[pspp] / spv-file-format.texi

@node SPSS Viewer File Format
@chapter SPSS Viewer File Format

SPSS Viewer or @file{.spv} files, here called SPV files, are written
by SPSS 16 and later to represent the contents of its output editor.
This chapter documents the format, based on examination of a corpus of
about 500 files from a variety of sources.  This description is
detailed enough to read SPV files, but probably not enough to write
them.

SPSS 15 and earlier versions use a completely different output format
based on the Microsoft Compound Document Format.  This format is not
documented here.

An SPV file is a Zip archive that can be read with @command{zipinfo}
and @command{unzip} and similar programs.  The final member in the Zip
archive is a file named @file{META-INF/MANIFEST.MF}.  This structure
makes SPV files resemble Java ``JAR'' files (and ODF files), but
whereas a JAR manifest contains a sequence of colon-delimited
key/value pairs, an SPV manifest contains the string
@samp{allowPivoting=true}, without a new-line.  (This string may be
the best way to identify an SPV file; it is invariant across the
corpus.)

The rest of the members in an SPV file's Zip archive fall into two
categories: @dfn{structure} and @dfn{detail} members.  Structure
member names begin with @file{outputViewer@var{nnnnnnnnnn}}, where
each @var{n} is a decimal digit, and end with @file{.xml}, and often
include the string @file{_heading} in between.  Each of these members
represents some kind of output item (a table, a heading, a block of
text, etc.) or a group of them.  The member whose output goes at the
beginning of the document is numbered 0, the next member in the output
is numbered 1, and so on.

Structure members contain XML.  This XML is sometimes self-contained,
but it often references detail members in the Zip archive, which are
named as follows:

@table @asis
@item @file{@var{prefix}_table.xml} and @file{@var{prefix}_tableData.bin}
@itemx @file{@var{prefix}_lightTableData.bin}
The structure of a table plus its data.  Older SPV files pair a
@file{@var{prefix}_table.xml} file that describes the table's
structure with a binary @file{@var{prefix}_tableData.bin} file that
gives its data.  Newer SPV files (the majority of those in the corpus)
instead include a single @file{@var{prefix}_lightTableData.bin} file
that incorporates both into a single binary format.

@item @file{@var{prefix}_warning.xml} and @file{@var{prefix}_warningData.bin}
@itemx @file{@var{prefix}_lightWarningData.bin}
Same format used for tables, with a different name.

@item @file{@var{prefix}_notes.xml} and @file{@var{prefix}_notesData.bin}
@itemx @file{@var{prefix}_lightNotesData.bin}
Same format used for tables, with a different name.

@item @file{@var{prefix}_chartData.bin} and @file{@var{prefix}_chart.xml}
The structure of a chart plus its data.  Charts do not have a
``light'' format.

@item @file{@var{prefix}_pmml.scf}
@itemx @file{@var{prefix}_stats.scf}
@item @file{@var{prefix}_model.xml}
Not yet investigated.  The corpus contains few examples.
@end table

The @file{@var{prefix}} in the names of the detail members is
typically an 11-digit decimal number that increases for each item,
tending to skip values.  Older SPV files use different naming
conventions.  Structure member refer to detail members by name, and so
their exact names do not matter to readers as long as they are unique.

@menu
* SPV Structure Member Format::
* SPV Light Detail Member Format::
* SPV Legacy Detail Member Binary Format::
* SPV Legacy Detail Member XML Format::
@end menu

@node SPV Structure Member Format
@section Structure Member Format

Structure members' XML files claim conformance with a collection of
XML Schemas.  These schemas are distributed, under a nonfree license,
with SPSS binaries.  Fortunately, the schemas are not necessary to
understand the structure members.  To a degree, the schemas can even
be deceptive because they document elements and attributes that are
not in the corpus and do not document elements and attributes that are
commonly found there.

Structure members use a different XML namespace for each schema, but
these namespaces are not entirely consistent.  In some SPV files, for
example, the @code{viewer-tree} schema is associated with namespace
@indicateurl{http://xml.spss.com/spss/viewer-tree} and in others with
@indicateurl{http://xml.spss.com/spss/viewer/viewer-tree} (note the
additional @file{viewer/}).  Under either name, the schema URIs are
not resolvable to obtain the schemas themselves.

One may ignore all of the above in interpreting a structure member.
The actual XML has a simple and straightforward form that does not
require a reader to take schemas or namespaces into account.

The elements found in structure members are documented below.  For
each element, we note the possible parent elements and the element's
contents.  The contents are specified as pseudo-regular expressions
with the following conventions:

@table @asis
@item text
XML text content.

@item CDATA
XML CDATA content.

@item @code{element}
The named element.

@item (@dots{})
Grouping multiple elements.

@item [@var{x}]
An optional @var{x}.

@item @var{a} @math{|} @var{b}
A choice between @var{a} and @var{b}.

@item @var{x}*
Zero or more @var{x}.
@end table

@ifnottex
For a diagram illustrating the hierarchy of elements within an SPV
structure member, please refer to a PDF version of the manual.
@end ifnottex

@iftex
The following diagram shows the hierarchy of elements within an SPV
structure member.  Edges point from parent to child elements.
Unlabeled edges indicate that the child appears exactly once; edges
labeled with *, zero or more times; edges labeled with ?, zero or one
times.
@center @image{dev/spv-structure, 5in}
@end iftex

@menu
* SPV Structure heading Element::
* SPV Structure label Element::
* SPV Structure container Element::
* SPV Structure text Element (Inside @code{container})::
* SPV Structure html Element::
* SPV Structure table Element::
* SPV Structure tableStructure Element::
* SPV Structure dataPath Element::
* SPV Structure pageSetup Element::
* SPV Structure pageHeader and pageFooter Elements::
* SPV Structure pageParagraph Element::
* SPV Structure @code{text} Element (Inside @code{pageParagraph})::
@end menu

@node SPV Structure heading Element
@subsection The @code{heading} Element

Parent: Document root or @code{heading} @*
Contents: @code{pageSetup}? @code{label} (@code{container} @math{|} @code{heading})*

The root of a structure member is a @code{heading}, which represents a
section of output beginning with a title (the @code{label}) and
ordinarily followed by content containers or further nested
(sub)-sections of output.

The document root heading, only, may also contain a @code{pageSetup}
element.

The following attributes have been observed on both document root and
nested @code{heading} elements.

@defvr {Optional} creator-version
The version of the software that created this SPV file.  A string of
the form @code{xxyyzzww} represents software version xx.yy.zz.ww,
e.g.@: @code{21000001} is version 21.0.0.1.  Trailing pairs of zeros
are sometimes omitted, so that @code{21}, @code{210000}, and
@code{21000000} are all version 21.0.0.0 (and the corpus contains all
three of those forms).
@end defvr

@noindent
The following attributes have been observed on document root
@code{heading} elements only:

@defvr {Optional} @code{creator}
The directory in the file system of the software that created this SPV
file.
@end defvr

@defvr {Optional} @code{creation-date-time}
The date and time at which the SPV file was written, in a
locale-specific format, e.g.@: @code{Friday, May 16, 2014 6:47:37 PM
PDT} or @code{lunedì 17 marzo 2014 3.15.48 CET} or even @code{Friday,
December 5, 2014 5:00:19 o'clock PM EST}.
@end defvr

@defvr {Optional} @code{lockReader}
Whether a reader should be allowed to edit the output.  The possible
values are @code{true} and @code{false}, but the corpus only contains
@code{false}.
@end defvr

@defvr {Optional} @code{schemaLocation}
This is actually an XML Namespace attribute.  A reader may ignore it.
@end defvr

@noindent
The following attributes have been observed only on nested
@code{heading} elements:

@defvr {Required} @code{commandName}
The locale-invariant name of the command that produced the output,
e.g.@: @code{Frequencies}, @code{T-Test}, @code{Non Par Corr}.
@end defvr

@defvr {Optional} @code{visibility}
To what degree the output represented by the element is visible.  The
only observed value is @code{collapsed}.
@end defvr

@defvr {Optional} @code{locale}
The locale used for output, in Windows format, which is similar to the
format used in Unix with the underscore replaced by a hyphen, e.g.@:
@code{en-US}, @code{en-GB}, @code{el-GR}, @code{sr-Cryl-RS}.
@end defvr

@defvr {Optional} @code{olang}
The output language, e.g.@: @code{en}, @code{it}, @code{es},
@code{de}, @code{pt-BR}.
@end defvr

@node SPV Structure label Element
@subsection The @code{label} Element

Parent: @code{heading} or @code{container} @*
Contents: text

Every @code{heading} and @code{container} holds a @code{label} as its
first child.  The root @code{heading} in a structure member always
contains the string ``Output''.  Otherwise, the text in @code{label}
describes what it labels, often by naming the statistical procedure
that was executed, e.g.@: ``Frequencies'' or ``T-Test''.  Labels are
often very generic, especially within a @code{container}, e.g.@:
``Title'' or ``Warnings'' or ``Notes''.  Label text is localized
according to the output language, e.g.@: in Italian a frequency table
procedure is labeled ``Frequenze''.

The corpus contains one example of an empty label, one that contains
no text.

This element has no attributes.

@node SPV Structure container Element
@subsection The @code{container} Element

Parent: @code{heading} @*
Contents: @code{label} (@code{table} @math{|} @code{text})?

A @code{container} serves to label a @code{table} or a @code{text}
item.

This element has the following attributes.

@defvr {Required} @code{visibility}
Either @code{visible} or @code{hidden}, this indicates whether the
container's content is displayed.
@end defvr

@defvr {Optional} @code{text-align}
Presumably indicates the alignment of text within the container.  The
only observed value is @code{left}.  Observed with nested @code{table}
and @code{text} elements.
@end defvr

@defvr {Optional} @code{width}
The width of the container in the form @code{@var{n}px}, e.g.@:
@code{1097px}.
@end defvr

@node SPV Structure text Element (Inside @code{container})
@subsection The @code{text} Element (Inside @code{container})

Parent: @code{container} @*
Contents: @code{html}

This @code{text} element is nested inside a @code{container}.  There
is a different @code{text} element that is nested inside a
@code{pageParagraph}.

This element has the following attributes.

@defvr {Required} @code{type}
One of @code{title}, @code{log}, or @code{text}.
@end defvr

@defvr {Optional} @code{commandName}
As on the @code{heading} element.  For output not specific to a
command, this is simply @code{log}.  The corpus contains one example
of where @code{commandName} is present but set to the empty string.
@end defvr

@defvr {Optional} @code{creator-version}
As on the @code{heading} element.
@end defvr

@node SPV Structure html Element
@subsection The @code{html} Element

Parent: @code{text} @*
Contents: CDATA

The CDATA contains an HTML document.  In some cases, the document
starts with @code{<html>} and ends with @code{</html}; in others the
@code{html} element is implied.  Generally the HTML includes a
@code{head} element with a CSS stylesheet.  The HTML body often begins
with @code{<BR>}.  The actual content ranges from trivial to simple:
just discarding the CSS and tags yields readable results.

This element has the following attributes.

@defvr {Required} @code{lang}
This always contains @code{en} in the corpus.
@end defvr

@node SPV Structure table Element
@subsection The @code{table} Element

Parent: @code{container} @*
Contents: @code{tableStructure}

This element has the following attributes.

