table - Create and manipulate tables

table pathName ?options? -anchor -background -cursor -exportselection -font -foreground -highlightbackground -highlightcolor -highlightthickness -insertbackground -insertborderwidth -insertofftime -insertontime -insertwidth -invertselected -relief -takefocus -xscrollcommand -yscrollcommand

A boolean value which specifies whether the first keypress in a cell will delete whatever text was previously there. Defaults to 0. Specifies the name of the cursor to show when over borders, a visual indication that interactive resizing is allowed (it is thus affect by the value of -resizeborders). Defaults to crosshair. Specifies a non-negative pixel value or list of values indicating the width of the 3-D border to draw on interior table cells (if such a border is being drawn; the relief option typically determines this). If one value is specified, a rectangle of this width will be drawn. If two values are specified, then only the left and right edges of the cell will have borders. If four values are specified, then the values correspond to the {left right top bottom} edges. This can be overridden by the a tag's borderwidth option. It can also be affected by the defined -drawmode for the table. Each value in the list must have one of the forms acceptable to Tk_GetPixels. Specifies a command which will be evaluated anytime the active cell changes. It uses the %-substition model described in COMMAND SUBSTITUTION below. A boolean value that specifies whether an internal cache of the table contents should be kept. This greatly enhances speed performance when used with -command but uses extra memory. Can maintain state when both -command and -variable are empty. The cache is automatically flushed whenever the value of -cache or -variable changes, otherwise you have to explicitly call clear on it. Defaults to off. Specifies what column index to interpret as the leftmost column in the table. This value is used for user indices in the table. Defaults to 0. Number of cols in the table. Defaults to 10. Specifies a separator character that will be interpreted as the column separator when cutting or pasting data in a table. By default, columns are separated as elements of a tcl list. Specifies one of the following stretch modes for columns to fill extra allocated window space:

none

Columns will not stretch to fill the assigned window space. If the columns are too narrow, there will be a blank space at the right of the table. This is the default.

unset

Only columns that do not have a specific width set will be stretched.

all

All columns will be stretched by the same number of pixels to fill the window space allocated to the table. This mode can interfere with interactive border resizing which tries to force column width.

last

The last column will be stretched to fill the window space allocated to the table.

fill (only valid for -rowstretch currently)

The table will get more or less columns according to the window space allocated to the table. This mode has numerous quirks and may disappear in the future. Provides the name of a procedure that will be evaluated by the widget to determine the tag to be used for a given column. When displaying a cell, the table widget will first check to see if a tag has been defined using the tag col widget method. If no tag is found, it will evaluate the named procedure passing the column number in question as the sole argument. The procedure is expected to return the name of a tag to use, or a null string. Errors occuring during the evaluation of the procedure, or the return of an invalid tag name are silently ignored. Default column width, interpreted as characters in the default font when the number is positive, or pixels if it is negative. Defaults to 10. Specified a command to use as a procedural interface to cell values. If -usecommand is true, this command will be used instead of any reference to the -variable array. When retrieving cell values, the return value of the command is used as the value for the cell. It uses the %-substition model described in COMMAND SUBSTITUTION below. Sets the table drawing mode to one of the following options:

slow

The table is drawn to an offscreen pixmap using the Tk bordering functions (double-buffering). This means there will be no flashing, but this mode is slow for larger tables.

compatible

The table is drawn directly to the screen using the Tk border functions. It is faster, but the screen may flash on update. This is the default.

fast

The table is drawn directly to the screen and the borders are done with fast X calls, so they are always one pixel wide only. As a side effect, it restricts -borderwidth to a range of 0 or 1. This mode provides best performance for large tables, but can flash on redraw and is not 100% Tk compatible on the border mode.

