Supercite User's Manual
Supercite Version 3.1
Manual Revision: 3.47 August 1993
Barry A. Warsaw bwarsaw@cen.com ...!uunet!cen.com!bwarsaw Copyright (C) 1993 Barry A. Warsaw
Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies.
The preferred way to spell Supercite is with a capital `S', lowercase `upercite'. There are a few alternate spellings out there and I won't be terribly offended if you use them. People often ask though...
Supercite is only useful in conjunction with MUAs and NUAs such as VM, GNUS, RMAIL, etc. (hereafter referred to collectively as MUAs). Supercite is typically called by the MUA after a reply buffer has been setup. Thereafter, Supercite's many commands and formatting styles are available in that reply buffer until the reply is sent. Supercite is re-initialized in each new reply buffer.
Supercite is currently at major revision 3.1, and is known to work in the following environments:
Another potentially useful thing would be for Supercite to set up the
outgoing mail headers with information it gleans from the reply buffer.
But by previously agreed upon convention, any text above the
mail-header-separator which separates mail headers from message
bodies cannot be modified by Supercite. Supercite, in fact, doesn't
know anything about the meaning of these headers, and never ventures
outside the designated region. See section Hints to MUA Authors for more
details.
sc-cite-original has been added. When sc-cite-original is
executed, the original message must be set up in a very specific way,
but this is handled automatically by the MUA. See section Hints to MUA Authors.
The first thing Supercite does, via sc-cite-original, is to parse
through the original message's mail headers. It saves this data in an
information association list, or info alist. The information
in this list is used in a number of places throughout Supercite.
See section Information Keys and the Info Alist.
After the mail header info is extracted, the headers are optionally removed (nuked) from the reply. Supercite then writes a reference header into the buffer. This reference header is a string carrying details about the citation it is about to perform.
Next, Supercite visits each line in the reply, transforming the line according to a customizable "script". Lines which were not previously cited in the original message are given a citation, while already cited lines remain untouched, or are coerced to your preferred style. Finally, Supercite installs a keymap into the reply buffer so that you have access to Supercite's post-yank formatting and reciting commands as you subsequently edit your reply. You can tell that Supercite has been installed into the reply buffer because that buffer's modeline will display the minor mode string `SC'.
When the original message is cited by sc-cite-original, it will
(optionally) be filled by Supercite. However, if you manually edit the
cited text and want to re-fill it, you must use an add-on package such
as filladapt or gin-mode. These packages can recognize
Supercited text and will fill them appropriately. Emacs' built-in
filling routines, e.g. fill-paragraph, do not recognize cited
text and will not re-fill them properly because it cannot guess the
fill-prefix being used.
See section Post-yank Formatting Commands for details.
As mentioned above, Supercite provides commands to recite or uncite regions of text in the reply buffer, and commands to perform other beautifications on the cited original text, maintaining consistent and informative citations throughout. Supercite tries to be as configurable as possible to allow for a wide range of personalized citation styles, but it is also immediately useful with the default configuration, once it has been properly connected to your MUA. See section Getting Connected for more details.
>> John originally wrote this >> and this as well > Jane said that John didn't know > what he was talking about And that's what I think too.
Note that multiple inclusions of the original messages result in a
nesting of the `>' characters. This can sometimes be quite
confusing when many levels of citations are included since it may be
difficult or impossible to figure out who actually participated in the
thread, and multiple nesting of `>' characters can sometimes
make the message very difficult for the eye to scan.
In non-nested citations, each cited line begins with an informative string attributing that line to the original author. Only the first level of attribution will be shown; subsequent citations don't nest the citation strings. The above dialog might look like this when non-nested citations are used:
John> John originally wrote this John> and this as well Jane> Jane said that John didn't know Jane> what he was talking about And that's what I think too.
Notice here that my inclusion of Jane's inclusion of John's original message did not result in a line cited with `Jane>John>'.
Supercite supports both styles of citation, and the variable
sc-nested-citation-p controls which style it will use when citing
previously uncited text. When this variable is nil (the default),
non-nested citations are used. When non-nil, nested citations
are used.
sc-citation-leader, and has the default value of a
string containing four spaces.
sc-citation-delimiter visually separates the citation from the
text of the line. This variable has a default value of ">" and
for best results, the string should consist of only a single character.
sc-citation-separator, and has the default value of
a string containing a single space.
For example, suppose you were using the default values for the above
variables, and Supercite provided the attribution string `Jane'.
In this case, the composed, non-nested citation string used might be
something like
" Jane> ".
This citation string will be inserted in front of
every line in the original message that is not already cited.
Nested citations, being simpler than non-nested citations, are composed of the same elements, sans the attribution string. Supercite is smart enough to not put additional spaces between citation delimiters for multi-level nested citations.
The variable sc-citation-leader-regexp describes how citation
leaders can look, by default it matches any number of spaces or tabs.
Note that since the lisp function looking-at is used to do the
matching, if you change this variable it need not start with a leading
"^".
Similarly, the variables sc-citation-delimiter-regexp and
sc-citation-separator-regexp respectively describe how citation
delimiters and separators can look. They follow the same rule as
sc-citation-leader-regexp above.
When Supercite composes a citation string, it provides the attribution
automatically. The analogous variable which handles recognition of the
attribution part of citation strings is sc-citation-root-regexp.
