man/l_roff-char.7.in: sprinkle *** (Dave Kemper)...

[s-roff.git] / TODO

TODO reminder.

Generic
=======

- Get rid of all the short options which choose preprocessors, instead
  use a multiplexer

-- Offer a builtin mode which guesses required preprocessors, maybe add
   configuration file option which can be used to specify global
   / per-user additions by means of COMMAND-PAIR=PREPROC-PATH.
   Of course also add to the command-line multiplexer option this
   possibility..

   This "auto" preprocessor should then reconsider it's guesses on
   a per-file base!?

   I really would like to get rid of the necessity to parse all files
   with all defined preprocessors completely.  Instead we should, once
   we see the starting command, hand over to a known builtin
   preprocessor and parse only that much until we see the end command.
   So convert all builtin preprocessors to libraries and offer jumpin
   points for that purpose only, let the standalone versions be simple
   wrappers.  S-roff can use dlopen(3) if available, otherwise don't
   offer this optimization but continue to use the full-parse as usual,
   that must be supported anyway for non-builtin preprocessors.
   (XXX For [s-]troff we may not do so until the user has chosen the
   preprocessor on the command line, since..)

-- Introduce a new command that can be used to select the preprocessors
   explicitly, e.g., ".preprocessors tbl,eqn" etc.
   If so, how to deal with included files?  Offer "-" mode which turns
   off preprocessors for a specific file?

   I think the best would be some kind of "shebang", for compatibilities
   sake it must be comment-based, as in '.\"!troff eqn tbl enc=latin1'?
   For example, newer man(1)s read the first line of the manual and
   check for a syntax <^'\" >followed by concat of [egprtv]+ (and in
   fact  *join in* $MANROFFSEQ environment [egprtv]+)
                while getopts 'egprtv' preproc_arg; do
                        case "${preproc_arg}" in
                        e)      pipeline="$pipeline | $EQN" ;;
                        g)      GRAP  ;; # Ignore for compatibility.
                        p)      pipeline="$pipeline | $PIC" ;;
                        r)      pipeline="$pipeline | $REFER" ;;
                        t)      pipeline="$pipeline | $TBL" ;;
                        v)      pipeline="$pipeline | $VGRIND" ;;
                        *)      usage ;;
                        esac

   This is good, but pretty much crypto and nothing for normal users.
   But extending this, why not check for <'\" preprocess: tbl eqn"> etc.?
   And as above.
   And this is also documented in groff_tmac.man.  But do not bet on
   this.

- Be sensible in respect to locale environment a.k.a. -Tlocale and offer
  per-file input encodings, chooseable via a new command.
  If so, how to deal with included files?  Offer "-" mode which turns
  off character set conversion for a specific file?
  Anyway, extend file_case to allow "stacking in" a character
  conversion.

-- Get rid of preconv(1) again, this will simply be a layer of
   file_case.  Yet still allow to specify some encoding on the command
   line.

- What about extending searchpath:: (and file_case) so that an archive
  could be opened as such, think .dmg, .JAR or tar (pax) and/or even
  better SQLite3?

= Unicode input:
  Making groff 21bit input-clean.
-> Use S-CText and make the entire toolchain UTF-8 -based and -clean.
  This could be implemented through file_case, too, and when it finally
  is possible usage of the \[uXX] escapes could be cut down drastically.

-- This also includes hyphenation etc.

- Directly support TTF/OTF fonts.

- Offer more typographic control a.k.a paragraph-wise formatting, but in
  a way that remains compatibility with trap handling.  Is this
  possible (maybe a stripped down try-and-error forward path bouquet
  algorithm as suggested by Doug McIlroy is feasible?).

= Make -Tlj4 work with -X

= Guess man5ext and man7ext variables

= Provide man.sun implementing .TX

= Provide a `check' target

= Implement tmac.bib in terms of tmac.s

= Catch the following error in -me:
  .(z
  .(l C
  .)z

= Arrows for next/previous page from R5 xditview

dev-ps
======

Read PFB files directly.

Generate %%For comment.

Generate %%Title comment.

Angles in arc command: don't generate more digits after the decimal
point than are necessary.

Possibly generate BoundingBox comment.

Per font composite character mechanism (sufficient for fractions).

Consider whether we ought to do rounding of graphical objects other
than lines. What's the point?

Error messages should refer to output page number.

Search for downloadable fonts using their PostScript names if not
found in download file.

Separate path for searching for downloadable font files.

Clip to the BoundingBox when importing documents.

dev-tty
=======

Document font and device description file usage of grotty.

With -h avoid using a tab when a single space will do.

pre-eqn
=======

Use the same size increases for sum prod int as eqn does.

Perhaps chartype should be renamed.

TeX makes {sub,super}script on a single character with an accent
into an accent onto the (character with the script). Should we do this?

