Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from https://docs.ruby-lang.org/en/3.4/RubyVM/../String.html below:

class String - Documentation for Ruby 3.4

A String object has an arbitrary sequence of bytes, typically representing text or binary data. A String object may be created using String::new or as literals.

String objects differ from Symbol objects in that Symbol objects are designed to be used as identifiers, instead of text or data.

You can create a String object explicitly with:

• A string literal.

• A heredoc literal.

You can convert certain objects to Strings with:

• Method String.

Some String methods modify self. Typically, a method whose name ends with ! modifies self and returns self; often, a similarly named method (without the !) returns a new string.

In general, if both bang and non-bang versions of a method exist, the bang method mutates and the non-bang method does not. However, a method without a bang can also mutate, such as String#replace.

These methods perform substitutions:

• String#sub: One substitution (or none); returns a new string.

• String#sub!: One substitution (or none); returns self if any changes, nil otherwise.

• String#gsub: Zero or more substitutions; returns a new string.

• String#gsub!: Zero or more substitutions; returns self if any changes, nil otherwise.

Each of these methods takes:

• A first argument, pattern (String or Regexp), that specifies the substring(s) to be replaced.

• Either of the following:

  • A second argument, replacement (String or Hash), that determines the replacing string.

  • A block that will determine the replacing string.

The examples in this section mostly use the String#sub and String#gsub methods; the principles illustrated apply to all four substitution methods.

Argument pattern

Argument pattern is commonly a regular expression:

s = 'hello'
s.sub(/[aeiou]/, '*') 
s.gsub(/[aeiou]/, '*') 
s.gsub(/[aeiou]/, '')  
s.sub(/ell/, 'al')     
s.gsub(/xyzzy/, '*')   
'THX1138'.gsub(/\d+/, '00') 

When pattern is a string, all its characters are treated as ordinary characters (not as Regexp special characters):

'THX1138'.gsub('\d+', '00') 

String replacement

If replacement is a string, that string determines the replacing string that is substituted for the matched text.

Each of the examples above uses a simple string as the replacing string.

String replacement may contain back-references to the pattern’s captures:

• \n (n is a non-negative integer) refers to $n.

• \k<name> refers to the named capture name.

See Regexp for details.

Note that within the string replacement, a character combination such as $& is treated as ordinary text, not as a special match variable. However, you may refer to some special match variables using these combinations:

• \& and \0 correspond to $&, which contains the complete matched text.

• \' corresponds to $', which contains the string after the match.

• \` corresponds to $`, which contains the string before the match.

• \+ corresponds to $+, which contains the last capture group.

See Regexp for details.

Note that \\ is interpreted as an escape, i.e., a single backslash.

Note also that a string literal consumes backslashes. See String Literals for details about string literals.

A back-reference is typically preceded by an additional backslash. For example, if you want to write a back-reference \& in replacement with a double-quoted string literal, you need to write "..\\&..".

If you want to write a non-back-reference string \& in replacement, you need to first escape the backslash to prevent this method from interpreting it as a back-reference, and then you need to escape the backslashes again to prevent a string literal from consuming them: "..\\\\&..".

You may want to use the block form to avoid excessive backslashes.

Hash replacement

If the argument replacement is a hash, and pattern matches one of its keys, the replacing string is the value for that key:

h = {'foo' => 'bar', 'baz' => 'bat'}
'food'.sub('foo', h) 

Note that a symbol key does not match:

h = {foo: 'bar', baz: 'bat'}
'food'.sub('foo', h) 

Block

In the block form, the current match string is passed to the block; the block’s return value becomes the replacing string:

s = '@'
'1234'.gsub(/\d/) { |match| s.succ! } 

Special match variables such as $1, $2, $`, $&, and $' are set appropriately.

In the class String, whitespace is defined as a contiguous sequence of characters consisting of any mixture of the following:

• NL (null): "\x00", "\u0000".

• HT (horizontal tab): "\x09", "\t".

• LF (line feed): "\x0a", "\n".

• VT (vertical tab): "\x0b", "\v".

• FF (form feed): "\x0c", "\f".

• CR (carriage return): "\x0d", "\r".

• SP (space): "\x20", " ".

Whitespace is relevant for the following methods:

• lstrip, lstrip!: Strip leading whitespace.

• rstrip, rstrip!: Strip trailing whitespace.

• strip, strip!: Strip leading and trailing whitespace.

A slice of a string is a substring selected by certain criteria.

These instance methods utilize slicing:

• String#[] (aliased as String#slice): Returns a slice copied from self.

• String#[]=: Mutates self with the slice replaced.

• String#slice!: Mutates self with the slice removed and returns the removed slice.

Each of the above methods takes arguments that determine the slice to be copied or replaced.

The arguments have several forms. For a string string, the forms are:

• string[index]

• string[start, length]

• string[range]

• string[regexp, capture = 0]

• string[substring]

string[index]

When a non-negative integer argument index is given, the slice is the 1-character substring found in self at character offset index:

'bar'[0]      
'bar'[2]      
'bar'[20]     
'тест'[2]     
'こんにちは'[4] 

When a negative integer index is given, the slice begins at the offset given by counting backward from the end of self:

'bar'[-3]      
'bar'[-1]      
'bar'[-20]     

string[start, length]

When non-negative integer arguments start and length are given, the slice begins at character offset start, if it exists, and continues for length characters, if available:

'foo'[0, 2]      
'тест'[1, 2]     
'こんにちは'[2, 2] 

'foo'[2, 0]      

'foo'[1, 200]    

'foo'[4, 2]      

Special case: if start equals the length of self, the slice is a new empty string:

'foo'[3, 2]    
'foo'[3, 200]  

When a negative start and non-negative length are given, the slice begins by counting backward from the end of self, and continues for length characters, if available:

'foo'[-2, 2]     
'foo'[-2, 200]   

'foo'[-4, 2]     

When a negative length is given, there is no slice:

'foo'[1, -1]   
'foo'[-2, -1]  

string[range]

When a Range argument range is given, it creates a substring of string using the indices in range. The slice is then determined as above:

'foo'[0..1]     
'foo'[0, 2]     

'foo'[2...2]    
'foo'[2, 0]     

'foo'[1..200]   
'foo'[1, 200]   

'foo'[4..5]     
'foo'[4, 2]     

'foo'[-4..-3]   
'foo'[-4, 2]    

'foo'[3..4]     
'foo'[3, 2]     

'foo'[-2..-1]   
'foo'[-2, 2]    

'foo'[-2..197]  
'foo'[-2, 200]  

string[regexp, capture = 0]

When the Regexp argument regexp is given, and the capture argument is 0, the slice is the first matching substring found in self:

'foo'[/o/]                
'foo'[/x/]                
s = 'hello there'
s[/[aeiou](.)\1/]        
s[/[aeiou](.)\1/, 0]     

If the argument capture is provided and not 0, it should be either a capture group index (integer) or a capture group name (String or Symbol); the slice is the specified capture (see Groups at Regexp and Captures):

s = 'hello there'
s[/[aeiou](.)\1/, 1] 
s[/(?<vowel>[aeiou])(?<non_vowel>[^aeiou])/, "non_vowel"] 
s[/(?<vowel>[aeiou])(?<non_vowel>[^aeiou])/, :vowel]      

If an invalid capture group index is given, there is no slice. If an invalid capture group name is given, IndexError is raised.

string[substring]

When the single String argument substring is given, it returns the substring from self if found, otherwise nil:

'foo'['oo'] 
'foo'['xx'] 

First, what’s elsewhere. Class String:

• Inherits from the Object class.

• Includes the Comparable module.

Here, class String provides methods that are useful for:

• Creating a String

• Frozen/Unfrozen Strings

• Querying

• Comparing

• Modifying a String

• Converting to New String

• Converting to Non-String

• Iterating

• ::new: Returns a new string.

• ::try_convert: Returns a new string created from a given object.

• +@: Returns a string that is not frozen: self if not frozen; self.dup otherwise.

• -@ (aliased as dedup): Returns a string that is frozen: self if already frozen; self.freeze otherwise.

• freeze: Freezes self if not already frozen; returns self.

Counts

• length (aliased as size): Returns the count of characters (not bytes).

• empty?: Returns true if self.length is zero; false otherwise.

• bytesize: Returns the count of bytes.

• count: Returns the count of substrings matching given strings.

Substrings

• =~: Returns the index of the first substring that matches a given Regexp or other object; returns nil if no match is found.

• index: Returns the index of the first occurrence of a given substring; returns nil if none found.

• rindex: Returns the index of the last occurrence of a given substring; returns nil if none found.

• include?: Returns true if the string contains a given substring; false otherwise.

• match: Returns a MatchData object if the string matches a given Regexp; nil otherwise.

• match?: Returns true if the string matches a given Regexp; false otherwise.

• start_with?: Returns true if the string begins with any of the given substrings.

• end_with?: Returns true if the string ends with any of the given substrings.

Encodings

• encoding: Returns the Encoding object that represents the encoding of the string.

• unicode_normalized?: Returns true if the string is in Unicode normalized form; false otherwise.

• valid_encoding?: Returns true if the string contains only characters that are valid for its encoding.

• ascii_only?: Returns true if the string has only ASCII characters; false otherwise.

Other

• sum: Returns a basic checksum for the string: the sum of each byte.

• hash: Returns the integer hash code.

• == (aliased as ===): Returns true if a given other string has the same content as self.

• eql?: Returns true if the content is the same as the given other string.

• <=>: Returns -1, 0, or 1 as a given other string is smaller than, equal to, or larger than self.

• casecmp: Ignoring case, returns -1, 0, or 1 as a given other string is smaller than, equal to, or larger than self.

• casecmp?: Returns true if the string is equal to a given string after Unicode case folding; false otherwise.

Each of these methods modifies self.

Insertion

• insert: Returns self with a given string inserted at a specified offset.

• <<: Returns self concatenated with a given string or integer.

• append_as_bytes: Returns self concatenated with strings without performing any encoding validation or conversion.

Substitution

• sub!: Replaces the first substring that matches a given pattern with a given replacement string; returns self if any changes, nil otherwise.

• gsub!: Replaces each substring that matches a given pattern with a given replacement string; returns self if any changes, nil otherwise.

• succ! (aliased as next!): Returns self modified to become its own successor.

• initialize_copy (aliased as replace): Returns self with its entire content replaced by a given string.

• reverse!: Returns self with its characters in reverse order.

• setbyte: Sets the byte at a given integer offset to a given value; returns the argument.

• tr!: Replaces specified characters in self with specified replacement characters; returns self if any changes, nil otherwise.

• tr_s!: Replaces specified characters in self with specified replacement characters, removing duplicates from the substrings that were modified; returns self if any changes, nil otherwise.

Casing

• capitalize!: Upcases the initial character and downcases all others; returns self if any changes, nil otherwise.

• downcase!: Downcases all characters; returns self if any changes, nil otherwise.

• upcase!: Upcases all characters; returns self if any changes, nil otherwise.

• swapcase!: Upcases each downcase character and downcases each upcase character; returns self if any changes, nil otherwise.

Encoding

• encode!: Returns self with all characters transcoded from one encoding to another.

• unicode_normalize!: Unicode-normalizes self; returns self.

• scrub!: Replaces each invalid byte with a given character; returns self.

• force_encoding: Changes the encoding to a given encoding; returns self.

Deletion

• clear: Removes all content, so that self is empty; returns self.

• slice!, []=: Removes a substring determined by a given index, start/length, range, regexp, or substring.

• squeeze!: Removes contiguous duplicate characters; returns self.

• delete!: Removes characters as determined by the intersection of substring arguments.

• lstrip!: Removes leading whitespace; returns self if any changes, nil otherwise.

• rstrip!: Removes trailing whitespace; returns self if any changes, nil otherwise.

• strip!: Removes leading and trailing whitespace; returns self if any changes, nil otherwise.

• chomp!: Removes the trailing record separator, if found; returns self if any changes, nil otherwise.

• chop!: Removes trailing newline characters if found; otherwise removes the last character; returns self if any changes, nil otherwise.

Each of these methods returns a new String based on self, often just a modified copy of self.

Extension

• *: Returns the concatenation of multiple copies of self.

• +: Returns the concatenation of self and a given other string.

• center: Returns a copy of self centered between pad substrings.

• concat: Returns the concatenation of self with given other strings.

• prepend: Returns the concatenation of a given other string with self.

• ljust: Returns a copy of self of a given length, right-padded with a given other string.

• rjust: Returns a copy of self of a given length, left-padded with a given other string.

Encoding

• b: Returns a copy of self with ASCII-8BIT encoding.

• scrub: Returns a copy of self with each invalid byte replaced with a given character.