This variable describes the attribution root for both nested and
non-nested citations. By default it can match zero-to-many alphanumeric
characters (also ".", "-", and "_"). But in some situations,
Supercite has to determine whether it is looking at a nested or
non-nested citation. Thus the variable
sc-citation-nonnested-root-regexp is used to describe only
non-nested citation roots. It is important to remember that if you
change sc-citation-root-regexp you should always also change
sc-citation-nonnested-root-regexp.
Nemacs users: For best results, try setting
sc-citation-root-regexp to:
"\\([-._a-zA-Z0-9]\\|\\cc\\|\\cC\\|\\ch\\|\\cH\\|\\ck\\|\\cK\\|\\ca\\|\\cg\\|\\cr\\|\\cu\\)*"
Mule users: For best results, try setting
sc-citation-root-regexp to:
"\\([-._a-zA-Z0-9]\\|\\cj\\)*"
sc-mail-field function. Thus, if the following fields were
present in the original article:
Date: 08 April 1991, 17:32:09 EST Subject: Better get out your asbestos suit
then, the following lisp constructs return:
(sc-mail-field "date") ==> "08 April 1991, 17:32:09 EST" (sc-mail-field "subject") ==> "Better get out your asbestos suit"
Since the argument to sc-mail-field can be any string, it is
possible that the mail field will not be present on the info alist
(possibly because the mail header was not present in the original
message). In this case, sc-mail-field will return the value of
the variable sc-mumble.
Supercite always places all mail fields found in the yanked original article into the info alist. If possible, Supercite will also places the following keys into the info alist:
"sc-attribution"
"sc-citation"
"sc-from-address"
"sc-reply-address"
"sc-sender-address"
"sc-emailname"
"sc-initials"
"sc-author"
"sc-firstname"
"sc-lastname"
"sc-middlename-1"
If the author's name has more than one middle name, they will appear as
info keys with the appropriate index (e.g., "sc-middlename-2",
...). See section Selecting an Attribution.
There are a number of built-in header rewrite functions supplied
by Supercite, but you can write your own custom header rewrite functions
(perhaps using the built-in ones as examples). The variable
sc-rewrite-header-list contains the list of such header rewrite
functions. This list is consulted both when inserting the initial
reference header, and when displaying electric references.
See section Electric References.
When Supercite is initially run on a reply buffer (via
sc-cite-original), it will automatically call one of these
functions. The one it uses is defined in the variable
sc-preferred-header-style. The value of this variable is an
integer which is an index into the sc-rewrite-header-list,
beginning at zero.
sc-header-on-said
below, date and from correspond to the values of the
`Date:' and `From:' mail headers respectively.
Also, the string ">>>>>" below is really the value of the
variable sc-reference-tag-string. This variable is used in all
built-in header rewrite functions, and you can customize its value to
change the tag string globally.
Finally, the references headers actually written may omit certain parts of the header if the info key associated with infokey is not present in the info alist. In fact, for all built-in headers, if the `From:' field is not present in the mail headers, the entire reference header will be omitted (but this usually signals a serious problem either in your MUA or in Supercite's installation).
sc-no-header
nil to produce a blank header. This header can possibly contain
a blank line after the mail-header-separator line.
sc-no-blank-line-or-header
sc-no-header except that any blank
line after the mail-header-separator line will be removed.
sc-header-on-said
>>>>> On date, from said:
sc-header-inarticle-writes
>>>>> In article message-id, from writes:
sc-header-regarding-adds
>>>>> Regarding subject; from adds:
sc-header-attributed-writes
>>>>> "sc-attribution" == sc-author <sc-reply-address> writes:
sc-header-author-writes
>>>>> sc-author writes:
sc-header-verbose
>>>>> On date,>>>>> sc-author>>>>> from the organization of organization>>>>> who can be reached at: sc-reply-address>>>>> (whose comments are cited below with: "sc-cite")>>>>> had this to say in article message-id>>>>> in newsgroups newsgroups>>>>> concerning the subject of subject>>>>> see references for more details
sc-preferred-header-style. However, you may want to select
different reference headers based on the type of reply or forwarding you
are doing. You may also want to preview the reference header before
deciding whether to insert it into the reply buffer or not. Supercite
provides an optional electric reference mode which you can drop
into to give you this functionality.
If the variable sc-electric-references-p is non-nil,
Supercite will bring up an electric reference mode buffer and place you
into a recursive edit. The electric reference buffer is read-only, so
you cannot directly modify the reference text until you exit electric
references and insert the text into the reply buffer. But you can cycle
through all the reference header rewrite functions in your
sc-rewrite-header-list.
You can also set a new preferred header style, jump to any header, or jump to the preferred header. The header will be shown in the electric reference buffer and the header index and function name will appear in the echo area.
The following commands are available while in electric reference mode (shown here with their default key bindings):
sc-eref-next (n)
sc-electric-circular-p is non-nil, invoking
sc-eref-next while viewing the last reference header in the list
will wrap around to the first header.
sc-eref-prev (p)
sc-electric-circular-p is non-nil,
invoking sc-eref-prev will wrap around to the last header.
sc-eref-goto (g)
sc-rewrite-header-list) can be specified as a numeric argument to
the command. Otherwise, Supercite will query you for the index in the
minibuffer.
sc-eref-jump (j)
sc-preferred-header-style.
sc-eref-setn (s)
sc-preferred-header-style) to the currently displayed header.
sc-eref-exit (LFD, RET, and ESC C-c)
sc-eref-abort (q, x)
Supercite will execute the hook sc-electric-mode-hook before
entering electric reference mode.
nil value, which the MUA recognizes to mean,
"use your default citation function". When you add Supercite's
citation function to the hook, thereby giving the variable a
non-nil value, it tells the MUA to run the hook via
run-hooks instead of using the default citation.Early in Supercite's development, the Supercite author, a few MUA authors, and some early Supercite users got together and agreed upon a standard interface between MUAs and citation packages (of which Supercite is currently the only known add-on :-). With the recent release of the Free Software Foundation's GNU Emacs 19, the interface has undergone some modification and it is possible that not all MUAs support the new interface yet. Some support only the old interface and some do not support the interface at all. Still, it is possible for all known MUAs to use Supercite, and the following sections will outline the procedures you need to follow.
