TODO

   1 ************************************************************************
   2 * Official mandoc TODO.
   3 * $Id: TODO,v 1.98 2011/03/30 00:02:40 schwarze Exp $
   4 ************************************************************************
   5
   6 ************************************************************************
   7 * parser bugs
   8 ************************************************************************
   9
  10 - .TP before .SH is still FATAL in man(7)
  11   reported by brad@  Sat, 15 Jan 2011 15:54:54 -0500
  12
  13 - the roff parser doesn't tolerate additional characters between
  14   a macro and the \} terminating a conditional block, e.g.
  15   .if n \{
  16   .br \}
  17   reported by ulrich spoerlein  Tue, 19 Oct 2010 20:39:50 +0200
  18
  19 ************************************************************************
  20 * formatter bugs
  21 ************************************************************************
  22
  23 ************************************************************************
  24 * missing features
  25 ************************************************************************
  26
  27 --- missing roff features ----------------------------------------------
  28
  29 - .ad (adjust margins)
  30   .ad l -- adjust left margin only (flush left)
  31   .ad r -- adjust right margin only (flush right)
  32   .ad c -- center text on line
  33   .ad b -- adjust both margins (alias: .ad n)
  34   .na   -- temporarily disable adjustment without changing the mode
  35   .ad   -- re-enable adjustment without changing the mode
  36   Adjustment mode is ignored while in no-fill mode (.nf).
  37
  38 - .it (line traps) occur in mysql(1), yasm_arch(7)
  39   generated by DocBook XSL Stylesheets v1.71.1 <http://docbook.sf.net/>
  40   reported by brad@  Sat, 15 Jan 2011 15:48:18 -0500
  41
  42 - .ns (no-space mode) occurs in xine-config(1)
  43   reported by brad@  Sat, 15 Jan 2011 15:45:23 -0500
  44
  45 - xloadimage(1) wants .ti (temporary indent), rep by naddy@
  46
  47 - .ta (tab settings) occurs in ircbug(1) and probably gnats(1)
  48   reported by brad@  Sat, 15 Jan 2011 15:50:51 -0500
  49
  50 - pod2man expects `tr' to be implemented for \*(-- to work
  51
  52 - \c (interrupted text) occurs in chat(8)
  53
  54 - \f(CW (constant width font) occurs in rsyncd.conf(5).
  55   I think we should treat unknown/unavailable fonts as \fR
  56   such that switching back with \fP works correctly -
  57   and doesn't revert the _previous_ \fP.
  58
  59 --- missing mdoc features ----------------------------------------------
  60
  61 - fix bad block nesting involving multiple identical explicit blocks
  62   see the OpenBSD mdoc_macro.c 1.47 commit message
  63
  64 - .Bl -column .Xo support is missing
  65   ultimate goal:
  66   restore .Xr and .Dv to
  67   lib/libc/compat-43/sigvec.3
  68   lib/libc/gen/signal.3
  69   lib/libc/sys/sigaction.2
  70
  71 - edge case: decide how to deal with blk_full bad nesting, e.g.
  72   .Sh .Nm .Bk .Nm .Ek .Sh found by jmc@ in ssh-keygen(1)
  73   from jmc@  Wed, 14 Jul 2010 18:10:32 +0100
  74
  75 - \\ is now implemented correctly
  76   * when defining strings and macros using .ds and .de
  77   * when parsing roff(7) and man(7) macro arguments
  78   It does not yet work in mdoc(7) macro arguments
  79   because libmdoc does not yet use mandoc_getarg().
  80   Also check what happens in plain text, it must be identical to \e.
  81
  82 - .Bd -filled should not be the same as .Bd -ragged, but align both
  83   the left and right margin.  In groff, it is implemented in terms
  84   of .ad b, which we don't have either.  Found in cksum(1).
  85
  86 - implement blank `Bl -column', such as
  87   .Bl -column
  88   .It foo Ta bar
  89   .El
  90
  91 - explicitly disallow nested `Bl -column', which would clobber internal
  92   flags defined for struct mdoc_macro
  93
  94 - inside `.Bl -column' phrases, punctuation is handled like normal
  95   text, e.g. `.Bl -column .It Fl x . Ta ...' should give "-x -."
  96
  97 - inside `.Bl -column' phrases, TERMP_IGNDELIM handling by `Pf'
  98   is not safe, e.g. `.Bl -column .It Pf a b .' gives "ab."
  99   but should give "ab ."