@defvr {Required} @code{commandName}
As on the @code{heading} element.
@end defvr

@defvr {Required} @code{type}
One of @code{table}, @code{note}, or @code{warning}.
@end defvr

@defvr {Required} @code{subType}
The locale-invariant name for the particular kind of output that this
table represents in the procedure.  This can be the same as
@code{commandName} e.g.@: @code{Frequencies}, or different, e.g.@:
@code{Case Processing Summary}.  Generic subtypes @code{Notes} and
@code{Warnings} are often used.
@end defvr

@defvr {Required} @code{tableId}
A number that uniquely identifies the table within the SPV file,
typically a large negative number such as @code{-4147135649387905023}.
@end defvr

@defvr {Optional} @code{creator-version}
As on the @code{heading} element.  In the corpus, this is only present
for version 21 and up and always includes all 8 digits.
@end defvr

@node SPV Structure tableStructure Element
@subsection The @code{tableStructure} Element

Parent: @code{table} @*
Contents: @code{dataPath}

This element has no attributes.

@node SPV Structure dataPath Element
@subsection The @code{dataPath} Element

Parent: @code{tableStructure} @*
Contents: text

Contains the name of the Zip member that holds the table details,
e.g.@: @code{0000000001437_lightTableData.bin}.

This element has no attributes.

@node SPV Structure pageSetup Element
@subsection The @code{pageSetup} Element

Parent: @code{heading} @*
Contents: @code{pageHeader} @code{pageFooter}

This element has the following attributes.

@defvr {Required} @code{initial-page-number}
Always @code{1}.
@end defvr

@defvr {Optional} @code{chart-size}
Always @code{as-is} or a localization (!) of it (e.g.@: @code{dimensione
attuale}, @code{Wie vorgegeben}).
@end defvr

@defvr {Optional} @code{margin-left}
@defvrx {Optional} @code{margin-right}
@defvrx {Optional} @code{margin-top}
@defvrx {Optional} @code{margin-bottom}
Margin sizes in the form @code{@var{size}in}, e.g.@: @code{0.25in}.
@end defvr

@defvr {Optional} @code{paper-height}
@defvrx {Optional} @code{paper-width}
Paper sizes in the form @code{@var{size}in}, e.g.@: @code{8.5in} by
@code{11in} for letter paper or @code{8.267in} by @code{11.692in} for
A4 paper.
@end defvr

@defvr {Optional} @code{reference-orientation}
Always @code{0deg}.
@end defvr

@defvr {Optional} @code{space-after}
Always @code{12pt}.
@end defvr

@node SPV Structure pageHeader and pageFooter Elements
@subsection The @code{pageHeader} and @code{pageFooter} Elements

Parent: @code{pageSetup} @*
Contents: @code{pageParagraph}*

This element has no attributes.

@node SPV Structure pageParagraph Element
@subsection The @code{pageParagraph} Element

Parent: @code{pageHeader} or @code{pageFooter} @*
Contents: @code{text}

Text to go at the top or bottom of a page, respectively.

This element has no attributes.

@node SPV Structure @code{text} Element (Inside @code{pageParagraph})
@subsection The @code{text} Element (Inside @code{pageParagraph})

Parent: @code{pageParagraph}
Contents: CDATA?

This @code{text} element is nested inside a @code{pageParagraph}.  There
is a different @code{text} element that is nested inside a
@code{container}.

The element is either empty, or contains CDATA that holds almost-XHTML
text: in the corpus, either an @code{html} or @code{p} element.  It is
@emph{almost}-XHTML because the @code{html} element designates the
default namespace as
@code{http://xml.spss.com/spss/viewer/viewer-tree} instead of an XHTML
namespace, and because the CDATA can contain substitution variables:
@code{&[Page]} for the page number and @code{&[PageTitle]} for the
page title.

Typical contents (indented for clarity):

@example
<html xmlns="http://xml.spss.com/spss/viewer/viewer-tree">
    <head></head>
    <body>
        <p style="text-align:right; margin-top: 0">Page &[Page]</p>
    </body>
</html>
@end example

This element has the following attributes.

@defvr {Required} @code{type}
Always @code{text}.
@end defvr

@node SPV Light Detail Member Format
@section Light Detail Member Format

This section describes the format of ``light'' detail @file{.bin}
members.  These members have a binary format which we describe here in
terms of a context-free grammar using the following conventions:

@table @asis
@item NonTerminal @result{} @dots{}
Nonterminals have CamelCaps names, and @result{} indicates a
production.  The right-hand side of a production is often broken
across multiple lines.  Break points are chosen for aesthetics only
and have no semantic significance.

@item 00, 01, @dots{}, ff.
A bytes with a fixed value, written as a pair of hexadecimal digits.

@item i0, i1, @dots{}, i9, i10, i11, @dots{}
@itemx b0, b1, @dots{}, b9, b10, b11, @dots{}
A 32-bit integer in little-endian or big-endian byte order,
respectively, with a fixed value, written in decimal, prefixed by
@samp{i}.

@item byte
A byte.

@item bool
A byte with value 0 or 1.

@item int16
@itemx be16
A 16-bit integer in little-endian or big-endian byte order,
respectively.

@item int
@itemx be32
A 32-bit integer in little-endian or big-endian byte order,
respectively.

@item int64
@itemx be64
A 64-bit integer in little-endian or big-endian byte order,
respectively.

@item double
A 64-bit IEEE floating-point number.

@item float
A 32-bit IEEE floating-point number.

@item string
@itemx bestring
A 32-bit integer, in little-endian or big-endian byte order,
respectively, followed by the specified number of bytes of character
data.  (The encoding is indicated by the Formats nonterminal.)

@item @var{x}?
@var{x} is optional, e.g.@: 00? is an optional zero byte.

@item @var{x}*@var{n}
@var{x} is repeated @var{n} times, e.g. byte*10 for ten arbitrary bytes.

@item @var{x}[@var{name}]
Gives @var{x} the specified @var{name}.  Names are used in textual
explanations.  They are also used, also bracketed, to indicate counts,
e.g.@: int[@t{n}] byte*[@t{n}] for a 32-bit integer followed by the
specified number of arbitrary bytes.

@item @var{a} @math{|} @var{b}
Either @var{a} or @var{b}.

@item (@var{x})
Parentheses are used for grouping to make precedence clear, especially
in the presence of @math{|}, e.g.@: in 00 (01 @math{|} 02 @math{|} 03)
00.

@item count(@var{x})
A 32-bit integer that indicates the number of bytes in @var{x},
followed by @var{x} itself.

@item v1(@var{x})
In a version 1 @file{.bin} member, @var{x}; in version 3, nothing.
(The @file{.bin} header indicates the version.)

@item v3(@var{x})
In a version 3 @file{.bin} member, @var{x}; in version 1, nothing.
@end table

Little-endian byte order is far more common in this format, but a few
pieces of the format use big-endian byte order.

A ``light'' detail member @file{.bin} consists of a number of sections
concatenated together, terminated by a byte 01:

@cartouche
@format
LightMember @result{}
    Header Title
    Caption Footnotes
    Fonts Borders PrintSettings TableSettings Formats
    Dimensions Data
    01
@end format
@end cartouche

The following sections go into more detail.

@menu
* SPV Light Member Header::
* SPV Light Member Title::
* SPV Light Member Caption::
* SPV Light Member Footnotes::
* SPV Light Member Fonts::
* SPV Light Member Borders::
* SPV Light Member Print Settings::
* SPV Light Member Table Settings::
* SPV Light Member Formats::
* SPV Light Member Dimensions::
* SPV Light Member Categories::
* SPV Light Member Data::
* SPV Light Member Value::
* SPV Light Member ValueMod::
@end menu

@node SPV Light Member Header
@subsection Header

An SPV light member begins with a 39-byte header:

@cartouche
@format
Header @result{}
    01 00
    (i1 @math{|} i3)[@t{version}]
    bool
    bool[@t{show-numeric-markers}]
    bool[@t{rotate-inner-column-labels}]
    bool[@t{rotate-outer-row-labels}]
    bool
    int
    int[@t{min-column-width}] int[@t{max-column-width}]
    int[@t{min-row-width}] int[@t{max-row-width}]
    int64[@t{table-id}]
@end format
@end cartouche

@code{version} is a version number that affects the interpretation of
some of the other data in the member.  We will refer to ``version 1''
and ``version 3'' later on and use v1(@dots{}) and v3(@dots{}) for
version-specific formatting (as described previously).

If @code{show-numeric-markers} is 1, footnote markers are shown as
numbers, starting from 1; otherwise, they are shown as letters,
starting from @samp{a}.

If @code{rotate-inner-column-labels} is 1, then column labels closest
to the data are rotated to be vertical; otherwise, they are shown
in the normal way.

If @code{rotate-outer-row-labels} is 1, then row labels farthest from
the data are rotated to be vertical; otherwise, they are shown in the
normal way.

@code{table-id} is a binary version of the @code{tableId} attribute in
the structure member that refers to the detail member.  For example,
if @code{tableId} is @code{-4122591256483201023}, then @code{table-id}
would be 0xc6c99d183b300001.

@code{min-column-width} is the minimum width that a column will be
assigned automatically.  @code{max-column-width} is the maximum width
that a column will be assigned to accommodate a long column label.
@code{min-row-width} and @code{max-row-width} are a similar range for
the width of row labels.  All of these measurements are in 1/96 inch
units.

The meaning of the other variable parts of the header is not known.

@node SPV Light Member Title
@subsection Title

@cartouche
@format
Title @result{}
    Value[@t{title1}] 01?
    Value[@t{c}] 01? 31
    Value[@t{title2}] 01?
@end format
@end cartouche

The Title, which follows the Header, specifies the pivot table's title
twice, as @code{title1} and @code{title2}.  In the corpus, they are
always the same.

Whereas the Value in @code{title1} and in @code{title2} are
appropriate for presentation, and localized to the user's language,
@code{c} is in English, sometimes less specific, and sometimes less
well formatted.  For example, for a frequency table, @code{title1} and
@code{title2} name the variable and @code{c} is simply ``Frequencies''.

@node SPV Light Member Caption
@subsection Caption

@cartouche
@format
Caption @result{} Caption1 Caption2
Caption1 @result{} 31 Value @math{|} 58
Caption2 @result{} 31 Value @math{|} 58
@end format
@end cartouche

The Caption, if present, is shown below the table.  Caption2 is
normally present.  Caption1 is only rarely nonempty; it might reflect
user editing of the caption.

@node SPV Light Member Footnotes
@subsection Footnotes

@cartouche
@format
Footnotes @result{} int[@t{n}] Footnote*[@t{n}]
Footnote @result{} Value[@t{text}] (58 @math{|} 31 Value[@t{marker}]) byte*4
@end format
@end cartouche

Each footnote has @code{text} and an optional customer @code{marker}
(such as @samp{*}).

@node SPV Light Member Fonts
@subsection Fonts

@cartouche
@format
Fonts @result{} 00 Font*8
Font @result{}
    byte[@t{index}] 31
    string[@t{typeface}] float[@t{size}] int[@t{style}] bool[@t{underline}]
    int[@t{halign}] int[@t{valign}]
    string[@t{fgcolor}] string[@t{bgcolor}]
    byte[@t{alternate}] string[@t{altfg}] string[@t{altbg}]
    v3(int[@t{left-margin}] int[@t{right-margin}] int[@t{top-margin}] int[@t{bottom-margin}])
@end format
@end cartouche

Each Font represents the font style for a different element, in the
following order: title, caption, footer, corner, column
labels, row labels, data, and layers.