• unicode_normalize: Returns a copy of self with each character Unicode-normalized.

• encode: Returns a copy of self with all characters transcoded from one encoding to another.

Substitution

• dump: Returns a copy of self with all non-printing characters replaced by xHH notation and all special characters escaped.

• undump: Returns a copy of self with all \xNN notations replaced by \uNNNN notations and all escaped characters unescaped.

• sub: Returns a copy of self with the first substring matching a given pattern replaced with a given replacement string.

• gsub: Returns a copy of self with each substring that matches a given pattern replaced with a given replacement string.

• succ (aliased as next): Returns the string that is the successor to self.

• reverse: Returns a copy of self with its characters in reverse order.

• tr: Returns a copy of self with specified characters replaced with specified replacement characters.

• tr_s: Returns a copy of self with specified characters replaced with specified replacement characters, removing duplicates from the substrings that were modified.

• %: Returns the string resulting from formatting a given object into self.

Casing

• capitalize: Returns a copy of self with the first character upcased and all other characters downcased.

• downcase: Returns a copy of self with all characters downcased.

• upcase: Returns a copy of self with all characters upcased.

• swapcase: Returns a copy of self with all upcase characters downcased and all downcase characters upcased.

Deletion

• delete: Returns a copy of self with characters removed.

• delete_prefix: Returns a copy of self with a given prefix removed.

• delete_suffix: Returns a copy of self with a given suffix removed.

• lstrip: Returns a copy of self with leading whitespace removed.

• rstrip: Returns a copy of self with trailing whitespace removed.

• strip: Returns a copy of self with leading and trailing whitespace removed.

• chomp: Returns a copy of self with a trailing record separator removed, if found.

• chop: Returns a copy of self with trailing newline characters or the last character removed.

• squeeze: Returns a copy of self with contiguous duplicate characters removed.

• [] (aliased as slice): Returns a substring determined by a given index, start/length, range, regexp, or string.

• byteslice: Returns a substring determined by a given index, start/length, or range.

• chr: Returns the first character.

Duplication

• to_s (aliased as to_str): If self is a subclass of String, returns self copied into a String; otherwise, returns self.

Each of these methods converts the contents of self to a non-String.

Characters, Bytes, and Clusters

• bytes: Returns an array of the bytes in self.

• chars: Returns an array of the characters in self.

• codepoints: Returns an array of the integer ordinals in self.

• getbyte: Returns the integer byte at the given index in self.

• grapheme_clusters: Returns an array of the grapheme clusters in self.

Splitting

• lines: Returns an array of the lines in self, as determined by a given record separator.

• partition: Returns a 3-element array determined by the first substring that matches a given substring or regexp.

• rpartition: Returns a 3-element array determined by the last substring that matches a given substring or regexp.

• split: Returns an array of substrings determined by a given delimiter – regexp or string – or, if a block is given, passes those substrings to the block.

Matching

• scan: Returns an array of substrings matching a given regexp or string, or, if a block is given, passes each matching substring to the block.

• unpack: Returns an array of substrings extracted from self according to a given format.

• unpack1: Returns the first substring extracted from self according to a given format.

Numerics

• hex: Returns the integer value of the leading characters, interpreted as hexadecimal digits.

• oct: Returns the integer value of the leading characters, interpreted as octal digits.

• ord: Returns the integer ordinal of the first character in self.

• to_i: Returns the integer value of leading characters, interpreted as an integer.

• to_f: Returns the floating-point value of leading characters, interpreted as a floating-point number.

Strings and Symbols

• inspect: Returns a copy of self, enclosed in double quotes, with special characters escaped.

• intern (aliased as to_sym): Returns the symbol corresponding to self.

• each_byte: Calls the given block with each successive byte in self.

• each_char: Calls the given block with each successive character in self.

• each_codepoint: Calls the given block with each successive integer codepoint in self.

• each_grapheme_cluster: Calls the given block with each successive grapheme cluster in self.

• each_line: Calls the given block with each successive line in self, as determined by a given record separator.

• upto: Calls the given block with each string value returned by successive calls to succ.

static VALUE
rb_str_init(int argc, VALUE *argv, VALUE str)
{
    static ID keyword_ids[2];
    VALUE orig, opt, venc, vcapa;
    VALUE kwargs[2];
    rb_encoding *enc = 0;
    int n;

    if (!keyword_ids[0]) {
        keyword_ids[0] = rb_id_encoding();
        CONST_ID(keyword_ids[1], "capacity");
    }

    n = rb_scan_args(argc, argv, "01:", &orig, &opt);
    if (!NIL_P(opt)) {
        rb_get_kwargs(opt, keyword_ids, 0, 2, kwargs);
        venc = kwargs[0];
        vcapa = kwargs[1];
        if (!UNDEF_P(venc) && !NIL_P(venc)) {
            enc = rb_to_encoding(venc);
        }
        if (!UNDEF_P(vcapa) && !NIL_P(vcapa)) {
            long capa = NUM2LONG(vcapa);
            long len = 0;
            int termlen = enc ? rb_enc_mbminlen(enc) : 1;

            if (capa < STR_BUF_MIN_SIZE) {
                capa = STR_BUF_MIN_SIZE;
            }
            if (n == 1) {
                StringValue(orig);
                len = RSTRING_LEN(orig);
                if (capa < len) {
                    capa = len;
                }
                if (orig == str) n = 0;
            }
            str_modifiable(str);
            if (STR_EMBED_P(str) || FL_TEST(str, STR_SHARED|STR_NOFREE)) {
                /* make noembed always */
                const size_t size = (size_t)capa + termlen;
                const char *const old_ptr = RSTRING_PTR(str);
                const size_t osize = RSTRING_LEN(str) + TERM_LEN(str);
                char *new_ptr = ALLOC_N(char, size);
                if (STR_EMBED_P(str)) RUBY_ASSERT((long)osize <= str_embed_capa(str));
                memcpy(new_ptr, old_ptr, osize < size ? osize : size);
                FL_UNSET_RAW(str, STR_SHARED|STR_NOFREE);
                RSTRING(str)->as.heap.ptr = new_ptr;
            }
            else if (STR_HEAP_SIZE(str) != (size_t)capa + termlen) {
                SIZED_REALLOC_N(RSTRING(str)->as.heap.ptr, char,
                        (size_t)capa + termlen, STR_HEAP_SIZE(str));
            }
            STR_SET_LEN(str, len);
            TERM_FILL(&RSTRING(str)->as.heap.ptr[len], termlen);
            if (n == 1) {
                memcpy(RSTRING(str)->as.heap.ptr, RSTRING_PTR(orig), len);
                rb_enc_cr_str_exact_copy(str, orig);
            }
            FL_SET(str, STR_NOEMBED);
            RSTRING(str)->as.heap.aux.capa = capa;
        }
        else if (n == 1) {
            rb_str_replace(str, orig);
        }
        if (enc) {
            rb_enc_associate(str, enc);
            ENC_CODERANGE_CLEAR(str);
        }
    }
    else if (n == 1) {
        rb_str_replace(str, orig);
    }
    return str;
}

Returns a new String that is a copy of string.

With no arguments, returns the empty string with the Encoding ASCII-8BIT:

s = String.new
s 
s.encoding 

With optional argument string and no keyword arguments, returns a copy of string with the same encoding:

String.new('foo')               
String.new('тест')              
String.new('こんにちは')          

(Unlike String.new, a string literal like '' or a here document literal always has script encoding.)

With optional keyword argument encoding, returns a copy of string with the specified encoding; the encoding may be an Encoding object, an encoding name, or an encoding name alias:

String.new('foo', encoding: Encoding::US_ASCII).encoding 
String.new('foo', encoding: 'US-ASCII').encoding         
String.new('foo', encoding: 'ASCII').encoding            

The given encoding need not be valid for the string’s content, and that validity is not checked:

s = String.new('こんにちは', encoding: 'ascii')
s.valid_encoding? 

But the given encoding itself is checked:

String.new('foo', encoding: 'bar') 

With optional keyword argument capacity, returns a copy of string (or an empty string, if string is not given); the given capacity is advisory only, and may or may not set the size of the internal buffer, which may in turn affect performance:

String.new(capacity: 1)
String.new('foo', capacity: 4096)

Note that Ruby strings are null-terminated internally, so the internal buffer size will be one or more bytes larger than the requested capacity depending on the encoding.

The string, encoding, and capacity arguments may all be used together:

String.new('hello', encoding: 'UTF-8', capacity: 25)

static VALUE
rb_str_s_try_convert(VALUE dummy, VALUE str)
{
    return rb_check_string_type(str);
}

If object is a String object, returns object.

Otherwise if object responds to :to_str, calls object.to_str and returns the result.

Returns nil if object does not respond to :to_str.

Raises an exception unless object.to_str returns a String object.

static VALUE
rb_str_format_m(VALUE str, VALUE arg)
{
    VALUE tmp = rb_check_array_type(arg);

    if (!NIL_P(tmp)) {
        return rb_str_format(RARRAY_LENINT(tmp), RARRAY_CONST_PTR(tmp), str);
    }
    return rb_str_format(1, &arg, str);
}

Returns the result of formatting object into the format specification self (see Kernel#sprintf for formatting details):

"%05d" % 123 

If self contains multiple substitutions, object must be an Array or Hash containing the values to be substituted:

"%-5s: %016x" % [ "ID", self.object_id ] 
"foo = %{foo}" % {foo: 'bar'} 
"foo = %{foo}, baz = %{baz}" % {foo: 'bar', baz: 'bat'} 

VALUE
rb_str_times(VALUE str, VALUE times)
{
    VALUE str2;
    long n, len;
    char *ptr2;
    int termlen;

    if (times == INT2FIX(1)) {
        return str_duplicate(rb_cString, str);
    }
    if (times == INT2FIX(0)) {
        str2 = str_alloc_embed(rb_cString, 0);
        rb_enc_copy(str2, str);
        return str2;
    }
    len = NUM2LONG(times);
    if (len < 0) {
        rb_raise(rb_eArgError, "negative argument");
    }
    if (RSTRING_LEN(str) == 1 && RSTRING_PTR(str)[0] == 0) {
        if (STR_EMBEDDABLE_P(len, 1)) {
            str2 = str_alloc_embed(rb_cString, len + 1);
            memset(RSTRING_PTR(str2), 0, len + 1);
        }
        else {
            str2 = str_alloc_heap(rb_cString);
            RSTRING(str2)->as.heap.aux.capa = len;
            RSTRING(str2)->as.heap.ptr = ZALLOC_N(char, (size_t)len + 1);
        }
        STR_SET_LEN(str2, len);
        rb_enc_copy(str2, str);
        return str2;
    }
    if (len && LONG_MAX/len <  RSTRING_LEN(str)) {
        rb_raise(rb_eArgError, "argument too big");
    }

    len *= RSTRING_LEN(str);
    termlen = TERM_LEN(str);
    str2 = str_enc_new(rb_cString, 0, len, STR_ENC_GET(str));
    ptr2 = RSTRING_PTR(str2);
    if (len) {
        n = RSTRING_LEN(str);
        memcpy(ptr2, RSTRING_PTR(str), n);
        while (n <= len/2) {
            memcpy(ptr2 + n, ptr2, n);
            n *= 2;
        }
        memcpy(ptr2 + n, ptr2, len-n);
    }
    STR_SET_LEN(str2, len);
    TERM_FILL(&ptr2[len], termlen);
    rb_enc_cr_str_copy_for_substr(str2, str);

    return str2;
}

Returns a new String containing integer copies of self:

"Ho! " * 3 
"Ho! " * 0 

VALUE
rb_str_plus(VALUE str1, VALUE str2)
{
    VALUE str3;
    rb_encoding *enc;
    char *ptr1, *ptr2, *ptr3;
    long len1, len2;
    int termlen;

    StringValue(str2);
    enc = rb_enc_check_str(str1, str2);
    RSTRING_GETMEM(str1, ptr1, len1);
    RSTRING_GETMEM(str2, ptr2, len2);
    termlen = rb_enc_mbminlen(enc);
    if (len1 > LONG_MAX - len2) {
        rb_raise(rb_eArgError, "string size too big");
    }
    str3 = str_enc_new(rb_cString, 0, len1+len2, enc);
    ptr3 = RSTRING_PTR(str3);
    memcpy(ptr3, ptr1, len1);
    memcpy(ptr3+len1, ptr2, len2);
    TERM_FILL(&ptr3[len1+len2], termlen);

    ENCODING_CODERANGE_SET(str3, rb_enc_to_index(enc),
                           ENC_CODERANGE_AND(ENC_CODERANGE(str1), ENC_CODERANGE(str2)));
    RB_GC_GUARD(str1);
    RB_GC_GUARD(str2);
    return str3;
}

Returns a new String containing other_string concatenated to self:

"Hello from " + self.to_s 

static VALUE
str_uplus(VALUE str)
{
    if (OBJ_FROZEN(str) || CHILLED_STRING_P(str)) {
        return rb_str_dup(str);
    }
    else {
        return str;
    }
}

Returns self if self is not frozen and can be mutated without warning issuance.

