is now obsolete. So you’ll probably no more find it in recent Code::Blocks versions.

The support of the version control system SVN is included in the Code::Blocks plugin TortoiseSVN. Via the menu ’TortoiseSVN’ ’Plugin settings’ you can configure the accessible svn commands in the tab ’Integration’ .

Menu integration
Add an entry TortoiseSVN with different settings in the menu bar.
Project manger
Activate the TortoiseSVN commands in the context menu of the project manager.
Editor
Active the TortoiseSVN commands in the context menu of the editor.

In the plugin settings you can configure which svn commands are accessible via the menu or the context menu. The tab integration provides the entry ’Edit main menu’ and ’Edit popup menu’ to configure these commands.

Note:
The File Explorer in Code::Blocks uses different icon overlays for indicating the svn status. The TortoiseSVN commands are included here in the context menu.

2.16  ToDo List

In complex software projects, where different users are involved, there is often the requirement of different tasks to be performed by different users. For this purpose, Code::Blocks offers a Todo List. This list can be opened via ’View’ ’To-Do list’ , and contains the tasks to be performed, together with their priorities, types and the responsible users. The list can be filtered for tasks, users and/or source files. A sorting by columns can be achieved by clicking the caption of the corresponding column.


PIC

Figure 2.20: Displaying the ToDo List

Note:
The To-Do list can be docked in the message console. Select the option ’Include the To-Do list in the message pane’ via the menu ’Settings’ ’Environment’ .

If the sources are opened in Code::Blocks, a Todo can be added to the list via the context menu command ’Add To-Do item’. A comment will be added in the selected line of the source code.

// TODO (user#1#): add new dialog for next release

When adding a To-Do, a dialogue box will appear where the following settings can be made (see Figure 2.21 on page 167).


PIC

Figure 2.21: Dialogue for adding a ToDo

User
User name <user> in the operating system. Tasks for other users can also be created here. In doing so, the corresponding user name has to be created by Add new. The assignment of a Todo is then made via the selection of entries listed for the User.

Note:
Note that the Users have nothing to do with the Personalities used in Code::Blocks.
Type
By default, type is set to Todo.
Priority
The importance of tasks can be expressed by priorities (1 - 9) in Code::Blocks.
Position
This setting specifies whether the comment is to be included before, after or at the exact position of the cursor.
Comment Style
A selection of formats for comments (e.g. doxygen).

2.17  Tools+

Creating a new tool is fairly simple, and can be completed in a few simple steps. First open ’Tools(+)’ ’Configure Tools...’ to access the User-defined Tools dialog.


PIC

Figure 2.22: User-defined Tools dialog

Tool Name

This is the name that will be displayed in the Tools(+) drop-down menu. It will also be displayed as the tab name for tools that redirect to the Tools output window.

Command Line

Any valid command line function and switches can be placed here. Variable substitution is also accepted. The following list contains the more useful variables (see section 3.2 on page 225 for the full list).

$relfile, $file

Respectively, the relative and absolute name of a selected file.

$reldir, $dir

Respectively, the relative and absolute name of a selected directory.

$relpath, $path

The relative and absolute name of the selected file or directory.

$mpaths

A list of selected files or directories (absolute paths only).

$fname, $fext

The name without extension and the extension without the name of the selected file.

$inputstr{prompt}

Prompts the user to enter a string of text which is substituted into the command line.

$if(condition){true clause}{false clause}

Resolves to false clause if condition is empty, 0, or false; otherwise true clause.

File Types

Wildcard expressions separated by semicolons will restrict population of the right click menu of a file, directory, or multiple paths in the Project Tree, File Explorer, or Editor Pane to the specified type(s). Leave blank to handle all file/directory types.

Working Directory

The directory from which the command is executed. Code::Blocks variables, project variables, and global variables are available. Also,

  1. If $dir is specified in the command line then $dir may be used here as well.
  2. $parentdir is available for $relfile, $file, $reldir, $dir, $relpath, $path, $fname, $fext, evaluating into the absolute path of the directory containing the item.

Tools Menu Path

Controls the placement of the command in the Tools(+) menu, giving the option of adding sub-menus (multiple levels are allowed).

Will create this structure.


PIC

Figure 2.23: Tools menu structure

The command name will be used if this entry is blank. If the first character is a period, the command will be hidden.

Context Menu Path

This controls the command’s placement in the right-click menu of the Projects and Files tabs of the Management pane. The same rules of structure with the Tools Menu Path apply here.


PIC

Figure 2.24: Context menu structure

Please note that the command will not show up in the context menu unless the Command Line contains one or more of the following: $relfile, $file, $reldir, $dir, $relpath, $path, $fname, and $fext.

Output to

This determines where the output of the command will be redirected. The purpose and function of the command will determine which is best to select. Tools Output Window

Tools that only output results command (and require no input) line generally use this setting. The program will be run invisibly and any output will be redirected to the appropriate tab of the Tools Output Window. The text [DONE] will be added upon the tool’s completion.


PIC

Figure 2.25: Tool Output window

Note:
If the Tools Output window is open when Code::Blocks is closed, it may trigger Code::Blocks to crash.

Code::Blocks Console

This will cause the program to be run through the executable cb_console_runner (the same program that is launched after Build and run). This is generally used for command line tools with more advanced user interactions, although GUI programs can also be used (especially if the program is unstable and/or also leaves messages in the standard output). Console runner will pause the window (prevent it from closing), display the run time, and the exit code when the program finishes.

Standard Shell

This is the same as placing the command in a batch or shell script, then running it. The program will run in whatever its default method is, and when it finishes, its window will close. This setting is useful for running a program (for example a file or web browser) that must remain open after Code::Blocks is closed.

Note:
As the Tools+ plugin is not yet complete, some functionality - specifically Menu Priority and Environment Variables - are not available.

2.17.1  Example Tools

Open file browser to selected file

Update Subversion directory

Export makefile

Note:
this uses the command line-tool cbp2make.

Compress active project to archive

2.18  Thread Search

Via the ’Search’ ’Thread Search’ menu, the appropriate plug-in can be shown or hidden as a tab in the Messages Console. In Code::Blocks, a preview can be displayed for the occurrence of a character string in a file, workspace or directory. In doing so, the list of search results will be displayed on the right-hand side of the ThreadSearch Console. By clicking an entry in the list, a preview is displayed on the left-hand side. By double-clicking in the list, the selected file is opened in the Code::Blocks editor.

Note:
The scope of file extensions to be included in the search, is preset and might have to be adjusted.

2.18.1  Features

ThreadSearch plugin offers the following features:


PIC

Figure 2.26: Thread Search Panel

2.18.2  Usage

  1. Configure your search preferences (see Figure 2.27 on page 209)

    Once the plugin is installed, there are 4 ways to run a search:

    1. Type/Select a word in the search combo box and press enter or click on Search on the Thread search panel of the Messages notebook.
    2. Type/Select a word in the toolbar search combo box and press enter or click on Search button.
    3. Right click on any ’word’ in active editor and click on ’Find occurrences’.
    4. Click on Search/Thread search to find the current word in active editor.

      Note:
      Items 1, 2 and 3 may not be available according to current configuration.
  2. Click again on the search button to cancel current search.
  3. A single click on a result item displays it on the preview editor at right location.
  4. A double click on a result item opens or set an editor in editors notebook at right location.

2.18.3  Configuration

To access ThreadSearch plugin configuration panel click on (see Figure 2.27 on page 209):


PIC

Figure 2.27: Configuration of Thread Search

  1. Options button on Messages notebook Thread search panel.
  2. Options button on Thread search toolbar.
  3. Settings/Environment menu item and then on the Thread search item on the left columns.

Note:
Items 1, 2 and 3 may not be available according to current configuration.

Search in part defines the set of files that will be analysed.

2.18.4  Options

Whole word
if checked, line matches search expression if search expression is found with no alpha-numeric +_ before and after.
Start word
if checked, line matches search expression if search expression is found at the beginning of a word, ie no alpha-numeric +_ before search expression.
Match case
if checked, the search is case sensitive.
Regular expression
the search expression is a regular expression.

Note:
If you want to search for regular expressions like n you will have to set the option ’Use Advanced RegEx searches’ via the menu ’Settings’ ’Editor’ ’General Settings’ .

2.18.5  Thread search options

Enable ’Find occurrences contextual menu item’
If checked, the Find occurrences of ’Focused word’ entry is added to the editor contextual menu.
Use default options when running ’Find occurrences’
If checked, a set of default options is applied to the searches launched with the ’Find occurrences’ contextual menu item. Per defaut option ’Whole word’ and ’Match case’ is enabled.
Delete previous results at search begin
If ThreadSearch is configured with ’Tree View’ then the search results will be listed hierarchically,

If you search different terms the list will become confusing, therefore previous search results can be cleared at search begin using this option.

Note:
In the list of occurences single items or all items can be deleted via the context menu ’Delete item’ or ’Delete all items’ .

2.18.6  Layout

Display header in log window
if checked, the header are displayed in the results list control.

Note:
If unchecked, the columns are no longer resizeable but space is spared.
Draw lines between columns
Draws lines between columns in list mode.
Show ThreadSearch toolbar
Display the toolbar of Thread Search plugin.
Show search widgets in ThreadSearch Messages panel
If checked, only the results list control and the preview editor are displayed. All other search widgets are hidden (spares space).
Show code preview editor
Code preview can be hidden either with this check box or with a double click on the splitter window middle border. This is where it can be shown again.

2.18.7  Panel Management

You can choose different modes how the the ThreadSearch window is managed. With the setting ’Message Notebook’ the ThreadSearch window will be a dockable window in the message panel. If you choose the setting ’Layout’ you will be able to undock the window from the message panel and put it somewhere else.

2.18.8  Logger Type

The view of the search results can be displayed in different ways. The setting ’List’ displays all occurrences as list. The other mode ’Tree’ gathers all occurrences within a file as a node.

2.18.9  Splitter Window Mode

The user can configure a horizontal or vertical splitting of the preview window and the output window of the search results.

2.18.10  Sort Search Results

The view of the search results may be sorted by path or file name.

2.19  Code statistics


PIC

Figure 2.28: Configuration for Code Statistics

Based on the entries in the configuration mask, this simple plugin detects the proportions of code, commentaries and blank lines for a project. The evaluation is called via the menu command ’Plugins’ ’Code statistics’ .

2.20  Code profiler

A simple graphical interface to the GNU GProf Profiler.

2.21  Projects Importer plugin

ProjectsImporter imports foreign projects and workspaces from Dev-C++, MSVC6, MSVC7, and MSVC8 for use as a Code::Blocks project.

2.22  Searching Available Source Code

This plugin makes it possible to select a term within the editor and to search for this term via the context menu ’Search at Koders’ in the [?] database. The dialogue offers the additional possibilities to of filtering for program languages and licences.

Note:
Koders and it’s successor BlackDuck seem to have disappeared or changed their website ! So this plugin does not work anymore. Waiting for an update ...

This database search will help you find source code originating from other world-wide projects of universities, consortiums and organisations such as Apache, Mozilla, Novell Forge, SourceForge and many others, which can be re-used without having to reinvent the wheel every time. Please observe the licence of the source code in each individual case.

2.23  Symbol Table Plugin

This plugin makes it possible to search for symbols in objects and libraries. The options and the path for the command line program nm are defined in the Options tab.


PIC

Figure 2.29: Configuring the Symbol Table

Clicking the ’Search’ stats the search, the results of the NM program are displayed in a separate window caleld ’SymTabs Result’. The name of the objects or libraries containing the symbol are listed under the title ’NM’s Output’.

3  Variable Expansion

Code::Blocks differentiates between several types of variables. These types serve the purpose of configuring the environment for creating a program, and at the same of improving the maintainability and portability. Access to the Code::Blocks variables is achieved via $<name>.

Environment Variable
are set during the startup of Code::Blocks. They can modify system environment variables such as PATH. This can be useful in cases where a defined environment is necessary for the creation of projects. The settings for environment variables in Code::Blocks are made at ’Settings’ ’Environment’ ’Environment Variables’ .
Builtin Variables
are predefined in Code::Blocks, and can be accessed via their names (see section 3.2 on page 225 for details).
Command Macros
This type of variables is used for controlling the build process. For further information please refer to section 3.4 on page 236.
Custom Variables
are user-defined variables which can be specified in the build options of a project. Here you can, for example define your derivative as a variable MCU and assign a corresponding value to it. Then set the compiler option -mcpu=$(MCU), and Code::Blocks will automatically replace the content. By this method, the settings for a project can be further parametrised.
Global Variables
are mainly used for creating Code::Blocks from the sources or developments of wxWidgets applications. These variables have a very special meaning. In contrast to all others if you setup such a variables and share your project file with others that have *not* setup this GV Code::Blocks will ask the user to setup the variable. This is a very easy way to ensure the ’other developer’ knows what to setup easily. Code::Blocks will ask for all path’s usually necessary.