@code{index} is the 1-based index of the Font, i.e. 1 for the first
Font, through 8 for the final Font.

@code{typeface} is the string name of the font.  In the corpus, this
is @code{SansSerif} in over 99% of instances and @code{Times New
Roman} in the rest.

@code{size} is the size of the font, in points.  The most common size
in the corpus is 12 points.

@code{style} is a bit mask.  Bit 0 (with value 1) is set for bold, bit
1 (with value 2) is set for italic.

@code{underline} is 1 if the font is underlined, 0 otherwise.

@code{halign} specifies horizontal alignment: 0 for center, 2 for
left, 4 for right, 61453 for decimal, 64173 for mixed.  Mixed
alignment varies according to type: string data is left-justified,
numbers and most other formats are right-justified.

@code{valign} specifies vertical alignment: 0 for center, 1 for top, 3
for bottom.

@code{fgcolor} and @code{bgcolor} are the foreground color and
background color, respectively.  In the corpus, these are always
@code{#000000} and @code{#ffffff}, respectively.

@code{alternate} is 01 if rows should alternate colors, 00 if all rows
should be the same color.  When @code{alternate} is 01, @code{altfg}
and @code{altbg} specify the colors for the alternate rows.

@code{left-margin}, @code{right-margin}, @code{top-margin}, and
@code{bottom-margin} are measured in multiples of 1/96 inch.

@node SPV Light Member Borders
@subsection Borders

@cartouche
@format
Borders @result{}
    b1[@t{endian}]
    be32[@t{n-borders}] Border*[@t{n-borders}]
    bool[@t{show-grid-lines}]
    00 00 00

Border @result{}
    be32[@t{border-type}]
    be32[@t{stroke-type}]
    be32[@t{color}]
@end format
@end cartouche

The Borders reflect how borders between regions are drawn.

The fixed value of @code{endian} can be used to validate the
endianness.

@code{show-grid-lines} is 1 to draw grid lines, otherwise 0.

Each Border describes one kind of border.  @code{n-borders} seems to
always be 19.  Each @code{border-type} appears once (although in an
unpredictable order) and correspond to the following borders:

@table @asis
@item 0
Title.
@item 1@dots{}4
Left, top, right, and bottom outer frame.
@item 5@dots{}8
Left, top, right, and bottom inner frame.
@item 9, 10
Left and top of data area.
@item 11, 12
Horizontal and vertical dimension rows.
@item 13, 14
Horizontal and vertical dimension columns.
@item 15, 16
Horizontal and vertical category rows.
@item 17, 18
Horizontal and vertical category columns.
@end table

@code{stroke-type} describes how a border is drawn, as one of:

@table @asis
@item 0
No line.
@item 1
Solid line.
@item 2
Dashed line.
@item 3
Thick line.
@item 4
Thin line.
@item 5
Double line.
@end table

@code{color} is an RGB color.  Bits 24--31 are alpha, bits 16--23 are
red, 8--15 are green, 0--7 are blue.  An alpha of 255 indicates an
opaque color, therefore opaque black is 0xff000000.

@node SPV Light Member Print Settings
@subsection Print Settings

@cartouche
@format
PrintSettings @result{}
    b1[@t{endian}]
    bool[@t{all-layers}]
    bool[@t{paginate-layers}]
    bool[@t{fit-width}]
    bool[@t{fit-length}]
    bool[@t{top-continuation}]
    bool[@t{bottom-continuation}]
    be32[@t{n-orphan-lines}]
    bestring[@t{continuation-string}]
@end format
@end cartouche

The PrintSettings reflect settings for printing.  The fixed value of
@code{endian} can be used to validate the endianness.

@code{all-layers} is 1 to print all layers, 0 to print only the
visible layers.

@code{paginate-layers} is 1 to print each layer at the start of a new
page, 0 otherwise.  (This setting is honored only @code{all-layers} is
1, since otherwise only one layer is printed.)

@code{fit-width} and @code{fit-length} control whether the table is
shrunk to fit within a page's width or length, respectively.

@code{n-orphan-lines} is the minimum number of rows or columns to put
in one part of a table that is broken across pages.

If @code{top-continuation} is 1, then @code{continuation-string} is
printed at the top of a page when a table is broken across pages for
printing; similarly for @code{bottom-continuation} and the bottom of a
page.  Usually, @code{continuation-string} is empty.

@node SPV Light Member Table Settings
@subsection Table Settings

@cartouche
@format
TableSettings @result{}
    be32[@t{endian}]
    be32
    be32[@t{current-layer}]
    bool[@t{omit-empty}]
    bool[@t{show-row-labels-in-corner}]
    bool[@t{show-alphabetic-markers}]
    bool[@t{footnote-marker-position}]
    v3(
      byte
      count(
        Breakpoints[@t{row-breaks}] Breakpoints[@t{column-breaks}]
        Keeps[@t{row-keeps}] Keeps[@t{column-keeps}]
        PointKeeps[@t{row-keeps}] PointKeeps[@t{column-keeps}]
      )
      bestring[@t{notes}]
      bestring[@t{table-look}]
      00...
    )

Breakpoints @result{} be32[@t{n-breaks}] be32*[@t{n-breaks}]

Keeps @result{} be32[@t{n-keeps}] Keep*@t{n-keeps}
Keep @result{} be32[@t{offset}] be[@t{n}]

PointKeeps @result{} be32[@t{n-point-keeps}] PointKeep*@t{n-point-keeps}
PointKeep @result{} be32[@t{offset}] be32 be32

@end format
@end cartouche

The TableSettings reflect display settings.  The fixed value of
@code{endian} can be used to validate the endianness.

@code{current-layer} is the displayed layer.

If @code{omit-empty} is 1, empty rows or columns (ones with nothing in
any cell) are hidden; otherwise, they are shown.

If @code{show-row-labels-in-corner} is 1, then row labels are shown in
the upper left corner; otherwise, they are shown nested.

If @code{show-alphabetic-markers} is 1, markers are shown as letters
(e.g. @samp{a}, @samp{b}, @samp{c}, @dots{}); otherwise, they are
shown as numbers starting from 1.

When @code{footnote-marker-position} is 1, footnote markers are shown
as superscripts, otherwise as subscripts.

The Breakpoints are rows or columns after which there is a page break;
for example, a row break of 1 requests a page break after the second
row.  Usually no breakpoints are specified, indicating that page
breaks should be selected automatically.

The Keeps are ranges of rows or columns to be kept together without a
page break; for example, a row Keep with @code{offset} 1 and @code{n}
10 requests that the 10 rows starting with the second row be kept
together.  Usually no Keeps are specified.

The PointKeeps seem to be generated automatically based on
user-specified Keeps.  They seems to indicate a conversion from rows
or columns to pixel or point offsets.

@code{notes} is a text string that contains user-specified notes.  It
is displayed when the user hovers the cursor over the table, like
``alt text'' on a webpage.  It is not printed.  It is usually empty.

@code{table-look} is the name of a SPSS ``TableLook'' table style,
such as ``Default'' or ``Academic''; it is often empty.

TableSettings ends with an arbitrary number of null bytes.

@node SPV Light Member Formats
@subsection Formats

@cartouche
@format
Formats @result{}
    int[@t{n-widths}] int*[@t{n-widths}]
    string[@t{encoding}]
    int[@t{current-layer}]
    bool[@t{digit-grouping}] bool[@t{leading-zero}] bool
    int[@t{epoch}]
    byte[@t{decimal}] byte[@t{grouping}]
    CustomCurrency
    count(
      v1(X0?)
      v3(count(X1 count(X2)) count(X3))

X0 @result{}
    byte*14
    string[@t{command}] string[@t{command-local}]
    string[@t{language}] string[@t{charset}] string[@t{locale}]
    bool 00 bool bool
    int[@t{epoch}]
    byte[@t{decimal}] byte[@t{grouping}]
    CustomCurrency
    byte[@t{missing}] bool

X1 @result{}
    byte*2
    byte[@t{lang}]
    byte[@t{variable-mode}]
    byte[@t{value-mode}]
    int*2
    00*17
    bool
    01
X2 @result{}
    int[@t{n-heights}] int*[@t{n-heights}]
    int[@t{n-style-map}] BlankMap*[@t{n-style-map}]
    int[@t{n-styles}] StylePair*[@t{n-styles}]
    count((i0 i0)?)
StyleMap @result{} int64[@t{cell-index}] int16[@t{style-index}]
X3 @result{}
    01 00 (03 @math{|} 04) 00 00 00
    string[@t{command}] string[@t{command-local}]
    string[@t{language}] string[@t{charset}] string[@t{locale}]
    bool 00 bool bool
    int[@t{epoch}]
    byte[@t{decimal}] byte[@t{grouping}]
    double[@t{small}] 01
    (string[@t{dataset}] string[@t{datafile}] i0 int[@t{date}] i0)?
    CustomCurrency
    byte[@t{missing}] bool (i2000000 i0)?

CustomCurrency @result{} int[@t{n-ccs}] string*[@t{n-ccs}]
@end format
@end cartouche

If @code{n-widths} is nonzero, then the accompanying integers are
column widths as manually adjusted by the user.  (Row heights are
computed automatically based on the widths.)

@code{encoding} is a character encoding, usually a Windows code page
such as @code{en_US.windows-1252} or @code{it_IT.windows-1252}.  The
rest of the character strings in the member use this encoding.  The
encoding string is itself encoded in US-ASCII.

@code{epoch} is the year that starts the epoch.  A 2-digit year is
interpreted as belonging to the 100 years beginning at the epoch.  The
default epoch year is 69 years prior to the current year; thus, in
2017 this field by default contains 1948.  In the corpus, @code{epoch}
ranges from 1943 to 1948, plus some contain -1.

@code{decimal} is the decimal point character.  The observed values
are @samp{.} and @samp{,}.

@code{grouping} is the grouping character.  Usually, it is @samp{,} if
@code{decimal} is @samp{.}, and vice versa.  Other observed values are
@samp{'} (apostrophe), @samp{ } (space), and zero (presumably
indicating that digits should not be grouped).

@code{command} describes the statistical procedure that generated the
output, in English.  It is not necessarily the literal syntax name of
the procedure: for example, NPAR TESTS becomes ``Nonparametric
Tests.''  @code{command-local} is the procedure's name, translated
into the output language; it is often empty and, when it is not,
sometimes the same as @code{command}.

@code{dataset} is the name of the dataset analyzed to produce the
output, e.g.@: @code{DataSet1}, and @code{datafile} the name of the
file it was read from, e.g.@: @file{C:\Users\foo\bar.sav}.  The latter
is sometimes the empty string.

@code{date} is a date, as seconds since the epoch, i.e.@: since
January 1, 1970.  Pivot tables within an SPV files often have dates a
few minutes apart, so this is probably a creation date for the tables
rather than for the file.

Sometimes @code{dataset}, @code{datafile}, and @code{date} are present
and other times they are absent.  The reader can distinguish by
assuming that they are present and then checking whether the
presumptive @code{dataset} contains a null byte (a valid string never
will).

