ISO 8879:1986/Amd 1:1988
(Amendment)Information processing — Text and office systems — Standard Generalized Markup Language (SGML) — Amendment 1
Information processing — Text and office systems — Standard Generalized Markup Language (SGML) — Amendment 1
Enhances the technical content of ISO 8879. Changes are listed in order by page number and clause or sub-clause number. Clauses of concern are: 4, 7, 9, 10, 11, 12, and 15.
Traitement de l'information — Systèmes bureautiques — Langage normalisé de balisage généralisé (SGML) — Amendement 1
General Information
Relations
Standards Content (Sample)
INTERNATIONAL STANDARD IS0 8879:1986/AMENDMENT 1
Published 1988-07-01
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION- MEXAYHAPOAHAfi OPrAHM3ALWl no CTAH~APTM3A~Wl. ORGANISATION INTERNATIONALE DE NORMALISATION
Information processing - Text and office systems -
Standard Generalized Markup Language (SGML)
AMENDMENT 1
Traitemen t de I ‘information - S yst&mes bureautiques - Langage standard g&&ah34 de balisage (SGML J
AMENDEMENT I
Amendment 1 to International Standard IS0 8879:1986 was prepared by Technical Committee ISO/TC 97, information processing
systems.
1 Introduction
This amendment enhances the technical content of IS0 8879. The purpose of these enhancements is to
improve the expression of the design of SGML, not to change that design. Typographic errors and omissions
are also corrected, clarifying notes are provided for unclear areas, and the following ambiguities are resolved
(principal clauses affected are in parentheses):
-
Entities rnay be parsed with respect to the base document type as well as active document and link types
(4.2, 4.3, 7.1, 12.1.1)
-
Data attributes (notation parameters) have the same syntax as other attributes (10.5.5, 11.3, 11.4)
--
Link syntax allows element nesting (12.1.4, 12.2, 12.3) and general entities (9.4.4, 10.4.2, 12.1.4.1)
--
Potential error conditions may be reported (15.4)
-
Conditions under which start-tag omission is ambiguous are clarified (7.3.1.1)
--
SHORTTAG and DATATAG have priority over OMiT’TAG (7.3.2, 7.4.1.1)
--
Forms of mixed content in which treatment of separator characters could cause anomalous results are
clarified (7.6.1) and warned against (I 1.2.4)
-
A tokenized attribute value has no extraneous spaces (7.9.3, 7.9.4)
-
CDATA and SDATA (like NDATA) are data entities (4.75.1, 7.9.4.3, 10.5.3, 10.5.5)
--
General entity name attribute, like other tokenized attributes, has a list form (7.9.4, 11.3.3)
--
System declaration enhanced and clarified (15.6)
- General rules for delimiter recognition were modified to avoid conflict with existing exceptions (9.6, 9.6.1,
Figure 3, 9.6.2, 9.6.6, 10.5.1, 11.2.2)
2 Changes to IS0 88794986
Changes are listed in order by page number and clause or sub-clause number. Page numbers are prefixed
by the word “Page”; clause and sub-clause numbers are not prefixed. All notes are part of the corrected
text, not part of the instructions for correction. Text appearing affer a colon (:) is unquoted literal text to be
inserted or deleted, as indicated.
UDC 681.3.06: 655.53
Ref. No. IS0 8879:1986/Al:1988 (E)
Descriptors : data processing, text processing, information interchange, documentation, documents, logical structure, programming
(computers), artificial languages, programming languages.
0 International Organization for Standardization, 1988 l
Printed in Switzerland Price based on 15 pages
---------------------- Page: 1 ----------------------
IS0 8879:1986/A1:1988(E)
Page 4
2 Change “formatted” to: imaged
Page 5
3 After the em dash in the reference to IS0 646 inset-t: IS0
4.2 Replace definition with:
A document type that the system has identified as being active.
NOTE - An SGML entity is parsed with respect to its active document types, if any, or if not, with respect to its base
document type and any active link types.
4.3 Replace definition with:
A link process that the system has identified as being active.
