11 Tables

11.1 Introduction to tables

The HTML table model allows authors to arrange data -- text, preformatted text, images, links, forms, form fields, other tables, etc. -- into rows and columns of cells.

Each table may have an associated caption (see the CAPTION element) that provides a short description of the table's purpose. A longer description may also be provided (via the summary attribute) for the benefit of people using speech or Braille-based user agents.

Table rows may be grouped into a head, foot, and body sections, (via the THEAD, TFOOT and TBODY elements, respectively). Row groups convey additional structural information and may be rendered by user agents in ways that emphasize this structure. User agents may exploit the head/body/foot division to support scrolling of body sections independently of the head and foot sections. When long tables are printed, the head and foot information may be repeated on each page that contains table data.

Authors may also group columns to provide additional structural information that may be exploited by user agents. Furthermore, authors may declare column properties at the start of a table definition (via the COLGROUP and COL elements) in a way that enables user agents to render the table incrementally rather than having to wait for all the table data to arrive before rendering.

Table cells may either contain "header" information (see the TH element) or "data" (see the TD element). Cells may span multiple rows and columns. The HTML 4.0 table model allows authors to label each cell so that non-visual user agents may more easily communicate heading information about the cell to the user. Not only do these mechanisms greatly assist users with visual disabilities, they make it possible for multi-modal wireless browsers with limited display capabilities (e.g., Web-enabled pagers and phones) to handle tables.

Tables should not be used purely as a means to layout document content as this may present problems when rendering to non-visual media. Additionally, when used with graphics, these tables may force users to scroll horizontally to view a table designed on a system with a larger display. To minimize these problems, authors should use style sheets to control layout rather than tables.

<TABLE border="1"
          summary="This table gives some statistics about fruit
                   flies: average height and weight, and percentage
                   with red eyes (for both males and females).">
<CAPTION><EM>A test table with merged cells</EM></CAPTION>
<TR><TH rowspan="2"><TH colspan="2">Average
    <TH rowspan="2">Red<BR>eyes
<TR><TH>height<TH>weight
<TR><TH>Males<TD>1.9<TD>0.003<TD>40%
<TR><TH>Females<TD>1.7<TD>0.002<TD>43%
</TABLE>

          A test table with merged cells
    /-----------------------------------------\
    |          |      Average      |   Red    |
    |          |-------------------|  eyes    |
    |          |  height |  weight |          |
    |-----------------------------------------|
    |  Males   | 1.9     | 0.003   |   40%    |
    |-----------------------------------------|
    | Females  | 1.7     | 0.002   |   43%    |
    \-----------------------------------------/

11.2 Elements for constructing tables

11.2.1 The TABLE element

<!ELEMENT TABLE - -
     (CAPTION?, (COL*|COLGROUP*), THEAD?, TFOOT?, TBODY+)>
<!ATTLIST TABLE                        -- table element --
  %attrs;                              -- %coreattrs, %i18n, %events --
  summary     %Text;         #IMPLIED  -- purpose/structure for speech output--
  width       %Length;       #IMPLIED  -- table width --
  border      %Pixels;       #IMPLIED  -- controls frame width around table --
  frame       %TFrame;       #IMPLIED  -- which parts of frame to render --
  rules       %TRules;       #IMPLIED  -- rulings between rows and cols --
  cellspacing %Length;       #IMPLIED  -- spacing between cells --
  cellpadding %Length;       #IMPLIED  -- spacing within cells --
  >

Start tag: required, End tag: required

The TABLE element contains all other elements that specify caption, rows, content, and formatting.

The following informative list describes what operations user agents may carry out when rendering a table:

• Make the table summary available to the user. Authors should provide a summary of a table's content and structure so that people using non-visual user agents may better understand it.
• Render the caption, if one is defined.
• Render the table header, if one is specified. Render the table footer, if one is specified. User agents must know where to render the header and footer. For instance, if the output medium is paged, user agents may put the header at the top of each page and the footer at the bottom. Similarly, if the user agent provides a mechanism to scroll the rows, the header may appear at the top of the scrolled area and the footer at the bottom.
• Calculate the number of columns in the table. Note that the number of rows in a table is equal to the number of TR elements contained by the TABLE element.
• Group the columns according to any column group specifications.
• Render the cells, row by row and grouped in appropriate columns, between the header and footer. Visual user agents should format the table according to HTML attributes and style sheet specification.

The HTML table model has been designed so that, with author assistance, user agents may render tables incrementally (i.e., as table rows arrive) rather than having to wait for all the data before beginning to render.

In order for a user agent to format a table in one pass, authors must tell the user agent:

• The number of columns in the table. Please consult the section on calculating the number of columns in a table for details on how to supply this information.
• The widths of these columns. Please consult the section on calculating the width of columns for details on how to supply this information.

More precisely, a user agent may render a table in a single pass when the column widths are specified using a combination of COLGROUP and COL elements. If any of the columns are specified in relative or percentage terms (see the section on calculating the width of columns), authors must also specify the width of the table itself.

Table directionality 

The directionality of a table is either the inherited directionality (the default is left-to-right) or that specified by the dir attribute for the TABLE element.

For a left-to-right table, column zero is on the left side and row zero is at the top. For a right-to-left table, column zero is on the right side and row zero is at the top.

