mdoc.3

   1 .\"     $Id: mdoc.3,v 1.55 2011/01/07 15:07:21 kristaps Exp $
   2 .\"
   3 .\" Copyright (c) 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
   4 .\" Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
   5 .\"
   6 .\" Permission to use, copy, modify, and distribute this software for any
   7 .\" purpose with or without fee is hereby granted, provided that the above
   8 .\" copyright notice and this permission notice appear in all copies.
   9 .\"
  10 .\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  11 .\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  12 .\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
  13 .\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  14 .\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  15 .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  16 .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  17 .\"
  18 .Dd $Mdocdate: January 7 2011 $
  19 .Dt MDOC 3
  20 .Os
  21 .Sh NAME
  22 .Nm mdoc ,
  23 .Nm mdoc_alloc ,
  24 .Nm mdoc_endparse ,
  25 .Nm mdoc_free ,
  26 .Nm mdoc_meta ,
  27 .Nm mdoc_node ,
  28 .Nm mdoc_parseln ,
  29 .Nm mdoc_reset
  30 .Nd mdoc macro compiler library
  31 .Sh SYNOPSIS
  32 .In mandoc.h
  33 .In mdoc.h
  34 .Vt extern const char * const * mdoc_macronames;
  35 .Vt extern const char * const * mdoc_argnames;
  36 .Ft int
  37 .Fo mdoc_addspan
  38 .Fa "struct mdoc *mdoc"
  39 .Fa "const struct tbl_span *span"
  40 .Fc
  41 .Ft "struct mdoc *"
  42 .Fo mdoc_alloc
  43 .Fa "struct regset *regs"
  44 .Fa "void *data"
  45 .Fa "mandocmsg msgs"
  46 .Fc
  47 .Ft int
  48 .Fn mdoc_endparse "struct mdoc *mdoc"
  49 .Ft void
  50 .Fn mdoc_free "struct mdoc *mdoc"
  51 .Ft "const struct mdoc_meta *"
  52 .Fn mdoc_meta "const struct mdoc *mdoc"
  53 .Ft "const struct mdoc_node *"
  54 .Fn mdoc_node "const struct mdoc *mdoc"
  55 .Ft int
  56 .Fo mdoc_parseln
  57 .Fa "struct mdoc *mdoc"
  58 .Fa "int line"
  59 .Fa "char *buf"
  60 .Fc
  61 .Ft int
  62 .Fn mdoc_reset "struct mdoc *mdoc"
  63 .Sh DESCRIPTION
  64 The
  65 .Nm mdoc
  66 library parses lines of
  67 .Xr mdoc 7
  68 input
  69 into an abstract syntax tree (AST).
  70 .Pp
  71 In general, applications initiate a parsing sequence with
  72 .Fn mdoc_alloc ,
  73 parse each line in a document with
  74 .Fn mdoc_parseln ,
  75 close the parsing session with
  76 .Fn mdoc_endparse ,
  77 operate over the syntax tree returned by
  78 .Fn mdoc_node
  79 and
  80 .Fn mdoc_meta ,
  81 then free all allocated memory with
  82 .Fn mdoc_free .
  83 The
  84 .Fn mdoc_reset
  85 function may be used in order to reset the parser for another input
  86 sequence.
  87 .Ss Types
  88 .Bl -ohang
  89 .It Vt struct mdoc
  90 An opaque type.
  91 Its values are only used privately within the library.
  92 .It Vt struct mdoc_node
  93 A parsed node.
  94 See
  95 .Sx Abstract Syntax Tree
  96 for details.
  97 .El
  98 .Ss Functions
  99 If
 100 .Fn mdoc_addspan ,
 101 .Fn mdoc_parseln ,
 102 or
 103 .Fn mdoc_endparse
 104 return 0, calls to any function but
 105 .Fn mdoc_reset
 106 or
 107 .Fn mdoc_free
 108 will raise an assertion.
 109 .Bl -ohang
 110 .It Fn mdoc_addspan
 111 Add a table span to the parsing stream.
 112 Returns 0 on failure, 1 on success.
 113 .It Fn mdoc_alloc
 114 Allocates a parsing structure.
 115 The
 116 .Fa data
 117 pointer is passed to
 118 .Fa msgs .
