Tkinter Reference A GUI For Python

Tkinter reference: a GUI
for Python
John W. Shipman
2004-06-03 16:13
Table of Contents
1. What is Tkinter? ...................................................................................................................... 2
2. A minimal application .............................................................................................................. 3
3. Layout management ................................................................................................................ 3
3.1. The .grid() method .................................................................................................... 4
3.2. Other grid management methods ................................................................................... 5
4. Standard attributes .................................................................................................................. 5
4.1. Dimensions ................................................................................................................... 5
4.2. Coordinate system ......................................................................................................... 6
4.3. Colors ........................................................................................................................... 6
4.4. Type fonts ..................................................................................................................... 6
4.5. Anchors ........................................................................................................................ 8
4.6. Relief styles ................................................................................................................... 8
4.7. Bitmaps ........................................................................................................................ 8
4.8. Cursors ......................................................................................................................... 9
4.9. Images ........................................................................................................................ 11
4.10. Geometry strings ....................................................................................................... 11
4.11. Window names .......................................................................................................... 12
5. The Button widget ................................................................................................................ 12
6. The Canvas widget ................................................................................................................ 14
6.1. Canvas concepts .......................................................................................................... 15
6.2. Methods on Canvas objects ......................................................................................... 16
6.3. The canvas arc object ................................................................................................... 20
6.4. The canvas bitmap object ............................................................................................. 20
6.5. The canvas image object ............................................................................................... 21
6.6. The canvas line object .................................................................................................. 21
6.7. The canvas oval object .................................................................................................. 22
6.8. The canvas polygon object ............................................................................................ 23
6.9. The canvas rectangle object .......................................................................................... 24
6.10. The canvas text object ................................................................................................. 24
6.11. The canvas window object .......................................................................................... 25
7. The Checkbutton widget ...................................................................................................... 25
8. The Entry widget .................................................................................................................. 28
8.1. Scrolling an Entry widget ........................................................................................... 31
9. The Frame widget .................................................................................................................. 32
10. The Label widget ................................................................................................................ 33
11. The Listbox widget ............................................................................................................ 34
11.1. Scrolling a Listbox widget ....................................................................................... 37
12. The Menu widget .................................................................................................................. 37
New Mexico Tech Computer Center Tkinter reference 1

12.1. Menu item creation (coption) options ....................................................................... 39
13. The Menubutton widget ...................................................................................................... 40
14. The Radiobutton widget .................................................................................................... 42
15. The Scale widget ................................................................................................................ 45
16. The Scrollbar widget ........................................................................................................ 48
16.1. The scrollbar command callback .................................................................................. 50
16.2. Connecting scrollbars to other widgets ........................................................................ 50
17. The Text widget .................................................................................................................. 51
17.1. Indices in text widgets ................................................................................................ 53
17.2. Marks in text widgets ................................................................................................. 53
17.3. Images in text widgets ................................................................................................ 54
17.4. Windows in text widgets ............................................................................................ 54
17.5. Tags in text widgets ................................................................................................... 54
17.6. Setting tabs in a Text widget ...................................................................................... 55
17.7. Methods on Text widgets .......................................................................................... 55
18. Toplevel: Top-level window methods ................................................................................. 61
19. Universal widget methods .................................................................................................... 62
20. Standardizing appearance ..................................................................................................... 69
20.1. How to name a widget class ....................................................................................... 70
20.2. How to name a widget instance .................................................................................. 70
20.3. Resource specification lines ........................................................................................ 70
20.4. Rules for resource matching ........................................................................................ 71
21. Connecting your application logic to the widgets .................................................................... 72
22. Control variables: the values behind the widgets .................................................................... 72
23. Focus: routing keyboard input ............................................................................................... 74
24. Events .................................................................................................................................. 75
24.1. Levels of binding ....................................................................................................... 76
24.2. Event sequences ......................................................................................................... 76
24.3. Event types ................................................................................................................ 77
24.4. Event modifiers ......................................................................................................... 78
24.5. Key names ................................................................................................................. 78
24.6. Writing your handler ................................................................................................. 81
24.7. The extra arguments trick ........................................................................................... 82
24.8. Virtual events ............................................................................................................ 83
1.What is Tkinter?
Tkinter is a GUI (graphical user interface) widget set for Python. This document contains only the
commoner features.
This document applies to Python 1.5 and Tkinter 8.0.4 running in the X Window system under Linux.
Your version may vary.
Pertinent references:
An Introduction to Tkinter [http://www.pythonware.com/library/tkinter/introduction/index.htm]
by Fredrik Lundh
Python and Tkinter Programming by John Grayson (Manning, 2000, ISBN 1-884777-81-3).
Python 2.2 quick reference [http://www.nmt.edu/tcc/help/pubs/python22/]: general information
about the Python language.
This publication is available in Web form [http://www.nmt.edu/tcc/help/pubs/tkinter/] and also
as a PDF document (82 pp.) [http://www.nmt.edu/tcc/help/pubs/tkinter/tkinter.pdf].
2 Tkinter reference New Mexico Tech Computer Center

We'll start by looking at the visible part of Tkinter: creating the widgets and arranging them on the
screen. Later we will talk about how to connect the facethe front panelof the application to the
logic behind it.
2.A minimal application

Here is a trivial Tkinter program containing only a Quit button:
#!/usr/local/bin/python 1
from Tkinter import * 2
class Application(Frame): 3
def __init__(self, master=None):
Frame.__init__(self, master) 4
self.grid() 5
self.createWidgets()
def createWidgets(self):
self.quitButton = Button ( self, text="Quit",
command=self.quit ) 6
self.quitButton.grid() 7
app = Application() 8
app.master.title("Sample application") 9
app.mainloop() 10
1 This line makes the script self-executing, assuming that your system has the Python interpreter at
path /usr/local/bin/python.
2 This line imports the entire Tkinter package into your program's namespace.
3 Your application class must inherit from Tkinter's Frame class.
4 Calls the constructor for the parent class, Frame.
5 Necessary to make the application actually appear on the screen.
6 Creates a button labeled Quit.
7 Places the button on the application.
8 The main program starts here by instantiating the Application class.
9 This method call sets the title of the window to Sample application.
10 Starts the application's main loop, waiting for mouse and keyboard events.
3.Layout management
Later we will discuss the widgets, the building blocks of your GUI application. How do widgets get
arranged in a window?
Although there are three different geometry managers in Tkinter, the author strongly prefers the
.grid() geometry manager for pretty much everything. This manager treats every window or frame
as a tablea gridwork of rows and columns.
A cell is the area at the intersection of one row and one column.
The width of each column is the width of the widest cell in that column.
The height of each row is the height of the largest cell in that row.

For widgets that do not fill the entire cell, you can specify what happens to the extra space. You can
either leave the extra space outside the widget, or stretch the widget to fit it, in either the horizontal
or vertical dimension.
You can combine multiple cells into one larger area, a process called spanning.
When you create a widget, it does not appear until you register it with a geometry manager. Hence,
construction and placing of a widget is a two-step process that goes something like this:
thing = Constructor(master, ...)
thing.grid(...)
where Constructor is one of the widget classes like Button, Frame, and so on, and master is the
parent widget in which this child widget is being constructed. All widgets have a .grid() method
that you can use to tell the geometry manager where to put it.
3.1.The .grid() method

To display a widget w on your application screen:
w.grid(option, ...)
This method registers a widget w with the grid geometry managerif you don't do this, the widget will
exist internally, but it will not be visible on the screen.
Here are the options to the .grid() geometry management method:
column The column number where you want the widget gridded, counting from
zero. The default value is zero.
columnspan Normally a widget occupies only one cell in the grid. However, you can grab
multiple cells of a row and merge them into one larger cell by setting the
columnspan option to the number of cells. For example, w.grid(row=0,
column=2, columnspan=3) would place widget w in a cell that spans
columns 2, 3, and 4 of row 0.
ipadx Internal x padding. This dimension is added inside the widget inside its left
and right sides.
ipady Internal y padding. This dimension is added inside the widget inside its top
and bottom borders.
padx External x padding. This dimension is added to the left and right outside the
widget.
pady External y padding. This dimension is added above and below the widget.
row The row number into which you want to insert the widget, counting from 0.
The default is the next higher-numbered unoccupied row.
rowspan Normally a widget occupies only one cell in the grid. You can grab multiple
adjacent cells of a column, however, by setting the rowspan option to the
number of cells to grab. This option can be used in combination with the
columnspan option to grab a block of cells. For example, w.grid(row=3,
column=2, rowspan=4, columnspan=5) would place widget w in an
area formed by merging 20 cells, with row numbers 36 and column numbers
26.
sticky This option determines how to distribute any extra space within the cell that
is not taken up by the widget at its natural size. See below.
If you do not provide a sticky attribute, the default behavior is to center the widget in the cell.

You can position the widget in a corner of the cell by using sticky=NE (top right), SE (bottom
right), SW (bottom left), or NW (top left).
You can position the widget centered against one side of the cell by using sticky=N (top center),
E (right center), S (bottom center), or W (left center).
Use sticky=N+S to stretch the widget vertically but leave it centered horizontally.
Use sticky=E+W to stretch it horizontally but leave it centered vertically.
Use sticky=N+E+S+W to stretch the widget both horizontally and vertically to fill the cell.
The other combinations will also work. For example, sticky=N+S+W will stretch the widget vertically
and place it against the west (left) wall.
3.2.Other grid management methods

These grid-related methods are defined on all widgets:
w.grid_forget()
This method makes widget w disappear from the screen. It still exists, it just isn't visible. You can
use .grid() it to make it appear again, but it won't remember its grid options.
w.grid_propagate()
Normally, all widgets propagate their dimensions, meaning that they adjust to fit the contents.
However, sometimes you want to force a widget to be a certain size, regardless of the size of its
contents. To do this, call w.grid_propagate(0) where w is the widget whose size you want to
force.
w.grid_remove()
This method is like .grid_forget(), but its grid options are remembered, so if you .grid() it
again, it will use the same grid configuration options.
4.Standard attributes
Before we look at the widgets, let's take a look at how some of their common attributessuch as sizes,
colors and fontsare specified.
Each widget has a set of options that affect its appearance and behaviorattributes such as fonts,
colors, sizes, text labels, and such.
You can specify options when calling the widget's constructor using keyword arguments such as
text="PANIC!" or height=20.
After you have created a widget, you can later change any option by using the widget's .config()
method, and retrieve the current setting of any option by using the widget's .cget() method. See
Section19, Universal widget methods (p. 62) for more on these methods.
4.1.Dimensions
Various lengths, widths, and other dimensions of widgets can be described in many different units.
If you set a dimension to an integer, it is assumed to be in pixels.
You can specify units by setting a dimension to a string containing a number followed by:
c Centimeters
i Inches

m Millimeters
p Printer's points (about 1/72")
4.2.Coordinate system
As in most contemporary display systems, the origin of each coordinate system is at its upper left corner,
with the x coordinate increasing toward the right, and the y coordinate increasing toward the bottom:
+x
+y
The base unit is the pixel, with the top left pixel having coordinates (0,0). Coordinates that you specify
as integers are always expressed in pixels, but any coordinate may be specified as a dimensioned
quantity; see Section4.1, Dimensions (p. 5).
4.3.Colors
There are two general ways to specify colors in Tkinter.
You can use a string specifying the proportion of red, green, and blue in hexadecimal digits:
#rgb Four bits per color

#rrggbb Eight bits per color
#rrrgggbbb Twelve bits per color
For example, "#fff" is white, "#000000" is black, "#000fff000" is pure green, and "#00ffff"
is pure cyan (green plus blue).
You can also use any locally defined standard color name. The colors "white", "black", "red",
"green", "blue", "cyan", "yellow", and "magenta" will always be available. Other names
may work, depending on your local installation.
4.4.Type fonts
Depending on your platform, there may be up to three ways to specify type style.
As a tuple whose first element is the font family, followed by a size in points, optionally followed
by a string containing one or more of the style modifiers bold, italic, underline, and over-
strike.
Examples: ("Helvetica", "16") for a 16-point Helvetica regular; ("Times", "24", "bold
italic") for a 24-point Times bold italic.
You can create a font object by importing the tkFont module and using its Font class constructor:
import tkFont
font = tkFont.Font ( option, ... )

where the options include:
family The font family name as a string.

size The font height as an integer in points. To get a font n pixels high, use -n.

weight "bold" for boldface, "normal" for regular weight.
slant "italic" for italic, "roman" for unslanted.
underline 1 for underlined text, 0 for normal.
overstrike 1 for overstruck text, 0 for normal.
For example, to get a 36-point bold Helvetica italic face:

helv36 = tkFont.Font ( family="Helvetica",
size=36, weight="bold" )
If you are running under the X Window System, you can use any of the X font names. For example,
the font named "-*-lucidatypewriter-medium-r-*-*-*-140-*-*-*-*-*-*" is the author's
favorite fixed-width font for onscreen use. Use the xfontsel program to help you select pleasing fonts.
To get a list of all the families of fonts available on your platform, call this function:
tkFont.families()
The return value is a list of strings. Note: You must create your root window before calling this function.
These methods are defined on all Font objects:
.actual ( option=None )
If you pass no arguments, you get back a dictionary of the font's actual attributes, which may differ
from the ones you requested. To get back the value of an attribute, pass its name as an argument.
.cget ( option )
Returns the value of the given option.
.configure ( option, ... )
Use this method to change one or more options on a font. For example, if you have a Font object
called titleFont, if you call titleFont.configure ( family="times", size=18 ), that
font will change to 18pt Times and any widgets that use that font will change too.
.copy()
Returns a copy of a Font object.
.measure ( text )
Pass this method a string, and it will return the number of pixels of width that string will take in
the font. Warning: some slanted characters may extend outside this area.
.metrics ( option )
If you call this method with no arguments, it returns a dictionary of all the font metrics. You can re-
trieve the value of just one metric by passing its name as an argument. Metrics include:
ascent Number of pixels of height between the baseline and the top of the highest as-
cender.
descent Number of pixels of height between the baseline and the bottom of the lowest
ascender.
fixed This value is 0 for a variable-width font and 1 for a monospaced font.
linespace Number of pixels of height total. This is the leading of type set solid in the
given font.

4.5.Anchors
Constants are defined in the Tkinter package that you can use to control where items are positioned
relative to their context. For example, anchors can specify where a widget is located inside a frame when
the frame is bigger than the widget.
These constants are given as compass points, where north is up and west is to the left. We apologize to
our Southern Hemisphere readers for this Northern Hemisphere chauvinism.
The anchor constants are shown in this diagram:
NW N NE
W CENTER E
SW S SE
For example, if you create a small widget inside a large frame and use the anchor=SE option, the
widget will be placed in the bottom right corner of the frame. If you used anchor=N instead, the widget
would be centered along the top edge.
Anchors are also used to define where text is positioned relative to a reference point. For example, if
you use CENTER as a text anchor, the text will be centered horizontally and vertically around the reference
point. Anchor NW will position the text so that the reference point coincides with the northwest (top left)
corner of the box containing the text. Anchor W will center the text vertically around the reference point,
with the left edge of the text box passing through that point, and so on.
4.6.Relief styles
The relief style of a widget refers to certain simulated 3-D effects around the outside of the widget. Here
is a screen shot of a row of buttons exhibiting all the possible relief styles:
The width of these borders depends on the borderwidth attribute of the widget. The above graphic
shows what they look like with a 5-pixel border; the default border width is 2.
4.7.Bitmaps
For bitmap options in widgets, these bitmaps are guaranteed to be available:
The graphic above shows Button widgets bearing the standard bitmaps. From left to right, they are
"error", "gray75", "gray50", "gray25", "gray12", "hourglass", "info", "questhead",
"question", and "warning".
You can use your own bitmaps. Any file in .xbm (X bit map) format will work. In place of a standard
bitmap name, use the string "@" followed by the pathname of the .xbm file.

4.8.Cursors
There are quite a number of different mouse cursors available. Their names are given below. The author
hasn't yet figured out how to get graphics for them, since cursors are omitted from screen captures, so
rough descriptions have been included. Try them out to see what they look like.
arrow Northeast-pointing arrow.

based_arrow_down Downward-pointing arrow with a doubled baseline.
based_arrow_up Upward-pointing arrow with a doubled baseline.
boat The front half of a ship.
bogosity A horizontal rectangle intersected by a right bracket, a vertical line, and
a left bracket.
bottom_left_corner A two-headed arrow oriented SW-NE (southwest to northeast) with its
reference point at the center.
bottom_right_corner A two-headed arrow oriented NW-SE with its reference point at the center.
bottom_side A vertical two-headed arrow with its referenc point at the bottom.
bottom_tee A horizontal line with a perpendicular bisector above it.
box_spiral A squared-off spiral.
center_ptr An up-arrow with the reference point at the tip.
circle A thick-walled circle.
clock A grandfather clock.
coffee_mug A coffee mug.
cross Vertical and horizontal lines with a dot positioned in a gap at the center.
cross_reverse Double lines radiating in the eight principal directions from the center.
crosshair Same as cross.
diamond_cross Four hollow right triangles positioned with their right angles at the center.
dot A thick black dot.
dotbox A dot inside a box.
double_arrow A vertical two-headed arrow.
draft_large An diagonal arrow with the head pointing NE.
draft_small Same as draft_large.
draped_box A box with a dot in the middle, and four small circular arcs, one centered
on each corner.
exchange Two curved arrows chasing each other's tails.
fleur Four arrows emanating from the center.
gobbler A turkey.
gumby The cartoon character, riding his famous horse Pokey.
hand1 A right hand pointing NE.
hand2 A right hand pointing up.
heart Like the card suit hearts.

icon A box with narrow gray lines along each side.
iron_cross The symbol of Imperial Germany.
left_ptr The default cursor in graphic contexts, an arrow pointing a bit west of
north.
left_side A horizontal two-headed arrow with its reference point in the center.
left_tee A vertical line with a perpendicular bisector pointing to the right, the
reference point at the junction.
leftbutton A stylized mouse with the left button pressed.
ll_angle Lower left corner, a vertical line sharing its bottom vertex with the left
vertex of a horizontal line.
lr_angle Lower right corner, similar to ll_angle.
man A stick figure of a human.
middlebutton A mouse with the middle button pressed.
mouse A three-button mouse.
pencil A pencil pointing down and a bit to the right, with its reference point at
the tip.
pirate The death's head symbol, suitable for functions like delete or crash
and burn.
plus Like left_ptr, but with a plus sign displayed to its right.
question_arrow Like left_ptr, but with a question mark displayed to its right.
right_ptr Like left_ptr but leaning to the right.
right_side Same as left_side.
right_tee A vertical line with a perpendicular bisector extended to its left, the refer-
ence point at the junction.
rightbutton A three-button mouse with the right-hand button pressed.
rtl_logo A square with some logo in it, unknown to the author.
sailboat Two sails of a small sailboat.
sb_down_arrow A down-arrow with a broad head.
sb_h_double_arrow Same as left_side.
sb_left_arrow A left-pointing arrow.
sb_right_arrow A right-pointing arrow.
sb_up_arrow Same as sb_down_arrow, strangely enough.
sb_v_double_arrow Same as bottom_side.
shuttle The U.S. Space Shuttle with boosters.
sizing Same as trek.
spider A spider, head down.
spraycan A can of spray paint.
star A five-pointed star.
target A diamond with a dot in the middle.

tcross Same as cross.
top_left_arrow Same as left_arrow.
top_left_corner Same as bottom_right_corner.
top_right_corner Same as bottom_left_corner.
top_side Same as bottom_side.
top_tee A horizontal line with a perpendicular bisector extending down; reference
point at the intersection.
trek A rocket ship straight out of the 1950s.
ul_angle Like ll_angle but intersecting at the upper left.
umbrella An umbrella.
ur_angle Like ll_angle but intersecting at the upper right.
watch An hourglass.
X_cursor The logo of the X Window System project.
xterm The default cursor in textual contexts, an I-beam with notches at the top
and bottom.
4.9.Images
To use graphic images in a Tkinter application, Tkinter must be configured to include the Python Imaging
Library (PIL).
Refer to the author's companion document for PIL documentation: Python Imaging Library (PIL) quick
reference [http://www.nmt.edu/tcc/help/pubs/pil/]. Objects of the ImageTk class can be used in
Tkinter applications.
4.10.Geometry strings
A geometry string is a standard way of describing the size and location of a top-level window on a
desktop.
A geometry string has this general form:
"wxhxy"
where:
The w and h parts give the window width and height in pixels. They are separated by the character
"x".
If the next part has the form +x, it specifies that the left side of the window should be x pixels from
the left side of the desktop. If it has the form -x, the right side of the window is x pixels from the
right side of the desktop.
If the next part has the form +y, it specifies that the top of the window should be y pixels below the
top of the desktop. If it has the form -y, the bottom of the window will be y pixels above the bottom
edge of the desktop.
For example, a window created with geometry="120x50-0+20" would be 120 pixels wide by 50
pixels high, and its top right corner will be along the right edge of the desktop and 20 pixels below the
top edge.