@code{n-ccs} is observed as either 0 or 5.  When it is 5, the
following strings are CCA through CCE format strings.  @xref{Custom
Currency Formats,,, pspp, PSPP}.  Most commonly these are all
@code{-,,,} but other strings occur.

@code{missing} is the character used to indicate that a cell contains
a missing value.  It is always observed as @samp{.}.

@node SPV Light Member Dimensions
@subsection Dimensions

A pivot table presents multidimensional data.  A Dimension identifies
the categories associated with each dimension.

@cartouche
@format
Dimensions @result{} int[@t{n-dims}] Dimension*[@t{n-dims}]
Dimension @result{} Value[@t{name}] DimProperties int[@t{n-categories}] Category*[@t{n-categories}]
DimProperties @result{}
    byte[@t{d1}]
    (00 @math{|} 01 @math{|} 02)[@t{d2}]
    (i0 @math{|} i2)[@t{d3}]
    bool[@t{show-dim-label}]
    bool[@t{hide-all-labels}]
    01 int[@t{dim-index}]
@end format
@end cartouche

@code{name} is the name of the dimension, e.g. @code{Variables},
@code{Statistics}, or a variable name.

The meanings of @code{d1}, @code{d2}, and @code{d3} are unknown.
@code{d1} is usually 0 but many other values have been observed.

If @code{show-dim-label} is 01, the pivot table displays a label for
the dimension itself.  Because usually the group and category labels
are enough explanation, it is usually 00.

If @code{hide-all-labels} is 01, the pivot table omits all labels for
the dimension, including group and category labels.  It is usually 00.
When @code{hide-all-labels} is 01, @code{show-dim-label} is ignored.

@code{dim-index} is usually the 0-based index of the dimension, e.g.@:
0 for the first dimension, 1 for the second, and so on.  Sometimes it
is -1.  There is no visible difference.

@node SPV Light Member Categories
@subsection Categories

Categories are arranged in a tree.  Only the leaf nodes in the tree
are really categories; the others just serve as grouping constructs.

@cartouche
@format
Category @result{} Value[@t{name}] (Leaf @math{|} Group)
Leaf @result{} 00 00 00 i2 int[@t{cat-index}] i0
Group @result{}
    bool[@t{merge}] 00 01 (i0 @math{|} i2)[@t{data}]
    i-1 int[@t{n-subcategories}] Category*[@t{n-subcategories}]
@end format
@end cartouche

@code{name} is the name of the category (or group).

A Leaf represents a leaf category.  The Leaf's @code{cat-index} is a
nonnegative integer less than @code{n-categories} in the Dimension in
which the Category is nested (directly or indirectly).  These
categories represent the original order in which the categories were
sorted; if the user sorted or rearranged the categories, then the
order of categories in the file reflects that without changing the
@code{cat-index} values.

A Group is a group of nested categories.  Usually a Group contains at
least one Category, so that @code{n-subcategories} is positive, but a
few Groups with @code{n-subcategories} 0 has been observed.

If a Group's @code{merge} is 00, the most common value, then the group
is really a distinct group that should be represented as such in the
visual representation and user interface.  If @code{merge} is 01, the
categories in this group should be shown and treated as if they were
direct children of the group's containing group (or if it has no
parent group, then direct children of the dimension), and this group's
name is irrelevant and should not be displayed.  (Merged groups can be
nested!)

A Group's @code{data} appears to be i2 when all of the categories
within a group are leaf categories that directly represent data values
for a variable (e.g. in a frequency table or crosstabulation, a group
of values in a variable being tabulated) and i0 otherwise.

@node SPV Light Member Data
@subsection Data

The final part of an SPV light member contains the actual data.

@cartouche
@format
Data @result{}
    int[@t{layers}] int[@t{rows}] int[@t{columns}] int*[@t{n-dimensions}]
    int[@t{n-data}] Datum*[@t{n-data}]
Datum @result{} int64[@t{index}] v1(00?) Value
@end format
@end cartouche

The values of @code{n-layers}, @code{n-rows}, and @code{n-columns}
each specifies the number of dimensions displayed in layers, rows, and
columns, respectively.  Any of them may be zero.  Their values sum to
@code{n-dimensions} from Dimensions (@pxref{SPV Light Member
Dimensions}).

The @code{n-dimensions} integers are a permutation of the 0-based
dimension numbers.  The first @code{n-layers} integers specify each of
the dimensions represented by layers, the next @code{n-rows} integers
specify the dimensions represented by rows, and the final
@code{n-columns} integers specify the dimensions represented by
columns.  When there is more than one dimension of a given kind, the
inner dimensions are given first.

The format of a Datum varies slightly from version 1 to version 3: in
version 1 it allows for an extra optional 00 byte.

A Datum consists of an @code{index} and a Value.  Suppose there are
@math{d} dimensions and dimension @math{i}, @math{0 \le i < d}, has
@math{n_i} categories.  Consider the datum at coordinates @math{x_i},
@math{0 \le i < d}, and note that @math{0 \le x_i < n_i}.  Then the
index is calculated by the following algorithm:

@display
let @i{index} = 0
for each @math{i} from 0 to @math{d - 1}:
    @i{index} = (@math{n_i \times} @i{index}) @math{+} @math{x_i}
@end display

For example, suppose there are 3 dimensions with 3, 4, and 5
categories, respectively.  The datum at coordinates (1, 2, 3) has
index @math{5 \times (4 \times (3 \times 0 + 1) + 2) + 3 = 33}.
Within a given dimension, the index is the @code{cat-index} in a Leaf.

@node SPV Light Member Value
@subsection Value

Value is used throughout the SPV light member format.  It boils down
to a number or a string.

@cartouche
@format
Value @result{} 00? 00? 00? 00? RawValue
RawValue @result{}
    01 ValueMod int[@t{format}] double[@t{x}]
  @math{|} 02 ValueMod int[@t{format}] double[@t{x}]
    string[@t{varname}] string[@t{vallab}] (01 @math{|} 02 @math{|} 03)
  @math{|} 03 string[@t{local}] ValueMod string[@t{id}] string[@t{c}] bool[@t{type}]
  @math{|} 04 ValueMod int[@t{format}] string[@t{vallab}] string[@t{varname}]
    (01 @math{|} 02 @math{|} 03) string[@t{s}]
  @math{|} 05 ValueMod string[@t{varname}] string[@t{varlabel}] (01 @math{|} 02 @math{|} 03)
  @math{|} ValueMod string[@t{format}] int[@t{n-args}] Argument*[@t{n-args}]
Argument @result{}
    i0 Value
  @math{|} int[@t{x}] i0 Value*[@t{x}@math{+}1]      /* @t{x} @math{>} 0 */
@end format
@end cartouche

There are several possible encodings, which one can distinguish by the
first nonzero byte in the encoding.

@table @asis
@item 01
The numeric value @code{x}, intended to be presented to the user
formatted according to @code{format}, which is in the format described
for system files.  @xref{System File Output Formats}, for details.
Most commonly, @code{format} has width 40 (the maximum).

An @code{x} with the maximum negative double value @code{-DBL_MAX}
represents the system-missing value SYSMIS.  (HIGHEST and LOWEST have
not been observed.)  @xref{System File Format}, for more about these
special values.

@item 02
Similar to @code{01}, with the additional information that @code{x} is
a value of variable @code{varname} and has value label @code{vallab}.
Both @code{varname} and @code{vallab} can be the empty string, the
latter very commonly.

The meaning of the final byte is unknown.  Possibly it is connected to
whether the value or the label should be displayed.

@item 03
A text string, in two forms: @code{c} is in English, and sometimes
abbreviated or obscure, and @code{local} is localized to the user's
locale.  In an English-language locale, the two strings are often the
same, and in the cases where they differ, @code{local} is more
appropriate for a user interface, e.g.@: @code{c} of ``Not a PxP table
for MCN...'' versus @code{local} of ``Computed only for a PxP table,
where P must be greater than 1.''

@code{c} and @code{local} are always either both empty or both
nonempty.

@code{id} is a brief identifying string whose form seems to resemble a
programming language identifier, e.g.@: @code{cumulative_percent} or
@code{factor_14}.  It is not unique.

@code{type} is 00 for text taken from user input, such as syntax
fragment, expressions, file names, data set names, and 01 for fixed
text strings such as names of procedures or statistics.  In the former
case, @code{id} is always the empty string; in the latter case,
@code{id} is still sometimes empty.

@item 04
The string value @code{s}, intended to be presented to the user
formatted according to @code{format}.  The format for a string is not
too interesting, and the corpus contains many clearly invalid formats
like A16.39 or A255.127 or A134.1, so readers should probably ignore
the format entirely.

@code{s} is a value of variable @code{varname} and has value label
@code{vallab}.  @code{varname} is never empty but @code{vallab} is
commonly empty.

The meaning of the final byte is unknown.

@item 05
Variable @code{varname}, which is rarely observed as empty in the
corpus, with variable label @code{varlabel}, which is often empty.

The meaning of the final byte is unknown.

@item 31 or 58
(These bytes begin a ValueMod.)  A format string, analogous to
@code{printf}, followed by one or more Arguments, each of which has
one or more values.  The format string uses the following syntax:

@table @code
@item \%
@itemx \:
@itemx \[
@itemx \]
Each of these expands to the character following @samp{\\}, to escape
characters that have special meaning in format strings.  These are
effective inside and outside the @code{[@dots{}]}  syntax forms
described below.

@item \n
Expands to a new-line, inside or outside the @code{[@dots{}]} forms
described below.

@item ^@var{i}
Expands to a formatted version of argument @var{i}, which must have
only a single value.  For example, @code{^1} expands to the first
argument's @code{value}.

@item [:@var{a}:]@var{i}
Expands @var{a} for each of the values in @var{i}.  @var{a}
should contain one or more @code{^@var{j}} conversions, which are
drawn from the values for argument @var{i} in order.  Some examples
from the corpus:

@table @code
@item [:^1:]1
All of the values for the first argument, concatenated.

@item [:^1\n:]1
Expands to the values for the first argument, each followed by
a new-line.

@item [:^1 = ^2:]2
Expands to @code{@var{x} = @var{y}} where @var{x} is the second
argument's first value and @var{y} is its second value.  (This would
be used only if the argument has two values.  If there were more
values, the second and third values would be directly concatenated,
which would look funny.)
@end table

@item [@var{a}:@var{b}:]@var{i}
This extends the previous form so that the first values are expanded
using @var{a} and later values are expanded using @var{b}.  For an
unknown reason, within @var{a} the @code{^@var{j}} conversions are
instead written as @code{%@var{j}}.  Some examples from the corpus:

@table @code
@item [%1:*^1:]1
Expands to all of the values for the first argument, separated by
@samp{*}.

@item [%1 = %2:, ^1 = ^2:]1
Given appropriate values for the first argument, expands to @code{X =
1, Y = 2, Z = 3}.

@item [%1:, ^1:]1
Given appropriate values, expands to @code{1, 2, 3}.
@end table
@end table

The format string is localized to the user's locale.
@end table

@node SPV Light Member ValueMod
@subsection ValueMod

A ValueMod can specify special modifications to a Value.

@cartouche
@format
ValueMod @result{}
    31 i0 (i0 @math{|} i1 string[@t{subscript}])
    v1(00 (i1 @math{|} i2) 00 00 int 00 00)
    v3(count(FormatString StylePair))
  @math{|} 31 int[@t{n-refs}] int16*[@t{n-refs}] Format
  @math{|} 58

Format @result{} 00 00 count(FormatString Style 58)
FormatString @result{} count((count((i0 58)?) (58 @math{|} 31 string))?)

