Keyword that starts an entity declaration.

The name of the general entity being declared.

<!ENTITY ent ...>

Keyword indicating that the declaration is for the “default” entity, which will be used whenever an entity is referenced that has not been explicited declared.

<!ENTITY #DEFAULT ...>

A quoted string containing data and possibly SGML markup that serves as the replacement string for this entity when it is referenced. If it contains markup, the markup should be balanced; that is, it should close any elements it opens.

<!ENTITY ent "text">

The ISO name for this is an internal entity specification.

Keyword indicating that the entity's replacement string is to be treated as CDATA, that is, that entity references and element markup won't be recognized. The ISO name for this is an internal entity specification with data text.

<!ENTITY ent  CDATA "chars">

Keyword indicating that the entity's replacement string is “system-specific” data containing instructions that are nonportable across computer systems. Typically, SDATA entities are used for special characters and symbols. This is an internal entity specification with data text.

<!ENTITY ent SDATA "instrucs">

Keyword indicating that the entity's replacement string should be treated as a processing instruction. The ISO name for this is an internal entity specification with data text.

<!ENTITY ent PI "instrucs">

Keyword indicating that the entity's replacement string should be interpreted as having start-tag delimiters around it, so that it can function as start-tag markup: <literal>.

<!ENTITY ent STARTTAG "elemname att="value">

Using the STARTTAG keyword rather than including the delimiters directly in the replacement string is helpful for protecting against changes to the concrete syntax in which SGML markup is expressed.

The ISO name for this is an internal entity specification with bracketed text.

Keyword indicating that the entity's replacement string should be interpreted as having end-tag delimiters around it, so that it can function as end-tag markup: </literal>.

<!ENTITY ent ENDTAG "elemname">

Using the ENDTAG keyword rather than including the delimiters directly in the replacement string is helpful for protecting against changes to the concrete syntax in which SGML markup is expressed.

The ISO name for this is an internal entity specification with bracketed text.

Keyword indicating that the entity's replacement string should be interpreted as having marked section delimiters around it, so that it can function as a marked section region: <![literal]]> .

<!ENTITY ent MS "IGNORE [ignored-stuff">

Using the MS keyword rather than including the delimiters directly in the replacement string may not help protect against changes to the concrete syntax in which SGML markup is expressed, because the literal must actually contain the inner opening square bracket (or alternate character, if you have redefined the concrete syntax).

The ISO name for this is an internal entity specification with bracketed text.

Keyword indicating that the entity's replacement string should be interpreted as having markup declaration delimiters around it, so that it can function as a markup declaration: <!literal> .

<!ENTITY ent MD "--cmt--">

Using the MD keyword rather than including the delimiters directly in the replacement string is helpful for protecting against changes to the concrete syntax in which SGML markup is expressed.

The ISO name for this is an internal entity specification with bracketed text.

Keyword indicating that the entity's contents can be found on a computer system, stored separately from the entity declaration.

<!ENTITY ent SYSTEM>

A file name, pathname, or other specification for how to locate the entity content on the computer system.

<!ENTITY ent SYSTEM "file.sgm">

Keyword indicating that the entity's contents can be found on a computer system and that the contents can be located by mapping the supplied public identifier to location instructions that are stored outside the SGML document. An entity with a public identifer can also optionally have system location instructions stored with it.

<!ENTITY ent PUBLIC "logical-name">

Symbolic identifier that maps to location instructions. The mapping is done by an entity manager, often through the use of a catalog file that lists public IDs and their corresponding physical locations.

If the FORMAL feature is set to YES in the SGML declaration, the public identifier must be constructed according to certain syntactic rules, described in Section A.10, “Formal Public Identifiers and Catalogs”.

Keyword indicating that the entity contents are considered to be a separate SGML document, with its own DOCTYPE declaration pointing to a potentially different DTD.

<!ENTITY ent SYSTEM SUBDOC>

The SUBDOC keyword can be used only if the SUBDOC feature in the SGML declaration is set to YES.

If an entity type of SUBDOC, CDATA, NDATA, or SDATA is not specified, the entity is assumed to contain SGML text and markup.

Keyword indicating that the entity contents are to be treated as CDATA, that is, that entity references and element markup won't be recognized. The notation must have been declared in a NOTATION declaration.

<!ENTITY ent SYSTEM CDATA notn>

If an entity type of SUBDOC, CDATA, NDATA, or SDATA is not specified, the entity is assumed to contain SGML text and markup.

Keyword and value indicating that the entity contents are to be treated as non-SGML data of the specified notation. The notation must have been declared in a NOTATION declaration.