single

The table is drawn to the screen as in fast mode, but only single pixel lines are drawn (not square borders). This specifies a string to display at the end of a line that would be clipped by its cell, like ``...''. An ellipsis will be displayed only on non-wrapping, non-multiline cells that would be clipped. The ellipsis will display on the left for east anchored cells, otherwise it displays on the right. Defaults to "" (no ellipsis). A boolean value which specifies whether cells should flash when their value changes. The table tag flash will be applied to these cells for the duration specified by -flashtime. Defaults to 0. The amount of time, in 1/4 second increments, for which a cell should flash when its value has changed. -flashmode must be on. Defaults to 2. Specifies the desired height for the window, in rows. If zero or less, then the desired height for the window is made just large enough to hold all the rows in the table. The height can be further limited by -maxheight. Specifies whether the foreground and background of an item should simply have their values swapped instead of merging the sel tag options when the cell is selected. Defaults to 0 (merge sel tag). A pixel value specifying the internal offset X padding for text in a cell. This value does not grow the size of the cell, it just causes the text to be drawn further from the cell border. It only affects one side (depending on anchor). Defaults to 0. See -padx for an alternate padding style. A pixel value specifying the internal offset Y padding for text in a cell. This value does not grow the size of the cell, it just causes the text to be drawn further from the cell border. It only affects one side (depending on anchor). Defaults to 0. See -pady for an alternate padding style. How to justify multi-line text in a cell. It must be one of left, right, or center. Defaults to left. The max height in pixels that the window will request. Defaults to 600. The max width in pixels that the window will request. Defaults to 800. Specifies the default setting for the multiline tag option. Defaults to 1. A pixel value specifying the offset X padding for a cell. This value causes the default size of the cell to increase by two times the value (one for each side), unless a specific pixel size is chosen for the cell with the width command. This will force an empty area on the left and right of each cell edge. This padding affects all types of data in the cell. Defaults to 0. See -ipadx for an alternate padding style. A pixel value specifying the offset Y padding for a cell. This value causes the default size of the cell to increase by two times the value (one for each side), unless a specific pixel size is chosen for the cell with the height command. This will force an empty area on the top and bottom of each cell edge. This padding affects all types of data in the cell. Defaults to 0. See -ipadx for an alternate padding style. Specifies what kind of interactive border resizing to allow, must be one of row, col, both (default) or none. Default row height, interpreted as lines in the default font when the number is positive, or pixels if it is negative. Defaults to 1. Specifies what row index to interpret as the topmost row in the table. This value is used for user indices in the table. Defaults to 0. Number of rows in the table. Defaults to 10. Specifies a separator character that will be interpreted as the row separator when cutting or pasting data in a table. By default, rows are separated as tcl lists. Specifies the stretch modes for rows to fill extra allocated window space. See -colstretchmode for valid options. Provides the name of a procedure that can evaluated by the widget to determine the tag to be used for a given row. The procedure must be defined by the user to accept a single argument (the row number), and return a tag name or null string. This operates in a similar manner as -coltagcommand, except that it applies to row tags. Specifies a command to evaluate when the selection is retrieved from a table via the selection mechanism (ie: evaluating "selection get"). The return value from this command will become the string passed on by the selection mechanism. It uses the %-substition model described in COMMAND SUBSTITUTION below. If an error occurs, a Tcl background error is generated and nothing is returned. Specifies one of several styles for manipulating the selection. The value of the option may be arbitrary, but the default bindings expect it to be either single, browse, multiple, or extended; the default value is browse. These styles are like those for the Tk listbox, except expanded for 2 dimensions. Specifies whether title cells should be allowed in the selection. Defaults to 0 (disallowed). Specifies one of several types of selection for the table. The value of the option may be one of row, col, cell, or both (meaning row && col); the default value is cell. These types define whether an entire row/col is affected when a cell's selection is changed (set or clear). A boolean value that specifies whether an associated Tcl array should be kept as a sparse array (1, the default) or as a full array (0). If true, then cell values that are empty will be deleted from the array (taking less memory). If false, then all values in the array will be maintained. Specifies one of two states for the entry: normal or disabled. If the table is disabled then the value may not be changed using widget commands and no insertion cursor will be displayed, even if the input focus is in the widget. Also, all insert or delete methods will be ignored. Defaults to normal. Number of columns to use as a title area. Defaults to 0. Number of rows to use as a title area. Defaults to 0. A boolean value which specifies whether to use the command option. This value sets itself to zero if command is used and returns an error. Defaults to 1 (will use command if specified). A boolean specifying whether validation should occur for the active buffer. Defaults to 0. Specifies a command to execute when the active cell is edited. This command is expected to return a Tcl boolean. If it returns true, then it is assumed the new value is OK, otherwise the new value is rejected (the edition will not take place). Errors in this command are handled in the background. It uses the %-substition model described in COMMAND SUBSTITUTION below. Global Tcl array variable to attach to the table's C array. It will be created if it doesn't already exist or is a simple variable. Keys used by the table in the array are of the form row,col for cells and the special key active which contains the value of the active cell buffer. The Tcl array is managed as a sparse array (the table doesn't require all valid indices have values). No stored value for an index is equivalent to the empty string, and clearing a cell will remove that index from the Tcl array, unless the -sparsearray options is set to 0. Specifies the desired width for the window, in columns. If zero or less, then the desired width for the window is made just large enough to hold all the columns in the table. The width can be further limited by -maxwidth. Specifies the default wrap value for tags. Defaults to 0.