Otherwise returns self.dup, which is not frozen.

static VALUE
str_uminus(VALUE str)
{
    if (!BARE_STRING_P(str) && !rb_obj_frozen_p(str)) {
        str = rb_str_dup(str);
    }
    return rb_fstring(str);
}

Returns a frozen, possibly pre-existing copy of the string.

The returned String will be deduplicated as long as it does not have any instance variables set on it and is not a String subclass.

Note that -string variant is more convenient for defining constants:

FILENAME = -'config/database.yml'

while dedup is better suitable for using the method in chains of calculations:

@url_list.concat(urls.map(&:dedup))

VALUE
rb_str_concat(VALUE str1, VALUE str2)
{
    unsigned int code;
    rb_encoding *enc = STR_ENC_GET(str1);
    int encidx;

    if (RB_INTEGER_TYPE_P(str2)) {
        if (rb_num_to_uint(str2, &code) == 0) {
        }
        else if (FIXNUM_P(str2)) {
            rb_raise(rb_eRangeError, "%ld out of char range", FIX2LONG(str2));
        }
        else {
            rb_raise(rb_eRangeError, "bignum out of char range");
        }
    }
    else {
        return rb_str_append(str1, str2);
    }

    encidx = rb_ascii8bit_appendable_encoding_index(enc, code);

    if (encidx >= 0) {
        rb_str_buf_cat_byte(str1, (unsigned char)code);
    }
    else {
        long pos = RSTRING_LEN(str1);
        int cr = ENC_CODERANGE(str1);
        int len;
        char *buf;

        switch (len = rb_enc_codelen(code, enc)) {
          case ONIGERR_INVALID_CODE_POINT_VALUE:
            rb_raise(rb_eRangeError, "invalid codepoint 0x%X in %s", code, rb_enc_name(enc));
            break;
          case ONIGERR_TOO_BIG_WIDE_CHAR_VALUE:
          case 0:
            rb_raise(rb_eRangeError, "%u out of char range", code);
            break;
        }
        buf = ALLOCA_N(char, len + 1);
        rb_enc_mbcput(code, buf, enc);
        if (rb_enc_precise_mbclen(buf, buf + len + 1, enc) != len) {
            rb_raise(rb_eRangeError, "invalid codepoint 0x%X in %s", code, rb_enc_name(enc));
        }
        rb_str_resize(str1, pos+len);
        memcpy(RSTRING_PTR(str1) + pos, buf, len);
        if (cr == ENC_CODERANGE_7BIT && code > 127) {
            cr = ENC_CODERANGE_VALID;
        }
        else if (cr == ENC_CODERANGE_BROKEN) {
            cr = ENC_CODERANGE_UNKNOWN;
        }
        ENC_CODERANGE_SET(str1, cr);
    }
    return str1;
}

Concatenates object to self and returns self:

s = 'foo'
s << 'bar' 
s          

If object is an Integer, the value is considered a codepoint and converted to a character before concatenation:

s = 'foo'
s << 33 

If that codepoint is not representable in the encoding of string, RangeError is raised.

s = 'foo'
s.encoding              
s << 0x00110000         
s = 'foo'.encode('EUC-JP')
s << 0x00800080         

If the encoding is US-ASCII and the codepoint is 0..0xff, string is automatically promoted to ASCII-8BIT.

s = 'foo'.encode('US-ASCII')
s << 0xff
s.encoding              

Related: String#concat, which takes multiple arguments.

static VALUE
rb_str_cmp_m(VALUE str1, VALUE str2)
{
    int result;
    VALUE s = rb_check_string_type(str2);
    if (NIL_P(s)) {
        return rb_invcmp(str1, str2);
    }
    result = rb_str_cmp(str1, s);
    return INT2FIX(result);
}

Compares self and other_string, returning:

• -1 if other_string is larger.

• 0 if the two are equal.

• 1 if other_string is smaller.

• nil if the two are incomparable.

Examples:

'foo' <=> 'foo' 
'foo' <=> 'food' 
'food' <=> 'foo' 
'FOO' <=> 'foo' 
'foo' <=> 'FOO' 
'foo' <=> 1 

VALUE
rb_str_equal(VALUE str1, VALUE str2)
{
    if (str1 == str2) return Qtrue;
    if (!RB_TYPE_P(str2, T_STRING)) {
        if (!rb_respond_to(str2, idTo_str)) {
            return Qfalse;
        }
        return rb_equal(str2, str1);
    }
    return rb_str_eql_internal(str1, str2);
}

Returns true if object has the same length and content; as self; false otherwise:

s = 'foo'
s == 'foo' 
s == 'food' 
s == 'FOO' 

Returns false if the two strings’ encodings are not compatible:

"\u{e4 f6 fc}".encode("ISO-8859-1") == ("\u{c4 d6 dc}") 

If object is not an instance of String but responds to to_str, then the two strings are compared using object.==.

Returns true if object has the same length and content; as self; false otherwise:

s = 'foo'
s == 'foo' 
s == 'food' 
s == 'FOO' 

Returns false if the two strings’ encodings are not compatible:

"\u{e4 f6 fc}".encode("ISO-8859-1") == ("\u{c4 d6 dc}") 

If object is not an instance of String but responds to to_str, then the two strings are compared using object.==.

static VALUE
rb_str_match(VALUE x, VALUE y)
{
    switch (OBJ_BUILTIN_TYPE(y)) {
      case T_STRING:
        rb_raise(rb_eTypeError, "type mismatch: String given");

      case T_REGEXP:
        return rb_reg_match(y, x);

      default:
        return rb_funcall(y, idEqTilde, 1, x);
    }
}

Returns the Integer index of the first substring that matches the given regexp, or nil if no match found:

'foo' =~ /f/ 
'foo' =~ /o/ 
'foo' =~ /x/ 

Note: also updates Global Variables at Regexp.

If the given object is not a Regexp, returns the value returned by object =~ self.

Note that string =~ regexp is different from regexp =~ string (see Regexp#=~):

number= nil
"no. 9" =~ /(?<number>\d+)/
number 
/(?<number>\d+)/ =~ "no. 9"
number 

static VALUE
rb_str_aref_m(int argc, VALUE *argv, VALUE str)
{
    if (argc == 2) {
        if (RB_TYPE_P(argv[0], T_REGEXP)) {
            return rb_str_subpat(str, argv[0], argv[1]);
        }
        else {
            return rb_str_substr_two_fixnums(str, argv[0], argv[1], TRUE);
        }
    }
    rb_check_arity(argc, 1, 2);
    return rb_str_aref(str, argv[0]);
}

Returns the substring of self specified by the arguments. See examples at String Slices.

static VALUE
rb_str_aset_m(int argc, VALUE *argv, VALUE str)
{
    if (argc == 3) {
        if (RB_TYPE_P(argv[0], T_REGEXP)) {
            rb_str_subpat_set(str, argv[0], argv[1], argv[2]);
        }
        else {
            rb_str_update(str, NUM2LONG(argv[0]), NUM2LONG(argv[1]), argv[2]);
        }
        return argv[2];
    }
    rb_check_arity(argc, 2, 3);
    return rb_str_aset(str, argv[0], argv[1]);
}

Replaces all, some, or none of the contents of self; returns new_string. See String Slices.

A few examples:

s = 'foo'
s[2] = 'rtune'     
s                  
s[1, 5] = 'init'   
s                  
s[3..4] = 'al'     
s                  
s[/e$/] = 'ly'     
s                  
s['lly'] = 'ncial' 
s                  

VALUE
rb_str_append_as_bytes(int argc, VALUE *argv, VALUE str)
{
    long needed_capacity = 0;
    volatile VALUE t0;
    enum ruby_value_type *types = ALLOCV_N(enum ruby_value_type, t0, argc);

    for (int index = 0; index < argc; index++) {
        VALUE obj = argv[index];
        enum ruby_value_type type = types[index] = rb_type(obj);
        switch (type) {
          case T_FIXNUM:
          case T_BIGNUM:
            needed_capacity++;
            break;
          case T_STRING:
            needed_capacity += RSTRING_LEN(obj);
            break;
          default:
            rb_raise(
                rb_eTypeError,
                "wrong argument type %"PRIsVALUE" (expected String or Integer)",
                rb_obj_class(obj)
            );
            break;
        }
    }

    str_ensure_available_capa(str, needed_capacity);
    char *sptr = RSTRING_END(str);

    for (int index = 0; index < argc; index++) {
        VALUE obj = argv[index];
        enum ruby_value_type type = types[index];
        switch (type) {
          case T_FIXNUM:
          case T_BIGNUM: {
            argv[index] = obj = rb_int_and(obj, INT2FIX(0xff));
            char byte = (char)(NUM2INT(obj) & 0xFF);
            *sptr = byte;
            sptr++;
            break;
          }
          case T_STRING: {
            const char *ptr;
            long len;
            RSTRING_GETMEM(obj, ptr, len);
            memcpy(sptr, ptr, len);
            sptr += len;
            break;
          }
          default:
            rb_bug("append_as_bytes arguments should have been validated");
        }
    }

    STR_SET_LEN(str, RSTRING_LEN(str) + needed_capacity);
    TERM_FILL(sptr, TERM_LEN(str)); /* sentinel */

    int cr = ENC_CODERANGE(str);
    switch (cr) {
      case ENC_CODERANGE_7BIT: {
        for (int index = 0; index < argc; index++) {
            VALUE obj = argv[index];
            enum ruby_value_type type = types[index];
            switch (type) {
              case T_FIXNUM:
              case T_BIGNUM: {
                if (!ISASCII(NUM2INT(obj))) {
                    goto clear_cr;
                }
                break;
              }
              case T_STRING: {
                if (ENC_CODERANGE(obj) != ENC_CODERANGE_7BIT) {
                    goto clear_cr;
                }
                break;
              }
              default:
                rb_bug("append_as_bytes arguments should have been validated");
            }
        }
        break;
      }
      case ENC_CODERANGE_VALID:
        if (ENCODING_GET_INLINED(str) == ENCINDEX_ASCII_8BIT) {
            goto keep_cr;
        }
        else {
            goto clear_cr;
        }
        break;
      default:
        goto clear_cr;
        break;
    }

    RB_GC_GUARD(t0);

  clear_cr:
    // If no fast path was hit, we clear the coderange.
    // append_as_bytes is predominently meant to be used in
    // buffering situation, hence it's likely the coderange
    // will never be scanned, so it's not worth spending time
    // precomputing the coderange except for simple and common
    // situations.
    ENC_CODERANGE_CLEAR(str);
  keep_cr:
    return str;
}

Concatenates each object in objects into self without any encoding validation or conversion and returns self:

s = 'foo'
s.append_as_bytes(" \xE2\x82")  
s.valid_encoding?               
s.append_as_bytes("\xAC 12")
s.valid_encoding?               