To learn exactly how to connect Supercite to the software systems you are using, read the appropriate following sections. For details on the interface specifications, or if you are writing or maintaining an MUA, see section Hints to MUA Authors.
The first thing that everyone should do, regardless of the MUA you are using is to set up Emacs so it will load Supercite at the appropriate time. You can either dump Supercite into your Emacs binary (ask your local Emacs guru how to do this if you don't know), or you can set up an autoload for Supercite. To do the latter, put the following in your `.emacs' file:
(autoload 'sc-cite-original "supercite" "Supercite 3.1" t) (autoload 'sc-submit-bug-report "supercite" "Supercite 3.1" t)
The function sc-cite-original is the top-level Supercite function
designed to be run from the citation hook. It expects
`point' and `mark' to be set around the region to cite, and it
expects the original article's mail headers to be present within this
region. Note that Supercite never touches any text outside this
region. Note further that for Emacs 19, the region need not be active
for sc-cite-original to do its job.
See section Hints to MUA Authors.
The other step in the getting connected process is to make sure your
MUA calls sc-cite-original at the right time. As mentioned
above, some MUAs handle this differently. Read the sections that follow
pertaining to the MUAs you are using.
One final note. After Supercite is loaded into your Emacs session, it
runs the hook sc-load-hook. You can put any customizations into
this hook since it is only run once. This will not work, however, if
your Emacs maintainer has put Supercite into your dumped Emacs' image.
In that case, you can use the sc-pre-hook variable, but this will
get executed every time sc-cite-original is called. See section Reply Buffer Initialization.
mail-citation-hook. By default, this hook's value is nil,
but by adding the following to your `.emacs' file, you can tell
these MUAs to use Supercite to perform the citing of the original
message:
(add-hook 'mail-citation-hook 'sc-cite-original)
GNUS users may also want to add the following bit of lisp as well. This prevents GNUS from inserting its default attribution header. Otherwise, both GNUS and Supercite will insert an attribution header:
(setq news-reply-header-hook nil)
Note that the mail-citation-hook interface described above was
not supported in FSF Emacs 19 until version 19.16 and in Lucid Emacs 19
until version 19.8. If you are running an earlier version of one of
these Emacsen, you will need to either upgrade to the latest version, or
use the unsupported overloading feature provided with Supercite.
See section Overloading for Non-conforming MUAs.
Note that those instructions will tell you to use the function
add-hook. This function is new with Emacs 19 and you will not
have it by default if you are running Emacs 18 or Epoch 4. You can
either substitute the appropriate call to setq, or you can use
the add-hook function that is provided in the `sc-unsupp.el'
file of unsupported Supercite hacks and ideas. Or you can upgrade to
some Emacs 19 variant! :-)
To use setq instead of add-hook, you would, for example,
change this:
(add-hook 'mail-citation-hook 'sc-cite-original)
to:
(setq mail-citation-hook 'sc-cite-original)
Note the lack of of a single quote on the first argument to setq.
mail-citation-hook interface supported
by other MUAs. At the time of this writing, MH-E 4.0 has not been
released, but if you have it, put this in your `.emacs' file to
connect Supercite and MH-E 4.x:
(add-hook 'mail-citation-hook 'sc-cite-original)
Note that if you are using Emacs 18 or Epoch 4, you will not have the
add-hook function. See section GNUS, RMAIL, PCMAIL, RNEWS with Emacs 18 or Epoch 4 for details on how to
proceed without add-hook.
MH-E version 3.x uses a slightly different interface than other MUAs.
MH-E provides a hook variable mh-yank-hooks, but it doesn't act
like a hook, and doing an add-hook will not work.
To connect Supercite to MH-E 3.x, you should instead add the following
to your .emacs file:
(add-hook 'mh-yank-hooks 'sc-cite-original)
You also need to make sure that MH-E includes all the original mail
headers in the yanked message. The variable that controls this is
mh-yank-from-start-of-msg. By default, this variable has the
value t, which tells MH-E to include all the mail headers when
yanking the original message. Before you switched to using Supercite,
you may have set this variable to other values so as not to include the
mail headers in the yanked message. Since Supercite requires these
headers (and cleans them out for you), you need to make sure the value
is t. This lisp, in your `.emacs' file will do the trick:
(setq mh-yank-from-start-of-msg t)
Note that versions of MH-E before 3.7 did not provide the
mh-yank-hooks variable. Your only option is to upgrade to MH-E
version 3.7 or later.
If you are running any release of VM after 4.40, you can add the following to your `.emacs' to connect Supercite with VM:
(add-hook 'mail-yank-hooks 'sc-cite-original)
Note that if you are using Emacs 18 or Epoch 4, you will not have the
add-hook function. See section GNUS, RMAIL, PCMAIL, RNEWS with Emacs 18 or Epoch 4 for details on how to
proceed without add-hook.