\
4.8.1 Insert new definition after 4.8:
4.8.1 associated notation (name): A notation name associated with the subject of a markup
deciaration by its assor=iatecrJ notation nar,je parameter.
Page 8
4.67 Change “is its default” to: becomes its default
4.69 In the first sentence, change “link type definition” to: by a link process definition
4.72.1
Inset-t new definition after 4.72:
4.72.1 data attribute: An attribute of the data conforming to a particular data content notation.
NOTE - In most cases, the value of the data attributes must be known before the data can be interpreted in accor-
dance with the notation.
Page 9
4.75 Delete both occurrences of: non-SGML
4.75.1 Insert new definition after 4.75:
4.75.1 data entity: An entity that was declared to be data and therefore is not parsed when ref-
erenced.
NOTE -- There are three kinds: character data entity, specific character data entity, and non-SGML data entity.
Page 12
4.143 Change “is’! to: can be
Renumber (and reorder) these sub-clauses as follows: 4.156 becomes 4.158, 4.158 becomes
4.156
4.157, and 4.157 becomes 4.156.
4.159.1 Insert new definition after 4.159:
4.159.1 initial link set: The link set that is current at the start of the document instance.
Page 15
Change “is” to: can be
4.225
Page 17
4.267 In the definition, insert before the colon: a semantic error (such as a generic identifier that
does not identify an element type) or
4.267 In item b), change “group’s” to: token’s
4.267 Re-letter items e) and f), respectively, to f) and g), and insert new item e): an otherwise allow-
able omission of a tag that creates an ambiguity:
4.269 Replace period at end of note with: (see 5.1).
2 Changes to IS0 8879-1986
2
c
---------------------- Page: 2 ----------------------
IS0 8879:19861A1:1988(E)
Page 21
5.2 Insert new paragraph before last paragraph:
Successive instances of a syntactic token are deemed to be repetitions of a repeatable token,
where permissible, rather than instances of multiple tokens.
6.1 In [I], insert before “non-SGML”: character data entity 1 specific character data entity 1
6.2 In [2], move the first “s*, ” to immediately after the equals sign and delete the second.
6.2 In [3], delete: s*
“appropriate delimiter action is performed” to: ensuing markup is parsed
6.2 In item a), change
(see item d, below)
Page 22
6.2 After item c), add new item d):
Within markup, a character is tested to see if it is part of a delimiter string, a separator, deiim-
ited literal data, or a token (see 9.6.6).
6.3 Change heading to: Data Entities
6.3 Insert new productions before [S]:
[5.1] character data entity -- SGML character*, Ee
[5.2] specific character data entity
-
-
SGML character*, Ee
6.3 In the note, delete: non-SGML
An SGML entity is parsed with respect to its
7.1 Replace last sentence of first paragraph with:
base document type and any active link types unless one or more active document types are
identified, in which case it is parsed only with respect to them.
7.1 Add to last paragraph:
There can be an active document type only if “CONCUR YES” is specified on the SGML decla-
ration. The base document type can be active only if at least one other document type is
active.
document
NOTE -- Parsing with respect to the base document type alone is accomplished by not identifying an
type.
There can be an active link type for an entity if it has at least one link process definition. The
possibility of simultaneous active link types depends on the class of link type, as follows:
More than one simple link type can be active, in addition to any implicit and explicit
simple
links.
Only one can be active.
implicit
explicit More than one can be active only if they form a chain of processes in which the
source document type of the first is the base document type, the source document
type of the second is the result document type of the first, etc. The last link type
in the chain can be an implicit link.
base document element, other prolog*
7.2 In [IO], replace the expression with:
Replace “definitions” with: declarations
7.2.1
7.3.1 Replace first paragraph with:
A tag can be omitted only as provided in this sub-sub-clause, and only if the omission would
not create an ambiguity, and if “OMITTAG YES” is specified on the SGML declaration.