The table command creates a 2-dimensional grid of cells. The table can use a Tcl array variable or Tcl command for data storage and retrieval, as well as optionally cache data in memory for speed. One of these data sources must be configured before any data is retained by the table. The widget has an active cell, the contents of which can be edited (when the state is normal). The widget supports a default style for the cells and also multiple tags, which can be used to change the style of a row, column or cell (see TAGS for details). A cell flash can be set up so that changed cells will change color for a specified amount of time ("blink"). Cells can have embedded images or windows, as described in TAGS and "EMBEDDED WINDOWS" respectively.

One or more cells may be selected as described below. If a table is exporting its selection (see -exportselection option), then it will observe the standard X11 protocols for handling the selection. See THE SELECTION for details.

It is not necessary for all the cells to be displayed in the table window at once; commands described below may be used to change the view in the window. Tables allow scrolling in both directions using the standard -xscrollcommand and -yscrollcommand options. They also support scanning, as described below.

In order to obtain good performance, the table widget supports multiple drawing modes, two of which are fully Tk compatible.

When the table command is loaded into an interpreter, a built-in Tcl command, tkTableInit, is evaluated. This will search for the appropriate table binding init file to load. The directories searched are those in $tcl_pkgPath, both with Tktable(version) appended and without, $tk_library and [pwd] (the current directory). You can also define an $env(TK_TABLE_LIBRARY) to head this search list. By default, the file searched for is called tkTable.tcl, but this can be overridden by setting $env(TK_TABLE_LIBRARY_FILE).

This entire init script can be overridden by providing your own tkTableInit procedure before the library is loaded. Otherwise, the aforementioned env(TK_TABLE_LIBRARY) variable will be set with the directory in which $env(TK_TABLE_LIBRARY_FILE) was found.

Many of the widget commands for tables take one or more indices as arguments. An index specifies a particular cell of the table, in any of the following ways:

number,number

Specifies the cell as a numerical index of row,col which corresponds to the index of the associated Tcl array, where -roworigin,-colorigin corresponds to the first cell in the table (0,0 by default). The values for row and column will be constrained to actual values in the table, which means a valid cell is always found.

active

Indicates the cell that has the location cursor. It is specified with the activate widget command.

anchor

Indicates the anchor point for the selection, which is set with the selection anchor widget command.

bottomright

Indicates the bottom-rightmost cell visible in the table.

end

Indicates the bottom right cell of the table.

origin

Indicates the top-leftmost editable cell of the table, not necessarily in the display. This takes into account the user specified origin and title area.

topleft

Indicates the top-leftmost editable cell visible in the table (this excludes title cells).

@x,y

Indicates the cell that covers the point in the table window specified by x and y (in pixel coordinates). If no cell covers that point, then the closest cell to that point is used.

In the widget command descriptions below, arguments named index, first, and last always contain text indices in one of the above forms.

A tag is a textual string that is associated with zero or more rows, columns or cells in a table. Tags may contain arbitrary characters, but it is probably best to avoid using names which look like indices to reduce coding confusion. A tag can apply to an entire row or column, or just a single cell. There are several permanent tags in each table that can be configured by the user and will determine the attributes for special cells:

active

This tag is given to the active cell

flash

If flash mode is on, this tag is given to any recently edited cells.

sel

This tag is given to any selected cells.

title

This tag is given to any cells in the title rows and columns. This tag has -state disabled by default.

Tags control the way cells are displayed on the screen. Where appropriate, the default for displaying cells is determined by the options for the table widget. However, display options may be associated with individual tags using the ``pathName tag configure'' widget command. If a cell, row or column has been tagged, then the display options associated with the tag override the default display style. The following options are currently supported for tags:

-anchor anchor

Anchor for item in the cell space.

-background or -bg color

Background color of the cell.

-borderwidth or -bd pixelList

Borderwidth of the cell, of the same format for the table, but may also be empty to inherit the default table borderwidth value (the default).

-ellipsis string

String to display at the end of a line that would be clipped by its cell, like ``...''. An ellipsis will be displayed only on non-wrapping, non-multiline cells that would be clipped. The ellipsis will display on the left for east anchored cells, otherwise it displays on the right.

-font fontName

Font for text in the cell.

-foreground or -fg color

Foreground color of the cell.

-justify justify

How to justify multi-line text in a cell. It must be one of left, right, or center.

-image imageName

An image to display in the cell instead of text.

-multiline boolean

Whether to display text with newlines on multiple lines.

-relief relief

The relief for the cell. May be the empty string to cause this tag to not disturb the value.

-showtext boolean

Whether to show the text over an image.

-state state

The state of the cell, to allow for certain cells to be disabled. This prevents the cell from being edited by the insert or delete methods, but a direct set will not be prevented.

-wrap boolean

Whether characters should wrap in a cell that is not wide enough.

