![]() |
![]() |
![]() |
![]() |
![]() |
Displays the print dialog box in order to print the explicitly selected element if any, and the whole document otherwise.
Preferences related to printing may be specified in the Print tab of the Options dialog box.
The Search/Replace dialog box
The document is searched from caret position to the end of the document.
If replacements are to be performed along with the search and an element has been explicitly selected, the search/replace starts at the beginning of this element, whatever is the position of the caret, and ends at the end of this element. (A backward search/replace starts at the end of the selected element and ends at its beginning.)
In such case, $1, $2, ..., $9 may be used in the replacement string to refer to the substrings matching the parenthesized groups of the regular expression.
$0 is replaced by the string matching the regular expression in its entirety. $$ may used to quote character '$'.
Examples:
Backward regular expression search/replace may be slow to the point of being unusable.
Preferences related to searching may be specified in the Search tab of the Options dialog box.
The Spell Checker dialog box
Tools->CheckSpellingErrors (keyboard shortcut Ctrl-Shift-S) checks the text of the document from the word containing the caret to the end of the document.
If an element has been explicitly selected, the spell checker starts at the beginning of this element, whatever is the location of the caret, and stops at the end of the element.
Note that if the part of the document to be checked contains no spelling errors the dialog box shown above is not displayed at all. Instead message "Found no spelling errors" is displayed at the bottom of XXE window.
This text field can also be used to modify this word. After editing the word, typing on the Enter key replaces the original word by the modified one.
It is possible to switch from a language to another at any time. In such case, the spell checker is automatically restarted with the other dictionary, beginning at the word displayed in the text field.
The last selected language is recorded in the user preference file in order to be automatically chosen in subsequent XXE sessions.
Note that, in this milestone version, the spell checker ignores xml:lang and lang attributes and that it cannot be configured to automatically skip certain types of elements.
An empty text field may be used to delete the original word.
This button is disabled for errors different from "Unknown word" or "Improperly capitalized word".
This button is disabled for errors different from "Unknown word" or "Improperly capitalized word".
This button is disabled for errors different from "Unknown word".
Personal dictionaries are text files named
dict_LL.txt, where LL is an ISO code for a
language. These files are created in
directory
~/.xxespell
under Unix and in
directory
C:\winnt\Profiles\<user>\xxespell
under
Windows.
Personal dictionaries are automatically saved when XXE is exited.
The Options dialog box
Default: no.
This heuristic is often a good one for XML-data.
Default: no.
However, if you need to deliver an XML file created by XXE to a person using a text editor to view or modify it, it is nicer to use a human-readable encoding such as ISO-8859-1 (Western character set). In such case, check the ISO-8859-1 toggle.
Default: UTF-8.
Notes:
However, if you need to deliver an XML file created by XXE to a person using a text editor to view or modify it, it is nicer to provide this person with an indented XML file:
Default: no.
Default: no.
In indented files, elements with an empty content model are saved as (XHTML example) <br /> rather than <br/> and empty elements with a non-empty content model are saved as <p></p> rather than <p/>.
This has been recommended by the W3C to help Web browsers (HTML user agents as they say) to correctly display XHTML. For DTDs other than XHTML, this behavior causes no harm. It even improves the readability of the indented XML files created by XXE.
Default: 2.
Default: 78.
This number is only used as a hint: XML files created by XXE may contain lines much longer than this number.
Default: 100dpi.
Default: empty.
Each area can contain a mix of text and variables:
Variable | Description |
%F | File name of the document being edited |
%P | Equivalent to "page %I of %C"; see below |
%I | Current page number |
%C | Total page count |
Default: empty.
Default: empty.
Default: gray.
Note that the font used for the page header is the default font of the style sheet (see the Style tab below).
Default: yes.
Default: 10.
This option specifies how a new element is recursively created when the content type for this type of element has several possible choices for the child elements.
DocBook example: the new element to be inserted is a figure.
<!ELEMENT figure ((title,titleabbrev?),(literallayout|programlisting| programlistingco|screen|screenco|screenshot|synopsis|cmdsynopsis| funcsynopsis|classsynopsis|fieldsynopsis|constructorsynopsis| destructorsynopsis|methodsynopsis|address|blockquote|graphic| graphicco|mediaobject|mediaobjectco|informalequation|informalexample| informalfigure|informaltable|indexterm|beginpage|link|olink|ulink)+)>
With this option, the newly created element may be invalid. Therefore, use this option if you really know what you are doing.
The, very questionable, heuristic selects the text container with the shortest name. This is the default value for this option.
This behavior can be specified by choosing one of the 3 following modes (each mode corresponds to a radio button in the Search tab):
This mode uses the DTD to recognize logically contiguous text across different types of elements.
A text search behaves as if 2 linefeed characters have been inserted after each contiguous piece of text. It is therefore possible using regular expression "\n\n" to find and visualize these text boundaries in the document.