3.1  Syntax

Code::Blocks treats the following functionally identical character sequences inside pre-build, post-build, or build steps as variables:

Variable names must consist of alphanumeric characters and are not case-sensitive. Variables starting with a single hash sign (#) are interpreted as global user variables (see section 3.7 on page 238 for details). The names listed below are interpreted as built-in types.

Variables which are neither global user variables nor built-in types, will be replaced with a value provided in the project file, or with an environment variable if the latter should fail.

The use of these variables can follow the following example for the date :

#include "include/manager.h"
wxString strdate = Manager::Get()->GetMacrosManager()->ReplaceMacros(_T("$TODAY"));

Note:
Per-target definitions have precedence over per-project definitions.

3.2  List of available built-ins

The variables listed here are built-in variables of Code::Blocks. They cannot be used within source files.

3.2.1  Code::Blocks workspace

$(WORKSPACE_FILENAME), $(WORKSPACE_FILE_NAME), $(WORKSPACEFILE), $(WORKSPACEFILENAME)

The filename of the current workspace project (.workspace).

$(WORKSPACENAME), $(WORKSPACE_NAME)

The name of the workspace that is displayed in tab Projects of the Management panel.

$(WORKSPACE_DIR), $(WORKSPACE_DIRECTORY), $(WORKSPACEDIR), $(WORKSPACEDIRECTORY)

The location of the workspace directory.

3.2.2  Files and directories

$(PROJECT_FILENAME), $(PROJECT_FILE_NAME), $(PROJECT_FILE), $(PROJECTFILE)

The filename of the currently compiled project.

$(PROJECT_NAME), $(PROJECTNAME)

The name of the currently compiled project.

$(PROJECT_DIR), $(PROJECTDIR), $(PROJECT_DIRECTORY), $(PROJECTDIRECTORY)

The common top-level directory of the currently compiled project.

$(ACTIVE_EDITOR_FILENAME)

The filename of the file opened in the currently active editor.

$(ACTIVE_EDITOR_LINE)

Return the current line in the active editor.

$(ACTIVE_EDITOR_COLUMN

Return the column of the current line in the active editor.

$(ACTIVE_EDITOR_DIRNAME)

the directory containing the currently active file (relative to the common top level path).

$(ACTIVE_EDITOR_STEM)

The base name (without extension) of the currently active file.

$(ACTIVE_EDITOR_EXT)

The extension of the currently active file.

$(ALL_PROJECT_FILES)

A string containing the names of all files in the current project.

$(MAKEFILE)

The filename of the makefile.

$(CODEBLOCKS), $(APP_PATH), $(APPPATH), $(APP-PATH)

The path to the currently running instance of Code::Blocks.

$(DATAPATH), $(DATA_PATH), $(DATA-PATH)

The ’shared’ directory of the currently running instance of Code::Blocks.

$(PLUGINS)

The plugins directory of the currently running instance of Code::Blocks.

$(TARGET_COMPILER_DIR)

The compiler installation directory so-called master path.

3.2.3  Build targets

replace FOOBAR with the target name

$(FOOBAR_OUTPUT_FILE)

The output file of a specific target.

$(FOOBAR_OUTPUT_DIR)

The output directory of a specific target.

$(FOOBAR_OUTPUT_BASENAME)

The output file’s base name (no path, no extension) of a specific target.

$(FOOBAR_PARAMETERS)

A specific target’s execution parameters

$(TARGET_OUTPUT_DIR)

The output directory of the current target.

$(TARGET_OBJECT_DIR)

The object directory of the current target.

$(TARGET_NAME)

The name of the current target.

$(TARGET_OUTPUT_FILE)

The output file of the current target.

$(TARGET_OUTPUT_BASENAME)

The output file’s base name (no path, no extension) of the current target.

$(TARGET_CC), $(TARGET_CPP), $(TARGET_LD), $(TARGET_LIB)

The build tool executable (compiler, linker, etc) of the current target.

$(TARGET_COMPILER_DIR)

The current target’s build tool directory (compiler, linker, etc).

3.2.4  Language and encoding

$(LANGUAGE)

The system language in plain language.

$(ENCODING)

The character encoding in plain language.

3.2.5  Time and date

$(TDAY)

Current date in the form YYYYMMDD (for example 20051228)

$(TODAY)

Current date in the form YYYY-MM-DD (for example 2005-12-28)

$(NOW)

Timestamp in the form YYYY-MM-DD-hh.mm (for example 2005-12-28-07.15)

$(NOW_L)

Timestamp in the form YYYY-MM-DD-hh.mm.ss (for example 2005-12-28-07.15.45)

$(WEEKDAY)

Plain language day of the week (for example ’Wednesday’)

$(TDAY_UTC), $(TODAY_UTC), $(NOW_UTC), $(NOW_L_UTC), $(WEEKDAY_UTC)

These are identical to the preceding types, but are expressed relative to UTC.

$(DAYCOUNT)

The number of the days passed since an arbitrarily chosen day zero (January 1, 2009). Useful as last component of a version/build number.

3.2.6  Platform dependence

$(PLATFORM)

Expands to msw on windows and unix on linux and mac (Since revision r11793)

3.2.7  Command line expansions

This commands can be used in the command line for the specific platform.

$(CMD_CP)

Will expand to a copy command for the specific platform. (on windows copy and on unix cp --preserve=timestamps)

$(CMD_RM)

delete command

$(CMD_MV)

move command

$(CMD_NULL)

NULL device (for redirecting streams)

$(CMD_MKDIR)

create a directory

$(CMD_RMDIR)

delete a directory

3.2.8  Random values

$(COIN)

This variable tosses a virtual coin (once per invocation) and returns 0 or 1.

$(RANDOM)

A 16-bit positive random number (0-65535)

3.2.9  Standard path

$(GET_DATA_DIR)

Unix: prefix/share/appname ; Windows: EXE path

$(GET_LOCAL_DATA_DIR)

Unix: /etc/appname ; Windows: EXE path

$(GET_CONFIG_DIR)

Unix: /etc ; Windows: C:\Documents and Settings\All Users\Application Data

$(GET_USER_CONFIG_DIR)

Unix:   ; Windows: C:\Documents and Settings\username\Application Data\appname

$(GET_USER_DATA_DIR)

Unix:  /.appname ; Windows: C:\Documents and Settings\username\Application Data

$(GET_USER_LOCAL_DATA_DIR)

Unix:  /.appname ; Windows: C:\Documents and Settings\username\Local Settings\Application Data\appname

$(GET_TEMP_DIR)

ALL platforms: A writable, temporary directory

3.2.10  Build in functions for path conversion

There are build in macro functions to simplify path generation

$TO_UNIX_PATH{}

convert path to unix path (use ’/’ as separator)

$TO_WINDOWS_PATH{}

convert path to windows (use ’\’ as separator)

$TO_NATIVE_PATH{}

convert to native path form the codeblocks instance is running on

Usage

$TO_UNIX_PATH{$(TARGET_OUTPUT_FILE)}

returns the current target output file as unix path

3.2.11  Conditional Evaluation

$if(condition){true clause}{false clause}

Conditional evaluation will resolve to its true clause if

Conditional evaluation will resolve to its false clause if

Note:
Please do note that neither the variable syntax variants %if(...) nor $(if)(...) are supported for this construct.

Example

For example if you are using several platforms and you want to set different parameters depending on the operating system. In the following code the script commands of [[ ]] are evaluated and the <command> will be executed. This could be useful in a post-built step (on a single line).

[[ if (PLATFORM ==  PLATFORM_MSW) { print (_T("cmd /c")); } else 
     { print (_T("sh ")); } ]] <command>

3.3  Script expansion

For maximum flexibility, you can embed scripts using the [[ ]] operator as a special case of variable expansion. Embedded scripts have access to all standard functionalities available to scripts and work pretty much like bash backticks (except for having access to Code::Blocks namespace). As such, scripts are not limited to producing text output, but can also manipulate Code::Blocks state (projects, targets, etc.).

Note:
Manipulating Code::Blocks state should be implemented rather with a pre-build script than with a script.

Example with Backticks

objdump -D find . -name *.elf > name.dis

The expression in backticks returns a list of all executables *.elf in any subdirectories. The result of this expression can be used directly by objdump. Finally the output is piped to a file named name.dis. Thus, processes can be automatted in a simple way without having to program any loops.

Example using Script

The script text is replaced by any output generated by your script, or discarded in case of a syntax error.

Since conditional evaluation runs prior to expanding scripts, conditional evaluation can be used for preprocessor functionalities. Built-in variables (and user variables) are expanded after scripts, so it is possible to reference variables in the output of a script.

[[ print(GetProjectManager().GetActiveProject().GetTitle()); ]]

inserts the title of the active project into the command line.

3.4  Command Macros

$compiler

Access to name of the compiler executable.

$linker

Access to name of the linker executable.

$options

Compiler flags

$link_options

Linker flags

$includes

Compiler include paths

$c

Linker include paths

$libs

Linker libraries

$file

Source file (full name)

$file_dir

Source file directory without file name and file name extension.

$file_name

Source file name without path info and file name extension.

$exe_dir

Directory of executable without file name and file name extension.

$exe_name

File name of executable without path and file name extension.

$exe_ext

File name extension of executable without path and file name.

$object

Object file

$exe_output

Executable output file

$objects_output_dir

Object Output Directory

3.5  Compile single file

$compiler $options $includes -c $file -o $object

3.6  Link object files to executable

$linker $libdirs -o $exe_output $link_objects $link_resobjects 
$link_options $libs

3.7  Global compiler variables

This section describes how to work with global variables.

3.7.1  Synopsis

Working as a developer on a project which relies on 3rd party libraries involves a lot of unnecessary repetitive tasks, such as setting up build variables according to the local file system layout. In the case of project files, care must be taken to avoid accidentially committing a locally modified copy. If one does not pay attention, this can happen easily for example after changing a build flag to make a release build.

The concept of global compiler variables is a unique new solution for Code::Blocks which addresses this problem. Global compiler variables allow you to set up a project once, with any number of developers using any number of different file system layouts being able to compile and develop this project. No local layout information ever needs to be changed more than once.

3.7.2  Names and Members

Global compiler variables in Code::Blocks are discriminated from per-project variables by a leading hash sign. Global compiler variables are structured; every variable consists of a name and an optional member. Names are freely definable, while some of the members are built into the IDE. Although you can choose anything for a variable name in principle, it is advisable to pick a known identifier for common packages. Thus the amount of information that the user needs to provide is minimised. The Code::Blocks team provides a list of recommended variables for known packages.

The member base resolves to the same value as the variable name uses without a member (alias).

The members include and lib are by default aliases for base/include and base/lib, respectively. However, a user can redefine them if another setup is desired.

It is generally recommended to use the syntax $(#variable.include) instead of $(#variable)/include, as it provides additional flexibility and is otherwise exactly identical in functionality (see subsection 3.7.6 on page 246 and Figure 3.1 on page 241 for details).

The members cflags and lflags are empty by default and can be used to provide the ability to feed the same consistent set of compiler/linker flags to all builds on one machine. Code::Blocks allows you to define custom variable members in addition to the built-in ones.


PIC

Figure 3.1: Global Variable Environment

3.7.3  Constraints

Code::Blocks will detect the most obvious cases of recursive definitions (which may happen by accident), but it will not perform an in-depth analysis of every possible abuse. If you enter crap, then crap is what you will get; you are warned now.

Examples

Defining wx.include as $(#wx)/include is redundant, but perfectly legal Defining wx.include as $(#wx.include) is illegal and will be detected by Code::Blocks Defining wx.include as $(#cb.lib) which again is defined as $(#wx.include) will create an infinite loop

3.7.4  Using Global Compiler Variables

All you need to do for using global compiler variables is to put them in your project! Yes, it’s that easy.

When the IDE detects the presence of an unknown global variable, it will prompt you to enter its value. The value will be saved in your settings, so you never need to enter the information twice.

If you need to modify or delete a variable at a later time, you can do so from the settings menu.

Example


PIC

Figure 3.2: Global Variables

The above image shows both per-project and global variables. WX_SUFFIX is defined in the project, but WX is a global user variable.

3.7.5  Variable Sets

Sometimes, you want to use different versions of the same library, or you develop two branches of the same program. Although it is possible to get along with a global compiler variable, this can become tedious. For such a purpose, Code::Blocks supports variable sets. A variable set is an independent collection of variables identified by a name (set names have the same constraints as variable names).

If you wish to switch to a different set of variables, you simply select a different set from the menu. Different sets are not required to have the same variables, and identical variables in different sets are not required to have the same values, or even the same custom members.

Another positive thing about sets is that if you have a dozen variables and you want to have a new set with one of these variables pointing to a different location, you are not required to re-enter all the data again. You can simply create a clone of your current set, which will then duplicate all of your variables.

Deleting a set also deletes all variables in that set (but not in another set). The default set is always present and cannot be deleted.

3.7.6  Custom Members Mini-Tutorial

As stated above, writing $(#var.include) and $(#var)/include is exactly the same thing by default. So why would you want to write something as unintuitive as $(#var.include)?