A priority order is defined among tags based on creation order (first created tag has highest default priority), and this order is used in implementing some of the tag-related functions described below. When a cell is displayed, its properties are determined by the tags which are assigned to it. The priority of a tag can be modified by the ``pathName tag lower'' and ``pathName tag raise'' widget commands.

If a cell has several tags associated with it that define the same display options (eg - a title cell with specific row and cell tags), then the options of the highest priority tag are used. If a particular display option hasn't been specified for a particular tag, or if it is specified as an empty string, then that option will not be used; the next-highest-priority tag's option will be used instead. If no tag specifies a particular display option, then the default style for the widget will be used.

Images are used for display purposes only. Editing in that cell will still be enabled and any querying of the cell will show the text value of the cell, regardless of the value of -showtext.

There may be any number of embedded windows in a table widget (one per cell), and any widget may be used as an embedded window (subject to the usual rules for geometry management, which require the table window to be the parent of the embedded window or a descendant of its parent). The embedded window's position on the screen will be updated as the table is modified or scrolled, and it will be mapped and unmapped as it moves into and out of the visible area of the table widget. Each embedded window occupies one cell's worth of space in the table widget, and it is referred to by the index of the cell in the table. Windows associated with the table widget are destroyed when the table widget is destroyed.

Windows are used for display purposes only. A value still exists for that cell, but will not be shown unless the window is deleted in some way. If the window is destroyed or lost by the table widget to another geometry manager, then any data associated with it is lost (the cell it occupied will no longer appear in window names).

When an embedded window is added to a table widget with the window configure widget command, several configuration options may be associated with it. These options may be modified with later calls to the window configure widget command. The following options are currently supported:

-create script

NOT CURRENTLY SUPPORTED. Specifies a Tcl script that may be evaluated to create the window for the annotation. If no -window option has been specified for this cell then this script will be evaluated when the cell is about to be displayed on the screen. Script must create a window for the cell and return the name of that window as its result. If the cell's window should ever be deleted, the script will be evaluated again the next time the cell is displayed.

-background or -bg color

Background color of the cell. If not specified, it uses the table's default background.

-borderwidth or -bd pixelList

Borderwidth of the cell, of the same format for the table, but may also be empty to inherit the default table borderwidth value (the default).

-padx pixels

As defined in the Tk options man page.

-pady pixels

As defined in the Tk options man page.

-relief relief

The relief to use for the cell in which the window lies. If not specified, it uses the table's default relief.

-sticky sticky

Stickiness of the window inside the cell, as defined by the grid command.

-window pathName

Specifies the name of a window (widget) to display in the annotation. It must exist before being specified here. When an empty string is specified, if a window was displayed it will cease to be managed by the table widget.

Table selections are available as type STRING. By default, the value of the selection will be the values of the selected cells in nested Tcl list form where each row is a list and each column is an element of a row list. You can change the way this value is interpreted by setting the -rowseparator and -colseparator options. For example, default Excel format would be to set -rowseparator to "\n" and -colseparator to "\t". Changing these values affects both how the table sends out the selection and reads in pasted data, ensuring that the table should always be able to cut and paste to itself. It is possible to change how pastes are handled by editing the table library procedure tk_tablePasteHandler. This might be necessary if -selectioncommand is set.

Individual cells can span multiple rows and/or columns. This is done via the spans command (see below for exact arguments). Cells in the title area that span are not permitted to span beyond the title area, and will be constrained accordingly. If the title area shrinks during a configure, sanity checking will occur to ensure the above. You may set spans on regular cells that extend beyond the defined row/col area. These spans will not be constrained, so that when the defined row/col area expands, the span will expand with it.

When setting a span, checks are made as to whether the span would overlap an already spanning or hidden cell. This is an error and it not allowed. Spans can affect the overall speed of table drawing, although not significantly. If spans are not used, then there is no performance loss.

Cells hidden by spanning cells still have valid data. This will be seen during cut and paste operations that involve hidden cells, or through direct access by a command like get or set.

The drawing properties of spanning cells apply to only the visual area of the cell. For example, if a cell is center justified over 5 columns, then when viewing any portion of those columns, it will appear centered in the visible area. The non-visible column area will not be considered in the centering calculations.

The various option based commands that the table supports all support the familiar Tk %-substitution model (see bind for more details). The following %-sequences are recognized and substituted by the table widget:

%c

For SelectionCommand, it is the maximum number of columns in any row in the selection. Otherwise it is the column of the triggered cell.

%C

A convenience substitution for %r,%c.

%i

For SelectionCommand, it is the total number of cells in the selection. For Command, it is 0 for a read (get) and 1 for a write (set). Otherwise it is the current cursor position in the cell.