 119 Always returns a valid pointer.
 120 The pointer must be freed with
 121 .Fn mdoc_free .
 122 .It Fn mdoc_reset
 123 Reset the parser for another parse routine.
 124 After its use,
 125 .Fn mdoc_parseln
 126 behaves as if invoked for the first time.
 127 If it returns 0, memory could not be allocated.
 128 .It Fn mdoc_free
 129 Free all resources of a parser.
 130 The pointer is no longer valid after invocation.
 131 .It Fn mdoc_parseln
 132 Parse a nil-terminated line of input.
 133 This line should not contain the trailing newline.
 134 Returns 0 on failure, 1 on success.
 135 The input buffer
 136 .Fa buf
 137 is modified by this function.
 138 .It Fn mdoc_endparse
 139 Signals that the parse is complete.
 140 Returns 0 on failure, 1 on success.
 141 .It Fn mdoc_node
 142 Returns the first node of the parse.
 143 .It Fn mdoc_meta
 144 Returns the document's parsed meta-data.
 145 .El
 146 .Ss Variables
 147 .Bl -ohang
 148 .It Va mdoc_macronames
 149 An array of string-ified token names.
 150 .It Va mdoc_argnames
 151 An array of string-ified token argument names.
 152 .El
 153 .Ss Abstract Syntax Tree
 154 The
 155 .Nm
 156 functions produce an abstract syntax tree (AST) describing input in a
 157 regular form.
 158 It may be reviewed at any time with
 159 .Fn mdoc_nodes ;
 160 however, if called before
 161 .Fn mdoc_endparse ,
 162 or after
 163 .Fn mdoc_endparse
 164 or
 165 .Fn mdoc_parseln
 166 fail, it may be incomplete.
 167 .Pp
 168 This AST is governed by the ontological
 169 rules dictated in
 170 .Xr mdoc 7
 171 and derives its terminology accordingly.
 172 .Qq In-line
 173 elements described in
 174 .Xr mdoc 7
 175 are described simply as
 176 .Qq elements .
 177 .Pp
 178 The AST is composed of
 179 .Vt struct mdoc_node
 180 nodes with block, head, body, element, root and text types as declared
 181 by the
 182 .Va type
 183 field.
 184 Each node also provides its parse point (the
 185 .Va line ,
 186 .Va sec ,
 187 and
 188 .Va pos
 189 fields), its position in the tree (the
 190 .Va parent ,
 191 .Va child ,
 192 .Va nchild ,
 193 .Va next
 194 and
 195 .Va prev
 196 fields) and some type-specific data, in particular, for nodes generated
 197 from macros, the generating macro in the
 198 .Va tok
 199 field.
 200 .Pp
 201 The tree itself is arranged according to the following normal form,
 202 where capitalised non-terminals represent nodes.
 203 .Pp
 204 .Bl -tag -width "ELEMENTXX" -compact
 205 .It ROOT
 206 \(<- mnode+
 207 .It mnode
 208 \(<- BLOCK | ELEMENT | TEXT
 209 .It BLOCK
 210 \(<- HEAD [TEXT] (BODY [TEXT])+ [TAIL [TEXT]]
 211 .It ELEMENT
 212 \(<- TEXT*
 213 .It HEAD
 214 \(<- mnode*
 215 .It BODY
 216 \(<- mnode* [ENDBODY mnode*]
 217 .It TAIL
 218 \(<- mnode*
 219 .It TEXT
 220 \(<- [[:printable:],0x1e]*
 221 .El
 222 .Pp
 223 Of note are the TEXT nodes following the HEAD, BODY and TAIL nodes of
 224 the BLOCK production: these refer to punctuation marks.
 225 Furthermore, although a TEXT node will generally have a non-zero-length
 226 string, in the specific case of
 227 .Sq \&.Bd \-literal ,
 228 an empty line will produce a zero-length string.
 229 Multiple body parts are only found in invocations of
 230 .Sq \&Bl \-column ,
 231 where a new body introduces a new phrase.