4.11.Window names
The term window describes a rectangular area on the desktop.
A top-level or root window is a window that has an independent existence under the window manager.
It is decorated with the window manager's decorations, and can be moved and resized independently.
Your application can use any number of top-level windows.
The term window also applies to any widget that is part of a top-level window.
Tkinter names all these windows using a hierarchical window path name.
The root window's name is ".".
Child windows have names of the form ".n", where n is some integer in string form. For example,
a window named ".135932060" is a child of the root window (".").
Child windows within child windows have names of the form "p.n" where p is the name of the
parent window and n is some integer. For example, a window named ".135932060.137304468"
has parent window ".135932060", so it is a grandchild of the root window.
The relative name of a window is the part past the last "." in the path name. To continue the previous
example, the grandchild window has a relative name "137304468".
The path name for any widget w can be determined by calling str(w).
See also Section19, Universal widget methods (p. 62) for methods you can use to operate on window
names, especially the .winfo_name, .winfo_parent, and .winfo_pathname methods.
5.The Button widget

To create a pushbutton in a top-level window or frame named master:
w = Button ( master, option=value, ... )
The constructor returns the new button object. Its options include:
activebackground Background color when the button is under the cursor.

activeforeground Background color when the button is under the cursor.
anchor Where the text is positioned on the button. See Section4.5, An-
chors (p. 8). For example, anchor=NE would position the text at the
top right corner of the button.
bd or borderwidth Border width in pixels. Default is 2.
bg or background Normal background color.
bitmap Name of one of the standard bitmaps to display on the button (instead
of text).
command Function or method to be called when the button is clicked.
cursor Selects the cursor to be shown when the mouse is over the button.
default NORMAL is the default; use DISABLED if the button is to be initially dis-
abled (grayed out, unresponsive to mouse clicks).
disabledforeground Foreground color used when the button is disabled.
fg or foreground Normal foreground (text) color.
font Text font to be used for the button's label.

height Height of the button in text lines (for textual buttons) or pixels (for images).
highlightbackground Color of the focus highlight when the widget does not have focus.
highlightcolor The color of the focus highlight when the widget has focus.
highlightthickness Thickness of the focus highlight.
image Image to be displayed on the button (instead of text).
justify How to show multiple text lines: LEFT to left-justify each line; CENTER
to center them; or RIGHT to right-justify.
padx Additional padding left and right of the text. See Section4.1, Dimen-
sions (p. 5) for the possible values for padding.
pady Additional padding above and below the text.
relief Specifies the relief type for the button (see Section4.6, Relief
styles (p. 8)). The default relief is RAISED.
state Set this option to DISABLED to gray out the button and make it unrespons-
ive. Has the value ACTIVE when the mouse is over it. Default is NORMAL.
takefocus Normally, keyboard focus does visit buttons (see Section23, Focus:
routing keyboard input (p. 74)), and a space character acts as the same
as a mouse click, pushing the button. You can set the takefocus option
to zero to prevent focus from visiting the button.
text Text displayed on the button. Use internal newlines to display multiple
text lines.
textvariable An instance of StringVar() that is associated with the text on this
button. If the variable is changed, the new value will be displayed on the
button. See Section22, Control variables: the values behind the wid-
gets (p. 72).
underline Default is -1, meaning that no character of the text on the button will be
underlined. If nonnegative, the corresponding text character will be un-
derlined. For example, underline=1 would underline the second char-
acter of the button's text.
width Width of the button in letters (if displaying text) or pixels (if displaying
an image).
wraplength If this value is set to a positive number, the text lines will be wrapped to
fit within this length. For possible values, see Section4.1, Dimen-
sions (p. 5).
Methods on Button objects:
.flash()
Causes the button to flash several times between active and normal colors. Leaves the button in the
state it was in originally. Ignored if the button is disabled.
.invoke()
Calls the button's callback, and returns what that function returns. Has no effect if the button is
disabled or there is no callback.

6.The Canvas widget
A canvas is a rectangular area intended for drawing pictures or other complex layouts. On it you can
place graphics, text, widgets, or frames. See the following sections for methods that create objects on
canvases:
.create_arc(): A slice out of an ellipse. See Section6.3, The canvas arc object (p. 20).
.create_bitmap(): An image as a bitmap. See Section6.4, The canvas bitmap object (p. 20).
.create_image(): A graphic image. See Section6.5, The canvas image object (p. 21).
.create_line(): One or more line segments. See Section6.6, The canvas line object (p. 21).
.create_oval(): An ellipse; use this also for drawing circles, which are a special case of an ellipse.
See Section6.7, The canvas oval object (p. 22).
.create_polygon(): A polygon. See Section6.8, The canvas polygon object (p. 23).
.create_rectangle(): A rectangle. See Section6.9, The canvas rectangle object (p. 24).
.create_text(): Text annotation. See Section6.10, The canvas text object (p. 24).
.create_window(): A rectangular window. See Section6.11, The canvas window object (p. 25).
To create a Canvas object:

w = Canvas ( master, option=value, ... )
The constructor returns the new canvas widget. Supported options include:
bd or borderwidth Border width in pixels.

bg or background Background color of the canvas. Default is a light gray, about "#E4E4E4".
closeenough A float that specifies how close the mouse must be to an item to be considered
inside it. Default is 1.0.
confine If true (the default), the canvas cannot be scrolled outside of the scrollre-
gion (see below).
cursor Cursor used in the canvas. See Section4.8, Cursors (p. 9).
height Size of the canvas in the Y dimension. See Section4.1, Dimensions (p. 5).
highlightback- Color of the focus highlight when the widget does not have focus. See Sec-
ground tion23, Focus: routing keyboard input (p. 74).
highlightcolor Color shown in the focus highlight.
relief The relief style of the canvas. Default is FLAT. See Section4.6, Relief
styles (p. 8).
scrollregion A tuple (w, n, e, s) that defines over how large an area the canvas can
be scrolled, where w is the left side, n the top, e the right side, and s the bot-
tom.
selectbackground The background color to use displaying selected items.
selectborderwidth The width of the border to use around selected items.
selectforeground The foreground color to use displaying selected items.

takefocus Normally, focus (see Section23, Focus: routing keyboard input (p. 74)) will
cycle through this widget with the tab key only if there are keyboard bindings
set for it (see Section24, Events (p. 75) for an overview of keyboard bind-
ings). If you set this option to 1, focus will always visit this widget. Set it to
"" to get the default behavior.
width Size of the canvas in the X dimension. See Section4.1, Dimensions (p. 5).
xscrollincrement Normally, canvases can be scrolled horizontally to any position. You can get
this behavior by setting xscrollincrement to zero. If you set this option
to some positive dimension, the canvas can be positioned only on multiples
of that distance, and the value will be used for scrolling by scrolling units,
such as when the user clicks on the arrows at the ends of a scrollbar. For more
information on scrolling units, see Section16, The Scrollbar wid-
get (p. 48).
xscrollcommand If the canvas is scrollable, this attribute should be the .set() method of the
horizontal scrollbar.
yscrollincrement Works like xscrollincrement, but governs vertical movement.
yscrollcommand If the canvas is scrollable, this attribute should be the .set() method of the
vertical scrollbar.
6.1.Canvas concepts
Before we discuss the methods on canvases, we need to define some terms:
6.1.1.Canvas and window coordinates

Because the canvas may be larger than the window, and equipped with scrollbars to move the overall
canvas around in the window, there are two coordinate systems for each canvas:
The window coordinates of a point are relative to the top left corner of the area on the display where
the canvas appears.
The canvas coordinates of a point are relative to the top left corner of the total canvas.
6.1.2.The display list; above and below

The display list refers to the sequence of all the objects on the canvas, from background (the bottom
of the display list) to foreground (the top).
If two objects overlap, the one above the other in the display list means the one closer to the foreground,
which will appear in the area of overlap and obscure the one below. By default, new objects are always
created at the top of the display list (and hence in front of all other objects), but you can re-order the
display list.
6.1.3.Object ID
The object ID of an object on the canvas is the value returned by the constructor for that object. All object
ID values are simple integers, and the object ID of an object is unique within that canvas.
6.1.4.Canvas tags
A tag is a string that you can associate with objects on the canvas.
A tag can be associated with any number of objects on the canvas, including zero.
An object can have any number of tags associated with it, including zero.

Tags have many uses. For example, if you are drawing a map on a canvas, and there are text objects for
the labels on rivers, you could attach the tag "riverLabel" to all those text objects. This would allow
you to perform operations on all the objects with that tag, such as changing their color or deleting them.
6.1.5.Canvas tagOrId arguments

A tagOrId argument specifies one or more objects on the canvas.
If a tagOrId argument is an integer, it is treated as an object ID, and it applies only to the unique
object with that ID.
If such an argument is a string, it is interpreted as a tag, and selects all the objects that have that tag
(if there are any).
6.2.Methods on Canvas objects

All Canvas objects support these methods:
.addtag_above ( newTag, tagOrId )

Attaches a new tag to the object just above the one specified by tagOrId in the display list. The
newTag argument is the tag you want to attach, as a string.
.addtag_all ( newTag )
Attaches the given tag newTag to all the objects on the canvas.
.addtag_below ( newTag, tagOrID )
Attaches a new tag to the object just below the one specified by tagOrId in the display list. The
newTag argument is a tag string.
.addtag_closest ( newTag, x, y, halo=None, start=None )
Adds a tag to the object closest to screen coordinate (x,y). If there are two or more objects at the
same distance, the one higher in the display list is selected.
Use the halo argument to increase the effective size of the point. For example, a value of 5 would
treat any object within 5 pixels of (x,y) as overlapping.
If an object ID is passed in the start argument, this method tags the highest qualifying object that
is below start in the display list. This allows you to search through all the overlapping objects
sequentially.
.addtag_enclosed ( newTag, x1, y1, x2, y2)
Add tag newTag to all objects that occur completely within the rectangle whose top left corner is
(x1, y1) and whose bottom right corner is (x2, y2).
.addtag_overlapping ( newTag, x1, y1, x2, y2)
Like the previous method, but affects all objects that share at least one point with the given rectangle.
.addtag_withtag ( newTag, tagOrId )
Adds tag newTag to the object or objects specified by tagOrId.
.bbox ( tagOrId=None )
Returns a tuple (x1, y1, x2, y2) describing a rectangle that encloses all the objects specified by
tagOrId. If the argument is omitted, returns a rectangle enclosing all objects on the canvas. The
top left corner of the rectangle is (x1, y1) and the bottom right corner is (x2, y2).
.canvasx ( screenx, gridspacing=None )
Translates a window x coordinate screenx to a canvas coordinate. If gridspacing is supplied,
the canvas coordinate is rounded to the nearest multiple of that value.

.canvasy ( screenx, gridspacing=None )
Translates a window y coordinate screeny to a canvas coordinate. If gridspacing is supplied,
the canvas coordinate is rounded to the nearest multiple of that value.
.coords ( tagOrId, x0, y0, x1, y1, ..., xn, yn )
If you pass only the tagOrId argument, returns a tuple of the coordinates of the lowest or only
object specified by that argument. The number of coordinates depends on the type of object. In most
cases it will be a 4-tuple (x1, y1, x2, y2) describing the bounding box of the object.
You can move an object by passing in new coordinates.
.dchars ( tagOrId, first=0, last=first )
Deletes characters from a text item or items. Characters between first and last are deleted,
where those values can be integer indices or the string "end" to mean the end of the text.
.delete ( tagOrId )
Deletes the object or objects selected by tagOrId.
.dtag ( tagOrId, tagToDelete )
Removes the tag specified by tagToDelete from the object or objects specified by tagOrId.
.find_above ( tagOrId )
Returns the ID number of the object just above the object specified by tagOrId. If multiple objects
match, you get the highest one.
.find_all()
Returns a list of the object ID numbers for all objects on the canvas, from lowest to highest.
.find_below ( tagOrId )
Returns the object ID of the object just below the one specified by tagOrId. If multiple objects
match, you get the lowest one.
.find_closest ( x, y, halo=None, start=None )
Returns the object ID of the object closest to point (x, y).
Use the halo argument to increase the effective size of the point. For example, halo=5 would treat
any object within 5 pixels of (x, y) as overlapping.
If an object ID is passed as the start argument, this method returns the highest qualifying object
that is below start in the display list.
.find_enclosed ( x1, y1, x2, y2 )
Returns a list of the object IDs of all objects that occur completely within the rectangle whose top
left corner is (x1, y1) and bottom right corner is (x2, y2).
.find_overlapping ( x1, y1, x2, y2 )
Like the previous method, but returns a list of the object IDs of all the objects that share at least one
point with the given rectangle.
.find_withtag ( tagOrId )
Returns a list of the object IDs of the object or objects specified by tagOrId.
.focus ( tagOrId=None )
Moves the focus to the object specified by tagOrId. If there are multiple such objects, moves the
focus to the first one in the display list that allows an insertion cursor. If there are no qualifying
items, or the canvas does not have focus, focus does not move.
If the argument is omitted, returns the ID of the object that has focus, or "" if none of them do.

.gettags ( tagOrId )
If tagOrId is an object ID, returns a list of all the tags associated with that object. If the argument
is a tag, returns all the tags for the lowest object that has that tag.
.icursor ( tagOrId, index )
Assuming that the selected item allows text insertion and has the focus, sets the insertion cursor to
index, which may be either an integer index or the string "end". Has no effect otherwise.
.index ( tagOrId, index )
Returns the integer index of the given index in the object specified by tagOrId (the lowest one
that allows text insertion, if tagOrId specifies multiple objects). The index argument is an integer
or the string "end".
.insert ( tagOrId, beforeThis, text )
Inserts the given string in the object or objects specified by tagOrId, before index beforethis
(which can be an integer or the string "end").
.itemcget ( tagOrId, option )
Returns the value of the given configuration option in the selected object (or the lowest object if
tagOrId specifies more than one). This is similar to the .cget() method for Tkinter objects.
.itemconfigure ( tagOrId, option, ... )
If no option arguments are supplied, returns a dictionary whose keys are the options of the object
specified by tagOrId (the lowest one, if tagOrId specifies multiple objects).
To change the configuration option of the specified item, supply one or more keyword arguments
of the form option=value.
.move ( tagOrId, xAmount, yAmount )
Moves the items specified by tagOrId by adding xAmount to their x coordinates and yAmount to
their y coordinates.
.postscript ( option, ... )
Generates an Encapsulated PostScript representation of the canvas's current contents. The options
include:
colormode Use "color" for color output, "gray" for grayscale, or "mono" for black and
white.
file If supplied, names a file where the PostScript will be written. If this option is
not given, the PostScript is returned as a string.
height How much of the Y size of the canvas to print. Default is all.
rotate If false, the page will be rendered in portrait orientation; if true, in landscape.
x Leftmost canvas coordinate of the area to print.
y Topmost canvas coordinate of the area to print.
width How much of the X size of the canvas to print. Default is all.
.scale ( tagOrId, xOrigin, yOrigin, xScale, yScale )

Scale all objects according to their distance from a point P=(xOrigin, yOrigin). The scale factors
xScale and yScale are based on a value of 1.0, which means no scaling. Every point in the objects
selected by tagOrId is moved so that its x distance from P is multiplied by xScale and its y distance
is multiplied by yScale.

.tag_bind ( tagOrId, sequence=None, function=None, add=None )
Binds events to objects on the canvas. For the object or objects selected by tagOrId, associates the
handler function with the event sequence. If the add argument is a string starting with "+",
the new binding is added to existing bindings for the given sequence, otherwise the new binding
replaces that for the given sequence.
For general information on event bindings, see Section24, Events (p. 75).
.tag_lower ( tagOrId, belowThis )
Moves the object or objects selected by tagOrId within the display list to a position just below the
first or only object specied by the tag or ID belowThis.
.tag_raise ( tagOrId, aboveThis )
Moves the object or objects selected by tagOrId within the display list to a position just above the
first or only object specied by the tag or ID aboveThis.
.tag_unbind ( tagOrId, sequence, funcId=None )
Removes bindings for handler funcId and event sequence from the canvas object or objects
specified by tagOrId. See Section24, Events (p. 75).
.type ( tagOrId )
Returns the type of the first or only object specified by tagOrId. The return value will be one of
the strings "arc", "bitmap", "image", "line", "oval", "polygon", "rectangle", "text",
or "window".
.xview ( MOVETO, fraction )
This method scrolls the canvas relative to its image, and is intended for binding to the command
option of a related scrollbar. The canvas is scrolled horizontally to a position given by offset,
where 0.0 moves the canvas to its leftmost position and 1.0 to its rightmost position.
.xview ( SCROLL, n, what )
This method moves the canvas left or right: the what argument specifies how much to move and
can be either UNITS or PAGES, and n tells how many units to move the canvas to the right relative
to its image (or left, if negative).
The size of the move for UNITS is given by the value of the canvas's xscrollincrement option;
see Section16, The Scrollbar widget (p. 48).
For movements by PAGES, n is multiplied by nine-tenths of the width of the canvas.
.xview_moveto ( fraction )
This method scrolls the canvas in the same way as .xview(MOVETO, fraction).
.xview_scroll ( n, what )
Same as .xview(SCROLL, n, what).
.yview ( MOVETO, fraction )
.yview ( SCROLL, n, what )
.yview_moveto ( fraction )
.yview_scroll ( n, what )
The vertical scrolling equivalents of .xview(), .xview_moveto(), and .xview_scroll().

6.3.The canvas arc object
An arc object on a canvas, in its most general form, is a wedge-shaped slice taken out of an ellipse. This
includes whole ellipses and circles as special cases. See Section6.7, The canvas oval object (p. 22) for
more on the geometry of the ellipse drawn.
To create an arc object on a canvas C, use:
id = C.create_arc ( x0, y0, x1, y1, option, ... )
The constructor returns the object ID of the new arc object on canvas C.
Point (x0, y0) is the top left corner and (x1, y1) the lower right corner of a rectangle into which the ellipse
is fit. If this rectangle is square, you get a circle.
The various options include:
extent Width of the slice in degrees. The slice starts at the angle given by the start option
and extends counterclockwise for extent degrees.
fill By default, the interior of an arc is transparent, and fill="" will select this beha-
vior. You can also set this option to any color and the interior of the arc will be
filled with that color.
outline The color of the border around the outside of the slice. Default is black.
outlinestipple If the outline option is used, this option specifies a bitmap used to stipple the
border. Default is black, and that default can be specified by setting outlines-
tipple="".
start Starting angle for the slice, in degrees, measured from +x direction. If omitted,
you get the entire ellipse.
stipple A bitmap indicating how the interior fill of the arc will be stippled. Default is
stipple="" (solid). You'll probably want something like stipple="gray25".
Has no effect unless fill has been set to some color.
style The default is to draw the whole arc; use style=PIESLICE for this style. To draw
only the circular arc at the edge of the slice, use style=ARC. To draw the circular
arc and the chord (a straight line connecting the endpoints of the arc), use
style=CHORD.
PIESLICE CHORD ARC
width Width of the border around the outside of the arc. Default is 1 pixel.
6.4.The canvas bitmap object

A bitmap object on a canvas is shown as two colors, the background color (for 0 data values) and the
foreground color (for 1 values).
To create a bitmap object on a canvas C, use:
id = C.create_bitmap ( x, y, *options ... )
which returns the integer ID number of the image object for that canvas.
The x and y values are the reference point that specifies where the bitmap is placed.