%r

For SelectionCommand, it is the number of rows in the selection. Otherwise it is the row of the triggered cell.

%s

For ValidateCommand, it is the current value of the cell being validated. For SelectionCommand, it is the default value of the selection. For BrowseCommand, it is the index of the last active cell. For Command, it is empty for reads (get) and the current value of the cell for writes (set).

%S

For ValidateCommand, it is the potential new value of the cell being validated. For BrowseCommand, it is the index of the new active cell.

%W

The pathname to the window for which the command was generated.

The table command creates a new Tcl command whose name is pathName. This command may be used to invoke various operations on the widget. It has the following general form: pathName option ?arg arg ...? Option and the args determine the exact behavior of the command.

The following commands are possible for table widgets:

pathName activate index

Sets the active cell to the one indicated by index.

pathName bbox first ?last?

It returns the bounding box for the specified cell (range) as a 4-tuple of x, y, width and height in pixels. It clips the box to the visible portion, if any, otherwise an empty string is returned.

pathName border option args

This command is a voodoo hack to implement border sizing for tables. This is normally called through bindings, with the following as valid options:

pathName border mark x y ?row|col?

Records x and y and the row and/or column border under that point in the table window, if any; used in conjunction with later border dragto commands. Typically this command is associated with a mouse button press in the widget. If row or col is not specified, it returns a tuple of both border indices (an empty item means no border). Otherwise, just the specified item is returned.

pathName border dragto x y

This command computes the difference between its x and y arguments and the x and y arguments to the last border mark command for the widget. It then adjusts the previously marked border by the difference. This command is typically associated with mouse motion events in the widget, to produce the effect of interactive border resizing.

pathName cget option

Returns the current value of the configuration option given by option. Option may have any of the values accepted by the table command.

pathName clear option ?first? ?last?

This command is a convenience routine to clear certain state information managed by the table. first and last represent valid table indices. If neither are specified, then the command operates on the whole table. The following options are recognized:

pathName clear cache ?first? ?last?

Clears the specified section of the cache, if the table has been keeping one.

pathName clear sizes ?first? ?last?

Clears the specified row and column areas of specific height/width dimensions. When just one index is specified, for example 2,0, that is interpreted as row 2 and column 0.

pathName clear tags ?first? ?last?

Clears the specified area of tags (all row, column and cell tags).

pathName clear all ?first? ?last?

Performs all of the above clear functions on the specified area.

pathName configure ?option? ?value option value ...?

Query or modify the configuration options of the widget. If no option is specified, returns a list describing all of the available options for pathName (see Tk_ConfigureInfo for information on the format of this list). If option is specified with no value, then the command returns a list describing the one named option (this list will be identical to the corresponding sublist of the value returned if no option is specified). If one or more option-value pairs are specified, then the command modifies the given widget option(s) to have the given value(s); in this case the command returns an empty string. Option may have any of the values accepted by the table command.

pathName curselection ?value?

With no arguments, it returns the sorted indices of the currently selected cells. Otherwise it sets all the selected cells to the given value. The set has no effect if there is no associated Tcl array or the state is disabled.

pathName curvalue ?value?

If no value is given, the value of the cell being edited (indexed by active) is returned, else it is set to the given value.

pathName delete option arg ?arg?

This command is used to delete various things in a table. It has several forms, depending on the option:

pathName delete active index ?index?

Deletes text from the active cell. If only one index is given, it deletes the character after that index, otherwise it deletes from the first index to the second. index can be a number, insert or end.

pathName delete cols ?switches? index ?count?

Deletes count cols starting at (and including) col index. The index will be constrained to the limits of the tables. If count is negative, it deletes cols to the left. Otherwise it deletes cols to the right. count defaults to 1 (meaning just the column specified). At the moment, spans are not adjusted with this action. Optional switches are:

-holddimensions

Causes the table cols to be unaffected by the deletion (empty cols may appear). By default the dimensions are adjusted by count.

-holdselection

Causes the selection to be maintained on the absolute cells values. Otherwise, the selection will be cleared..

-holdtags

Causes the tags specified by the tag method to not move along with the data. Also prevents specific widths set by the width method from being adjusted. By default, these tags are properly adjusted.

-holdwindows

Causes the embedded windows created with the window method to not move along with the data. By default, these windows are properly adjusted.

-keeptitles

Prevents title area cells from being changed. Otherwise they are treated just like regular cells and will move as specified.

--

Signifies the end of the switches.

pathName delete rows ?switches? index ?count?

Deletes count rows starting at (and including) row index. If count is negative, it deletes rows going up. Otherwise it deletes rows going down. The selection will be cleared. The switches are the same as those for column deletion.

pathName get first ?last?

