************************************************************************
* Official mandoc TODO.
* $Id$
************************************************************************

************************************************************************
* missing features
************************************************************************

- explicit blocks with missing end macro should be implicitely closed
  at the end of the enclosing block, e.g. .Bl It (El) Sh
  reminded by stsp@  in net/pptp pptp.8  Fri, 23 Apr 2010 20:32:39 +0200

- fix bad block nesting involving multiple identical explicit blocks
  see the OpenBSD mdoc_macro.c 1.47 commit message

- .Bl -column .Xo support is missing
  ultimate goal:
  restore .Xr and .Dv to
  lib/libc/compat-43/sigvec.3
  lib/libc/gen/signal.3
  lib/libc/sys/sigaction.2

- edge case: decide how to deal with blk_full bad nesting, e.g.
  .Sh .Nm .Bk .Nm .Ek .Sh found by jmc@ in ssh-keygen(1)
  from jmc@  Wed, 14 Jul 2010 18:10:32 +0100

- auto-Bk in the SYNOPSIS
  patch from kristaps@  Fri, 16 Jul 2010 14:51:24 +0200
  to be revisited after OpenBSD 4.8 tree unlock

- implement \\
  in plain text, identical to \e
  as a macro argument, identical to \ i.e. escaping the next character
  We do not have macro definitions yet; if we implement them,
  \\ must behave in a macro def like in a macro argument,
  and when using the macro, it must expand yet again.

- look at bsd.lv tbl(1)
  from kristaps@  Fri, 11 Sep 2009 17:10:53 +0200
  also look at the mail from Thomas Klausner wiz at NetBSD
    on Wed, 2 Jun 2010 11:01:29 +0200
  joerg@ has patches for this somewhere...

- look at pages generated from reStructeredText, e.g. devel/mercurial hg(1)
  These are a weird mixture of man(7) and custom autogenerated low-level
  roff stuff.  Figure out to what extent we can cope.
  noted by stsp@  Sat, 24 Apr 2010 09:17:55 +0200
  reminded by nicm@  Mon, 3 May 2010 09:52:41 +0100

- implement blank `Bl -column', such as
  .Bl -column
  .It foo Ta bar
  .El

- explicitly disallow nested `Bl -column', which would clobber internal
  flags defined for struct mdoc_macro

- set a meaningful default if no `Bl' list type is assigned

- have a blank `It' head for `Bl -tag' not puke

- prohibit `Nm' from having non-text HEAD children
  (e.g., NetBSD mDNSShared/dns-sd.1)
  (mdoc_html.c and mdoc_term.c `Nm' handlers can be slightly simplified)

- allow `Qq', `Dq', `Sq', `Aq', `Bq' to have 0 arguments
  noted by Alex Kozlov 08/06/10 23:05
  
************************************************************************
* formatting issues: ugly output
************************************************************************

- perl(1) SYNOPSIS looks bad; reported by deraadt@
  1) man(7) seems to need SYNOPSIS .Nm blocks, too

- In .Bl -column,
  .It Em Authentication<tab>Key Length
  ought to render "Key Length" with emphasis, too,
  see OpenBSD iked.conf(5).

- empty phrases in .Bl column produce too few blanks
  try e.g. .Bl -column It Ta Ta
  reported by millert Fri, 02 Apr 2010 16:13:46 -0400

- %A doesn't put an "and" before the final author name.

************************************************************************
* formatting issues: gratuitious differences
************************************************************************

- .%T should be quoted, not underlined, when .%J is also present,
  to better distinguish the contents of .%T and .%J,
  see for example OpenBSD cat(1)

- .It ${name Ns [ selector ] Ns }
  should be "${name[selector]}" not "${name [selector]}"
  This is parsed as
  text("${name") text("[") Ns() text(selector)...
  Opening punctuation should not fall out of .Ns.
  see for example OpenBSD csh(1)

- .%A should append the last author with " and " (if there are two)
  or ", and " (if there are more), not ", "
  see for example OpenBSD csh(1)

- In .Bl -bullet, the groff bullet is "+\b+\bo\bo", the mandoc bullet
  is just "o\bo".
  see for example OpenBSD ksh(1)

- .No text No ) is "text )", not "text)"
  see the terrible example
    case word in [[(]  pattern [| pattern] ... ) list ;; ] ... esac
  in OpenBSD ksh(1)

- .Sm should *not* produce as a blank line in .Bd -literal
  see for example "Brace expansion" in OpenBSD ksh(1)

- The characters "|" and "\*(Ba" should never be bold,
  not even in the middle of a word, e.g. ".Cm b\*(Bac" in
  "mknod [-m mode] name b|c major minor"
  in OpenBSD ksh(1)

- A bogus .Pp between two .It must not produce a double blank line,
  see between -R and -r in OpenBSD rm(1), before "update" in mount(8),
  or in DIAGNOSTICS in init(8).

- .Bd -literal and .Bd -unfilled are *not* identical.
  In -literal, tabs are 8 spaces.
  In -unfilled, tabs are 5 spaces, just like in -filled and -ragged.
  See the CCDF_* display in OpenBSD ccdconfig(8).

- In .Bd -unfilled, .Pp should produce one blank line, not two;
  see the ccd.conf display in OpenBSD ccdconfig(8).

- .Nx 1.0a
  should be "NetBSD 1.0A", not "NetBSD 1.0a",
  see OpenBSD ccdconfig(8).

- In .Bl -tag, if a tag exceeds the right margin and must be continued
  on the next line, it must be indented by -width, not width+1;
  see "rule block|pass" in OpenBSD ifconfig(8).

- When .%T is used outside an .Rs context and with a trailing comma,
  there is no point in rendering two commata,
  see the first paragraph of the DESCRIPTION in OpenBSD mount_nfs(8).

- When .%T is used outside an .Rs context and without a trailing comma,
  no comma should be rendered at all,
  see the first paragraph of the DESCRIPTION in OpenBSD exports(5).

- Bogus .Pp before .Bl should not cause a double blank line,
  see "The route utility provides the following simple commands:"
  in OpenBSD route(8).

************************************************************************
* performance issues
************************************************************************

Several areas can be cleaned up to make mandoc even faster.  These are 

- improve hashing mechanism for macros (quite important: performance)

- improve hashing mechanism for characters (not as important)

- the PDF file is HUGE: this can be reduced by using relative offsets

************************************************************************
* structural issues
************************************************************************

- rendering frontend code can calculate widths only for plain strings,
  not for strings containing escape sequences.  For example, this
  hinders calculation of the indent required for .Nm \&[ in text(1).
  comments from kristaps@  Wed, 21 Jul 2010 23:26:08 +0200

- another example of the same problem:
  .Bl -tag -width "\eD{format}XX" -compact
  in OpenBSD ksh(1) gives the wrong width
  because "\e" is one character in groff, two in mandoc

- Now that `ds' is minimally supported, we can get rid of some
  predefined strings.  \*(C+ has already been thrown out.  Track these
  down and whack them.  Look in e.g. gcc.1 for the top-level `ds'
  invocations.  These are reproduced across most crappy GNU manuals.