Since version 5.34, VM has supported the newer mail-citation-hook
interface, but mail-yank-hooks is still being supported for
backward compatibility. If you are running a newer version of VM and
you want to maintain consistency with other MUAs, use this bit of code
instead:
(add-hook 'mail-citation-hook 'sc-cite-original)
After you have followed the directions in that section. You should add the following lisp code to your `.emacs' file:
(add-hook 'mail-citation-hook 'sc-cite-original)
Note that if you are using Emacs 18 or Epoch 4, you will not have the
add-hook function. See section GNUS, RMAIL, PCMAIL, RNEWS with Emacs 18 or Epoch 4 for details on how to
proceed without add-hook.
mail-citation-hook
variable instead of the MUA's default hard-coded citing routines. Since
most newer versions of the known MUAs support the
mail-citation-hook variable, it is recommended that you upgrade
if at all possible. But if you can't upgrade, at least you're not out
of luck! Once you set up overloading properly, you should follow the
directions for connecting Supercite to the Emacs 19 MUAs.
See section GNUS, RMAIL, or RNEWS with any Emacs 19.
Users of Bob Weiner's Hyperbole package take note. Hyperbole provides
the necessary overloads (and a whole lot more!) and you can potentially
clobber it if you were to load Supercite's overloading after
Hyperbole's. For this reason, Supercite will not perform any
overloading if it finds the variable hyperb:version is
boundp (i.e. it exists because Hyperbole has been loaded into
your Emacs session). If this is the case, Supercite will display a
warning message in the minibuffer. You should consult the Hyperbole
manual for further details.
Overloading involves the re-definition of the citing function with the
new, mail-citation-hook savvy version. The function in
`sc-oloads.el' that does this is sc-perform-overloads. This
function is smart enough to only overload the MUA functions when it is
absolutely necessary, based on the version numbers it can figure out.
Also, sc-perform-overloads will only install the new functions
once. It is also smart enough to do nothing if the MUA is not yet
loaded.
The tricky part is finding the right time and place to perform the overloading. It must be done after the MUA has been loaded into your Emacs session, but before the first time you try to yank in a message. Fortunately, this has been figured out for you.
If you must overload, you should put the following lisp code in your `.emacs' file, to make sure the `sc-oloads.el' file gets loaded at the right time:
(autoload 'sc-perform-overloads "sc-oloads" "Supercite 3.1" t)
Then you must make sure that the function sc-perform-overloads
gets run at the right time. For GNUS, put this in your `.emacs'
file:
(setq news-reply-mode-hook 'sc-perform-overloads) (setq mail-setup-hook 'sc-perform-overloads)
If you are using RNEWS, put this in your `.emacs' file:
(setq news-reply-mode-hook 'sc-perform-overloads)
If you are using RMAIL or PCMAIL, put this in your `.emacs' file:
(setq mail-setup-hook 'sc-perform-overloads)
If you are using GNEWS, put this in your `.emacs' file:
(setq news-reply-mode-hook 'sc-perform-overloads) (setq gnews-ready-hook 'sc-perform-overloads)
Now go back and follow the directions for getting the Emacs 19 MUAs
connected to Supercite. Be sure to see section GNUS, RMAIL, PCMAIL, RNEWS with Emacs 18 or Epoch 4 on substitutes
for Emacs 19's add-hook function.
sc-cite-original performs the following steps as it
initializes the reply buffer:
sc-pre-hook.
This hook variable is run before sc-cite-original does any other
work. You could conceivably use this hook to set certain Supercite
variables based on the reply buffer's mode or name (i.e., to do
something different based on whether you are replying or following up to
an article).
sc-mode-map-prefix. By default, this variable is
set to C-c C-p; a finger twister perhaps, but unfortunately the
best default due to the scarcity of available keybindings in many MUAs.
sc-mail-warn-if-non-rfc822-p is
non-nil, then Supercite will warn you if it finds a mail header
that does not conform to RFC822. This is rare and indicates a problem
either with your MUA or the original author's MUA, or some MTA (mail
transport agent) along the way.
Once the info keys have been extracted from the mail headers, the
headers are nuked from the reply buffer. You can control exactly which
headers are removed or kept, but by default, all headers are removed.
There are two variables which control mail header nuking. The variable
sc-nuke-mail-headers controls the overall behavior of the header
nuking routines. By setting this variable to 'all, you
automatically nuke all mail headers. Likewise, setting this variable to
'none inhibits nuking of any mail headers. In between these
extremes, you can tell Supercite to nuke only a specified list of mail
headers by setting this variable to 'specified, or to keep only a
specified list of headers by setting it to 'keep.
If sc-nuke-mail-headers is set to 'specified or
'keep, then the variable sc-nuke-mail-header-list is
consulted for the list of headers to nuke or keep. This variable
contains a list of regular expressions. If the mail header line matches
a regular expression in this list, the header will be nuked or kept.
The line is matched against the regexp using looking-at rooted at
the beginning of the line.
If the variable sc-blank-lines-after-headers is non-nil,
it contains the number of blank lines remaining in the buffer after mail
headers are nuked. By default, only one blank line is left in the buffer.
sc-cite-region-limit. If this variable contains an
integer, messages with more lines that this will not be cited at all,
and a warning message will be displayed. Supercite has performed
everything necessary, though, for you to manually cite only the small
portion of the original message that you want to use.
If sc-cite-region-limit contains a non-nil value, the
original message will always be cited, regardless of its size. If the
variable contains the value nil, the region will never be cited
automatically. Use this if you always want to be able to edit and cite
the message manually.