Options include:
anchor The bitmap is positioned relative to point (x, y). The default is anchor=CEN-
TER, meaning that the bitmap is centered on the (x, y) position. See Sec-
tion4.5, Anchors (p. 8) for the various anchor option values. For ex-
ample, if you specify anchor=NE, the bitmap will be positioned so that
point (x, y) is located at the northeast (top right) corner of the bitmap.
background The color that will appear where there are 0 values in the bitmap. The
default is background="", meaning transparent.
bitmap The bitmap to be displayed; see Section4.7, Bitmaps (p. 8).
foreground The color that will appear where there are 1 values in the bitmap. The
default is foreground="black".
tags The tags to be associated with the object, as a sequence of strings.
6.5.The canvas image object

To display a graphics image on a canvas C, use:
id = C.create_image ( x, y, option, ... )
This constructor returns the integer ID number of the image object for that canvas.
The image is positioned relative to point (x, y). Options include:
anchor The default is anchor=CENTER, meaning that the image is centered on the (x, y)
position. See Section4.5, Anchors (p. 8) for the possible values of this option.
For example, if you specify anchor=S, the image will be positioned so that point
(x, y) is located at the center of the bottom (south) edge of the image.
image The image to be displayed. See Section4.9, Images (p. 11), above, for information
about how to create images that can be loaded onto canvases.
tags The tags to be associated with the object, as a sequence of strings. See Section6.1.4,
Canvas tags (p. 15).
6.6.The canvas line object

In general, a line can consist of any number of segments connected end to end, and each segment can
be straight or curved. To create a canvas line object on a canvas C, use:
C.create_line ( x0, y0, x1, y1, ..., xn, yn, option, ... )
The line goes through the series of points (x0, y0), (x1, y1), (xn, yn). Options include:
arrow The default is for the line to have no arrowheads. Use arrow=FIRST to get an
arrowhead at the (x0, y0) end of the line. Use arrow=LAST to get an arrowhead
at the far end. Use arrow=BOTH for arrowheads at both ends.

arrowshape A tuple (d1, d2, d3) that describes the shape of the arrowheads added by the
arrow option. Default is (8,10,3).
d2
d3
d1
fill The color to use in drawing the line. Default is fill="black".

smooth If true, the line is drawn as a series of parabolic splines fitting the point set. Default
is false, which renders the line as a set of straight segments.
splinesteps If the smooth option is true, each spline is rendered as a number of straight line
segments. The splinesteps option specifies the number of segments used to
approximate each section of the line; the default is splinesteps=12.
stipple To draw a stippled line, set this option to a bitmap that specifies the stippling
pattern, such as stipple="gray25". See Section4.7, Bitmaps (p. 8) for the
possible values.
width The line's width. Default is 1 pixel. See Section4.1, Dimensions (p. 5) for possible
values.
6.7.The canvas oval object

Ovals, mathematically, are ellipses, including circles as a special case. The ellipse is fit into a rectangle
defined by the coordinates (x0, y0) of the top left corner and the coordinates (x1, y1) of the bottom
right corner:
(x0, y0)
(x1,y1)
The oval will coincide with the top and left-hand lines of this box, but will fit just inside the bottom and
right-hand sides.
To create an ellipse on a canvas C, use:
id = C.create_oval ( x0, y0, x1, y1, option, ... )
which returns the object ID of the new oval object on canvas C.
Options for ovals:
fill The default appearance of ellipse is transparent, and a value of fill="" will select
this behavior. You can also set this option to any color and the interior of the ellipse
will be filled with that color; see Section4.3, Colors (p. 6).

outline The color of the border around the outside of the ellipse. Default is out-
line="black".
stipple A bitmap indicating how the interior of the ellipse will be stippled. Default is
stipple="", which means a solid color. A typical value would be
stipple="gray25". Has no effect unless the fill has been set to some color.
See Section4.7, Bitmaps (p. 8).
width Width of the border around the outside of the ellipse. Default is 1 pixel; see Sec-
tion4.1, Dimensions (p. 5) for possible values. If you set this to zero, the border
will not appear. If you set this to zero and make the fill transparent, you can make
the entire oval disappear.
6.8.The canvas polygon object

As displayed, a polygon has two parts: its outline and its interior. Its geometry is specified as a series
of vertices [(x0, y0), (x1, y1), (xn, yn)], but the actual perimeter includes one more segment from (xn,
yn) back to (x0, y0). In this example, there are five vertices:
(x2,y2)
(x0,y0)
(x1,y1)
(x4,y4) (x3,y3)
To create a new polygon object on a canvas C:

id = C.create_polygon ( x0, y0, x1, y1, ..., option, ... )
The constructor returns the object ID for that object. Options:
fill You can color the interior by setting this option to a color. The default appearance
for the interior of a polygon is transparent, and you can set fill="" to get this
behavior. See Section4.3, Colors (p. 6).
outline Color of the outline; defaults to outline="black". Use outline="" to make
the outline transparent.
smooth The default outline uses straight lines to connect the vertices; use smooth=0 to
get that behavior. If you use smooth=1, you get a continuous spline curve.
Moreover, if you set smooth=1, you can make any segment straight by duplicating
the coordinates at each end of that segment.
splinesteps If the smooth option is true, each spline is rendered as a number of straight line
segments. The splinesteps option specifies the number of segments used to
approximate each section of the line; the default is splinesteps=12.
stipple A bitmap indicating how the interior of the polygon will be stippled. Default is

width Width of the outline; defaults to 1. See Section4.1, Dimensions (p. 5).
6.9.The canvas rectangle object

Each rectangle is specified as two points: (x0, y0) is the top left corner, and (x0, y0) is the bottom right
corner.
Rectangles are drawn in two parts:
The outline lies outside the canvas on its top and left sides, but inside the canvas on its bottom and
right side. The default appearance is a 1-pixel-wide black outline.
The fill is the area inside the outline. Its default appearance is transparent.
To create a rectangle object on canvas C:

id = C.create_rectangle ( x0, y0, x1, y1, option, ... )
This constructor returns the object ID of the rectangle on that canvas. Options include:
fill By default, the interior of a rectangle is empty, and you can get this behavior with
fill="". You can also set the option to a color; see Section4.3, Colors (p. 6).
outline The color of the border. Default is outline="black".
stipple A bitmap indicating how the interior of the rectangle will be stippled. Default is
width Width of the border. Default is 1 pixel. Use width=0 to make the border invisible.
See Section4.1, Dimensions (p. 5).
6.10.The canvas text object

You can display one or more lines of text on a canvas C by creating a text object:
id = C.create_text ( x, y, option, ... )
This returns the object ID of the text object on canvas C. Options include:
anchor The default is anchor=CENTER, meaning that the text is centered vertically and
horizontally around position (x, y). See Section4.5, Anchors (p. 8) for possible
values. For example, if you specify anchor=SW, the text will be positioned so its
lower left corner is at point (x, y).
fill The default text color is black, but you can render it in any color by setting the
fill option to that color. See Section4.3, Colors (p. 6).
font If you don't like the default font, set this option to any font value. See Section4.4,
Type fonts (p. 6).
justify For multi-line textual displays, this option controls how the lines are justified:
LEFT (the default), CENTER, or RIGHT.

stipple A bitmap indicating how the text will be stippled. Default is stipple="", which
means solid. A typical value would be stipple="gray25". See Section4.7,
Bitmaps (p. 8).
text The text to be displayed in the object, as a string. Use newline characters ("\n")
to force line breaks.
width If you don't specify a width option, the text will be set inside a rectangle as long
as the longest line. However, you can also set the width option to a dimension,
and each line of the text will be broken into shorter lines, if necessary, or even
broken within words, to fit within the specified width. See Section4.1, Dimen-
sions (p. 5).
6.11.The canvas window object

You can place any Tkinter widget onto a canvas by using a canvas window object. A window is a rectan-
gular area that can hold one Tkinter widget. The widget must be the child of the same top-level window
as the canvas, or the child of some widget located in the same top-level window.
If you want to put complex multi-widget objects on a canvas, you can use this method to place a Frame
widget on the canvas, and then place other widgets inside that frame.
To create a new canvas window object on a canvas C:
id = C.create_window ( x, y, option, ... )
This returns the object ID for the window object. Options include:
anchor The default is anchor=CENTER, meaning that the window is centered on the (x,
y) position. See Section4.5, Anchors (p. 8) for the possible values. For example,
if you specify anchor=E, the window will be positioned so that point (x, y) is on
the midpoint of its right-hand (east) edge.
height The height of the area reserved for the window. If omitted, the window will be
sized to fit the height of the contained widget. See Section4.1, Dimensions (p. 5)
for possible values.
width The width of the area reserved for the window. If omitted, the window will be
sized to fit the width of the contained widget.
window Use window=w where w is the widget you want to place onto the canvas. If this is
omitted initially, you can later call C.itemconfigure ( id, window=w) to
place the widget w onto the canvas, where id is the window's object ID..
7.The Checkbutton widget
The purpose of a checkbutton widget (sometimes called checkbox) is to allow the user to read and
select a two-way choice. The graphic above shows how checkbuttons look in the off (0) and on (1) state
in one implementation: this is a screen shot of two checkbuttons using 24-point Times font.

The indicator is the part of the checkbutton that shows its state, and the label is the text that appears beside
it.
You will need to create a control variable, an instance of the IntVar class, so your program can
query and set the state of the checkbutton. See Section22, Control variables: the values behind the
widgets (p. 72), below.
You can also use event bindings to react to user actions on the checkbutton; see Section24,
Events (p. 75), below.
You can disable a checkbutton. This changes its appearance to grayed out and makes it unresponsive
to the mouse.
You can get rid of the checkbutton indicator and make the whole widget a push-push button that
looks recessed when it is set, and looks raised when it is cleared.
To create a checkbutton in an existing parent window or frame master:

w = Checkbutton ( master, option, ... )
The constructor returns a new Checkbutton object. Options include:
activebackground Background color when the checkbutton is under the cursor. See Sec-
tion4.3, Colors (p. 6).
activeforeground Foreground color when the checkbutton is under the cursor.
anchor If the widget inhabits a space larger than it needs, this option specifies
where the checkbutton will sit in that space. The default is anchor=CEN-
TER. See Section4.5, Anchors (p. 8) for the allowable values. For ex-
ample, if you use anchor=NW, the widget will be placed in the upper left
corner of the space.
bg or background The normal background color displayed behind the label and indicator.
See Section4.3, Colors (p. 6).
bitmap To display a monochrome image on a button, set this option to a bitmap;
see Section4.7, Bitmaps (p. 8).
bd or borderwidth The size of the border around the indicator. Default is 2 pixels. For possible
values, see Section4.1, Dimensions (p. 5).
command A procedure to be called every time the user changes the state of this
checkbutton.
cursor If you set this option to a cursor name (see Section4.8, Cursors (p. 9)),
the mouse cursor will change to that pattern when it is over the checkbut-
ton.
disabledforeground The foreground color used to render the text of a disabled checkbutton.
The default is a stippled version of the default foreground color.
font The font used for the text. See Section4.4, Type fonts (p. 6).
fg or foreground The color used to render the text.
height The number of lines of text on the checkbutton. Default is 1.
highlightbackground The color of the focus highlight when the checkbutton does not have focus.
See Section23, Focus: routing keyboard input (p. 74).
highlightcolor The color of the focus highlight when the checkbutton has the focus.
highlightthickness The thickness of the focus highlight. Default is 1. Set to 0 to suppress
display of the focus highlight.

image To display a graphic image on the button, set this option to an image ob-
ject. See Section4.9, Images (p. 11).
indicatoron Normally a checkbutton displays as its indicator a box that shows
whether the checkbutton is set or not. You can get this behavior by setting
indicatoron=1. However, if you set indicatoron=0, the indicator
disappears, and the entire widget becomes a push-push button that looks
raised when it is cleared and sunken when it is set. You may want to in-
crease the borderwidth value to make it easier to see the state of such
a control.
justify If the text contains multiple lines, this option controls how the text is
justified: CENTER, LEFT, or RIGHT.
offvalue Normally, a checkbutton's associated control variable will be set to 0 when
it is cleared (off). You can supply an alternate value for the off state by
setting offvalue to that value.
onvalue Normally, a checkbutton's associated control variable will be set to 1 when
it is set (on). You can supply an alternate value for the on state by setting
onvalue to that value.
padx How much space to leave to the left and right of the checkbutton and text.
Default is 1 pixel. For possible values, see Section4.1, Dimensions (p. 5).
pady How much space to leave above and below the checkbutton and text.
Default is 1 pixel.
relief With the default value, relief=FLAT, the checkbutton does not stand
out from its background. You may set this option to any of the other styles
(see Section4.6, Relief styles (p. 8)), or use relief=SOLID, which
gives you a solid black frame around it.
selectcolor The color of the checkbutton when it is set. Default is selectcol-
or="red".
selectimage If you set this option to an image, that image will appear in the checkbut-
ton when it is set. See Section4.9, Images (p. 11).
state The default is state=NORMAL, but you can use state=DISABLED to
gray out the control and make it unresponsive. If the cursor is currently
over the checkbutton, the state is ACTIVE.
takefocus The default is that the input focus (see Section23, Focus: routing key-
board input (p. 74)) will pass through a checkbutton. If you set takefo-
cus=0, focus will not pass through it.
text The label displayed next to the checkbutton. Use newlines ("\n") to dis-
play multiple lines of text.
textvariable If you need to change the label on a checkbutton during execution, create
a StringVar (see Section22, Control variables: the values behind the
widgets (p. 72)) to manage the current value, and set this option to that
control variable. Whenever the control variable's value changes, the
checkbutton's annotation will automatically change as well.
underline With the default value of -1, none of the characters of the text label are
underlined. Set this option to the index of a character in the text (counting
from zero) to underline that character.
variable The control variable that tracks the current state of the checkbutton; see
Section22, Control variables: the values behind the widgets (p. 72).

Normally this variable is an IntVar, and 0 means cleared and 1 means
set, but see the offvalue and onvalue options above.
width The default width of a checkbutton is determined by the size of the dis-
played image or text. You can set this option to a number of characters
and the checkbutton will always have room for that many characters.
wraplength Normally, lines are not wrapped. You can set this option to a number of
characters and all lines will be broken into pieces no longer than that
number.
Methods on checkbuttons include:
.deselect()
Clears (turns off) the checkbutton.
.flash()
Flashes the checkbutton a few times between its active and normal colors, but leaves it the way it
started.
.invoke()
You can call this method to get the same actions that would occur if the user clicked on the check-
button to change its state.
.select()
Sets (turns on) the checkbutton.
.toggle()
Clears the checkbutton if set, sets it if cleared.
8.The Entry widget

The purpose of an Entry widget is to let the user see and modify a single line of text.
If you want to display multiple lines of text that can be edited, see Section17, The Text wid-
get (p. 51).
If you want to display one or more lines of text that cannot be modified by the user, see Section10,
The Label widget (p. 33).
Some definitions:
The selection is a highlighted region of the text in an Entry widget, if there is one.
Typically the selection is made by the user with the mouse, and selected text is copied to the system's
clipboard. However, Tkinter allows you to control whether or not selected text gets copied to the
clipboard. You can also select text in an Entry under program control.
The insertion cursor shows where new text will be inserted. It is displayed only when the user clicks
the mouse somewhere in the widget. It usually appears as a blinking vertical line inside the widget.
You can customize its appearance in several ways.
Positions within the widget's displayed text are given as an index. There are several ways to specify
an index:
As normal Python indexes, starting from 0.
The constant END refers to the position after the existing text.
The constant INSERT refers to the current position of the insertion cursor.

The constant ANCHOR refers to the first character of the selection, if there is a selection.
You may need to figure out which character position in the widget corresponds to a given mouse
position. To simplify that process, you can use as an index a string of the form "@n", where n is
the horizontal distance in pixels between the left edge of the Entry widget and the mouse. Such
an index will specify the character at that horizontal mouse position.
To create a new Entry widget in a root window or frame named master:

w = Entry ( master, option, ... )
This constructor returns the entry widget object. Options include:
bg or background The background color inside the entry area. Default is a light gray.
bd or borderwidth The width of the border around the entry area. Default is 2.
cursor The cursor used when the mouse is within the entry widget; see Sec-
tion4.8, Cursors (p. 9).
font The font used for text entered in the widget by the user. See Section4.4,
Type fonts (p. 6).
exportselection By default, if you select text within an Entry widget, it is automatically
exported to the clipboard. To avoid this exportation, use exportselec-
tion=0.
fg or foreground The color used to render the text. Default is black.
highlightbackground Color of the focus highlight when the widget does not have focus. See
Section23, Focus: routing keyboard input (p. 74).
highlightcolor Color shown in the focus highlight when the widget has the focus.
insertbackground By default, the insertion cursor (which shows the point within the text
where new keyboard input will be inserted) is black. To get a different
color of insertion cursor, set insertbackground to any color; see Sec-
insertborderwidth By default, the insertion cursor is a simple rectangle. You can get the
cursor with the RAISED relief effect (see Section4.6, Relief styles (p. 8))
by setting insertborderwidth to the dimension of the 3-d border. If
you do, make sure that the insertwidth attribute is at least twice that
value.
insertofftime By default, the insertion cursor blinks. You can set insertofftime to
a value in milliseconds to specify how much time the insertion cursor
spends off. Default is 300. If you use insertofftime=0, the insertion
cursor won't blink at all.
insertontime Similar to insertofftime, this attribute specifies how much time the
cursor spends on per blink. Default is 600 (milliseconds).
insertwidth By default, the insertion cursor is 2 pixels wide. You can adjust this by
setting insertwidth to any dimension.
justify This option controls how the text is justified when the text doesn't fill the
widget's width. The value can be LEFT (the default), CENTER, or RIGHT.
relief Selects three-dimensional shading effects around the text entry. See Sec-
tion4.6, Relief styles (p. 8). The default is relief=SUNKEN

selectbackground The background color to use displaying selected text. See Section4.3,
Colors (p. 6).
selectborderwidth The width of the border to use around selected text. The default is one
pixel.
selectforeground The foreground (text) color of selected text.
show Normally, the characters that the user types appear in the entry. To make
a password entry that echoes each character as an asterisk, set
show="*".
state Use this attribute to disable the Entry widget so that the user can't type
anything into it. Use state=DISABLE to disable the widget,
state=NORMAL to allow user input again. Your program can also find
out whether the cursor is currently over the widget by interrogating this
attribute; it will have the value ACTIVE when the mouse is over it.
takefocus By default, the focus will tab through entry widgets. Set this option to 0
to take the widget out of the sequence. For a discussion of focus, see Sec-
tion23, Focus: routing keyboard input (p. 74).
textvariable In order to be able to retrieve the current text from your entry widget,
you must set this option to an instance of the StringVar class; see Sec-
tion22, Control variables: the values behind the widgets (p. 72). You
can retrieve the text using v.get(), or set it using v.set(), where v is
the associated control variable.
width The size of the entry in characters. The default is 20. For proportional
fonts, the physical length of the widget will be based on the average width
of a character times the value of the width option.
xscrollcommand If you expect that users will often enter more text than the onscreen size
of the widget, you can link your entry widget to a scrollbar. Set this option
to the .set method of the scrollbar. For more information, see Section8.1,
Scrolling an Entry widget (p. 31).
Methods on Entry objects include:
.delete ( first, last=None )

Deletes characters from the widget, starting with the one at index first, up to but not including
the character at position last. If the second argument is omitted, only the single character at position
first is deleted.
.get()
Returns the entry's current text as a string.
.icursor ( index )
Set the insertion cursor just before the character at the given index.
.index ( index )
Shift the contents of the entry so that the character at the given index is the leftmost visible character.
Has no effect if the text fits entirely within the entry.
.insert ( index, s )
Inserts string s before the character at the given index.

.select_adjust ( index )
This method is used to make sure that the selection includes the character at the specified index.
If the selection already includes that character, nothing happens. If not, the selection is expanded
from its current position (if any) to include position index.
.select_clear()
Clears the selection. If there isn't currently a selection, has no effect.
.select_from ( index )
Sets the ANCHOR index position to the character selected by index, and selects that character.
.select_present()
If there is a selection, returns true, else returns false.
.select_range ( start, end )
Sets the selection under program control. Selects the text starting at the start index, up to but
not including the character at the end index. The start position must be before the end position.
To select all the text in an entry widget e, use e.select_range(0, END).
.select_to ( index )
Selects all the text from the ANCHOR position up to but not including the character at the given index.
.xview ( index )
Same as .xview(). This method is useful in linking the Entry widget to a horizontal scrollbar.
See Section8.1, Scrolling an Entry widget (p. 31).
.xview_moveto ( f )
Positions the text in the entry so that the character at position f, relative to the entire text, is positioned
at the left edge of the window. The f argument must be in the range [0,1], where 0 means the left
end of the text and 1 the right end.
.xview_scroll ( number, what )
Used to scroll the entry horizontally. The what argument must be either UNITS, to scroll by character
widths, or PAGES, to scroll by chunks the size of the entry widget. The number is positive to scroll
left to right, negative to scroll right to left. For example, for an entry widget e, e.xview_scroll(-
1, PAGES) would move the text one page to the right, and e.xview_scroll(4, UNITS)
would move the text four characters to the left.
8.1.Scrolling an Entry widget

Making an Entry widget scrollable requires a little extra code on your part to adapt the Scrollbar
widget's callback to the methods available on the Entry widget. Here are some code fragments illus-
trating the setup. First, the creation and linking of the Entry and Scrollbar widgets:
self.entry = Entry ( self, width=10 )
self.entry.grid(row=0, sticky=E+W)
self.entryScroll = Scrollbar ( self, orient=HORIZONTAL,

command=self.__scrollHandler )
self.entryScroll.grid(row=1, sticky=E+W)
self.entry["xscrollcommand"] = self.entryScroll.set
Here's the adapter function referred to above:
def __scrollHandler(self, *L):
op, howMany = L[0], L[1]
if op == "scroll":

units = L[2]
self.entry.xview_scroll ( howMany, units )
elif op == "moveto":
self.entry.xview_moveto ( howMany )
9.The Frame widget

A frame is basically just a container for other widgets.
Your application's root window is basically a frame.
Each frame has its own grid layout, so the gridding of widgets within each frame works independ-
ently.
Frame widgets are a valuable tool in making your application modular. You can group a set of related
widgets into a compound widget by putting them into a frame. Better yet, you can declare a new
class that inherits from Frame, adding your own interface to it. This is a good way to hide the details
of interactions within a group of related widgets from the outside world.
To create a new frame widget in a root window or frame named master:

w = Frame ( master, option, ... )
The constructor returns the new frame widget. Options:
bg or background The frame's background color. See Section4.3, Colors (p. 6).
bd or borderwidth Width of the frame's border. The default is 0 (no border). For permitted
cursor The cursor used when the mouse is within the frame widget; see Sec-
height The vertical dimension of the new frame. This will be ignored unless you
also call .grid_propagate(0) on the frame.
highlightbackground Color of the focus highlight when the frame does not have focus. See
highlightcolor Color shown in the focus highlight when the frame has the focus.
relief The default relief for a frame is FLAT, which means the frame will blend
in with its surroundings. To put a border around a frame, set its border-
width to a positive value and set its relief to one of the standard relief
types; see Section4.6, Relief styles (p. 8).
takefocus Normally, frame widgets are not visited by input focus (see Section23,
Focus: routing keyboard input (p. 74) for an overview of this topic).
However, you can set takefocus=1 if you want the frame to receive
keyboard input. To handle such input, you will need to create bindings
for keyboard events; see Section24, Events (p. 75) for more on events
and bindings.
width The horizontal dimension of the new frame. See Section4.1, Dimen-
sions (p. 5). This value be ignored unless you also call
.grid_propagate(0) on the frame.

