Basic functionality: inserting cookies
The main function of html-helper-mode is to make it easy to insert
HTML cookies. html-helper-mode does its best to do this consistently
and conveniently. For every HTML and HTML+ cookie there should be a
template which contains the text of the cookie itself and information
as to where the fields are that the user might want to fill in.
Each template can be activated in one of two ways: by typing the key
sequence for the cookie, or by completing on the beginning of the
cookie. For example, C-cC-bt will insert the string
<title></title>, and position point where you
want to type the title. Or, you could type <kb and hit
M-TAB, and the tag <kbd></kbd> will be
If a command is called with a prefix argument (by pressing
C-u first), then instead of inserting the cookie at the
point the cookie is inserted around the region. Thus,
C-uC-cC-pb will wrap the <b></b> token
around region, thereby making it bold.
More complicated cookies have several fields that need filling in. For example,
a hyperlink (<a href=""></a>) has two fields - the URL and
the link description. When a multiple-field cookie is inserted,
point will be placed at the first field and then the other fields will
be entered on a special list. M-C-f and M-C-b
will make the point skip back and forth on this list.
html-helper-mode can prompt you in the minibuffer for all of the
fields a cookie need. For example, you can have it ask you to type in
the URL: and link description fields in a hyperlink. This feature is
turned off by default - for information on how to turn it on, read
about configuring html-helper-mode.
HTML documents can have nested lists: these are a bit hard to read
in source form. html-helper-mode has code to indent lists cookies to
the appropriate depth depending on how deeply nested the list is. List
cookies are automatically indented when they are inserted. To indent a
line by hand, hit TAB.
The indentation code is a bit of a hack - it doesn't really parse the
HTML document very carefully. The main restriction is that cookies that
are parts of lists (for example, <ul> or <li>)
should appear at the beginning of their lines for the code to work best.
New document skeletons
When a new HTML document is created, html-helper-mode can insert a
skeleton with all the tags every HTML document should have. This
feature is turned off by default - see the documentation for configuring html-helper-mode to turn it on
and to customize the skeleton.
Good HTML documents should all have timestamps stating the last time
they were modified. html-helper-mode can make this easy by
automatically updating your timestamp. This feature is turned off by
default - see the documentation for configuring html-helper-mode to turn it on
and to customize the timestamp.
Emacs editing support
One of the joys of emacs is that it is pretty smart about parsing the
contents of the buffer. However, HTML is fairly outside of the realm
of what emacs thinks is a programming language, so the syntax support
html-helper-mode provides isn't that helpful. It does define
> as a match for <, and emacs knows about
HTML comment syntax, so you can type M-; to get a
If you have hilit19 loaded in your emacs, then html-helper-mode will
automatically set up regular expressions to colour links, HTML
cookies, included images, and comments. I think this is one of the most
useful features of html-helper-mode. For these patterns to be
installed correctly hilit19 must be loaded before
I do not use font-lock, so currently there are no font-lock patterns
defined. However, there is user-contributed
font-lock code on the net.
Once you have html-helper-mode installed, you probably will want to
customize a few things. For more information, see the document on
I don't use emacs18 anymore - 19 is quite stable and much nicer. I've
tried to make sure html-helper-mode works in emacs18, and it seems to
right now, but it is not being well supported. If you can, upgrade to
Nelson Minar <email@example.com>
Last modified: Fri Oct 6 12:54:28 CEST 2000