The variable sc-cite-blank-lines-p controls whether blank lines
in the original message should be cited or not. If this variable is
non-nil, blank lines will be cited just like non-blank lines.
Otherwise, blank lines will be treated as paragraph separators.
Citing of the original message is highly configurable. Supercite's
default setup does a pretty good job of citing many common forms of
previously cited messages. But there are as many citation styles out
there as people on the net, or just about! It would be impossible for
Supercite to anticipate every style in existence, and you probably
wouldn't encounter them all anyway. But you can configure Supercite to
recognize those styles you see often.
See section Configuring the Citation Engine for details.
sc-post-hook.
This variable is very similar to sc-pre-hook, except that it runs
after sc-cite-original is finished. This hook is provided mostly
for completeness and backward compatibility. Perhaps it could be used to
reset certain variables set in sc-pre-hook.
sc-auto-fill-region-p has a
nil value. Supercite will also re-fill paragraphs when you
manually cite or re-cite text.
However, during normal editing, Supercite itself cannot be used to fill
paragraphs. This is a change from version 2. There are other add-on
lisp packages which do filling much better than Supercite ever did. The
two best known are filladapt and gin-mode. Both work well
with Supercite and both are available at the normal Emacs Lisp archive
sites. gin-mode works pretty well out of the box, but if you use
filladapt, you may want to run the function
sc-setup-filladapt from your sc-load-hook. This simply
makes filladapt a little more Supercite savvy than its default
setup.
Also, Supercite will collapse leading whitespace between the citation
string and the text on a line when the variable
sc-fixup-whitespace-p is non-nil. The default value for
this variable is nil.
Its important to understand that Supercite's automatic filling (during
the initial citation of the reply) is very fragile. That is because
figuring out the fill-prefix for a particular paragraph is a
really hard thing to do automatically. This is especially the case when
the original message contains code or some other text where leading
whitespace is important to preserve. For this reason, many Supercite
users typically run with sc-auto-fill-region-p (and possibly also
sc-fixup-whitespace-p) set to nil. They then manually
fill each cited paragraph in the reply buffer.
I usually run with both these variables containing their default values. When Supercite's automatic filling breaks on a particular message, I will use Emacs' undo feature to undo back before the citation was applied to the original message. Then I'll toggle the variables and manually cite those paragraphs that I don't want to fill or collapse whitespace on. See section Variable Toggling Shortcuts.
If you find that Supercite's automatic filling is just too fragile for
your tastes, you might consider one of these alternate approaches.
Also, to make life easier, a shortcut function to toggle the state of
both of these variables is provided on the key binding
C-c C-p C-p (with the default value of sc-mode-map-prefix;
see section Post-yank Formatting Commands).
You will noticed that the minor mode string will
show the state of these variables as qualifier characters. When both
variables are nil, the Supercite minor mode string will display
`SC'. When just sc-auto-fill-region-p is non-nil, the
string will display `SC:f', and when just
sc-fixup-whitespace-p is non-nil, the string will display
`SC:w'. When both variables are non-nil, the string will
display `SC:fw'. Note that the qualifiers chosen are mnemonics for
the default bindings of the toggling function for each respective
variable.
See section Variable Toggling Shortcuts.
Why are these variables not set to nil by default? It is because
many users won't manually fill paragraphs that are Supercited, and there
have been widespread complaints on the net about mail and news messages
containing lines greater than about 72 characters. So the default is to
fill cited text.
sc-preferred-attribution-list controls this; it contains
keys which are matched against the attribution alist in the given order.
The first value of a key that produces a non-nil, non-empty
string match is used as the attribution string, and if no keys match, a
secondary mechanism is used to generate the attribution.
See section Anonymous Attributions.
The following preferences are always available in the attribution alist (barring error):
"emailname"
"initials"
"firstname"
"lastname"
"middlename-1"
"sc-lastchoice"
"sc-consult"
sc-attrib-selection-list which can
be used to select special attributions based on the value of any info
key. See below for details.
"x-attribution"
Middle name indexes can be any positive integer greater than zero, though it is unlikely that many authors will have more than one middle name, if that many.
At this point, let me digress into a discussion of etiquette. It is my belief that while the style of the citations is a reflection of the personal tastes of the replier (i.e., you), the attribution selection is ultimately the personal choice of the original author. In a sense it is his or her "net nickname", and therefore the author should have some say in the selection of attribution string. Imagine how you would feel if someone gave you a nickname that you didn't like?
For this reason, Supercite recognizes a special mail header,
`X-Attribution:', which if present, tells Supercite the attribution
string preferred by the original author. It is the value of this header
that is associated with the "x-attribution" key in the
attribution alist. Currently, you can override the preference of this
key by changing sc-preferred-attribution-list, but that isn't
polite, and in the future Supercite may hard-code this. For now, it is
suggested that if you change the order of the keys in this list, that
"x-attribution" always be first, or possible second behind only
"sc-lastchoice". This latter is the default.
The value "sc-consult" in sc-preferred-attribution-list
has a special meaning during attribution selection. When Supercite
encounters this preference, it begins processing a customizable list of
attributions, contained in the variable sc-attrib-selection-list.
Each element in this list contains lists of the following form:
(infokey ((regexp . attribution)
(regexp . attribution)
(...)))
where infokey is a key for sc-mail-field and regexp
is a regular expression to match against the infokey's value. If
regexp matches the infokey's value, the attribution is
used as the attribution string. Actually, attribution can be a
string or a list; if it is a list, it is evaluated and the return
value (which must be a string), is used as the attribution.
