Class: Xlsxrb::Ooxml::WorksheetParser

Inherits:
Object
  • Object
show all
Defined in:
lib/xlsxrb/ooxml/worksheet_parser.rb,
sig/generated/xlsxrb/ooxml/worksheet_parser.rbs

Overview

Streaming parser for xl/worksheets/sheetN.xml. Uses a fast string-scanning approach for row/cell extraction, falling back to REXML SAX only for column definitions and unmapped data.

Defined Under Namespace

Classes: ColumnsListener

Constant Summary collapse

EMPTY_ARRAY =

Returns:

  • (Object) 
[].freeze
EMPTY_HASH =

Returns:

  • (Object) 
{}.freeze
XML_ENTITIES =

Returns:

  • (Object) 
{ "&amp;" => "&", "&lt;" => "<", "&gt;" => ">", "&quot;" => '"', "&apos;" => "'" }.freeze

Class Method Summary collapse

Class Method Details

.each_event(xml_string, shared_strings: [], part_name: "xl/worksheets/sheet1.xml", &block) ⇒ void

This method returns an undefined value.

Yields Event objects for columns, rows, cells, and hyperlinks.

Parameters:

  • xml_string (Object)
  • shared_strings: (Object) (defaults to: [])
  • part_name: (Object) (defaults to: "xl/worksheets/sheet1.xml") 


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
 
# File 'lib/xlsxrb/ooxml/worksheet_parser.rb', line 57

def self.each_event(xml_string, shared_strings: [], part_name: "xl/worksheets/sheet1.xml", &block)
  return enum_for(:each_event, xml_string, shared_strings: shared_strings, part_name: part_name) unless block
  return if xml_string.nil? || xml_string.empty?

  # 1. Parse and yield column events
  if xml_string.include?("<cols")
    listener = ColumnsListener.new
    XmlParser.parse(xml_string, listener)
    listener.columns.each do |col|
      block.call(Event.new(
                   type: :column,
                   args: [col[:min], col[:max], col[:width], col[:hidden], col[:custom_width], col[:outline_level]],
                   source: { part: part_name }
                 ))
    end
  end

  # 2. Parse and yield row/cell events
  fast_scan_events(xml_string, shared_strings, part_name, &block)

  # 3. Parse and yield hyperlink events
  return unless xml_string.include?("hyperlink")

  xml = xml_string.b
  hpos_term = xml.index("hyperlinks")
  return unless hpos_term

  hpos = xml.rindex("<", hpos_term)
  return unless hpos

  h_end_term = xml.index("/hyperlinks", hpos)
  h_end = h_end_term ? xml.index(">", h_end_term) : xml.size
  h_end ||= xml.size

  pos = hpos
  while pos < h_end
    hl_start_term = xml.index("hyperlink", pos)
    break unless hl_start_term && hl_start_term < h_end

    hl_start = xml.rindex("<", hl_start_term)
    break unless hl_start && hl_start < h_end

    hl_tag_end = xml.index(">", hl_start + 1)
    break unless hl_tag_end

    hl_tag = xml.byteslice(hl_start, hl_tag_end - hl_start)
    ref = tag_attr(hl_tag, ' ref="')
    rid = tag_attr(hl_tag, ' r:id="') || tag_attr(hl_tag, ' id="')
    display = tag_attr(hl_tag, ' display="')
    tooltip = tag_attr(hl_tag, ' tooltip="')
    location = tag_attr(hl_tag, ' location="')

    if ref
      block.call(Event.new(
                   type: :hyperlink,
                   args: [ref, rid, display, tooltip, location],
                   source: { part: part_name, cell: ref }
                 ))
    end
    pos = hl_tag_end + 1
  end
end

.each_row(xml_string, shared_strings: [], part_name: "xl/worksheets/sheet1.xml", &block) ⇒ void

This method returns an undefined value.

Streaming parse: yields one raw row hash at a time.

Parameters:

  • xml_string (Object)
  • shared_strings: (Object) (defaults to: [])
  • part_name: (Object) (defaults to: "xl/worksheets/sheet1.xml") 


28
29
30
31
32
33
 
# File 'lib/xlsxrb/ooxml/worksheet_parser.rb', line 28

def self.each_row(xml_string, shared_strings: [], part_name: "xl/worksheets/sheet1.xml", &block)
  return enum_for(:each_row, xml_string, shared_strings: shared_strings, part_name: part_name) unless block
  return if xml_string.nil? || xml_string.empty?

  fast_scan_rows_direct(xml_string, shared_strings, part_name, &block)
end

.fast_scan_cells_direct(xml, from, to, shared_strings, row_source, &block) ⇒ void

This method returns an undefined value.

