about summary refs log tree commit diff
path: root/Completion/Base/_regex_arguments
blob: 598911bc559b61fedb89ebaf8c4d5f06222c667d (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
#autoload

## usage: _regex_arguments funcname regex

# _regex_arguments compiles `regex' and emit the result of the state
# machine into the function `funcname'. `funcname' parses a command line
# according to `regex' and evaluate appropriate actions in `regex'. Before
# parsing the command line string is genereted by concatinating `words'
# (before `PREFIX') and `PREFIX' with a separator NUL ($'\0').

# The `regex' is defined as follows.

## regex word definition:

# elt-pattern = "/" ( pattern | "[]" )	# cutoff
#	      | "%" pattern		# non-cutoff
# lookahead = "@" pattern
# parse-action = "-" zsh-code-to-eval
# complete-action = "!" zsh-code-to-eval

## regex word sequence definition:

# element = elt-pattern [ lookahead ] [ parse-action ] [ complete-action ]
#
# regex = element 
#	| "(" regex ")"
#	| regex "#"
#	| regex regex
#	| regex "|" regex
#	| void
#	| null
#
# NOTE: void and null has no explicit representation. However null can
# be represent with empty words such as \( \).

# example: (in zsh quoted form)

# $'[^\0]#\0' \#	: zero or more words

## auxiliary functions definition:

# fst : a * b -> a
# snd : a * b -> b
# fst( (x, y) ) = x
# snd( (x, y) ) = y

# nullable : regex -> bool
# first : regex -> list of element
# match : string * list of element -> element + {bottom}
# right : string * element -> string
# left : string * element -> string
# next : regex * element -> regex + {bottom}
# trans : string * string * regex -> (string * string * regex) + {bottom}

# nullable(void) = false
# nullable(null) = true
# nullable(e) = false
# nullable(r #) = true
# nullable(r1 r2) = nullable(r1) and nullable(r2)
# nullable(r1 | r2) = nullable(r1) or nullable(r2)

# first(void) = {}
# first(null) = {}
# first(e) = [ e ]
# first(r #) = first(r)
# first(r1 r2) = nullable(r1) ? first(r1) ++ first(r2) : first(r1)
# first(r1 | r2) = first(r1) ++ first(r2)

# match(s, []) = bottom
# match(s, [e1, e2, ...]) = e	if [[ $s = $elt-pattern[e]$lookahead[e]* ]]
#		   	  | match(s, [e2, ...])	otherwise

# right(s, e) = ${s##$elt-pattern[e]}
# left(s, e) = ${(M)s##$elt-pattern[e]}

### XXX: It can treat lookaheads if zsh provide $1, $2, ... in perl.

# next(void, e) = bottom
# next(null, e) = bottom
# next(e1, e0) = e1 eq e0 ? null : bottom	# eq is test operator of identity equality.
# next(r #, e) = next(r, e) != bottom ? next(r, e) (r #) : bottom
# next(r1 r2, e) = next(r1, e) != bottom ? next(r1, e) r2 : next(r2, e)
# next(r1 | r2, e) = next(r1, e) != bottom ? next(r1, e) : next(r2, e)

# trans( (t, s, r) ) = ( (cutoff(e) ? '' : t ++ left(s, e)), right(s, e), next(r, e) )
#   where e = match(s, first(r))

# NOTE: This `next' definition is slightly different to ordinaly one.
# This definition uses only one element of first(r) for transition
# instead of all elements of first(r).

# If _regex_arguments takes the regex r0, the first state of the state
# machine is r0.  The state of the state machine transit as follows.

# ('', s0, r0) -> trans('', s0, r0) = (t1, s1, r1) -> trans(t1, s1, r1) -> ... 

# If the state is reached to bottom, the state transition is stopped.

# ... -> (tN, sN, rN) -> bottom

# For each transitions (tI, sI, rI) to trans(tI, sI, rI), the state
# machine evaluate parse-action bound to match(sI, first(rI)).

# In parse-action bound to match(sI, first(rI)) = e, it can refer variables:
#  _ra_left : tI+1
#  _ra_match : left(sI, e)
#  _ra_right : sI+1

# If the state transition is stopped, the state machine evaluate
# complete-actions bound to first(rN) if tN and sN does not contain NUL.
# When complete-actions are evaluated, completion focus is restricted to
# tN ++ sN. (This is reason of tN and sN cannot contain NUL when
# completion.)
# Also, if there are last transitions that does not cut off the string
# (tJ ++ sJ = tJ+1 ++ sJ+1 = ... = tN-1 ++ sN-1 = tN ++ sN),
# complete-actions bound to them
# --- match(sJ, first(rJ)), ..., match(sN-1, first(rN-1)) --- are also
# evaluated before complete-actions bound to first(rN).