When a user agent allots extra cells to a row (see the section on calculating the number of columns in a table), extra row cells are added to the right of the table for left-to-right tables and to the left side for right-to-left tables.

Note that TABLE is the only element on which dir reverses the visual order of the columns; a single table row (TR) or a group of columns (COLGROUP) cannot be independently reversed.

When set for the TABLE element, the dir attribute also affects the direction of text within table cells (since the dir attribute is inherited by block-level elements).

<TABLE dir="RTL">
...the rest of the table...
</TABLE>

The direction of text in individual cells can be changed by setting the dir attribute in an element that defines the cell. Please consult the section on bidirectional text for more information on text direction issues.

11.2.2 Table Captions: The CAPTION element

<!ELEMENT CAPTION  - - (%inline;)*     -- table caption -->
<!ENTITY % CAlign "(top|bottom|left|right)">

<!ATTLIST CAPTION
  %attrs;                              -- %coreattrs, %i18n, %events --
  >

Start tag: required, End tag: required

When present, the CAPTION element's text should describe the nature of the table. The CAPTION element is only permitted immediately after the TABLE start tag. A TABLE element may only contain one CAPTION element.

Visual user agents allow sighted people to quickly grasp the structure of the table from the headings as well as the caption. A consequence of this is that captions will often be inadequate as a summary of the purpose and structure of the table from the perspective of people relying on non-visual user agents.

Authors should therefore take care to provide additional information summarizing the purpose and structure of the table using the summary attribute of the TABLE element. This is especially important for tables without captions. Examples below illustrate the use of the summary attribute.

Visual user agents should avoid clipping any part of the table including the caption, unless a means is provided to access all parts, e.g., by horizontal or vertical scrolling. We recommend that the caption text be wrapped to the same width as the table. (See also the section on recommended layout algorithms.)

11.2.3 Row groups: the THEAD, TFOOT, and TBODY elements

<!ELEMENT THEAD    - O (TR)+           -- table header -->
<!ELEMENT TFOOT    - O (TR)+           -- table footer -->

Start tag: required, End tag: optional

<!ELEMENT TBODY    O O (TR)+           -- table body -->

Start tag: optional, End tag: optional

<!ATTLIST (THEAD|TBODY|TFOOT)          -- table section --
  %attrs;                              -- %coreattrs, %i18n, %events --
  %cellhalign;                         -- horizontal alignment in cells --
  %cellvalign;                         -- vertical alignment in cells --
  >

Table rows may be grouped into a table head, table foot, and one or more table body sections, using the THEAD, TFOOT and TBODY elements, respectively. This division enables user agents to support scrolling of table bodies independently of the table head and foot. When long tables are printed, the table head and foot information may be repeated on each page that contains table data.

The table head and table foot should contain information about the table's columns. The table body should contain rows of table data.

When present, each THEAD, TFOOT, and TBODY contains a row group. Each row group must contain at least one row, defined by the TR element.

<TABLE>
<THEAD>
     <TR> ...header information...
</THEAD>
<TFOOT>
     <TR> ...footer information...
</TFOOT>
<TBODY>
     <TR> ...first row of block one data...
     <TR> ...second row of block one data...
</TBODY>
<TBODY>
     <TR> ...first row of block two data...
     <TR> ...second row of block two data...
     <TR> ...third row of block two data...
</TBODY>
</TABLE>

TFOOT must appear before TBODY within a TABLE definition so that user agents can render the foot before receiving all of the (potentially numerous) rows of data. The following summarizes which tags are required and which may be omitted:

• The TBODY start tag is always required except when the table contains only one table body and no table head or foot sections. The TBODY end tag may always be safely omitted.
• The start tags for THEAD and TFOOT are required when the table head and foot sections are present respectively, but the corresponding end tags may always be safely omitted.

Conforming user agent parsers must obey these rules for reasons of backward compatibility.

<TABLE>
<THEAD>
     <TR> ...header information...
<TFOOT>
     <TR> ...footer information...
<TBODY>
     <TR> ...first row of block one data...
     <TR> ...second row of block one data...
<TBODY>
     <TR> ...first row of block two data...
     <TR> ...second row of block two data...
     <TR> ...third row of block two data...
</TABLE>

The THEAD, TFOOT, and TBODY sections must contain the same number of columns.

11.2.4 Column groups: the COLGROUP and COL elements

Column groups allow authors to create structural divisions within a table. Authors may highlight this structure through style sheets or HTML attributes (e.g., the rules attribute for the TABLE element). For an example of the visual presentation of column groups, please consult the sample table.

A table may either contain a single implicit column group (no COLGROUP element delimits the columns) or any number of explicit column groups (each delimited by an instance of the COLGROUP element).

The COL element allows authors to share attributes among several columns without implying any structural grouping. The "span" of the COL element is the number of columns that will share the element's attributes.

The COLGROUP element 

<!ELEMENT COLGROUP - O (col)*          -- table column group -->
<!ATTLIST COLGROUP
  %attrs;                              -- %coreattrs, %i18n, %events --
  span        NUMBER         1         -- default number of columns in group --
  width       %MultiLength;  #IMPLIED  -- default width for enclosed COLs --
  %cellhalign;                         -- horizontal alignment in cells --
  %cellvalign;                         -- vertical alignment in cells --
  >

Start tag: required, End tag: optional

The COLGROUP element creates an explicit column group. The number of columns in the column group may be specified in two, mutually exclusive ways:

1. The element's span attribute (default value 1) specifies the number of columns in the group.
2. Each COL element in the COLGROUP represents one or more columns in the group.

