OpenOffice::OODoc::Image - Image manipulation methods
The OpenOffice::OODoc::Image class is a derivative of OpenOffice::OODoc::XPath designed for the manipulation of graphics objects contained in documents. It mainly allows you to modify the size and position of an image and exchange its content outside the document.
This class should not be explictly used in an ordinary application, because all its features are available in the OpenOffice::OODoc::Document class, in combination with other features. So, each time an application needs to get an image-focused access to a document, it should use the general odfDocument() constructor instead of the odfImage() one.
Practically, the present manual is provided to describe the image-container processing features of OpenOffice::OODoc::Document (knowing that these features are technically supported by the OpenOffice::OODoc::Image component of the API).
Knowing that an image is displayed or printed according to a style, the OpenOffice::OODoc::Image features should be used in conjunction with the OpenOffice::OODoc::Styles ones. The OpenOffice::OODoc::Document class allows the user to invoke text-, style- and image-focused methods from the same object.
All the methods described here can equally be used with images contained in style sheets (headers, footers) as with images contained in the body of a document. It can therefore be associated just as well with a "styles.xml" member as with a "content.xml" member of an OpenOffice.org file.
This class works with all types of document (text, presentation, etc.).
For all methods where the first argument is given below as "image", it is (unless otherwise stated) either the name of an image as it appears to the end user when editing its properties in OpenOffice.org or StarOffice or the image's element reference obtained previously by the program. All these methods fail and return a null value (or in some cases produce an error message) if the argument does not correspond to a known image contained in the document.
Note: This module is not an image-processing tool. It can insert or remove images, and control the way the images are displayed in the documents. But it can't process the images themselves.
Short Form: odfImage(<parameters>)
This constructor should not be explicitly used in ordinary applications
knowing that all the features of the returned object are inherited by
any Document object.
See OpenOffice::OODoc::XPath->new for commun arguments.
The XML member loaded by default is 'content.xml', but only member
=> 'styles' is required if you want to work with page background
my $doc = odfImage
file => 'my_presentation.odp',
part => 'styles.xml'
Real applications should not expressly use this constructor,
knowing that the compound OpenOffice::OODoc::Document (whose
usual constructor is ooDocument) inherits all the features
Creates an element which represents an image and inserts it into the
document according to given parameters. The image element created is
only an anchor. It then needs to have a graphical content loaded
into it and, if needed, be given parameters using other methods.
inserts a default image element called "Logo".
Normally, the image name is unique in an OpenOffice.org document but
no checking of its uniqueness is performed here. Things you should
know, however, are:
- if several images have the same name, one only of them can be
retrieved and handled by methods which use the name to identify
them. It is almost impossible to know which of them it will be,
without a thorough knowledge of the OpenOffice.org format and
the internal logic of OODoc.
- if the document is subsequently read and saved by
OpenOffice.org (which controls the uniqueness of image names),
sequential numbers will be given to all but one of the repeated
names in order to make them unique.
Attributes can be passed in pairs [parameter => value]. Such as:
style => image style name
It should be noted that an image should normally be associated with
a graphic style. The presence of a style is not obligatory when
calling createImageElement (and this is not checked), but the image
should preferably be actually linked to an existing style before
displaying or printing the document. See OpenOffice::OODoc::Styles
for style creation or, better, OpenOffice::OODoc::Document for image
attachment => anchor element
indicates if the image is attached to a text element (for ex. a
paragraph), and which one. This parameter must be an existing
element reference (obtained, for example, using getElement or
selectElementxxx). It is useless if the image is linked to a page.
The OASIS OpenDocument specification doesn't provide the list of
possible attachments for an image, and OpenOffice.org make some
differences in this area between the different document classes.
For example, OOo Writer doesn't display images which are directly
attached to table cells, while OOo Calc does (in text documents,
an image which appears in a table cell is attached to a paragraph,
possibly empty, belonging to the cell; see the example below).
page => anchor page
If this parameter is used, it indicates that the image will be
anchored to a page, and the given value is a page number. It does
not matter if, when createImageElement is called, this number is
beyond the end of the document or not. If the content class of the
document is "presentation" (Impress) or "drawing" (Draw), then the
page option is mandatory and must be either the visible name or the
element reference of an existing draw page. In a "text" content,
all that matters is that the particular page exists when it is opened
by OpenOffice.org and if this parameter is absent, the image is
anchored to a paragraph.
position => coordinates
This parameter indicates the x,y coordinates of the image in
relation to its anchor point. By default and generally, if the page
parameter is given, the origin (0,0) is the top left corner of the
physical page. When attached to a text element, if there is no given
position, the image is appended to the text.