10.The Label widget
Label widgets can display one or more lines of text in the same style, or a bitmap or image. To create a
label widget in a root window or frame master:
w = Label ( master, option, ... )
The constructor returns the label widget. Options include:
anchor This options controls where the text is positioned if the widget has more
space than the text needs. The default is anchor=CENTER, which centers
the text in the available space. For other values, see Section4.5, An-
chors (p. 8). For example, if you use anchor=NW, the text would be
positioned in the upper left-hand corner of the available space.
bg or background The background color of the label area. See Section4.3, Colors (p. 6).
bitmap Set this option equal to a bitmap or image object and the label will display
that graphic. See Section4.7, Bitmaps (p. 8) and Section4.9, Im-
ages (p. 11).
bd or borderwidth Width of the border around the label. Default is 2.
cursor Cursor that appears when the mouse is over this label. See Section4.8,
Cursors (p. 9).
font If you are displaying text in this label (with the text or textvariable
option, the font option specifies in what font that text will be displayed.
See Section4.4, Type fonts (p. 6).
fg or foreground If you are displaying text or a bitmap in this label, this option specifies
the color of the text. If you are displaying a bitmap, this is the color that
will appear at the position of the 1-bits in the bitmap. See Section4.3,
Colors (p. 6).
height Height of the label in lines (not pixels!). If this option is not set, the label
will be sized to fit its contents.
image To display a static image in the label widget, set this option to an image
object. See Section4.9, Images (p. 11).
justify Specifies how multiple lines of text will be aligned with respect to each
other: LEFT for flush left, CENTER for centered (the default), or RIGHT
for right-justified.
padx Extra space added to the left and right of the text within the widget. De-
fault is 1.
pady Extra space added above and below the text within the widget. Default
is 1.
relief Specifies the appearance of a decorative border around the label. The
default is FLAT; for other values, see Section4.6, Relief styles (p. 8).
takefocus Normally, focus does not cycle through Label widgets; see Section23,
Focus: routing keyboard input (p. 74). If you want this widget to be
visited by the focus, set takefocus=1.
text To display one or more lines of text in a label widget, set this option to a
string containing the text. Internal newlines ("\n") will force a line break.

textvariable To slave the text displayed in a label widget to a control variable of class
StringVar, set this option to that variable. SeeSection22, Control
variables: the values behind the widgets (p. 72).
underline You can display an underline (_) below the nth letter of the text, counting
from 0, by setting this option to n. The default is underline=-1, which
means no underlining.
width Width of the label in characters (not pixels!). If this option is not set, the
label will be sized to fit its contents.
wraplength You can limit the number of characters in each line by setting this option
to the desired number. The default value, 0, means that lines will be
broken only at newlines.
There are no special methods for label widgets other than the common ones (see Section19, Universal
widget methods (p. 62)).
11.The Listbox widget

The purpose of a listbox widget is to display a set of lines of text. Generally they are intended to allow
the user to select one or more items from a list. If you need something more like a text editor, see Sec-
tion17, The Text widget (p. 51).
To create a new listbox widget inside a root window or frame master:
w = Listbox ( master, option, ... )
This constructor returns the new listbox widget. Options:
bg or background The background color in the listbox.

bd or borderwidth The width of the border around the listbox. Default is 2. For possible
cursor The cursor that appears when the mouse is over the listbox. See Section4.8,
Cursors (p. 9).
font The font used for the text in the listbox. See Section4.4, Type
fonts (p. 6).
fg or foreground The color used for the text in the listbox. See Section4.3, Colors (p. 6).
height Number of lines (not pixels!) shown in the listbox. Default is 10.
relief Selects three-dimensional border shading effects. The default is SUNKEN.
For other values, see Section4.6, Relief styles (p. 8).
selectbackground The background color to use displaying selected text.
selectborderwidth The width of the border to use around selected text. The default is that
the selected item is shown in a solid block of color selectbackground;
if you increase the selectborderwidth, the entries are moved farther
apart and the selected entry shows RAISED relief (see Section4.6, Relief
styles (p. 8)).

selectforeground The foreground color to use displaying selected text.
selectmode Determines how many items can be selected, and how mouse drags affect
the selection:
BROWSE: Normally, you can only select one line out of a listbox. If you
click on an item and then drag to a different line, the selection will
follow the mouse. This is the default.
SINGLE: You can only select one line, and you can't drag the
mousewherever you click button 1, that line is selected.
MULTIPLE: You can select any number of lines at once. Clicking on
any line toggles whether or not it is selected.
EXTENDED: You can select any adjacent group of lines at once by
clicking on the first line and dragging to the last line.
takefocus Normally, the focus will tab through listbox widgets. Set this option to 0
to take the widget out of the sequence. See Section23, Focus: routing
keyboard input (p. 74).
width The width of the widget in characters (not pixels!). The width is based on
an average character, so some strings of this length in proportional fonts
may not fit. The default is 20.
xscrollcommand If you want to allow the user to scroll the listbox horizontally, you can
link your listbox widget to a horizontal scrollbar. Set this option to the
.set method of the scrollbar. See Section11.1, Scrolling a Listbox
widget (p. 37) for more on scrollable listbox widgets.
yscrollcommand If you want to allow the user to scroll the listbox vertically, you can link
your listbox widget to a vertical scrollbar. Set this option to the .set
method of the scrollbar. See Section11.1, Scrolling a Listbox wid-
get (p. 37).
A special set of index forms is used for many of the methods on listbox objects:
If you specify an index as an integer, it refers to the line in the listbox with that index, counting from
0.
Index END refers to the last line in the listbox.
Index ACTIVE refers to the selected line. If the listbox allows multiple selections, it refers to the line
that was last selected.
An index string of the form "@x,y" refers to the line closest to coordinate (x,y) relative to the widget's
upper left corner.
Methods on listbox objects include:
.activate ( index )
Selects the line specifies by the given index.
.curselection()
Returns a tuple containing the line numbers of the selected element or elements, counting from 0.
If nothing is selected, returns an empty tuple.

.delete ( first, last=None )
Deletes the lines whose indices are in the range [first, last], inclusive (contrary to the usual
Python idiom, where deletion stops short of the last index), counting from 0. If the second argument
is omitted, the single line with index first is deleted.
.get ( first, last=None )
Returns a tuple containing the text of the lines with indices from first to last, inclusive. If the
second argument is omitted, returns the text of the line closest to first.
.index ( i )
If possible, positions the visible part of the listbox so that the line containing index i is at the top
of the widget.
.insert ( index, *elements )
Insert one or more new lines into the listbox before the line specified by index. Use END as the first
argument if you want to add new lines to the end of the listbox.
.nearest ( y )
Return the index of the visible line closest to the y-coordinate y relative to the listbox widget.
.see ( index )
Adjust the position of the listbox so that the line referred to by index is visible.
.selection_clear ( first, last=None )
Unselects all of the lines between indices first and last, inclusive. If the second argument is
omitted, unselects the line with index first.
.selection_includes ( index )
Returns 1 if the line with the given index is selected, else returns 0.
.selection_set ( first, last=None )
Selects all of the lines between indices first and last, inclusive. If the second argument is omitted,
selects the line with index first.
.size()
Returns the number of lines in the listbox.
.xview()
To make the listbox horizontally scrollable, set the command option of the associated horizontal
scrollbar to this method. See Section11.1, Scrolling a Listbox widget (p. 37).
Scroll the listbox so that the leftmost fraction of the width of its longest line is outside the left
side of the listbox. Fraction is in the range [0,1].
.xview_scroll ( number, what )
Scrolls the listbox horizontally. For the what argument, use either UNITS to scroll by characters, or
PAGES to scroll by pages, that is, by the width of the listbox. The number argument tells how many
to scroll; negative values move the text to the right within the listbox, positive values leftward.
.yview()
To make the listbox vertically scrollable, set the command option of the associated vertical scrollbar
to this method. See Section11.1, Scrolling a Listbox widget (p. 37).
.yview_moveto ( fraction )
Scroll the listbox so that the top fraction of the width of its longest line is outside the left side of
the listbox. Fraction is in the range [0,1].

.yview_scroll ( number, what )
Scrolls the listbox vertically. For the what argument, use either UNITS to scroll by lines, or PAGES
to scroll by pages, that is, by the height of the listbox. The number argument tells how many to
scroll; negative values move the text downward inside the listbox, and positive values move the
text up.
11.1.Scrolling a Listbox widget

Here is a code fragment illustrating the creation and linking of a listbox to both a horizontal and a ver-
tical scrollbar.
self.yScroll = Scrollbar ( self, orient=VERTICAL )
self.yScroll.grid ( row=0, column=1, sticky=N+S )
self.xScroll = Scrollbar ( self, orient=HORIZONTAL )

self.xScroll.grid ( row=1, column=0, sticky=E+W )
self.listbox = Listbox ( self,

xscrollcommand=self.xScroll.set,
yscrollcommand=self.yScroll.set )
self.listbox.grid ( row=0, column=0, sticky=N+S+E+W )
self.xScroll["command"] = self.listbox.xview
self.yScroll["command"] = self.listbox.yview
12.The Menu widget

Drop-down menus are a popular way to present the user with a number of choices, yet take up min-
imal space on the face of the application the rest of the time.
A menubutton is the part that always appears on the application.
A menu is the list of choices that appears only after the user clicks on the menubutton.
To select a choice, the user can drag the mouse from the menubutton down onto one of the choices.
Alternatively, they can click and release the menubutton: the choices will appear and stay until the
user clicks one of them.
The Unix version of Tkinter (at least) supports tear-off menus. If you as the designer wish it, a
dotted line will appear above the choices. The user can click on this line to tear off the menu: a
new, separate, independent window appears containing the choices.
Refer to Section13, The Menubutton widget (p. 40), below, to see how to create a menubutton and
connect it to a menu widget. First let's look at the Menu widget, which displays the list of choices.
The choices displayed on a menu may be any of these things:
A simple command: a text string (or image) that the user can select to perform some operation.
A cascade: a text string or image that the user can select to show another whole menu of choices.
A checkbutton (see Section7, The Checkbutton widget (p. 25)).
A group of radiobuttons (see Section14, The Radiobutton widget (p. 42)).
To create a menu widget, you must first have created a Menubutton, which we will call mb:
w = Menu ( mb, option, ... )
This constructor returns a new menu widget. Options include:

activebackground The background color that will appear on a choice when it is under the
mouse. See Section4.3, Colors (p. 6).
activeborderwidth Specifies the width of a border drawn around a choice when it is under
the mouse. Default is 1 pixel. For possible values, see Section4.1, Dimen-
sions (p. 5).
activeforeground The foreground color that will appear on a choice when it is under the
mouse.
bg or background The background color for choices not under the mouse.
bd or borderwidth The width of the border around all the choices. Default is 1.
cursor The cursor that appears when the mouse is over the choices, but only
when the menu has been torn off. See Section4.8, Cursors (p. 9).
disabledforeground The color of the text for items whose state is DISABLED.
font The default font for textual choices. See Section4.4, Type fonts (p. 6).
fg or foreground The foreground color used for choices not under the mouse.
postcommand You can set this option to a procedure, and that procedure will be called
every time someone brings up this menu.
relief The default 3-D effect for menus is relief=RAISED. For other options,
see Section4.6, Relief styles (p. 8).
selectcolor Specifies the color displayed in checkbuttons and radiobuttons when they
are selected.
tearoff Normally, a menu can be torn off, the first position (position 0) in the list
of choices is occupied by the tear-off element, and the additional choices
are added starting at position 1. If you set tearoff=0, the menu will not
have a tear-off feature, and choices will be added starting at position 0.
title Normally, the title of a tear-off menu window will be the same as the text
of the menubutton or cascade that lead to this menu. If you want to change
the title of that window, set the title option to that string.
tearoffcommand If you would like your program to be notified when the user clicks on
the tear-off entry in a menu, set this option to your procedure. It will be
called with two arguments: the window ID of the parent window, and
the window ID of the new tear-off menu's root window.
These methods are available on Menu objects. The ones that create choices on the menu have their own
particular options; see Section12.1, Menu item creation (coption) options (p. 39).
.add_cascade ( coption, ... )

Add a new cascade element as the next available choice in self. Use the menu option in this call to
connect the cascade to the next level's menu, an object of type Menu.
.add_checkbutton ( coption, ... )
Add a new checkbutton as the next available choice in self. The options allow you to set up the
checkbutton much the same way as you would set up a Checkbutton object; see Section12.1, Menu
item creation (coption) options (p. 39).
.add_command ( coption, ... )
Add a new command as the next available choice in self. Use the label, bitmap, or image option
to place text or an image on the menu; use the command option to connect this choice to a procedure
that will be called when this choice is picked.

.add_radiobutton ( coption, ... )
Add a new radiobutton as the next available choice in self. The options allow you to set up the ra-
diobutton in much the same way as you would set up a Radiobutton object; see Section14, The
Radiobutton widget (p. 42).
.add_separator()
Add a separator after the last currently defined option. This is just a ruled horizontal line you can
use to set off groups of choices. Separators are counted as choices, so if you already have three
choices, and you add a separator, the separator will occupy position 3 (counting from 0).
.delete ( index1, index2=None )
This method deletes the choices numbered from index1 through index2, inclusive. To delete one
choice, omit the index2 argument. You can't use this method to delete a tear-off choice, but you
can do that by setting the menu object's tearoff option to 0.
.entrycget ( index, coption )
To retrieve the current value of some coption for a choice, call this method with index set to the
index of that choice and coption set to the name of the desired option.
.entryconfigure ( index, coption, ... )
To change the current value of some coption for a choice, call this method with index set to the
index of that choice and one or more coption=value arguments.
.index ( i )
Returns the position of the choice specified by index i. For example, you can use .index(END) to
find the index of the last choice (or None if there are no choices).
.insert_cascade ( index, coption, ... )
Inserts a new cascade at the position given by index, counting from 0. Any choices after that position
move down one. The options are the same as for .add_cascade(), above.
.insert_checkbutton ( index, coption, ... )
Insert a new checkbutton at the position specified by index. Options are the same as for
.add_checkbutton(), above.
.insert_command ( index, coption, ... )
Insert a new command at position index. Options are the same as for .add_command(), above.
.insert_radiobutton ( index, coption, ... )
Insert a new radiobutton at position index. Options are the same as for .add_radiobutton(),
above.
.insert_separator ( index )
Insert a new separator at the position specified by index.
.invoke ( index )
Calls the command callback associated with the choice at position index. If a checkbutton, its state
is toggled between set and cleared; if a radiobutton, that choice is set.
.type ( index )
Returns the type of the choice specified by index: either "cascade", "checkbutton", "command",
"radiobutton", "separator", or "tearoff".
12.1.Menu item creation (coption) options

These are the possible values of the coption choice options used in the methods above:

activebackground The background color used for choices when they are under the mouse.
activeforeground The foreground color used for choices when they are under the mouse.
background The background color used for choices when they are not under the mouse.
Note that this cannot be abbreviated as bg.
bitmap Display a bitmap for this choice; see Section4.7, Bitmaps (p. 8).
columnbreak Normally all the choices are displayed in one long column. If you set
columnbreak=1, this choice will start a new column to the right of the
one containing the previous choice.
command A procedure to be called when this choice is activated.
font The font used to render the label text. See Section4.4, Type fonts (p. 6)
foreground The foreground color used for choices when they are not under the mouse.
Note that this cannot be abbreviated as fg.
image Display an image for this choice; see Section4.9, Images (p. 11).
label The text string to appear for this choice.
menu This option is used only for cascade choices. Set it to a Menu object that
displays the next level of choices.
offvalue Normally, the control variable for a checkbutton is set to 0 when the
checkbutton is off. You can change the off value by setting this option to
the desired value. See Section22, Control variables: the values behind
the widgets (p. 72).
onvalue Normally, the control variable for a checkbutton is set to 1 when the
checkbutton is on. You can change the on value by setting this option to
the desired value.
selectcolor Normally, the color displayed in a set checkbutton or radiobutton is red.
Change that color by setting this option to the color you want; see Sec-
state Normally, all choices react to mouse clicks, but you can set state=DIS-
ABLED to gray it out and make it unresponsive.
underline Normally none of the letters in the label are underlined. Set this option
to the index of a letter to underline that letter.
value Specifies the value of the associated control variable (see Section22,
Control variables: the values behind the widgets (p. 72)) for a radiobut-
ton. This can be an integer if the control variable is an IntVar, or a string
if the control variable is a StringVar.
variable For checkbuttons or radiobuttons, this option should be set to the control
variable associated with the checkbutton or group of radiobuttons. See
Section22, Control variables: the values behind the widgets (p. 72).
13.The Menubutton widget

A menubutton is the part of a drop-down menu that stays on the screen all the time. Every menubutton
is associated with a Menu widget (see above) that can display the choices for that menubutton when
the user clicks on it.
To create a menubutton within a root window or frame master:
w = Menubutton ( master, option, ... )

The constructor returns the new menubutton widget. Options:
activebackground The background color when the mouse is over the menubutton. See Sec-
activeforeground The foreground color when the mouse is over the menubutton.
anchor This options controls where the text is positioned if the widget has more
space than the text needs. The default is anchor=CENTER, which centers
the text. For other options, see Section4.5, Anchors (p. 8). For example,
if you use anchor=W, the text would be centered against the left side of
the widget.
bg or background The background color when the mouse is not over the menubutton.
bitmap To display a bitmap on the menubutton, set this option to a bitmap name;
see Section4.7, Bitmaps (p. 8).
bd or borderwidth Width of the border around the menubutton. Default is 2 pixels. For
possible values, see Section4.1, Dimensions (p. 5).
cursor The cursor that appears when the mouse is over this menubutton. See
Section4.8, Cursors (p. 9).
direction Normally, the menu will appear below the menubutton. Set direc-
tion=LEFT to display the menu to the left of the button; use direc-
tion=RIGHT to display the menu to the right of the button; or use dir-
ection=ABOVE to place the menu above the button.
disabledforeground The foreground color shown on this menubutton when it is disabled.
fg or foreground The foreground color when the mouse is not over the menubutton.
height The height of the menubutton in lines of text (not pixels!). The default is
to fit the menubutton's size to its contents.
image To display an image on this menubutton, set this option to the image
object. See Section4.9, Images (p. 11).
justify This option controls where the text is located when the text doesn't fill
the menubutton: use justify=LEFT to left-justify the text (this is the
default); use justify=CENTER to center it, or justify=RIGHT to right-
justify.
menu To associate the menubutton with a set of choices, set this option to the
Menu object containing those choices. That menu object must have been
created by passing the associated menubutton to the constructor as its
first argument. See below for an example showing how to associate a
menubutton and menu.
padx How much space to leave to the left and right of the text of the
menubutton. Default is 1.
pady How much space to leave above and below the text of the menubutton.
Default is 1.

relief Normally, menubuttons will have RAISED appearance. For other 3-d ef-
fects, see Section4.6, Relief styles (p. 8).
state Normally, menubuttons respond to the mouse. Set state=DISABLED to
gray out the menubutton and make it unresponsive.
text To display text on the menubutton, set this option to the string containing
the desired text. Newlines ("\n") within the string will cause line breaks.
textvariable You can associate a control variable of class StringVar with this
menubutton. Setting that control variable will change the displayed text.
See Section22, Control variables: the values behind the widgets (p. 72).
underline Normally, no underline appears under the text on the menubutton. To
underline one of the characters, set this option to the index of that charac-
ter.
width Width of the menubutton in characters (not pixels!). If this option is not
set, the label will be sized to fit its contents.
number.
Here is a brief example showing the creation of a menubutton and its associated menu with three
checkboxes:
self.mb = Menubutton ( self, text="condiments",
relief=RAISED )
self.mb.grid()
self.mb.menu = Menu ( self.mb, tearoff=0 )

self.mb["menu"] = self.mb.menu
self.mayoVar = IntVar()
self.ketchVar = IntVar()
self.mb.menu.add_checkbutton ( label="mayo",
variable=self.mayoVar )
self.mb.menu.add_checkbutton ( label="ketchup",
variable=self.ketchVar )
This example creates a menubutton labeled condiments. When clicked, two checkbuttons labeled
mayo and ketchup will drop down.
14.The Radiobutton widget

Radiobuttons are sets of related widgets that allow the user to select only one of a set of choices. Each
radiobutton consists of two parts, the indicator and the label:
The indicator is the diamond-shaped part that turns red in the selected item.
The label is the text, although you can use an image or bitmap as the label.