# example:

# compdef _tst tst

# _regex_arguments _tst /$'[^\0]#\0' /$'[^\0]#\0' '!compadd aaa'
#  _tst complete `aaa' for first argument.
#  First $'[^\0]#\0' is required to match with command name.

# _regex_arguments _tst /$'[^\0]#\0' \( /$'[^\0]#\0' '!compadd aaa' /$'[^\0]#\0' !'compadd bbb' \) \#
#  _tst complete `aaa' for (2i+1)th argument and `bbb' for (2i)th argument.

# _regex_arguments _tst /$'[^\0]#\0' \( /$'[^\0]#\0' '!compadd aaa' \| /$'[^\0]#\0' !'compadd bbb' \) \#
#  _tst complete `aaa' or `bbb'.

## Recursive decent regex parser

_ra_parse_elt () {
  : index=$index "[$regex[$index]]"
  local state
  if (( $#regex < index )); then
    return 1
  else
    case "$regex[index]" in
      [/%]*) state=$index
          first=($state)
	  last=($state)
	  nullable=
	  case "${regex[index][1]}" in
	    /) cutoff[$state]=yes ;;
	    %) cutoff[$state]= ;;
	  esac
          pattern[$state]="${regex[index++][2,-1]}"
	  [[ -n "$pattern[$state]" ]] && pattern[$state]="($pattern[$state])"
	  if [[ $index -le $#regex && $regex[index] = @* ]]; then
	    lookahead[$state]="${regex[index++][2,-1]}"
	    [[ -n "$lookahead[$state]" ]] && lookahead[$state]="($lookahead[$state])"
	  else
	    lookahead[$state]=""
	  fi
	  if [[ $index -le $#regex && $regex[index] = -* ]]; then
	    parse_action[$state]="${regex[index++][2,-1]}"
	  else
	    parse_action[$state]=""
	  fi
	  if [[ $index -le $#regex && $regex[index] = \!* ]]; then
	    complete_action[$state]="${regex[index++][2,-1]}"
	  else
	    complete_action[$state]=""
	  fi
	  ;;
      \() (( index++ ))
          _ra_parse_alt || return 1
	  [[ $index -le $#regex && "$regex[$index]" = \) ]] || return 1
	  (( index++ ))
	  ;;
      *)  return 1
          ;;
    esac
  fi

  return 0
}

_ra_parse_clo () {
  : index=$index "[$regex[$index]]"
  _ra_parse_elt || return 1

  if (( index <= $#regex )) && [[ "$regex[$index]" = \# ]]; then
    (( index++ ))
    nullable=yes

    for i in $last; do tbl[$i]="$tbl[$i] $first"; done
  fi

  return 0
}

_ra_parse_seq () {
  : index=$index "[$regex[$index]]"
  local last_seq
  local first_seq nullable_seq
  first_seq=()
  nullable_seq=yes

  _ra_parse_clo || {
    first=()
    last=()
    nullable=yes
    return 0
  }
  first_seq=($first)
  last_seq=($last)
  [[ -n "$nullable" ]] || nullable_seq=

  while :; do
    _ra_parse_clo || break
    for i in $last_seq; do tbl[$i]="${tbl[$i]} $first"; done
    [[ -n "$nullable_seq" ]] && first_seq=($first_seq $first)
    [[ -n "$nullable" ]] || { nullable_seq= last_seq=() }
    last_seq=($last_seq $last)
  done

  first=($first_seq)
  nullable=$nullable_seq
  last=($last_seq)
  return 0
}

_ra_parse_alt () {
  : index=$index "[$regex[$index]]"
  local last_alt
  local first_alt nullable_alt 
  first_alt=()
  nullable_alt=

  _ra_parse_seq || return 1
  first_alt=($first_alt $first)
  last_alt=($last_alt $last)
  [[ -n "$nullable" ]] && nullable_alt=yes

  while :; do
    (( index <= $#regex )) || break
    [[ "$regex[$index]" = \| ]] || break
    (( index++ ))

