These define the fonts, colours, styles and keywords for languages that Programmer's Notepad (via Scintilla) has built-in support for. PN contains code that parses/lexes and understands a lot of languages already. These files surface and configure that support.

Document element for scheme.

Child elements:

• style-classes ?
• keyword-classes ?
• base-language *
• language +

Container element for style classes.

Child elements:

• style-class +

Purpose unknown, probably defines style settings to be inherited by other style elements. A scheme can override predefined styles which can be found in master.scheme. All styles inherit the default style class, default style class inherits system defaults.

Attributes:

• name. Identifier.
• description. String. User-friendly description.
• inherit-style. Identifier.
• Other visual style settings applicable to style element.

Container element for keyword classes.

Child elements:

• keyword-class +

Defines named keyword group.

Attributes:

• name. Identifier.

Child elements:

• include-class *

Text content:

• whitespace-separated list of keywords.

Adds a keyword class to this class.

Attributes:

• name. Identifier of keyword class being included.

Defines language settings to be inherited by language elements.

Defines language settings.

Attributes:

• name. Identifier.
• base. Identifier of the base language to inherit settings from. Optional.
• title. String. User-friendly language name to be presented to the user.
• folding. Bool. Optional.
• foldcomments. Bool. Optional.
• foldelse. Bool. Optional. Fold if and else blocks as two separate blocks.
• foldcompact. Bool. Optional. Include blank lines after the block being folded into the folding region.
• foldpreproc. Bool. Optional.
• wordchars. String. Optional. Example: wordchars=“abcdeABCDE0123456789_”.

Child elements:

• comments ?
• lexer ?
• property *
• use-keywords ?
• use-styles ?

Defines comment delimiters. This doesn't affect how the lexer highlights comments, it only affects PN functions of commenting and uncommenting blocks of code.

Attributes:

• line. String. Optional. For C++: line="//".
• streamStart. String. Optional. For C++: streamStart="/*".
• streamEnd. String. Optional. For C++: streamEnd="*/".
• blockStart. String. Optional. For C++: blockStart="/**".
• blockLine. String. Optional. For C++: blockLine=" *".
• blockEnd. String. Optional. For C++: blockEnd=" */".

Assigns lexer for the language.

Attributes:

• name. Identifier of lexer to use. For .scheme file this is an identifier of a built-in scintilla lexer. This affects, how the text can be highlighted, how many keyword lists and styles can be used.
• stylebits. String. Optional. Purpose unknown.

Key-value pair for lexer property.

Attributes:

• name. Identifier of lexer property.
• value. String. Property value to set.

Container element for keywords used by the language.

Child elements:

• keyword +

Defines binding between keyword class and keyword list index.

Attributes:

• name. String. User-fiendly name for the keyword class.
• key. Zero-based index of corresponding WordList as seen by lexer. Lexer can assign a special meaning to a particular word list, for example, C++ lexer uses word list 2 for doxygen keywords to be highlighted in documentation comments.
• class. Identifier. Name of keyword class to bind.

Container element for styles used by the language.

Child elements:

• style *
• group *

Defines style settings.

Attributes:

• name. String. User-fiendly name for the style.
• key. Number. Number of corresponding style as seen by lexer. Style keys are lexer-specific. For concrete values see SCE_ definitions in SciLexer.h for the lexer of interest. For their meaning see the lexer source (LexLLL.cxx file, ColouriseLLLDoc function). See also scintilla docs on style numbers.
• class. Identifier. Optional. Purpose unknown, probably name of default or custom style class to inherit settings from.
• fore. Color. Optional.
• back. Color. Optional.
• bold. Bool. Optional.
• italic. Bool. Optional.
• font. String. Optional. Font name.
• eolfilled. Bool. Optional. Determines whether the style's background will extend past end of line.

Groups styles. Purpose unknown, probably defines common style settings for enclosed styles.

Attributes:

• name. String. User-fiendly name for the style.
• description. String. Optional.
• class. Identifier. Optional. Purpose unknown, probably identifies a style class to inherit style settings from.

Child elements:

• style *

These define the syntax for a language as well as the fonts, colours… The files are interpreted by a plugin for PN called customscheme.dll which provides code to tell PN how to interpret text according to the rules in the .schemedef. Scintilla built-in lexers (used by .scheme files) usually highlight text according to hardcoded syntax rules, customscheme lexer supports simple configuration of the language syntax.

As PN scans the text in a file it assigns every character in the file a “style”, represented by a key. Unfortunately each different built-in language can and often does use the keys to mean different things. So in C++ style 5 is used for keywords whereas HTML uses style 5 for numbers. There is one definitive source of information mapping key numbers to meanings, and that's the .properties files shipped with the SciTE distribution - SciTE is the test editor developed alongside Scintilla.

One style number that is almost always unchanged is that 32 is the default text style - normally this means text with no specific meaning.

When using a .schemedef file, the style keys are fixed:

• 32: Default
• 1: Line Comment Style
• 2: Block Comment Style
• 3: First Identifiers Type
• 4: Numbers
• 5: Keywords
• 6: Keywords 2
• 7: Keywords 3
• 8: Keywords 4
• 9: Keywords 5
• 10: First String Type
• 11: Second String Type
• 12: Pre-processor
• 13: Second Identifiers Type

The limitation to keyword sets is defined by two hard limits: The first is that Scintilla supports a fixed maximum number of keyword sets, currently 8. The second is that each language lexer only uses the keyword sets that the author thought were necessary - this means that often a language will support only one or two sets of keywords. The .schemedef lexer supports 5 as when it was developed that was the Scintilla limit.

The changes a user makes to fonts and colours are not stored in the original .scheme or .schemedef files, the configuration of these things is instead stored in a file called UserSettings.xml in the per-user PN application data directory. Unless you're using the portable edition of PN, this will likely be here:

Vista: c:\users\[your username]\AppData\Roaming\Echo Software\PN2

XP: c:\documents and settings\[your username]\Application Data\Echo Software\PN2

The configuration in usersettings.xml is applied to the schemes you have installed, and the result is cached in a binary format in the .cscheme files you'll find in that directory.