If you prefer, you can dispense with the indicator. This makes the radiobuttons look like push-
push buttons, with the selected entry appearing sunken and the rest appearing raised.
To form several radiobuttons into a functional group, create a single control variable (see Section22,
Control variables: the values behind the widgets (p. 72), below), and set the variable option of
each radiobutton to that variable.
The control variable can be either an IntVar or a StringVar. If two or more radiobuttons share
the same control variable, setting any of them will clear the others.
Each radiobutton in a group must have a unique value option of the same type as the control
variable. For example, a group of three radiobuttons might share an IntVar and have values of 0,
1, and 99. Or you can use a StringVar control variable and give the radiobuttons value options
like "too hot", "too cold", and "just right".
To create a new radiobutton widget as the child of a root window or frame named master:
w = Radiobutton ( master, option, ... )
This constructor returns the new radiobutton widget. Options:
activebackground The background color when the mouse is over the radiobutton. See Sec-
activeforeground The foreground color when the mouse is over the radiobutton.
anchor If the widget inhabits a space larger than it needs, this option specifies
where the radiobutton will sit in that space. The default is anchor=CEN-
TER. For other positioning options, see Section4.5, Anchors (p. 8). For
example, if you set anchor=NE, the radiobutton will be placed in the top
right corner of the available space.
bg or background The normal background color behind the indicator and label.
bitmap To display a monochrome image on a radiobutton, set this option to a
bitmap; see Section4.7, Bitmaps (p. 8).
borderwidth The size of the border around the indicator part itself. Default is 2 pixels.
For possible values, see Section4.1, Dimensions (p. 5).
command A procedure to be called every time the user changes the state of this ra-
diobutton.
cursor If you set this option to a cursor name (see Section4.8, Cursors (p. 9)),
the mouse cursor will change to that pattern when it is over the radiobut-
ton.
disabledforeground The foreground color used to render the text of a disabled radiobutton.
The default is a stippled version of the default foreground color.
font The font used for the text. See Section4.4, Type fonts (p. 6).
fg or foreground The color used to render the text.
height The number of lines (not pixels) of text on the radiobutton. Default is 1.
highlightbackground The color of the focus highlight when the radiobutton does not have focus.
highlightcolor The color of the focus highlight when the radiobutton has the focus.
highlightthickness The thickness of the focus highlight. Default is 1. Set highlightthick-
ness=0 to suppress display of the focus highlight.

image To display a graphic image instead of text for this radiobutton, set this
option to an image object. See Section4.9, Images (p. 11). The image
appears when the radiobutton is not selected; compare selectimage,
below.
indicatoron Normally a radiobutton displays its indicator. If you set this option to
zero, the indicator disappears, and the entire widget becomes a push-
push button that looks raised when it is cleared and sunken when it is
set. You may want to increase the borderwidth value to make it easier
to see the state of such a control.
justify If the text contains multiple lines, this option controls how the text is
justified: CENTER (the default), LEFT, or RIGHT.
padx How much space to leave to the left and right of the radiobutton and text.
Default is 1.
pady How much space to leave above and below the radiobutton and text.
Default is 1.
relief By default, a radiobutton will have FLAT relief, so it doesn't stand out
from its background. See Section4.6, Relief styles (p. 8) for more 3-d
effect options. You can also use relief=SOLID, which displays a solid
black frame around the radiobutton.
selectcolor The color of the radiobutton when it is set. Default is red.
selectimage If you are using the image option to display a graphic instead of text
when the radiobutton is cleared, you can set the selectimage option
to a different image that will be displayed when the radiobutton is set.
See Section4.9, Images (p. 11).
state The default is state=NORMAL, but you can set state=DISABLED to gray
out the control and make it unresponsive. If the cursor is currently over
the radiobutton, the state is ACTIVE.
takefocus By default, the input focus (see Section23, Focus: routing keyboard in-
put (p. 74)) will pass through a radiobutton. If you set takefocus=0,
focus will not visit this radiobutton.
text The label displayed next to the radiobutton. Use newlines ("\n") to dis-
play multiple lines of text.
textvariable If you need to change the label on a radiobutton during execution, create
a StringVar (see Section22, Control variables: the values behind the
widgets (p. 72)) to manage the current value, and set this option to that
control variable. Whenever the control variable's value changes, the radi-
obutton's annotation will automatically change to that text as well.
underline With the default value of -1, none of the characters of the text label are
underlined. Set this option to the index of a character in the text (counting
from zero) to underline that character.
value When a radiobutton is turned on by the user, its control variable is set to
its current value option. If the control variable is an IntVar, give each
radiobutton in the group a different integer value option. If the control
variable is a StringVar, give each radiobutton a different string value
option.

variable The control variable that this radiobutton shares with the other radiobut-
tons in the group; see Section22, Control variables: the values behind
the widgets (p. 72). This can be either an IntVar or a StringVar.
width The default width of a radiobutton is determined by the size of the dis-
played image or text. You can set this option to a number of characters
(not pixels) and the radiobutton will always have room for that many
characters.
number.
Methods on radiobutton objects include:
.deselect()
Clears (turns off) the radiobutton.
.flash()
Flashes the radiobutton a few times between its active and normal colors, but leaves it the way it
started.
.invoke()
You can call this method to get the same actions that would occur if the user clicked on the radiobut-
ton to change its state.
.select()
Sets (turns on) the radiobutton.
15.The Scale widget

The purpose of a scale widget is to allow the user to set some integer or float value within a specified
range. Here are two scale widgets, one horizontal and one vertical:
Each scale displays a slider that the user can drag along a trough to change the value. In the figure, the
first slider is currently at -0.38 and the second at 7.
You can drag the slider to a new value with mouse button 1.
If you click button 1 in the trough, the slider will move one increment in that direction per click.
Holding down button 1 in the trough will, after a delay, start to auto-repeat its function.
If the scale has keyboard focus, left arrow and up arrow keystrokes will move the slider up (for
vertical scales) or left (for horizontal scales). Right arrow and down arrow keystrokes will move the
slider down or to the right.
To create a new scale widget as the child of a root window or frame named master:
w = Scale ( master, option, ... )

The constructor returns the new scale widget. Options:
activebackground The color of the slider when the mouse is over it. See Section4.3, Col-
ors (p. 6).
bg or background The background color of the parts of the widget that are outside the
trough.
bd or borderwidth Width of the 3-d border around the trough and slider. Default is 2 pixels.
For acceptable values, see Section4.1, Dimensions (p. 5).
command A procedure to be called every time the slider is moved. If the slider is
moved rapidly, you may not get a callback for every possible position,
but you'll certainly get a callback when it settles.
cursor The cursor that appears when the mouse is over the scale. See Section4.8,
Cursors (p. 9).
digits The way your program reads the current value shown in a scale widget
is through a control variable; see Section22, Control variables: the values
behind the widgets (p. 72). The control variable for a scale can be an
IntVar, a DoubleVar (float), or a StringVar. If it is a string variable,
the digits option controls how many digits to use when the numeric
scale value is converted to a string.
font The font used for the label and annotations. See Section4.4, Type
fonts (p. 6).
fg or foreground The color of the text used for the label and annotations.
from_ A float or integer value that defines one end of the scale's range. For ver-
tical scales, this is the top end; for horizontal scales, the left end. The un-
derbar (_) is not a typo: because from is a reserved word in Python, this
option is spelled from_. The default is 0. See the to option, below, for
the other end of the range.
highlightbackground The color of the focus highlight when the scale does not have focus. See
highlightcolor The color of the focus highlight when the scale has the focus.
label You can display a label within the scale widget by setting this option to
the label's text. The label appears in the top left corner if the scale is hori-
zontal, or the top right corner if vertical. The default is no label.
length The length of the scale widget. This is the x dimension if the scale is hori-
zontal, or the y dimension if vertical. The default is 100 pixels. For allow-
able values, see Section4.1, Dimensions (p. 5).
orient Set orient=HORIZONTAL if you want the scale to run along the x dimen-
sion, or orient=VERTICAL to run parallel to the y-axis. Default is hori-
zontal.
relief With the default relief=FLAT, the scale does not stand out from its
background. You may also use relief=SOLID to get a solid black frame
around the scale, or any of the other relief types described in Section4.6,
Relief styles (p. 8).

repeatdelay This option controls how long button 1 has to be held down in the trough
before the slider starts moving in that direction repeatedly. Default is
repeatdelay=300, and the units are milliseconds.
resolution Normally, the user will only be able to change the scale in whole units.
Set this option to some other value to change the smallest increment of
the scale's value. For example, if from_=-1.0 and to=1.0, and you set
resolution=0.5, the scale will have 5 possible values: -1.0, -0.5, 0.0,
+0.5, and +1.0. All smaller movements will be ignored.
showvalue Normally, the current value of the scale is displayed in text form by the
slider (above it for horizontal scales, to the left for vertical scales). Set this
option to 0 to suppress that label.
sliderlength Normally the slider is 30 pixels along the length of the scale. You can
change that length by setting the sliderlength option to your desired
length; see Section4.1, Dimensions (p. 5).
state Normally, scale widgets respond to mouse events, and when they have
the focus, also keyboard events. Set state=DISABLED to make the widget
unresponsive.
takefocus Normally, the focus will cycle through scale widgets. Set this option to 0
if you don't want this behavior. See Section23, Focus: routing keyboard
input (p. 74).
tickinterval Normally, no ticks are displayed along the scale. To display periodic
scale values, set this option to a number, and ticks will be displayed on
multiples of that value. For example, if from_=0.0, to=1.0, and tick-
interval=0.25, labels will be displayed along the scale at values 0.0,
0.25, 0.50, 0.75, and 1.00. These labels appear below the scale if horizontal,
to its left if vertical. Default is 0, which suppresses display of ticks.
to A float or integer value that defines one end of the scale's range; the other
end is defined by the from_ option, discussed above. The to value can
be either greater than or less than the from_ value. For vertical scales,
the to value defines the bottom of the scale; for horizontal scales, the
right end.
troughcolor The color of the trough.
variable The control variable for this scale, if any; see Section22, Control variables:
the values behind the widgets (p. 72). Control variables may be from
class IntVar, DoubleVar (float), or StringVar. In the latter case, the
numerical value will be converted to a string. See the the digits option,
above, for more information on this conversion.
width The width of the trough part of the widget. This is the x dimension for
vertical scales and the y dimension if the scale has orient=HORIZONTAL.
Default is 15 pixels.
Scale objects have these methods:
.get()
This method returns the current value of the scale.
.set ( value )
Sets the scale's value.

16.The Scrollbar widget
A number of widgets, such as listboxes and canvases, can act like sliding windows into a larger virtual
area. You can connect scrollbar widgets to them to give the user a way to slide the view around relative
to the contents. Here's a screen shot of an entry widget with an associated scrollbar widget:
The slider, or scroll thumb, is the raised-looking rectangle that shows the current scroll position.
The two triangular arrowheads at each end are used for moving the position by small steps.
The trough is the sunken-looking area visible behind the arrowheads and slider.
Scrollbars can be horizontal, like the one shown above, or vertical. A widget that has two scrollable
dimensions, such as a canvas or listbox, can have both a horizontal and a vertical scrollbar.
The slider's size and position, relative to the length of the entire widget, show the size and position
of the view relative to its total size. For example, if a vertical scrollbar is associated with a listbox,
and its slider extends from 50% to 75% of the height of the scrollbar, that means that the visible part
of the listbox shows that portion of the overall list starting at the halfway mark and ending at the
three-quarter mark.
In a horizontal scrollbar, clicking B1 (button 1) on the left arrowhead moves the view by a small
amount to the left. Clicking B1 on the right arrowhead moves the view by that amount to the right.
For a vertical scrollbar, clicking the upward- and downward-pointing arrowheads moves the view
small amounts up or down. Refer to the discussion of the associated widget to find out the exact
amount that these actions move the view.
The user can drag the slider with B1 or B2 (the middle button) to move the view.
For a horizontal scrollbar, clicking B1 in the trough to the left of the slider moves the view left by a
page, and clicking B1 in the trough to the right of the slider moves the view a page to the right. For
a vertical scrollbar, the corresponding actions move the view a page up or down.
Clicking B2 anywhere along the trough moves the slider so that its left or top end is at the mouse,
or as close to it as possible.
To create a new scrollbar widget as the child of a root window or frame master:
w = Scrollbar ( master, option, ... )
The constructor returns the new scrollbar widget. Options for scrollbars include:
activebackground The color of the slider and arrowheads when the mouse is over them. See
Section4.3, Colors (p. 6).
bg or background The color of the slider and arrowheads when the mouse is not over them.
bd or borderwidth The width of the 3-d borders around the entire perimeter of the trough,
and also the width of the 3-d effects on the arrowheads and slider. Default
is no border around the trough, and a 2-pixel border around the arrow-
heads and slider.
command A procedure to be called whenever the scrollbar is moved. For a discussion
of the calling sequence, see Section16.1, The scrollbar command call-
back (p. 50).
cursor The cursor that appears when the mouse is over the scrollbar. See Sec-

elementborderwidth The width of the borders around the arrowheads and slider. The default
is elementborderwidth=-1, which means to use the value of the
borderwidth option.
highlightbackground The color of the focus highlight when the scrollbar does not have focus.
highlightcolor The color of the focus highlight when the scrollbar has the focus.
highlightthickness The thickness of the focus highlight. Default is 1. Set to 0 to suppress
display of the focus highlight.
jump This option controls what happens when a user drags the slider. Normally
(jump=0), every small drag of the slider causes the command callback to
be called. If you set this option to 1, the callback isn't called until the user
releases the mouse button.
orient Set orient=HORIZONTAL for a horizontal scrollbar, orient=VERTICAL
for a vertical one.
repeatdelay This option controls how long button 1 has to be held down in the trough
before the slider starts moving in that direction repeatedly. Default is
repeatdelay=300, and the units are milliseconds.
repeatinterval This option controls how often slider movement will repeat when button
1 is held down in the trough. Default is repeatinterval=100, and the
units are milliseconds.
takefocus Normally, you can tab the focus through a scrollbar widget; see Section23,
Focus: routing keyboard input (p. 74). Set takefocus=0 if you don't
want this behavior. The default key bindings for scrollbars allow the user
to use the and arrow keys to move horizontal scrollbars, and they
can use the and keys to move vertical scrollbars.
troughcolor The color of the trough.
width Width of the scrollbar (its y dimension if horizontal, and its x dimension
if vertical). Default is 16. For possible values, see Section4.1, Dimen-
sions (p. 5).
Methods on scrollbar objects include:
.get()
Returns two numbers (a, b) describing the current position of the slider. The a value gives the pos-
ition of the left or top edge of the slider, for horizontal and vertical scrollbars respectively; the b
value gives the position of the right or bottom edge. Each value is in the interval [0.0, 1.0] where 0.0
is the leftmost or top position and 1.0 is the rightmost or bottom position. For example, if the slider
extends from halfway to three-quarters of the way along the trough, you might get back the tuple
(0.5,0.75).
.set ( first, last )
To connect a scrollbar to another widget w, set w's xscrollcommand or yscrollcommand to the
scrollbar's .set method. The arguments have the same meaning as the values returned by the
.get() method. Please note that moving the scrollbar's slider does not move the corresponding
widget.

16.1.The scrollbar command callback
When the user manipulates a scrollbar, the scrollbar calls its command callback. The arguments to this
call depend on what the user does:
When the user requests a movement of one unit left or up, for example by clicking button B1 on
the left or top arrowhead, the arguments to the callback look like:
command("scroll", -1, "units")
When the user requests a movement of one unit right or down, the arguments are:
command("scroll", 1, "units")
When the user requests a movement of one page left or up:
command("scroll", -1, "pages")
When the user requests a movement of one page right or down:
command("scroll", 1, "pages")
When the user drags the slider to a value f in the range [0,1], where 0 means all the way left or up
and 1 means all the way right or down, the call is:
command("moveto", f)
These calling sequences match the arguments expected by the .xview() and .yview() methods of
canvases, listboxes, and text widgets. The Entry widget does not have an .xview() method. See
Section8.1, Scrolling an Entry widget (p. 31).
16.2.Connecting scrollbars to other widgets

Here is a code fragment showing the creation of a canvas with horizontal and vertical scrollbars. In this
fragment, self is assumed to be a Frame widget.
self.canv = Canvas ( self, width=600, height=400,
scrollregion=(0, 0, 1200, 800) )
self.canv.grid ( row=0, column=0 )
self.scrollY = Scrollbar ( self, orient=VERTICAL,

command=self.canv.yview )
self.scrollY.grid ( row=0, column=1, sticky=N+S )
self.scrollX = Scrollbar ( self, orient=HORIZONTAL,

command=self.canv.xview )
self.scrollX.grid ( row=1, column=0, sticky=E+W )
self.canv["xscrollcommand"] = self.scrollX.set
self.canv["yscrollcommand"] = self.scrollY.set
Notes:
The connection goes both ways. The canvas's xscrollcommand option has to be connected to the
horizontal scrollbar's .set method, and the scrollbar's command option has to be connected to the
canvas's .xview method. The vertical scrollbar and canvas must have the same mutual connection.
The sticky options on the .grid() method calls for the scrollbars force them to stretch just enough
to fit the corresponding dimension of the canvas.

17.The Text widget
Text widgets are a much more generalized method for handling multiple lines of text than the Label
widget. Text widgets are pretty much a complete text editor in a window:
You can mix text with different fonts, colors, and backgrounds.
You can intersperse embedded images with text. An image is treated as a single character. See Sec-
tion17.3, Images in text widgets (p. 54).
You can even embed in it a window containing any Tkinter widgeteven a frame widget containing
other widgets. A window is also treated as a single character. See Section17.4, Windows in text
widgets (p. 54).
An index is a way of describing a specific position between two characters of a text widget. See Sec-
tion17.1, Indices in text widgets (p. 53).
Text widgets allow you to define names for regions of the text called tags. You can change the ap-
pearance of a tagged region, changing its font, foreground and background colors, and other attributes.
See Section17.5, Tags in text widgets (p. 54).
A text widget may contain invisible mark objects between character positions. See Section17.2, Marks
in text widgets (p. 53).
To create a text widget as the child of a root window or frame named master:
w = Text ( master, option, ... )
The constructor returns the new text widget. Options include:
bg or background The default background color of the text widget. See Section4.3, Col-
ors (p. 6).
bd or borderwidth The width of the border around the text widget. Default is 2 pixels.
cursor The cursor that will appear when the mouse is over the text widget. See
Section4.8, Cursors (p. 9).
exportselection Normally, text selected within a text widget is exported to be the selection
in the window manager. Set exportselection=0 if you don't want
that behavior.
font The default font for text inserted into the widget. Note that you can have
multiple fonts in the widgets by using tags to change the properties of
some text. See Section4.4, Type fonts (p. 6).
fg or foreground The color used for text (and bitmaps) within the widget. You can change
the color for tagged regions; this option is just the default.
height The height of the widget in lines (not pixels!), measured according to the
current font size.
highlightbackground The color of the focus highlight when the text widget does not have focus.
highlightcolor The color of the focus highlight when the text widget has the focus.
insertbackground The color of the insertion cursor. Default is black.
insertborderwidth Size of the 3-D border around the insertion cursor. Default is 0.

insertofftime The number of milliseconds the insertion cursor is off during its blink
cycle. Set this option to zero to suppress blinking. Default is 300.
insertontime The number of milliseconds the insertion cursor is on during its blink
cycle. Default is 600.
insertwidth Width of the insertion cursor (its height is determined by the tallest item
in its line). Default is 2 pixels.
padx The size of the internal padding added to the left and right of the text
area. Default is one pixel. For possible values, see Section4.1, Dimen-
sions (p. 5).
pady The size of the internal padding added above and below the text area.
Default is one pixel.
relief The 3-D appearance of the text widget. Default is relief=SUNKEN; for
other values, see Section4.6, Relief styles (p. 8).
selectbackground The background color to use displaying selected text.
selectborderwidth The width of the border to use around selected text.
selectforeground The foreground color to use displaying selected text.
spacing1 This option specifies how much extra vertical space is put above each line
of text. If a line wraps, this space is added only before the first line it oc-
cupies on the display. Default is 0.
spacing2 This option specifies how much extra vertical space to add between dis-
played lines of text when a logical line wraps. Default is 0.
spacing3 This option specifies how much extra vertical space is added below each
line of text. If a line wraps, this space is added only after the last line it
occupies on the display. Default is 0.
state Normally, text widgets respond to keyboard and mouse events; set
state=NORMAL to get this behavior. If you set state=DISABLED, the
text widget will not respond, and you won't be able to modify its contents
programmatically either.
tabs This option controls how tab characters position text. See Section17.6,
Setting tabs in a Text widget (p. 55).
takefocus Normally, focus will visit a text widget (see Section23, Focus: routing
keyboard input (p. 74)). Set takefocus=0 if you do not want focus in
the widget.
width The width of the widget in characters (not pixels!), measured according
to the current font size.
wrap This option controls the display of lines that are too wide.
With the default behavior, wrap=CHAR, any line that gets too long
will be broken at any character.
Set wrap=WORD and it will break the line after the last word that will
fit.
If you want to be able to create lines that are too long to fit in the
window, set wrap=NONE and provide a horizontal scrollbar.
xscrollcommand To make the text widget horizontally scrollable, set this option to the
.set method of the horizontal scrollbar.

yscrollcommand To make the text widget vertically scrollable, set this option to the .set
method of the vertical scrollbar.
17.1.Indices in text widgets

An index is a general method of specifying a position in the content of a text widget. An index is a string
with one of these forms:
"line.column"
The position just before the given column (counting from zero) on the given line (counting from
one). Examples: "1.0" is the position of the beginning of the text; "2.3" is the position before the
fourth character of the second line.
"line.end"
The position just before the newline at the end of the given line (counting from one). So, for example,
index "10.end" is the position at the end of the tenth line.
INSERT
The position of the insertion cursor in the text widget.
CURRENT
The position of the character closest to the mouse pointer.
END
The position after the last character of the text.
SEL_FIRST
If some of the text in the widget is currently selection (as by dragging the mouse over it), this is the
position before the start of the selection. If you try to use this index and nothing is selected, a
TclError exception will be raised.
SEL_LAST
The position after the end of the selection, if any. As with SEL_FIRST, you'll get a TclError ex-
ception if you use such an index and there is no selection.
"markname"
You can use a mark as an index; just pass its name where an index is expected. See Section17.2,
Marks in text widgets (p. 53).
"tag.first"
The position before the first character of the region tagged with name tag; see Section17.5, Tags
in text widgets (p. 54).
"tag.last"
The position after the last character of a tagged region.
"@x,y"
The position before the character closest to the coordinate (x, y).
embedded-object
If you have an image or window embedded in the text widget, you can use the PhotoImage,
BitmapImage, or embedded widget as an index. See Section17.3, Images in text widgets (p. 54)
and Section17.4, Windows in text widgets (p. 54).
17.2.Marks in text widgets

A mark represents a floating position somewhere in the contents of a text widget.