The advantage of using the span attribute is that authors may group together information about column widths. Thus, if a table contains forty columns, all of which have a width of 20 pixels, it is easier to write:

   <COLGROUP span="40" width="20">
   </COLGROUP>

than:

   <COLGROUP>
      <COL width="20">
      <COL width="20">
      ...a total of forty COL elements...
   </COLGROUP>

When it is necessary to single out a column (e.g., for style information, to specify width information, etc.) within a group, authors must identify that column with a COL element. Thus, to apply special style information to the last column of the previous table, we single it out as follows:

   <COLGROUP width="20">
      <COL span="39">
      <COL id="format-me-specially">
   </COLGROUP>

The width attribute of the COLGROUP element is inherited by all 40 columns. The first COL element refers to the first 39 columns (doing nothing special to them) and the second one assigns an id value to the fortieth columns so that style sheets may refer to it.

<TABLE>
<COLGROUP span="10" width="50">
<COLGROUP span="5" width="0*">
<THEAD>
<TR><TD> ...
</TABLE>

The COL element 

<!ELEMENT COL      - O EMPTY           -- table column -->
<!ATTLIST COL                          -- column groups and properties --
  %attrs;                              -- %coreattrs, %i18n, %events --
  span        NUMBER         1         -- COL attributes affect N columns --
  width       %MultiLength;  #IMPLIED  -- column width specification --
  %cellhalign;                         -- horizontal alignment in cells --
  %cellvalign;                         -- vertical alignment in cells --
  >

Start tag: required, End tag: forbidden

The COL element allows authors to group together attribute specifications for table columns. The COL does not group columns together structurally -- that is the role of the COLGROUP element. COL elements are empty and serve only as a support for attributes. They may appear inside or outside an explicit column group (i.e., COLGROUP element).

The width attribute for COL refers to the width of each column in the element's span.

Calculating the number of columns in a table 

There are two ways to determine the number of columns in a table (in order of precedence):

1. If the TABLE element contains any COLGROUP or COL elements, user agents should calculate the number of columns by summing the following:
   • For each COL element, take the value of its span attribute (default value 1).
   • For each COLGROUP element containing at least one COL element, ignore the span attribute for the COLGROUP element. For each COL element, perform the calculation of step 1.
   • For each empty COLGROUP element, take the value of its span attribute (default value 1).
2. Otherwise, if the TABLE element contains no COLGROUP or COL elements, user agents should base the number of columns on what is required by the rows. The number of columns is equal to the number of columns required by the row with the most columns, including cells that span multiple columns. For any row that has fewer than this number of columns, the end of that row should be padded with empty cells. The "end" of a row depends on the table directionality.

It is an error if a table contains COLGROUP or COL elements and the two calculations do not result in the same number of columns.

Once the user agent has calculated the number of columns in the table, it may group them into column groups.

<TABLE>
<COLGROUP span="3"></COLGROUP>
<TR><TD> ...
...rows...
</TABLE>

<TABLE>
<COLGROUP>
<COL>
<COL span="2">
</COLGROUP>
<TR><TD> ...
...rows...
</TABLE>

<TABLE>
<COLGROUP>
<COL>
</COLGROUP>
<COLGROUP span="2">
<TR><TD> ...
...rows...
</TABLE>

<TABLE>
<TR>
  <TD><TD><TD>
</TR>
</TABLE>

Calculating the width of columns 

Authors may specify column widths in three ways:

Fixed

A fixed width specification is given in pixels (e.g., width="30"). A fixed-width specification enables incremental rendering.

Percentage

A percentage specification (e.g., width="20%") is based on the percentage of the horizontal space available to the table (between the current left and right margins, including floats). Note that this space does not depend on the table itself, and thus percentage specifications enable incremental rendering.

Proportional

Proportional specifications (e.g., width="3*") refer to portions of the horizontal space required by a table. If the table width is given a fixed value via the width attribute of the TABLE element, user agents may render the table incrementally even with proportional columns.

However, if the table does not have a fixed width, user agents must receive all table data before they can determine the horizontal space required by the table. Only then may this space be allotted to proportional columns.

If an author specifies no width information for a column, a user agent may not be able to incrementally format the table since it must wait for the entire column of data to arrive in order to allot an appropriate width.

If column widths prove to be too narrow for the contents of a particular table cell, user agents may choose to reflow the table.

    
<TABLE>
<COLGROUP>
   <COL width="30">
<COLGROUP>
   <COL width="30">
   <COL width="0*">
   <COL width="2*">
<COLGROUP align="center">
   <COL width="1*">
   <COL width="3*" align="char" char=":">
<THEAD>
<TR><TD> ...
...rows...
</TABLE>

    
<TABLE width="200">
<COLGROUP span="10" width="15">
<COLGROUP width="*">
   <COL id="penultimate-column">
   <COL id="last-column">
<THEAD>
<TR><TD> ...
...rows...
</TABLE>

11.2.5 Table rows: The TR element

<!ELEMENT TR       - O (TH|TD)+        -- table row -->
<!ATTLIST TR                           -- table row --
  %attrs;                              -- %coreattrs, %i18n, %events --
  %cellhalign;                         -- horizontal alignment in cells --
  %cellvalign;                         -- vertical alignment in cells --
  >

Start tag: required, End tag: optional

The TR elements acts as a container for a row of table cells. The end tag may be omitted.