StylePair @result{}
    (31 Style | 58)
    (31 Style2 | 58)

Style @result{}
    bool[@t{bold}] bool[@t{italic}] bool[@t{underline}] bool[@t{show}]
    string[@t{fgcolor}] string[@t{bgcolor}]
    string[@t{typeface}] byte[@t{size}]

Style2 @result{}
    int[@t{halign}] int[@t{valign}] double[@t{offset}]
    int16[@t{left-margin}] int16[@t{right-margin}]
    int16[@t{top-margin}] int16[@t{bottom-margin}]
@end format
@end cartouche

A ValueMod that begins with ``31 i0'' specifies a string to append to
the main text of the Value, as a subscript.  The subscript text is a
brief indicator, e.g.@: @samp{a} or @samp{a,b}, with its meaning
indicated by the table caption.  In this usage, subscripts are similar
to footnotes.  One apparent difference is that a Value can only
reference one footnote but a subscript can list more than one letter.

A ValueMod that begins with 31 followed by a nonzero ``int'' specifies
a footnote or footnotes that the Value references.  Footnote markers
are shown appended to the main text of the Value, as superscripts.

The Format, if present, is a format string for substitutions using the
syntax explained previously.  It appears to be an English-language
version of the localized format string in the Value in which the
Format is nested.

Style and Style2, if present, change the style for this individual
Value.  @code{bold}, @code{italic}, and @code{underline} control the
particular style.  @code{fgcolor} and @code{bgcolor} are strings, such
as @code{#ffffff}.  The @code{size} is a font size in units of 1/96
inch.

@code{halign} is 0 for center, 2 for left, 4 for right, 6 for decimal,
0xffffffad for mixed.  For decimal alignment, @code{offset} is the
decimal point's offset from the right side of the cell, in units of
1/72 inch.

@code{valign} specifies vertical alignment: 0 for center, 1 for top, 3
for bottom.

@code{left-margin}, @code{right-margin}, @code{top-margin}, and
@code{bottom-margin} are in units of 1/72 inch.

@node SPV Legacy Detail Member Binary Format
@section Legacy Detail Member Binary Format

Whereas the light binary format represents everything about a given
pivot table, the legacy binary format conceptually consists of a
number of named sources, each of which consists of a number of named
variables, each of which is a 1-dimensional array of numbers or
strings or a mix.  Thus, the legacy binary member format is quite
simple.

This section uses the same context-free grammar notation as in the
previous section, with the following additions:

@table @asis
@item vAF(@var{x})
In a version 0xaf legacy member, @var{x}; in other versions, nothing.
(The legacy member header indicates the version; see below.)

@item vB0(@var{x})
In a version 0xb0 legacy member, @var{x}; in other versions, nothing.
@end table

A legacy detail member @file{.bin} has the following overall format:

@cartouche
@format
LegacyBinary @result{}
    00 byte[@t{version}] int16[@t{n-sources}] int[@t{member-size}]
    Metadata*[@t{n-sources}] Data*[@t{n-sources}]
@end format
@end cartouche

@code{version} is a version number that affects the interpretation of
some of the other data in the member.  Versions 0xaf and 0xb0 are
known.  We will refer to ``version 0xaf'' and ``version 0xb0'' members
later on.

A legacy member consists of @code{n-sources} data sources, each of
which has Metadata and Data.

@code{member-size} is the size of the legacy binary member, in bytes.

The following sections go into more detail.

@menu
* SPV Legacy Member Metadata::
* SPV Legacy Member Data::
@end menu

@node SPV Legacy Member Metadata
@subsection Metadata

@cartouche
@format
Metadata @result{}
    int[@t{n-data}] int[@t{n-variables}] int[@t{offset}]
    vAF(byte*32[@t{source-name}])
    vB0(byte*64[@t{source-name}] int[@t{x}])
@end format
@end cartouche

A data source has @code{n-variables} variables, each with
@code{n-data} data values.

@code{source-name} is a 32- or 64-byte string padded on the right with
zero bytes.  The names that appear in the corpus are very generic:
usually @code{tableData} for pivot table data or @code{source0} for
chart data.

A given Metadata's @code{offset} is the offset, in bytes, from the
beginning of the member to the start of the corresponding Data.  This
allows programs to skip to the beginning of the data for a particular
source; it is also important to determine whether a source includes
any string data (@pxref{SPV Legacy Member Data}).

The meaning of @code{x} in version 0xb0 is unknown.

@node SPV Legacy Member Data
@subsection Data

@cartouche
@format
Data @result{} NumericData*[@t{n-variables}] StringData?
NumericData @result{} byte*288[@t{variable-name}] double*[@t{n-data}]
@end format
@end cartouche

Data follow the Metadata in the legacy binary format, with sources in
the same order.  Each NumericSeries begins with a @code{variable-name}
that generally indicates its role in the pivot table, e.g.@: ``cell'',
``cellFormat'', ``dimension0categories'', ``dimension0group0'',
followed by the numeric data, one double per datum.  A double with the
maximum negative double @code{-DBL_MAX} represents the system-missing
value SYSMIS.

@cartouche
@format
StringData @result{} i1 string[@t{source-name}] Pairs Labels

Pairs @result{} int[@t{n-string-vars}] PairSeries*[@t{n-string-vars}]
PairVar @result{} string[@t{pair-var-name}] int[@t{n-pairs}] Pair*[@t{n-pairs}]
Pair @result{} int[@t{i}] int[@t{j}]

Labels @result{} int[@t{n-labels}] Label*[@t{n-labels}]
Label @result{} int[@t{frequency}] int[@t{s}]
@end format
@end cartouche

A source may include a mix of numeric and string data values.  When a
source includes any string data, the data values that are strings are
set to SYSMIS in the NumericData, and StringData follows the
NumericData.  A source that contains no string data omits the
StringData.  To reliably determine whether a source includes
StringData, the reader should check whether the offset following the
NumericData is the offset of the next source, as indicated by its
Metadata (or the end of the member, in the case of the last source).

StringData repeats the name of the source (from Metadata).

The string data overlays the numeric data.  @code{n-string-vars} is
the number of variables in the source that include string data.  More
precisely, it is the 1-based index of the last variable in the source
that includes any string data; thus, it would be 4 if there are 5
variables and only the fourth one includes string data.

Each PairVar consists a sequence of 0 or more Pair nonterminals, each
of which maps from a 0-based index within variable @code{i} to a
0-based label index @code{j}, e.g.@: pair @code{i} = 2, @code{j} = 3,
means that the third data value (with value SYSMIS) is to be replaced
by the string of the fourth Label.

The labels themselves follow the pairs.  The valuable part of each
label is the string @code{s}.  Each label also includes a
@code{frequency} that reports the number of pairs that reference it
(although this is not useful).

@node SPV Legacy Detail Member XML Format
@section Legacy Detail Member XML Format

This format is still under investigation.

The design of the detail XML format is not what one would end up with
for describing pivot tables.  This is because it is a special case
of a much more general format (``visualization XML'' or ``VizML'')
that can describe a wide range of visualizations.  Most of this
generality is overkill for tables, and so we end up with a funny
subset of a general-purpose format.

The important elements of the detail XML format are:

@itemize @bullet
@item
Variables.  Variables in detail XML roughly correspond to the
dimensions in a light detail member.  There is one variable for each
dimension, plus one variable for each level of labeling along an axis.

The bulk of variables are defined with @code{sourceVariable} elements.
The data for these variables comes from the associated
@code{tableData.bin} member.  Some variables are defined, with
@code{derivedVariable} elements, as a constant or in terms of a
mapping function from a source variable.

@item
Assignment of variables to axes.  A variable can appear as columns, or
rows, or layers.  The @code{faceting} element and its sub-elements
describe this assignment.
@end itemize

All elements have an optional @code{id} attribute.  In practice many
elements are assigned @code{id} attributes that are never referenced.

@menu
* SPV Detail visualization Element::
* SPV Detail userSource Element::
* SPV Detail sourceVariable Element::
* SPV Detail derivedVariable Element::
* SPV Detail extension Element::
* SPV Detail graph Element::
* SPV Detail location Element::
* SPV Detail coordinates Element::
* SPV Detail faceting Element::
* SPV Detail facetLayout Element::
* SPV Detail style Element::
@end menu

@node SPV Detail visualization Element
@subsection The @code{visualization} Element

@format
Parent: Document root
Contents:
     extension?
     userSource
     (sourceVariable @math{|} derivedVariable)@math{+}
     graph
     labelFrame@math{+}
     container?
     style@math{+}
     layerController?
@end format

This element has the following attributes.

@defvr {Required} creator
The version of the software that created this SPV file, as a string of
the form @code{xxyyzz}, which represents software version xx.yy.zz,
e.g.@: @code{160001} is version 16.0.1.  The corpus includes major
versions 16 through 19.
@end defvr

@defvr {Required} date
The date on the which the file was created, as a string of the form
@code{YYYY-MM-DD}.
@end defvr

@defvr {Required} lang
The locale used for output, in Windows format, which is similar to the
format used in Unix with the underscore replaced by a hyphen, e.g.@:
@code{en-US}, @code{en-GB}, @code{el-GR}, @code{sr-Cryl-RS}.
@end defvr

@defvr {Required} name
The title of the pivot table, localized to the output language.
@end defvr

@defvr {Required} style
The @code{id} of a @code{style} element (@pxref{SPV Detail style
Element}).  This is the base style for the entire pivot table.  In
every example in the corpus, the value is @code{visualizationStyle}
and the corresponding @code{style} element has no attributes other
than @code{id}.
@end defvr

@defvr {Required} type
A floating-point number.  The meaning is unknown.
@end defvr

@defvr {Required} version
The visualization schema version number.  In the corpus, the value is
one of 2.4, 2.5, 2.7, and 2.8.
@end defvr

@node SPV Detail userSource Element
@subsection The @code{userSource} Element

Parent: @code{visualization} @*
Contents:

This element has the following attributes.

@defvr {Optional} missing
Always @code{listwise}.
@end defvr

@node SPV Detail sourceVariable Element
@subsection The @code{sourceVariable} Element

Parent: @code{visualization} @*
Contents: @code{extension}* (@code{format} @math{|} @code{stringFormat})?

This element defines a variable whose values can be used elsewhere in
the visualization.  It ties this element's @code{id} to a variable
from the @file{tableData.bin} member that corresponds to this
@file{.xml}.

This element has the following attributes.

@defvr {Required} categorical
Always set to @code{true}.
@end defvr

@defvr {Required} source
Always set to @code{tableData}, the @code{source-name} in the
corresponding @file{tableData.bin} member (@pxref{SPV Legacy Member
Metadata}).
@end defvr

@defvr {Required} sourceName
The name of a variable within the source, the @code{variable-name} in
the corresponding @file{tableData.bin} member (@pxref{SPV Legacy
Member Data}).
@end defvr

@defvr {Optional} dependsOn
The @code{variable-name} of a variable linked to this one, so that a
viewer can work with them together.  For a group variable, this is the
name of the corresponding categorical variable.
@end defvr

@defvr {Optional} label
The variable label, if any
@end defvr

@defvr {Optional} labelVariable
The @code{variable-name} of a variable whose string values correspond
one-to-one with the values of this variable and are suitable for use
as value labels.
@end defvr

@node SPV Detail derivedVariable Element
@subsection The @code{derivedVariable} Element

Parent: @code{visualization} @*
Contents: @code{extension}* (@code{format} @math{|} @code{stringFormat} @code{valueMapEntry}*)

Like @code{sourceVariable}, this element defines a variable whose
values can be used elsewhere in the visualization.  Instead of being
read from a data source, the variable's data are defined by a
mathematical expression.