Let’s take a standard Boost installation under Windows for an example. Generally, you would expect a fictional package ACME to have its include files under ACME/include and its libraries under ACME/lib. Optionally, it might place its headers into yet another subfolder called acme. So after adding the correct paths to the compiler and linker options, you would expect to #include <acme/acme.h> and link to libacme.a (or whatever it happens to be).

Boost, however, installs headers into C:\Boost\include\boost-1_33_1\boost and its libraries under C:\Boost\lib by default. It seems impossible to get this under one hood without having to adjust everything on every new PC, especially if you have to work under Linux or some other OS, too.

This is where the true power of global user variables is unveiled. When defining the value of the #boost variable, you go one step further than usual. You define the member include as C:\Boost\include\boost-1_33_1\boost and the member lib as C:\Boost\lib, respectively. Your projects using $(#boost.include) and $(#boost.lib) will magically work on every PC without any modifications. You don’t need to know why, you don’t want to know why.

4  Working With Code::Blocks

This chapter deals with some basic knowledge to be able to work with Code::Blocks. Some paragraphs, here directly taken from Wiki, overlap but with a slightly different presentation with the content of the first chapter.

4.1  The build process of Code::Blocks

In this pages, the build process is explained in detail. What goes on behind the scenes and ”when” is covered. I hope it makes an interesting read :).

4.1.1  Build order

As you have probably guessed, Code::Blocks does not launch build commands at random but rather as a well thought out and prepared sequence. But first let’s see what components are the subject of the build:

Workspace:
contains one or more projects
Project:
contains one or more build targets. It also contains the project’s files.
Build target:
project files are assigned to it, which are built as a group and generate one binary output. This output can either be an executable, a dynamic library or a static library. NOTE: there is one kind of build target that does not produce a binary output directly but rather just runs its pre/post build steps (which may generate a binary output externally).

Let’s break these subjects in sections and explain them in more detail.

4.1.2  Workspace

The workspace is the top-level container item for organizing your projects. Since there can be only one workspace open at a time, there really is no build order issue for them. It’s only one, so it’s just built ;).

Use the menu ’Build’ ’Build workspace’ to build a workspace (i.e. all the projects contained in it).

4.1.3  Projects

Here, things start getting interesting :).

Projects build order is different depending if the user has set project dependencies or not. Please read on.

Without project dependencies

In this case, projects are built in the order of appearence, from top to bottom. Most projects though (at least not the ”hello world” ones), would want to setup project dependencies.

Using project dependencies

Project dependencies are a simple way to tell Code::Blocks that a given project ”depends” on another (in the same workspace, always).

Thus imagine in your workspace you have a library project and an executable project which depends on the library. Then you could (and should) tell Code::Blocks about this dependency. To do this, you select ’Project’ ’Properties’ and click the ”Project’s dependencies..” button.

Please notice that the dependency information is saved within the workspace file, not the project file as it is a dependency between two projects within a workspace.


PIC

Figure 4.1: Setting up project dependencies

It is very easy to use this dialog. Select the project you want to add a dependency and then put a checkmark on all projects that this project depends on. This will ensure that all the projects you checked will be built before the project that depends on them, ensuring a synchronized build.

Tip: You don’t have to close this dialog and launch the other project’s properties again to set their dependencies. You can set all projects dependencies from this same dialog. Just select a different project in the drop-down box :).

Some things to note:

4.1.4  Build targets

Build targets build order depend on a couple of things.

  1. If the user has selected a specific build target in the compiler toolbar’s drop-down box, then only this build target is built. If project dependencies are setup for the project containing this target, all projects it depends on will also build their target with the same name. If no such target exists, that project is skipped.
  2. If the virtual ”All” target is selected, then all targets in the project (and all the projects it depends on) are built in order, top to bottom. There are a couple of exceptions to this:

4.1.5  Preprocessing phase

Before the actual build process starts (i.e. the compiler/linker commands start executing), a preprocessing step runs which generates all required command-lines for the entire build process. This step caches much of the information it generates, making subsequent builds faster in effect.

This step also runs any attached build scripts.

4.1.6  Actual commands execution

This is the stage that the build process actually starts, from the user’s point of view. The files start getting compiled and finally linked to generate the various binary outputs the build targets define.

In this step, the pre-build and post-build steps are also executed.

4.1.7  Pre-build and post-build steps

These are commands that can be setup on project and/or target level. They are shell commands that e.g. copy files around or any other operation you can do with a usual OS system shell.

The variables specified in the article Variable expansion (chapter 3 on page 224) can be used in the scripts to get things like target output directory, project directory, target type and others.

Here’s a breakdown of the pre/post build steps execution order for an imaginary project with two targets (Debug/Release):

  1. Project pre-build steps
    1. Target ”Debug” pre-build steps
    2. Target ”Debug” compile files
    3. Target ”Debug” link files and generate binary output
    4. Target ”Debug” post-build steps (see notes below)
    5. Target ”Release” pre-build steps
    6. Target ”Release” compile files
    7. Target ”Release” link files and generate binary output
    8. Target ”Release” post-build steps (see notes below)
  2. Project post-build steps

I hope this is self-explaining :)

Note:
Pre-build steps are always ran. Post-build steps will run only if the project/target they ’re attached to is not up-to-date (i.e. is going to be built). You can change this by selecting ”Always execute, even if target is up to date” in the build options.

Script Samples

Post-build script that copies the output file into the folder C:\Program\bin in Windows:

cmd /c copy "$(PROJECT_DIR)$(TARGET_OUTPUT_FILE)" "C:\Program\bin"

Execute the bash script ”copyresources.sh” in Linux:

/bin/sh copyresources.sh

Create a new directory in the output directory:

mkdir $(TARGET_OUTPUT_DIR)/data

4.2  Creating a New Project

These pages are a guide to many of the beginning (and some intermediate) features of the creation and modification of a Code::Blocks project. If this is your first experience with Code::Blocks, here is a good starting point.

4.2.1  The project wizard

Launch the Project Wizard through ’File’ ’New’ ’Project...’ to start a new project. Here there are many pre-configured templates for various types of projects, including the option to create custom templates. Select Console application, as this is the most common for general purposes, and click Go.


PIC

Figure 4.2: Project Wizard

Note:
red text instead of black text below any of the icons signifies it is using a customized wizard script.

The console application wizard will appear next. Continue through the menus, selecting C++ when prompted for a language. In the next screen, give the project a name and type or select a destination folder. As seen below, Code::Blocks will generate the remaining entries from these two.


PIC

Figure 4.3: Console Application

Finally, the wizard will ask if this project should use the default compiler (normally GCC) and the two default builds: Debug and Release. All of these settings are fine. Press finish and the project will be generated. The main window will turn gray, but that is not a problem, the source file needs only to be opened. In the Projects tab of the Management pane on the left expand the folders and double click on the source file main.cpp to open it in the editor.


PIC

Figure 4.4: Select a Source

This file contains the following standard code.

main.cpp

#include <iostream> 
using namespace std; 
int main() 
{ 
   cout << "Hello world!" << endl; 
   return 0; 
}

4.2.2  Changing file composition

A single source file is of little uses in programs of any useful complexity. In order to handle this, Code::Blocks has several very simple methods of adding additional files to the project.

Adding a blank file

In this example, we will be splitting the function

main.cpp

   cout << "Hello world!" << endl;

into a separate file.

Note:
it is generally improper programming style to create a function this small; it is done here to give a simple example.

To add the new file to the project, bring up the file template wizard through either ’File’ ’New’ ’File...’ or ’Main Toolbar’ ’New file (button)’ ’File...’ Use the menu ’Build’ ’Build workspace’ to build a workspace (i.e. all the projects contained in it).


PIC

Figure 4.5: New File

Select C/C++ source and click Go. Continue through the following dialogs very much like the original project creation, selecting C++ when prompted for a language. On the final page, you will be presented with several options. The first box will determine the new filename and location (as noted, the full path is required). You may optionally use the ... button to bring up a file browser window to save the file’s location. Checking Add file to active project will store the filename in the Sources folder of the Projects tab of the Management panel. Checking any of the build targets will alert Code::Blocks that the file should be compiled and linked into the selected target(s). This can be useful if, for example, the file contains debug specific code, as it will allow the inclusion to (or exclusion from) the appropriate build target(s). In this example, however, the hello function is of key importance, and is required in each target, so select all the boxes and click Finish to generate the file.


PIC

Figure 4.6: Hello Program Configurations

The newly created file should open automatically; if it does not, open it by double clicking on its file in the Projects tab of the Management panel. Now add in code for the function main.cpp will call.

hello.cpp

#include <iostream> 
using namespace std; 
 
void hello() 
{ 
   cout << "Hello world!" << endl; 
}

Adding a pre-existing file

Now that the hello() function is in a separate file, the function must be declared for main.cpp to use it. Launch a plain text editor (for example Notepad or Gedit), and add the following code:

hello.h

#ifndef HELLO_H_INCLUDED 
#define HELLO_H_INCLUDED 
 
void hello(); 
 
#endif // HELLO_H_INCLUDED

Save this file as a header (hello.h) in the same directory as the other source files in this project. Back in Code::Blocks, click ’Project’ ’Add files...’ to open a file browser. Here you may select one or multiple files (using combinations of Ctrl and Shift). (The option ’Project’ ’Add files recursively...’ will search through all the subdirectories in the given folder, selecting the relevant files for inclusion.) Select hello.h, and click Open to bring up a dialog requesting to which build targets the file(s) should belong. For this example, select both targets.


PIC

Figure 4.7: Target Belonging

Note:
if the current project has only one build target, this dialog will be skipped.

Returning to the main source (main.cpp), include the header file and replace the cout function to match the new setup of the project.

main.cpp

#include "hello.h" 
 
int main() 
{ 
   hello(); 
   return 0; 
}

Press Ctrl-F9, ’Build’ ’Build’ , or ’Compiler Toolbar’ ’Build (button - the gear)’ to compile the project. If the following output is generated in the build log (in the bottom panel) then all steps were followed correctly.

-------------- Build: Debug in HelloWorld --------------- 
 
Compiling: main.cpp 
Compiling: hello.cpp 
Linking console executable: bin\Debug\HelloWorld.exe 
Output size is 923.25 KB 
Process terminated with status 0 (0 minutes, 0 seconds) 
0 errors, 0 warnings (0 minutes, 0 seconds)

The executable may now be run by either clicking the Run button or hitting Ctrl-F10.

Note:
the option F9 (for build and run) combines these commands, and may be more useful in some situations.

See the build process of Code::Blocks for what occurs behind the scenes during a compile.

Removing a file

Using the above steps, add a new C++ source file, useless.cpp, to the project. Removing this unneeded file from the project is straightforward. Simply right-click on useless.cpp in the Projects tab of the Management pane and select Remove file from project.


PIC

Figure 4.8: Remove a File from a Project

Note:
removing a file from a project does not physically delete it; Code::Blocks only removes it from the project management.

4.2.3  Modifying build options

Build targets have come up several times so far. Changing between the two default generated ones - Debug and Release - can simply be done through the drop-down list on the Compiler Toolbar. Each of these targets has the ability to be a different type (for example: static library; console application), contain a different set of source files, custom variables, different build flags (for example: debug symbols -p; size optimization -Os; link time optimization -flto), and several other options.


PIC

Figure 4.9: Target Selection

’Open Project’ ’Properties...’ to access the main properties of the active project, HelloWorld. Most of the settings on the first tab, Project settings, are rarely changed. Title: allows the name of the project to be changed. If Platforms: is changed to something other than its default All, Code::Blocks will only allow the project to build on the selected targets. This is useful if, for example, the source code contains Windows API, and would therefore be invalid anywhere but Windows (or any other operating system specific situations). The Makefile: options are used only if the project should use a makefile instead of Code::Blocks’ internal build system (see Code::Blocks and Makefiles [section 4.5 on page 316] for further details).

Adding a new build target

Switch to the Build targets tab. Click Add to create a new build target and name it Release Small. The highlight in the left hand column should automatically switch to the new target (if not, click on it to change the focus). As the default setting for Type: - ”GUI application” - is incorrect for the HelloWorld program, change it to ”Console application” via the drop-down list. The output filename HelloWorld.exe is fine except in that it will cause the executable to be output in the main directory. Add the path ”bin\ReleaseSmall\” (Windows) or ”bin/ReleaseSmall/” (Linux) in front of it to change the directory (it is a relative path from the root of the project). The Execution working dir: refers to where the program will be executed when Run or Build and run are selected. The default setting ”.” is fine (it refers to the project’s directory). The Objects output dir: needs to be changed to ”obj\ReleaseSmall\” (Windows) or ”obj/ReleaseSmall/” (Linux) in order to be consistent with the remainder of the project. The Build target files: currently has nothing selected. This is a problem, as nothing will be compiled if this target is built. Check all the boxes.