<TABLE summary="This table charts the number of cups
                   of coffee consumed by each senator, the type 
                   of coffee (decaf or regular), and whether 
                   taken with sugar.">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR> ...A header row...
<TR> ...First row of data...
<TR> ...Second row of data...
...the rest of the table...
</TABLE>

11.2.6 Table cells: The TH and TD elements

<!ELEMENT (TH|TD)  - O (%flow;)*       -- table header cell, table data cell-->

<!-- Scope is simpler than axes attribute for common tables -->
<!ENTITY % Scope "(row|col|rowgroup|colgroup)">

<!-- TH is for headers, TD for data, but for cells acting as both use TD -->
<!ATTLIST (TH|TD)                      -- header or data cell --
  %attrs;                              -- %coreattrs, %i18n, %events --
  abbr        %Text;         #IMPLIED  -- abbreviation for header cell --
  axis        CDATA          #IMPLIED  -- names groups of related headers--
  headers     IDREFS         #IMPLIED  -- list of id's for header cells --
  scope       %Scope;        #IMPLIED  -- scope covered by header cells --
  rowspan     NUMBER         1         -- number of rows spanned by cell --
  colspan     NUMBER         1         -- number of cols spanned by cell --
  %cellhalign;                         -- horizontal alignment in cells --
  %cellvalign;                         -- vertical alignment in cells --
  >

Start tag: required, End tag: optional

Table cells may contain two types of information: header information and data. This distinction enables user agents to render header and data cells distinctly, even in the absence of style sheets. For example, visual user agents may present header cell text with a bold font. Speech synthesizers may render header information with a distinct voice inflection.

The TH element defines a cell that contains header information. User agents have two pieces of header information available: the contents of the TH element and the value of the abbr attribute. User agents must render either the contents of the cell or the value of the abbr attribute. For visual media, the latter may be appropriate when there is insufficient space to render the full contents of the cell. For non-visual media abbr may be used as an abbreviation for table headers when these are rendered along with the contents of the cells to which they apply.

The headers and scope attributes also allow authors to help non-visual user agents process header information. Please consult the section on labeling cells for non-visual user agents for information and examples.

The TD element defines a cell that contains data.

Cells may be empty (i.e., contain no data).

<TABLE summary="This table charts the number of cups
                   of coffee consumed by each senator, the type 
                   of coffee (decaf or regular), and whether 
                   taken with sugar.">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR>
   <TH>Name</TH>
   <TH>Cups</TH>
   <TH>Type of Coffee</TH>
   <TH>Sugar?</TH>
<TR>
   <TD>T. Sexton</TD>
   <TD>10</TD>
   <TD>Espresso</TD>
   <TD>No</TD>
<TR>
   <TD>J. Dinnen</TD>
   <TD>5</TD>
   <TD>Decaf</TD>
   <TD>Yes</TD>
</TABLE>

Name         Cups       Type of Coffee   Sugar?
T. Sexton    10         Espresso         No
J. Dinnen    5          Decaf            Yes

Cells that span several rows or columns 

Cells may span several rows or columns. The number of rows or columns spanned by a cell is set by the rowspan and colspan attributes for the TH and TD elements.

<TABLE border="1">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR><TH>Name<TH>Cups<TH>Type of Coffee<TH>Sugar?
<TR><TD>T. Sexton<TD>10<TD>Espresso<TD>No
<TR><TD>J. Dinnen<TD>5<TD>Decaf<TD>Yes
<TR><TD>A. Soria<TD colspan="3"><em>Not available</em>
</TABLE>

Cups of coffee consumed by each senator
 --------------------------------------
 |   Name  |Cups|Type of Coffee|Sugar?|
 --------------------------------------
 |T. Sexton|10  |Espresso      |No    |
 --------------------------------------
 |J. Dinnen|5   |Decaf         |Yes   |
 --------------------------------------
 |A. Soria |Not available             |
 --------------------------------------

<TABLE border="1">
<TR><TD>1 <TD rowspan="2">2 <TD>3
<TR><TD>4 <TD>6
<TR><TD>7 <TD>8 <TD>9
</TABLE>

-------------
| 1 | 2 | 3 | 
----|   |----
| 4 |   | 6 |
----|---|----
| 7 | 8 | 9 |
-------------

<TABLE border="1">
<TR><TD>1 <TD>2 <TD>3
<TR><TD colspan="2">4 <TD>6
<TR><TD>7 <TD>8 <TD>9
</TABLE>

-------------
| 1 | 2 | 3 | 
--------|----
| 4     | 6 |
--------|----
| 7 | 8 | 9 |
-------------

Defining overlapping cells is an error. User agents may vary in how they handle this error (e.g., rendering may vary).