For each given object object that is an Integer, the value is considered a Byte. If the Integer is bigger than one byte, only the lower byte is considered, similar to String#setbyte:

s = ""
s.append_as_bytes(0, 257)             

Related: String#<<, String#concat, which do an encoding aware concatenation.

static VALUE
rb_str_is_ascii_only_p(VALUE str)
{
    int cr = rb_enc_str_coderange(str);

    return RBOOL(cr == ENC_CODERANGE_7BIT);
}

Returns true if self contains only ASCII characters, false otherwise:

'abc'.ascii_only?         
"abc\u{6666}".ascii_only? 

static VALUE
rb_str_b(VALUE str)
{
    VALUE str2;
    if (STR_EMBED_P(str)) {
        str2 = str_alloc_embed(rb_cString, RSTRING_LEN(str) + TERM_LEN(str));
    }
    else {
        str2 = str_alloc_heap(rb_cString);
    }
    str_replace_shared_without_enc(str2, str);

    if (rb_enc_asciicompat(STR_ENC_GET(str))) {
        // BINARY strings can never be broken; they're either 7-bit ASCII or VALID.
        // If we know the receiver's code range then we know the result's code range.
        int cr = ENC_CODERANGE(str);
        switch (cr) {
          case ENC_CODERANGE_7BIT:
            ENC_CODERANGE_SET(str2, ENC_CODERANGE_7BIT);
            break;
          case ENC_CODERANGE_BROKEN:
          case ENC_CODERANGE_VALID:
            ENC_CODERANGE_SET(str2, ENC_CODERANGE_VALID);
            break;
          default:
            ENC_CODERANGE_CLEAR(str2);
            break;
        }
    }

    return str2;
}

Returns a copy of self that has ASCII-8BIT encoding; the underlying bytes are not modified:

s = "\x99"
s.encoding   
t = s.b      
t.encoding   

s = "\u4095" 
s.encoding   
s.bytes      
t = s.b      
t.encoding   
t.bytes      

static VALUE
rb_str_byteindex_m(int argc, VALUE *argv, VALUE str)
{
    VALUE sub;
    VALUE initpos;
    long pos;

    if (rb_scan_args(argc, argv, "11", &sub, &initpos) == 2) {
        long slen = RSTRING_LEN(str);
        pos = NUM2LONG(initpos);
        if (pos < 0 ? (pos += slen) < 0 : pos > slen) {
            if (RB_TYPE_P(sub, T_REGEXP)) {
                rb_backref_set(Qnil);
            }
            return Qnil;
        }
    }
    else {
        pos = 0;
    }

    str_ensure_byte_pos(str, pos);

    if (RB_TYPE_P(sub, T_REGEXP)) {
        if (rb_reg_search(sub, str, pos, 0) >= 0) {
            VALUE match = rb_backref_get();
            struct re_registers *regs = RMATCH_REGS(match);
            pos = BEG(0);
            return LONG2NUM(pos);
        }
    }
    else {
        StringValue(sub);
        pos = rb_str_byteindex(str, sub, pos);
        if (pos >= 0) return LONG2NUM(pos);
    }
    return Qnil;
}

Returns the Integer byte-based index of the first occurrence of the given substring, or nil if none found:

'foo'.byteindex('f') 
'foo'.byteindex('o') 
'foo'.byteindex('oo') 
'foo'.byteindex('ooo') 

Returns the Integer byte-based index of the first match for the given Regexp regexp, or nil if none found:

'foo'.byteindex(/f/) 
'foo'.byteindex(/o/) 
'foo'.byteindex(/oo/) 
'foo'.byteindex(/ooo/) 

Integer argument offset, if given, specifies the byte-based position in the string to begin the search:

'foo'.byteindex('o', 1) 
'foo'.byteindex('o', 2) 
'foo'.byteindex('o', 3) 

If offset is negative, counts backward from the end of self:

'foo'.byteindex('o', -1) 
'foo'.byteindex('o', -2) 
'foo'.byteindex('o', -3) 
'foo'.byteindex('o', -4) 

If offset does not land on character (codepoint) boundary, IndexError is raised.

Related: String#index, String#byterindex.

static VALUE
rb_str_byterindex_m(int argc, VALUE *argv, VALUE str)
{
    VALUE sub;
    VALUE initpos;
    long pos, len = RSTRING_LEN(str);

    if (rb_scan_args(argc, argv, "11", &sub, &initpos) == 2) {
        pos = NUM2LONG(initpos);
        if (pos < 0 && (pos += len) < 0) {
            if (RB_TYPE_P(sub, T_REGEXP)) {
                rb_backref_set(Qnil);
            }
            return Qnil;
        }
        if (pos > len) pos = len;
    }
    else {
        pos = len;
    }

    str_ensure_byte_pos(str, pos);

    if (RB_TYPE_P(sub, T_REGEXP)) {
        if (rb_reg_search(sub, str, pos, 1) >= 0) {
            VALUE match = rb_backref_get();
            struct re_registers *regs = RMATCH_REGS(match);
            pos = BEG(0);
            return LONG2NUM(pos);
        }
    }
    else {
        StringValue(sub);
        pos = rb_str_byterindex(str, sub, pos);
        if (pos >= 0) return LONG2NUM(pos);
    }
    return Qnil;
}

Returns the Integer byte-based index of the last occurrence of the given substring, or nil if none found:

'foo'.byterindex('f') 
'foo'.byterindex('o') 
'foo'.byterindex('oo') 
'foo'.byterindex('ooo') 

Returns the Integer byte-based index of the last match for the given Regexp regexp, or nil if none found:

'foo'.byterindex(/f/) 
'foo'.byterindex(/o/) 
'foo'.byterindex(/oo/) 
'foo'.byterindex(/ooo/) 

The last match means starting at the possible last position, not the last of longest matches.

'foo'.byterindex(/o+/) 
$~ 

To get the last longest match, needs to combine with negative lookbehind.

'foo'.byterindex(/(?<!o)o+/) 
$~ 

Or String#byteindex with negative lookforward.

'foo'.byteindex(/o+(?!.*o)/) 
$~ 

Integer argument offset, if given and non-negative, specifies the maximum starting byte-based position in the string to end the search:

'foo'.byterindex('o', 0) 
'foo'.byterindex('o', 1) 
'foo'.byterindex('o', 2) 
'foo'.byterindex('o', 3) 

If offset is a negative Integer, the maximum starting position in the string to end the search is the sum of the string’s length and offset:

'foo'.byterindex('o', -1) 
'foo'.byterindex('o', -2) 
'foo'.byterindex('o', -3) 
'foo'.byterindex('o', -4) 

If offset does not land on character (codepoint) boundary, IndexError is raised.

Related: String#byteindex.

static VALUE
rb_str_bytes(VALUE str)
{
    VALUE ary = WANTARRAY("bytes", RSTRING_LEN(str));
    return rb_str_enumerate_bytes(str, ary);
}

Returns an array of the bytes in self:

'hello'.bytes 
'тест'.bytes  
'こんにちは'.bytes

VALUE
rb_str_bytesize(VALUE str)
{
    return LONG2NUM(RSTRING_LEN(str));
}

Returns the count of bytes (not characters) in self:

'foo'.bytesize        
'тест'.bytesize       
'こんにちは'.bytesize   

Contrast with String#length:

'foo'.length       
'тест'.length      
'こんにちは'.length  

static VALUE
rb_str_byteslice(int argc, VALUE *argv, VALUE str)
{
    if (argc == 2) {
        long beg = NUM2LONG(argv[0]);
        long len = NUM2LONG(argv[1]);
        return str_byte_substr(str, beg, len, TRUE);
    }
    rb_check_arity(argc, 1, 2);
    return str_byte_aref(str, argv[0]);
}

Returns a substring of self, or nil if the substring cannot be constructed.

With integer arguments index and length given, returns the substring beginning at the given index of the given length (if possible), or nil if length is negative or index falls outside of self:

s = '0123456789' 
s.byteslice(2)   
s.byteslice(200) 
s.byteslice(4, 3)  
s.byteslice(4, 30) 
s.byteslice(4, -1) 
s.byteslice(40, 2) 

In either case above, counts backwards from the end of self if index is negative:

s = '0123456789'   
s.byteslice(-4)    
s.byteslice(-4, 3) 

With Range argument range given, returns byteslice(range.begin, range.size):

s = '0123456789'    
s.byteslice(4..6)   
s.byteslice(-6..-4) 
s.byteslice(5..2)   
s.byteslice(40..42) 

In all cases, a returned string has the same encoding as self:

s.encoding              
s.byteslice(4).encoding 

static VALUE
rb_str_bytesplice(int argc, VALUE *argv, VALUE str)
{
    long beg, len, vbeg, vlen;
    VALUE val;
    int cr;

    rb_check_arity(argc, 2, 5);
    if (!(argc == 2 || argc == 3 || argc == 5)) {
        rb_raise(rb_eArgError, "wrong number of arguments (given %d, expected 2, 3, or 5)", argc);
    }
    if (argc == 2 || (argc == 3 && !RB_INTEGER_TYPE_P(argv[0]))) {
        if (!rb_range_beg_len(argv[0], &beg, &len, RSTRING_LEN(str), 2)) {
            rb_raise(rb_eTypeError, "wrong argument type %s (expected Range)",
                     rb_builtin_class_name(argv[0]));
        }
        val = argv[1];
        StringValue(val);
        if (argc == 2) {
            /* bytesplice(range, str) */
            vbeg = 0;
            vlen = RSTRING_LEN(val);
        }
        else {
            /* bytesplice(range, str, str_range) */
            if (!rb_range_beg_len(argv[2], &vbeg, &vlen, RSTRING_LEN(val), 2)) {
                rb_raise(rb_eTypeError, "wrong argument type %s (expected Range)",
                         rb_builtin_class_name(argv[2]));
            }
        }
    }
    else {
        beg = NUM2LONG(argv[0]);
        len = NUM2LONG(argv[1]);
        val = argv[2];
        StringValue(val);
        if (argc == 3) {
            /* bytesplice(index, length, str) */
            vbeg = 0;
            vlen = RSTRING_LEN(val);
        }
        else {
            /* bytesplice(index, length, str, str_index, str_length) */
            vbeg = NUM2LONG(argv[3]);
            vlen = NUM2LONG(argv[4]);
        }
    }
    str_check_beg_len(str, &beg, &len);
    str_check_beg_len(val, &vbeg, &vlen);
    str_modify_keep_cr(str);

    if (RB_UNLIKELY(ENCODING_GET_INLINED(str) != ENCODING_GET_INLINED(val))) {
        rb_enc_associate(str, rb_enc_check(str, val));
    }

    rb_str_update_1(str, beg, len, val, vbeg, vlen);
    cr = ENC_CODERANGE_AND(ENC_CODERANGE(str), ENC_CODERANGE(val));
    if (cr != ENC_CODERANGE_BROKEN)
        ENC_CODERANGE_SET(str, cr);
    return str;
}

Replaces some or all of the content of self with str, and returns self. The portion of the string affected is determined using the same criteria as String#byteslice, except that length cannot be omitted. If the replacement string is not the same length as the text it is replacing, the string will be adjusted accordingly.

If str_index and str_length, or str_range are given, the content of self is replaced by str.byteslice(str_index, str_length) or str.byteslice(str_range); however the substring of str is not allocated as a new string.

The form that take an Integer will raise an IndexError if the value is out of range; the Range form will raise a RangeError. If the beginning or ending offset does not land on character (codepoint) boundary, an IndexError will be raised.

static VALUE
rb_str_capitalize(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE | ONIGENC_CASE_TITLECASE;
    VALUE ret;

    flags = check_case_options(argc, argv, flags);
    enc = str_true_enc(str);
    if (RSTRING_LEN(str) == 0 || !RSTRING_PTR(str)) return str;
    if (flags&ONIGENC_CASE_ASCII_ONLY) {
        ret = rb_str_new(0, RSTRING_LEN(str));
        rb_str_ascii_casemap(str, ret, &flags, enc);
    }
    else {
        ret = rb_str_casemap(str, &flags, enc);
    }
    return ret;
}

Returns a string containing the characters in self; the first character is upcased; the remaining characters are downcased:

s = 'hello World!' 
s.capitalize       

The casing may be affected by the given options; see Case Mapping.

Related: String#capitalize!.

static VALUE
rb_str_capitalize_bang(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE | ONIGENC_CASE_TITLECASE;

    flags = check_case_options(argc, argv, flags);
    str_modify_keep_cr(str);
    enc = str_true_enc(str);
    if (RSTRING_LEN(str) == 0 || !RSTRING_PTR(str)) return Qnil;
    if (flags&ONIGENC_CASE_ASCII_ONLY)
        rb_str_ascii_casemap(str, str, &flags, enc);
    else
        str_shared_replace(str, rb_str_casemap(str, &flags, enc));

    if (ONIGENC_CASE_MODIFIED&flags) return str;
    return Qnil;
}

Upcases the first character in self; downcases the remaining characters; returns self if any changes were made, nil otherwise:

s = 'hello World!' 
s.capitalize!      
s                  
s.capitalize!      

The casing may be affected by the given options; see Case Mapping.

Related: String#capitalize.

static VALUE
rb_str_casecmp(VALUE str1, VALUE str2)
{
    VALUE s = rb_check_string_type(str2);
    if (NIL_P(s)) {
        return Qnil;
    }
    return str_casecmp(str1, s);
}

Compares self.downcase and other_string.downcase; returns:

• -1 if other_string.downcase is larger.

• 0 if the two are equal.

• 1 if other_string.downcase is smaller.

• nil if the two are incomparable.

Examples:

'foo'.casecmp('foo') 
'foo'.casecmp('food') 
'food'.casecmp('foo') 
'FOO'.casecmp('foo') 
'foo'.casecmp('FOO') 
'foo'.casecmp(1) 

See Case Mapping.