PIC

Figure 4.10: Target Options

The next step is to change the target’s settings. Click Build options... to access the settings. The first tab the comes up has a series of compiler flags accessible through check boxes. Select ”Strip all symbols from binary” and ”Optimize generated code for size”. The flags here contain many of the more common options, however, custom arguments may be passed. Switch to the Other options sub-tab and add the following switches:

-fno-rtti 
-fno-exceptions 
-ffunction-sections 
-fdata-sections 
-flto

Now switch to the Linker settings tab. The Link libraries: box provides a spot to add various libraries (for example, wxmsw28u for the Windows Unicode version of the wxWidgets monolithic dll). This program does not require any such libraries. The custom switches from the previous step require their link-time counterparts. Add

-flto 
-Os 
-Wl,--gc-sections 
-shared-libgcc 
-shared-libstdc++

to the Other linker options: tab. (For further details on what these switches do, see the GCC documentation on optimization options and linker options.)

Virtual Targets

Click OK to accept these changes and return to the previous dialog. Now that there are two release builds, it will take two separate runs of Build or Build and run to compile both. Fortunately, Code::Blocks provides the option to chain multiple builds together. Click Virtual targets..., then Add. Name the virtual target Releases and click OK. In the right-hand Build targets contained box, select both Release and Release small. Close out of this box and hit OK on the main window.


PIC

Figure 4.11: Virtual Targets

The virtual target ”Releases” will now be available from the Compiler Toolbar; building this should result in the following output.

-------------- Build: Release in HelloWorld --------------- 
 
Compiling: main.cpp 
Compiling: hello.cpp 
Linking console executable: bin\Release\HelloWorld.exe 
Output size is 457.50 KB 
 
-------------- Build: Release Small in HelloWorld --------------- 
 
Compiling: main.cpp 
Compiling: hello.cpp 
Linking console executable: bin\ReleaseSmall\HelloWorld.exe 
Output size is 8.00 KB 
Process terminated with status 0 (0 minutes, 1 seconds) 
0 errors, 0 warnings (0 minutes, 1 seconds)

4.3  Debugging with Code::Blocks

This section descibes how to work in debug mode.

4.3.1  Build debug version of your project

Make sure that the project is compiled with the -g (debugging symbols) compiler option on, and the -s (strip symbols) option off. This ensures that the executable has debug symbols included.

Compiler optimization switches should be turned off, stripping symbols (-s) must be turned off.

Keep in mind that you may have to re-build your project as up-to-date object files might not be re-compiled with -g otherwise. Please be aware that in compilers other than GCC, -g and/or -s might be a different switch (-s might not be available at all).

’Menu’ ’Project’ ’Build Options’


PIC

Figure 4.12: Debugger Project Build Options

4.3.2  Add Watches

In version 10.05

Note:
It’s a very old version. You should no more use it

Open The Debugger Watches Window.


PIC

Figure 4.13: Open a debugger Watch Window

The list of watches can be saved to a file and later re-loaded. To do so, right click in the list of watches and select ”save watch file” (and ”load watch file” to re-load them again).


PIC

Figure 4.14: Save a Watch Window

From 12.11 or latest nightly builds

In the latest nightly builds the watches window has been redesigned and works differently that the one in 10.05.

Currently there are three ways of adding watches in it:

  1. Click in the empty last row in the watches window, type the name of the variable (or full expression) and hit enter.
  2. While the debugger has stopped on a breakpoint select a variable name or full expression, right click to open the context menu and then select ”Add watch ’expression’”.
  3. Select an expression in the editor and drag’n’drop it in the watches window.

The automatic inclusion of local variables and function arguments have being reimplemented in 13.12.

4.3.3  Double-clicking in the Call stack window

Note:
when debugging, double-clicking on a frame in the ”call stack” debug window does not automatically update the variables displayed in the ”watches” debug window.

You have to right-click on a frame in the ”call stack” debug window and select ”Switch to this frame”.


PIC

Figure 4.15: A Watch Window

4.3.4  Set Breakpoints

Find the line containing the variable to be watched. Set a breakpoint in a position that will allow you to observe the variable value.

’Menu’ ’Debug’ ’Toggle Breakpoint’


PIC

Figure 4.16: Set Watch Variables

Run the debugger until the breakpoint is reached. Right click the variable to set a watch in the Watch Window.

Breakpoints may also be toggled with a left click in the left editor margin.

4.3.5  Notes

Script support

Code::Blocks natively use squirrel script language to deal with gdb, see: Debugger scripts (section 4.4 on page 310). As gdb 7.x support python pretty printer, so, it can also use gdb(with python support) to show some complex variable values. see forum thread unofficial MinGW GDB gdb with python released and Use GDB python under Codeblocks for more details.

Single file debugging

To debug your program you need to setup a project. Single file programs (without associated project) are not supported.

Path with spaces

Breakpoints could not work if the path/folder you’ve placed your project contains spaces or other special characters. To be safe use English letters, digits and ’_’.

Forking

If your application uses the ’fork’ system call you’ll have trouble stopping the debugged program or setting breakpoints on the fly. Here is a link explaining the forking modes of GDB: http://sourceware.org/gdb/onlinedocs/gdb/Forks.html

Update to the newest version of MinGW

From gdb 6.8 released on April 2008, it supports many features which does not exist in early versions. You can update by installing binaries from SourceForge MinGW64 packages.

Note:
TDM-Mingw package was a good choice until 5.1 version, but development is now abandonned.

Use 32bit CDB for 32bit programs and 64bit CDB for 64bit programs

It seems that debugging a 32bit program with 64bit CDB doesn’t work on Windows 7 (and more?), but 32bit CDB works perfectly.

Note:
This should no longer be the case with Code::Blocks rev>=10920. See the ticket for details: #429

Limits on the early version of MinGW

If your are still using the MinGW and gdb 6.7 from 8.02 setup files, setting breakpoints in the constructor will not work. Here are some tricks.

Breakpoints do not work in constructors or destructors in GDB 6.7 and earlier version. They do, however, work in routines called from them. This is an early GDB restriction, not a bug. So you could do something like:


PIC

Figure 4.17: Debugging With an old GDB

...and place a breakpoint in ”DebugCtorDtor” at the line "int i = 0;". The debugger will break at that line. If you then step the debugger (’Menu Debug’ ’Next Line’ ; or alternatively F7) you’ll reach the code in the constructor/destructor (”is_initialized = true/false;”).

4.4  Debugger scripts

This section descibes the debugger scripts.

4.4.1  Basic principle of the debugger script


PIC

Figure 4.18: Debugger command

Look at the image above, this will give a brief introduction of how the debugger script works. For example, you want to view the variable ”msg”. There are two handshake between the debugger plugin and gdb.

First, the debugger plugin sends a command to gdb to query the type of msg

whatis msg

then, the gdb will return the type

type = wxString

Secondly, the debugger checks that wxString is already registered, and sends the command

output /c msg.m_pchData[0]@((wxStringData*)msg.m_pchData-1)->nDataLength

Then, gdb replies with the string below:

{119 w, 120 x, 87 W, 105 i, 100 d, 103 g, 101 e, 116 t, 
115 s, 32  , 50 2, 46 ., 56 8, 46 ., 49 1, 48 0, 45 -, 
87 W, 105 i, 110 n, 100 d, 111 o, 119 w, 115 s, 45 -, 
85 U, 110 n, 105 i, 99 c, 111 o, 100 d, 101 e, 32  , 
98 b, 117 u, 105 i, 108 l, 100 d}

Finally, the value is shown in the watch window.

4.4.2  Script functions

Debugger scripts are similar to Visual Studio Debugger Visualizer. It allows you to write a small piece of code that gets executed by the debugger whenever you try to view a specific type of variables and can be used to show custom text with the important information you need in it.

Quote from Game_Ender - March 23, 2006

I don’t think its possible to open up another window to visualize something.

Let’s see how this works. Everything is inside a single file in the scripts/ folder, named gdb_types.script :). Support for more (user-defined) scripts is planned for the future.

This script is called by Code::Blocks at two places:

  1. when GDB is launched. It calls the script function RegisterTypes() to register all user-defined types known by the Code::Blocks’ debugger.
  2. whenever GDB encounters your variable type, it calls the script functions specific to this datatype (registered in RegisterTypes() - more on that below).

That’s the overview. Let’s dissect the shipped gdb_types.script and see how it adds std::string support to GDB.

// Registers new types with driver 
function RegisterTypes(driver) 
{ 
//    signature: 
//    driver.RegisterType(type_name, regex, eval_func, parse_func); 
 
   // STL String 
   driver.RegisterType( 
       _T("STL String"), 
       _T("[^[:alnum:]_]+string[^[:alnum:]_]*"), 
       _T("Evaluate_StlString"), 
       _T("Parse_StlString") 
   ); 
}

The ”driver” parameter is the debugger driver but you don’t need to care about it :) (currently this only works with GDB). This class contains a single method: RegisterType. Here’s its C++ declaration:

void GDB_driver::RegisterType(const wxString& name, const wxString& regex, 
                     const wxString& eval_func, const wxString& parse_func)

So, in the above script code, the ”STL String” (just a name - doesn’t matter what) type is registered, providing a regular expression string for the debugger plugin to match against and, finally, it provides the names of the two mandatory functions needed for each registered type:

  1. evaluation function: this must return a command understood by the actual debugger (GDB in this case). For the ”STL string”, the evaluate function returns an GDB ”output” command which will be executed by the GDB debugger.
  2. parser function: once the debugger runs the command returned by the evaluation function, it passes its result to this function for further processing. What this function returns, is what is actually diplayed by Code::Blocks (usually in the watches window or in a tooltip).

Let’s see the evaluation function for std::string:

function Evaluate_StlString(type, a_str, start, count) 
{ 
   local oper = _T("."); 
 
   if (type.Find(_T("*")) > 0) 
       oper = _T("->"); 
 
   local result = _T("output ") + a_str + oper + _T("c_str()[") 
                  + start + _T("]@"); 
   if (count != 0) 
       result = result + count; 
   else 
       result = result + a_str + oper + _T("size()"); 
   return result; 
}

I’m not going to explain what this function returns. I’ll just tell you that it returns a GDB command that will make GDB print the actual std::string’s contents. Yes, you need to know your debugger and its commands before you try to extend it.

What I will tell you though, is what those function arguments are.

Let’s see the relevant parser function now:

function Parse_StlString(a_str, start) 
{ 
   // nothing needs to be done 
   return a_str; 
}

Well, in this example, nothing needs to be done. ”a_str” already contains the std:string’s contents so we just return it :)

I suggest you study how wxString is registered in that same file, as a more complex example.

4.5  Code::Blocks and Makefiles

This section describes how to use a makefile in Code::Blocks by using a wxWidgeys example.

4.5.1  Wiki article

Author : Gavrillo 22:34, 21 May 2010 (UTC)

CB does not, by default, use makefiles. It has its own .cbp project files which do the same thing automatically. There are a few reasons that you might want to use a makefile. You maybe migrating a project that has a makefile into Code::Blocks. Another possibility is wanting to take a project out of Code::Blocks.

Needing to use a pre-processor is not a valid reason to use a makefile as CB has a pre/post build option. From the menu ’Project’ ’Build options’ there appears a tab with pre/post build steps that can be used for this purpose.

This chapter deals more specifically with makefiles using mingw32-make 3.81, CB 8.02 and Wxwidgets 2.8 on Windows Vista, although I am sure most of it will apply to other configurations.

If you decide that you want to use your own makefile, you need to enter the screen from ’Project’ ’Properties’ and you will see a tick box for ’this is a custom makefile’. Tick this box, make sure the name just above it is the one you want for your makefile.

You should also look at ’Project’ ’Build options’ . There is a tab called ’Make commands’ (you have to scroll the tabs horizontally to get to it). In the field ’build project/target’ you should see the line $make -f $makefile $target. Assuming you’re in debug mode $target will probably be called ’debug’ which is probably not what you want. You should change $target to your output file’s name (with the .exe extension and without the leading $).

One other useful addition is in ’Project’ ’Project Tree’ ’Edit File types and categories’ . If you add makefiles with the mask *.mak (CB seems to prefer .mak to .mk) you will be able to add your makefile with the extension .mak to the project and it will appear in the project management pane on the left.

Assuming you are going to edit the makefile within CB you should make sure that the editor uses tabs (as opposed to spaces). This is a generic problem with make as it needs to start command lines with a tab character and many editors replace tabs with spaces. To set this in CB, open the ’Settings’ ’Editor’ window and check the tickbox for use tab character.

The real problems start now however. CB’s automatic makefile adds all the headers for Wxwidgets, but if you use a makefile, all this is turned off and you have to do this yourself.

Fortunately CB has another feature that can come to your rescue. If you go to menu ’Settings’ ’Compiler and Debugger’ , scroll the tabs horizontally to the right end and you will find the tab ’other settings’. In there click on the tick box for ’Save build to HTML ...’. This will cause CB to create, at build time, an HTML file that records all the build commands.

Note:
this way to create an html buid file does not exist in recent CB version, but thera are other solutions

If you compile (without using a makefile - so if you’ve already reset everything - sorry) the default wxWwidgets minimum program, you can see the compiler and linker commands that produce this file.

Assuming that you are going to use this as the basis for your project, you can use the content of the HTML file produced as the basis of your makefile.