 100
 101 - set a meaningful default if no `Bl' list type is assigned
 102
 103 - have a blank `It' head for `Bl -tag' not puke
 104
 105 - prohibit `Nm' from having non-text HEAD children
 106   (e.g., NetBSD mDNSShared/dns-sd.1)
 107   (mdoc_html.c and mdoc_term.c `Nm' handlers can be slightly simplified)
 108
 109 --- missing man features -----------------------------------------------
 110
 111 - groff an-ext.tmac macros (.UR, .UE) occur in xine(5)
 112   reported by brad@  Sat, 15 Jan 2011 15:45:23 -0500
 113
 114 --- missing tbl features -----------------------------------------------
 115
 116 - implement basic non-parametric .de to support e.g. sox(1)
 117   reported by naddy@ Sat, 16 Oct 2010 23:51:57 +0200
 118   *** sox(1) still doesn't work, tbl(1) errors need investigation
 119
 120 --- missing misc features ----------------------------------------------
 121
 122 - clean up escape sequence handling, creating three classes:
 123   (1) fully implemented, or parsed and ignored without loss of content
 124   (2) unimplemented, potentially causing loss of content
 125       or serious mangling of formatting (e.g. \n) -> ERROR
 126       see textproc/mgdiff(1) for nice examples
 127   (3) undefined, just output the character -> perhaps WARNING
 128
 129 - The \t escape sequence is the same as a literal tab, see for example
 130   the ASCII table in hexdump(1) where
 131     .Bl -column \&000_nu \&001_so \&002_st \&003_et \&004_eo
 132     .It \&000\ nul\t001\ soh\t002\ stx\t003\ etx\t004\ eot\t005\ enq
 133   produces
 134     000 nul  001 soh  002 stx  003 etx  004 eot  005 enq
 135   and the example in oldrdist(1)
 136
 137 - look at pages generated from reStructeredText, e.g. devel/mercurial hg(1)
 138   These are a weird mixture of man(7) and custom autogenerated low-level
 139   roff stuff.  Figure out to what extent we can cope.
 140   For details, see http://docutils.sourceforge.net/rst.html
 141   noted by stsp@  Sat, 24 Apr 2010 09:17:55 +0200
 142   reminded by nicm@  Mon, 3 May 2010 09:52:41 +0100
 143
 144 - check compatibility with Plan9:
 145   http://swtch.com/usr/local/plan9/tmac/tmac.an
 146   http://swtch.com/plan9port/man/man7/man.html
 147   "Anthony J. Bentley" <anthonyjbentley@gmail.com> 28 Dec 2010 21:58:40 -0700
 148
 149 ************************************************************************
 150 * formatting issues: ugly output
 151 ************************************************************************
 152
 153 - a column list with blank `Ta' cells triggers a spurrious
 154   start-with-whitespace printing of a newline
 155
 156 - double quotes inside double quotes are escaped by doubling them
 157   implement this in mdoc(7), too
 158   so far, we only have it in roff(7) and man(7)
 159   reminded by millert@  Thu, 09 Dec 2010 17:29:52 -0500
 160
 161 - perl(1) SYNOPSIS looks bad; reported by deraadt@
 162   1) man(7) seems to need SYNOPSIS .Nm blocks, too
 163
 164 - In .Bl -column,
 165   .It Em Authentication<tab>Key Length
 166   ought to render "Key Length" with emphasis, too,
 167   see OpenBSD iked.conf(5).
 168
 169 - empty phrases in .Bl column produce too few blanks
 170   try e.g. .Bl -column It Ta Ta
 171   reported by millert Fri, 02 Apr 2010 16:13:46 -0400
 172
 173 - .%T can have trailing punctuation.  Currently, it puts the trailing
 174   punctuation into a trailing MDOC_TEXT element inside its own scope.
 175   That element should rather be outside its scope, such that the
 176   punctuation does not get underlines.  This is not trivial to
 177   implement because .%T then needs some features of in_line_eoln() -
 178   slurp all arguments into one single text element - and one feature
 179   of in_line() - put trailing punctuation out of scope.
 180   Found in mount_nfs(8) and exports(5), search for "Appendix".
 181
 182 - in enclosures, mandoc sometimes fancies a bogus end of sentence
 183   reminded by jmc@  Thu, 23 Sep 2010 18:13:39 +0059
 184
 185 ************************************************************************
 186 * formatting issues: gratuitious differences
 187 ************************************************************************
 188
 189 - .Rv (and probably .Ex) print different text if an `Nm' has been named
 190   or not (run a manual without `Nm blah' to see this).  I'm not sure
 191   that this exists in the wild, but it's still an error.
 192
 193 - In .Bl -bullet, the groff bullet is "+\b+\bo\bo", the mandoc bullet
 194   is just "o\bo".
 195   see for example OpenBSD ksh(1)
 196
 197 - The characters "|" and "\*(Ba" should never be bold,
 198   not even in the middle of a word, e.g. ".Cm b\*(Bac" in
 199   "mknod [-m mode] name b|c major minor"
 200   in OpenBSD ksh(1)
 201
 202 - A bogus .Pp between two .It must not produce a double blank line,
 203   see between -R and -r in OpenBSD rm(1), before "update" in mount(8),
 204   or in DIAGNOSTICS in init(8), or before "is always true" in ksh(1).
