mdoc.h

   1 /* $Id: mdoc.h,v 1.37 2009/03/06 14:13:47 kristaps Exp $ */
   2 /*
   3  * Copyright (c) 2008 Kristaps Dzonsons <kristaps@kth.se>
   4  *
   5  * Permission to use, copy, modify, and distribute this software for any
   6  * purpose with or without fee is hereby granted, provided that the
   7  * above copyright notice and this permission notice appear in all
   8  * copies.
   9  *
  10  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
  11  * WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED
  12  * WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE
  13  * AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
  14  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
  15  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
  16  * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
  17  * PERFORMANCE OF THIS SOFTWARE.
  18  */
  19 #ifndef MDOC_H
  20 #define MDOC_H
  21
  22 /*
  23  * This library implements a validating scanner/parser for ``mdoc'' roff
  24  * macro documents, a.k.a. BSD manual page documents.  The mdoc.c file
  25  * drives the parser, while macro.c describes the macro ontologies.
  26  * validate.c pre- and post-validates parsed macros, and action.c
  27  * performs actions on parsed and validated macros.
  28  */
  29
  30 /* What follows is a list of ALL possible macros. */
  31
  32 #define MDOC___          0
  33 #define MDOC_Dd          1
  34 #define MDOC_Dt          2
  35 #define MDOC_Os          3
  36 #define MDOC_Sh          4
  37 #define MDOC_Ss          5
  38 #define MDOC_Pp          6
  39 #define MDOC_D1          7
  40 #define MDOC_Dl          8
  41 #define MDOC_Bd          9
  42 #define MDOC_Ed          10
  43 #define MDOC_Bl          11
  44 #define MDOC_El          12
  45 #define MDOC_It          13
  46 #define MDOC_Ad          14
  47 #define MDOC_An          15
  48 #define MDOC_Ar          16
  49 #define MDOC_Cd          17
  50 #define MDOC_Cm          18
  51 #define MDOC_Dv          19
  52 #define MDOC_Er          20
  53 #define MDOC_Ev          21
  54 #define MDOC_Ex          22
  55 #define MDOC_Fa          23
  56 #define MDOC_Fd          24
  57 #define MDOC_Fl          25
  58 #define MDOC_Fn          26
  59 #define MDOC_Ft          27
  60 #define MDOC_Ic          28
  61 #define MDOC_In          29
  62 #define MDOC_Li          30
  63 #define MDOC_Nd          31
  64 #define MDOC_Nm          32
  65 #define MDOC_Op          33
  66 #define MDOC_Ot          34
  67 #define MDOC_Pa          35
  68 #define MDOC_Rv          36
  69 #define MDOC_St          37
  70 #define MDOC_Va          38
  71 #define MDOC_Vt          39
  72 #define MDOC_Xr          40
  73 #define MDOC__A          41
  74 #define MDOC__B          42
  75 #define MDOC__D          43
  76 #define MDOC__I          44
  77 #define MDOC__J          45
  78 #define MDOC__N          46
  79 #define MDOC__O          47
  80 #define MDOC__P          48
  81 #define MDOC__R          49
  82 #define MDOC__T          50
  83 #define MDOC__V          51
  84 #define MDOC_Ac          52
  85 #define MDOC_Ao          53
  86 #define MDOC_Aq          54
  87 #define MDOC_At          55
  88 #define MDOC_Bc          56
  89 #define MDOC_Bf          57
  90 #define MDOC_Bo          58
  91 #define MDOC_Bq          59
  92 #define MDOC_Bsx         60
  93 #define MDOC_Bx          61
  94 #define MDOC_Db          62
  95 #define MDOC_Dc          63
  96 #define MDOC_Do          64
  97 #define MDOC_Dq          65
  98 #define MDOC_Ec          66