Related: String#casecmp?.

static VALUE
rb_str_casecmp_p(VALUE str1, VALUE str2)
{
    VALUE s = rb_check_string_type(str2);
    if (NIL_P(s)) {
        return Qnil;
    }
    return str_casecmp_p(str1, s);
}

Returns true if self and other_string are equal after Unicode case folding, otherwise false:

'foo'.casecmp?('foo') 
'foo'.casecmp?('food') 
'food'.casecmp?('foo') 
'FOO'.casecmp?('foo') 
'foo'.casecmp?('FOO') 

Returns nil if the two values are incomparable:

'foo'.casecmp?(1) 

See Case Mapping.

Related: String#casecmp.

static VALUE
rb_str_center(int argc, VALUE *argv, VALUE str)
{
    return rb_str_justify(argc, argv, str, 'c');
}

Returns a centered copy of self.

If integer argument size is greater than the size (in characters) of self, returns a new string of length size that is a copy of self, centered and padded on both ends with pad_string:

'hello'.center(10)       
'  hello'.center(10)     
'hello'.center(10, 'ab') 
'тест'.center(10)        
'こんにちは'.center(10)    

If size is not greater than the size of self, returns a copy of self:

'hello'.center(5)  
'hello'.center(1)  

Related: String#ljust, String#rjust.

static VALUE
rb_str_chars(VALUE str)
{
    VALUE ary = WANTARRAY("chars", rb_str_strlen(str));
    return rb_str_enumerate_chars(str, ary);
}

Returns an array of the characters in self:

'hello'.chars     
'тест'.chars      
'こんにちは'.chars 

static VALUE
rb_str_chomp(int argc, VALUE *argv, VALUE str)
{
    VALUE rs = chomp_rs(argc, argv);
    if (NIL_P(rs)) return str_duplicate(rb_cString, str);
    return rb_str_subseq(str, 0, chompped_length(str, rs));
}

Returns a new string copied from self, with trailing characters possibly removed:

When line_sep is "\n", removes the last one or two characters if they are "\r", "\n", or "\r\n" (but not "\n\r"):

$/                    
"abc\r".chomp         
"abc\n".chomp         
"abc\r\n".chomp       
"abc\n\r".chomp       
"тест\r\n".chomp      
"こんにちは\r\n".chomp  

When line_sep is '' (an empty string), removes multiple trailing occurrences of "\n" or "\r\n" (but not "\r" or "\n\r"):

"abc\n\n\n".chomp('')           
"abc\r\n\r\n\r\n".chomp('')     
"abc\n\n\r\n\r\n\n\n".chomp('') 
"abc\n\r\n\r\n\r".chomp('')     
"abc\r\r\r".chomp('')           

When line_sep is neither "\n" nor '', removes a single trailing line separator if there is one:

'abcd'.chomp('d')  
'abcdd'.chomp('d') 

static VALUE
rb_str_chomp_bang(int argc, VALUE *argv, VALUE str)
{
    VALUE rs;
    str_modifiable(str);
    if (RSTRING_LEN(str) == 0 && argc < 2) return Qnil;
    rs = chomp_rs(argc, argv);
    if (NIL_P(rs)) return Qnil;
    return rb_str_chomp_string(str, rs);
}

Like String#chomp, but modifies self in place; returns nil if no modification made, self otherwise.

static VALUE
rb_str_chop(VALUE str)
{
    return rb_str_subseq(str, 0, chopped_length(str));
}

Returns a new string copied from self, with trailing characters possibly removed.

Removes "\r\n" if those are the last two characters.

"abc\r\n".chop      
"тест\r\n".chop     
"こんにちは\r\n".chop 

Otherwise removes the last character if it exists.

'abcd'.chop     
'тест'.chop     
'こんにちは'.chop 
''.chop         

If you only need to remove the newline separator at the end of the string, String#chomp is a better alternative.

static VALUE
rb_str_chop_bang(VALUE str)
{
    str_modify_keep_cr(str);
    if (RSTRING_LEN(str) > 0) {
        long len;
        len = chopped_length(str);
        STR_SET_LEN(str, len);
        TERM_FILL(&RSTRING_PTR(str)[len], TERM_LEN(str));
        if (ENC_CODERANGE(str) != ENC_CODERANGE_7BIT) {
            ENC_CODERANGE_CLEAR(str);
        }
        return str;
    }
    return Qnil;
}

Like String#chop, but modifies self in place; returns nil if self is empty, self otherwise.

Related: String#chomp!.

static VALUE
rb_str_chr(VALUE str)
{
    return rb_str_substr(str, 0, 1);
}

Returns a string containing the first character of self:

s = 'foo' 
s.chr     

static VALUE
rb_str_clear(VALUE str)
{
    str_discard(str);
    STR_SET_EMBED(str);
    STR_SET_LEN(str, 0);
    RSTRING_PTR(str)[0] = 0;
    if (rb_enc_asciicompat(STR_ENC_GET(str)))
        ENC_CODERANGE_SET(str, ENC_CODERANGE_7BIT);
    else
        ENC_CODERANGE_SET(str, ENC_CODERANGE_VALID);
    return str;
}

Removes the contents of self:

s = 'foo' 
s.clear   

static VALUE
rb_str_codepoints(VALUE str)
{
    VALUE ary = WANTARRAY("codepoints", rb_str_strlen(str));
    return rb_str_enumerate_codepoints(str, ary);
}

Returns an array of the codepoints in self; each codepoint is the integer value for a character:

'hello'.codepoints     
'тест'.codepoints      
'こんにちは'.codepoints 

static VALUE
rb_str_concat_multi(int argc, VALUE *argv, VALUE str)
{
    str_modifiable(str);

    if (argc == 1) {
        return rb_str_concat(str, argv[0]);
    }
    else if (argc > 1) {
        int i;
        VALUE arg_str = rb_str_tmp_new(0);
        rb_enc_copy(arg_str, str);
        for (i = 0; i < argc; i++) {
            rb_str_concat(arg_str, argv[i]);
        }
        rb_str_buf_append(str, arg_str);
    }

    return str;
}

Concatenates each object in objects to self and returns self:

s = 'foo'
s.concat('bar', 'baz') 
s                      

For each given object object that is an Integer, the value is considered a codepoint and converted to a character before concatenation:

s = 'foo'
s.concat(32, 'bar', 32, 'baz') 

Related: String#<<, which takes a single argument.

static VALUE
rb_str_count(int argc, VALUE *argv, VALUE str)
{
    char table[TR_TABLE_SIZE];
    rb_encoding *enc = 0;
    VALUE del = 0, nodel = 0, tstr;
    char *s, *send;
    int i;
    int ascompat;
    size_t n = 0;

    rb_check_arity(argc, 1, UNLIMITED_ARGUMENTS);

    tstr = argv[0];
    StringValue(tstr);
    enc = rb_enc_check(str, tstr);
    if (argc == 1) {
        const char *ptstr;
        if (RSTRING_LEN(tstr) == 1 && rb_enc_asciicompat(enc) &&
            (ptstr = RSTRING_PTR(tstr),
             ONIGENC_IS_ALLOWED_REVERSE_MATCH(enc, (const unsigned char *)ptstr, (const unsigned char *)ptstr+1)) &&
            !is_broken_string(str)) {
            int clen;
            unsigned char c = rb_enc_codepoint_len(ptstr, ptstr+1, &clen, enc);

            s = RSTRING_PTR(str);
            if (!s || RSTRING_LEN(str) == 0) return INT2FIX(0);
            send = RSTRING_END(str);
            while (s < send) {
                if (*(unsigned char*)s++ == c) n++;
            }
            return SIZET2NUM(n);
        }
    }

    tr_setup_table(tstr, table, TRUE, &del, &nodel, enc);
    for (i=1; i<argc; i++) {
        tstr = argv[i];
        StringValue(tstr);
        enc = rb_enc_check(str, tstr);
        tr_setup_table(tstr, table, FALSE, &del, &nodel, enc);
    }

    s = RSTRING_PTR(str);
    if (!s || RSTRING_LEN(str) == 0) return INT2FIX(0);
    send = RSTRING_END(str);
    ascompat = rb_enc_asciicompat(enc);
    while (s < send) {
        unsigned int c;

        if (ascompat && (c = *(unsigned char*)s) < 0x80) {
            if (table[c]) {
                n++;
            }
            s++;
        }
        else {
            int clen;
            c = rb_enc_codepoint_len(s, send, &clen, enc);
            if (tr_find(c, table, del, nodel)) {
                n++;
            }
            s += clen;
        }
    }

    return SIZET2NUM(n);
}

Returns the total number of characters in self that are specified by the given selectors (see Multiple Character Selectors):

a = "hello world"
a.count "lo"                   
a.count "lo", "o"              
a.count "hello", "^l"          
a.count "ej-m"                 

"hello^world".count "\\^aeiou" 
"hello-world".count "a\\-eo"   

c = "hello world\\r\\n"
c.count "\\"                   
c.count "\\A"                  
c.count "X-\\w"                

static VALUE
rb_str_crypt(VALUE str, VALUE salt)
{
#ifdef HAVE_CRYPT_R
    VALUE databuf;
    struct crypt_data *data;
#   define CRYPT_END() ALLOCV_END(databuf)
#else
    extern char *crypt(const char *, const char *);
#   define CRYPT_END() rb_nativethread_lock_unlock(&crypt_mutex.lock)
#endif
    VALUE result;
    const char *s, *saltp;
    char *res;
#ifdef BROKEN_CRYPT
    char salt_8bit_clean[3];
#endif

    StringValue(salt);
    mustnot_wchar(str);
    mustnot_wchar(salt);
    s = StringValueCStr(str);
    saltp = RSTRING_PTR(salt);
    if (RSTRING_LEN(salt) < 2 || !saltp[0] || !saltp[1]) {
        rb_raise(rb_eArgError, "salt too short (need >=2 bytes)");
    }

#ifdef BROKEN_CRYPT
    if (!ISASCII((unsigned char)saltp[0]) || !ISASCII((unsigned char)saltp[1])) {
        salt_8bit_clean[0] = saltp[0] & 0x7f;
        salt_8bit_clean[1] = saltp[1] & 0x7f;
        salt_8bit_clean[2] = '\0';
        saltp = salt_8bit_clean;
    }
#endif
#ifdef HAVE_CRYPT_R
    data = ALLOCV(databuf, sizeof(struct crypt_data));
# ifdef HAVE_STRUCT_CRYPT_DATA_INITIALIZED
    data->initialized = 0;
# endif
    res = crypt_r(s, saltp, data);
#else
    crypt_mutex_initialize();
    rb_nativethread_lock_lock(&crypt_mutex.lock);
    res = crypt(s, saltp);
#endif
    if (!res) {
        int err = errno;
        CRYPT_END();
        rb_syserr_fail(err, "crypt");
    }
    result = rb_str_new_cstr(res);
    CRYPT_END();
    return result;
}

Returns the string generated by calling crypt(3) standard library function with str and salt_str, in this order, as its arguments. Please do not use this method any longer. It is legacy; provided only for backward compatibility with ruby scripts in earlier days. It is bad to use in contemporary programs for several reasons:

• Behaviour of C’s crypt(3) depends on the OS it is run. The generated string lacks data portability.

• On some OSes such as Mac OS, crypt(3) never fails (i.e. silently ends up in unexpected results).

• On some OSes such as Mac OS, crypt(3) is not thread safe.

• So-called “traditional” usage of crypt(3) is very very very weak. According to its manpage, Linux’s traditional crypt(3) output has only 2**56 variations; too easy to brute force today. And this is the default behaviour.

• In order to make things robust some OSes implement so-called “modular” usage. To go through, you have to do a complex build-up of the salt_str parameter, by hand. Failure in generation of a proper salt string tends not to yield any errors; typos in parameters are normally not detectable.

  • For instance, in the following example, the second invocation of String#crypt is wrong; it has a typo in “round=” (lacks “s”). However the call does not fail and something unexpected is generated.

    "foo".crypt("$5$rounds=1000$salt$") 
    "foo".crypt("$5$round=1000$salt$")  
    

• Even in the “modular” mode, some hash functions are considered archaic and no longer recommended at all; for instance module $1$ is officially abandoned by its author: see phk.freebsd.dk/sagas/md5crypt_eol/ . For another instance module $3$ is considered completely broken: see the manpage of FreeBSD.

• On some OS such as Mac OS, there is no modular mode. Yet, as written above, crypt(3) on Mac OS never fails. This means even if you build up a proper salt string it generates a traditional DES hash anyways, and there is no way for you to be aware of.

  "foo".crypt("$5$rounds=1000$salt$") 
  

If for some reason you cannot migrate to other secure contemporary password hashing algorithms, install the string-crypt gem and require 'string/crypt' to continue using it.