This can be very useful for when you are replying to net acquaintances who do not use the `X-Attribution:' mail header. You may know what nickname they would prefer to use, and you can set up this list to match against a specific mail field, e.g., `From:', allowing you to cite your friend's message with the appropriate attribution.
sc-default-author-name and the fallback attribution string is
contained in the variable sc-default-attribution. Default values
for these variables are "Anonymous" and "Anon",
respectively. Note that in most circumstances, getting the default
author name or attribution is a sign that something is set up
incorrectly.
Also, if the preferred attribution, which you specified in your
sc-preferred-attribution-alist variable cannot be found, a
secondary method can be employed to find a valid attribution string. The
variable sc-use-only-preference-p controls what happens in this
case. If the variable's value is non-nil, then
sc-default-author-name and sc-default-attribution are
used, otherwise, the following steps are taken to find a valid
attribution string, and the first step to return a non-nil,
non-empty string becomes the attribution:
"x-attribution" key.
nil, non-empty attribution string in the
attribution alist.
sc-default-attribution is used.
Once the attribution string has been automatically selected, a number of
things can happen. If the variable sc-confirm-always-p is
non-nil, you are queried for confirmation of the chosen
attribution string. The possible values for completion are those strings
in the attribution alist, however you are not limited to these choices.
You can type any arbitrary string at the confirmation prompt. The string
you enter becomes the value associated with the "sc-lastchoice"
key in the attribution alist.
Once an attribution string has been selected, Supercite will force the
string to lower case if the variable sc-downcase-p is
non-nil.
Two hook variables provide even greater control of the attribution
selection process. The hook sc-attribs-preselect-hook is run
before any attribution is selected. Likewise, the hook
sc-attribs-postselect-hook is run after the attribution is
selected (and the corresponding citation string is built), but before
these values are committed for use by Supercite. During the
post-selection hook, the local variables attribution and
citation are bound to the appropriate strings. By changing these
variables in your hook functions, you change the attribution and
citation strings used by Supercite. One possible use of this would be
to override any automatically derived attribution string when it is only
one character long; e.g. you prefer to use "initials" but the
author only has one name.
There are a number of Supercite variables that control how author names are extracted from the `From:' header. Some headers may contain a descriptive title as in:
From: computer!speedy!doe (John Xavier-Doe -- Decent Hacker)
Supercite knows which part of the `From:' header is email address
and which part is author name, but in this case the string "Decent
Hacker" is not part of the author's name. You can tell Supercite to
ignore the title, while still recognizing hyphenated names through the
use of a regular expression in the variable sc-titlecue-regexp.
This variable has the default value of "\\\\s +-+\\\\s +". Any
text after this regexp is encountered is ignored as noise.
Some `From:' headers may contain extra titles in the name fields
not separated by a title cue, but which are nonetheless not part of the
author's name proper. Examples include the titles "Dr.", "Mr.",
"Ms.", "Jr.", "Sr.", and "III" (e.g., Thurston Howe, the Third).
Also, some companies prepend or append the name of the division,
organization, or project on the author's name. All of these titles are
noise which should be ignored. The variable sc-name-filter-alist
is used for this purpose. As implied by its name, this variable is an
association list, where each element is a cons cell of the form:
(regexp . position)
where regexp is a regular expression that is matched (using
string-match) against each element of the `From:' field's
author name. position is a position indicator, starting at zero.
Thus to strip out all titles of "Dr.", "Mr.", etc. from the name,
sc-name-filter-alist would have an entry such as:
("^\\(Mr\\|Mrs\\|Ms\\|Dr\\)[.]?$" . 0)
which only removes them if they appear as the first word in the name.
The position indicator is an integer, or one of the two special symbols
last or any. last always matches against the last
word in the name field, while any matches against every word in
the name field.
evaluated. In this way, an entire region of text
can be transformed in an awk-like manner. Regi is used
throughout Supercite, from mail header information extraction, to header
nuking, to citing text.
While the details of Regi are discussed below (see section Using Regi), only those who wish to customize certain aspects of Supercite need concern themselves with it. It is important to understand though, that any conceivable citation style that can be described by a regular expression can be recognized by Supercite. This leads to some interesting applications. For example, if you regularly recieve email from a co-worker that uses an uncommon citation style (say one that employs a `|' or `}' character at the front of the line), it is possible for Supercite to recognize this and coerce the citation to your preferred style, for consistency. In theory, it is possible for Supercite to recognize such things as uuencoded messages or C code and cite or fill those differently than normal text. None of this is currently part of Supercite, but contributions are welcome!
regi-interpret. A frame is a list of arbitrary size where each
element is a entry of the following form:
(pred func [negate-p [case-fold-search]])
Regi starts with the first entry in a frame, evaluating the pred
of that entry against the beginning of the line that `point' is on.
If the pred evaluates to true (or false if the optional
negate-p is non-nil), then the func for that entry is
evaluated. How processing continues is determined by the return
value for func, and is described below. If pred was false
the next entry in the frame is checked until all entries have been
matched against the current line. If no entry matches, `point' is
moved forward one line and the frame is reset to the first entry.
pred can be a string, a variable, a list or one of the following
symbols: t, begin, end, or every. If
pred is a string, or a variable or list that evaluates to a
string, it is interpreted as a regular expression. This regexp is
matched against the current line, from the beginning, using
looking-at. This match folds case if the optional
case-fold-search is non-nil. If pred is not a
string, or does not evaluate to a string, it is interpreted as a
binary value (nil or non-nil).