  99 #define MDOC_Ef          67
 100 #define MDOC_Em          68
 101 #define MDOC_Eo          69
 102 #define MDOC_Fx          70
 103 #define MDOC_Ms          71
 104 #define MDOC_No          72
 105 #define MDOC_Ns          73
 106 #define MDOC_Nx          74
 107 #define MDOC_Ox          75
 108 #define MDOC_Pc          76
 109 #define MDOC_Pf          77
 110 #define MDOC_Po          78
 111 #define MDOC_Pq          79
 112 #define MDOC_Qc          80
 113 #define MDOC_Ql          81
 114 #define MDOC_Qo          82
 115 #define MDOC_Qq          83
 116 #define MDOC_Re          84
 117 #define MDOC_Rs          85
 118 #define MDOC_Sc          86
 119 #define MDOC_So          87
 120 #define MDOC_Sq          88
 121 #define MDOC_Sm          89
 122 #define MDOC_Sx          90
 123 #define MDOC_Sy          91
 124 #define MDOC_Tn          92
 125 #define MDOC_Ux          93
 126 #define MDOC_Xc          94
 127 #define MDOC_Xo          95
 128 #define MDOC_Fo          96
 129 #define MDOC_Fc          97
 130 #define MDOC_Oo          98
 131 #define MDOC_Oc          99
 132 #define MDOC_Bk          100
 133 #define MDOC_Ek          101
 134 #define MDOC_Bt          102
 135 #define MDOC_Hf          103
 136 #define MDOC_Fr          104
 137 #define MDOC_Ud          105
 138 #define MDOC_Lb          106
 139 #define MDOC_MAX         107
 140
 141 /* What follows is a list of ALL possible macro arguments. */
 142
 143 #define MDOC_Split       0
 144 #define MDOC_Nosplit     1
 145 #define MDOC_Ragged      2
 146 #define MDOC_Unfilled    3
 147 #define MDOC_Literal     4
 148 #define MDOC_File        5
 149 #define MDOC_Offset      6
 150 #define MDOC_Bullet      7
 151 #define MDOC_Dash        8
 152 #define MDOC_Hyphen      9
 153 #define MDOC_Item        10
 154 #define MDOC_Enum        11
 155 #define MDOC_Tag         12
 156 #define MDOC_Diag        13
 157 #define MDOC_Hang        14
 158 #define MDOC_Ohang       15
 159 #define MDOC_Inset       16
 160 #define MDOC_Column      17
 161 #define MDOC_Width       18
 162 #define MDOC_Compact     19
 163 #define MDOC_Std         20
 164 #define MDOC_Filled      21
 165 #define MDOC_Words       22
 166 #define MDOC_Emphasis    23
 167 #define MDOC_Symbolic    24
 168 #define MDOC_ARG_MAX     25
 169
 170 /* Warnings are either syntax or groff-compatibility. */
 171 enum    mdoc_warn {
 172         WARN_SYNTAX,
 173         WARN_COMPAT
 174 };
 175
 176 /* An argument to a macro (multiple values = `It -column'). */
 177 struct  mdoc_arg {
 178         int               arg;
 179         int               line;
 180         int               pos;
 181         size_t            sz;
 182         char            **value;
 183 };
 184
 185 /* Type of a syntax node. */
 186 enum    mdoc_type {
 187         MDOC_TEXT,
 188         MDOC_ELEM,
 189         MDOC_HEAD,
 190         MDOC_TAIL,
 191         MDOC_BODY,
 192         MDOC_BLOCK,
 193         MDOC_ROOT
 194 };
 195
 196 /* Section (named/unnamed) of `Sh'. */
 197 enum    mdoc_sec {
 198         SEC_PROLOGUE            = 0,
 199         SEC_BODY                = 1,
 200         SEC_NAME                = 2,
 201         SEC_LIBRARY             = 3,
 202         SEC_SYNOPSIS            = 4,
 203         SEC_DESCRIPTION         = 5,
 204         SEC_IMPLEMENTATION      = 6,
 205         SEC_RETURN_VALUES       = 7,
 206         SEC_ENVIRONMENT         = 8,
 207         SEC_FILES               = 9,
 208         SEC_EXAMPLES            = 10,
 209         SEC_DIAGNOSTICS         = 11,
 210         SEC_COMPATIBILITY       = 12,