You can’t just copy it from the HTML viewer in CB (there’s no such facility in CB) but you can load the file into a browser or editor and copy it from there. It can be found in your project directory with <the_same_name_as_your_project>_build_log.HTML. Sadly it will require a little tweaking as shown below.

Here’s a copy of a build file for the basic Wxwidgets program as described above.

Note:
for a better lisibility, long lines have been split. The ^ sign indicates a continuation line in DOS mode, and \ sign indicates a continuation line in a makefile. But you should have commands on only one line as long as you remove continuation line signs

mingw32-make.exe -f test.mak test.exe

mingw32-g++.exe -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL         ^
    -DwxUSE_UNICODE -Wall -g -D__WXDEBUG__ -IC:\PF\wxWidgets2.8\include         ^
    -IC:\PF\wxWidgets2.8\contrib\include -IC:\PF\wxWidgets2.8\lib\gcc_dll\mswud ^
    -c C:\Development\test\testApp.cpp -o obj\Debug\testApp.o

mingw32-g++.exe -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL         ^
    -DwxUSE_UNICODE -Wall -g -D__WXDEBUG__ -IC:\PF\wxWidgets2.8\include         ^
    -IC:\PF\wxWidgets2.8\contrib\include -IC:\PF\wxWidgets2.8\lib\gcc_dll\mswud ^
    -c C:\Development\test\testMain.cpp -o obj\Debug\testMain.o

windres -IC:\PF\wxWidgets2.8\include -IC:\PF\wxWidgets2.8\contrib\include       ^
    -IC:\PF\wxWidgets2.8\lib\gcc_dll\mswud -iC:\Development\test\resource.rc    ^
    -o obj\Debug\resource.coff

mingw32-g++.exe -LC:\PF\wxWidgets2.8\lib\gcc_dll -o bin\Debug\test.exe          ^
    obj\Debug\testApp.o obj\Debug\testMain.o obj\Debug\resource.coff            ^
    -lwxmsw28ud -mwindows

Process terminated with status 0 (0 minutes, 12 seconds)
0 errors, 0 warnings

The above can be converted to the makefile below. I have deliberately left it fairly close to the HTML file output.

# test program makefile

Incpath1 = C:\PF\wxWidgets2.8\include
Incpath2 = C:\PF\wxWidgets2.8\contrib\include
Incpath3 = C:\PF\wxWidgets2.8\lib\gcc_dll\mswud

Libpath = C:\PF\wxWidgets2.8\lib\gcc_dll

flags = -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL     \
        -DwxUSE_UNICODE -Wall -g -D__WXDEBUG__

CXX = mingw32-g++.exe

test.exe : obj\Debug\testApp.o obj\Debug\testMain.o obj\Debug\resource.coff
    $(CXX) -L$(Libpath) -o bin\Debug\test.exe obj\Debug\testApp.o           \
    obj\Debug\testMain.o obj\Debug\resource.coff -lwxmsw28ud -mwindows

obj\Debug\testMain.o : C:\Development\test\testMain.cpp
    $(CXX) $(flags) -I$(Incpath1) -I$(Incpath2) -I$(Incpath3)               \
    -c C:\Development\test\testMain.cpp -o obj\Debug\testMain.o

obj\Debug\testApp.o : C:\Development\test\testApp.cpp
    $(CXX) $(flags) -I$(Incpath1) -I$(Incpath2) -I$(Incpath3)               \
    -c C:\Development\test\testApp.cpp -o obj\Debug\testApp.o

obj\Debug\resource.coff : C:\Development\test\resource.rc
    windres -I$(Incpath1) -I$(Incpath2) -I$(Incpath3)                       \
    -iC:\Development\test\resource.rc -oobj\Debug\resource.coff

# original output from codeblocks compilation
# note I’ve had to add compiling the .res file
#
# mingw32-g++.exe -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL       ^
#   -DwxUSE_UNICODE -Wall -Wall -g -D__WXDEBUG__                                ^
#   -Wall -g -IC:\PF\wxWidgets2.8\include -IC:\PF\wxWidgets2.8\contrib\include  ^
#   -IC:\PF\wxWidgets2.8\lib\gcc_dll\mswud                                      ^
#   -c C:\Development\test\testApp.cpp -o obj\Debug\testApp.o

# mingw32-g++.exe -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL       ^
#   -DwxUSE_UNICODE -Wall -Wall -g -D__WXDEBUG__                                ^
#   -Wall -g -IC:\PF\wxWidgets2.8\include -IC:\PF\wxWidgets2.8\contrib\include  ^
                                                                                       
                                                                                       
#   -IC:\PF\wxWidgets2.8\lib\gcc_dll\mswud                                      ^
#   -c C:\Development\test\testMain.cpp -o obj\Debug\testMain.o

# mingw32-g++.exe -LC:\PF\wxWidgets2.8\lib\gcc_dll -o bin\Debug\test.exe        ^
#    obj\Debug\testApp.o obj\Debug\testMain.o                                   ^
#    obj\Debug\resource.res -lwxmsw28ud -mwindows

I have written a generic makefile which I have only tested on Windows Vista but that should work with any project started as described above. You have to change the name of the project and set the paths as appropriate (you will probably only need to change Ppath and WXpath).


# Generic program makefile
# -- assumes that you name your directory with same name as the project file
# -- eg project test is in <development path>\test\

# Project name and version
Proj := test
Version := Debug

#paths for Project (Ppath) Object files (Opath) and binary path (Bpath)
Ppath := C:\Development\$(Proj)
Opath := obj\$(Version)
Bpath := bin\$(Version)

#Library & header paths
WXpath := C:\PF\wxWidgets2.8
IncWX := $(WXpath)\include
IncCON := $(WXpath)\contrib\include
IncMSW := $(WXpath)\lib\gcc_dll\mswud
Libpath := $(WXpath)\lib\gcc_dll

flags = -pipe -mthreads -D__GNUWIN32__ -D__WXMSW__ -DWXUSINGDLL -DwxUSE_UNICODE \
        -Wall -g -D__WXDEBUG__

CXX = mingw32-g++.exe

Obj := $(Opath)\$(Proj)App.o $(Opath)\$(Proj)Main.o $(Opath)\resource.coff

$(Proj).exe : $(Obj)
    $(CXX) -L$(Libpath) -o $(Bpath)\$(Proj).exe $(Obj) -lwxmsw28ud -mwindows

$(Opath)\$(Proj)Main.o : $(Ppath)\$(Proj)Main.cpp
    $(CXX) $(flags) -I$(IncWX) -I$(IncCON) -I$(IncMSW) -c $^ -o $@

$(Opath)\$(Proj)App.o : C:\Development\$(Proj)\$(Proj)App.cpp
    $(CXX) $(flags) -I$(IncWX) -I$(IncCON) -I$(IncMSW) -c $^ -o $@

$(Opath)\resource.coff : C:\Development\$(Proj)\resource.rc
    windres -I$(IncWX) -I$(IncCON) -I$(IncMSW) -i$^ -o $@

.PHONEY : clean
                                                                                       
                                                                                       

clean:
    del $(Bpath)\$(Proj).exe $(Obj) $(Opath)\resource.coff

Note:
exporting a makefile of a Code::Blocks project is indirectly possible. This can be achieved with cbp2make utility (see the description in section 4.6 on page 324 and/or a usage exemple via Tool+ subsection 2.17.1 on page 196.

One final remark. Once you have used a makefile, attempts to compile anything without a makefile result in a number of warnings. The only way I have found of solving this is to reinstall CB. If you need both perhaps it is possible to install 2 versions of CB.

Note:
this is a remark of the original text in the Wiki. But, as gd_on user, I’ve never had such behaviour

In general it is not recommended to use a makefile, but if there is some compelling reason to do so, then this is how.

4.5.2  Complements

By default, Code::Blocks builds a ”Release” and a ”Debug” target. In your Makefile, these targets may be not present. But may be you have an ”All” target (or ”all”). You can rename in Code::Blocks the target (or add one) with this name given in the Makefile.

More, your Makefile builds the executable with a specific name in a specific folder. You should adjust in Code::Blocks the path and the name of the executable. Like that, Code::Blocks, as it does not know nor analyse the Makefile, will find the executable, and the Execute Green arrow in the menu can work (or Ctrl-F10).

4.6  Cbp2make Utility

A Makefile generation tool for Code::Blocks IDE by Mirai Computing. The text of this section comes from his cbp2make Wiki on SourceForge.

Note:
Cbp2make is not a Code::Blocks plugin, but a full console application, located in the main Code::Blocks directory, to generate makefile(s) from the internal Code::Blocks generating system

4.6.1  About

”cbp2make” is a stand-alone build tool that allows you to generate makefile(s) for GNU Make out of Code::Blocks IDE project or workspace file. (See also http://forums.codeblocks.org/index.php/topic,13675.0.html])

4.6.2  Usage

Create makefile for a single project or workspace

Let’s assume you have a project ”my_project.cbp” and you need a makefile for this project. In this simplest case all you have to do is:

cbp2make -in my_project.cbp

Same thing applies to workspaces.

cbp2make -in my_projects.workspace

Create makefile with another filename

By default, ”cbp2make” will append ”.mak” extension to the project name to compose a filename for makefile. If you want to change that, use following command:

cbp2make -in my_project.cbp -out Makefile

Create makefile for another platform

If you are working in GNU/Linux and you want to generate a makefile for Windows or the other way around, you can specify one or more platforms for which you need makefiles.

cbp2make -in my_project.cbp -windows 
cbp2make -in my_project.cbp -unix 
cbp2make -in my_project.cbp -unix -windows -mac 
cbp2make -in my_project.cbp --all-os

”cbp2make” will append ”.unix” or ”.windows” or ”.mac” suffix to makefile name for each platform respectively.

Create makefile for multiple projects or workspaces

If you have more than one independent project or workspace, you can process them at once, but you will need a text file containing the list of projects, e.g., ”projects.lst”, with one project filename per line.

   my_project.cbp 
   my_other_project.cbp

And then you can process them using command:

cbp2make -list -in projects.lst

4.6.3  Configuration

Some installation-specific or project-specific options, primarily toolchain settings, can be saved to a configuration file. By default (since rev.110), cbp2make does not save any settings to a configuration file unless the user explicitly specifies the "--config" option. A configuration file can be either global (stored in user profile / home directory) or local (stored in current directory).

Please, keep in mind that since cbp2make is in early stage of development, an old configuration file may become incompatible with new tool version and it may be necessary to update it manually or initialize a new one.

Initialization

cbp2make --config options --global 
cbp2make --config options --local

Later use

When cbp2make is invoked, first it tries to load a local configuration file. If a local configuration is missing, next attempt will be to load a global one. If this attempt is not successful either, the default built-in configuration is used. Configuration lookup order can be overridden with "--local" or "--global" command line options. If one of options is supplied to cbp2make, non-specified configuration is not picked up even if the specified one is missing and non-specified do exists.

Default lookup order

cbp2make -in project.cbp -out Makefile}

Explicitly specified configuration

cbp2make --local -in project.cbp -out Makefile 
cbp2make --global -in project.cbp -out Makefile

4.6.4  Command line syntax

Generate makefile:

cbp2make -in <project_file> [-cfg <configuration>] [-out <makefile>]
[-unix] [-windows] [-mac] [--all-os] [-targets "<target1>[,<target2>[, ...]]"]
[--flat-objects] [--flat-objpath] [--wrap-objects] [--wrap-options]
[--with-deps] [--keep-objdir] [--keep-outdir] [--target-case keep|lower|upper]

cbp2make -list -in <project_file_list> [-cfg <configuration>]
[-unix] [-windows] [-mac] [--all-os] [-targets "<target1>[,<target2>[, ...]]"]
[--flat-objects] [--flat-objpath] [--wrap-objects] [--wrap-options]
[--with-deps] [--keep-objdir] [--keep-outdir] [--target-case keep|lower|upper]

Manage toolchains:

cbp2make --config toolchain --add \[-unix|-windows|-mac\] -chain <toolchain>
cbp2make --config toolchain --remove \[-unix|-windows|-mac\] -chain <toolchain>

Manage build tools:

cbp2make --config tool --add \[-unix|-windows|-mac\] -chain <toolchain>
         -tool <tool> -type <type> <tool options>

cbp2make --config tool --remove \[-unix|-windows|-mac\] -chain <toolchain>
         -tool <tool>

Tool types:

    pp=preprocessor as=assembler cc=compiler rc=resource compiler
    sl=static linker dl=dynamic linker el=executable linker
    nl=native linker

Tool options (common):

    -desc <description> -program <executable> -command <command_template>
    -mkv <make_variable> -srcext <source_extensions> -outext <output_extension>
    -quotepath <yes|no> -fullpath <yes|no> -unixpath <yes|no>

Tool options (compiler):

    -incsw <include_switch> -defsw <define_switch> -deps <yes|no>
\end{lstlisting}

Tool options (linker):
\begin{verbatim}
    -ldsw <library_dir_switch> -llsw <link_library_switch> -lpfx <library_prefix>
    -lext <library_extension> -objext <object_extension> -lflat <yes|no>