Parameters:

  • xml (Object)
  • from (Object)
  • to (Object)
  • shared_strings (Object)
  • row_source (Object) 


544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
 
# File 'lib/xlsxrb/ooxml/worksheet_parser.rb', line 544

def self.fast_scan_cells_direct(xml, from, to, shared_strings, row_source, &block)
  pos = from
  cell_count = 0

  while pos < to
    c_start = xml.index("<c", pos)
    break unless c_start && c_start < to

    nb = xml.getbyte(c_start + 2)
    unless [32, 62, 9, 10, 13, 47].include?(nb)
      pos = c_start + 2
      next
    end

    c_tag_end = xml.index(">", c_start + 2)
    break unless c_tag_end

    # Zero-allocation attribute scan directly within the tag
    type = nil
    style_index = nil
    col_idx = nil
    row_idx = nil

    ai = c_start + 2
    while ai < c_tag_end
      b = xml.getbyte(ai)
      if [32, 9, 10, 13].include?(b)
        ai += 1
        next
      end

      if b == 114 && xml.getbyte(ai + 1) == 61 && xml.getbyte(ai + 2) == 34 # r="
        ai += 3
        col_idx = 0
        while ai < c_tag_end
          cb = xml.getbyte(ai)
          if cb.between?(65, 90)
            col_idx = (col_idx * 26) + (cb - 64)
            ai += 1
          elsif cb.between?(97, 122)
            col_idx = (col_idx * 26) + (cb - 96)
            ai += 1
          else
            break
          end
        end
        col_idx -= 1
        row_idx = 0
        while ai < c_tag_end
          cb = xml.getbyte(ai)
          break unless cb.between?(48, 57)

          row_idx = (row_idx * 10) + (cb - 48)
          ai += 1
        end
        row_idx -= 1
        ai += 1 if xml.getbyte(ai) == 34
      elsif b == 116 && xml.getbyte(ai + 1) == 61 && xml.getbyte(ai + 2) == 34 # t="
        ai += 3
        type_b = xml.getbyte(ai)
        if type_b == 115 && xml.getbyte(ai + 1) == 34 # "s"
          type = "s"
          ai += 2
        elsif type_b == 98 && xml.getbyte(ai + 1) == 34 # "b"
          type = "b"
          ai += 2
        elsif type_b == 101 && xml.getbyte(ai + 1) == 34 # "e"
          type = "e"
          ai += 2
        else
          t_end = xml.index('"', ai)
          type = xml.byteslice(ai, t_end - ai) if t_end
          ai = t_end ? t_end + 1 : c_tag_end
        end
      elsif b == 115 && xml.getbyte(ai + 1) == 61 && xml.getbyte(ai + 2) == 34 # s="
        ai += 3
        style_index = 0
        while ai < c_tag_end
          cb = xml.getbyte(ai)
          break unless cb.between?(48, 57)

          style_index = (style_index * 10) + (cb - 48)
          ai += 1
        end
        ai += 1 if xml.getbyte(ai) == 34
      else
        ai += 1
      end
    end

    # Self-closing <c ... />
    if xml.getbyte(c_tag_end - 1) == 47
      cell_obj = Elements::Cell.new(
        row_index: row_idx || row_source[:row],
        column_index: col_idx || cell_count,
        value: nil,
        style_index: style_index,
        errors: Elements::EMPTY_ERRORS
      )
      cell_count += 1
      block.call(cell_obj)
      pos = c_tag_end + 1
      next
    end

    c_end = xml.index("</c>", c_tag_end + 1)
    break unless c_end

    # Parse cell content sequentially (bounded to c_end - avoid unbounded scans)
    value = nil
    formula = nil
    inline_str = nil
    cpos = c_tag_end + 1
    while cpos < c_end
      tag_pos = xml.index("<", cpos)
      break unless tag_pos && tag_pos < c_end