This element has the following attributes.

@defvr {Required} categorical
Always set to @code{true}.
@end defvr

@defvr {Required} value
An expression that defines the variable's value.  In theory this could
be an arbitrary expression in terms of constants, functions, and other
variables, e.g.@: @math{(@var{var1} + @var{var2}) / 2}.  In practice,
the corpus contains only the following forms of expressions:

@table @code
@item constant(@var{number})
@itemx constant(@var{variable})
A constant.  The meaning when a variable is named is unknown.
Sometimes the ``variable name'' has spaces in it.

@item map(@var{variable})
Transforms the values in the named @var{variable} using the
@code{valueMapEntry}s contained within the element.
@end table
@end defvr

@defvr {Optional} dependsOn
The @code{variable-name} of a variable linked to this one, so that a
viewer can work with them together.  For a group variable, this is the
name of the corresponding categorical variable.
@end defvr

@menu
* SPV Detail valueMapEntry Element::
@end menu

@node SPV Detail valueMapEntry Element
@subsubsection The @code{valueMapEntry} Element

Parent: @code{derivedVariable} @*
Contents: empty

A @code{valueMapEntry} element defines a mapping from one or more
values of a source expression to a target value.  (In the corpus, the
source expression is always just the name of a variable.)  Each target
value requires a separate @code{valueMapEntry}.  If multiple source
values map to the same target value, they can be combined or separate.

@code{valueMapEntry} has the following attributes.

@defvr {Required} from
A source value, or multiple source values separated by semicolons,
e.g.@: @code{0} or @code{13;14;15;16}.
@end defvr

@defvr {Required} to
The target value.
@end defvr

@node SPV Detail extension Element
@subsection The @code{extension} Element

This is a general-purpose ``extension'' element.  Readers that don't
understand a given extension should be able to safely ignore it.  The
attributes on this element, and their meanings, vary based on the
context.  Each known usage is described separately below.  The current
extensions use attributes exclusively, without any nested elements.

@subsubheading @code{visualization} Parent Element

With @code{visualization} as its parent element, @code{extension} has
the following attributes.

@defvr {Optional} numRows
An integer that presumably defines the number of rows in the displayed
pivot table.
@end defvr

@defvr {Optional} showGridline
Always set to @code{false} in the corpus.
@end defvr

@defvr {Optional} minWidthSet
@defvrx {Optional} maxWidthSet
Always set to @code{true} in the corpus.
@end defvr

@subsubheading @code{container} Parent Element

With @code{container} as its parent element, @code{extension} has the
following attributes.

@defvr {Required} combinedFootnotes
Always set to @code{true} in the corpus.
@end defvr

@subsubheading @code{sourceVariable} and @code{derivedVariable} Parent Element

With @code{sourceVariable} or @code{derivedVariable} as its parent
element, @code{extension} has the following attributes.  A given
parent element often contains several @code{extension} elements that
specify the meaning of the source data's variables or sources, e.g.@:

@example
<extension from="0" helpId="corrected_model"/>
<extension from="3" helpId="error"/>
<extension from="4" helpId="total_9"/>
<extension from="5" helpId="corrected_total"/>
@end example

@defvr {Required} from
An integer or a name like ``dimension0''.
@end defvr

@defvr {Required} helpId
An identifier.
@end defvr

@node SPV Detail graph Element
@subsection The @code{graph} Element

Parent: @code{visualization} @*
Contents: @code{location}@math{+} @code{coordinates} @code{faceting} @code{facetLayout} @code{interval}

@code{graph} has the following attributes.

@defvr {Required} cellStyle
@defvrx {Required} style
Each of these is the @code{id} of a @code{style} element (@pxref{SPV
Detail style Element}).  The former is the default style for
individual cells, the latter for the entire table.
@end defvr

@node SPV Detail location Element
@subsection The @code{location} Element

Parent: @code{graph} @*
Contents: empty

Each instance of this element specifies where some part of the table
frame is located.  All the examples in the corpus have four instances
of this element, one for each of the parts @code{height},
@code{width}, @code{left}, and @code{top}.  Some examples in the
corpus add a fifth for part @code{bottom}, even though it is not clear
how all of @code{top}, @code{bottom}, and @code{heigth} can be honored
at the same time.  In any case, @code{location} seems to have little
importance in representing tables; a reader can safely ignore it.

@defvr {Required} part
One of @code{height}, @code{width}, @code{top}, @code{bottom}, or
@code{left}.  Presumably @code{right} is acceptable as well but the
corpus contains no examples.
@end defvr

@defvr {Required} method
How the location is determined:

@table @code
@item sizeToContent
Based on the natural size of the table.  Observed only for
parts @code{height} and @code{width}.

@item attach
Based on the location specified in @code{target}.  Observed only for
parts @code{top} and @code{bottom}.

@item fixed
Using the value in @code{value}.  Observed only for parts @code{top},
@code{bottom}, and @code{left}.

@item same
Same as the specified @code{target}.  Observed only for part
@code{left}.
@end table
@end defvr

@defvr {Optional} min
Minimum size.  Only observed with value @code{100pt}.  Only observed
for part @code{width}.
@end defvr

@defvr {Dependent} target
Required when @code{method} is @code{attach} or @code{same}, not
observed otherwise.  This is the ID of an element to attach to.
Observed with the ID of @code{title}, @code{footnote}, @code{graph},
and other elements.
@end defvr

@defvr {Dependent} value
Required when @code{method} is @code{fixed}, not observed otherwise.
Observed values are @code{0%}, @code{0px}, @code{1px}, and @code{3px}
on parts @code{top} and @code{left}, and @code{100%} on part
@code{bottom}.
@end defvr

@node SPV Detail coordinates Element
@subsection The @code{coordinates} Element

Parent: @code{graph} @*
Contents: empty

This element is always present and always empty, with no attributes
(except @code{id}).

@node SPV Detail faceting Element
@subsection The @code{faceting} Element

Parent: @code{graph} @*
Contents: @code{cross} @code{layer}*

The @code{faceting} element describes the row, column, and layer
structure of the table.  Its @code{cross} child determines the row and
column structure, and each @code{layer} child (if any) represents a
layer.

@code{faceting} has no attributes (other than @code{id}).

@subsubheading The @code{cross} Element

Parent: @code{faceting} @*
Contents: @code{nest} @code{nest}

The @code{cross} element describes the row and column structure of the
table.  It has exactly two @code{nest} children, the first of which
describes the table's rows and the second the table's columns.

@code{cross} has no attributes (other than @code{id}).

@subsubheading The @code{nest} Element

Parent: @code{cross} @*
Contents: @code{variableReference}@math{+}

A given @code{nest} usually consists of one or more dimensions, each
of which is represented by @code{variableReference} child elements.
Minimally, a dimension has two @code{variableReference} children, one
for the categories, one for the data, e.g.:

@example
<nest>
  <variableReference ref="dimension0categories"/>
  <variableReference ref="dimension0"/>
</nest>
@end example

@noindent
Groups of categories introduce additional variable references, e.g.@:

@example
<nest>
  <variableReference ref="dimension0categories"/>
  <variableReference ref="dimension0group0"/>
  <variableReference ref="dimension0"/>
</nest>
@end example

@noindent
Grouping can be hierarchical, e.g.@:

@example
<nest>
  <variableReference ref="dimension0categories"/>
  <variableReference ref="dimension0group1"/>
  <variableReference ref="dimension0group0"/>
  <variableReference ref="dimension0"/>
</nest>
@end example

@noindent
XXX what are group maps?

@example
<nest id="nest_1973">
  <variableReference ref="dimension1categories"/>
  <variableReference ref="dimension1group1map"/>
  <variableReference ref="dimension1group0map"/>
  <variableReference ref="dimension1"/>
</nest>
<nest>
  <variableReference ref="dimension0categories"/>
  <variableReference ref="dimension0group0map"/>
  <variableReference ref="dimension0"/>
</nest>
@end example

@noindent
A @code{nest} can contain multiple dimensions:

@example
<nest>
  <variableReference ref="dimension1categories"/>
  <variableReference ref="dimension1group0"/>
  <variableReference ref="dimension1"/>
  <variableReference ref="dimension0categories"/>
  <variableReference ref="dimension0"/>
</nest>
@end example

One @code{nest} within a given @code{cross} may have no dimensions, in
which case it still has one @code{variableReference} child, which
references a @code{derivedVariable} whose @code{value} attribute is
@code{constant(0)}.  In the corpus, such a @code{derivedVariable} has
@code{row} or @code{column}, respectively, as its @code{id}.

@code{nest} has no attributes (other than @code{id}).

@subsubheading The @code{variableReference} Element

Parent: @code{nest} @*
Contents: empty

@code{variableReference} has one attribute.

@defvr {Required} ref
The @code{id} of a @code{sourceVariable} or @code{derivedVariable}
element.
@end defvr

@subsubheading The @code{layer} Element

Parent: @code{faceting} @*
Contents: empty

Each layer is represented by a pair of @code{layer} elements.  The
first of this pair is for a category variable, the second for the data
variable, e.g.:

@example
<layer value="0" variable="dimension0categories" visible="true"/>
<layer value="dimension0" variable="dimension0" visible="false"/>
@end example

@noindent
@code{layer} has the following attributes.

@defvr {Required} variable
The @code{id} of a @code{sourceVariable} or @code{derivedVariable}
element.
@end defvr

@defvr {Required} value
The value to select.  For a category variable, this is always
@code{0}; for a data variable, it is the same as the @code{variable}
attribute.
@end defvr

@defvr {Optional} visible
Whether the layer is visible.  Generally, category layers are visible
and data layers are not, but sometimes this attribute is omitted.
@end defvr

@defvr {Optional} method
When present, this is always @code{nest}.
@end defvr

@node SPV Detail facetLayout Element
@subsection The @code{facetLayout} Element

Parent: @code{graph} @*
Contents: @code{tableLayout} @code{facetLevel}@math{+} @code{setCellProperties}*

@subsubheading The @code{tableLayout} Element

Parent: @code{facetLayout} @*
Contents: empty

@defvr {Required} verticalTitlesInCorner
Always set to @code{true}.
@end defvr

@defvr {Optional} style
The @code{id} of a @code{style} element.
@end defvr

@defvr {Optional} fitCells
Always set to @code{ticks}.
@end defvr

@subsubheading The @code{facetLevel} Element

Parent: @code{facetLayout} @*
Contents: @code{axis}

Each @code{facetLevel} describes a @code{variableReference} or
@code{layer}, and a table has one @code{facetLevel} element for
each such element.  For example, an SPV detail member that contains
four @code{variableReference} elements and two @code{layer} elements
will contain six @code{facetLevel} elements.