You handle each mark by giving it a name. This name can be any string that doesn't include
whitespace or periods.
There are two special marks. INSERT is the current position of the insertion cursor, and CURRENT
is the position closest to the mouse cursor.
Marks float along with the adjacent content. If you modify text somewhere away from a mark, the
mark stays at the same position relative to its immediate neighbors.
Marks have a property called gravity that controls what happens when you insert text at a mark.
The default gravity is RIGHT, which means that when new text is inserted at that mark, the mark
stays after the end of the new text. If you set the gravity of a mark to LEFT (using the text widget's
.mark_gravity() method), the mark will stay at a position just before text newly inserted at that
mark.
Deleting the text all around a mark does not remove the mark. If you want to remove a mark, use
the .mark_unset() method on the text widget.
Refer to Section17.7, Methods on Text widgets (p. 55), below, to see how to use marks.
17.3.Images in text widgets

You can put an image or bitmap into a text widget. It is treated as a single character whose size is the
natural size of the object. See Section4.9, Images (p. 11) andSection4.7, Bitmaps (p. 8).
Images are placed into the text widget by calling that widget's .image_create() method. See below
for the calling sequence and other methods for image manipulation.
Images are manipulated by passing their name to methods on the text widget. You can give Tkinter a
name for an image, or you can just let Tkinter generate a default name for that image.
17.4.Windows in text widgets

You can put any Tkinter widgeteven a frame containing other widgetsinto a text widget. For example,
you can put a fully functional button or a set of radiobuttons into a text widget.
Use the .window_create() method on the text widget to add the embedded widget. For the calling
sequence and related methods, see Section17.7, Methods on Text widgets (p. 55).
17.5.Tags in text widgets

There are lots of ways to change both the appearance and functionality of the items in a text widget.
For text, you can change the font, size, and color. Also, you can make text, widgets, or embedded images
respond to keyboard or mouse actions.
To control these appearance and functional features, you associate each feature with a tag. You can then
associate a tag with any number of pieces of text in the widget.
The name of a tag can be any string that does not contain white space or periods.
There is one special predefined tag called SEL. This is the region currently selected, if any.
Since any character may be part of more than one tag, there is a tag stack that orders all the tags.
Entries are added at the end of the tag list, and later entries have priority over earlier entries.
So, for example, if there is a character c that is part of two tagged regions t1 and t2, and t1 is
deeper in the tag stack than t2, and t1 wants the text to be green and t2 wants it to be blue, c will
be rendered in blue because t2 has precedence over t1.
You can change the ordering of tags in the tag stack.

Tags are created by using the .tag_add() method on the text widget. See Section17.7, Methods on
Text widgets (p. 55), below, for information on this and related methods.
17.6.Setting tabs in a Text widget

The tabs option for Text widgets gives you a number of ways to set tab stops within the widget.
The default is to place tabs every eight characters.
To set specific tab stops, set this option to a sequence of one or more distances. For example, setting
tabs=("3c", "5c", "12c") would put tab stops 3, 5, and 12cm from the left side. Past the last
tab you set, tabs have the same width as the distance between the last two existing tab stops. So,
continuing our example, because 12c-5c is 7 cm, if the user keeps pressing the Tab key, the cursor
would be positioned at 19cm, 26cm, 33cm, and so on.
Normally, text after a tab character is aligned with its left edge on the tab stop, but you can include
any of the keywords LEFT, RIGHT, CENTER, or NUMERIC in the list after a distance, and that will
change the positioning of the text after each tab.
A LEFT tab stop has the default behavior.
A RIGHT tab stop will position the text so its right edge is on the stop.
A CENTER tab will center the following text on the tab stop.
A NUMERIC tab stop will place following text to the left of the stop up until the first period (".")
in the textafter that, the period will be centered on the stop, and the rest of the text will posi-
tioned to its right.
For example, setting tabs=("0.5i", "0.8i", RIGHT, "1.2i", CENTER, "2i", NUMERIC)
would set four tab stops: a left-aligned tab stop half an inch from the left side, a right-aligned tab
stop 0.8" from the left side, a center-aligned tab stop 1.2" from the left, and a numeric-aligned tab
stop 2" from the left.
17.7.Methods on Text widgets

These methods are available on all text widgets:
.compare ( index1, op, index2 )

Compares the positions of two indices in the text widget, and returns true if the relational op holds
between index1 and index2. The op specifies what comparison to use, one of: "<", "<=", "==",
"!=", ">=", or ">".
For example, for a text widget t, t.compare("2.0", "<=", END) returns true if the beginning
of the second line is before or at the end of the text in t.
.delete ( index1, index2=None )
Deletes text starting just after index1. If the second argument is omitted, only one character is de-
leted. If a second index is given, deletion proceeds up to, but not including, the character after in-
dex2. Recall that indices sit between characters.
.get ( index1, index2=None )
Use this method to retrieve the current text from the widget. Retrieval starts at index index1. If
the second argument is omitted, you get the character after index1. If you provide a second index,
you get the text between those two indices. Embedded images and windows (widgets) are ignored.

.image_cget ( index, option )
To retrieve the current value of an option set on an embedded image, call this method with an index
pointing to the image and the name of the option.
.image_configure ( index, option, ... )
To set one or more options on an embedded image, call this method with an index pointing to the
image as the first argument, and one or more option=value pairs.
If you specify no options, you will get back a dictionary defining all the options on the image, and
the corresponding values.
.image_names()
This method returns a tuple of the names of all the text widget's embedded images.
.index ( i )
For an index i, this method returns the equivalent position in the form "line.char".
.insert ( index, text, tags=None )
Inserts the given text at the given index.
If you omit the tags argument, the newly inserted text will be tagged with any tags that apply to
the characters both before and after the insertion point.
If you want to apply one or more tags to the text you are inserting, provide as a third argument a
sequence of tag strings. Any tags that apply to existing characters around the insertion point are
ignored.
.mark_gravity ( mark, gravity=None )
Changes or queries the gravity of an existing mark; see Section17.2, Marks in text widgets (p. 53),
above, for an explanation of gravity.
To set the gravity, pass in the name of the mark, followed by either LEFT or RIGHT. To find the
gravity of an existing mark, omit the second argument and the method returns LEFT or RIGHT.
.mark_names()
Returns a sequence of the names of all the marks in the window, including INSERT and CURRENT.
.mark_set ( mark, index )
If no mark with name mark exists, one is created with RIGHT gravity and placed where index
points. If the mark already exists, it is moved to the new location.
.mark_unset ( mark )
Removes the named mark.
.search ( pattern, index, option, ... )
Searches for pattern (which can be either a string or a regular expression) in the buffer starting
at the given index. If it succeeds, it returns an index of the "line.char" form; if it fails, it returns
an empty string.
The allowable options for this method are:
backwards Set this option to 1 to search backwards from the index. Default is
forwards.
count If you set this option to an IntVar control variable, when there is a
match you can retrieve the length of the text that matched by using
the .get() method on that variable after the method returns.
regexp Set this option to 1 to interpret the pattern as a Tcl-style regular ex-
pression. The default is to look for an exact match to pattern. Tcl

regular expressions are a subset of Python regular expressions, sup-
porting these features: . ^ [c1] () * + ? e1|e2
nocase Set this option to 1 to ignore case. The default is a case-sensitive search.
stopindex To limit the search, set this option to the index beyond which the search
should not go.
.see ( index )
If the text containing the given index is not visible, scroll the text until that text is visible.
.tag_add ( tagName, index1, index2=None )
This method associates the tag named tagName with a region of the contents starting just after index
index1 and extending up to index index2. If you omit index2, only the character after index1
is tagged.
.tag_bind ( tagName, sequence, func, add=None )
This method binds an event to all the text tagged with tagName. See Section24, Events (p. 75),
below, for more information on event bindings.
To create a new binding for tagged text, use the first three arguments: sequence identifies the
event, and func is the function you want it to call when that event happens.
To add another binding to an existing tag, pass the same first three arguments and "+" as the fourth
argument.
To find out what bindings exist for a given sequence on a tag, pass only the first two arguments;
the method returns the associated function.
To find all the bindings for a given tag, pass only the first argument; the method returns a list of all
the tag's sequence arguments.
.tag_cget ( tagName, option )
Use this method to retrieve the value of the given option for the given tagName.
.tag_config ( tagName, option, ... )
To change the value of options for the tag named tagName, pass in one or more option=value
pairs.
If you pass only one argument, you will get back a dictionary defining all the options and their
values currently in force for the named tag.
Here are the options for tag configuration:
background The background color for text with this tag. Note that you can't use
bg as an abbreviation.
bgstipple To make the background appear grayish, set this option to one of the
standard bitmap names (see Section4.7, Bitmaps (p. 8)). This has
no effect unless you also specify a background.
borderwidth Width of the border around text with this tag. Default is 0. Note that
you can't use bd as an abbreviation.
fgstipple To make the text appear grayish, set this option a bitmap name.
font The font used to display text with this tag. See Section4.4, Type
fonts (p. 6).
foreground The color used for text with this tag. Note that you can't use the fg
abbreviation here.

justify The justify option set on the first character of each line determines
how that line is justified: LEFT (the default), CENTER, or RIGHT.
lmargin1 How much to indent the first line of a chunk of text that has this tag.
The default is 0. See Section4.1, Dimensions (p. 5)for allowable
values.
lmargin2 How much to indent successive lines of a chunk of text that has this
tag. The default is 0.
offset How much to raise (positive values) or lower (negative values) text
with this tag relative to the baseline. Use this to get superscripts or
subscripts, for example. For allowable values, see Section4.1, Dimen-
sions (p. 5).
overstrike Set overstrike=1 to draw a horizontal line through the center of
text with this tag.
relief Which 3-D effect to use for text with this tag. The default is re-
lief=FLAT; for other possible values see Section4.6, Relief
styles (p. 8).
rmargin Size of the right margin for chunks of text with this tag. Default is 0.
spacing1 This option specifies how much extra vertical space is put above each
line of text with this tag. If a line wraps, this space is added only before
the first line it occupies on the display. Default is 0.
spacing2 This option specifies how much extra vertical space to add between
displayed lines of text with this tag when a logical line wraps. Default
is 0.
spacing3 This option specifies how much extra vertical space is added below
each line of text with this tag. If a line wraps, this space is added only
after the last line it occupies on the display. Default is 0.
tabs How tabs are expanded on lines with this tag. See Section17.6, Setting
tabs in a Text widget (p. 55).
underline Set underline=1 to underline text with this tag.
wrap How long lines are wrapped in text with this tag. See the description
of the wrap option for text widgets, above.
.tag_delete ( tagName, ... )

To delete one or more tags, pass their names to this method. Their options and bindings go away,
and the tags are removed from all regions of text.
.tag_lower ( tagName, belowThis=None )
Use this method to change the order of tags in the tag stack (see Section17.5, Tags in text wid-
gets (p. 54), above, for an explanation of the tag stack). If you pass two arguments, the tag with
name tagName is moved to a position just below the tag with name belowThis. If you pass only
one argument, that tag is moved to the bottom of the tag stack.
.tag_names ( index=None )
If you pass an index argument, this method returns a sequence of all the tag names that are associated
with the character after that index. If you pass no argument, you get a sequence of all the tag names
defined in the text widget.

.tag_nextrange ( tagName, index1, index2=None )
This method searches a given region for places where a tag named tagName starts. The region
searched starts at index index1 and ends at index index2. If the index2 argument is omitted,
the search goes all the way to the end of the text.
If there is a place in the given region where that tag starts, the method returns a sequence [i0,
i1], where i0 is the index of the first tagged character and i1 is the index of the position just after
the last tagged character.
If no tag starts are found in the region, the method returns an empty string.
.tag_prevrange ( tagName, index1, index2=None )
This method searches a given region for places where a tag named tagName starts. The region
searched starts before index index1 and ends at index index2. If the index2 argument is omitted,
the search goes all the way to the end of the text.
The return values are as in .tag_nextrange().
.tag_raise ( tagName, aboveThis=None )
Use this method to change the order of tags in the tag stack (see Section17.5, Tags in text wid-
gets (p. 54), above, for an explanation of the tag stack). If you pass two arguments, the tag with
name tagName is moved to a position just above the tag with name aboveThis. If you pass only
one argument, that tag is moved to the top of the tag stack.
.tag_ranges ( tagName )
This method finds all the ranges of text in the widget that are tagged with name tagName, and returns
a sequence [s0, e0, s1, e1, ], where each si is the index just before the first character of the
range and ei is the index just after the last character of the range.
.tag_remove ( tagName, index1, index2=None )
Removes the tag named tagName from all characters between index1 and index2. If index2 is
omitted, the tag is removed from the single character after index1.
.tag_unbind ( tagName, sequence, funcid=None )
Remove the event binding for the given sequence from the tag named tagName. If there are
multiple handlers for this sequence and tag, you can remove only one handler by passing it as the
third argument.
.window_cget ( index, option )
Returns the value of the given option for the embedded widget at the given index.
.window_configure ( index, option )
To change the value of options for embedded widget at the given index, pass in one or more op-
tion=value pairs.
If you pass only one argument, you will get back a dictionary defining all the options and their
values currently in force for the given widget.
.window_create ( index, option, ... )
This method creates a window where a widget can be embedded within a text widget. There are
two ways to provide the embedded widget:
a. you can use pass the widget to the window option in this method, or
b. you can define a procedure that will create the widget and pass that procedure as a callback
to the create option.
Options for .window_create() are:

align Specifies how to position the embedded widget vertically in its line,
if it isn't as tall as the text on the line. Values include: align=CENTER
(the default), which centers the widget vertically within the line;
align=TOP, which places the top of the image at the top of the line;
align=BOTTOM, which places the bottom of the image at the bottom
of the line; and align=BASELINE, which aligns the bottom of the
image with the text baseline.
create A procedure that will create the embedded widget on demand. This
procedure takes no arguments and must create the widget as a child
of the text widget and return the widget as its result.
padx Extra space added to the left and right of the widget within the text
line. Default is 0.
pady Extra space added above and below the widget within the text line.
Default is 0.
stretch This option controls what happens when the line is higher than the
embedded widget. Normally this option is 0, meaning that the embed-
ded widget is left at its natural size. If you set stretch=1, the widget
is stretched vertically to fill the height of the line, and the align option
is ignored.
window The widget to be embedded. This widget must be a child of the text
widget.
.window_names()
Returns a sequence containing the names of all embedded widgets.
.xview ( MOVETO, fraction )
This method scrolls the text widget horizontally, and is intended for binding to the command option
of a related horizontal scrollbar.
This method can be called in two different ways. The first call positions the text at a value given by
fraction, where 0.0 moves the text to its leftmost position and 1.0 to its rightmost position.
.xview ( SCROLL, n, what )
The second call moves the text left or right: the what argument specifies how much to move and
can be either UNITS or PAGES, and n tells how many characters or pages to move the text to the
right relative to its image (or left, if negative).
This method scrolls the text in the same way as .xview(MOVETO, fraction).
.xview_scroll ( n, what )
Same as .xview(SCROLL, n, what).
.yview(MOVETO, fraction)
.yview(SCROLL, n, what)
.yview_moveto(fraction)
.yview_scroll(n, what)
These are the same as the .xview() and related methods, only for vertical scrolling. When scrolling
vertically by UNITS, the units are lines.

18.Toplevel: Top-level window methods
A top-level window is a window that has an independent existence under the window manager. It is
decorated with the window manager's decorations, and can be moved and resized independently. Your
application can use any number of top-level windows.
For any widget w, you can get to its top-level widget using w.winfo_toplevel().
To create a new top-level window:
w = Toplevel ( option, ... )
Options include:
bg or background The background color of the window. See Section4.3, Colors (p. 6).
bd or borderwidth Border width in pixels; default is 0. For possible values, see Section4.1,
Dimensions (p. 5). See also the relief option, below.
class_ You can give a Toplevel window a class name. Such names are
matched against the option database, so your application can pick up the
user's configuration preferences (such as colors) by class name. For ex-
ample, you might design a series of popups called screamers, and set
them all up with class_="Screamer". Then you can put a line in your
option database like this:
*Screamer*background:
red
and then, if you use the .option_readfile() method to read your
option database, all widgets with that class name will default to a red
background. This option is named class_ because class is a reserved
word in Python.
cursor The cursor that appears when the mouse is in this window. See Section4.8,
Cursors (p. 9).
height Window height; see Section4.1, Dimensions (p. 5).
relief Normally, a top-level window will have no 3-d borders around it. To get
a shaded border, set the bd option larger that its default value of zero,
and set the relief option to one of the constants discussed under Sec-
tion4.6, Relief styles (p. 8).
width The desired width of the window; see Section4.1, Dimensions (p. 5).
These methods are available for top-level windows:
.aspect ( nmin, dmin, nmax, dmax )

Constrain the root window's width:length ratio to the range [ nmin / dmin, nmax / dmax ].
.deiconify()
If this window is iconified, expand it.
.geometry ( newGeometry=None )
Set the window geometry. For the form of the argument, see Section4.10, Geometry strings (p. 11).
If the argument is omitted, the current geometry string is returned.
.iconify()
Iconify the window.

.lift ( aboveThis=None )
To raise this window to the top of the stacking order in the window manager, call this method with
no arguments. You can also raise it to a position in the stacking order just above another Toplevel
window by passing that window as an argument.
.lower ( belowThis=None )
If the argument is omitted, moves the window to the bottom of the stacking order in the window
manager. You can also move the window to a position just under some other top-level window by
passing that Toplevel widget as an argument.
.maxsize ( width=None, height=None )
Set the maximum window size. If the arguments are omitted, returns the current (width, height).
.minsize ( width=None, height=None )
Set the minimum window size. If the arguments are omitted, returns the current minima as a 2-
tuple.
.resizable ( width=None, height=None )
If width is true, allow horizontal resizing. If height is true, allow vertical resizing. If the arguments
are omitted, returns the current size as a 2-tuple.
.title ( text=None )
Set the window title. If the argument is omitted, returns the current title.
.withdraw()
Hides the window. Restore it with .deiconify() or .iconify().
19.Universal widget methods

The methods are defined below on all widgets. In the descriptions, w can be any widgeta frame, a
top-level window, whatever.
w.after ( ms, func=None, *args )

Requests Tkinter to call function func with arguments args after a delay of at least ms milliseconds.
There is no upper limit to how long it will actually take, but your callback won't be called sooner
than you request, and it will be called only once.
This method returns an integer after identifier that can be passed to the .after_cancel()
method if you want to cancel the callback.
w.after_cancel ( id )
Cancels a request for callback set up earlier .after(). The id argument is the result returned by
the original .after() call.
w.after_idle ( func, *args )
Requests that Tkinter call function func with arguments args next time the system is idle, that is,
next time there are no events to be processed. The callback will be called only once.
w.bell()
Makes a noise, usually a beep.
w.bind ( sequence=None, func=None, add=None )
This method is used to attach an event binding to a widget. See Section24, Events (p. 75) for the
overview of event bindings.
The sequence argument describes what event we expect, and the func argument is a function to
be called when that event happens to the widget. If there was already a binding for that event for

this widget, normally the old callback is replaced with func, but you can preserve both callbacks
by passing add="+".
w.bind_all ( sequence=None, func=None, add=None )
Like .bind(), but applies to all widgets in the entire application.
w.bind_class ( className, sequence=None, func=None, add=None )
Like .bind(), but applies to all widgets named className (e.g., "Button").
w.bindtags ( tagList=None )
If you call this method, it will return the binding tags for the widget as a sequence of strings. A
binding tag is the name of a window (starting with ".") or the name of a class (e.g., "Listbox").
You can change the order in which binding levels are called by passing as an argument the sequence
of binding tags you want the widget to use.
See Section24, Events (p. 75) for a discussion of binding levels and their relationship to tags.
w.cget ( option )
Returns the current value of option as a string. You can also get the value of an option for widget
w as w[option].
w.clipboard_append ( text )
Appends the given text string to the display's clipboard, where cut and pasted strings are stored
for all that display's applications.
w.clipboard_clear()
Clears the display's clipboard (see .clipboard_append() above).
w.config( option=value, ... )
w.configure ( option=value, ... )
Set the values of one or more options. For the options whose names are Python reserved words
(class, from, in), use a trailing underbar: "class_", "from_", "in".
You can also set the value of an option for widget w with the statement
w[option] = value
If you call the .config() method on a widget with no arguments, you'll get a dictionary of all the
widget's current options. The keys are the option names (including aliases like bd for borderwidth).
The value for each key is:
for most entries, a five-tuple: (option name, option database key, option database class, default
value, current value); or,
for alias names (like "fg"), a two-tuple: (alias name, equivalent standard name).
w.destroy()
Calling w.destroy() on a widget w destroys w and all its children.
w.event_add ( virtual, *sequences )
This method creates a virtual event whose name is given by the virtual string argument. Each
additional argument describes one sequence, that is, the description of a physical event. When that
event occurs, the new virtual event is triggered.
See Section24, Events (p. 75) for a general description of virtual events.
w.event_delete ( virtual, *sequences )
Deletes physical events from the virtual event whose name is given by the string virtual. If all
the physical events are removed from a given virtual event, that virtual event won't happen anymore.