Returns the value of the cells specified by the table indices first and (optionally) last in a list.

pathName height ?row? ?value row value ...?

If no row is specified, returns a list describing all rows for which a height has been set. If row is specified with no value, it prints out the height of that row in characters (positive number) or pixels (negative number). If one or more row-value pairs are specified, then it sets each row to be that height in lines (positive number) or pixels (negative number). If value is default, then the row uses the default height, specified by -rowheight.

pathName hidden ?index? ?index ...?

When called without args, it returns all the hidden cells (those cells covered by a spanning cell). If one index is specified, it returns the spanning cell covering that index, if any. If multiple indices are specified, it returns 1 if all indices are hidden cells, 0 otherwise.

pathName icursor ?arg?

With no arguments, prints out the location of the insertion cursor in the active cell. With one argument, sets the cursor to that point in the string. 0 is before the first character, you can also use insert or end for the current insertion point or the end of the text. If there is no active cell, or the cell or table is disabled, this will return -1.

pathName index index ?row|col?

Returns the integer cell coordinate that corresponds to index in the form row,col. If row or col is specified, then only the row or column index is returned.

pathName insert option arg arg

This command is used to into various things into a table. It has several forms, depending on the option:

pathName insert active index value

The value is a text string which is inserted at the index postion of the active cell. The cursor is then positioned after the new text. index can be a number, insert or end.

pathName insert cols ?switches? index ?count?

Inserts count cols starting at col index. If count is negative, it inserts before the specified col. Otherwise it inserts after the specified col. The selection will be cleared. The switches are the same as those for column deletion.

pathName insert rows ?switches? index ?count?

Inserts count rows starting at row index. If count is negative, it inserts before the specified row. Otherwise it inserts after the specified row. The selection will be cleared. The switches are the same as those for column deletion.

pathName reread

Rereads the old contents of the cell back into the editing buffer. Useful for a key binding when <Escape> is pressed to abort the edit (a default binding).

pathName scan option args

This command is used to implement scanning on tables. It has two forms, depending on option:

pathName scan mark x y

Records x and y and the current view in the table window; used in conjunction with later scan dragto commands. Typically this command is associated with a mouse button press in the widget. It returns an empty string.

pathName scan dragto x y.

This command computes the difference between its x and y arguments and the x and y arguments to the last scan mark command for the widget. It then adjusts the view by 5 times the difference in coordinates. This command is typically associated with mouse motion events in the widget, to produce the effect of dragging the list at high speed through the window. The return value is an empty string.

pathName see index

Adjust the view in the table so that the cell given by index is positioned as the cell one off from top left (excluding title rows and columns) if the cell is not currently visible on the screen. The actual cell may be different to keep the screen full.

pathName selection option arg

This command is used to adjust the selection within a table. It has several forms, depending on option:

pathName selection anchor index

Sets the selection anchor to the cell given by index. The selection anchor is the end of the selection that is fixed while dragging out a selection with the mouse. The index anchor may be used to refer to the anchor cell.

pathName selection clear first ?last?

If any of the cells between first and last (inclusive) are selected, they are deselected. The selection state is not changed for cells outside this range. first may be specified as all to remove the selection from all cells.

pathName selection includes index

Returns 1 if the cell indicated by index is currently selected, 0 if it isn't.

pathName selection set first ?last?

Selects all of the cells in the range between first and last, inclusive, without affecting the selection state of cells outside that range.

pathName set ?row|col? index ?value? ?index value ...?

Sets the specified index to the associated value. Table validation will not be triggered via this method. If row or col precedes the list of index/value pairs, then the value is assumed to be a Tcl list whose values will be split and set into the subsequent columns (if row is specified) or rows (for col). For example, set row 2,3 {2,3 2,4 2,5} will set 3 cells, from 2,3 to 2,5. The setting of cells is silently bounded by the known table dimensions.

pathName spans ?index? ?rows,cols index rows,cols ...?

This command is used to manipulate row/col spans. When called with no arguments, all known spans are returned as a list of tuples of the form {index span}. When called with only the index, the span for that index only is returned, if any. Otherwise an even number of index rows,cols pairs are used to set spans. A span starts at the index and continues for the specified number of rows and cols. Negative spans are not supported. A span of 0,0 unsets any span on that cell. See EXAMPLES for more info.

pathName tag option ?arg arg ...?

This command is used to manipulate tags. The exact behavior of the command depends on the option argument that follows the tag argument. cget, cell, and row|col complain about unknown tag names. The following forms of the command are currently supported:

pathName tag cell tagName ?index ...?