Returns a frozen, possibly pre-existing copy of the string.

The returned String will be deduplicated as long as it does not have any instance variables set on it and is not a String subclass.

Note that -string variant is more convenient for defining constants:

FILENAME = -'config/database.yml'

while dedup is better suitable for using the method in chains of calculations:

@url_list.concat(urls.map(&:dedup))

static VALUE
rb_str_delete(int argc, VALUE *argv, VALUE str)
{
    str = str_duplicate(rb_cString, str);
    rb_str_delete_bang(argc, argv, str);
    return str;
}

Returns a copy of self with characters specified by selectors removed (see Multiple Character Selectors):

"hello".delete "l","lo"        
"hello".delete "lo"            
"hello".delete "aeiou", "^e"   
"hello".delete "ej-m"          

static VALUE
rb_str_delete_bang(int argc, VALUE *argv, VALUE str)
{
    char squeez[TR_TABLE_SIZE];
    rb_encoding *enc = 0;
    char *s, *send, *t;
    VALUE del = 0, nodel = 0;
    int modify = 0;
    int i, ascompat, cr;

    if (RSTRING_LEN(str) == 0 || !RSTRING_PTR(str)) return Qnil;
    rb_check_arity(argc, 1, UNLIMITED_ARGUMENTS);
    for (i=0; i<argc; i++) {
        VALUE s = argv[i];

        StringValue(s);
        enc = rb_enc_check(str, s);
        tr_setup_table(s, squeez, i==0, &del, &nodel, enc);
    }

    str_modify_keep_cr(str);
    ascompat = rb_enc_asciicompat(enc);
    s = t = RSTRING_PTR(str);
    send = RSTRING_END(str);
    cr = ascompat ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
    while (s < send) {
        unsigned int c;
        int clen;

        if (ascompat && (c = *(unsigned char*)s) < 0x80) {
            if (squeez[c]) {
                modify = 1;
            }
            else {
                if (t != s) *t = c;
                t++;
            }
            s++;
        }
        else {
            c = rb_enc_codepoint_len(s, send, &clen, enc);

            if (tr_find(c, squeez, del, nodel)) {
                modify = 1;
            }
            else {
                if (t != s) rb_enc_mbcput(c, t, enc);
                t += clen;
                if (cr == ENC_CODERANGE_7BIT) cr = ENC_CODERANGE_VALID;
            }
            s += clen;
        }
    }
    TERM_FILL(t, TERM_LEN(str));
    STR_SET_LEN(str, t - RSTRING_PTR(str));
    ENC_CODERANGE_SET(str, cr);

    if (modify) return str;
    return Qnil;
}

Like String#delete, but modifies self in place. Returns self if any changes were made, nil otherwise.

static VALUE
rb_str_delete_prefix(VALUE str, VALUE prefix)
{
    long prefixlen;

    prefixlen = deleted_prefix_length(str, prefix);
    if (prefixlen <= 0) return str_duplicate(rb_cString, str);

    return rb_str_subseq(str, prefixlen, RSTRING_LEN(str) - prefixlen);
}

Returns a copy of self with leading substring prefix removed:

'hello'.delete_prefix('hel')      
'hello'.delete_prefix('llo')      
'тест'.delete_prefix('те')        
'こんにちは'.delete_prefix('こん')  

Related: String#delete_prefix!, String#delete_suffix.

static VALUE
rb_str_delete_prefix_bang(VALUE str, VALUE prefix)
{
    long prefixlen;
    str_modify_keep_cr(str);

    prefixlen = deleted_prefix_length(str, prefix);
    if (prefixlen <= 0) return Qnil;

    return rb_str_drop_bytes(str, prefixlen);
}

Like String#delete_prefix, except that self is modified in place. Returns self if the prefix is removed, nil otherwise.

static VALUE
rb_str_delete_suffix(VALUE str, VALUE suffix)
{
    long suffixlen;

    suffixlen = deleted_suffix_length(str, suffix);
    if (suffixlen <= 0) return str_duplicate(rb_cString, str);

    return rb_str_subseq(str, 0, RSTRING_LEN(str) - suffixlen);
}

Returns a copy of self with trailing substring suffix removed:

'hello'.delete_suffix('llo')      
'hello'.delete_suffix('hel')      
'тест'.delete_suffix('ст')        
'こんにちは'.delete_suffix('ちは')  

Related: String#delete_suffix!, String#delete_prefix.

static VALUE
rb_str_delete_suffix_bang(VALUE str, VALUE suffix)
{
    long olen, suffixlen, len;
    str_modifiable(str);

    suffixlen = deleted_suffix_length(str, suffix);
    if (suffixlen <= 0) return Qnil;

    olen = RSTRING_LEN(str);
    str_modify_keep_cr(str);
    len = olen - suffixlen;
    STR_SET_LEN(str, len);
    TERM_FILL(&RSTRING_PTR(str)[len], TERM_LEN(str));
    if (ENC_CODERANGE(str) != ENC_CODERANGE_7BIT) {
        ENC_CODERANGE_CLEAR(str);
    }
    return str;
}

Like String#delete_suffix, except that self is modified in place. Returns self if the suffix is removed, nil otherwise.

static VALUE
rb_str_downcase(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_DOWNCASE;
    VALUE ret;

    flags = check_case_options(argc, argv, flags);
    enc = str_true_enc(str);
    if (case_option_single_p(flags, enc, str)) {
        ret = rb_str_new(RSTRING_PTR(str), RSTRING_LEN(str));
        str_enc_copy_direct(ret, str);
        downcase_single(ret);
    }
    else if (flags&ONIGENC_CASE_ASCII_ONLY) {
        ret = rb_str_new(0, RSTRING_LEN(str));
        rb_str_ascii_casemap(str, ret, &flags, enc);
    }
    else {
        ret = rb_str_casemap(str, &flags, enc);
    }

    return ret;
}

Returns a string containing the downcased characters in self:

s = 'Hello World!' 
s.downcase         

The casing may be affected by the given options; see Case Mapping.

Related: String#downcase!, String#upcase, String#upcase!.

static VALUE
rb_str_downcase_bang(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_DOWNCASE;

    flags = check_case_options(argc, argv, flags);
    str_modify_keep_cr(str);
    enc = str_true_enc(str);
    if (case_option_single_p(flags, enc, str)) {
        if (downcase_single(str))
            flags |= ONIGENC_CASE_MODIFIED;
    }
    else if (flags&ONIGENC_CASE_ASCII_ONLY)
        rb_str_ascii_casemap(str, str, &flags, enc);
    else
        str_shared_replace(str, rb_str_casemap(str, &flags, enc));

    if (ONIGENC_CASE_MODIFIED&flags) return str;
    return Qnil;
}

Downcases the characters in self; returns self if any changes were made, nil otherwise:

s = 'Hello World!' 
s.downcase!        
s                  
s.downcase!        

The casing may be affected by the given options; see Case Mapping.

Related: String#downcase, String#upcase, String#upcase!.

VALUE
rb_str_dump(VALUE str)
{
    int encidx = rb_enc_get_index(str);
    rb_encoding *enc = rb_enc_from_index(encidx);
    long len;
    const char *p, *pend;
    char *q, *qend;
    VALUE result;
    int u8 = (encidx == rb_utf8_encindex());
    static const char nonascii_suffix[] = ".dup.force_encoding(\"%s\")";

    len = 2;                    /* "" */
    if (!rb_enc_asciicompat(enc)) {
        len += strlen(nonascii_suffix) - rb_strlen_lit("%s");
        len += strlen(enc->name);
    }

    p = RSTRING_PTR(str); pend = p + RSTRING_LEN(str);
    while (p < pend) {
        int clen;
        unsigned char c = *p++;

        switch (c) {
          case '"':  case '\\':
          case '\n': case '\r':
          case '\t': case '\f':
          case '\013': case '\010': case '\007': case '\033':
            clen = 2;
            break;

          case '#':
            clen = IS_EVSTR(p, pend) ? 2 : 1;
            break;

          default:
            if (ISPRINT(c)) {
                clen = 1;
            }
            else {
                if (u8 && c > 0x7F) {   /* \u notation */
                    int n = rb_enc_precise_mbclen(p-1, pend, enc);
                    if (MBCLEN_CHARFOUND_P(n)) {
                        unsigned int cc = rb_enc_mbc_to_codepoint(p-1, pend, enc);
                        if (cc <= 0xFFFF)
                            clen = 6;  /* \uXXXX */
                        else if (cc <= 0xFFFFF)
                            clen = 9;  /* \u{XXXXX} */
                        else
                            clen = 10; /* \u{XXXXXX} */
                        p += MBCLEN_CHARFOUND_LEN(n)-1;
                        break;
                    }
                }
                clen = 4;       /* \xNN */
            }
            break;
        }

        if (clen > LONG_MAX - len) {
            rb_raise(rb_eRuntimeError, "string size too big");
        }
        len += clen;
    }

    result = rb_str_new(0, len);
    p = RSTRING_PTR(str); pend = p + RSTRING_LEN(str);
    q = RSTRING_PTR(result); qend = q + len + 1;

    *q++ = '"';
    while (p < pend) {
        unsigned char c = *p++;

        if (c == '"' || c == '\\') {
            *q++ = '\\';
            *q++ = c;
        }
        else if (c == '#') {
            if (IS_EVSTR(p, pend)) *q++ = '\\';
            *q++ = '#';
        }
        else if (c == '\n') {
            *q++ = '\\';
            *q++ = 'n';
        }
        else if (c == '\r') {
            *q++ = '\\';
            *q++ = 'r';
        }
        else if (c == '\t') {
            *q++ = '\\';
            *q++ = 't';
        }
        else if (c == '\f') {
            *q++ = '\\';
            *q++ = 'f';
        }
        else if (c == '\013') {
            *q++ = '\\';
            *q++ = 'v';
        }
        else if (c == '\010') {
            *q++ = '\\';
            *q++ = 'b';
        }
        else if (c == '\007') {
            *q++ = '\\';
            *q++ = 'a';
        }
        else if (c == '\033') {
            *q++ = '\\';
            *q++ = 'e';
        }
        else if (ISPRINT(c)) {
            *q++ = c;
        }
        else {
            *q++ = '\\';
            if (u8) {
                int n = rb_enc_precise_mbclen(p-1, pend, enc) - 1;
                if (MBCLEN_CHARFOUND_P(n)) {
                    int cc = rb_enc_mbc_to_codepoint(p-1, pend, enc);
                    p += n;
                    if (cc <= 0xFFFF)
                        snprintf(q, qend-q, "u%04X", cc);    /* \uXXXX */
                    else
                        snprintf(q, qend-q, "u{%X}", cc);  /* \u{XXXXX} or \u{XXXXXX} */
                    q += strlen(q);
                    continue;
                }
            }
            snprintf(q, qend-q, "x%02X", c);
            q += 3;
        }
    }
    *q++ = '"';
    *q = '\0';
    if (!rb_enc_asciicompat(enc)) {
        snprintf(q, qend-q, nonascii_suffix, enc->name);
        encidx = rb_ascii8bit_encindex();
    }
    /* result from dump is ASCII */
    rb_enc_associate_index(result, encidx);
    ENC_CODERANGE_SET(result, ENC_CODERANGE_7BIT);
    return result;
}

Returns a printable version of self, enclosed in double-quotes, with special characters escaped, and with non-printing characters replaced by hexadecimal notation:

"hello \n ''".dump    
"\f\x00\xff\\\"".dump 

Related: String#undump (inverse of String#dump).

static VALUE
rb_str_each_byte(VALUE str)
{
    RETURN_SIZED_ENUMERATOR(str, 0, 0, rb_str_each_byte_size);
    return rb_str_enumerate_bytes(str, 0);
}

Calls the given block with each successive byte from self; returns self:

'hello'.each_byte {|byte| print byte, ' ' }
print "\n"
'тест'.each_byte {|byte| print byte, ' ' }
print "\n"
'こんにちは'.each_byte {|byte| print byte, ' ' }
print "\n"

Output:

104 101 108 108 111
209 130 208 181 209 129 209 130
227 129 147 227 130 147 227 129 171 227 129 161 227 129 175

Returns an enumerator if no block is given.

static VALUE
rb_str_each_char(VALUE str)
{
    RETURN_SIZED_ENUMERATOR(str, 0, 0, rb_str_each_char_size);
    return rb_str_enumerate_chars(str, 0);
}

Calls the given block with each successive character from self; returns self:

'hello'.each_char {|char| print char, ' ' }
print "\n"
'тест'.each_char {|char| print char, ' ' }
print "\n"
'こんにちは'.each_char {|char| print char, ' ' }
print "\n"

Output:

h e l l o
т е с т
こ ん に ち は

Returns an enumerator if no block is given.

static VALUE
rb_str_each_codepoint(VALUE str)
{
    RETURN_SIZED_ENUMERATOR(str, 0, 0, rb_str_each_char_size);
    return rb_str_enumerate_codepoints(str, 0);
}

Calls the given block with each successive codepoint from self; each codepoint is the integer value for a character; returns self:

'hello'.each_codepoint {|codepoint| print codepoint, ' ' }
print "\n"
'тест'.each_codepoint {|codepoint| print codepoint, ' ' }
print "\n"
'こんにちは'.each_codepoint {|codepoint| print codepoint, ' ' }
print "\n"

Output:

104 101 108 108 111
1090 1077 1089 1090
12371 12435 12395 12385 12399

Returns an enumerator if no block is given.

static VALUE
rb_str_each_grapheme_cluster(VALUE str)
{
    RETURN_SIZED_ENUMERATOR(str, 0, 0, rb_str_each_grapheme_cluster_size);
    return rb_str_enumerate_grapheme_clusters(str, 0);
}

Calls the given block with each successive grapheme cluster from self (see Unicode Grapheme Cluster Boundaries); returns self:

s = "\u0061\u0308-pqr-\u0062\u0308-xyz-\u0063\u0308" 
s.each_grapheme_cluster {|gc| print gc, ' ' }

Output:

ä - p q r - b̈ - x y z - c̈

Returns an enumerator if no block is given.

static VALUE
rb_str_each_line(int argc, VALUE *argv, VALUE str)
{
    RETURN_SIZED_ENUMERATOR(str, argc, argv, 0);
    return rb_str_enumerate_lines(argc, argv, str, 0);
}

With a block given, forms the substrings (“lines”) that are the result of splitting self at each occurrence of the given line separator line_sep; passes each line to the block; returns self:

s = <<~EOT
This is the first line.
This is line two.