<!ENTITY ent SYSTEM NDATA notn>

If an entity type of SUBDOC, CDATA, NDATA, or SDATA is not specified, the entity is assumed to contain SGML text and markup.

Keyword indicating that the entity contents consist of data that is “system-specific”, that is, nonportable across computer systems. The notation must have been declared in a NOTATION declaration.

<!ENTITY ent SYSTEM SDATA notn>

If an entity type of SUBDOC, CDATA, NDATA, or SDATA is not specified, the entity is assumed to contain SGML text and markup.

Keyword that starts an entity declaration.

Key letter that indicates that the declaration is for a parameter entity, an entity that can be referenced only within markup declarations. The percent sign must be surrounded by spaces.

The name of the parameter entity.

<!ENTITY % ent ...>

A quoted string containing replacement data for the parameter entity.

<!ENTITY % ent "string">

Keyword indicating that the entity's contents can be found on a computer system, stored separately from the entity declaration.

<!ENTITY % ent SYSTEM>

A file name, pathname, or other specification for how to locate the entity contents on the computer system.

<!ENTITY % ent SYSTEM "file.dtd">

Keyword indicating that the entity's contents can be found on a computer system and that the contents can be located by mapping the supplied public identifier to location instructions that are stored outside the SGML document. An entity with a public identifer can also optionally have system location instructions stored with it.

<!ENTITY % ent PUBLIC "logical-name">

Symbolic identifier that maps to location instructions. The mapping is done by an entity manager, often through the use of a catalog file that lists public IDs and their corresponding physical locations.

If the FORMAL feature is set to YES in the SGML declaration, the public identifier must be constructed according to certain syntactic rules, described in Section A.10, “Formal Public Identifiers and Catalogs”.

Keyword for the character set parameter. This parameter is meant for human consumption. Parsers have a “bootstrapping” problem in using the CHARSET parameter to parse SGML declarations; because an SGML declaration itself is stored electronically, parsers must use a “default” understanding of the character codes within in order to read the declaration in the first place.

Keyword introducing the public identifier of a well-known repertoire of logical characters to which to map the actual character codes found in the document. For example, “the ABC's” is a well-known character repertoire, and it might be commonly understood to be represented by the numbers 0 to 25.

Multiple BASESET/DESCSET pairs can appear here.

Symbolic identifier that maps to a description of a character repertoire and the numeric codes representing those characters. The mapping is done by an entity manager, often through the use of a catalog file that lists public IDs and their corresponding physical locations.

If the FORMAL feature is set to YES in the SGML declaration (see Section A.9.5, “Feature Use”), the public identifier must be constructed according to certain syntactic rules, described in Section A.10, “Formal Public Identifiers and Catalogs”.

Keyword introducing the “described character set,” a succession of three-field entries explaining how the numeric codes found in this document should be interpreted to map to the characters represented by the base character set provided above.

The simplest example might be a case where ISO 646 is both the base character set and the set actually used in the document:

DESCSET
0 128 0

Numeric character code that might be found in the document.

If the count field is greater than 1, this field serves as a starting point for a contiguous series of numeric codes corresponding to a contiguous series of the same length in the base character set.

The number of codes, counting upwards, in succession that map exactly between the described and base character sets.

If the number is 1, there is a simple mapping between the described code and the base code. If the number is greater than 1, a contiguous series of described codes corresponds to a contiguous series of the same length of base codes.

Numeric character code in the base character set that is associated with a logical character. For example, in ISO 646, the number 65 represents the letter A.

If the count field is greater than 1, this field serves as a starting point for a contiguous series of numeric codes corresponding to a contiguous series of the same length in the described character set.

Prose description of the logical character intended to be represented, for example, “happy face.” Such a description might need to be used if the character set used in the document needs characters not contained in the base character set.

Indicator that this numeric code should not be found in the document. Such a code is considered a non-SGML character and is not allowed to appear in the document.

Keyword for the capacity set parameter.

Keyword indicating that what follows is a public identifier for an externally stored collection of capacity settings.

Symbolic identifier that maps to a collection of capacity settings. The mapping is done by an entity manager, often through the use of a catalog file that lists public IDs and their corresponding physical locations.

If the FORMAL feature is set to YES in the SGML declaration (see Section A.9.5, “Feature Use”), the public identifier must be constructed according to certain syntactic rules, described in Section A.10, “Formal Public Identifiers and Catalogs”.

Keyword indicating that the reference capacity set should be used, except for any capacities provided below. The reference capacity set is explained in Table A.3, “Reference Capacity Set”.