Coordinates go from left to right and top to bottom, however
everything really depends on the image style. Coordinates should be
given here in the form of a string "x,y", and the default unit is
centimeters. You can choose millimeters instead by attaching the
usual abbreviation, such as "12.5cm, 35mm" which is the same as
"125mm, 3.5cm" or "12.5,3.5", etc. The point ("pt") unit is allowed
size => size
The image's size (width, height) is given here in the same syntax as
for position. Caution: if no size is given, you must not assume that
the image will be spontaneously displayed in the document in a
convenient size. Remember that the "original" size of the image is
not automatically selected by default; the application must provide
a display size. If an image is to be displayed according to its
original size which is not known in advance, you can get it using,
for example, the imgsize() function of Image::Size (this function
returns the size in points, so the "pt" unit must be expressly
selected unless you convert the size in centimeters).
description => label
This optional parameter gives the descriptive text (long label) for
the image as it will appear to the end user when editing the image's
link => link
The link parameter contains a reference to the image's physical
content which can be inserted into an OpenOffice.org file (internal
link) or a reference to an external file or even the URL of an image
accessible using a communication protocol supported by the
Remember that, with OpenOffice.org, physical images are "referenced",
and not "loaded" into XML elements. The reverse of course would not
only be against "good practice" but would result in outrageously
large documents. According to the OpenDocument specification, the
conforming applications could embed base-64-encoded graphics in the
XML elements, but this option is not used when you insert an image
in a document through the OOo GUI.
If the target link is unavailable when the document is displayed,
an error message is displayed in place of the image. (See imageLink
about links.) This parameter is useless if import is given.
Relative paths in the local filesystem are allowed, but should be
used with care, due to differences between operating systems... and
between editing applications (for example, in a Unix environment,
OOo 2.x requires "../image.jpg" while OOo 1.x allows "image.jpg" for
a link to an image file in the current directory).
import => image file
The presence of this parameter indicates that the image content
should be imported from an external file and, implicitly, that the
image link (which it is useless to give here) points to the image
imported into the OpenOffice.org file. The import will only be made
when all the updates are validated by a save(). See importImage()
about importing images. An imported image is *not* embedded in the
XML image element; it's always referred to through a link, and stored
in a separate, non-XML member of the ODF archive. However, with the
"import" option, the appropriate internal link is automatically
created and the new member, containing the image, is automatically
built later, when the save() method is executed from the current
document (provided that the image file is available).
For those who know the appropriate XML vocabulary, there are other
parameters you can pass. Parameters other than those described above
are written to the image's XML element as is and without any checks.
All of these attributes, and others, can be read or modified later
by other methods.
We would discourage you from relying on the default choices in a
serious application, but it is still a possibility. With no
parameters (other than mandatory image name), the createImageElement
method chooses its own often arbitrary course according to the
- if the OODoc::Image object is associated with a document body
(document-content) then a new paragraph is created at the end of
the document and the image is inserted into this paragraph
(appears at the end of the text). Repeated image creations
without parameters will therefore add images one after the other
at the end of the document.
- if the OODoc::Image object is associated with a background
(document-styles) then it tries to create a paragraph in the
first available header and insert the image into it. If no page
style contains a header, the same is attempted in the first
available footer. If there is no footer either, the creation
fails with the message "No valid attachment".
This method otherwise tries to be "intelligent" whenever the set of
parameters is incomplete. If the results are useful, then the rest
is up to you...
The method returns the new image element's reference (undef if it
The following example attaches an image to a paragraph, gives it a
size, loads its content into it from an external file and attributes
a style to it which has been defined elsewhere. You will note that
this example combines an OODoc::Image method (createImageElement)
with an OODoc::Text method (getParagraph). This means that, here,
the $doc object is a OODoc::Document class. See also
createImageStyle in OODoc::Document.
description => "Kilimanjaro in winter",
attachment => $doc->getParagraph(4),
size => "5cm, 3.5cm",
style => "gr1",
import => "C:\Images\Landscape.jpg"
The same image element could be inserted in a table cell. To do
so, in a spreadsheet document, the "attachment" option could be
set with a $doc->getCell($table, $row, $column) value. But the
present version of OpenOffice.org doesn't allow direct cell
attachments in text documents; the image element must be attached
to a paragraph which is in turn attached to the target cell. So
a possible approach consists of 1) issuing a setText($cell, "") in
order to ensure the target cell contains an empty text paragraph
and 2) provide an "attachment" option set to
$doc->getCellParagraph($cell) in order to anchor the image to this
Remember: creating an image element will only make that image appear
in the document if 1) the image has a valid link which points to a
valid image, 2) the corresponding graphics file exists (loaded
possibly using importImage) in the archive if it is an internal
link, and 3) the image has a style actually defined elsewhere
(either pre-existing, created using the createStyle method of
OODoc::Styles, copied from another document using replicateElement,
or coming from another source).