 211         SEC_ERRORS              = 13,
 212         SEC_SEE_ALSO            = 14,
 213         SEC_STANDARDS           = 15,
 214         SEC_HISTORY             = 16,
 215         SEC_AUTHORS             = 17,
 216         SEC_CAVEATS             = 18,
 217         SEC_BUGS                = 19,
 218         SEC_CUSTOM
 219 };
 220
 221 /* Information from prologue. */
 222 struct  mdoc_meta {
 223         int               msec;
 224         char             *vol;
 225         char             *arch;
 226         time_t            date;
 227         char             *title;
 228         char             *os;
 229         char             *name;
 230 };
 231
 232 /* Text-only node. */
 233 struct  mdoc_text {
 234         char             *string;
 235 };
 236
 237 /* Block (scoped) node. */
 238 struct  mdoc_block {
 239         size_t            argc;
 240         struct mdoc_arg  *argv;
 241         struct mdoc_node *head;
 242         struct mdoc_node *body;
 243         struct mdoc_node *tail;
 244 };
 245
 246 /* In-line element node. */
 247 struct  mdoc_elem {
 248         size_t            argc;
 249         struct mdoc_arg  *argv;
 250 };
 251
 252 /* Typed nodes of an AST node. */
 253 union   mdoc_data {
 254         struct mdoc_text  text;
 255         struct mdoc_elem  elem;
 256         struct mdoc_block block;
 257 };
 258
 259 /* Node in AST. */
 260 struct  mdoc_node {
 261         struct mdoc_node *parent;
 262         struct mdoc_node *child;
 263         struct mdoc_node *next;
 264         struct mdoc_node *prev;
 265         int               line;
 266         int               pos;
 267         int               tok;
 268         int               flags;
 269 #define MDOC_VALID       (1 << 0)
 270 #define MDOC_ACTED       (1 << 1)
 271         enum mdoc_type    type;
 272         union mdoc_data   data;
 273         enum mdoc_sec     sec;
 274 };
 275
 276 /* Call-backs for parse messages. */
 277 struct  mdoc_cb {
 278         void    (*mdoc_msg)(void *, int, int, const char *);
 279         int     (*mdoc_err)(void *, int, int, const char *);
 280         int     (*mdoc_warn)(void *, int, int,
 281                         enum mdoc_warn, const char *);
 282 };
 283
 284 /* Global table of macro names (`Bd', `Ed', etc.). */
 285 extern  const char *const *mdoc_macronames;
 286
 287 /* Global table of argument names (`column', `tag', etc.). */
 288 extern  const char *const *mdoc_argnames;
 289
 290 __BEGIN_DECLS
 291
 292 struct  mdoc;
 293
 294 /* Free memory allocated with mdoc_alloc. */
 295 void              mdoc_free(struct mdoc *);
 296
 297 /* Allocate a new parser instance. */
 298 struct  mdoc     *mdoc_alloc(void *data, const struct mdoc_cb *);
 299
 300 /* Parse a single line in a stream (boolean retval). */
 301 int               mdoc_parseln(struct mdoc *, int, char *buf);
 302
 303 /* Get result first node (after mdoc_endparse!). */
 304 const struct mdoc_node *mdoc_node(const struct mdoc *);
 305
 306 /* Get result meta-information (after mdoc_endparse!). */
 307 const struct mdoc_meta *mdoc_meta(const struct mdoc *);
 308
 309 /* Signal end of parse sequence (boolean retval). */
 310 int               mdoc_endparse(struct mdoc *);
 311
 312 /* The following are utility functions. */
 313
 314 const char       *mdoc_a2att(const char *);
 315 const char       *mdoc_a2lib(const char *);
 316 const char       *mdoc_a2st(const char *);
 317
 318 int               mdoc_isdelim(const char *);
 319
 320 __END_DECLS
 321
 322 #endif /*!MDOC_H*/