Page 23
7.3.1.1 Insert new paragraph after list:
\
It is ambiguous to omit the start-tag of an element whose content begins with a short reference
string whose mapping is changed by the element’s associated short reference map.
7.3.1.2 Delete: for a document element, or
7.3.1.2 Re-letter items a) and b), respectively, to b) and c), and insert new item a): by the end of an
SGML document entity or SGML subdocument entity;
3
2 Chanqes to IS0 88794986
c
---------------------- Page: 3 ----------------------
IS0 8879:1986/A1:1988(E)
7.3.2 In the first sentence, change “the end-tag can be omitted for” to: data can serve as the end-tag
of
7.3.2 Add new paragraph:
An element that has a data tag is not treated as having an omitted end-fag.
7.4.1 .I
In [IS], delete: document type specification, w,
Page 24
7.4.1 .I insert note after list:
NOTE -- A generic icW?tifier speciflicatiol7 is implied for an empty start-tag prior to determining whether any tags
were omitted before it.
7.4.1.3 In [ 18 J, delete: document type specification,
7.4.2 Change ” resolution of references” to: interpretation of attribute value literals
7.51 .I
In [21], delete: s*!
7.6 In [ 271, change “link type” to: link set
- _
Page 25
7.6.1 Add note after second paragraph (after the list):
NOTE - The determination that an RE is data is made during recognition of markup. Markup recognition, which
inciudes recognition of omitted tags, occurs before the determination of whether an RE can be ignored according to the
above rules. This sequence produces intuitive results in the normal case, where data can occur anywhere in the con-
tent of an element, as in:
(quote 1 #PGnATA)+f-
In content models where the occurrence of data is restricted, however, as in
(x, #PCDATA)
the situations described in itetns a) and b ) would be treated as errors during tn arkup recog nition. The use of such con-
tt-lerefo Ire not normall y recommended 11.24)
tent models is normally unneces sary and (see
7.7 In the last paragraph, delete: or “EXPLICIT YES”
7.7.1 Delete sub-sub-clause.
7.7.2 Delete sub-sub-clause.
Page 26
7.8 Insert after “valid”: in a generic identifier specification
7.8.1 .I After “specified for”, inset-t: the start-tag of
7.9.3 Add to paragraph:
NOTE --- Intet pretation of an &trihtrte W?/IE liter& occfjrs a3 though the attribute were character cl&a, regardless of
its actual declared value.
An attribute value other than characfer data is tokenixd by replacing a sequence of SPACE
characters with a single SPACE character and ignoring leading or trailing SPACE characters.
NOTE - Tokenization is performed Whout regard to the original literal: for example, whether CDATA or SDATA enti-
ties were used is irrelevant.
Page 27
Replace paraqraph with:
7.9.3.1
.
An attribute value specification can be an attribute value (that is, not an attribute value literal)
only if it contains nothing but name characters and either:
a) it occurs in an atfribute definition list; or
“SHORTTAG YES” is specified on the SGML deciaratior,.
W
7.9.4 In [35], change “general entity name” to: general entity name f general entity name list
c
2 Changes to IS0 88794986
4
---------------------- Page: 4 ----------------------
IS0 88?9:1986/A1:1988(E)
7.9.4 insert new production after [35]:
[35.1] general entity name list = name list
7.9.4 In [39], [40], [42], and [43], replace “s+ ” with: SPACE
7.9.4 Change “fourteen” to: fifteen
7.9.4.3
Replace sub-clause with:
The value of a general entity name attribute, and of each name in a general entity name list,
must be the name of a data entity or an SGML subdocument entity. Such a name must be
declared as though it occurred in a general entity reference with no name group specified (see
9.4.4.1).
In item a), change “for an” to: for a general entity name list,
7.9.4.5
7.9.4.5 Add new paragraph after last paragraph:
In a single attribute specificafion list, the total number of names in genera/ entity name and
general entity name list attribute values, whether defaulted or specified, cannot exceed the
“GRPCNT” quantity.