      tag_char = xml.getbyte(tag_pos + 1)
      case tag_char
      when 118 # 'v'
        if xml.getbyte(tag_pos + 2) == 62 # <v>
          v_val_start = tag_pos + 3
          v_end = xml.index("</v>", v_val_start)
          if v_end
            if type == "s"
              # Zero-allocation integer parse for SST index
              s_idx = 0
              v_i = v_val_start
              while v_i < v_end
                s_idx = (s_idx * 10) + (xml.getbyte(v_i) - 48)
                v_i += 1
              end
              value = shared_strings[s_idx] || ""
            elsif type == "b"
              value = xml.getbyte(v_val_start) == 49 # '1'
            else
              raw_value = xml.byteslice(v_val_start, v_end - v_val_start)
              value = resolve_fast_value(raw_value, type, shared_strings)
            end
            cpos = v_end + 4
          else
            cpos = tag_pos + 3
          end
        elsif xml.getbyte(tag_pos + 2) == 47 && xml.getbyte(tag_pos + 3) == 62 # <v/>
          cpos = tag_pos + 4
        else
          cpos = tag_pos + 2
        end
      when 102 # 'f'
        f_tag_end = xml.index(">", tag_pos + 2)
        if f_tag_end && f_tag_end < c_end
          if xml.getbyte(f_tag_end - 1) == 47 # self-closing <f ... />
            cpos = f_tag_end + 1
          else
            f_end = xml.index("</f>", f_tag_end + 1)
            if f_end && f_end <= c_end
              formula = xml.byteslice(f_tag_end + 1, f_end - f_tag_end - 1).force_encoding("UTF-8")
              formula = decode_xml_entities(formula) if formula.include?("&")
              cpos = f_end + 4
            else
              cpos = f_tag_end + 1
            end
          end
        else
          cpos = tag_pos + 2
        end
      when 105 # 'i' - <is>
        if xml.byteslice(tag_pos, 4) == "<is>"
          is_end = xml.index("</is>", tag_pos + 4)
          if is_end && is_end <= c_end
            inline_str = extract_inline_text(xml, tag_pos + 4, is_end)
            cpos = is_end + 5
          else
            cpos = tag_pos + 4
          end
        else
          cpos = tag_pos + 2
        end
      else
        # Skip unknown tag
        close = xml.index(">", tag_pos + 1)
        cpos = close ? close + 1 : c_end
      end
    end

    val_to_use = inline_str || value
    cell_obj = Elements::Cell.new(
      row_index: row_idx || row_source[:row],
      column_index: col_idx || cell_count,
      value: val_to_use,
      formula: formula,
      style_index: style_index,
      errors: Elements::EMPTY_ERRORS
    )
    cell_count += 1
    block.call(cell_obj)

    pos = c_end + 4
  end
end

.parse(xml_string, shared_strings: []) ⇒ Object

Parses all rows from a worksheet XML string. Returns an Array of raw row hashes:

{ index:, cells: [{ ref:, type:, style_index:, value:, formula: }], attrs:, unmapped: }

Parameters:

  • xml_string (Object)
  • shared_strings: (Object) (defaults to: [])

Returns:

  • (Object) 


19
20
21
22
23
24
25
 
# File 'lib/xlsxrb/ooxml/worksheet_parser.rb', line 19

def self.parse(xml_string, shared_strings: [])
  return [] if xml_string.nil? || xml_string.empty?

  rows = []
  each_row(xml_string, shared_strings: shared_strings) { |row| rows << row }
  rows
end

.parse_columns(xml_string, part_name: "xl/worksheets/sheet1.xml") ⇒ Object

Parses column definitions () from a worksheet.

Parameters:

  • xml_string (Object)
  • part_name: (Object) (defaults to: "xl/worksheets/sheet1.xml")

Returns:

  • (Object) 


36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
 
# File 'lib/xlsxrb/ooxml/worksheet_parser.rb', line 36

def self.parse_columns(xml_string, part_name: "xl/worksheets/sheet1.xml")
  return [] if xml_string.nil? || xml_string.empty?

  columns = []
  each_event(xml_string, part_name: part_name) do |event|
    next unless event.type == :column

    min, max, width, hidden, custom_width, outline_level = event.args
    columns << {
      min: min,
      max: max,
      width: width,
      hidden: hidden,
      custom_width: custom_width,
      outline_level: outline_level
    }
  end
  columns
end