 205   The same happens with .Pp just before .El, see bgpd.conf(5).
 206   Also have `It' complain if `Pp' is invoked at certain times (not
 207   -compact?).
 208
 209 - If the *first* line after .It is .Pp, break the line right after
 210   the tag, do not pad with space characters before breaking.
 211   See the description of the a, c, and i commands in sed(1).
 212
 213 - If the first line after .It is .D1, do not assert a blank line
 214   in between, see for example tmux(1).
 215   reported by nicm@  13 Jan 2011 00:18:57 +0000
 216
 217 - .Nx 1.0a
 218   should be "NetBSD 1.0A", not "NetBSD 1.0a",
 219   see OpenBSD ccdconfig(8).
 220
 221 - In .Bl -tag, if a tag exceeds the right margin and must be continued
 222   on the next line, it must be indented by -width, not width+1;
 223   see "rule block|pass" in OpenBSD ifconfig(8).
 224
 225 - When the -width string contains macros, the macros must be rendered
 226   before measuring the width, for example
 227     .Bl -tag -width ".Dv message"
 228   in magic(5), located in src/usr.bin/file, is the same
 229   as -width 7n, not -width 11n.
 230
 231 - The \& zero-width character counts as output.
 232   That is, when it is alone on a line between two .Pp,
 233   we want three blank lines, not two as in mandoc.
 234
 235 - When .Fn arguments exceed one output line, all but the first
 236   should be indented, see e.g. rpc(3);
 237   reported by jmc@ on discuss@  Fri, 29 Oct 2010 13:48:33 +0100
 238
 239 - .Ns should work when called at the end of an input line, see
 240   the following code in vi(1):
 241     .It Xo
 242     .Op Ar line
 243     .Cm a Ns Op Cm ppend Ns
 244     .Op Cm !\&
 245     .Xc
 246     The input text is appended after the specified line.
 247
 248 - Header lines of excessive length:
 249   Port OpenBSD man_term.c rev. 1.25 to mdoc_term.c
 250   and document it in mdoc(7) and man(7) COMPATIBILITY
 251   found while talking to Chris Bennett
 252
 253 - In man(7), the sequence
 254     .HP
 255     one line of regular text
 256     .SH
 257   should not produce two blank lines before the .SH,
 258   see for example named-checkconf(8).
 259
 260 - In man(7), the sequence
 261     regular text
 262     .RS
 263     .IP
 264   should have a blank line between the text and the beginning of the
 265   indented paragraph, see for example sudo(1).
 266
 267 - In man(7), the sequence
 268     .SH HEADER
 269     <blank line>
 270     .PP
 271     regular text
 272   should not produce any blank lines between the header and the text,
 273   see for example rsync(1).
 274   Reported by naddy@  Mon, 28 Mar 2011 20:45:42 +0200
 275
 276 - In man(7), the sequence
 277     regular text
 278     .PP
 279     .RS
 280     indented text
 281   should produce one blank line between the regular and indented texts,
 282   see for example rsync(1), and
 283     .RE
 284     <blank line>
 285     .PP
 286     .RS
 287   should produce two,not one blank lines.
 288   Reported by naddy@  Mon, 28 Mar 2011 20:45:42 +0200
 289
 290 - In man(7), the sequence
 291     regular text
 292     .IP
 293     .IP "tag"
 294     indented text
 295   should produce one, not four blank lines between the regular text
 296   and the tag, see for example rsync(1).
 297   Likewise,
 298     regular text
 299     .IP
 300     indented text
 301   should produce one, not two blank lines in between, and
 302     regular text
 303     .IP
 304     .RS
 305     .IP tag
 306     indented text
 307   should produce one, not three blank lines.
 308   Reported by naddy@  Mon, 28 Mar 2011 20:45:42 +0200
 309
 310 - At least sometimes, .sp is ignored right after .SH; see named.conf(5).
 311
 312 - trailing whitespace must be ignored even when followed by a font escape,
 313   see for example
 314     makes
 315     \fBdig \fR
 316     operate in batch mode
 317   in dig(1).
 318
 319 ************************************************************************
 320 * error reporting issues
 321 ************************************************************************
 322
 323 ************************************************************************
 324 * performance issues
 325 ************************************************************************
 326
 327 Several areas can be cleaned up to make mandoc even faster.  These are
 328
 329 - improve hashing mechanism for macros (quite important: performance)
 330
 331 - improve hashing mechanism for characters (not as important)
 332
 333 - the PDF file is HUGE: this can be reduced by using relative offsets
 334
 335 ************************************************************************
 336 * structural issues
 337 ************************************************************************
 338
 339 - Find better ways to prevent endless loops
 340   in roff(7) macro and string expansion.
 341
 342 - Finish cleanup of date handling.
 343   Decide which formats should be recognized where.
 344   Update both mdoc(7) and man(7) documentation.
 345   Triggered by  Tim van der Molen  Tue, 22 Feb 2011 20:30:45 +0100