Manage platforms:

cbp2make --config platform \[-unix|-windows|-mac\] \[-pwd <print_dir_command>\]
         \[-cd <change_dir_command>\] \[-rm <remove_file_command>\]
         \[-rmf <remove_file_forced>\] \[-rmd <remove_dir_command>\]
         \[-cp <copy_file_command>\] \[-mv <move_file_command>\]
         \[-md <make_dir_command>\] \[-mdf <make_dir_forced>\]
         \[-make <default_make_tool>\]

Manage global compiler variables:

cbp2make --config variable --add \[-set <set_name>\] -name <var_name>
        \[-desc <description>\] \[-field <field_name>\] -value <var_value>

cbp2make --config variable --remove \[-set <set_name>\] \[-name <var_name>\]
        \[-field <field_name>\]

Manage options:

cbp2make --config options --default-options "<options>"
cbp2make --config show

Common options:

cbp2make --local         // use configuration from current directory
cbp2make --global        // use configuration from home directory
cbp2make --verbose       // show project information
cbp2make --quiet         // hide all messages
cbp2make --help          // display this message
cbp2make --version       // display version information

Options

"Makefile generation" 
 
   -in <project_file>   // specifies an input file or a list of files; 
 
   -cfg <configuration> // specifies a configuration file, see also "--local" 
                          and "--global" options; 
 
   -out <makefile>     // specifies the name of a makefile or a list of 
                          makefiles; 
 
   -unix               // enables Unix / Linux compatible makefile generation; 
 
   -windows            // enables Windows compatible makefile generation; 
 
   -mac                // enables Macintosh compatible makefile generation; 
 
   --all-os            // enables all target platforms at once; 
 
   -targets "<target1>[,<target2>[, ...]]" // specifies the only build targets 
                                             that a makefile will be made for; 
 
   --flat-objects      // forces "flat" names for object files with limited 
                          character set; 
 
   --flat-objpath      // forces "flat" paths for object files with no 
                          subdirectories; 
 
   --wrap-objects      // allows to use multiline lists of objects which 
                          makes a makefile easier to read; 
 
   --wrap-options      // allows to use multiline macros; 
 
   --with-deps         // allows a built-in dependency scanner for C/C++ 
                          projects; 
 
   --keep-objdir       // disables command that erase directories for 
                          object files in clean target; 
 
   --keep-outdir       // disables command that erase directory for an 
                          output binary file in clean target; 
 
   --target-case [keep|lower|upper] // specifies style for makefile targets;

4.7  Code::Blocks Interface Internationalization

This section describes how to obtain and use a localized version of Code::Blocks.

Code::Blocks interface can be presented in your own language. Many of the strings used internally in Code::Blocks interface are introduced with a wxWidgets macro : _. Strings that do not change with the language are normally introduced by the macro wxT() or _T(). To obtain a Code::Blocks interface presented in your own language, you must simply tell to Code::Blocks that a language file is available. To be understandable by Code::Blocks, it must be a .mo file obtained after ”compilation” of a .po file. Such files are available on the forum for ”French” and in Launchpad web site for a wider set of languages.

The original site on Launchpad is now obsolete : https://launchpad.net/codeblocks
The forum topic concerning translation is http://forums.codeblocks.org/index.php/topic,1022.0.html. If you are interested, you’ll can also find tools to extract strings from Code::Blocks source files. These tools create a .pot file, which can then be completed with the translations to create a .po file.
A new web site has been created recently on https://launchpad.net/codeblocks-gd. It contains more than 9300 strings though the original site had only 2173! Many works has been done on Code::Blocks!

In the translation page choose ”View All languages”, at the bottom, right.

Old translations have been imported in that new page, only the most used languages (currently 14 languages). On request, it’s possible to add new languages (but translators will have a little bit more work !).
Sorry for that, but original translators names have been lost in many cases :-[.
French language has the greatest number of translated strings. The template (.pot file) has been updated for recent svn versions and Launchpad contain the translation work done until now. For Russian language, a quite recent web page has also been used but is not up to date. Many translations need to be approved, but I’m not the right guy to do that !
The launchpad page is opened as ”structured”. So you should be able to propose new translations or to correct them. In some cases, they should be approved by somebody before publication.
I’ll try to maintain the ”template” when new English strings will be available.

You (translators) should be able to participate. You only need to have (or create) a launchpad (Ubuntu) account.

Other users can request a download for the .po or .mo file. It’s this last one (.mo file), the binary form, that you can use to obtain Code::Blocks interface in your own language : simply put it in your ”codeblocks installation directory”/share/CodeBlocks/locale/”language string” (for me, on Windows, it’s
C:\Program Files\CodeBlocks_wx313_64\share\CodeBlocks\locale\fr_FR. Then in the menu Settings/Environment.../View you should be able to choose the language.

Some more details for using translated menu strings in Code::Blocks.

For users of translations only :

Download the .mo format file in the requested language button. The launchpad name may be something like : de_LC_MESSAGES_All_codeblocks.mo (here for german).

You should put this file inside your codeblocks installation directory.

On Windows, it’s something like :
C:\Program Files (x86)\CodeBlocks\share\CodeBlocks\locale\xxxx for 32 bits
or
C:\Program Files\CodeBlocks\share\CodeBlocks\locale\xxxx for 64 bits.

Path under Linux is quite similar.

xxxx has to be adapted to your language. It’s :

Create the sub-directories if needed. Then place your downloaded .mo file here. You can leave the filename as it is or only keep the first letters (as you want), but keep the extension .mo.

Then start Code::Blocks. In Setting/Environment/View you should be able to check the language box (internationalization) and choose your language. If not, you have probably forgotten something or made an error.
Restart Code::Blocks to activate the new language.

If you want to switch back to English, simply uncheck the language box.

For translators :

You can directly work in launchpad.

Problem : the interface is not really friendly.

You can also download the .po file, work on it with poedit for example (free version is OK). You can test your translations locally by compiling (creating a .mo file) and installing the .mo file in the right sub-directory of Code::Blocks.

When you have made sufficient progress (it’s your decision), you can upload you .po file in launchpad. It may be necessary to approve your work or mark it as to be reviewed.

Don’t be afraid: it’s quite a long work. In the old site, there was 2173 string to translate. Now it’s more than 9300 ! But the job can be shared, Launchpad is done for that !

Tip : Begin with menus that you use often : you’ll see progress faster.

5  Installing and Configuring CodeBlocks with MinGW

This chapter describes how to install and configure Code::Blocks. Install process is here described for Windows, but may be adapted to other OS.

5.1  Installing the latest official version of Code::Blocks on Windows

Install steps:

We’ll see in the next section how to install and configure an other compiler.

Notes:

5.2  Configuring MinGW

This sections describes how to install and configure MinGW.

5.2.1  Overview

A compiler toolchain is what Code::Blocks uses to turn the code you type into it into numbers that the computer understands. As a compiler toolchain is a very complex undertaking it is not part of Code::Blocks itself but rather is a separate project that Code::Blocks then uses. The kind of compiler toolchains talked about on this page are ”MinGW” toolchains. Which means ”Minimalist GNU for Windows.” And ”GNU” expands to ”GNU’s Not Unix.” More information about the GNU project can be found on the GNU Home Page.

For most MinGW-based compiler toolchains, having your toolchain in your PATH is important because it means that during development the toolchain libraries will be accessible by default to your programs as you develop them and also makes it easier to use utilities such as CMake as they will be able to find your compiler toolchain. When you actually distribute your programs to other computers then you will copy the needed .dll files out of your toolchain directory and include them as part of your installer. On your machine they are in your PATH so you always have them, on your users computers they won’t have the compiler toolchain so there you provide the .dll files with your program.

5.2.2  MinGW Compiler Toolchains

You can find on the web several MinGW distributions. Here is a list of distributions, which is not exhaustive.

MinGW - The original project
http://www.mingw.org/: 32 bit only compilers;
TDM distribution
http://tdm-gcc.tdragon.net/: 32 and 64 bit, but old 5.1 distribution, now obsolete. Was used and distributed with Code::Blocks 17.12;
New TDM distribution
http://tdm-gcc.tdragon.net/: based on 9.2 distribution. Seems to have sometime problems, at least to compile Code::Blocks itself.
MinGW 64
https://sourceforge.net/projects/mingw-w64/files/: 32 and 64 bit, 8.1 distribution in Toolchains targetting. The parent project of MinGW-Builds, includes much more than is necessary - MinGW-Builds will usually suffice instead of the full works. Several choices are available: for a 32 bit compiler, you can choose the posix, sjlj version (i686-posix-sjlj) and for a 64 bits compiler you can choose the posix, seh version (x86_64-posix-seh) (choices compatibles with the old TDM version). This is this last 64 bits-posix-seh version which is used in recent nightlies compiled versions of Code::Blocks and distributed with 20.03 version. Other choices also work: depends on your needs. gcc, g++, gfortran, gdb, lto, omp, mingw32-make are in the distribution;
MinGW 64 Ray_linn Personal build
https://sourceforge.net/projects/mingw-w64/files/Multilib%20Toolchains%28Targetting%20Win32%20and%20Win64%29/ray_linn/gcc-9.x-with-ada/: a 64 bit, 9.1 distribution in the personal build sub-directory. Use static libraries, so no need to distribute dlls with your own executables, but they are bigger. ada, gcc, g++, gfortran, lto, objc, obj-c++, omp are in the distribution. Problem : gdb and make are not included.
MinGW Equation
http://www.equation.com/servlet/equation.cmd?fa=fortran: 32 and 64 bit recent versions (several choices). Use static libraries, so as with above version, produces bigger executables but no need to distribute dlls with your own executables. gcc, g++, gfortran, gdb, lto, omp, make are in the distribution;
MinGW LH_Mouse version
https://gcc-mcf.lhmouse.com/: recent 32 and 64 bits versions (but not the last one). No gfortran (?). Special thread model (mcf). gcc, g++, gdb, lto, omp, mingw32-make are in the distribution. No gfortran;
MinGW on Winlibs
http://winlibs.com/: recent 64 bit compiler (but not the last one). gcc, g++, gfortran, gdb, lto, objc, obj-c++, omp, mingw32-make are in the distribution;
Msys2
http://www.msys2.org/ and https://packages.msys2.org/group/mingw-w64-x86_64-toolchain: 32 and/or 64 bit versions, installed in C:\msys64\mingw32 and C:\msys64\mingw64. Unlike the above all-in-one installers, additional work is necessary to adjust to your needs, your compilers, and your toolchains. Read carefully the documentation. Nevertheless, Msys2 produces recent compiler versions and gives access to an update facility: pacman. ada, gcc, g++, gfortran, gdb, lto, objc, obj-c++, omp, mingw32-make are in the distribution or may be installed via pacman;

5.2.3  Code::Blocks Configuration

Go to your Compiler settings:


PIC

Figure 5.1: Compiler Settings

And then under the ”Toolchain executables” tab (red arrow), click on the ellipsis (”...”, blue arrow) and choose the root directory where you installed MinGW (64-bit here). Once you have that directory chosen, in the ”Program Files” sub-tab (green arrow) area fill out the fields as shown. If you aren’t using the MinGW 64-bit toolchain there might be minor variation in the executable names. If you choose the blue arrow ellipsis first then for each ellipsis you click on under ”Program Files” you will already be in your MinGW 64-bit bin directory where the actual programs are.


PIC

Figure 5.2: Code::Blocks Toolchain Configuration

Note: You can enter name as gcc.exe or x86_64-w64-mingw32-gcc.exe or mingw32-gcc.exe (depends of the distribution): it’s the same executable. Idem for g++.

Note:
To configure a new compiler, gfortran for example, enter gfortran.exe in the 3 first text fields in the tab ”Program Files”, or the exact name as it is in your distribution.

Now, go to your Debugger settings:


PIC

Figure 5.3: Settings Debugger

Choose your default debugger (red arrow), and then fill in the Executable path for it as shown for MinGW 64-bit (blue arrow).


PIC

Figure 5.4: Debugger Default Configuration

Summary

You now have a Code::Blocks environment that is configured to use MinGW 64-bit properly. Using this guide as a template you can easily set up alternative compiler toolchains no matter the source - just follow the same basic procedure.

Development Tools

Normally you should not need many of these tools. ZIP is convenient, especially when: building Code::Blocks itself, is often already included with MinGW, but other than that these tools only serve specialized purposes.

5.3  Nightly version of Code::Blocks on Windows

Nightly builds are provided ”as is”. They are ”binary” distributions, normally provided daily, representing the latest and greatest state of the Code::Blocks sources. Normally they are pretty stable, but however they can introduce new bugs, regressions, and on the other hand they can introduce new features, bug fixes, ...

