StringScanner

Class StringScanner supports processing a stored string as a stream; this code creates a new StringScanner object with string 'foobarbaz':

require 'strscan'
scanner = StringScanner.new('foobarbaz')

About the Examples

All examples here assume that StringScanner has been required:

require 'strscan'

Some examples here assume that these constants are defined:

MULTILINE_TEXT = <<~EOT
Go placidly amid the noise and haste,
and remember what peace there may be in silence.
EOT

HIRAGANA_TEXT = 'こんにちは'

ENGLISH_TEXT = 'Hello'

Some examples here assume that certain helper methods are defined:

• put_situation(scanner): Displays the values of the scanner’s methods pos, charpos, rest, and rest_size.

• put_match_values(scanner): Displays the scanner’s match values.

• match_values_cleared?(scanner): Returns whether the scanner’s match values are cleared.

See examples at helper methods.

The StringScanner Object

This code creates a StringScanner object (we’ll call it simply a scanner), and shows some of its basic properties:

scanner = StringScanner.new('foobarbaz')
scanner.string # => "foobarbaz"
put_situation(scanner)
# Situation:
#   pos:       0
#   charpos:   0
#   rest:      "foobarbaz"
#   rest_size: 9

The scanner has:

• A stored string, which is:

  • Initially set by StringScanner.new(string) to the given string ('foobarbaz' in the example above).

  • Modifiable by methods string=(new_string) and concat(more_string).

  • Returned by method string.

  More at Stored String below.

• A position; a zero-based index into the bytes of the stored string (not into its characters):

  • Initially set by StringScanner.new to 0.

  • Returned by method pos.

  • Modifiable explicitly by methods reset, terminate, and pos=(new_pos).

  • Modifiable implicitly (various traversing methods, among others).

  More at Byte Position below.

• A target substring, which is a trailing substring of the stored string; it extends from the current position to the end of the stored string:

  • Initially set by StringScanner.new(string) to the given string ('foobarbaz' in the example above).

  • Returned by method rest.

  • Modified by any modification to either the stored string or the position.

  Most importantly: the searching and traversing methods operate on the target substring, which may be (and often is) less than the entire stored string.

  More at Target Substring below.

Stored String

The stored string is the string stored in the StringScanner object.

Each of these methods sets, modifies, or returns the stored string:

Positions

A StringScanner object maintains a zero-based byte position and a zero-based character position.

Each of these methods explicitly sets positions:

Byte Position (Position)

The byte position (or simply position) is a zero-based index into the bytes in the scanner’s stored string; for a new StringScanner object, the byte position is zero.

When the byte position is:

• Zero (at the beginning), the target substring is the entire stored string.

• Equal to the size of the stored string (at the end), the target substring is the empty string ''.

To get or set the byte position:

• pos: returns the byte position.

• pos=(new_pos): sets the byte position.

Many methods use the byte position as the basis for finding matches; many others set, increment, or decrement the byte position:

scanner = StringScanner.new('foobar')
scanner.pos # => 0
scanner.scan(/foo/) # => "foo" # Match found.
scanner.pos         # => 3     # Byte position incremented.
scanner.scan(/foo/) # => nil   # Match not found.
scanner.pos # => 3             # Byte position not changed.

Some methods implicitly modify the byte position; see:

• Setting the Target Substring.

• Traversing the Target Substring.

The values of these methods are derived directly from the values of pos and string:

• charpos: the character position.

• rest: the target substring.

• rest_size: rest.size.

Character Position

The character position is a zero-based index into the characters in the stored string; for a new StringScanner object, the character position is zero.

Method charpos returns the character position; its value may not be reset explicitly.

Some methods change (increment or reset) the character position; see:

• Setting the Target Substring.

• Traversing the Target Substring.

Example (string includes multi-byte characters):

scanner = StringScanner.new(ENGLISH_TEXT) # Five 1-byte characters.
scanner.concat(HIRAGANA_TEXT)             # Five 3-byte characters
scanner.string # => "Helloこんにちは"       # Twenty bytes in all.
put_situation(scanner)
# Situation:
#   pos:       0
#   charpos:   0
#   rest:      "Helloこんにちは"
#   rest_size: 20
scanner.scan(/Hello/) # => "Hello" # Five 1-byte characters.
put_situation(scanner)
# Situation:
#   pos:       5
#   charpos:   5
#   rest:      "こんにちは"
#   rest_size: 15
scanner.getch         # => "こ"    # One 3-byte character.
put_situation(scanner)
# Situation:
#   pos:       8
#   charpos:   6
#   rest:      "んにちは"
#   rest_size: 12

Target Substring

The target substring is the part of the stored string that extends from the current byte position to the end of the stored string; it is always either:

• The entire stored string (byte position is zero).

• A trailing substring of the stored string (byte position positive).

The target substring is returned by method rest, and its size is returned by method rest_size.

Examples:

scanner = StringScanner.new('foobarbaz')
put_situation(scanner)
# Situation:
#   pos:       0
#   charpos:   0
#   rest:      "foobarbaz"
#   rest_size: 9
scanner.pos = 3
put_situation(scanner)
# Situation:
#   pos:       3
#   charpos:   3
#   rest:      "barbaz"
#   rest_size: 6
scanner.pos = 9
put_situation(scanner)
# Situation:
#   pos:       9
#   charpos:   9
#   rest:      ""
#   rest_size: 0

Setting the Target Substring

The target substring is set whenever:

• The stored string is set (position reset to zero; target substring set to stored string).