Default: mode 1.
Default: yes.
Default: no.
Default: no.
These preferences parameterize the CSS style sheet used to visualize the document or to print it.
Setting some of these preferences will have no visible effect if the style sheet author has specified the corresponding properties in the style sheet. For example, if the user preferred background is specified in the Style pane as being light yellow and if the style sheet author has specified the root element background-color as being white, the document will be rendered with a white background.
Default: Serif (the Java default serif font family).
Default: SansSerif (the Java default sans-serif font family).
Default: Monospaced (the Java default monospaced font family).
Default: SansSerif.
Default: 12pt.
Default: white.
Default: black.
XXE can be used to edit non-XML structured formats like HTML, Javadoc(TM) comments, APT, /etc/passwd, /etc/printcap, http.conf (the Apache configuration file), etc, if plug-ins are provided for such structured formats.
A format plug-in works by converting the data on the fly from the non-XML format to XML when loading a document and by converting them back to the non-XML format each time the document is saved.
XXE relies on the file name extension to find out which plug-in is needed to load a document. Documents whose file name extension has not been associated with any format plug-in are loaded as XML (unless downloaded from a http server which returns the MIME type of the document).
XXE is bundled with two format plug-ins: one for Javadoc and one for APT. See below for a description of these plug-ins.
Before being able to load a non-XML document into XXE, the user needs to register the corresponding plug-in with the application.
The format plug-ins bundled with XXE are automatically registered the first time the user starts XXE. You may need to manually register these plug-ins before being able to use them if you have an existing XXE user preference file (for example if you have upgraded your XXE distribution). The XXE user preference file is stored in file ~/.xxe under Unix and in file C:\winnt\Profiles\<user>\xxe.ini under Windows NT.
Procedure:
These first 3 steps are not needed for the bundled plug-ins.
The dialog box used to register a format plug-in with XXE
The dialog box used to configure a format plug-in
Note that the plug-in parameters are documented in this dialog box.
A Java file as displayed by XXE
This plug-in adds Javadoc(TM) support to XXE.
It must be registered as com.xmlmind.javadocformat.Format.
The intended audience for this plug-in is Java programmers and Javadoc writers. With XXE and this plug-in, it becomes possible to edit the Javadoc comments contained in a Java file using a word processor-like view. Writing Javadoc this way is less tedious (no manual formatting of comment lines) and is no longer error-prone (DTD-directed editing).
This plug-in has full support for all Javadoc 1.3 tags plus most of HTML 3.2 tags.
The following HTML 3.2 tags and attributes are not supported:
The Javadoc 1.3 tags are modeled in XML as follows:
The file name extension required for a ``Javadoc document'' is .java.
Unlike Web browsers, this plug-in is not designed to load broken HTML 3.2. However, this plug-in can help Javadoc writers to easily spot and fix the HTML errors contained in Javadoc comments. See next section for a real world case study.
the Javadoc plug-in refuses to load Sun's String.java for the above reason
When XXE refuses to load a Java file, an error dialog is displayed with
Try to guess what the error message means and fix the problem using your favorite text editor (use xmleditor-support@xmlmind.com if you are really stuck), then reload the Java file into XXE.
Warning: before using this plug-in for the first time, please take the time to configure it properly to make sure that its newline and tab policies are compatible with yours.
The Javadoc plug-in has been tested on all the Java sources given by Sun for the Linux JDK 1.3.1.
The plug-in has succeeded to load 1750 out of 1877 Java files. It has failed 127 times generally for the following reasons:
Example: putting the name of a @param between <code></code>.
Note that it is not useful to do so because Javadoc automatically adds a sensible style to this kind of plain text.
This is the most important discrepancy between the Javadoc plug-in and Javadoc: Javadoc intelligently discards <code></code>, while the Javadoc plug-in stubbornly refuses to load the Java file.
Example 1: putting plain text or inline elements such as <code></code> directly into a blockquote.
Example 2: putting a pre block between <code></code>.
Note that all the errors described above are impossible to do if XXE is used to edit the Javadoc comments.
Having succeeded to load a Java file into XXE does not mean that this file is valid: 45 out the 1750 loaded files were found having validity problems.
Generally these validity problems are minor and are caused by missing attributes or invalid attribute values. It is possible to quickly fix them by using menu command Tools->CheckValidity.
This plug-in adds APT (Almost Plain Text) format support to XXE.
It must be registered as com.xmlmind.aptformat.APTFormat.
APT is a text format for technical articles which is simple, structured, tag-less, and which is convertible to HTML, XHTML, PDF, RTF, PostScript, SGML or XML DocBook using the Open Source tool called aptconvert. Aptconvert can be downloaded from http://www.pixware.fr/.
The file name extension required for an APT document is .apt or .txt.