<TABLE border="1">
<TR><TD>1 <TD>2 <TD>3
<TR><TD>4 <TD rowspan="2">5 <TD>6
<TR><TD colspan="2">7 <TD>9
</TABLE>

11.3 Table formatting by visual user agents

HTML 4.0 includes mechanisms to control:

• border styles
• horizontal and vertical alignment of cell contents
• and cell margins

11.3.1 Borders and rules

The following attributes affect a table's external frame and internal rules.

<TABLE border="1" 
       summary="This table charts the number of cups
                of coffee consumed by each senator, the type 
                of coffee (decaf or regular), and whether 
                taken with sugar.">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR>
   <TH>Name</TH>
   <TH>Cups</TH>
   <TH>Type of Coffee</TH>
   <TH>Sugar?</TH>
<TR>
   <TD>T. Sexton</TD>
   <TD>10</TD>
   <TD>Espresso</TD>
   <TD>No</TD>
<TR>
   <TD>J. Dinnen</TD>
   <TD>5</TD>
   <TD>Decaf</TD>
   <TD>Yes</TD>
</TABLE>

<TABLE border="5" frame="vsides" rules="cols">
<TR> <TD>1 <TD>2 <TD>3
<TR> <TD>4 <TD>5 <TD>6
<TR> <TD>7 <TD>8 <TD>9
</TABLE>

The following settings should be observed by user agents for backwards compatibility.

• Setting border="0" implies frame="void" and, unless otherwise specified, rules="none".
• Other values of border imply frame="border" and, unless otherwise specified, rules="all".
• The value "border" in the start tag of the TABLE element should be interpreted as the value of the frame attribute. It implies rules="all" and some default (non-zero) value for the border attribute.

<TABLE border="2">
<TABLE border="2" frame="border" rules="all">

<TABLE border>
<TABLE frame="border" rules="all">

11.3.2 Horizontal and vertical alignment

The following attributes may be set for different table elements (see their definitions).

<!-- horizontal alignment attributes for cell contents -->
<!ENTITY % cellhalign
  "align      (left|center|right|justify|char) #IMPLIED
   char       %Character;    #IMPLIED  -- alignment char, e.g. char=':' --
   charoff    %Length;       #IMPLIED  -- offset for alignment char --"
  >
<!-- vertical alignment attributes for cell contents -->
<!ENTITY % cellvalign
  "valign     (top|middle|bottom|baseline) #IMPLIED"
  >

<TABLE border="1">
<COLGROUP>
<COL><COL align="char" char=".">
<THEAD>
<TR><TH>Vegetable <TH>Cost per kilo
<TBODY>
<TR><TD>Lettuce        <TD>$1
<TR><TD>Silver carrots <TD>$10.50
<TR><TD>Golden turnips <TD>$100.30
</TABLE>

------------------------------
|   Vegetable  |Cost per kilo|
|--------------|-------------|
|Lettuce       |        $1   |
|--------------|-------------|
|Silver carrots|       $10.50|
|--------------|-------------|
|Golden turnips|      $100.30|
------------------------------

When the contents of a cell contain more than one instance of the alignment character specified by char and the contents wrap, user agent behavior is undefined. Authors should therefore be attentive in their use of char.

Inheritance of alignment specifications 

The alignment of cell contents can be specified on a cell by cell basis, or inherited from enclosing elements, such as the row, column or the table itself.

The order of precedence (from highest to lowest) for the attributes align, char, and charoff is the following:

1. An alignment attribute set on an element within a cell's data (e.g., P).
2. An alignment attribute set on a cell (TH and TD).
3. An alignment attribute set on a column grouping element (COL and COLGROUP). When a cell is part of a multi-column span, the alignment property is inherited from the cell definition at the beginning of the span.
4. An alignment attribute set on a row or row grouping element (TR, THEAD, TFOOT, and TBODY). When a cell is part of a multi-row span, the alignment property is inherited from the cell definition at the beginning of the span.
5. An alignment attribute set on the table (TABLE).
6. The default alignment value.

The order of precedence (from highest to lowest) for the attribute valign (as well as the other inherited attributes lang, dir, and style) is the following:

1. An attribute set on an element within a cell's data (e.g., P).
2. An attribute set on a cell (TH and TD).
3. An attribute set on a row or row grouping element (TR, THEAD, TFOOT, and TBODY). When a cell is part of a multi-row span, the attribute value is inherited from the cell definition at the beginning of the span.
4. An attribute set on a column grouping element (COL and COLGROUP). When a cell is part of a multi-column span, the attribute value is inherited from the cell definition at the beginning of the span.
5. An attribute set on the table (TABLE).
6. The default attribute value.

Furthermore, when rendering cells, horizontal alignment is determined by columns in preference to rows, while for vertical alignment, rows are given preference over columns.

The default alignment for cells depends on the user agent. However, user agents should substitute the default attribute for the current directionality (i.e., not just "left" in all cases).

User agents that do not support the "justify" value of the align attribute should use the value of the inherited directionality in its place.

11.3.3 Cell margins

These two attributes control spacing between and within cells. The following illustration explains how they relate:

<TABLE cellspacing="20" cellpadding="20%">
<TR> <TD>Data1 <TD>Data2 <TD>Data3
</TABLE>

If a table or given column has a fixed width, cellspacing and cellpadding may demand more space than assigned. User agents may give these attributes precedence over the width attribute when a conflict occurs, but are not required to.