 232 .Ss Badly-nested Blocks
 233 The ENDBODY node is available to end the formatting associated
 234 with a given block before the physical end of that block.
 235 It has a non-null
 236 .Va end
 237 field, is of the BODY
 238 .Va type ,
 239 has the same
 240 .Va tok
 241 as the BLOCK it is ending, and has a
 242 .Va pending
 243 field pointing to that BLOCK's BODY node.
 244 It is an indirect child of that BODY node
 245 and has no children of its own.
 246 .Pp
 247 An ENDBODY node is generated when a block ends while one of its child
 248 blocks is still open, like in the following example:
 249 .Bd -literal -offset indent
 250 \&.Ao ao
 251 \&.Bo bo ac
 252 \&.Ac bc
 253 \&.Bc end
 254 .Ed
 255 .Pp
 256 This example results in the following block structure:
 257 .Bd -literal -offset indent
 258 BLOCK Ao
 259         HEAD Ao
 260         BODY Ao
 261                 TEXT ao
 262                 BLOCK Bo, pending -> Ao
 263                         HEAD Bo
 264                         BODY Bo
 265                                 TEXT bo
 266                                 TEXT ac
 267                                 ENDBODY Ao, pending -> Ao
 268                                 TEXT bc
 269 TEXT end
 270 .Ed
 271 .Pp
 272 Here, the formatting of the
 273 .Sq \&Ao
 274 block extends from TEXT ao to TEXT ac,
 275 while the formatting of the
 276 .Sq \&Bo
 277 block extends from TEXT bo to TEXT bc.
 278 It renders as follows in
 279 .Fl T Ns Cm ascii
 280 mode:
 281 .Pp
 282 .Dl <ao [bo ac> bc] end
 283 .Pp
 284 Support for badly-nested blocks is only provided for backward
 285 compatibility with some older
 286 .Xr mdoc 7
 287 implementations.
 288 Using badly-nested blocks is
 289 .Em strongly discouraged :
 290 the
 291 .Fl T Ns Cm html
 292 and
 293 .Fl T Ns Cm xhtml
 294 front-ends are unable to render them in any meaningful way.
 295 Furthermore, behaviour when encountering badly-nested blocks is not
 296 consistent across troff implementations, especially when using  multiple
 297 levels of badly-nested blocks.
 298 .Sh EXAMPLES
 299 The following example reads lines from stdin and parses them, operating
 300 on the finished parse tree with
 301 .Fn parsed .
 302 This example does not error-check nor free memory upon failure.
 303 .Bd -literal -offset indent
 304 struct regset regs;
 305 struct mdoc *mdoc;
 306 const struct mdoc_node *node;
 307 char *buf;
 308 size_t len;
 309 int line;
 310
 311 bzero(&regs, sizeof(struct regset));
 312 line = 1;
 313 mdoc = mdoc_alloc(&regs, NULL, NULL);
 314 buf = NULL;
 315 alloc_len = 0;
 316
 317 while ((len = getline(&buf, &alloc_len, stdin)) >= 0) {
 318     if (len && buflen[len - 1] = '\en')
 319         buf[len - 1] = '\e0';
 320     if ( ! mdoc_parseln(mdoc, line, buf))
 321         errx(1, "mdoc_parseln");
 322     line++;
 323 }
 324
 325 if ( ! mdoc_endparse(mdoc))
 326     errx(1, "mdoc_endparse");
 327 if (NULL == (node = mdoc_node(mdoc)))
 328     errx(1, "mdoc_node");
 329
 330 parsed(mdoc, node);
 331 mdoc_free(mdoc);
 332 .Ed
 333 .Pp
 334 To compile this, execute
 335 .Pp
 336 .Dl % cc main.c libmdoc.a libmandoc.a
 337 .Pp
 338 where
 339 .Pa main.c
 340 is the example file.
 341 .Sh SEE ALSO
 342 .Xr mandoc 1 ,
 343 .Xr mdoc 7
 344 .Sh AUTHORS
 345 The
 346 .Nm
 347 library was written by
 348 .An Kristaps Dzonsons Aq kristaps@bsd.lv .