In the corpus, @code{facetLevel} elements and the elements that they
describe are always in the same order.  The correspondence may also be
observed in two other ways.  First, one may use the @code{level}
attribute, described below.  Second, in the corpus, a
@code{facetLevel} always has an @code{id} that is the same as the
@code{id} of the element it describes with @code{_facetLevel}
appended.  One should not formally rely on this, of course, but it is
usefully indicative.

@defvr {Required} level
A 1-based index into the @code{variableReference} and @code{layer}
elements, e.g.@: a @code{facetLayout} with a @code{level} of 1
describes the first @code{variableReference} in the SPV detail member,
and in a member with four @code{variableReference} elements, a
@code{facetLayout} with a @code{level} of 5 describes the first
@code{layer} in the member.
@end defvr

@defvr {Required} gap
Always observed as @code{0pt}.
@end defvr

@subsubheading The @code{axis} Element

Parent: @code{facetLevel} @*
Contents: @code{label}? @code{majorTicks}

@defvr {Attribute} style
The @code{id} of a @code{style} element.
@end defvr

@subsubheading The @code{label} Element

Parent: @code{axis} or @code{labelFrame} @*
Contents: @code{text}@math{+} @math{|} @code{descriptionGroup}

This element represents a label on some aspect of the table.  For example,
the table's title is a @code{label}.

The contents of the label can be one or more @code{text} elements or a
@code{descriptionGroup}.

@defvr {Attribute} style
@defvrx {Optional} textFrameStyle
Each of these is the @code{id} of a @code{style} element.
@code{style} is the style of the label text, @code{textFrameStyle} the
style for the frame around the label.
@end defvr

@defvr {Optional} purpose
The kind of entity being labeled, one of @code{title},
@code{subTitle}, @code{layer}, or @code{footnote}.
@end defvr

@subsubheading The @code{descriptionGroup} Element

Parent: @code{label} @*
Contents: (@code{description} @math{|} @code{text})@math{+}

A @code{descriptionGroup} concatenates one or more elements to form a
label.  Each element can be a @code{text} element, which contains
literal text, or a @code{description} element that substitutes a value
or a variable name.

@defvr {Attribute} target
The @code{id} of an element being described.  In the corpus, this is
always @code{faceting}.
@end defvr

@defvr {Attribute} separator
A string to separate the description of multiple groups, if the
@code{target} has more than one.  In the corpus, this is always a
new-line.
@end defvr

Typical contents for a @code{descriptionGroup} are a value by itself:
@example
<description name="value"/>
@end example
@noindent or a variable and its value, separated by a colon:
@example
<description name="variable"/><text>:</text><description name="value"/>
@end example

@subsubheading The @code{description} Element

Parent: @code{descriptionGroup} @*
Contents: empty

A @code{description} is like a macro that expands to some property of
the target of its parent @code{descriptionGroup}.

@defvr {Attribute} name
The name of the property.  Only @code{variable} and @code{value}
appear in the corpus.
@end defvr

@subsubheading The @code{majorTicks} Element

Parent: @code{axis} @*
Contents: @code{gridline}?

@defvr {Attribute} labelAngle
@defvrx {Attribute} length
Both always defined to @code{0}.
@end defvr

@defvr {Attribute} style
@defvrx {Attribute} tickFrameStyle
Each of these is the @code{id} of a @code{style} element.
@code{style} is the style of the tick labels, @code{tickFrameStyle}
the style for the frames around the labels.
@end defvr

@subsubheading The @code{gridline} Element

Parent: @code{majorTicks} @*
Contents: empty

Represents ``gridlines,'' which for a table represents the lines
between the rows or columns of a table (XXX?).

@defvr {Attribute} style
The style for the gridline.
@end defvr

@defvr {Attribute} zOrder
Observed as a number between 28 and 31.  Does not seem to be
important.
@end defvr

@subsubheading The @code{setCellProperties} Element

Parent: @code{facetLayout} @*
Contents: @code{setMetaData} @code{setStyle}* @code{setFormat}@math{+} @code{union}?

This element sets style properties of cells designated by the
@code{target} attribute of its child elements, as further restricted
by the optional @code{union} element if present.  The @code{target}
values often used, e.g.@: @code{graph} or @code{labeling}, actually
affect every cell, so the @code{union} element is a useful
restriction.

@defvr {Optional} applyToConverse
If present, always @code{true}.  This appears to invert the meaning of
the @code{target} of sub-elements: the selected cells are the ones
@emph{not} designated by @code{target}.  This is confusing, given the
additional restrictions of @code{union}, but in the corpus
@code{applyToConverse} is never present along with @code{union}.
@end defvr

@subsubheading The @code{setMetaData} Element

Parent: @code{setCellProperties} @*
Contents: empty

This element is not known to have any visible effect.

@defvr {Required} target
The @code{id} of an element whose metadata is to be set.  In the
corpus, this is always @code{graph}, the @code{id} used for the
@code{graph} element.
@end defvr

@defvr {Required} key
@defvrx {Required} value
A key-value pair to set for the target.

In the corpus, @code{key} is @code{cellPropId} or, rarely,
@code{diagProps}, and @code{value} is always the @code{id} of the
parent @code{setCellProperties}.
@end defvr

@subsubheading The @code{setStyle} Element

Parent: @code{setCellProperties} @*
Contents: empty

This element associates a style with the target.

@defvr {Required} target
The @code{id} of an element whose style is to be set.  In the corpus,
this is always the @code{id} of an @code{interval}, @code{labeling},
or, rarely, @code{graph} element.
@end defvr

@defvr {Required} style
The @code{id} of a @code{style} element that identifies the style to
set on the target.
@end defvr

@subsubheading The @code{setFormat} Element

@format
Parent: @code{setCellProperties}
Contents:
    @code{format}
  @math{|} @code{numberFormat}
  @math{|} @code{stringFormat}@math{+}
  @math{|} @code{dateTimeFormat}
@end format

This element sets the format of the target, ``format'' in this case
meaning the SPSS print format for a variable.

The details of this element vary depending on the schema version, as
declared in the root @code{visualization} element's @code{version}
attribute (@pxref{SPV Detail visualization Element}).  In version 2.5
and earlier, @code{setFormat} contains one of a number of child
elements that correspond to the different varieties of print formats.
In version 2.7 and later, @code{setFormat} instead always contains a
@code{format} element.

XXX reinvestigate the above claim about versions: it appears to be
incorrect.

The @code{setFormat} element itself has the following attributes.

@defvr {Required} target
The @code{id} of an element whose style is to be set.  In the corpus,
this is always the @code{id} of an @code{majorTicks} or
@code{labeling} element.
@end defvr

@defvr {Optional} reset
If this is @code{true}, this format overrides the target's previous
format.  If it is @code{false}, the adds to the previous format.  In
the corpus this is always @code{true}.  The default behavior is
unknown.
@end defvr

@menu
* SPV Detail format Element::
* SPV Detail numberFormat Element::
* SPV Detail stringFormat Element::
* SPV Detail dateTimeFormat Element::
* SPV Detail affix Element::
* SPV Detail relabel Element::
* SPV Detail union Element::
@end menu

@node SPV Detail format Element
@subsubsection The @code{format} Element

Parent: @code{sourceVariable}, @code{derivedVariable}, @code{formatMapping}, @code{labeling}, @code{formatMapping}, @code{setFormat} @*
Contents: (@code{affix}@math{+} @math{|} @code{relabel}@math{+})?

This element appears only in schema version 2.7 (@pxref{SPV Detail
visualization Element}).

This element determines a format, equivalent to an SPSS print format.

@subsubheading Attributes for All Formats

These attributes apply to all kinds of formats.  The most important of
these attributes determines the high-level kind of formatting in use:

@defvr {Optional} baseFormat
Either @code{dateTime} or @code{elapsedTime}.  When this attribute is
omitted, this element is a numeric or string format.
@end defvr

@noindent
Whether, in the corpus, other attributes are always present (``yes''),
never present (``no''), or sometimes present (``opt'') depends on
@code{baseFormat}:

@multitable {maximumFractionDigits} {@code{dateTime}} {@code{elapsedTime}} {number} {string}
@headitem Attribute @tab @code{dateTime} @tab @code{elapsedTime} @tab number @tab string
@item errorCharacter        @tab yes @tab yes @tab yes @tab opt
@item @w{ }
@item separatorChars        @tab yes @tab  no @tab  no @tab no
@item @w{ }
@item mdyOrder              @tab yes @tab  no @tab  no @tab no
@item @w{ }
@item showYear              @tab yes @tab  no @tab  no @tab no
@item yearAbbreviation      @tab yes @tab  no @tab  no @tab no
@item @w{ }
@item showMonth             @tab yes @tab  no @tab  no @tab no
@item monthFormat           @tab yes @tab  no @tab  no @tab no
@item @w{ }
@item showDay               @tab yes @tab opt @tab  no @tab no
@item dayPadding            @tab yes @tab opt @tab  no @tab no
@item dayOfMonthPadding     @tab yes @tab  no @tab  no @tab no
@item dayType               @tab yes @tab  no @tab  no @tab no
@item @w{ }
@item showHour              @tab yes @tab opt @tab  no @tab no
@item hourFormat            @tab yes @tab opt @tab  no @tab no
@item hourPadding           @tab yes @tab yes @tab  no @tab no
@item @w{ }
@item showMinute            @tab yes @tab yes @tab  no @tab no
@item minutePadding         @tab yes @tab yes @tab  no @tab no
@item @w{ }
@item showSecond            @tab yes @tab yes @tab  no @tab no
@item secondPadding         @tab  no @tab yes @tab  no @tab no
@item @w{ }
@item showMillis            @tab  no @tab yes @tab  no @tab no
@item @w{ }
@item minimumIntegerDigits  @tab  no @tab  no @tab yes @tab no
@item maximumFractionDigits @tab  no @tab yes @tab yes @tab no
@item minimumFractionDigits @tab  no @tab yes @tab yes @tab no
@item useGrouping           @tab  no @tab opt @tab yes @tab no
@item scientific            @tab  no @tab  no @tab yes @tab no
@item small                 @tab  no @tab  no @tab opt @tab no
@item suffix                @tab  no @tab  no @tab opt @tab no
@item @w{ }
@item tryStringsAsNumbers   @tab  no @tab  no @tab  no @tab yes
@item @w{ }
@end multitable

@defvr {Attribute} errorCharacter
A character that replaces the formatted value when it cannot otherwise
be represented in the given format.  Always @samp{*}.
@end defvr

@subsubheading Date and Time Attributes

These attributes are used with @code{dateTime} and @code{elapsedTime}
formats or both.

@defvr {Attribute} separatorChars
Exactly four characters.  In order, these are used for: decimal point,
grouping, date separator, time separator.  Always @samp{.,-:}.
@end defvr

@defvr {Attribute} mdyOrder
Within a date, the order of the days, months, and years.
@code{dayMonthYear} is the only observed value, but one would expect
that @code{monthDayYear} and @code{yearMonthDay} to be reasonable as
well.
@end defvr

@defvr {Attribute} showYear
@defvrx {Attribute} yearAbbreviation
Whether to include the year and, if so, whether the year should be
shown abbreviated, that is, with only 2 digits.  Each is @code{true}
or @code{false}; only values of @code{true} and @code{false},
respectively, have been observed.
@end defvr

@defvr {Attribute} showMonth
@defvrx {Attribute} monthFormat
Whether to include the month (@code{true} or @code{false}) and, if so,
how to format it.  @code{monthFormat} is one of the following:

@table @code
@item long
The full name of the month, e.g.@: in an English locale,
@code{September}.