11.4 Table rendering by non-visual user agents

11.4.1 Associating header information with data cells

Non-visual user agents such as speech synthesizers and Braille-based devices may use the following TD and TH element attributes to render table cells more intuitively:

• For a given data cell, the headers attribute lists which cells provide pertinent header information. For this purpose, each header cell must be named using the id attribute. Note that its not always possible to make a clean division of cells into headers or data. You should use the TD element for such cells together with the id or scope attributes as appropriate.
• For a given header cell, the scope attribute tells the user agent the data cells for which this header provides information. Authors may choose to use this attribute instead of headers according to which is more convenient; the two attributes fulfill the same function. The headers attribute is generally needed when headers are placed in irregular positions with respect to the data they apply to.
• The abbr attribute specifies an abbreviated header for header cells so that user agents may render header information more rapidly.

<TABLE border="1" 
       summary="This table charts the number of cups
                of coffee consumed by each senator, the type 
                of coffee (decaf or regular), and whether 
                taken with sugar.">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR>
   <TH id="t1">Name</TH>
   <TH id="t2">Cups</TH>
   <TH id="t3" abbr="Type">Type of Coffee</TH>
   <TH id="t4">Sugar?</TH>
<TR>
   <TD headers="t1">T. Sexton</TD>
   <TD headers="t2">10</TD>
   <TD headers="t3">Espresso</TD>
   <TD headers="t4">No</TD>
<TR>
   <TD headers="t1">J. Dinnen</TD>
   <TD headers="t2">5</TD>
   <TD headers="t3">Decaf</TD>
   <TD headers="t4">Yes</TD>
</TABLE>

Caption: Cups of coffee consumed by each senator
Summary: This table charts the number of cups
         of coffee consumed by each senator, the type 
         of coffee (decaf or regular), and whether 
         taken with sugar.
Name: T. Sexton,   Cups: 10,   Type: Espresso,   Sugar: No
Name: J. Dinnen,   Cups: 5,    Type: Decaf,      Sugar: Yes

<TABLE border="1" 
       summary="This table charts the number of cups
                of coffee consumed by each senator, the type 
                of coffee (decaf or regular), and whether 
                taken with sugar.">
<CAPTION>Cups of coffee consumed by each senator</CAPTION>
<TR>
   <TH scope="col">Name</TH>
   <TH scope="col">Cups</TH>
   <TH scope="col" abbr="Type">Type of Coffee</TH>
   <TH scope="col">Sugar?</TH>
<TR>
   <TD>T. Sexton</TD>
   <TD>10</TD>
   <TD>Espresso</TD>
   <TD>No</TD>
<TR>
   <TD>J. Dinnen</TD>
   <TD>5</TD>
   <TD>Decaf</TD>
   <TD>Yes</TD>
</TABLE>

<TABLE border="1" cellpadding="5" cellspacing="2"
  summary="History courses offered in the community of
           Bath arranged by course name, tutor, summary, 
           code, and fee">
  <TR>
    <TH colspan="5" scope="colgroup">Community Courses -- Bath Autumn 1997</TH>
  </TR>
  <TR>
    <TH scope="col" abbr="Name">Course Name</TH>
    <TH scope="col" abbr="Tutor">Course Tutor</TH>
    <TH scope="col">Summary</TH>
    <TH scope="col">Code</TH>
    <TH scope="col">Fee</TH>
  </TR>
  <TR>
    <TD scope="row">After the Civil War</TD>
    <TD>Dr. John Wroughton</TD>
    <TD>
       The course will examine the turbulent years in England
       after 1646. <EM>6 weekly meetings starting Monday 13th
      October.</EM>
    </TD>
    <TD>H27</TD>
    <TD>&pound;32</TD>
  </TR>
  <TR>
    <TD scope="row">An Introduction to Anglo-Saxon England</TD>
    <TD>Mark Cottle</TD>
    <TD>
       One day course introducing the early medieval
       period reconstruction the Anglo-Saxons and
       their society. <EM>Saturday 18th October.</EM>
    </TD>
    <TD>H28</TD>
    <TD>&pound;18</TD>
  </TR>
  <TR>
    <TD scope="row">The Glory that was Greece</TD>
    <TD>Valerie Lorenz</TD>
    <TD>
     Birthplace of democracy, philosophy, heartland of theater, home of
     argument. The Romans may have done it but the Greeks did it
     first. <EM>Saturday day school 25th October 1997</EM>
    </TD>
    <TD>H30</TD>
    <TD>&pound;18</TD>
  </TR>
</TABLE>

11.4.2 Categorizing cells

Users browsing a table with a speech-based user agent may wish to hear an explanation of a cell's contents in addition to the contents themselves. One way the user might provide an explanation is by speaking associated header information before speaking the data cell's contents (see the section on associating header information with data cells).

Users may also want information about more than one cell, in which case header information provided at the cell level (by headers, scope, and abbr) may not provide adequate context. Consider the following table, which classifies expenses for meals, hotels, and transport in two locations (San Jose and Seattle) over several days:

Users might want to extract information from the table in the form of queries:

• "What did I spend for all my meals?"
• "What did I spend for meals on 25 August?"
• "What did I spend for all expenses in San Jose?"