• The byte position is set (target substring adjusted accordingly).

Querying the Target Substring

This table summarizes (details and examples at the links):

Searching the Target Substring

A search method examines the target substring, but does not advance the positions or (by implication) shorten the target substring.

This table summarizes (details and examples at the links):

Traversing the Target Substring

A traversal method examines the target substring, and, if successful:

• Advances the positions.

• Shortens the target substring.

This table summarizes (details and examples at links):

Querying the Scanner

Each of these methods queries the scanner object without modifying it (details and examples at links)

Matching

StringScanner implements pattern matching via Ruby class Regexp, and its matching behaviors are the same as Ruby’s except for the fixed-anchor property.

Matcher Methods

Each matcher method takes a single argument pattern, and attempts to find a matching substring in the target substring.

Which matcher you choose will depend on:

• Where you want to find a match:

  • Only at the beginning of the target substring: check, match?, scan, skip.

  • Anywhere in the target substring: check_until, exist?, scan_until, skip_until.

• Whether you want to:

  • Traverse, by advancing the positions: scan, scan_until, skip, skip_until.

  • Keep the positions unchanged: check, check_until, match?, exist?.

• What you want for the return value:

  • The matched substring: check, scan.

  • The substring: check_until, scan_until.

  • The match size: match?, skip.

  • The substring size: exist?, skip_until.

Match Values

The match values in a StringScanner object generally contain the results of the most recent attempted match.

Each match value may be thought of as:

• Clear: Initially, or after an unsuccessful match attempt: usually, false, nil, or {}.

• Set: After a successful match attempt: true, string, array, or hash.

Each of these methods clears match values:

• ::new(string).

• reset.

• terminate.

Each of these methods attempts a match based on a pattern, and either sets match values (if successful) or clears them (if not);

• check(pattern)

• check_until(pattern)

• exist?(pattern)

• match?(pattern)

• scan(pattern)

• scan_until(pattern)

• skip(pattern)

• skip_until(pattern)

Basic Match Values

Basic match values are those not related to captures.

Each of these methods returns a basic match value:

See examples below.

Captured Match Values

Captured match values are those related to captures.

Each of these methods returns a captured match value:

See examples below.

Match Values Examples

Successful basic match attempt (no captures):

scanner = StringScanner.new('foobarbaz')
scanner.exist?(/bar/)
put_match_values(scanner)
# Basic match values:
#   matched?:       true
#   matched_size:   3
#   pre_match:      "foo"
#   matched  :      "bar"
#   post_match:     "baz"
# Captured match values:
#   size:           1
#   captures:       []
#   named_captures: {}
#   values_at:      ["bar", nil]
#   []:
#     [0]:          "bar"
#     [1]:          nil

Failed basic match attempt (no captures);

scanner = StringScanner.new('foobarbaz')
scanner.exist?(/nope/)
match_values_cleared?(scanner) # => true

Successful unnamed capture match attempt:

scanner = StringScanner.new('foobarbazbatbam')
scanner.exist?(/(foo)bar(baz)bat(bam)/)
put_match_values(scanner)
# Basic match values:
#   matched?:       true
#   matched_size:   15
#   pre_match:      ""
#   matched  :      "foobarbazbatbam"
#   post_match:     ""
# Captured match values:
#   size:           4
#   captures:       ["foo", "baz", "bam"]
#   named_captures: {}
#   values_at:      ["foobarbazbatbam", "foo", "baz", "bam", nil]
#   []:
#     [0]:          "foobarbazbatbam"
#     [1]:          "foo"
#     [2]:          "baz"
#     [3]:          "bam"
#     [4]:          nil

Successful named capture match attempt; same as unnamed above, except for named_captures:

scanner = StringScanner.new('foobarbazbatbam')
scanner.exist?(/(?<x>foo)bar(?<y>baz)bat(?<z>bam)/)
scanner.named_captures # => {"x"=>"foo", "y"=>"baz", "z"=>"bam"}

Failed unnamed capture match attempt:

scanner = StringScanner.new('somestring')
scanner.exist?(/(foo)bar(baz)bat(bam)/)
match_values_cleared?(scanner) # => true

Failed named capture match attempt; same as unnamed above, except for named_captures:

scanner = StringScanner.new('somestring')
scanner.exist?(/(?<x>foo)bar(?<y>baz)bat(?<z>bam)/)
match_values_cleared?(scanner) # => false
scanner.named_captures # => {"x"=>nil, "y"=>nil, "z"=>nil}

Fixed-Anchor Property

Pattern matching in StringScanner is the same as in Ruby’s, except for its fixed-anchor property, which determines the meaning of '\A':

• false (the default): matches the current byte position.

  Example

  Run in Playground

  scanner = StringScanner.new('foobar')
  scanner.scan(/\A./) # => "f"
  scanner.scan(/\A./) # => "o"
  scanner.scan(/\A./) # => "o"
  scanner.scan(/\A./) # => "b"
  

• true: matches the beginning of the target substring; never matches unless the byte position is zero:

  Example

  Run in Playground

  scanner = StringScanner.new('foobar', fixed_anchor: true)
  scanner.scan(/\A./) # => "f"
  scanner.scan(/\A./) # => nil
  scanner.reset
  scanner.scan(/\A./) # => "f"
  

The fixed-anchor property is set when the StringScanner object is created, and may not be modified (see StringScanner.new); method fixed_anchor? returns the setting.