Class: Metrocot::OptSpacePattern

Inherits:
BasePattern show all
Defined in:
lib/metrocot.rb

Overview

matches zero or more white spaces

Instance Attribute Summary

Attributes inherited from BasePattern

#matched, #metrocot, #name, #node_scraper, #pattern_no, #pred, #source, #succ

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from BasePattern

#default_scanner, #dump, #dump_match_map, #log, #log_match_data, #with_scanned_match_data

Constructor Details

#initialize ⇒ OptSpacePattern

Returns a new instance of OptSpacePattern.



359
360
361
 
# File 'lib/metrocot.rb', line 359

def initialize
  super(" ")
end

Class Method Details

.parse(s) ⇒ Object 



371
372
373
374
 
# File 'lib/metrocot.rb', line 371

def self.parse( s ) 
  return nil unless s[0..0] == " "
  OptSpacePattern.new
end

Instance Method Details

#description ⇒ Object 



363
364
365
 
# File 'lib/metrocot.rb', line 363

def description
  "spaces"
end

#each_match(match_range, match_map) ⇒ Object 



380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
 
# File 'lib/metrocot.rb', line 380

def each_match( match_range, match_map )

  result = nil

  super(match_range, match_map)

  match_start_index  = match_range.start_index
  match_start_offset = match_range.start_offset
  match_end_index    = match_range.start_index
  match_end_offset   = match_range.start_offset

  raise "negative range #{match_range}" if match_start_index > match_end_index
  raise "negative range #{match_range}" if match_start_index == match_end_index && match_start_offset > match_end_offset

  # consume rest of first text node  

  hnodes = match_range.hnodes
  hnode_text = nil

  if hnodes[match_start_index] && hnodes[match_start_index].text?
  
    hnode_text = hnodes[match_start_index].inner_text
    while match_end_offset < hnode_text.size && (/^\s+$/.=== hnode_text[match_start_offset .. match_end_offset])
      match_end_offset += 1
    end
      
    if match_end_offset >= match_start_offset 
      if match_end_offset >= hnode_text.size
        log( "matched entire string of #{match_end_offset - match_start_offset} spaces" )
      else
        log( "matched first #{match_end_offset - match_start_offset} leading spaces" )
      end
    end
  end

  sp_match_range = match_range.crop( match_start_index, match_start_offset, match_end_index, match_end_offset )

  result = with_scanned_match_data( match_map, hnodes[match_start_index ... match_end_index] ) { |match_map|
    yield( sp_match_range, match_map ) 
  }
  
  result
end

#optional ⇒ Object 



367
368
369
 
# File 'lib/metrocot.rb', line 367

def optional
  true
end

#priority ⇒ Object 



376
377
378
 
# File 'lib/metrocot.rb', line 376

def priority
  -8
end