Exports the content of an image contained in a document if the
OODoc::Image object was linked to a file when it was created (with
file or archive parameters passed to the new constructor).
The first argument is either the name of the image (as it would
appear to the end user in the image's properties in OpenOffice.org),
or the image element's reference if the program already has it.
The second optional argument is the destination file. Example:
$doc->exportImage("Logo1", "C:\My Documents\logo.jpg");
It is up to the application to choose an appropriate extension for
the exported file (.jpg, .gif, .png, etc.). You can easily find out
which extension using the imageLink accessor.
Without the second argument, the image file is created in the
current directory, and its name is the name of the image in the
document, with an extension depending on the format (.jpg, .png,
etc), according to the information stored in the document (but
the format is not checked). Be careful, the export fails unless
such a construct provides a valid file name for the operating system.
If the image is unnamed in the document, exportImage() tries to
build a path/name which replicates the internal path/name of the
image in the archive and to use it under the current directory
(this path normally begins with "Pictures/").
Caution: this method only exports what is exportable i.e. internal
images (physically contained in the file). It has no effect if used
with an image inserted by an external link into the document.
Exports all or part of the images contained in a document.
By default, and with no parameters, each internal image is exported
to a file whose access path is the same as it would be if using
exportImage. This behaviour can be changed by parameters passed
in hash form (parameter => value).
Possible parameters are as follows:
selection => filtering of image names (regex)
filter => filtering of image names (regex)
name => filtering of image names (regex)
target => path and/or basic filename
path => path and/or basic filename
suffix => extension to be given to filenames
extension => extension to be given to filenames
start_count => begin count indicator
The "filter", "selection" or "name" parameters allow you to export
only those images whose names match the given regular expression.
Such filtering works on the name as the document "knows" it, i.e. as
it appears to the end user in the image's properties within the
document. It is not a technical filter and does not allow, for
example, selection of images according to their file type.
The "target" or "path" parameters allow you to choose the access
path and basic filename for exported files (e.g.
"/usr/local/images/img"). If given, a sequential number will
automatically be added to the basic filename to identify each file.
The "suffix" or "extension" parameters allow the application to
force a common extension for each file instead of leaving the
extension as it existed in the archive (which normally identifies the
physical image type).
By default, the sequential numbers given to filenames (between basic
name and extension) are reset to zero each time exportImages is
called. An application can however force the numbering to start at a
different value using the "start_count" parameter.
In a list context, this method returns a list of exported files
which the application can use later. In a scalar context, it returns
the number of exported files.
Returns the list of all image elements in the active context. If the
current OODoc::Image object is associated with document-content,
this will be the images contained in the body of the document. If
associated with document-styles, this will be the images linked to
headers and footers.
Returns the element which corresponds to the image whose name is
given as an argument. This name is usually unique as OpenOffice.org
does not allow the user to give two files the same name in a
document. It is the name which appears in the "Options" tab when
editing the object's properties in OpenOffice.org Writer, or in the
"Name object" dialog box when you right-click an image in
OpenOffice.org Impress. Returns undef if the image is not found.
Can also be used as a check method by specifying an element as the
argument instead of a name. In this case, the given element is
simply returned without modification if it is indeed an image or
undef if not.
Caution: images do not always have names. OpenOffice.org Writer gives
default names to images (e.g. Image1, Image2, ...) if the user does
not deliberately name them. This is not the case in Impress.
This method returns the given image's link (see definition of "link"
in the section on imageLink), but only if it is an internal link in
a form which is directly usable by a zip archive management tool
(without the initial "#").
Accessor which allows you to check or modify (even create) an
image's XML attributes directly. The attribute is modified or
created if a value is given as the third argument. If not, it
returns the current value of the attribute if found, or undef if
not. The name and value of the attribute must be given according to
the OpenOffice.org vocabulary.
This generic accessor remains invisible to most applications as
specialist accessors are available for the most useful attributes
(e.g. imageLink, imageName, etc.).
Returns an image's description, or if the "text" argument is given,
replaces it. This description corresponds to the optional text which
appears in the Options tab when editing the image's properties in
OpenOffice.org. It is not used in Impress or Draw documents.
Reads or modifies an image's link. A link is the address of the file
which physically contains the image.
If working with an image loaded into the document (often but not
always the case), the link is internal and the file is physically
stored in the zip archive containing the document. In this case, the
link is written as "#Pictures/xxxx". It corresponds to what the user
would see if opening the file using a zip archive tool instead of
opening it in OpenOffice.org.
If working with an external image, the link is then the URL of the
The same method allows you to read and modify the link. If a second
argument is given, it cancels and replaces the existing link, or
creates a link if the image was still "empty".