This is line four.
This is line five.
EOT

s.each_line {|line| p line }

Output:

"This is the first line.\n"
"This is line two.\n"
"\n"
"This is line four.\n"
"This is line five.\n"

With a different line_sep:

s.each_line(' is ') {|line| p line }

Output:

"This is "
"the first line.\nThis is "
"line two.\n\nThis is "
"line four.\nThis is "
"line five.\n"

With chomp as true, removes the trailing line_sep from each line:

s.each_line(chomp: true) {|line| p line }

Output:

"This is the first line."
"This is line two."
""
"This is line four."
"This is line five."

With an empty string as line_sep, forms and passes “paragraphs” by splitting at each occurrence of two or more newlines:

s.each_line('') {|line| p line }

Output:

"This is the first line.\nThis is line two.\n\n"
"This is line four.\nThis is line five.\n"

With no block given, returns an enumerator.

static VALUE
rb_str_empty(VALUE str)
{
    return RBOOL(RSTRING_LEN(str) == 0);
}

Returns true if the length of self is zero, false otherwise:

"hello".empty? 
" ".empty? 
"".empty? 

static VALUE
str_encode(int argc, VALUE *argv, VALUE str)
{
    VALUE newstr = str;
    int encidx = str_transcode(argc, argv, &newstr);
    return encoded_dup(newstr, str, encidx);
}

Returns a copy of self transcoded as determined by dst_encoding. By default, raises an exception if self contains an invalid byte or a character not defined in dst_encoding; that behavior may be modified by encoding options; see below.

With no arguments:

• Uses the same encoding if Encoding.default_internal is nil (the default):

  Encoding.default_internal 
  s = "Ruby\x99".force_encoding('Windows-1252')
  s.encoding                
  s.bytes                   
  t = s.encode              
  t.encoding                
  t.bytes                   
  

• Otherwise, uses the encoding Encoding.default_internal:

  Encoding.default_internal = 'UTF-8'
  t = s.encode              
  t.encoding                
  

With only argument dst_encoding given, uses that encoding:

s = "Ruby\x99".force_encoding('Windows-1252')
s.encoding            
t = s.encode('UTF-8') 
t.encoding            

With arguments dst_encoding and src_encoding given, interprets self using src_encoding, encodes the new string using dst_encoding:

s = "Ruby\x99"
t = s.encode('UTF-8', 'Windows-1252') 
t.encoding                            

Optional keyword arguments enc_opts specify encoding options; see Encoding Options.

Please note that, unless invalid: :replace option is given, conversion from an encoding enc to the same encoding enc (independent of whether enc is given explicitly or implicitly) is a no-op, i.e. the string is simply copied without any changes, and no exceptions are raised, even if there are invalid bytes.

static VALUE
str_encode_bang(int argc, VALUE *argv, VALUE str)
{
    VALUE newstr;
    int encidx;

    rb_check_frozen(str);

    newstr = str;
    encidx = str_transcode(argc, argv, &newstr);

    if (encidx < 0) return str;
    if (newstr == str) {
        rb_enc_associate_index(str, encidx);
        return str;
    }
    rb_str_shared_replace(str, newstr);
    return str_encode_associate(str, encidx);
}

Like encode, but applies encoding changes to self; returns self.

VALUE
rb_obj_encoding(VALUE obj)
{
    int idx = rb_enc_get_index(obj);
    if (idx < 0) {
        rb_raise(rb_eTypeError, "unknown encoding");
    }
    return rb_enc_from_encoding_index(idx & ENC_INDEX_MASK);
}

Returns the Encoding object that represents the encoding of obj.

static VALUE
rb_str_end_with(int argc, VALUE *argv, VALUE str)
{
    int i;

    for (i=0; i<argc; i++) {
        VALUE tmp = argv[i];
        const char *p, *s, *e;
        long slen, tlen;
        rb_encoding *enc;

        StringValue(tmp);
        enc = rb_enc_check(str, tmp);
        if ((tlen = RSTRING_LEN(tmp)) == 0) return Qtrue;
        if ((slen = RSTRING_LEN(str)) < tlen) continue;
        p = RSTRING_PTR(str);
        e = p + slen;
        s = e - tlen;
        if (!at_char_boundary(p, s, e, enc))
            continue;
        if (memcmp(s, RSTRING_PTR(tmp), tlen) == 0)
            return Qtrue;
    }
    return Qfalse;
}

Returns whether self ends with any of the given strings.

Returns true if any given string matches the end, false otherwise:

'hello'.end_with?('ello')               
'hello'.end_with?('heaven', 'ello')     
'hello'.end_with?('heaven', 'paradise') 
'тест'.end_with?('т')                   
'こんにちは'.end_with?('は')              

Related: String#start_with?.

VALUE
rb_str_eql(VALUE str1, VALUE str2)
{
    if (str1 == str2) return Qtrue;
    if (!RB_TYPE_P(str2, T_STRING)) return Qfalse;
    return rb_str_eql_internal(str1, str2);
}

Returns true if object has the same length and content; as self; false otherwise:

s = 'foo'
s.eql?('foo') 
s.eql?('food') 
s.eql?('FOO') 

Returns false if the two strings’ encodings are not compatible:

"\u{e4 f6 fc}".encode("ISO-8859-1").eql?("\u{c4 d6 dc}") 

static VALUE
rb_str_force_encoding(VALUE str, VALUE enc)
{
    str_modifiable(str);

    rb_encoding *encoding = rb_to_encoding(enc);
    int idx = rb_enc_to_index(encoding);

    // If the encoding is unchanged, we do nothing.
    if (ENCODING_GET(str) == idx) {
        return str;
    }

    rb_enc_associate_index(str, idx);

    // If the coderange was 7bit and the new encoding is ASCII-compatible
    // we can keep the coderange.
    if (ENC_CODERANGE(str) == ENC_CODERANGE_7BIT && encoding && rb_enc_asciicompat(encoding)) {
        return str;
    }

    ENC_CODERANGE_CLEAR(str);
    return str;
}

Changes the encoding of self to encoding, which may be a string encoding name or an Encoding object; returns self:

s = 'łał'
s.bytes                   
s.encoding                
s.force_encoding('ascii') 
s.encoding                

Does not change the underlying bytes:

s.bytes                   

Makes the change even if the given encoding is invalid for self (as is the change above):

s.valid_encoding?                 
s.force_encoding(Encoding::UTF_8) 
s.valid_encoding?                 

VALUE
rb_str_getbyte(VALUE str, VALUE index)
{
    long pos = NUM2LONG(index);

    if (pos < 0)
        pos += RSTRING_LEN(str);
    if (pos < 0 ||  RSTRING_LEN(str) <= pos)
        return Qnil;

    return INT2FIX((unsigned char)RSTRING_PTR(str)[pos]);
}

Returns the byte at zero-based index as an integer, or nil if index is out of range:

s = 'abcde'   
s.getbyte(0)  
s.getbyte(-1) 
s.getbyte(5)  

Related: String#setbyte.

static VALUE
rb_str_grapheme_clusters(VALUE str)
{
    VALUE ary = WANTARRAY("grapheme_clusters", rb_str_strlen(str));
    return rb_str_enumerate_grapheme_clusters(str, ary);
}

Returns an array of the grapheme clusters in self (see Unicode Grapheme Cluster Boundaries):

s = "\u0061\u0308-pqr-\u0062\u0308-xyz-\u0063\u0308" 
s.grapheme_clusters

static VALUE
rb_str_gsub_bang(int argc, VALUE *argv, VALUE str)
{
    str_modify_keep_cr(str);
    return str_gsub(argc, argv, str, 1);
}

Performs the specified substring replacement(s) on self; returns self if any replacement occurred, nil otherwise.

See Substitution Methods.

Returns an Enumerator if no replacement and no block given.

Related: String#sub, String#gsub, String#sub!.

static VALUE
rb_str_hash_m(VALUE str)
{
    st_index_t hval = rb_str_hash(str);
    return ST2FIX(hval);
}

Returns the integer hash value for self. The value is based on the length, content and encoding of self.

Related: Object#hash.

static VALUE
rb_str_hex(VALUE str)
{
    return rb_str_to_inum(str, 16, FALSE);
}

Interprets the leading substring of self as a string of hexadecimal digits (with an optional sign and an optional 0x) and returns the corresponding number; returns zero if there is no such leading substring:

'0x0a'.hex        
'-1234'.hex       
'0'.hex           
'non-numeric'.hex 

Related: String#oct.

VALUE
rb_str_include(VALUE str, VALUE arg)
{
    long i;

    StringValue(arg);
    i = rb_str_index(str, arg, 0);

    return RBOOL(i != -1);
}

Returns true if self contains other_string, false otherwise:

s = 'foo'
s.include?('f')    
s.include?('fo')   
s.include?('food') 

static VALUE
rb_str_index_m(int argc, VALUE *argv, VALUE str)
{
    VALUE sub;
    VALUE initpos;
    rb_encoding *enc = STR_ENC_GET(str);
    long pos;

    if (rb_scan_args(argc, argv, "11", &sub, &initpos) == 2) {
        long slen = str_strlen(str, enc); /* str's enc */
        pos = NUM2LONG(initpos);
        if (pos < 0 ? (pos += slen) < 0 : pos > slen) {
            if (RB_TYPE_P(sub, T_REGEXP)) {
                rb_backref_set(Qnil);
            }
            return Qnil;
        }
    }
    else {
        pos = 0;
    }

    if (RB_TYPE_P(sub, T_REGEXP)) {
        pos = str_offset(RSTRING_PTR(str), RSTRING_END(str), pos,
                         enc, single_byte_optimizable(str));

        if (rb_reg_search(sub, str, pos, 0) >= 0) {
            VALUE match = rb_backref_get();
            struct re_registers *regs = RMATCH_REGS(match);
            pos = rb_str_sublen(str, BEG(0));
            return LONG2NUM(pos);
        }
    }
    else {
        StringValue(sub);
        pos = rb_str_index(str, sub, pos);
        if (pos >= 0) {
            pos = rb_str_sublen(str, pos);
            return LONG2NUM(pos);
        }
    }
    return Qnil;
}

Returns the integer index of the first match for the given argument, or nil if none found; the search of self is forward, and begins at position offset (in characters).