w.event_generate ( sequence, **kw )
This method causes an event to trigger without any external stimulus. The handling of the event is
the same as if it had been triggered by an external stimulus. The sequence argument describes the
event to be triggered. You can set values for selected fields in the Event object by providing
keyword=value arguments, where the keyword specifies the name of a field in the Event object.
See Section24, Events (p. 75) for a full discussion of events.
w.event_info ( virtual=None )
If you call this method without an argument, you'll get back a sequence of all the currently defined
virtual event names.
To retrieve the physical events associated with a virtual event, pass this method the name of the
virtual event and you will get back a sequence of the physical sequence names, or None if the
given virtual event has never been defined.
w.focus_displayof()
Returns the name of the window that currently has input focus on the same display as the widget.
If no such window has input focus, returns None.
See Section23, Focus: routing keyboard input (p. 74) for a general description of input focus.
w.focus_force()
Force the input focus to the widget. This is impolite. It's better to wait for the window manager to
give you the focus. See also .grab_set_global() below.
w.focus_get()
Get the name of the widget that has focus in this application, if anyotherwise return None.
w.focus_lastfor()
This method retrieves the name of the widget that last had the input focus in the top-level window
that contains w. If none of this top-level's widgets have ever had input focus, it returns the name of
the top-level widget. If this application doesn't have the input focus, .focus_lastfor() will return
the name of the widget that will get the focus next time it comes back to this application.
w.focus_set()
If w's application has the input focus, the focus will jump to w. If w's application doesn't have focus,
Tk will remember to give it to w next the application gets focus.
w.grab_current()
If there is a grab in force for w's display, return its identifier, otherwise return None. Refer to Sec-
tion24, Events (p. 75) for a discussion of grabs.
w.grab_release()
If w has a grab in force, release it.
.grab_set()
Widget w grabs all events for w's application. If there was another grab in force, it goes away. See
Section24, Events (p. 75) for a discussion of grabs.
w.grab_set_global()
Widget w grabs all events for the entire screen. This is considered impolite and should be used only
in great need. Any other grab in force goes away. Try to use this awesome power only for the forces
of good, and never for the forces of evil, okay?
w.grab_status()
If there is a local grab in force (set by .grab_set()), this method returns the string "local". If
there is a global grab in force (from .grab_set_global()), it returns "global". If no grab is in
force, it returns None.

w.image_names()
Returns the names of all the images in w's application as a sequence of strings.
w.keys()
Returns the option names for the widget as a sequence of strings.
w.mainloop()
This method must be called, generally after all the static widgets are created, to start processing
events. You can leave the main loop with the .quit() method (below). You can also call this
method inside an event handler to resume the main loop.
w.nametowidget ( name )
This method returns the actual widget whose path name is name. See Section4.11, Window
names (p. 12).
w.option_add ( pattern, value, priority=None )
This method adds default option values to the Tkinter option database. The pattern is a string
that specifies a default value for options of one or more widgets. The priority values are one
of:
widgetDefault Level 20, for global default properties of widgets.

startupFile Level 40, for default properties of specific applications.
userDefault Level 60, for options that come from user files such as their .Xde-
faults file.
interactive Level 80, for options that are set after the application starts up. This is
the default priority level.
Higher-level priorities take precedence over lower-level ones. See Section20, Standardizing ap-
pearance (p. 69) for an overview of the option database. The syntax of the pattern argument to
.option_add() is the same as the option-pattern part of the resource specification line.
For example, to get the effect of this resource specification line:
*Button*font: times 24 bold
your application (self in this example) might include these lines:
self.bigFont = tkFont.Font ( family="times", size=24,
weight="bold" )
self.option_add ( "*Button*font", self.bigFont )
Any Button widgets created after executing these lines would default to bold Times 24 font (unless
overriden by a font option to the Button constructor).
w.option_clear()
This method removes all options from the Tkinter option database. This has the effect of going back
to all the default values.
w.option_get ( name, classname )
Use this method to retrieve the current value of an option from the Tkinter option database. The
first argument is the instance key and the second argument is the class key. If there are any matches,
it returns the value of the option that best matches. If there are no matches, it returns "".
Refer to Section20, Standardizing appearance (p. 69) for more about how keys are matched with
options.

w.option_readfile ( fileName, priority=None )
As a convenience for user configuration, you can designate a named file where users can put their
preferred options, using the same format as the .Xdefaults file. Then, when your application is
initializing, you can pass that file's name to this method, and the options from that file will be added
to the database. If the file doesn't exist, or its format is invalid, this method will raise TclError.
Refer to Section20, Standardizing appearance (p. 69) for an introduction to the options database
and the format of option files.
w.quit()
This method exits the main loop. See .mainloop(), above, for a discussion of main loops.
w.selection_clear()
If w currently has a selection (such as a highlighted segment of text in an entry widget), clear that
selection.
w.selection_get()
If w currently has a selection, this method returns the selected text. If there is no selection, it raises
TclError.
w.selection_own()
Make w the owner of the selection in w's display, stealing it from the previous owner, if any.
w.selection_own_get()
Returns the widget that currently owns the selection in w's display. Raises TclError if there is no
such selection.
w.tk_focusFollowsMouse()
Normally, the input focus cycles through a sequence of widgets determined by their hierarchy and
creation order; see Section23, Focus: routing keyboard input (p. 74). You can, instead, tell Tkinter
to force the focus to be wherever the mouse is; just call this method. There is no easy way to undo
it, however.
w.unbind ( sequence, funcid=None )
This method deletes bindings on w for the event described by sequence. If the second argument
is a callback bound to that sequence, that callback is removed and the rest, if any, are left in place.
If the second argument is omitted, all bindings are deleted.
See Section24, Events (p. 75), below, for a general discussion of event bindings.
w.unbind_all ( sequence )
Deletes all event bindings throughout the application for the event described by the given sequence.
w.unbind_class ( className, sequence )
Like .unbind(), but applies to all widgets named className (e.g., "Entry" or "Listbox").
w.update()
This method forces the updating of the display. It should be used only if you know what you're
doing, since it can lead to unpredictable behavior or looping. It should never be called from an event
callback or a function that is called from an event callback.
w.update_idletasks()
Some tasks in updating the display, such as resizing and redrawing widgets, are called idle tasks
because they are usually deferred until the application has finished handling events and has gone
back to the main loop to wait for new events.
If you want to force the display to be updated before the application next idles, call the w.update_id-
letasks() method on any widget.

w.winfo_children()
Returns a list of all w's children, in their stacking order from lowest (bottom) to highest (top).
w.winfo_class()
Returns w's class name (e.g., "Button").
w.winfo_colormapfull()
Returns true if w's window's color map is full, that is, if the last attempt to add a color to it failed
and no colors have been removed since then.
w.winfo_containing ( rootX, rootY, displayof=0 )
This method is used to find the window that contains point (rootX, rootY). If the displayof
option is false, the coordinates are relative to the application's root window; if true, the coordinates
are treated as relative to the top-level window that contains w. If the specified point is in one of the
application's top-level window, this method returns that window; otherwise it returns None.
w.winfo_depth()
Returns the number of bits per pixel in w's display.
w.winfo_fpixels ( number )
For any dimension number (see Section4.1, Dimensions (p. 5)), this method returns that distance
in pixels on w's display, as a floating-point number.
w.winfo_geometry()
Returns the geometry string describing the size and on-screen location of w. See Section4.10,
Geometry strings (p. 11).
Warning
The geometry is not accurate until the application has updated its idle tasks. In particular, all
geometries are initially "1x1+0+0" until the widgets and geometry manager have negotiated
their sizes and positions. See the .update_idletasks() method, above, in this section to
see how to insure that the widget's geometry is up to date.
w.winfo_height()
Returns the current height of w in pixels. See the remarks on geometry updating under
.winfo_geometry(), above.
w.winfo_id()
Returns an integer that uniquely identifies w within its top-level window. You will need this for the
.winfo_pathname() method, below.
w.winfo_ismapped()
This method returns true if w is mapped, false otherwise. A widget is mapped if it has been gridded
(or placed or packed, if you are using one of the other geometry managers) into its parent, and if
its parent is mapped, and so on up to the top-level window.
w.winfo_manager()
If w has not been gridded (or placed via one of the other geometry managers), this method returns
an empty string. If w has been gridded or otherwise placed, it returns a string naming the geometry
manager, such as "grid".
w.winfo_name()
This method returns w's name relative to its parent. See Section4.11, Window names (p. 12). Also
see .winfo_pathname(), below, to find out how to obtain a widget's path name.

w.winfo_parent()
Returns w's parent's path name, or an empty string if w is a top-level window. See Section4.11,
Window names (p. 12) above, for more on widget path names.
w.winfo_pathname ( id, displayof=0 )
If the displayof argument is false, returns the window path name of the widget with unique
identifier id in the application's main window. If displayof is true, the id number specifies a
widget in the same top-level window as w. See Section4.11, Window names (p. 12) for a discussion
of widget path names.
w.winfo_pixels ( number )
For any dimension number (see Dimensions, above), this method returns that distance in pixels on
w's display, as an integer.
w.winfo_pointerx()
w.winfo_pointery()
w.winfo_pointerxy()
The first two methods return the x and y coordinates, respectively, of the mouse pointer relative
to w's root window. If the mouse pointer isn't on the same screen, each will return -1. The third
method returns a tuple (x, y), or (-1, -1) if the mouse is off the screen.
w.winfo_reqheight()
w.winfo_reqwidth()
These methods return the requested height and width, respectively, of widget w. These dimensions
are the ones necessary so that all of w's contents have the room they need. The actual height and
width may be different due to negotiations with the geometry manager.
w.winfo_rgb ( color )
For any given color, this method returns the equivalent red-green-blue color specification as a 3-
tuple (r, g, b), where each number is an integer in the range [0, 65536). For example, if the
color is "green", this method returns the 3-tuple (0, 65535, 0).
For more on specifying colors, see Section4.3, Colors (p. 6).
w.winfo_rootx()
w.winfo_rooty()
These methods return the x and y coordinates, respectively, of the upper left-hand corner of w's
root window relative to w's parent.
If w has a border, this is the outer corner of the border.
w.winfo_screenheight()
w.winfo_screenwidth()
w.winfo_screenmmheight()
w.winfo_screenmmwidth()
These methods retrieve the size of the screen. The first two return the height and width, respectively,
in pixels. The latter two return the screen size in millimeters, as an integer.
w.winfo_screenvisual()
Returns a string that describes the display's method of color rendition. This is usually "truecolor"
for 16- or 24-bit displays, "pseudocolor" for 256-color displays.

w.winfo_toplevel()
Returns the top-level window containing w. That window supports all the methods on Toplevel
widgets; see Section18, Toplevel: Top-level window methods (p. 61).
w.winfo_x()
w.winfo_y()
These methods return the x and y coordinates, respectively, of the top left corner of w relative to its
parent. If w has a border, this is the outer corner of the border.
20.Standardizing appearance and the option database

It's easy to apply colors, fonts, and other options to the widgets when you create them. However,
if you want a lot of widgets to have the same background color or font, it's tedious to specify each
option each time, and
it's nice to let the user override your choices with their favorite color schemes, fonts, and other
choices.
Accordingly, we use the idea of an option database to set up default option values.
Your application can specify a file (such as the standard .Xdefaults file used by the X Window
System) that contains the user's preferences. You can set up your application to read the file and tell
Tkinter to use those defaults. See the section on the .option_readfile() method, above, in the
section on Section19, Universal widget methods (p. 62), for the structure of this file.
Your application can directly specify defaults for one or many types of widgets by using the .op-
tion_add() method; see this method under Section19, Universal widget methods (p. 62).
Before we discuss how options are set, consider the problem of customizing the appearance of GUIs in
general. We could give every widget in the application a name, and then ask the user to specify every
property of every name. But this is cumbersome, and would also make the application hard to reconfig-
ureif the designer adds new widgets, the user would have to describe every property of every new
widget.
So, the option database allows the programmer and the user to specify general patterns describing which
widgets to configure.
These patterns operate on the names of the widgets, but widgets are named using two parallel naming
schemes:
a. Every widget has a class name. By default, the class name is the same as the class constructor:
"Button" for buttons, "Frame" for a frame, and so on. However, you can create new classes of
widgets, usually inheriting from the Frame class, and give them new names of your own creation.
See Section20.1, How to name a widget class (p. 70) for details.
b. You can also give any widget an instance name. The default name of a widget is usually a meaningless
number (see Section4.11, Window names (p. 12)). However, as with widget classes, you can assign
a name to any widget. See the section Section20.2, How to name a widget instance (p. 70) for
details.
Every widget in every application therefore has two hierarchies of namesthe class name hierarchy
and the instance name hierarchy. For example, a button embedded in a text widget which is itself em-
bedded in a frame would have the class hierarchy Frame.Text.Button. It might also have an instance
hierarchy something like .mainFrame.messageText.panicButton if you so named all the instances.
The initial dot stands for the root window; see Section4.11, Window names (p. 12) for more information
on window path names.

The option database mechanism can make use of either class names or instance names in defining options,
so you can make options apply to whole classes (e.g., all buttons have a blue background) or to specific
instances (e.g., the Panic Button has red letters on it). After we look at how to name classes and instances,
in Section20.3, Resource specification lines (p. 70), we'll discuss how the options database really
works.
20.1.How to name a widget class

For example, suppose that Jukebox is a new widget class that you have created. It's probably best to
have new widget classes inherit from the Frame class, so to Tkinter it acts like a frame, and you can
arrange other widgets such as labels, entries, and buttons inside it.
You set the new widget's class name by passing the name as the class_ attribute to the parent con-
structor in your new class's constructor. Here is a fragment of the code that defines the new class:
class Jukebox(Frame):
def __init__(self, master):
"Constructor for the Jukebox class"
Frame.__init__ ( self, master, class_="Jukebox" )
self.__createWidgets()
...
20.2.How to name a widget instance

To give an instance name to a specific widget in your application, set that widget's name option to a
string containing the name.
Here's an example of an instance name. Suppose you are creating several buttons in an application, and
you want one of the buttons to have an instance name of panicButton. Your call to the constructor
might look like this:
self.panic = Button ( self, name="panicButton", text="Panic", ...)
20.3.Resource specification lines

Each line in an option file specifies the value of one or more options in one or more applications and
has one of these formats:
app option-pattern: value
option-pattern: value
The first form sets options only when the name of the application matches app; the second form sets
options for all applications.
For example, if your application is called xparrot, a line of the form
xparrot*background: LimeGreen
sets all background options in the xparrot application to lime green. (Use the -name option on the
command line when launching your application to set the name to "xparrot".)
The option-pattern part has this syntax:
{{*|.}name}...option
That is, each option-pattern is a list of zero or more names, each of which is preceded by an asterisk
or period. The last name in the series is the name of the option you are setting. Each of the rest of the
names can be either:
the name of a widget class (capitalized), or
the name of an instance (lowercased).

The way the option patterns work is a little complicated. Let's start with a simple example:
*font: times 24
This line says that all font options should default to 24-point Times. The * is called the loose binding
symbol, and means that this option pattern applies to any font option anywhere in any application.
Compare this example:
*Listbox.font: lucidatypewriter 14
The period between Listbox and font is called the tight binding symbol, and it means that this rule
applies only to font options for widgets in class Listbox.
As another example, suppose your xparrot application has instances of widgets of class Jukebox. In
order to set up a default background color for all widgets of that class Jukebox, you could put a line
in your options file like this:
xparrot*Jukebox*background: PapayaWhip
The loose-binding (*) symbol between Jukebox and background makes this rule apply to any
background attribute of any widget anywhere inside a Jukebox. Compare this option line:
xparrot*Jukebox.background: NavajoWhite
This rule will apply to the frame constituting the Jukebox widget itself, but because of the tight-binding
symbol it will not apply to widgets that are inside the Jukebox widget.
In the next section we'll talk about how Tkinter figures out exactly which option value to use if there
are multiple resource specification lines that apply.
20.4.Rules for resource matching

When you are creating a widget, and you don't specify a value for some option, and two or more resource
specifications apply to that option, the most specific one applies.
For example, suppose your options file has these two lines:
*background: LimeGreen
*Listbox*background: FloralWhite
Both specifications apply to the background option in a Listbox widget, but the second one is more
specific, so it will win.
In general, the names in a resource specification are a sequence n1, n2, n3, ..., o where each ni is a class
or instance name. The class names are ordered from the highest to the lowest level, and o is the name
of an option.
However, when Tkinter is creating a widget, all it has is the class name and the instance name of that
widget.
Here are the precedence rules for resource specifications:
1. The name of the option must match the o part of the option-pattern. For example, if the rule
is
xparrot*indicatoron: 0
this will match only options named indicatoron.
2. The tight-binding operator (.) is more specific than the loose-binding operator (*). For example, a
line for *Button.font is more specific than a line for *Button*font.
3. References to instances are more specific than references to classes. For example, if you have a
button whose instance name is panicButton, a rule for *panicButton*font is more specific
than a rule for *Button*font.

4. A rule with more levels is more specific. For example, a rule for *Button*font is more specific
than a rule for *font.
5. If two rules have same number of levels, names earlier in the list are more specific than later names.
For example, a rule for xparrot*font is more specific than a rule for *Button*font.
21.Connecting your application logic to the widgets

The preceding sections talked about how to arrange and configure the widgetsthe front panel of the
application.
Next, we'll talk about how to connect up the widgets to the logic that carries out the actions that the
user requests.
To make your application respond to events such as mouse clicks or keyboard inputs, there are two
methods:
Some controls such as buttons have a command attribute that lets you specify a procedure, called
a handler, that will be called whenever the user clicks that control.
The sequence of events for using a Button widget is very specific, though. The user must move
the mouse pointer onto the widget with mouse button 1 up, then press mouse button 1, and then
release mouse button 1 while still on the widget. No other sequence of events will press a
Button widget.
There is a much more general mechanism that can let your application react to many more kinds
of inputs: the press or release of any keyboard key or mouse button; movement of the mouse
into, around, or out of a widget; and many other events. As with command handlers, in this
mechanism you write handler procedures that will be called whenever certain types of events
occur. This mechanism is discussed under Section24, Events (p. 75).
Many widgets require you to use control variables, special objects that connect widgets together and
to your program, so that you can read and set properties of the widgets. Control variables will be
discussed in the next section.
22.Control variables: the values behind the widgets

A Tkinter control variable is a special object that acts like a regular Python variable in that it is a container
for a value, such as a number or string.
One special quality of a control variable is that it can be shared by a number of different widgets, and
the control variable can remember all the widgets that are currently sharing it. This means, in particular,
that if your program stores a value v into a control variable c with its c.set(v) method, any widgets that
are linked to that control variable are automatically updated on the screen.
Tkinter uses control variables for a number of important functions, for example:
Checkbuttons use a control variable to hold the current state of the checkbutton (on or off).
A single control variable is shared by a group of radiobuttons and can be used to tell which one of
them is currently set. When the user clicks on one radiobutton in a group, the sharing of this control
variable is the mechanism by which Tkinter groups radiobuttons so that when you set one, any
other set radiobutton in the group is cleared.
Control variables hold text string for several applications. Normally the text displayed in an Entry
widget is linked to a control variable. In several other controls, it is possible to use a string-valued
control variable to hold text such as the labels of checkbuttons and radiobuttons and the content of
Label widgets.

For example, you could link an Entry widget to a Label widget so that when the user changes the
text in the entry and presses the Enter key, the label is automatically updated to show that same
text.
To get a control variable, use one of these four class constructors, depending on what type of values
you need to store in it:
v = DoubleVar() # Holds a float; default value 0.0
v = IntVar() # Holds an integer; default value 0
v = StringVar() # Holds a string; default value ""
All control variables have these two methods:
.get()
Returns the current value of the variable.
.set ( value )
Changes the current value of the variable. If any widget options are slaved to this variable, those
widgets will be updated when the main loop next idles; see .update_idletasks() in Section19,
Universal widget methods (p. 62) for more information on controlling this update cycle.
Here are some comments on how control variables are used with specific widgets:
Button
You can set its textvariable to a StringVar. Anytime that variable is changed, the text on the
button will be updated to display the new value. This is not necessary unless the button's text is
actually going to change: use the text attribute if the button's label is static.
Checkbutton
Normally, you will set the widget's variable option to an IntVar, and that variable will be set
to 1 when the checkbutton is turned on and to 0 when it is turned off. However, you can pick different
values for those two states with the onvalue and offvalue options, respectively.
You can even use a StringVar as the checkbutton's variable, and supply string values for the
offvalue and onvalue. Here's an example:
self.spamVar = StringVar()
self.spamCB = Checkbutton ( self, text="Spam?",
variable=self.spamVar, onvalue="yes", offvalue="no" )
If this checkbutton is on, self.spamVar.get() will return the string "yes"; if the checkbutton
is off, that same call will return the string "no". Furthermore, your program can turn the checkbutton
on by calling .set("yes").
You can also the textvariable option of a checkbutton to a StringVar. Then you can change
the text label on that checkbutton using the .set() method on that variable.
Entry
Set its textvariable option to a StringVar. Use that variable's .get() method to retrieve the
text currently displayed in the widget. You can also the variable's .set() method to change the
text displayed in the widget.
Label
You can set its textvariable option to a StringVar. Then any call to the variable's .set()
method will change the text displayed on the label. This is not necessary if the label's text is static;
use the text attribute for labels that don't change while the application is running.
Menubutton
If you want to be able to change the text displayed on the menu button, set its textvariable option
to a StringVar and use that variable's .set() method to change the displayed text.