Caution: the actual physical existence of a valid image which
corresponds to the link is not checked.
Reads an image's name, or if the second argument is given, replaces
Returns undef if the first argument (name or reference) is not an
Allows you to read or modify the coordinates of an image in relation
to its anchor point. Coordinates are returned in the form of a pair
my ($x, $y) = $doc->imagePosition("Landscape");
If the coordinate arguments are given, they replace the image's old
Caution: coordinates are not numeric values in the classic sense of
the word. They are character strings starting with numeric values
and ending with units of measure (normally "cm" or "mm"). If an
application passes only numeric arguments without giving a unit of
measure, the default unit is assumed to be the centimetre.
$doc->imagePosition("Landscape", "2.5cm", "5.125cm");
Normally, coordinates are measured from left to right and from top
to bottom starting from the point of reference. The point of
reference is normally the top left corner of the page or paragraph
to which it is anchored (depending on the type of anchorage). All
this can depend on the page style.
Allows you to read or modify the display size of an image. Returns
the height and width as a pair of values. If height and width
arguments are provided, they replace the image's old size.
Note that this method deals with the display size only, and ignores
the original size of the image; in other words, it doesn't change
anything in the graphic object itself.
Caution: it is up to the application to maintain the aspect ratio of
the image if needed.
See the "size" option about the createImageElement() method.
See imagePosition() about measure formats.
Returns the name of the image's current style, or changes this style
if the second argument "style" is given. In the second case, the
presence or pertinence of the called style is not checked. This can
otherwise be created or imported later.
Loads an image's content from an external file.
This import cancels and replaces any previous image content, if any.
The first argument is either the image's name (as the user would see
it) or the image's corresponding element reference. In any case, the
image element must already exist, created if necessary using
The second argument is the full path of the image file to be
imported and associated to the ODF image object. If this argument is
omitted, importImage() assumes that the full path of the needed
resource file is indicated by an external link already associated
with the image object in the document (see imageLink).
Obviously, the imported file should preferably be in a graphic format
supported by most office software products, but this method actually
allows you to import anything, including data which could not be
Caution: this method does not carry out the import immediately and
does not check for the presence of the file to be imported. It only
sends the request to the associated OODoc::File object (via its
raw_import method) which performs the operation at the next save()
call. As a consequence, the image file is needed later, but its
availability is not required by the importImage() method itself.
Method added to OpenOffice::OODoc::Element class objects by the
OODoc::Image package and is an element method rather than a document
method. Allows you to "ask" an element if it is an image.
print "This is an image" if $element->isImage;
Deletes an image's optional image description.
Returns the first (or only) image (if found) which has an image
description matching the provided regular expression.
See also selectImageElementsByDescription().
Allows you to select an image element by its link (as in imageLink)
instead of by its name. The link can be an approximation and
represented in this case by a regular expression.
It should only be used if you are sure that the link is unique in
the document. This is not normally the case, as many images can
share the same content. Conversely, the returned element is the
first image, in the physical XML order, which has the given link
(and not necessarily in the logical order of the document).
Returns undef if no image has the given link and therefore allows
you to find out if a particular target (e.g. URL) is used in the
Note: if an application needs to get a list of individual or
multiple images which share the same target, the easiest way is to
loop through the list returned by getImageElementList() and perform an
imageLink() on each element.
See also selectImageElementsByLink().
Returns a list of images whose image descriptions match the given
Returns a list of images whose internal or external links match the
given regular expression.
Allows you, for example, to get a list of image elements which share
the same physical image file or even those which use a particular
transport protocol to access these images.
my @webimages = $doc->selectImageElementsByLink("^http:");
allows you to get a list of images referred to through an HTTP URL
(i.e. which will not be visible if the user's machine is not connected
to the internet).
Returns a list of images whose names match the given regular
No variable is exported.
There is a static class variable which contains the description of a
default image style (in hash form):
This hash is not used directly by the module itself (which offers no
styles functions), but it is available to any application which
needs to create image styles using either XPath base methods or the
Styles module. Its structure is exactly the same as the parameters
hash expected by the createStyle method. By default, it is
automatically used by the createImageStyle method of the Document
Developer/Maintainer: Jean-Marie Gouarne http://jean.marie.gouarne.online.fr
Copyright 2004-2009 by Genicorp, S.A. http://www.genicorp.com
Initial English version of the reference manual by Graeme A. Hunter (firstname.lastname@example.org).
License: GNU Lesser General Public License v2.1
To install OpenOffice::OODoc, copy and paste the appropriate command in to your terminal.
perl -MCPAN -e shell
For more information on module installation, please visit the detailed CPAN module installation guide.