Before we describe what the builds contain, it might be better to start with what the nightly builds DO NOT contain. For starters ask yourself the question : what is Code::Blocks?
Well it’s an IDE (Integrated Development Environment), it means it integrates different tools and makes them work together. So Code::Blocks is NOT a compiler (nor MS, nor Borland, nor GCC), it is not a debugger (nor MS, nor GDB), it is not a makefile system ! So these components are NOT part of Code::Blocks, and therefore they are also not part of the nightly build distributions. However several of the mentioned development components can be combined to work together nicely through Code::Blocks. For example, you can plug-in the GCC compiler and GDB debugger and compile and debug your hand written applications.
Code::Blocks itself is being compiled with GCC, on windows this is done by using the MinGW port. Since CB is a multi-threaded application it needs supporting routines providing it the multi-threading functionalities. This is provided by means of the MinGW port, more specifically the ”mingwm10.dll”, in every post announcing a new nightly build you can find a link to download this dll. Code::Blocks has a GUI (Graphical User Interface). There are several ways to create a GUI: you can code by using the core Windows API (works only on Windows) or you can use MS-MFC (works only on Windows) or you can use some 3rd party abstraction of GUI’s, like QT, wxWidgets, Tk, ...
Code::Blocks uses wxWidgets, next to GUI abstraction, wxWidgets brings many more abstractions (strings, files, streams, sockets, ...) and the good thing is : wxWidgets brings these abstractions for many different platforms (Windows, Linux, Apple, ...). This means Code::Blocks needs to be equipped with the real functionalities from wxWidgets (say the binary code doing the actual work), this is being achieved through a dll : ”wxmsw31u_gcc_cb.dll” (17.12 was built with ”wxmsw28u_gcc_cb.dll”). Again on every announcement post of a new nightly build you can find a download link for this dll and other prerequisites.

When, on Windows, you install an official version with MinGW included (the recommended version), you’ll find a MinGW directory in C:\Program Files\codeblocks. If it works well here for most of your usages, it’s not the best place, at least because there is a space inside the path name and this space may trouble some of the components of MinGW. So, on Windows, move this MinGW directory to C:, for example. You can even rename it MinGW32 for a 32 bit toolchain version or MinGW64 for a 64 bit version.

As told previously, a good solution to install a nightly is first to install an official version, then to configure and try it. Like that, your configuration files, associations, shortcuts, will be correctly set and they’ll be kept for your nightly install. The link to find last nightlies is http://forums.codeblocks.org/index.php/board,20.0.html.

If you install your nightly above a 20.03 version, you need to follow carefully this procedure because a few things have changed, at least they are compiled in 64 bits with a recent compiler, a different wxWidgets version and need some complementary dlls.

Normally, you won’t have such kind of problems with a 20.03 version.

Normally, that’s all. Your nightly is ready to use ...

6  Building CodeBlocks from sources

This chapter describes how to build Code::Blocks itself.

Note:
This chapter existed in the version 1 original .tex files but was not published in all languages. It needs to be reviewed and completed at least for Linux users.

6.1  Introduction

This section will describe the process used in creating the nightly builds, and can be used as a guideline if you want to build Code::Blocks yourself. It is described as a sequence of actions.

In order to perform our build tasks, we will need several tools. Let’s create an ingredient list for our cooking experiments

6.2  Windows or Linux

This section is written for a Windows syntax, but can be easily adapted to Linux. Other detailed instructions may be found in the Wiki: http://wiki.codeblocks.org/index.php/Installing_Code::Blocks_from_source_on_Linux

Since the Code::Blocks developers build Code::Blocks using GCC, we might as well use that one under windows. The easiest and cleanest port is MinGW. This is the compiler distributed with Code::Blocks when you download the FULL official package. With C::B 17.12 it was a version TDM 5.1.0 which was distributed, a rather old one. We will stick to version 8.1.0, which works nicely and which is now distributed with Code::Blocks 20.03, but you can find more recent ones (as shown in MinGW Compiler Toolchains (subsection 5.2.2 on page 346). You can obtain two MinGW versions : for generating 32 bits code, or for 64 bits code on [url :MinGW64]. Many sub-choices are available. For 32 bits, I would suggest i686-posix-sjlj and for 64 bits x86_64-posix-seh.

Note:
The 64 bits version can produce 64 bits code and 32 bits code. Simply add option -m64 or -m32 to your compilation AND link options. Normally, 32 bits version produces only 32 bits code. Just take care that if you use libraries, static or dynamic, they must have been generated with the same bits number

First, a brief explanation of MinGW components:

gcc-core
the core of the GCC suite
gcc-g++
the c and c++ compiler
gfortran
the fortran compiler. IMPORTANT : gfortran 5.1 has a bug : open instruction, to open a file for reading data, does not work ! Unfortunately, last TDM distribution use this gfortran and has not been corrected.
mingw Runtime
implementation of the run time libraries
mingw utils
several utilities (implementation of smaller programs that GCC itself uses)
win32Api
the APIs for creating Windows programs
binutils
several utilities used in build environments
make
the Gnu make program, so you can build from make files
GDB
the Gnu debugger

I would suggest extracting (and installing for the GDB if necessary) everything in the C:\MinGW directory. If you want to have both 32 bits and 64 bits version, you can install them respectively in C:\MinGW32 and C:\MinGW64. The remainder of this article will assume that this is where you have put it. If you already have an installation of Code::Blocks that came bundled with MinGW, I still advise you to install MinGW as described here. A compiler does not belong under the directory tree of an IDE; they are two separate things. Code::Blocks just brings it along in the official versions so that the average user does not need to bother with this process. Nevertheless, some Unix Like tools have problems when they are installed in a path containing spaces (or even accentuated characters) in the name of subdirectories. For a simple usage of C::B, you won’t have problems, but this may happen, so don’t hesitate to move your C:\Program Files\Codeblocks\MinGW to something like C:\MinGW.

You may need to add the bin directory of your MinGW (and/or MinGW32/MinGW64) installation to your path. An easy way to do this is with the following command at the command prompt:

set path=%PATH%;C:\MinGW32\bin;C:\MinGW32\i686-w64-mingw32\bin;

or (for 64 bits installation):

set path=%PATH%;C:\MinGW64\bin;C:\MinGW64\x86_64-w64-mingw32\bin;

It’s necessary to run your executable outside Code::Blocks IDE context.

You can also modify the global (or user) PATH enviroment variable.

6.2.1  Initial Build System

For [url :CODEBLOCKS] a project description CodeBlocks.cbp is available. If you load this project file in Code::Blocks then you are able to build Code::Blocks from sources. All we need to do is get hold of a pre-built version Code::Blocks.

First, download a nightly build. You can make your selection from here ([url :FORUM] in Nightly Builds). The nightly builds are unicode versions, containing the core and contributed plug-ins. Read carefully the first post of each nightly : it contains instructions to download and install additionnal dlls, necessary with recent MinGW versions, particularly version 8.1 used to build recent C::B nightly.

Next, unpack the 7-zip file to any directory you like. If you don’t have 7-zip, you can download it for free from [url :7Z].

Now, Code::Blocks needs one more dll to work correctly: the WxWidgets dll. You can also download it at the nightly builds forum. Just unzip it into the same directory that you unpacked the Code::Blocks nightly build. It also needs the mingwm10.dll. It’s in the bin directory of our MinGW installation. So, it’s important to make sure the bin directory of your MinGW installation is in your path variable.

Finally, start up this new nightly build of Code::Blocks. It should discover the MinGW compiler we just installed. If it’s not the case, go to the menu ”Settings / Compiler... / Toolchain executables” and adjust the MinGW path to your specific installation.

6.2.2  Version Control System

In order to be able to retrieve the latest and greatest Code::Blocks sources, we need to install a Version Control System.

The Code::Blocks developers provide their sources through the version control system[url :Subversion]. So, we need a client to access their svn repository of sources. A nice, easy client for Windows is [url :TortoiseSVN], which is freely available. Download and install it, keeping all suggested settings.

Now, go create a directory wherever you like, for example D:\projects\CodeBlocks. Right click on that directory and choose from the pop-up menu: svn-checkout. In the dialog that pops up, fill in the following information for Url of Repository:

svn://svn.code.sf.net/p/codeblocks/code/trunk

and leave all other settings as they are.

Now be patient while TortoiseSVN retrieves the most recent source files from the Code::Blocks repository into our local directory. Yes; all those Code::Blocks sources are coming your way!

For more info on SVN settings, see info on SVN settings (Nevertheless, this C::B plugin does not exist anymore in recent Code::Blocks versions). If you don’t like an Explorer integration or look for a cross-plattform client you might want to have a look at RapidSVN.

6.2.3  wxWidgets

[url :Wxwidgets] is a platform abstraction that provides an API to support many things such as GUI, sockets, files, registry functionality. By using this API, you can create a platform independent program.

Code::Blocks is a wxWidgets (here after: wx) application, that means if you want to run Code::Blocks you needed the wx functionality. This can be provided in a couple of ways. It could be a .dll or a static library. Code::Blocks uses wx as a dll and this dll can also be downloaded from the nightly build section of the forum.

However, if we want to build a wx application, we need to include the headers of the wx sources. They tell the compiler about the functionality of wx. In addition to those header files, our application needs to link to the wx import libraries. Well, let’s take it step by step.

Wx is provided as a zip file of it’s sources, so we need to build that ourselves. We already shopped for the MinGW compiler, so we have all the tools we need at hand.

Next, let’s unzip the wx sources into C:\Projects so we will end up with a wx root directory like this: C:\Projects\wxWidgets-3.1.3. Next unzip the patch into the same directory letting it overwrite files. Note that we are going to refer to the wx root directory from now on as <wxDir>

Now, we are going to build the wxWidgets. This is how we do it:

First, make sure C:\MingGW\bin is in your path, during the build some programs will be called that reside in the the MinGW\bin directory. Also, Make has to be version 3.80 or above.

Now it is time to compile wxWidgets. Open the command prompt and change to the wxWidgets directory:

cd <wxDir>\build\msw

We are now in the right place. We are first going to clean up the source:

mingw32-make -f makefile.gcc SHARED=1 MONOLITHIC=1 BUILD=release UNICODE=1 clean

Once everything is clean, we can compile wxWidgets:

mingw32-make -f makefile.gcc SHARED=1 MONOLITHIC=1 BUILD=release UNICODE=1

If you are using TDM-gcc 5.1.0, or gcc 8.1.0, you will have to add the following options on the same command line :

USE_XRC=1
CXXFLAGS+="-fpermissive -fno-keep-inline-dllexport -std=gnu++11
 -Wno-deprecated-declarations"
LDFLAGS="-Wl,--allow-multiple-definition"

Code::Blocks for Windows, since SVN rev 11701, is generated to support DIRECT2D activated/forced (for faster and better quality graphics). wxWidgets generation needs some adjustments to be done directly in setup.h file(s). Normally the only file which needs to be modified is in the lib subdirectory, the one created the first time wxWidgets is generated, and used in following generations. For a gcc, dll build, it’s something like : C:\wxWidgets-3.1.3\lib\gcc_dll\mswu\wx.
You can simply modify the line containing (line 1593 in wxWidgets 3.1.3):

    #define wxUSE_GRAPHICS_DIRECT2D 0

by

    #define wxUSE_GRAPHICS_DIRECT2D 1

Other setup.h may be modified, but it’s normally not required.

optionally :

USE_OPENGL=1
DEBUG_LEVEL=0

Note:
USE_OPENGL=1 creates an additional dll, used if your wxWidgets program requires an OpenGL windows, particularly when wxGLCanvas is used

It is also possible to personalise the name of the produced dll:

VENDOR=cb

or for a 64 bits build :

VENDOR=cb_64

Note:
If VENDOR is not specified, it’s equivalent to VENDOR=custom. VENDOR=cb is used by Code::Blocks team. So, to avoid confusion, it may be advisable to use an other ”vendor” identifier.

This is going to take some time.

For making the debug build, follow these same steps and options (the only change is BUILD=debug):

Eventually, add the same options as above.

Well have a little look in the directory (<wxDir>\lib\gcc_dll) now. The import libraries and the dll have shown up and there should also a mswu\wx subdirectory at that position containing setup.h.

You can use strip.exe (distributed with MinGW) to reduce the size of the dlls. for ex:

strip ..\..\lib\gcc_dll\wxmsw313u_gcc_cb_64.dll
strip ..\..\lib\gcc_dll\wxmsw313u_gl_gcc_cb_64.dll

Congratulations! You have just built wxWidgets!

Let’s do some more preliminary tasks before we get to the real deal of compiling Code::Blocks.

6.2.4  Zip

During the build of Code::Blocks, several resources are going to be zipped in zip files. Therefore, the build process should have access to a zip.exe. We have to download that zip.exe (if it’s not already in your MinGW distribution) and put it somewhere in our path. A good place is: MingW\bin.

You can download zip.exe for free from this site (http://www.info-zip.org/pub/infozip/Zip.html), or frome here (ftp://ftp.info-zip.org/pub/infozip/win32/) which is a direct ftp link to the directory containing the most recent versions.

Once downloaded, simply extract zip.exe to the appropriate location.

6.2.5  Workspace

This brings us to the last preliminary task. The Code::Blocks code can be divided into 2 major parts: the core with internal plug-ins, and the contributed plug-ins. You always need to build the core/internal parts before building the contrib part.