Keyword indicating the capacity whose maximum is being set. The available capacity categories are listed in Table A.3, “Reference Capacity Set”.

The maximum number of computer storage points for the indicated capacity category that this SGML document needs. The default capacity points provided by the reference capacity set are listed in Table A.3, “Reference Capacity Set”.

Keyword for the concrete syntax scope parameter.

Keyword indicating that the concrete syntax defined in the SGML declaration is used everywhere in the document (except for the SGML declaration itself), including both the instance and the prolog (the markup declarations making up the one or more DTD and any link set declarations).

Keyword indicating that the concrete syntax defined in the SGML declaration is used only in the document instance, with the reference concrete syntax being used in the prolog. There are some restrictions on the variant concrete syntax if this keyword is specified, relating to ensuring that parsers can properly distinguish between the prolog and the instance and can handle both syntaxes.

Keyword for the concrete syntax parameter.

Keyword indicating that what follows is a public identifier for an externally stored description of a concrete syntax.

Symbolic identifier that maps to a description of a concrete syntax. The mapping is done by an entity manager, often through the use of a catalog file that lists public IDs and their corresponding physical locations.

If the FORMAL feature is set to YES in the SGML declaration (see Section A.9.5, “Feature Use”), the public identifier must be constructed according to certain syntactic rules, described in Section A.10, “Formal Public Identifiers and Catalogs”.

Specification of pairs of character numbers whose logical meanings should be switched. This parameter allows a concrete syntax identified by a public ID to be used with only slight changes. An example of characters that might need to be switched is the line feed and carriage return characters, numbers 10 and 13 in the character set used by the reference concrete syntax.

Indication of the characters that should be prohibited from being used in markup (as opposed to data). The NONE keyword means that no characters are shunned. The CONTROLS keyword means that all “control characters” (as opposed to graphic characters) should be shunned. Any individual character numbers specified will be shunned as well.

Specification of the “syntax-reference character set,” the mapping of numeric character codes used in markup to logical character meanings for the purpose of this subparameter. Its presence allows all specifications of character numbers and numeric character entity references in the SYNTAX parameter to be independent of the character set specified for data content in CHARSET.

See Figure A.12, “CHARSET Parameter Syntax” for a description of how to supply BASESET/CHARSET information.

Specification of the character numbers that correspond to special function character roles in SGML markup. RE identifies the record end character, RS identifies the record start character, and SPACE identifies the space character. Additional character numbers can be named and assigned special function roles as follows:

Declaration of the characters that can be used in markup names (such as generic identifiers and entity names). LCNMSTRT and LCNMCHAR specify, respectively, additional lowercase characters that can be in the first position and subsequent positions of a name. UCNMSTRT and UCNMCHAR specify, respectively, the corresponding additional uppercase characters that can be in the first position and the subsequent positions of a name. If a “lowercase” character specified is a symbol that has no case, it should be specified in the same position in the character list for both the LC and UC fields. Often, an underscore ( _ ) is added to subsequent positions.

If no additional characters are specified, by dfeault the first character must be an uppercase or lowercase letter, and the subsequent characters can be a letter or a digit. The reference concrete syntax adds a period and a hyphen to the subsequent characters. The characters specified here are not raw character numbers; they are either real typed characters or numeric character references.

Setup of the case sensitivity of SGML markup names. GENERAL sets the case sensitivity of all names except for entity names, and ENTITY sets the case sensitivity of entity names. The reference concrete syntax specifies that GENERAL is NO (case insensitive) and ENTITY is YES (case sensitive).

Replaces the markup delimiters in the reference delimiter set with alternate delimiters. The SGMLREF keyword represents the specification of the reference delimiter set; the delimiter role keywords and their reference delimiters are shown in Table A.4, “Reference General Delimiter Set”.

Replaces the short reference markup delimiters in the reference delimiter set with alternate delimiters. The SGMLREF keyword represents the specification of the reference delimiter set (the reference delimiters for short references are not shown here). The NONE keyword means that no short reference delimiters are enabled, except any that are explicitly defined in this SGML declaration.

Replaces the reserved name keywords in the reference set with alternate keywords. The SGMLREF keyword represents the specification of the reference set; the reference keywords are listed in Table A.5, “Reference Reserved Name Set”. Note that the reserved names from the SGML declaration itself are not listed here, as they must be used in their reference form.

Sets the basic lengths and measurements of various SGML markup characteristics. The SGMLREF keyword represents the specification of the reference quantity set; the quantity keywords and reference values are listed in Table A.6, “Reference Quantity Set”.