Implement mark and lineups within scripts, matrices and piles, and accents.
(Why would this be useful?)

Perhaps push hmotions down through lists to avoid upsetting spacing
adjustments.

Possibly generate .lf commands during compute_metrics phase.

Consider whether there should be extra space at the side of piles.

Provide scriptstyle displaystyle etc.

Provide a nicer matrix syntax, eg
matrix ccc {
a then b then c above
e then f then g above
h then i then k
}

Perhaps generate syntax error messages using the style of gpic.

Wide accents.

More use of \Z.

Extensible square roots.

Vphantom

Smash.

Provide a variant of vec that extends over the length of the accentee.

Support vertical arrow delimiters.

Make the following work:
.EQ
delim @@
.EN
.EQ @<-@
some equation
.EN

pre-pic
=======

In troff mode, dotted and dashed splines.

Make DELIMITED have type lstr; this would allow us to give better
error messages for problems within the body of for and if constructs.

In troff mode without -x, fake \D't' with .ps commands.

Perhaps an option to set command char.

Add an output class for dumb line printers. It wouldn't be pretty but
it would be better than nothing. Integrate it with texinfo.  Useful
for groff -Tascii as well.

Option to allow better positioning of arrowheads on arcs.

Perhaps add PostScript output mode.

Change the interface to the output class so that output devices have
the opportunity to handle arrowheads themselves.

Consider whether the line thickness should scale.

Consider whether the test in a for loop should be fuzzy (as it
apparently is in grap).

Possibly change fillval so that zero is black.

Provide a way of getting text blocks (positioned with `.in' rather
than \h), into pic.  Should be possible to use block of diverted text
in pic.  Possibly something similar to T{ and T} in tbl.

Option to provide macro backtraces.

Have a path that is searched by `copy' statement.  Set by environment
variable or command line option.

pre-refer
=========

inline references

Some sort of macro/subroutine that can cover several references.

move-punctuation should ignore multiple punctuation characters.

Make the index files machine independent.

Allow search keys to be negated (with !) to indicate that the
reference should not contain the key.  Ignore negated keys during
indexed searching.

Provide an option with lkbib and lookbib that prints the location
(filename, position) of each reference.  Need to map filename_id's
back to filenames.

Rename join-authors to join-fields.  Have a separate label-join-fields
command used by @ and #.

Have some sort of quantifier: eg $.n#A means execute `$.n' for each
instance of an A field, setting $ to that field, and then join the
results using the join-authors command.

no-text-in-bracket command which says not to allow post_text and
pre_text when the [] flags has been given. Useful for superscripted
footnotes.

Make it possible to translate - to \(en in page ranges.

Trim eign a bit.

In indexed searching discard all numeric keys except dates.

Allow `\ ' to separate article from first word.

%also

Option automatically to supply [] flags in every reference.

See if we can avoid requiring a comma before jr. and so on
in find_last_name().

Cache sortified authors in authors string during tentative evaluation of
label specification.

Possibly don't allow * and % expressions in the first part of ?:, | or
& expressions.

Handle better the case where <> occurs inside functions and in the
first operand of ~. Or perhaps implement <> using some magic character
in the string.

Should special treatment be given to lines beginning with . in
references?  (Unix refer seems to treat them like `%').

Add global flag to control whether all files should be stat-ed after
loading, and whether they should be stat-ed before each search.
Perhaps make this dependent on the number of files there are.

Option to truncate keys to truncate_len in linear searching.

Allow multiple -f options in indxbib.

In indxbib, possibly store common words rather than common words
filename.  In this case store only words that are actually present in
the file.

Perhaps we should put out an obnoxious copyright message when lookbib
starts up.

Provide an option that writes a file containing just the references
actually used.  Useful if you want to distribute a document.

Have a magic token such that
%A <sort stuff><magic token><print stuff>
will print as though it were
%A <print stuff>
but sort as though it were
%A <sort stuff>
Do we need this if we can specify author alternatives for sorting?
No, provided we have separate alternatives for @.

In consider_authors when last names are ambiguous we might be able to
use just the first name and not Jr. bit. Or we might be able to
abbreviate the author.

It ought to be possible to specify an alternative field to sort on
instead of date. (ie if there's a field giving the type of document --
these references should sort after any years)

Provide a way to execute a command using a command-line option.

Option to set the label-spec as a command-line option (-L).

Command to to specify which fields can occur multiple times:
multiple AE

Command to specify how various fields sort:
aort-as-name A
sort-as-date D
sort-as-title T
sort-as-other O

Command to specify which fields are author fields:
# if we don't have A use field Q
author-fields AQ

Commands to set properties of tokens.
sortify-token \(ae ae
uppercase-token \[ae] \[AE]

Command to set the names of months:
months january february march april may ...