The four special symbol values for pred are recognized:
t
begin
end
every
Note that negate-p and case-fold-search are ignored if pred is one of these special symbols. Only the first occurance of each symbol in a frame is used; any duplicates are ignored. Also note that for performance reasons, the entries associated with these symbols are removed from the frame during the main interpreting loop.
Your func can return certain values which control continued Regi
processing. By default, if your func returns nil (as it
should be careful to do explicitly), Regi will reset the frame to the
first entry, and advance `point' to the beginning of the next line.
If a list is returned from your function, it can contain any combination
of the following elements:
continue
abort
end
entry is still processed.
(frame . newframe)
(step . step)
During execution of your func, the following variables will be temporarily bound to some useful information:
curline
looking-at, as a string.
curframe
curentry
sc-nuke-mail-header,
et al.
However, the citation frames Supercite uses provide a lot of customizing
power and are thus available to you to change to suit your needs. The
workhorse of citation is the frame contained in the variable
sc-default-cite-frame. This frame recognizes many situations,
such as blank lines, which it interprets as paragraph separators. It
also recognizes previously cited nested and non-nested citations in the
original message. By default it will coerce non-nested citations into
your preferred citation style, and it will add a level of citation to
nested citations. It will also simply cite uncited lines in your
preferred style.
In a similar vein, there are default frames for unciting and
reciting, contained in the variables
sc-default-uncite-frame and sc-default-recite-frame
respectively.
As mentioned earlier (see section Recognizing Citations), citations are
recognized through the values of the regular expressions
sc-citation-root-regexp, et al. To recognize odd styles, you
could modify these variables, or you could modify the default citing
frame. Alternatively, you could set up association lists of frames for
recognizing specific alternative forms.
For each of the actions -- citing, unciting, and reciting -- an alist is
consulted to find the frame to use (sc-cite-frame-alist,
sc-uncite-frame-alist, and sc-recite-frame-alist
respectively). These frames can contain alists of the form:
((infokey (regexp . frame) (regexp . frame) ...) (infokey (regexp . frame) (regexp . frame) ...) (...))
Where infokey is a key suitable for sc-mail-field,
regexp is a regular expression which is string-match'd
against the value of the sc-mail-field key, and frame is
the frame to use if a match occurred. frame can be a variable
containing a frame or a frame in-lined.
When Supercite is about to cite, uncite, or recite a region, it consults the appropriate alist and attempts to find a frame to use. If one is not found from the alist, then the appropriate default frame is used.
sc-cite-original has had a chance to do its thing, a number of
useful Supercite commands will be available to you. Since there is wide
variety in the keymaps that MUAs set up in their reply buffers, it is
next to impossible for Supercite to properly sprinkle its commands into
the existing keymap. For this reason Supercite places its commands on a
separate keymap, putting this keymap onto a prefix key in the reply
buffer. You can customize the prefix key Supercite uses by changing the
variable sc-mode-map-prefix. By default, the
sc-mode-map-prefix is C-c C-p; granted, not a great choice,
but unfortunately the best general solution so far. In the rest of this
chapter, we'll assume you've installed Supercite's keymap on the default
prefix.
sc-cite-region-limit to nil. The following commands
perform these functions on the region of text between `point' and
`mark'. Each of them sets the undo boundary before modifying
the region so that the command can be undone in the standard Emacs
way.A quick note about Emacs 19. Unlike in Emacs 18, the region delimited by `point' and `mark' can have two states. It can be active or inactive. Although the FSF Emacs 19 and Lucid Emacs 19 use different terminology and functions, both employ the same convention such that when the region is inactive, commands that modify the region should generate an error. The user needs to explicitly activate the region before successfully executing the command. All Supercite commands conform to this convention.
Here is the list of Supercite citing commands:
sc-cite-region (C-c C-p c)
sc-cite-frame-alist, or the default citing
frame sc-default-cite-frame. It runs the hook
sc-pre-cite-hook before interpreting the frame. With an optional
universal argument (C-u), it temporarily sets
sc-confirm-always-p to t so you can confirm the
attribution string for a single manual citing.
See section Configuring the Citation Engine.
sc-uncite-region (C-c C-p u)
sc-uncite-frame-alist, or the default unciting frame
sc-default-uncite-frame. It runs the hook
sc-pre-uncite-hook before interpreting the frame.
See section Configuring the Citation Engine.
sc-recite-region (C-c C-p r)
sc-recite-frame-alist, or the default reciting frame
sc-default-recite-frame. It runs the hook
sc-pre-recite-hook before interpreting the frame.
See section Configuring the Citation Engine.
Supercite will always ask you to confirm the attribution when reciting a
region, regardless of the value of sc-confirm-always-p.
sc-insert-reference (C-c C-p w)
sc-preferred-header-style is
inserted. An optional numeric argument is the index into
sc-rewrite-header-list indicating which reference header to
write.
With just the universal argument (C-u), electric reference mode is
entered, regardless of the value of sc-electric-references-p.
sc-insert-citation (C-c C-p i)
Like Supercite commands in general, the toggling commands are placed on
a keymap prefix within the greater Supercite keymap. For the default
value of sc-mode-map-prefix, this will be
C-c C-p C-t.
The following commands toggle the value of certain Supercite variables which take only a binary value:
sc-mail-nuke-blank-lines-p.
sc-confirm-always-p.
sc-downcase-p.
sc-electric-references-p.
sc-auto-fill-region-p.
sc-electric-circular-p.
sc-nested-citation-p.
sc-use-only-preferences-p.
sc-fixup-whitespace-p.