Each query involves a computation by the user agent that may involve zero or more cells. In order to determine, for example, the costs of meals on 25 August, the user agent must know which table cells refer to "Meals" (all of them) and which refer to "Dates" (specifically, 25 August), and find the intersection of the two sets.

To accommodate this type of query, the HTML 4.0 table model allows authors to place cell headers and data into categories. For example, for the travel expense table, an author could group the header cells "San Jose" and "Seattle" into the category "Location", the headers "Meals", "Hotels", and "Transport" in the category "Expenses", and the four days into the category "Date". The previous three questions would then have the following meanings:

• "What did I spend for all my meals?" means "What are all the data cells in the "Expenses=Meals" category?
• "What did I spend for meals on 25 August?" means "What are all the data cells in the "Expenses=Meals" and "Date=Aug-25-1997" categories?
• "What did I spend for all expenses in San Jose?" means "What are all the data cells in the "Expenses=Meals, Hotels, Transport" and "Location=San Jose" categories?

Authors categorize a header or data cell by setting the axis attribute for the cell. For instance, in the travel expense table, the cell containing the information "San Jose" could be placed in the "Location" category as follows:

  <TH id="a6" axis="location">San Jose</TH>

Any cell containing information related to "San Jose" should refer to this header cell via either the headers or the scope attribute. Thus, meal expenses for 25-Aug-1997 should be marked up to refer to id attribute (whose value here is "a6") of the "San Jose" header cell:

  
  <TD headers="a6">37.74</TD>

Each headers attribute provides a list of id references. Authors may thus categorize a given cell in any number of ways (or, along any number of "headers", hence the name).

Below we mark up the travel expense table with category information:

<TABLE border="1"
          summary="This table summarizes travel expenses
                   incurred during August trips to
                   San Jose and Seattle">
<CAPTION>
  Travel Expense Report
</CAPTION>
<TR>
  <TH></TH>
  <TH id="a2" axis="expenses">Meals</TH>
  <TH id="a3" axis="expenses">Hotels</TH>
  <TH id="a4" axis="expenses">Transport</TH>
  <TD>subtotals</TD>
</TR>
<TR>
  <TH id="a6" axis="location">San Jose</TH>
  <TH></TH>
  <TH></TH>
  <TH></TH>
  <TD></TD>
</TR>
<TR>
  <TD id="a7" axis="date">25-Aug-97</TD>
  <TD headers="a6 a7 a2">37.74</TD>
  <TD headers="a6 a7 a3">112.00</TD>
  <TD headers="a6 a7 a4">45.00</TD>
  <TD></TD>
</TR>
<TR>
  <TD id="a8" axis="date">26-Aug-97</TD>
  <TD headers="a6 a8 a2">27.28</TD>
  <TD headers="a6 a8 a3">112.00</TD>
  <TD headers="a6 a8 a4">45.00</TD>
  <TD></TD>
</TR>
<TR>
  <TD>subtotals</TD>
  <TD>65.02</TD>
  <TD>224.00</TD>
  <TD>90.00</TD>
  <TD>379.02</TD>
</TR>
<TR>
  <TH id="a10" axis="location">Seattle</TH>
  <TH></TH>
  <TH></TH>
  <TH></TH>
  <TD></TD>
</TR>
<TR>
  <TD id="a11" axis="date">27-Aug-97</TD>
  <TD headers="a10 a11 a2">96.25</TD>
  <TD headers="a10 a11 a3">109.00</TD>
  <TD headers="a10 a11 a4">36.00</TD>
  <TD></TD>
</TR>
<TR>
  <TD id="a12" axis="date">28-Aug-97</TD>
  <TD headers="a10 a12 a2">35.00</TD>
  <TD headers="a10 a12 a3">109.00</TD>
  <TD headers="a10 a12 a4">36.00</TD>
  <TD></TD>
</TR>
<TR>
  <TD>subtotals</TD>
  <TD>131.25</TD>
  <TD>218.00</TD>
  <TD>72.00</TD>
  <TD>421.25</TD>
</TR>
<TR>
  <TH>Totals</TH>
  <TD>196.27</TD>
  <TD>442.00</TD>
  <TD>162.00</TD>
  <TD>800.27</TD>
</TR>
</TABLE>

Note that marking up the table this way also allows user agents to avoid confusing the user with unwanted information. For instance, if a speech synthesizer were to speak all of the figures in the "Meals" column of this table in response to the query "What were all my meal expenses?", a user would not be able to distinguish a day's expenses from subtotals or totals. By carefully categorizing cell data, authors allow user agents to make important semantic distinctions when rendering.

Of course, there is no limit to how authors may categorize information in a table. In the travel expense table, for example, we could add the additional categories "subtotals" and "totals".

This specification does not require user agents to handle information provided by the axis attribute, nor does it make any recommendations about how user agents may present axis information to users or how users may query the user agent about this information.