With no arguments, prints out the list of cells that use the tag. Otherwise it sets the specified cells to use the named tag, replacing any tag that may have been set using this method before. If tagName is {}, the cells are reset to the default tag. Tags added during -*tagcommand evaluation do not register here. If tagName does not exist, it will be created with the default options.

pathName tag cget tagName option

This command returns the current value of the option named option associated with the tag given by tagName. Option may have any of the values accepted by the tag configure widget command.

pathName tag col tagName ?col ...?

With no arguments, prints out the list of cols that use the tag. Otherwise it sets the specified columns to use the named tag, replacing any tag that may have been set using this method before. If tagName is {}, the cols are reset to the default tag. Tags added during -coltagcommand evaluation do not register here. If tagName does not exist, it will be created with the default options.

pathName tag configure tagName ?option? ?value? ?option value ...?

This command is similar to the configure widget command except that it modifies options associated with the tag given by tagName instead of modifying options for the overall table widget. If no option is specified, the command returns a list describing all of the available options for tagName (see Tk_ConfigureInfo for information on the format of this list). If option is specified with no value, then the command returns a list describing the one named option (this list will be identical to the corresponding sublist of the value returned if no option is specified). If one or more option-value pairs are specified, then the command modifies the given option(s) to have the given value(s) in tagName; in this case the command returns an empty string. See TAGS above for details on the options available for tags.

pathName tag delete tagName

Deletes a tag. No error if the tag does not exist.

pathName tag exists tagName

Returns 1 if the named tag exists, 0 otherwise.

pathName tag includes tagName index

Returns 1 if the specified index has the named tag, 0 otherwise.

pathName tag lower tagName ?belowThis?

Lower the priority of the named tag. If belowThis is not specified, then the tag's priority is lowered to the bottom, otherwise it is lowered to one below belowThis.

pathName tag names ?pattern?

If no pattern is specified, shows the names of all defined tags. Otherwise the pattern is used as a glob pattern to show only tags matching that pattern. Tag names are returned in priority order (highest priority tag first).

pathName tag raise tagName ?aboveThis?

Raise the priority of the named tag. If aboveThis is not specified, then the tag's priority is raised to the top, otherwise it is raised to one above aboveThis.

pathName tag row tagName ?row ...?

With no arguments, prints out the list of rows that use the tag. Otherwise it sets the specified columns to use the named tag, replacing any tag that may have been set using this method before. If tagName is {}, the rows are reset to use the default tag. Tags added during -rowtagcommand evaluation do not register here. If tagName does not exist, it will be created with the default options.

pathName validate index

Explicitly validates the specified index based on the current -validatecommand and returns 0 or 1 based on whether the cell was validated.

pathName width ?col? ?value col value ...?

If no col is specified, returns a list describing all cols for which a width has been set. If col is specified with no value, it prints out the width of that col in characters (positive number) or pixels (negative number). If one or more col-value pairs are specified, then it sets each col to be that width in characters (positive number) or pixels (negative number). If value is default, then the col uses the default width, specified by -colwidth.

pathName window option ?arg arg ...?

This command is used to manipulate embedded windows. The exact behavior of the command depends on the option argument that follows the window argument. The following forms of the command are currently supported:

pathName window cget index option

This command returns the current value of the option named option associated with the window given by index. Option may have any of the values accepted by the window configure widget command.

pathName window configure index ?option? ?value? ?option value ...?

This command is similar to the configure widget command except that it modifies options associated with the embedded window given by index instead of modifying options for the overall table widget. If no option is specified, the command returns a list describing all of the available options for index (see Tk_ConfigureInfo for information on the format of this list). If option is specified with no value, then the command returns a list describing the one named option (this list will be identical to the corresponding sublist of the value returned if no option is specified). If one or more option-value pairs are specified, then the command modifies the given option(s) to have the given value(s) in index; in this case the command returns an empty string. See EMBEDDED WINDOWS above for details on the options available for windows.

pathName window delete index ?index ...?

Deletes an embedded window from the table. The associated window will also be deleted.

pathName window move indexFrom indexTo

Moves an embedded window from one cell to another. If a window already exists in the target cell, it will be deleted.

pathName window names ?pattern?