@item short
The abbreviated name of the month, e.g.@: in an English locale,
@code{Sep}.

@item number
The number representing the month, e.g.@: 9 for September.

@item paddedNumber
A two-digit number representing the month, e.g.@: 09 for September.
@end table

Only values of @code{true} and @code{short}, respectively, have been
observed.
@end defvr

@defvr {Attribute} dayPadding
@defvrx {Attribute} dayOfMonthPadding
@defvrx {Attribute} hourPadding
@defvrx {Attribute} minutePadding
@defvrx {Attribute} secondPadding
These attributes presumably control whether each field in the output
is padded with spaces to its maximum width, but the details are not
understood.  The only observed value for any of these attributes is
@code{true}.
@end defvr

@defvr {Attribute} showDay
@defvrx {Attribute} showHour
@defvrx {Attribute} showMinute
@defvrx {Attribute} showSecond
@defvrx {Attribute} showMillis
These attributes presumably control whether each field is displayed
in the output, but the details are not understood.  The only
observed value for any of these attributes is @code{true}.
@end defvr

@defvr {Attribute} dayType
This attribute is always @code{month} in the corpus, specifying that
the day of the month is to be displayed; a value of @code{year} is
supposed to indicate that the day of the year, where 1 is January 1,
is to be displayed instead.
@end defvr

@defvr {Attribute} hourFormat
@code{hourFormat}, if present, is one of:

@table @code
@item AMPM
The time is displayed with an @code{am} or @code{pm} suffix, e.g.@:
@code{10:15pm}.

@item AS_24
The time is displayed in a 24-hour format, e.g.@: @code{22:15}.

This is the only value observed in the corpus.

@item AS_12
The time is displayed in a 12-hour format, without distinguishing
morning or evening, e.g.@: @code{10;15}.
@end table

@code{hourFormat} is sometimes present for @code{elapsedTime} formats,
which is confusing since a time duration does not have a concept of AM
or PM.  This might indicate a bug in the code that generated the XML
in the corpus, or it might indicate that @code{elapsedTime} is
sometimes used to format a time of day.
@end defvr

@subsubheading Numeric Attributes

These attributes are used for formats when @code{baseFormat} is
@code{number}.  Attributes @code{maximumFractionDigits}, and
@code{minimumFractionDigits}, and @code{useGrouping} are also used
when @code{baseFormat} is @code{elapsedTime}.

@defvr {Attribute} minimumIntegerDigits
Minimum number of digits to display before the decimal point.  Always
observed as @code{0}.
@end defvr

@defvr {Attribute} maximumFractionDigits
@defvrx {Attribute} maximumFractionDigits
Maximum or minimum, respectively, number of digits to display after
the decimal point.  The observed values of each attribute range from 0
to 9.
@end defvr

@defvr {Attribute} useGrouping
Whether to use the grouping character to group digits in large
numbers.  It would make sense for the grouping character to come from
the @code{separatorChars} attribute, but that attribute is only
present when @code{baseFormat} is @code{dateTime} or
@code{elapsedTime}, in the corpus at least.  Perhaps that is because
this attribute has only been observed as @code{false}.
@end defvr

@defvr {Attribute} scientific
This attribute controls when and whether the number is formatted in
scientific notation.  It takes the following values:

@table @code
@item onlyForSmall
Use scientific notation only when the number's magnitude is smaller
than the value of the @code{small} attribute.

@item whenNeeded
Use scientific notation when the number will not otherwise fit in the
available space.

@item true
Always use scientific notation.  Not observed in the corpus.

@item false
Never use scientific notation.  A number that won't otherwise fit will
be replaced by an error indication (see the @code{errorCharacter}
attribute).  Not observed in the corpus.
@end table
@end defvr

@defvr {Optional} small
Only present when the @code{scientific} attribute is
@code{onlyForSmall}, this is a numeric magnitude below which the
number will be formatted in scientific notation.  The values @code{0}
and @code{0.0001} have been observed.  The value @code{0} seems like a
pathological choice, since no real number has a magnitude less than 0;
perhaps in practice such a choice is equivalent to setting
@code{scientific} to @code{false}.
@end defvr

@defvr {Optional} prefix
@defvrx {Optional} suffix
Specifies a prefix or a suffix to apply to the formatted number.  Only
@code{suffix} has been observed, with value @samp{%}.
@end defvr

@subsubheading String Attributes

These attributes are used for formats when @code{baseFormat} is
@code{string}.

@defvr {Attribute} tryStringsAsNumbers
When this is @code{true}, it is supposed to indicate that string
values should be parsed as numbers and then displayed according to
numeric formatting rules.  However, in the corpus it is always
@code{false}.
@end defvr

@node SPV Detail numberFormat Element
@subsubsection The @code{numberFormat} Element

Parent: @code{setFormat} @*
Contents: @code{affix}@math{+}

This element appears only in schema version 2.5 and earlier
(@pxref{SPV Detail visualization Element}).  Possibly this element
could also contain @code{relabel} elements in a more diverse corpus.

This element has the following attributes.

@defvr {Attribute} maximumFractionDigits
@defvrx {Attribute} minimumFractionDigits
@defvrx {Attribute} minimumIntegerDigits
@defvrx {Optional} scientific
@defvrx {Optional} small
@defvrx {Optional} suffix
@defvrx {Optional} useGroupging
The syntax and meaning of these attributes is the same as on the
@code{format} element for a numeric format.  @pxref{SPV Detail format
Element}.
@end defvr

@node SPV Detail stringFormat Element
@subsubsection The @code{stringFormat} Element

Parent: @code{setFormat} @*
Contents: (@code{affix}@math{+} @math{|} @code{relabel}@math{+})?

This element appears only in schema version 2.5 and earlier
(@pxref{SPV Detail visualization Element}).

This element has no attributes.

@node SPV Detail dateTimeFormat Element
@subsubsection The @code{dateTimeFormat} Element

Parent: @code{setFormat} @*
Contents: empty

This element appears only in schema version 2.5 and earlier
(@pxref{SPV Detail visualization Element}).  Possibly this element
could also contain @code{affix} and @code{relabel} elements in a more
diverse corpus.

The following attribute is required.

@defvr {Attribute} baseFormat
Either @code{dateTime} or @code{time}.
@end defvr

When @code{baseFormat} is @code{dateTime}, the following attributes
are available.

@defvr {Attribute} dayOfMonthPadding
@defvrx {Attribute} dayPadding
@defvrx {Attribute} dayType
@defvrx {Attribute} hourFormat
@defvrx {Attribute} hourPadding
@defvrx {Attribute} mdyOrder
@defvrx {Attribute} minutePadding
@defvrx {Attribute} monthFormat
@defvrx {Attribute} separatorChars
@defvrx {Attribute} showDay
@defvrx {Attribute} showHour
@defvrx {Attribute} showMinute
@defvrx {Attribute} showMonth
@defvrx {Attribute} showSecond
@defvrx {Attribute} showYear
@defvrx {Attribute} yearAbbreviation
The syntax and meaning of these attributes is the same as on the
@code{format} element when that element's @code{baseFormat} is
@code{dateTime}.  @pxref{SPV Detail format Element}.
@end defvr

When @code{baseFormat} is @code{time}, the following attributes are
available.

@defvr {Attribute} hourFormat
@defvrx {Attribute} hourPadding
@defvrx {Attribute} minutePadding
@defvrx {Attribute} monthFormat
@defvrx {Attribute} separatorChars
@defvrx {Attribute} showDay
@defvrx {Attribute} showHour
@defvrx {Attribute} showMinute
@defvrx {Attribute} showMonth
@defvrx {Attribute} showSecond
@defvrx {Attribute} showYear
@defvrx {Attribute} yearAbbreviation
The syntax and meaning of these attributes is the same as on the
@code{format} element when that element's @code{baseFormat} is
@code{elapsedTime}.  @pxref{SPV Detail format Element}.
@end defvr

@node SPV Detail affix Element
@subsubsection The @code{affix} Element

Parent: @code{format} or @code{numberFormat} or @code{stringFormat} @*
Contents: empty

Possibly this element could have @code{dateTimeFormat} as a parent in
a more diverse corpus.

This defines a suffix (or, theoretically, a prefix) for a formatted
value.  It is used to insert a reference to a footnote.  It has the
following attributes:

@defvr {Attribute} definesReference
This specifies the footnote number as a natural number: 1 for the
first footnote, 2 for the second, and so on.
@end defvr

@defvr {Attribute} position
Position for the footnote label.  Always @code{superscript}.
@end defvr

@defvr {Attribute} suffix
Whether the affix is a suffix (@code{true}) or a prefix
(@code{false}).  Always @code{true}.
@end defvr

@defvr {Attribute} value
The text of the suffix or prefix.  Typically a letter, e.g.@: @code{a}
for footnote 1, @code{b} for footnote 2, @enddots{}  The corpus
contains other values: @code{*}, @code{**}, and a few that begin with
at least one comma: @code{,b}, @code{,c}, @code{,,b}, and @code{,,c}.
@end defvr

@node SPV Detail relabel Element
@subsubsection The @code{relabel} Element

Parent: @code{format} or @code{stringFormat} @*
Contents: empty

Possibly this element could have @code{numberFormat} or
@code{dateTimeFormat} as a parent in a more diverse corpus.

This specifies how to display a given value.  It is used to implement
value labels and to display the system-missing value in a
human-readable way.  It has the following attributes:

@defvr {Attribute} from
The value to map.  In the corpus this is an integer or the
system-missing value @code{-1.797693134862316E300}.
@end defvr

@defvr {Attribute} to
The string to display in place of the value of @code{from}.  In the
corpus this is a wide variety of value labels; the system-missing
value is mapped to @samp{.}.
@end defvr

@node SPV Detail union Element
@subsubsection The @code{union} Element

Parent: @code{setCellProperties} @*
Contents: @code{intersect}@math{+}

This element represents a set of cells, computed as the union of the
sets represented by each of its children.

@subsubheading The @code{intersect} Element

Parent: @code{union} @*
Contents: @code{where}@math{+} @math{|} @code{intersectWhere}?

This element represents a set of cells, computed as the intersection
of the sets represented by each of its children.

Of the two possible children, in the corpus @code{where} is far more
common, appearing thousands of times, whereas @code{intersectWhere}
only appears 4 times.

Most @code{intersect} elements have two or more children.

@subsubheading The @code{where} Element

Parent: @code{intersect} @*
Contents: empty

This element represents the set of cells in which the value of a
specified variable falls within a specified set.

@defvr {Attribute} variable
The @code{id} of a variable, e.g.@: @code{dimension0categories} or
@code{dimension0group0map}.
@end defvr

@defvr {Attribute} include
A value, or multiple values separated by semicolons,
e.g.@: @code{0} or @code{13;14;15;16}.
@end defvr

@subsubheading The @code{intersectWhere} Element

Parent: @code{intersect} @*
Contents: empty

The meaning of this element is unknown.

@defvr {Attribute} variable
@defvrx {Attribute} variable2
The meaning of these attributes is unknown.  In the four examples in
the corpus they always take the values @code{dimension2categories} and
@code{dimension0categories}, respectively.
@end defvr

@node SPV Detail style Element
@subsection The @code{style} Element

TBD.