To build the internal part, you can use the Code::Blocks project file which you can find at: <cbDir>\src\CodeBlocks.cbp. Our Code::Blocks master directory is from now one mentioned as <cbDir>, by the way. A workspace is something that groups several projects together. To build the contrib plug-ins, they can be found at <cbDir>\src\ContribPlugins.workspace.

But, let’s create a workspace containing everything. Let’s put that workspace in the master directory <cbDir>. Just use a regular text editor and create a file with the name ”CbProjects.workspace”. Such a file already exists in recent C::B versions. You can give it the following content :

<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<CodeBlocks_workspace_file>
  <Workspace title="CodeBlocks Workspace wx3.1.x">
    <Project filename="CodeBlocks_wx31_64.cbp" active="1" />
    <Project filename="tools/Addr2LineUI/Addr2LineUI_wx31.cbp" />
    <Project filename="tools/cb_share_config/cb_share_config_wx31.cbp" />
    <Project filename="tools/CBLauncher/CbLauncher_wx31.cbp" />
    <Project filename="tools/cbp2make/cbp2make_wx31.cbp" />
    <Project filename="plugins/codecompletion/cctest_wx31.cbp" />
    <Project filename="plugins/contrib/wxContribItems/wxContribItems_wx31.cbp" />
    <Project filename="plugins/contrib/wxSmith/wxSmith_wx31.cbp" />
    <Project filename="plugins/contrib/wxSmithContribItems/wxSmithContribItems_wx31.cbp">
      <Depends filename="plugins/contrib/wxContribItems/wxContribItems_wx31.cbp" />
      <Depends filename="plugins/contrib/wxSmith/wxSmith_wx31.cbp" />
    </Project>
    <Project filename="plugins/contrib/wxSmithAui/wxSmithAui_wx31.cbp">
      <Depends filename="plugins/contrib/wxSmith/wxSmith_wx31.cbp" />
    </Project>
    <Project filename="plugins/contrib/AutoVersioning/AutoVersioning_wx31.cbp" />
    <Project filename="plugins/contrib/BrowseTracker/BrowseTracker_wx31.cbp" />
    <Project filename="plugins/contrib/byogames/byogames_wx31.cbp" />
    <Project filename="plugins/contrib/cb_koders/cb_koders_wx31.cbp" />
    <Project filename="plugins/contrib/Cccc/Cccc_wx31.cbp" />
    <Project filename="plugins/contrib/codesnippets/codesnippets_wx31.cbp" />
    <Project filename="plugins/contrib/codestat/codestat_wx31.cbp" />
    <Project filename="plugins/contrib/copystrings/copystrings_wx31.cbp" />
    <Project filename="plugins/contrib/CppCheck/CppCheck_wx31.cbp" />
    <Project filename="plugins/contrib/Cscope/Cscope_wx31.cbp" />
    <Project filename="plugins/contrib/devpak_plugin/DevPakPlugin_wx31.cbp" />
    <Project filename="plugins/contrib/DoxyBlocks/DoxyBlocks_wx31.cbp" />
    <Project filename="plugins/contrib/dragscroll/DragScroll_wx31.cbp" />
    <Project filename="plugins/contrib/EditorConfig/EditorConfig_wx31.cbp" />
    <Project filename="plugins/contrib/EditorTweaks/EditorTweaks_wx31.cbp" />
    <Project filename="plugins/contrib/envvars/envvars_wx31.cbp" />
    <Project filename="plugins/contrib/FileManager/FileManager_wx31.cbp" />
    <Project filename="plugins/contrib/FortranProject/FortranProject_cbsvn_wx31.cbp" />
    <Project filename="plugins/contrib/headerfixup/headerfixup_wx31.cbp" />
    <Project filename="plugins/contrib/help_plugin/help-plugin_wx31.cbp" />
    <Project filename="plugins/contrib/HexEditor/HexEditor_wx31.cbp" />
    <Project filename="plugins/contrib/IncrementalSearch/IncrementalSearch_wx31.cbp" />
    <Project filename="plugins/contrib/keybinder/keybinder_wx31.cbp" />
                                                                                       
                                                                                       
    <Project filename="plugins/contrib/lib_finder/lib_finder_wx31.cbp">
      <Depends filename="plugins/contrib/wxContribItems/wxContribItems_wx31.cbp" />
    </Project>
    <Project filename="plugins/contrib/MouseSap/MouseSap_wx31.cbp" />
    <Project filename="plugins/contrib/NassiShneiderman/NassiShneiderman_wx31.cbp" />
    <Project filename="plugins/contrib/profiler/cbprofiler_wx31.cbp" />
    <Project filename="plugins/contrib/ProjectOptionsManipulator/ProjectOptionsManipulator_wx31.cbp" />
    <Project filename="plugins/contrib/regex_testbed/RegExTestbed_wx31.cbp" />
    <Project filename="plugins/contrib/ReopenEditor/ReopenEditor_wx31.cbp" />
    <Project filename="plugins/contrib/rndgen/rndgen_wx31.cbp" />
    <Project filename="plugins/contrib/SmartIndent/SmartIndent_wx31.cbp" />
    <Project filename="plugins/contrib/source_exporter/Exporter_wx31.cbp" />
    <Project filename="plugins/contrib/SpellChecker/SpellChecker_wx31.cbp" />
    <Project filename="plugins/contrib/symtab/symtab_wx31.cbp" />
    <Project filename="plugins/contrib/ThreadSearch/ThreadSearch_wx31.cbp">
      <Depends filename="plugins/contrib/wxContribItems/wxContribItems_wx31.cbp" />
    </Project>
    <Project filename="plugins/contrib/ToolsPlus/ToolsPlus_wx31.cbp" />
  </Workspace>
</CodeBlocks_workspace_file>

Note:
Several variants of this file exists, depending on the OS, wxWidgets version, 32 or 64 bits build.

We will use this workspace to build all of Code::Blocks.

6.2.6  Building Codeblocks

This section describes the final process of building Code::Blocks.

6.2.6.1 Windows

Finally we have arrived at the final step; our final goal. Run the Code::Blocks executable from your nightly build download. Choose Open from the File menu and browse for our above created workspace, and open it up. Be a little patient while Code::Blocks is parsing everything, and Code::Blocks will ask us for 3 or 4 global variables, these global variables will tell the nightly Code::Blocks where it can find wxWidgets (remember : header files and import libraries) and where it can find ... Code::Blocks sources, this is needed for the contrib plug-ins, they need to know (as for any user created plug-in) where the sdk (Code::Blocks header files) are. These are the values in our case :

wx
<wxDir> base directory of wxWidgets.: the variable name may be wx31, wx31_64, ...
cb
<cbDir>/src Code::Blocks directory contaning the sources.
cb_release_type
: -O (for a release version, usual case. For developpers you can put -g to debug C::B)
boost
: main directory where boost is installed (ex: C:\boost). Used by the NassiShneiderman plugin, fill with the same value the subsections include and lib

Now go to the Project Menu and choose (re)build workspace, and off you go. Watch how Code::Blocks is building Code::Blocks.

After the creation of Code::Blocks, the generated files with the debug information can be found in the devel subdirectory. By calling, or executing in a console, the batch file update.bat from the source directory <cbDir>/src (or more specifically the version adapted to your specific generation, as for example update31_64.bat), the files are copied to the <cbDir>/src/output subdirectory (or the adapted version). In addition, it will strip out all debugging symbols. This step is very important - never ever forget it.

Now you can copy the wx dll in both that output and, optionally, the devel directory.

Then you can close Code::Blocks. That was the downloaded nightly remember?

Time to test it. In the output directory, start up the CodeBlocks.exe. If everything went well, you’ll have your very own home-built Code::Blocks running.

6.2.6.2 Linux

(Note: this section should be reviewed and completed. Does not seem to be totally updated)

linux
Starting update_revision.sh

With this function the SVN revision of the Nightly Builts is updated in the sources. The file can be found in the main directory of the Code::Blocks sources.

When generating under Linux, the following steps are necessary. In this example we assume that you are in the Code::Blocks source directory. Under Linux, the environment variable PKG_CONFIG_PATH must be set. The <prefix> directory has to contain the codeblocks.pc file.

PKG_CONFIG_PATH=$PKG_CONFIG_PATH:<prefix>

sh update_revsion.sh
./bootstrap
./configure --with-contrib=[all | plugin names separated with comma]
     --prefix=<install-dir>
make
make install (as root)

You can also build under Linux as under Windows, with a workspace file. Nevertheless, pkg_config and wx-config must be set correctly.

6.2.7  Generate plugins only

This section is for generating only plugins.

6.2.7.1 Windows

Configure the global variables via ’Settings’ ’Global Variables’ .

Variable cb

For the cb variable, set the base entry to the source directory of Code::Blocks.

<cbDir>/codeblocks/src

Variable wx

For the wx variable, set the base entry to the source directory of wx: e.g.

C:\wxWidgets-2.8.12

or to build with a more recent version, wx31 (or wx31_64)

C:\wxWidgets-3.1.4

In the Code::Blocks project, the project variable WX_SUFFIX is set to u. This means that, when generating Code::Blocks linking will be carried out against the *u_gcc_custom.dll library (by default). The official nightly Builts of Code::Blocks will be linked against gcc_cb.dll. In doing so, the layout is as follows.

gcc_<VENDOR>.dll

The <VENDOR> variable is set in the configuration file compiler.gcc or in the make command line as shown before. To ensure, that a distinction is possible between the officially generated Code::Blocks and those generated by yourself, the default setting VENDOR=custom should never be changed.

Afterwards create the workspace ContribPlugins.cbp via ’Project’ ’Build workspace’ . Then execute update.bat once more.

6.2.7.2 Linux

Configure the wx variable with the global variables.

base
/usr
include
/usr/include/wx-2.8 (or the version installed on your machine)
lib
/usr/lib

URL catalog

[`→7Z]

7z zip homepage.
http://www.7-zip.org

[`→ZIP]

Info-zip Ftp page.
ftp://ftp.info-zip.org/pub/infozip/win32/

[`→SourceForgeCB]

Codeblocks at Sourceforge.
https://sourceforge.net/p/codeblocks/code/log/

[`→SourceForgeCBSrc]

Trunk SVN C::B Source on Sourceforge.
https://sourceforge.net/p/codeblocks/code/HEAD/tree/trunk/

[`→FORUM]

Codeblocks forum.
http://forums.codeblocks.org/

[`→WIKI]

Codeblocks wiki.
http://wiki.codeblocks.org/index.php/Main_Page

[`→CODEBLOCKS]

Codeblocks homepage.
http://www.codeblocks.org/

[`→GCC]

GCC home page.
http://gcc.gnu.org/

[`→MinGW]

MinGW original page.
http://www.mingw.org/

[`→MinGW64]

MinGW64 files page.
https://sourceforge.net/projects/mingw-w64/files/

[`→MinGW64-Ray-Linn]

MinGW64 Ray Linn Personal page.
https://sourceforge.net/projects/mingw-w64/files/Multilib%20Toolchains%28Targetting%20Win32%20and%20Win64%29/ray_linn/gcc-9.x-with-ada/

[`→MinGW64-Equation]

MinGW64 Equation files page.
http://www.equation.com/servlet/equation.cmd?fa=fortran

[`→MinGW64-LHMouse]

MinGW64 LHMouse files page.
https://gcc-mcf.lhmouse.com/

[`→MinGW64-WinLibs]

MinGW64 WinLibs page.
http://winlibs.com/

[`→MSys2]

Msys2 home page.
http://www.msys2.org/

[`→MSys2-MinGW64-Toolchain]

Msys2 Toolchain for MinGW64.
https://packages.msys2.org/group/mingw-w64-x86_64-toolchain

[`→UnxUtils]

Unix Like Utilities files page.
https://sourceforge.net/projects/unxutils/

[`→GnuWin32]

An other Unix Like Utilities files page.
https://sourceforge.net/projects/gnuwin32/

[`→HIGHTEC]

HighTec homepage.
http://www.hightec-rt.com/

[`→TortoiseSVN]

TortoiseSVN homepage.
http://tortoisesvn.net/

[`→Subversion]

Subversion Apache homepage.
http://subversion.apache.org/

[`→Wxwidgets]

WxWidgets homepage.
http://www.wxwidgets.org/

[`→Wxcode]

WxCode homepage.
http://wxcode.sourceforge.net/

[`→Scripts]

Scripting commands.
http://wiki.codeblocks.org/index.php?title=Scripting_commands

[`→CBDoc]

C::B documentation on the wiki.
http://wiki.codeblocks.org/index.php/Category:Code::Blocks_Documentation

[`→UserDoc]

C::B User documentation on the wiki.
http://wiki.codeblocks.org/index.php/Category:User_Documentation

[`→NassiShneiderman]

NassiShneiderman WiKi page.
https://www.wikipedia.org/wiki/Nassi-Shneiderman_diagram

[`→Launchpad Old]

Old Translation page on Launchpad.
https://launchpad.net/codeblocks

[`→Launchpad New]

New Translation page on Launchpad.
https://launchpad.net/codeblocks-gd

[`→Translation]

Translation topic on CB Forum.
http://forums.codeblocks.org/index.php/topic,1022.0.html