    _ra_parse_seq || break
    first_alt=($first_alt $first)
    last_alt=($last_alt $last)
    [[ -n "$nullable" ]] && nullable_alt=yes
  done

  first=($first_alt)
  last=($last_alt)
  nullable=$nullable_alt
  return 0
}

## function generator

_ra_gen_func () {
  local old new
  local state next index
  local start="${(j/:/)first}"

  old=()
  new=($start)

  print -lr - \
    "$funcname () {" \
      'setopt localoptions extendedglob' \
      'local _ra_state _ra_left _ra_match _ra_right _ra_actions _ra_tmp' \
      "_ra_state='$start'" \
      '_ra_left=' \
      '_ra_right="${(pj:\0:)${(@)words[1,CURRENT - 1]:Q}}"$'\''\0'\''"$PREFIX"' \
      '_ra_actions=()' \
      'while :; do' \
	'case "$_ra_state" in'

  while (( $#new )); do
    state="$new[1]"
    shift new
    old=("$old[@]" "$state")

    print -lr - \
	"$state)" \
	  'case "$_ra_right" in'

    for index in ${(s/:/)state}; do
      if [[ "$pattern[$index]" != "([])" ]]; then
	next="${(j/:/)${(@)=tbl[$index]}}"
	print -lr - \
	      "$pattern[$index]$lookahead[$index]*)"
	if [[ -n "$pattern[$index]" ]]; then
	  if [[ -n "$cutoff[$index]" ]]; then
	    print -lr - \
		  '_ra_match="${(M)_ra_right##'"$pattern[$index]"'}"' \
		  '_ra_right="$_ra_right[$#_ra_match + 1, -1]"' \
		  '_ra_left=' \
		  'if (( $#_ra_match )); then' \
		    '_ra_actions=()'
	    if [[ -n "${complete_action[$index]:q}" ]]; then
	      print -lr - \
		  'else' \
		    '_ra_actions=("$_ra_actions[@]" '"${complete_action[$index]:q}"')'
	    fi
	    print -lr - \
		  'fi'
	  else
	    print -lr - \
		  '_ra_match="${(M)_ra_right##'"$pattern[$index]"'}"' \
		  '_ra_right="$_ra_right[$#_ra_match + 1, -1]"' \
		  '_ra_left="$_ra_left$_ra_match"'
	    if [[ -n "${complete_action[$index]:q}" ]]; then
	      print -lr - \
		  '_ra_actions=("$_ra_actions[@]" '"${complete_action[$index]:q}"')'
	    fi
	  fi
	else
	  print -lr - \
		'_ra_match=' \
		'_ra_actions=("$_ra_actions[@]" '"${complete_action[$index]:q}"')'
	fi
	print -lr - \
		"$parse_action[$index]"
	if [[ -n $next ]]; then
	  print -lr - \
		"_ra_state=$next"
	  (( $old[(I)$next] || $new[(I)$next] )) || new=($next "$new[@]")
	else
	  print -lr - \
		'_message "no arg"' \
		'break'
	fi
	print -lr - \
		';;'
      fi
    done

    print -lr - \
	    '*)' \
	      'if [[ "$_ra_left$_ra_right" = *$'\''\0'\''* ]]; then' \
		'_message "parse failed before current word"' \
	      'else' \
		'compset -p $(( $#PREFIX - $#_ra_right - $#_ra_left ))'

    print -lr - \
		'for _ra_tmp in $_ra_actions; do' \
		  'eval "$_ra_tmp"' \
		'done'
    for index in ${(s/:/)state}; do
      print -lr - \
		"$complete_action[$index]"
    done

    print -lr - \
	      'fi' \
	      'break' \
	      ';;' \
	  'esac' \
	  ';;'
  done

  print -lr - \
	'esac' \
      'done' \
    '}'
}

_regex_arguments () {
  setopt localoptions extendedglob

  local funcname="_regex_arguments_tmp"
  local funcdef

  typeset -A tbl cutoff pattern lookahead parse_action complete_action
  local regex index first last nullable
  local i state next

  funcname="$1"
  shift

  regex=("$@")
  index=1
  tbl=()
  pattern=()
  lookahead=()
  parse_action=()
  complete_action=()
  _ra_parse_alt

  funcdef="$(_ra_gen_func)"

  unfunction "$funcname" 2>/dev/null
  eval "${(F)funcdef}"
}

_regex_arguments "$@"