The following commands let you set the value of multi-value variables,
in the same way that Emacs' set-variable does:
sc-preferred-attribution-list.
sc-cite-region-limit.
sc-mail-nuke-mail-headers.
sc-mail-header-nuke-list.
sc-preferred-header-style.
One special command is provided to toggle both
sc-auto-fill-region-p and sc-fixup-whitespace-p together.
This is because you typically want to run Supercite with either variable
as nil or non-nil. The command to toggle these variables
together is bound on C-c C-p C-p.
Finally, the command C-c C-p C-t h (also C-c C-p C-t ?) brings up a Help message on the toggling keymap.
sc-mail-field-query (C-c C-p f)
sc-mail-process-headers (C-c C-p g)
sc-open-line (C-c C-p o)
open-line commands, but inserts the
citation string in front of the new line. As with open-line,
an optional numeric argument inserts that many new lines.
sc-describe (C-c C-p h and C-c C-p ?)
sc-version (C-c C-p v)
sc-submit-bug-report (C-c C-p C-b)
Martin> Each news/mail-reader should provide a form of Martin> mail-yank-original that Martin> 1: inserts the original message incl. header into the Martin> reply buffer; no indentation/prefixing is done, the header Martin> tends to be a "full blown" version rather than to be Martin> stripped down. Martin> 2: `point' is at the start of the header, `mark' at the Martin> end of the message body. Martin> 3: (run-hooks 'mail-yank-hooks) Martin> [Supercite] should be run as such a hook and merely Martin> rewrite the message. This way it isn't anymore Martin> [Supercite]'s job to gather the original from obscure Martin> sources. [...]
This specification was adopted, but with the recent release of FSF GNU
Emacs 19, it has undergone a slight modification. Instead of the
variable mail-yank-hooks, the new preferred hook variable that
the MUA should provide is mail-citation-hook.
mail-yank-hooks can be provided for backward compatibility, but
mail-citation-hook should always take precedence. Richard
Stallman (of the FSF) suggests that the MUAs should defvar
mail-citation-hook to nil and perform some default citing
when that is the case. Take a look at Emacs 19's `sendmail.el'
file, specifically the mail-yank-original defun for
details.
If you are writing a new MUA package, or maintaining an existing MUA package, you should make it conform to this interface so that your users will be able to link Supercite easily and seamlessly. To do this, when setting up a reply or forward buffer, your MUA should follow these steps:
mail-header-separator line and one copy below,
however there will probably be more headers below this line.
mail-citation-hook. You will probably want to
provide some kind of default citation functions in cases where the user
does not have Supercite installed. By default, your MUA should
defvar mail-citation-hook to nil, and in your
yanking function, check its value. If it finds
mail-citation-hook to be nil, it should perform some
default citing behavior. User who want to connect to Supercite then
need only add sc-cite-original to this list of hooks using
add-hook.If you do all this, your users will not need to overload your routines to use Supercite, and your MUA will join the ranks of those that conform to this interface "out of the box."
But most of this work is internal and not of very great importance to the casual user. There have been some changes at the user-visible level, but for the most part, the Supercite configuration variables from version 2 should still be relevant to version 3. Below, I briefly outline those user-visible things that have changed since version 2. For details, look to other sections of this manual.
sc-spacify-name-chars is gone in version 3.
sc-nickname-alist is gone in version 3. The
sc-attrib-selection-list is a more general construct supporting
the same basic feature.
sc-preferred-attribution has been changed
to sc-preferred-attribution-list, and has been expanded upon to
allow you to specify an ordered list of preferred attributions.
sc-mail-fields-list has been removed, and header nuking in
general has been greatly improved, giving you wider flexibility in
specifying which headers to keep and remove while presenting a
simplified interface to commonly chosen defaults.
sc-all-but-cite-p has been replaced by
sc-cite-region-limit.
In the version 2 manual I thanked some specific people for their help in
developing Supercite 2. You folks know who you are and your continued
support is greatly appreciated. I wish to thank everyone on the
Supercite mailing list, especially the brave alpha testers, who helped
considerably in testing out the concepts and implementation of Supercite
version 3. Special thanks go out to the MUA and Emacs authors Kyle
Jones, Stephen Gildea, Richard Stallman, and Jamie Zawinski for coming
to a quick agreement on the new mail-citation-hook interface, and
for adding the magic lisp to their code to support this.
All who have helped and contributed have been greatly appreciated.
Internet: supercite-request@anthem.nlm.nih.gov
UUCP: uunet!anthem.nlm.nih.gov!supercite-request
Please be sure to include the most reliable and shortest (preferably Internet) address back to you. To post articles to the list, send your message to this address (you do not need to be a member to post, but be sure to indicate this in your article or replies may not be CC'd to you):
Internet: supercite@anthem.nlm.nih.gov
UUCP: uunet!anthem.nlm.nih.gov!supercite
If you are sending bug reports, they should go to the following address,
but please! use the command sc-submit-bug-report since it
will be much easier for me to duplicate your problem if you do so. It
will set up a mail buffer automatically with this address on the
`To:' line:
Internet: supercite-help@anthem.nlm.nih.gov
UUCP: uunet!anthem.nlm.nih.gov!supercite-help
sc-", each appears under its sc-command name and
its command name.
sc-", each appears under its sc-variable name and
its variable name.