Radiobutton
The variable option must be set to a control variable, either an IntVar or a StringVar. All the
radiobuttons in a functional group must share the same control variable.
Set the value option of each radiobutton in the group to a different value. Whenever the user sets
a radiobutton, the variable will be set to the value option of that radiobutton, and all the other ra-
diobuttons that share the group will be cleared.
You might wonder, what state is a group of radiobuttons in when the control variable has never
been set and the user has never clicked on them? Each control variable has a default value: 0 for an
IntVar, 0.0 for a DoubleVar, and "" for a StringVar. If one of the radiobuttons has that value,
that radiobutton will be set initially. If no radiobutton's value option matches the value of the
variable, the radiobuttons will all appear to be cleared.
If you want to change the text label on a radiobutton during the execution of your application, set
its textvariable option to a StringVar. Then your program can change the text label by passing
the new label text to the variable's .set() method.
Scale
For a scale widget, set its variable option to a control variable of any class, and set its from_ and
to options to the limiting values for the opposite ends of the scale.
For example, you could use an IntVar and set the scale's from_=0 and to=100. Then every user
change to the widget would change the variable's value to some value between 0 and 100 inclusive.
Your program can also move the slider by using the .set() method on the control variable. To
continue the above example, .set(75) would move the slider to a position three-fourths of the
way along its trough.
To set up a Scale widget for floating values, use a DoubleVar.
You can use a StringVar as the control variable of a Scale widget. You will still need to provide
numeric from_ and to values, but the numeric value of the widget will be converted to a string
for storage in the StringVar. Use the scale's digits option to control the precision of this conver-
sion.
23.Focus: routing keyboard input

To say a widget has focus means that keyboard input is currently directed to that widget.
By focus traversal, we mean the sequence of widgets that will be visited as the user moves from
widget to widget with the tab key. See below for the rules for this sequence.
You can traverse backwards using shift-tab.
The Entry and Text widgets are intended to accept keyboard input, and if an entry or text widget
currently has the focus, any characters you type into it will be added to its text. The usual editing
characters such as and will have their usual effects.
Because Text widgets can contain tab characters, you must use the special key sequence control-tab
to move the focus past a text widget.
Most of the other types of widgets will normally be visited by focus traversal, and when they have
focus:
Button widgets can be pressed by pressing the spacebar.
Checkbutton widgets can be toggled between set and cleared states using the spacebar.

In Listbox widgets, the and keys scroll up or down one line; the PageUp and PageDown
keys scroll by pages; and the spacebar selects the current line, or de-selects it if it was already
selected.
You can set a Radiobutton widget by pressing the spacebar.
Horizontal Scale widgets respond to the and keys, and vertical ones respond to and .
In a Scrollbar widget, the PageUp and PageDown keys move the scrollbar by pageloads. The
and keys will move vertical scrollbars by units, and the and keys will move horizontal
scrollbars by units.
Many widgets are provided with an outline called the focus highlight that shows the user which
widget has the highlight. This is normally a thin black frame located just outside the widget's border
(if any). For widgets that don't normally have a focus highlight (specifically, frames, labels, and
menus), you can set the highlightthickness option to a nonzero value to make the focus highlight
visible.
You can also change the color of the focus highlight using the highlightcolor option.
Widgets of class Frame, Label, and Menu are not normally visited by the focus. However, you can
set their takefocus options to 1 to get them included in focus traversal. You can also take any
widget out of focus traversal by setting its takefocus option to 0.
The order in which the tab key traverses the widgets is:
For widgets that are children of the same parent, focus goes in the same order the widgets were
created.
For parent widgets that contain other widgets (such as frames), focus visits the parent widget first
(unless its takefocus option is 0), then it visits the child widgets, recursively, in the order they
were created.
To sum up: to set up the focus traversal order of your widgets, create them in that order. Remove widgets
from the traversal order by setting their takefocus options to 0, and for those whose default takefocus
option is 0, set it to 1 if you want to add them to the order.
The above describes the default functioning of input focus in Tkinter. There is another, completely dif-
ferent way to handle itlet the focus go wherever the mouse goes. Under Section19, Universal widget
methods (p. 62), refer to the .tk_focusFollowsMouse() method.
You can also add, change or delete the way any key on the keyboard functions inside any widget by
using event bindings. See Section24, Events (p. 75) for the details.
24.Events: responding to stimuli

An event is something that happens to your applicationfor example, the user presses a key or clicks
or drags the mouseto which the application needs to react.
The widgets normally have a lot of built-in behaviors. For example, a button will react to a mouse click
by calling its command callback. For another example, if you move the focus to an entry widget and
press a letter, that letter gets added to the content of the widget.
However, the event binding capability of Tkinter allows you to add, change, or delete behaviors.
First, some definitions:
An event is some occurrence that your application needs to know about.
An event handler is a function in your application that gets called when an event occurs.

We call it binding when your application sets up an event handler that gets called when an event
happens to a widget.
24.1.Levels of binding
You can bind a handler to an event at any of three levels:
1. Instance binding: You can bind an event to one specific widget. For example, you might bind the
PageUp key in a canvas widget to a handler that makes the canvas scroll up one page. To bind an
event of a widget, call the .bind() method on that widget (see Section19, Universal widget
methods (p. 62)).
For example, suppose you have a canvas widget named self.canv and you want to draw an orange
blob on the canvas whenever the user clicks the mouse button 2 (the middle button). To implement
this behavior:
self.canv.bind ( "<Button-2>", self.__drawOrangeBlob )
The first argument is a sequence descriptor that tells Tkinter that whenever the middle mouse button
goes down, it is to call the event handler named self.__drawOrangeBlob. (See Section24.6,
Writing your handler (p. 81), below, for an overview of how to write handlers such as .__dra-
wOrangeBlob()). Note that you omit the parentheses after the handler name, so that Python will
pass in a reference the handler instead of trying to call it right away.
2. Class binding: You can bind an event to all widgets of a class. For example, you might set up all
Button widgets to respond to middle mouse button clicks by changing back and forth between
English and Japanese labels. To bind an event to all widgets of a class, call the .bind_class()
method on any widget (see Section19, Universal widget methods (p. 62), above).
For example, suppose you have several canvases, and you want to set up mouse button 2 to draw
an orange blob in any of them. Rather than having to call .bind() for every one of them, you can
set them all up with one call something like this:
self.bind_class ( "Canvas", "<Button-2>",
self.__drawOrangeBlob )
3. Application binding: You can set up a binding so that a certain event calls a handler no matter what
widget has the focus or is under the mouse. For example, you might bind the PrintScrn key to all
the widgets of an application, so that it prints the screen no matter what widget gets that key. To
bind an event at the application level, call the .bind_all() method on any widget (see Section19,
Universal widget methods (p. 62)).
Here's how you might bind the PrintScrn key, whose key name is "Print":
self.bind_all ( "<Key-Print>", self.__printScreen )
24.2.Event sequences
Tkinter has a powerful and general method for allowing you to define exactly which events, both spe-
cific and general, you want to bind to handlers.
In general, an event sequence is a string containing one or more event patterns. Each event pattern describes
one thing that can happen. If there is more than one event pattern in a sequence, the handler will be
called only when all the patterns happen in that same sequence.
The general form of an event pattern is:
<[modifier-]...type[-detail]>
The entire pattern is enclosed inside <>.

The event type describes the general kind of event, such as a key press or mouse click. See Section24.3,
Event types (p. 77).
You can add optional modifier items before the type to specify combinations such as the shift or
control keys being depressed during other key presses or mouse clicks. Section24.4, Event modifi-
ers (p. 78)
You can add optional detail items to describe what key or mouse button you're looking for. For
mouse buttons, this is 1 for button 1, 2 for button 2, or 3 for button 3.
The usual setup has button 1 on the left and button 3 on the right, but left-handers can swap
these positions.
For keys on the keyboard, this is either the key's character (for single-character keys like the A
or * key) or the key's name; see Section24.5, Key names (p. 78) for a list of all key names.
Here are some examples to give you the flavor of event patterns:
<Button-1> The user pressed the first mouse button.

<KeyPress-H> The user pressed the H key.
<Control-Shift- The user pressed control-shift-H.
KeyPress-H>
24.3.Event types
The full set of event types is rather large, but a lot of them are not commonly used. Here are most of
the ones you'll need:
Activate A widget is changing from being inactive to being active. This refers to
changes in the state option of a widget such as a button changing from
inactive (grayed out) to active.
Button The user pressed one of the mouse buttons. The detail part specifies
which button.
ButtonRelease The user let up on a mouse button. This is probably a better choice in
most cases than the Button event, because if the user accidentally presses
the button, they can move it off the widget to avoid setting off the event.
Configure The user changed the size of a widget, for example by dragging a corner
or side of the window.
Deactivate A widget is changing from being active to being inactive. This refers to
changes in the state option of a widget such as a radiobutton changing
from active to inactive (grayed out).
Destroy A widget is being destroyed.
Enter The user moved the mouse pointer into a visible part of a widget. (This
is different than the enter key, which is a KeyPress event for a key whose
name is actually "return".)
Expose This event occurs whenever at least some part of your application or
widget becomes visible after having been covered up by another window.
FocusIn A widget got the input focus (see Section23, Focus: routing keyboard
input (p. 74) for a general introduction to input focus.) This can happen
either in response to a user event (like using the tab key to move focus

between widgets) or programmatically (for example, your program calls
the .focus_set() on a widget).
FocusOut The input focus was moved out of a widget. As with FocusIn, the user
can cause this event, or your program can cause it.
KeyPress The user pressed a key on the keyboard. The detail part specifies which
key. This keyword may be abbreviated Key.
KeyRelease The user let up on a key.
Leave The user moved the mouse pointer out of a widget.
Map A widget is being mapped, that is, made visible in the application. This
will happen, for example, when you call the widget's .grid() method.
Motion The user moved the mouse pointer entirely within a widget.
Unmap A widget is being unmapped and is no longer visible. This happens, for
example, when you use the widget's .grid_remove() method.
Visibility Happens when at least some part of the application window becomes
visible on the screen.
24.4.Event modifiers
The modifier names that you can use in event sequences include:
Alt True when the user is holding the alt key down.
Any This modifier generalizes an event type. For example, the event pattern
"<Any-KeyPress>" applies to the pressing of any key.
Control True when the user is holding the control key down.
Double Specifies two events happening close together in time. For example,
<Double-Button-1> describes two presses of button 1 in rapid succes-
sion.
Lock True when the user has pressed shift lock.
Shift True when the user is holding down the shift key.
Triple Like Double, but specifies three events in rapid succession.
You can use shorter forms of the events. Here are some examples:
"<1>" is the same as "<Button-1>".
"x" is the same as "<KeyPress-x>".
Note that you can leave out the enclosing "<>" for most single-character keypresses, but you can't do
that for the space character (whose name is "<space>") or the less-than (<) character (whose name is
"<less>").
24.5.Key names
The detail part of an event pattern for a KeyPress or KeyRelease event specifies which key you're
binding. (See the Any modifier, above, if you want to get all keypresses or key releases).
The table below shows several different ways to name keys. See Section24.6, Writing your hand-
ler (p. 81), below, for more information on Event objects, whose attributes will describe keys in these
same ways.

The .keysym column shows the key symbol, a string name for the key. This corresponds to the
.keysym attribute of the Event object.
The .keycode column is the key code. This identifies which key was pressed, but the code does
not reflect the state of various modifiers like the shift and control keys and the NumLock key. So, for
example, both a and A have the same key code.
The .keysym_num column shows a numeric code equivalent to the key symbol. Unlike .keycode,
these codes are different for different modifiers. For example, the digit 2 on the numeric keypad
(key symbol KP_2) and the down arrow on the numeric keypad (key symbol KP_Down) have the
same key code (88), but different .keysym_num values (65433 and 65458, respectively).
The Key column shows the text you will usually find on the physical key, such as tab.
There are many more key names for international character sets. This table shows only the Latin-1
set for the usual USA-type 101-key keyboard.
.keysym .keycode .keysym_num Key

Alt_L 64 65513 The left-hand alt key
Alt_R 113 65514 The right-hand alt key
BackSpace 22 65288 backspace
Cancel 110 65387 break
Caps_Lock 66 65549 CapsLock
Control_L 37 65507 The left-hand control key
Control_R 109 65508 The right-hand control key
Delete 107 65535 Delete
Down 104 65364
End 103 65367 end
Escape 9 65307 esc
Execute 111 65378 SysReq
F1 67 65470 Function key F1
Fi 66+i 65469+i Function key Fi
Home 97 65360 home
Insert 106 65379 insert
Left 100 65361
Linefeed 54 106 Linefeed (control-J)
KP_0 90 65438 0 on the keypad

.keysym .keycode .keysym_num Key
KP_Add 86 65451 + on the keypad
KP_Begin 84 65437 The center key (same key as 5) on the keypad
KP_Decimal 91 65439 Decimal (.) on the keypad
KP_Delete 91 65439 delete on the keypad
KP_Divide 112 65455 / on the keypad
KP_Down 88 65433 on the keypad
KP_End 87 65436 end on the keypad
KP_Enter 108 65421 enter on the keypad
KP_Home 79 65429 home on the keypad
KP_Insert 90 65438 insert on the keypad
KP_Left 83 65430 on the keypad
KP_Multiply 63 65450 on the keypad
KP_Next 89 65435 PageDown on the keypad
KP_Prior 81 65434 PageUp on the keypad
KP_Right 85 65432 on the keypad
KP_Subtract 82 65453 - on the keypad
KP_Up 80 65431 on the keypad
Next 105 65366 PageDown
Num_Lock 77 65407 NumLock
Pause 110 65299 pause
Print 111 65377 PrintScrn
Prior 99 65365 PageUp
Return 36 65293 The enter key (control-M). The name Enter refers to
a mouse-related event, not a keypress; see Section24,
Events (p. 75)
Right 102 65363
Scroll_Lock 78 65300 ScrollLock
Shift_L 50 65505 The left-hand shift key
Shift_R 62 65506 The right-hand shift key
Tab 23 65289 The tab key
Up 98 65362

24.6.Writing your handler
The sections above tell you how to describe what events you want to handle, and how to bind them.
Now let us turn to the writing of the handler that will be called when the event actually happens.
The handler will be passed an Event object that describes what happened. The handler can be either
a function or a method. Here is the calling sequence for a regular function:
def handlerName ( event ):
And as a method:
def handlerName ( self, event ):
The attributes of the Event object passed to the handler are described below. Some of these attributes
are always set, but some are set only for certain types of events.
.char If the event was related to a KeyPress or KeyRelease for a key that
produces a regular ASCII character, this string will be set to that character.
(For special keys like delete, see the .keysym attribute, below.)
.height If the event was a Configure, this attribute is set to the widget's new
height in pixels.
.keysym For KeyPress or KeyRelease events involving a special key, this attrib-
ute is set to the key's string name, e.g., "Prior" for the PageUp key. See
Section24.5, Key names (p. 78) for a complete list of .keysym names.
.keysym_num For KeyPress or KeyRelease events, this is set to a numeric version of
the .keysym field. For regular keys that produce a single character, this
field is set to the integer value of the key's ASCII code. For special keys,
refer to Section24.5, Key names (p. 78).
.num If the event was related to a mouse button, this attribute is set to the button
number (1, 2, or 3).
.serial An integer serial number that is incremented every time the server pro-
cesses a client request. You can use .serial values to find the exact time
sequence of events: those with lower values happened sooner.
.time This attribute is set to an integer which has no absolute meaning, but is
incremented every millisecond. This allows your application to determine,
for example, the length of time between two mouse clicks.
.widget Always set to the widget that caused the event. For example, if the event
was a mouse click that happened on a canvas, this attribute will be the
actual Canvas widget.
.width If the event was a Configure, this attribute is set to the widget's new
width in pixels.
.x The x coordinate of the mouse at the time of the event, relative to the
upper left corner of the widget.
.y The y coordinate of the mouse at the time of the event, relative to the
upper left corner of the widget.
.x_root The x coordinate of the mouse at the time of the event, relative to the
upper left corner of the screen.
.y_root The y coordinate of the mouse at the time of the event, relative to the
upper left corner of the screen.

Here's an example of an event handler. Under Section24.1, Levels of binding (p. 76), above, there is
an example showing how to bind mouse button 2 clicks on a canvas named self.canv to a handler
called self.__drawOrangeBlob(). Here is that handler:
def __drawOrangeBlob ( self, event ):
"Draws an orange blob in self.canv where the mouse is."
r = 5 # Blob radius
self.canv.create_oval ( event.x-r, event.y-r,
event.x+r, event.y+r, fill="orange" )
When this handler is called, the current mouse position is (event.x, event.y). The .cre-
ate_oval() method draws a circle whose bounding box is square and centered on that position and
has sides of length 2*r.
24.7.The extra arguments trick

Sometimes you would like to pass other arguments to a handler besides the event.
Here is an example. Suppose your application has an array of ten checkbuttons whose widgets are
stored in a list self.cbList, indexed by the checkbutton number in range(10).
Suppose further that you want to write one handler named .__cbHandler for <Button-1> events
in all ten of these checkbuttons. The handler can get the actual Checkbutton widget that triggered it
by referring to the .widget attribute of the Event object that gets passed in, but how does it find out
that checkbutton's index in self.cbList?
It would be nice to write our handler with an extra argument for the checkbutton number, something
like this:
def __cbHandler ( self, event, cbNumber ):
But event handlers are passed only one argument, the event. So we can't use the function above because
of a mismatch in the number of arguments.
Fortunately, Python's ability to provide default values for function arguments gives us a way out. Have
a look at this code:
def __createWidgets ( self ):

self.cbList = [] # Create the checkbutton list
for i in range(10):
cb = Checkbutton ( self, )
self.cbList.append ( cb )
cb.grid( row=1, column=i )
def handler ( event, self=self, i=i ): 1
return self.__cbHandler ( event, i )
cb.bind ( "<Button-1>", handler )

def __cbHandler ( self, event, cbNumber ):

1 These lines define a new function handler that expects three arguments. The first argument is
the Event object passed to all event handlers, and the second and third arguments will be set to
their default valuesthe extra arguments we need to pass it.
This technique can be extended to supply any number of additional arguments to handlers.

24.8.Virtual events
You can create your own new kinds of events called virtual events. You can give them any name you
want so long as it is enclosed in double pairs of <<>>.
For example, suppose you want to create a new event called <<panic>>, that is triggered either by
mouse button 3 or by the pause key. To create this event, call this method on any widget w:
w.event_add ( "<<panic>>", "<Button-3>",
"<KeyPress-Pause>" )
You can then use "<<panic>>" in any event sequence. For example, if you use this call:
w.bind ( "<<panic>>", h )
any mouse button 3 or pause keypress in widget w will trigger the handler h.
See .event_add(), .event_delete(), and .event_info() under Section19, Universal widget
methods (p. 62) for more information about creating and managing virtual events.

Tkinter Reference A GUI For Python

Uploaded by

Copyright:

Available Formats

Tkinter Reference A GUI For Python

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tkinter Reference A GUI For Python

Uploaded by

Copyright:

Available Formats

What are the main widgets described in the document?

What are the main widgets described in the document?

What are some of the standard attributes that can be set for widgets?

What are some of the standard attributes that can be set for widgets?

Tkinter reference: a GUI

New Mexico Tech Computer Center Tkinter reference 1

2 Tkinter reference New Mexico Tech Computer Center

2.A minimal application

New Mexico Tech Computer Center Tkinter reference 3

3.1.The .grid() method

4 Tkinter reference New Mexico Tech Computer Center

3.2.Other grid management methods

New Mexico Tech Computer Center Tkinter reference 5

#rgb Four bits per color

font = tkFont.Font ( option, ... )

family The font family name as a string.

6 Tkinter reference New Mexico Tech Computer Center

For example, to get a 36-point bold Helvetica italic face:

New Mexico Tech Computer Center Tkinter reference 7

8 Tkinter reference New Mexico Tech Computer Center

arrow Northeast-pointing arrow.

New Mexico Tech Computer Center Tkinter reference 9

10 Tkinter reference New Mexico Tech Computer Center

New Mexico Tech Computer Center Tkinter reference 11

5.The Button widget

activebackground Background color when the button is under the cursor.

12 Tkinter reference New Mexico Tech Computer Center

Methods on Button objects:

New Mexico Tech Computer Center Tkinter reference 13

To create a Canvas object:

bd or borderwidth Border width in pixels.

14 Tkinter reference New Mexico Tech Computer Center

6.1.1.Canvas and window coordinates

6.1.2.The display list; above and below

New Mexico Tech Computer Center Tkinter reference 15

6.1.5.Canvas tagOrId arguments

6.2.Methods on Canvas objects

.addtag_above ( newTag, tagOrId )

16 Tkinter reference New Mexico Tech Computer Center

New Mexico Tech Computer Center Tkinter reference 17

.scale ( tagOrId, xOrigin, yOrigin, xScale, yScale )

18 Tkinter reference New Mexico Tech Computer Center

New Mexico Tech Computer Center Tkinter reference 19

PIESLICE CHORD ARC

6.4.The canvas bitmap object

20 Tkinter reference New Mexico Tech Computer Center

6.5.The canvas image object

6.6.The canvas line object

New Mexico Tech Computer Center Tkinter reference 21

fill The color to use in drawing the line. Default is fill="black".

6.7.The canvas oval object

22 Tkinter reference New Mexico Tech Computer Center

6.8.The canvas polygon object

To create a new polygon object on a canvas C:

New Mexico Tech Computer Center Tkinter reference 23

6.9.The canvas rectangle object

To create a rectangle object on canvas C:

6.10.The canvas text object