However, user agents, particularly speech synthesizers, may want to factor out information common to several cells that are the result of a query. For instance, if the user asks "What did I spend for meals in San Jose?", the user agent would first determine the cells in question (25-Aug-1997: 37.74, 26-Aug-1997:27.28), then render this information. A user agent speaking this information might read it:

   Location: San Jose. Date: 25-Aug-1997. Expenses, Meals: 37.74
   Location: San Jose. Date: 26-Aug-1997. Expenses, Meals: 27.28

or, more compactly:

   San Jose, 25-Aug-1997, Meals: 37.74
   San Jose, 26-Aug-1997, Meals: 27.28

An even more economical rendering would factor the common information and reorder it:

   San Jose, Meals, 25-Aug-1997: 37.74
                    26-Aug-1997: 27.28

User agents that support this type of rendering should allow user agents a means to customize rendering (e.g., through style sheets).

11.4.3 Algorithm to find heading information

In the absence of header information from either the scope or headers attribute, user agents may construct header information according to the following algorithm. The goal of the algorithm is to find an ordered list of headers. (In the following description of the algorithm the table directionality is assumed to be left-to-right.)

• First, search left from the cell's position to find row header cells. Then search upwards to find column header cells. The search in a given direction stops when the edge of the table is reached or when a data cell is found after a header cell.
• Row headers are inserted into the list in the order they appear in the table. For left-to-right tables, headers are inserted from left to right.
• Column headers are inserted after row headers, in the order they appear in the table, from top to bottom.
• If a header cell has the headers attribute set, then the headers referenced by this attribute are inserted into the list and the search stops for the current direction.
• TD cells that set the axis attribute are also treated as header cells.

11.5 Sample table

This sample illustrates grouped rows and columns. The example is adapted from "Developing International Software", by Nadine Kano.

<TABLE border="2" frame="hsides" rules="groups"
          summary="Code page support in different versions
                   of MS Windows.">
<CAPTION>CODE-PAGE SUPPORT IN MICROSOFT WINDOWS</CAPTION>
<COLGROUP align="center">
<COLGROUP align="left">
<COLGROUP align="center" span="2">
<COLGROUP align="center" span="3">
<THEAD valign="top">
<TR>
<TH>Code-Page<BR>ID
<TH>Name
<TH>ACP
<TH>OEMCP
<TH>Windows<BR>NT 3.1
<TH>Windows<BR>NT 3.51
<TH>Windows<BR>95
<TBODY>
<TR><TD>1200<TD>Unicode (BMP of ISO/IEC-10646)<TD><TD><TD>X<TD>X<TD>*
<TR><TD>1250<TD>Windows 3.1 Eastern European<TD>X<TD><TD>X<TD>X<TD>X
<TR><TD>1251<TD>Windows 3.1 Cyrillic<TD>X<TD><TD>X<TD>X<TD>X
<TR><TD>1252<TD>Windows 3.1 US (ANSI)<TD>X<TD><TD>X<TD>X<TD>X
<TR><TD>1253<TD>Windows 3.1 Greek<TD>X<TD><TD>X<TD>X<TD>X
<TR><TD>1254<TD>Windows 3.1 Turkish<TD>X<TD><TD>X<TD>X<TD>X
<TR><TD>1255<TD>Hebrew<TD>X<TD><TD><TD><TD>X
<TR><TD>1256<TD>Arabic<TD>X<TD><TD><TD><TD>X
<TR><TD>1257<TD>Baltic<TD>X<TD><TD><TD><TD>X
<TR><TD>1361<TD>Korean (Johab)<TD>X<TD><TD><TD>**<TD>X
<TBODY>
<TR><TD>437<TD>MS-DOS United States<TD><TD>X<TD>X<TD>X<TD>X
<TR><TD>708<TD>Arabic (ASMO 708)<TD><TD>X<TD><TD><TD>X
<TR><TD>709<TD>Arabic (ASMO 449+, BCON V4)<TD><TD>X<TD><TD><TD>X
<TR><TD>710<TD>Arabic (Transparent Arabic)<TD><TD>X<TD><TD><TD>X
<TR><TD>720<TD>Arabic (Transparent ASMO)<TD><TD>X<TD><TD><TD>X
</TABLE>

                  CODE-PAGE SUPPORT IN MICROSOFT WINDOWS
===============================================================================
Code-Page | Name                         | ACP  OEMCP | Windows Windows Windows
    ID    |                              |            |  NT 3.1 NT 3.51    95
-------------------------------------------------------------------------------
   1200   | Unicode (BMP of ISO 10646)   |            |    X       X       *
   1250   | Windows 3.1 Eastern European |  X         |    X       X       X
   1251   | Windows 3.1 Cyrillic         |  X         |    X       X       X
   1252   | Windows 3.1 US (ANSI)        |  X         |    X       X       X
   1253   | Windows 3.1 Greek            |  X         |    X       X       X
   1254   | Windows 3.1 Turkish          |  X         |    X       X       X
   1255   | Hebrew                       |  X         |                    X
   1256   | Arabic                       |  X         |                    X
   1257   | Baltic                       |  X         |                    X
   1361   | Korean (Johab)               |  X         |            **      X
-------------------------------------------------------------------------------
    437   | MS-DOS United States         |        X   |    X       X       X
    708   | Arabic (ASMO 708)            |        X   |                    X
    709   | Arabic (ASMO 449+, BCON V4)  |        X   |                    X
    710   | Arabic (Transparent Arabic)  |        X   |                    X
    720   | Arabic (Transparent ASMO)    |        X   |                    X
===============================================================================