Page 28
9.1 Add new paragraph and note after last paragraph:
A reference to a non-SGML data entity, PI entity, or SGML subdocument entity, is prohibited in
A reference to a CDATA entity or SDATA entity is permitted.
replaceable character data.
- An effect of this sub-clause is to r .equire that an element or marked section declared replace-
NOTE
able character data must st art and end in the same entity.
9.3 Add note:
quantity but a character string, name or name therefore,
NOTE - An umber or number token is not a
cc 73
“01” and for example, are not equivalent.
1 Y
Page 29
9.4.3 In item “b” of the note, change “must (and other elements should)” to: , or of a marked sec-
tion that is ignored, must (and the content of other elements and marked sections should)
9.4.3 In item “c” of the note, replace everything up to the right parenthesis with: For other elements
and marked sections, the start-tag and end-tag (or marked section start and marked section end
9.4.4 In [59] and [SO], replace “document type specification” with: name group?
9.4.4 Replace paragraph with new text:
A name group can be specified only if “CONCUR YES”, “SIMPLE YES”, “IMPLICIT YES”, or
“EXPLICIT YES” is specified on the SGIIIIL declaration, and the reference does not occur within
a start-tag or prolog.
requirement does not proh ibit specifying a na/??e for a parameter entity reference within a
NOTE - This
instance.
marked section start in a document
If a name group is specified, the entity reference is ignored unless a name in the name group is
that of an active document type or link type.
9.4.4.1 Applicable Entity Declaration
used in the value of
N OTE - The following requirements apply to entity names general entity name and general enti-
ity references
name list attributes, a s well as to named ent
tY
Before an entity name can be used, it must be declared by an entity declaration in all applica-
ble DTDs, except that a general entity name that is undeclared in a particular DTD is treated as
though it had been declared for the default entity if that DTD has one.
The applicable DTDs depend on the context in which the entity name occurs, as follows:
2 Changes to IS0 88794986
---------------------- Page: 5 ----------------------
IS0 8879:1986/A1:1988(E)
a) Within a start-tag, all DTDs applicable to the tag.
NOTE - the base DTD or those of active document types named in the tag’s dowment type speci-
That is, either
fication.
b) Elsewhere in a document instance, the DTDs of active document types or the source DTDs
of active link types specified in a named entity reference’s name group, or, for any name
that is undeclared in such a DTD, or if no name group is specified, the base DTD.
NOTE - An effect of this requirement is that an entity declared in the base document type can be referenced in
an instance of any document type in which no entity with the same name was defined, and in which, for a general
entity, no default entity was defined.
C ) Within a DTD, the same DTD.
d ) Within a result attribute specification, the result DTD.
e Elsewhere in an LPD, the source DTD.
NOTE - An LPD is parsed as though entity declarations within it had occurred in its source DTD.
NOTES
ted by an
An effect of these requirements is that a defat Ilt entity declar ,ed in the base document type will be *eferen
declared general entity name in an instan ce of any document type that does not itself t lave a defaul entity.
2 If an entity declaration specifies a data content notation, the notation must be declared in the same DTD as the enti-
tY*
9.4.5 In the paragraph after the note, replace “name character” with: character that could occur
in the reference
Page 30
9.5 In the last paragraph before the notes, insert before the period: in the context in which the
replacement occurs.
9.6 Change the initial “A” to: Except as otherwise provided in this International Standard, a
9.6 Change Cc translation-reference character set” to: syntax-reference character set.
9.6
Add note and new paragraph:
- A named char acter reference can be used to enter the first character of a delimiter string or delirniter-in-
NOTE
first character.
context, but only the
All characters of a delimiter string or delimiter-in-context (see 9.6.2) must occur in the same
entity.
9.6.1 For “CON”, add note at end of Meaning:
NOTE --- Most delimiters will not be recocHzed when the content is character data or replaceable character data.