Perhaps provide some sort of macro capability:
# perhaps a macro capability
defmacro foo
annotation-field $1
endef

Command to control strings used in capitalization
capitalize-start \s+2
capitalize-end \s-2
(perhaps make these arguments to the capitalize command.)

pre-soelim
==========

understand .pso.

src: troff
==========

A line prefix request to make e.g. French quotation possible:

  He said: >> blablablabla
  >> blablabla blabla  bla
  >> blabla blabla bla bla
  >> bla bla bla blablabla
  >> blabla. <<

Give a more helpful error message when the indent is set to a value
greater than the line-length.

Tracing. This is a pain to implement because requests are responsible
for reading their own arguments.

Possibly implement -s option (stop every N pages). This functionality
would be more appropriate in a postprocessor.

Line breaking should be smarter.  In particular, it should be possible
to shrink spaces.  Also avoid having a line that's been shrunk a lot
next to a line that's been stretched a lot.  The difficulty is to
design a mechanism that allows the user complete control over the
decision of where to break the line.

Provide a mechanism to control the shape of the rag in non-justified
text.

Add a discretionary break escape sequence. \='...'...'...' like TeX.

Think about kerning between characters and spaces. (Need to implement
get_breakpoints and split methods for kern_pair_node class.)

In troff, if .L > 1 when a diversion is reread in no-fill mode, then
extra line-spacing is added on. Groff at the moment treats line-spacing
like vertical spacing and doesn't do this.

Suppose \(ch comes from a special font S, and that the current font is
R. Suppose that R contains a hyphen character and that S does not.
Suppose that the current font is R. Suppose that \(ch is in a word
and has a non-zero hyphen-type. Then we ought to be able to hyphenate,
but we won't be able to because we will look for the hyphen only in
font S and not in font R.

Perhaps the current input level should be accessible in a number register.

Should \w deal with a newline like \X?

Have another look at uses of token::delimiter.  Perhaps we need to
distinguish the case where we want to see if a token could start a
number, from the case where we want to see if it could occur somewhere
in a number expression.

Provide a facility like copy thru in pic.

Fancier implementation of font families which doesn't group fonts into
families purely on the basis of their names.

In the DESC file make the number of fonts optional if they are all on
one line.

Number register to give the diversion level.

Time various alternative implementations of scale (both in font.c and
number.c). On a sparc it's faster to always do it in floating point.

Devise a more compact representation for the hyphenation patterns trie.

Have a per-environment parameter to increase letter-spacing.

Request to set character height.

Request to set character slant.

Provide some way to upcase or downcase strings.

Support non-uniformly scalable fonts. Perhaps associate a suffix with
a particular range of sizes.  eg
  sizesuffix .display 14-512
Then is you ask for R at pointsize 16, groff will first look for
R.display and then R.  Probably necessary to be able to specify a
separate unitwidth for each sizesuffix (eg. for X).

Make it possible to suppress hyphenation on a word-by-word basis.
(Perhaps store hyphenation flags in tfont.)

Possibly allow multiple simultaneous input line traps.

Unpaddable, breakable space escape sequence.

Support hanging punctuation.

In justified text, if the last line of a paragraph is only a little
bit short it might be desirable to justify the line.  Allow the user
control over this.

The pm request could print where the macro was defined.  Also could
optionally print the contents of a macro.

Provide some way to round numbers to multiples of the current
horizontal or vertical resolution.

Better string-processing support (search).

Generalized ligatures.

Request to remove an environment. (Maintain a count of the references
to the environment from the environment table, environment dictionary
or environment stack.)

Perhaps in the nr request a leading `-' should only be recognized as a
decrement when it's at the same input level as the request.

Don't ever change a charinfo.  Create new variants instead and chain
them together.

Unix troff appears to read the first character of a request name in
copy mode.  Should we do the same?

Number register giving name of end macro.

More thorough range checking.

Provide syntax for octal and hexadecimal numeric constants.  Perhaps
o#100 and x#7f as per Scheme.  Or perhaps PostScript 16#7f.  Ambiguity
between whether `c' is treated as digit or scaling indicator.

Local variables.

tmac
====

Support multiple line-spacing.

If we have footnotes in the abstract in RP format, then the footnote
will appear on the cover sheet, which it should, but also on the first
page, which it should not.

Should we allow multi-page cover-sheets?

Warn about automatically numbered footnotes in floating keeps.

When we bring back the footnote overflow at the top of page, it would
be more efficient to avoid diverting it again. (Need to keep track of
footnote height.)

Possibly have a place above which the footnote trap must not be
placed.

Improved indexing, not using tm, controlled by string variable (eg
-dIDX=file.idx).

When changing from multi-column to narrower columns, we could avoid
doing a @super-eject. (This might not be a good idea.)

Think about cutmarks.  Possibly implement CM.

Implement thesis Mode (TM, CT).

Implement more V10 features.

Should this

.LP
.rs
.sp \n(.tu

print two pages?

# s-ts-mode