If no pattern is specified, shows the cells of all embedded windows. Otherwise the pattern is used as a glob pattern to show only cells matching that pattern.

pathName xview args

This command is used to query and change the horizontal position of the information in the widget's window. It can take any of the following forms:

pathName xview

Returns a list containing two elements. Each element is a real fraction between 0 and 1; together they describe the horizontal span that is visible in the window. For example, if the first element is .2 and the second element is .6, 20% of the table's text is off-screen to the left, the middle 40% is visible in the window, and 40% of the text is off-screen to the right. These are the same values passed to scrollbars via the -xscrollcommand option.

pathName xview index

Adjusts the view in the window so that the column given by index is displayed at the left edge of the window.

pathName xview moveto fraction

Adjusts the view in the window so that fraction of the total width of the table text is off-screen to the left. fraction must be a fraction between 0 and 1.

pathName xview scroll number what

This command shifts the view in the window left or right according to number and what. Number must be an integer. What must be either units or pages or an abbreviation of one of these. If what is units, the view adjusts left or right by number cells on the display; if it is pages then the view adjusts by number screenfuls. If number is negative then cells farther to the left become visible; if it is positive then cells farther to the right become visible.

pathName yview ?args?

This command is used to query and change the vertical position of the text in the widget's window. It can take any of the following forms:

pathName yview

Returns a list containing two elements, both of which are real fractions between 0 and 1. The first element gives the position of the table element at the top of the window, relative to the table as a whole (0.5 means it is halfway through the table, for example). The second element gives the position of the table element just after the last one in the window, relative to the table as a whole. These are the same values passed to scrollbars via the -yscrollcommand option.

pathName yview index

Adjusts the view in the window so that the row given by index is displayed at the top of the window.

pathName yview moveto fraction

Adjusts the view in the window so that the element given by fraction appears at the top of the window. Fraction is a fraction between 0 and 1; 0 indicates the first element in the table, 0.33 indicates the element one-third the way through the table, and so on.

pathName yview scroll number what

This command adjusts the view in the window up or down according to number and what. Number must be an integer. What must be either units or pages. If what is units, the view adjusts up or down by number cells; if it is pages then the view adjusts by number screenfuls. If number is negative then earlier elements become visible; if it is positive then later elements become visible.

The initialization creates class bindings that give the following default behaviour:

[1]

Clicking Button-1 in a cell activates that cell. Clicking into an already active cell moves the insertion cursor to the character nearest the mouse.

[2]

Moving the mouse while Button-1 is pressed will stroke out a selection area. Exiting while Button-1 is pressed causing scanning to occur on the table along with selection.

[3]

Moving the mouse while Button-2 is pressed causes scanning to occur without any selection.

[4]

Home moves the table to have the origin in view.

[5]

End moves the table to have the end cell in view.

[6]

Control-Home moves the table to the origin and activates that cell.

[7]

Control-End moves the table to the end and activates that cell.

[8]

Shift-Control-Home extends the selection to the origin.

[9]

Shift-Control-End extends the selection to the end.

[10]

The left, right, up and down arrows move the active cell.

[11]

Shift-<arrow> extends the selection in that direction.

[12]

Control-leftarrow and Control-rightarrow move the insertion cursor within the cell.

[13]

Control-slash selects all the cells.

[14]

Control-backslash clears selection from all the cells.

[15]

Backspace deletes the character before the insertion cursor in the active cell.

[16]

Delete deletes the character after the insertion cursor in the active cell.

[17]

Escape rereads the value of the active cell from the specified data source, discarding any edits that have may been performed on the cell.

[18]

Control-a moves the insertion cursor to the beginning of the active cell.

[19]

Control-e moves the insertion cursor to the end of the active cell.

[20]

Control-minus and Control-equals decrease and increase the width of the column with the active cell in it.

[21]

Moving the mouse while Button-3 (the right button on Windows) is pressed while you are over a border will cause interactive resizing of that row and/or column to occur, based on the value of -resizeborders.

Some bindings may have slightly different behavior dependent on the -selectionmode of the widget.

If the widget is disabled using the -state option, then its view can still be adjusted and cells can still be selected, but no insertion cursor will be displayed and no cell modifications will take place.

The behavior of tables can be changed by defining new bindings for individual widgets or by redefining the class bindings. The default bindings are either compiled in or read from a file expected to correspond to: "[lindex $tcl_pkgPath 0]/Tktable<version>/tkTable.tcl".

The number of rows and columns or a table widget should not significantly affect the speed of redraw. Recalculation and redraw of table parameters and cells is restricted as much as possible.

The display cell with the insert cursor is redrawn each time the cursor blinks, which causes a steady stream of graphics traffic. Set the -insertofftime option to 0 avoid this. The use of a -command with the table without a cache can cause significant slow-down, as the command is called once for each request of a cell value.

Set the topleft title area to be one spanning cell. This overestimates both row and column span by one, but the command does all the constraining for us. $table span [$table cget -roworigin],[$table cget -colorigin] [$table cget -titlerows],[$table cget -titlecols] Force a table window refresh (useful for the slight chance that a bug in the table is not causing proper refresh): $table configure -padx [$table cget -padx]

table, widget, extension