9.6.1 In “CXT” Meaning, change “OS” to: DSM
Page 31
Figure 3 Except for “DSC”, change all “OS” in the Mode column to: DSM
Figure 3 For “DSC”, add Constraint: ENT
Figure 3 For “ETAGO”, add an additional Mode: TAG
Figure 3 For “MINUS”, add Constraint: EX
Figure 3 For “MINUS”, change Description of Role to: Exclusion
Figure 3 For “PERO”, add an additional Mode: LIT
Figure 3 For “PLUS”, add Constraint: EX
9.6.1 In the list, change “OS” to “DSM” and insert new entry before it:
DS Recognized in a declaration subset.
9.6.1 In the Meaning of “DSM” (originally ‘DDS”), change “and” to: or
In the Meaning of “MD”, delete the parenthesized phrase.
9.6.1
9.6.1 Add note after list:
6 2 Changes to IS0 88794986
---------------------- Page: 6 ----------------------
IS0 8879:19861Al :I 988( E)
NOTE - Recognition modes can nest. For example, when a markup declaration begins, the recognition mode
becomes “MD”. If a GRPO occurs in the declaration, the mode will become “GRP”, and an MDC (for example) will not
be recognizable. When a GRPC occurs, the “‘RF”” mode ends and the recognition mode again becomes “MD”.
9.6.2 In the Meaning of “NMS”, insert before “is specified”: , “SIMPLE YES”, “IMPLlClT YES”, or
“EXPLICIT YES”,
9.6.2 Replace second paragraph with:
Other contextual constraints are:
9.6.2 Add to end of second list:
EX In “MD” mode, recognized only at the start of a delimiter-in-context in which it is followed
by grps; in “GRP” mode, no constraints.
ENT Recognized only in the same entity as ,the corresponding c3so.
Page 32
9.6.6 n the first paragraph, replace each occurrence of “name start character” with: name character
Page 34
10.1 .I In the second sentence of the first paragraph, change “intervening” to: surrounding or interven-
ing
Page 35
10.1.5 Add paragraph and note:
A generic idenfifier can be a valid associated element type whether or not it is specified as an
element fype in the document type definition.
makes it easier to use public definitions that name a large set of generic identifiers in con-
NOTE - This provision
ictive dot ument type that does not permit all of them.
junction with a more restr
10.1.6 Add sentence to 1st paragraph: The generated system identifier must be the same for all uses
of the external identifier, except in the case of a default entity, where it must be the same for
all references with a given undeclared entity name.
added to that of its public identifiel-
10.1.6.2 Replace the words “whether specified or generated” with:
component
10.1.7 Replace period at end of paragraph with: , except that such a sequence is ignored if it occurs
at the start or end of the minimum literal.
Page 36
10.2.2.2 Insert new production after [87]:
[87.1] IS0 text description r minimum data
10.2.2.3 Move “two-character” before “language code”.
10.2.2.3 In the second sentence, change “must” to: should
“minimum literal” to: minimum data
10.2.2.4 In [89] and in the paragraph that follows it, change
10.2.2.4 Change both occurrences of “must” to: should
10.2.2.4 Add to note I:
The full set of 128 characters, which is not registered, should be designated by:
ESC 2/S 4/O
Page 37
In note 2, change “it uniquely identifies” to: it should uniquely identify
10.2.2.4
In [97], replace “qualified status keyword I” with a left parenthesis and insert a second right
lOv4.2
parenthesis before the first asterisk.
10.4.2 Delete productions [98] and [99].
7
2 Changes to IS0 88794986
---------------------- Page: 7 ----------------------
IS0 8879:1986/A1:1988(E)
10.4.2 In the definition for “IGNORE”, replace “marked section declaration” with “marked section start
or marked section end” and delete: , but its status keyword specification is ignored
10.4.2 Delete the two paragraphs that begin with “A qualified status keyword”.
Page 38
10.4.2 Replace the first paragraph beginning “If the effective” with the following paragraph and note:
If the effective status is “CDATA” or “RCDATA”, a marked section declaration is terminated by
the first marked section end.
NOTE - A nested marked secti declaration with in the CDATA or RCDATA marked section in the same entity could
be recognized.
not occur, as the markup would not
IO.51 Insert paragraph aft
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.