With string argument substring, returns the index of the first matching substring in self:

'foo'.index('f')         
'foo'.index('o')         
'foo'.index('oo')        
'foo'.index('ooo')       
'тест'.index('с')        
'こんにちは'.index('ち')   

With Regexp argument regexp, returns the index of the first match in self:

'foo'.index(/o./) 
'foo'.index(/.o/) 

With positive integer offset, begins the search at position offset:

'foo'.index('o', 1)        
'foo'.index('o', 2)        
'foo'.index('o', 3)        
'тест'.index('с', 1)       
'こんにちは'.index('ち', 2)  

With negative integer offset, selects the search position by counting backward from the end of self:

'foo'.index('o', -1)  
'foo'.index('o', -2)  
'foo'.index('o', -3)  
'foo'.index('o', -4)  
'foo'.index(/o./, -2) 
'foo'.index(/.o/, -2) 

Related: String#rindex.

static VALUE
rb_str_insert(VALUE str, VALUE idx, VALUE str2)
{
    long pos = NUM2LONG(idx);

    if (pos == -1) {
        return rb_str_append(str, str2);
    }
    else if (pos < 0) {
        pos++;
    }
    rb_str_update(str, pos, 0, str2);
    return str;
}

Inserts the given other_string into self; returns self.

If the Integer index is positive, inserts other_string at offset index:

'foo'.insert(1, 'bar') 

If the Integer index is negative, counts backward from the end of self and inserts other_string at offset index+1 (that is, after self[index]):

'foo'.insert(-2, 'bar') 

VALUE
rb_str_inspect(VALUE str)
{
    int encidx = ENCODING_GET(str);
    rb_encoding *enc = rb_enc_from_index(encidx);
    const char *p, *pend, *prev;
    char buf[CHAR_ESC_LEN + 1];
    VALUE result = rb_str_buf_new(0);
    rb_encoding *resenc = rb_default_internal_encoding();
    int unicode_p = rb_enc_unicode_p(enc);
    int asciicompat = rb_enc_asciicompat(enc);

    if (resenc == NULL) resenc = rb_default_external_encoding();
    if (!rb_enc_asciicompat(resenc)) resenc = rb_usascii_encoding();
    rb_enc_associate(result, resenc);
    str_buf_cat2(result, "\"");

    p = RSTRING_PTR(str); pend = RSTRING_END(str);
    prev = p;
    while (p < pend) {
        unsigned int c, cc;
        int n;

        n = rb_enc_precise_mbclen(p, pend, enc);
        if (!MBCLEN_CHARFOUND_P(n)) {
            if (p > prev) str_buf_cat(result, prev, p - prev);
            n = rb_enc_mbminlen(enc);
            if (pend < p + n)
                n = (int)(pend - p);
            while (n--) {
                snprintf(buf, CHAR_ESC_LEN, "\\x%02X", *p & 0377);
                str_buf_cat(result, buf, strlen(buf));
                prev = ++p;
            }
            continue;
        }
        n = MBCLEN_CHARFOUND_LEN(n);
        c = rb_enc_mbc_to_codepoint(p, pend, enc);
        p += n;
        if ((asciicompat || unicode_p) &&
          (c == '"'|| c == '\\' ||
            (c == '#' &&
             p < pend &&
             MBCLEN_CHARFOUND_P(rb_enc_precise_mbclen(p,pend,enc)) &&
             (cc = rb_enc_codepoint(p,pend,enc),
              (cc == '$' || cc == '@' || cc == '{'))))) {
            if (p - n > prev) str_buf_cat(result, prev, p - n - prev);
            str_buf_cat2(result, "\\");
            if (asciicompat || enc == resenc) {
                prev = p - n;
                continue;
            }
        }
        switch (c) {
          case '\n': cc = 'n'; break;
          case '\r': cc = 'r'; break;
          case '\t': cc = 't'; break;
          case '\f': cc = 'f'; break;
          case '\013': cc = 'v'; break;
          case '\010': cc = 'b'; break;
          case '\007': cc = 'a'; break;
          case 033: cc = 'e'; break;
          default: cc = 0; break;
        }
        if (cc) {
            if (p - n > prev) str_buf_cat(result, prev, p - n - prev);
            buf[0] = '\\';
            buf[1] = (char)cc;
            str_buf_cat(result, buf, 2);
            prev = p;
            continue;
        }
        /* The special casing of 0x85 (NEXT_LINE) here is because
         * Oniguruma historically treats it as printable, but it
         * doesn't match the print POSIX bracket class or character
         * property in regexps.
         *
         * See Ruby Bug #16842 for details:
         * https://bugs.ruby-lang.org/issues/16842
         */
        if ((enc == resenc && rb_enc_isprint(c, enc) && c != 0x85) ||
            (asciicompat && rb_enc_isascii(c, enc) && ISPRINT(c))) {
            continue;
        }
        else {
            if (p - n > prev) str_buf_cat(result, prev, p - n - prev);
            rb_str_buf_cat_escaped_char(result, c, unicode_p);
            prev = p;
            continue;
        }
    }
    if (p > prev) str_buf_cat(result, prev, p - prev);
    str_buf_cat2(result, "\"");

    return result;
}

Returns a printable version of self, enclosed in double-quotes, and with special characters escaped:

s = "foo\tbar\tbaz\n"
s.inspect

VALUE
rb_str_intern(VALUE str)
{
    VALUE sym;

    GLOBAL_SYMBOLS_ENTER(symbols);
    {
        sym = lookup_str_sym_with_lock(symbols, str);

        if (sym) {
            // ok
        }
        else if (USE_SYMBOL_GC) {
            rb_encoding *enc = rb_enc_get(str);
            rb_encoding *ascii = rb_usascii_encoding();
            if (enc != ascii && sym_check_asciionly(str, false)) {
                str = rb_str_dup(str);
                rb_enc_associate(str, ascii);
                OBJ_FREEZE(str);
                enc = ascii;
            }
            else {
                str = rb_str_dup(str);
                OBJ_FREEZE(str);
            }
            str = rb_fstring(str);
            int type = rb_str_symname_type(str, IDSET_ATTRSET_FOR_INTERN);
            if (type < 0) type = ID_JUNK;
            sym = dsymbol_alloc(symbols, rb_cSymbol, str, enc, type);
        }
        else {
            ID id = intern_str(str, 0);
            sym = ID2SYM(id);
        }
    }
    GLOBAL_SYMBOLS_LEAVE();
    return sym;
}

Returns the Symbol corresponding to str, creating the symbol if it did not previously exist. See Symbol#id2name.

"Koala".intern         
s = 'cat'.to_sym       
s == :cat              
s = '@cat'.to_sym      
s == :@cat             

This can also be used to create symbols that cannot be represented using the :xxx notation.

'cat and dog'.to_sym   

VALUE
rb_str_length(VALUE str)
{
    return LONG2NUM(str_strlen(str, NULL));
}

Returns the count of characters (not bytes) in self:

'foo'.length        
'тест'.length       
'こんにちは'.length   

Contrast with String#bytesize:

'foo'.bytesize        
'тест'.bytesize       
'こんにちは'.bytesize   

static VALUE
rb_str_lines(int argc, VALUE *argv, VALUE str)
{
    VALUE ary = WANTARRAY("lines", 0);
    return rb_str_enumerate_lines(argc, argv, str, ary);
}

Forms substrings (“lines”) of self according to the given arguments (see String#each_line for details); returns the lines in an array.

static VALUE
rb_str_ljust(int argc, VALUE *argv, VALUE str)
{
    return rb_str_justify(argc, argv, str, 'l');
}

Returns a left-justified copy of self.

If integer argument size is greater than the size (in characters) of self, returns a new string of length size that is a copy of self, left justified and padded on the right with pad_string:

'hello'.ljust(10)       
'  hello'.ljust(10)     
'hello'.ljust(10, 'ab') 
'тест'.ljust(10)        
'こんにちは'.ljust(10)    

If size is not greater than the size of self, returns a copy of self:

'hello'.ljust(5)  
'hello'.ljust(1)  

Related: String#rjust, String#center.

static VALUE
rb_str_lstrip(VALUE str)
{
    char *start;
    long len, loffset;
    RSTRING_GETMEM(str, start, len);
    loffset = lstrip_offset(str, start, start+len, STR_ENC_GET(str));
    if (loffset <= 0) return str_duplicate(rb_cString, str);
    return rb_str_subseq(str, loffset, len - loffset);
}

Returns a copy of self with leading whitespace removed; see Whitespace in Strings:

whitespace = "\x00\t\n\v\f\r "
s = whitespace + 'abc' + whitespace
s        
s.lstrip 

Related: String#rstrip, String#strip.

static VALUE
rb_str_lstrip_bang(VALUE str)
{
    rb_encoding *enc;
    char *start, *s;
    long olen, loffset;

    str_modify_keep_cr(str);
    enc = STR_ENC_GET(str);
    RSTRING_GETMEM(str, start, olen);
    loffset = lstrip_offset(str, start, start+olen, enc);
    if (loffset > 0) {
        long len = olen-loffset;
        s = start + loffset;
        memmove(start, s, len);
        STR_SET_LEN(str, len);
        TERM_FILL(start+len, rb_enc_mbminlen(enc));
        return str;
    }
    return Qnil;
}

Like String#lstrip, except that any modifications are made in self; returns self if any modification are made, nil otherwise.

Related: String#rstrip!, String#strip!.

static VALUE
rb_str_match_m(int argc, VALUE *argv, VALUE str)
{
    VALUE re, result;
    if (argc < 1)
        rb_check_arity(argc, 1, 2);
    re = argv[0];
    argv[0] = str;
    result = rb_funcallv(get_pat(re), rb_intern("match"), argc, argv);
    if (!NIL_P(result) && rb_block_given_p()) {
        return rb_yield(result);
    }
    return result;
}

Returns a MatchData object (or nil) based on self and the given pattern.

Note: also updates Global Variables at Regexp.

• Computes regexp by converting pattern (if not already a Regexp).

  regexp = Regexp.new(pattern)
  

• Computes matchdata, which will be either a MatchData object or nil (see Regexp#match):

  matchdata = <tt>regexp.match(self)

With no block given, returns the computed matchdata:

'foo'.match('f') 
'foo'.match('o') 
'foo'.match('x') 

If Integer argument offset is given, the search begins at index offset:

'foo'.match('f', 1) 
'foo'.match('o', 1) 

With a block given, calls the block with the computed matchdata and returns the block’s return value:

'foo'.match(/o/) {|matchdata| matchdata } 
'foo'.match(/x/) {|matchdata| matchdata } 
'foo'.match(/f/, 1) {|matchdata| matchdata } 

static VALUE
rb_str_match_m_p(int argc, VALUE *argv, VALUE str)
{
    VALUE re;
    rb_check_arity(argc, 1, 2);
    re = get_pat(argv[0]);
    return rb_reg_match_p(re, str, argc > 1 ? NUM2LONG(argv[1]) : 0);
}

Returns true or false based on whether a match is found for self and pattern.

Note: does not update Global Variables at Regexp.

Computes regexp by converting pattern (if not already a Regexp).

regexp = Regexp.new(pattern)

Returns true if self+.match(regexp) returns a MatchData object, false otherwise:

'foo'.match?(/o/) 
'foo'.match?('o') 
'foo'.match?(/x/) 

If Integer argument offset is given, the search begins at index offset:

'foo'.match?('f', 1) 
'foo'.match?('o', 1) 

static VALUE
rb_str_oct(VALUE str)
{
    return rb_str_to_inum(str, -8, FALSE);
}

Interprets the leading substring of self as a string of octal digits (with an optional sign) and returns the corresponding number; returns zero if there is no such leading substring:

'123'.oct             
'-377'.oct            
'0377non-numeric'.oct 
'non-numeric'.oct     

If self starts with 0, radix indicators are honored; see Kernel#Integer.

Related: String#hex.

static VALUE
rb_str_ord(VALUE s)
{
    unsigned int c;

    c = rb_enc_codepoint(RSTRING_PTR(s), RSTRING_END(s), STR_ENC_GET(s));
    return UINT2NUM(c);
}

Returns the integer ordinal of the first character of self:

'h'.ord         
'hello'.ord     
'тест'.ord      
'こんにちは'.ord  

static VALUE
rb_str_partition(VALUE str, VALUE sep)
{
    long pos;

    sep = get_pat_quoted(sep, 0);
    if (RB_TYPE_P(sep, T_REGEXP)) {
        if (rb_reg_search(sep, str, 0, 0) < 0) {
            goto failed;
        }
        VALUE match = rb_backref_get();
        struct re_registers *regs = RMATCH_REGS(match);

        pos = BEG(0);
        sep = rb_str_subseq(str, pos, END(0) - pos);
    }
    else {
        pos = rb_str_index(str, sep, 0);
        if (pos < 0) goto failed;
    }
    return rb_ary_new3(3, rb_str_subseq(str, 0, pos),
                          sep,
                          rb_str_subseq(str, pos+RSTRING_LEN(sep),
                                             RSTRING_LEN(str)-pos-RSTRING_LEN(sep)));

  failed:
    return rb_ary_new3(3, str_duplicate(rb_cString, str), str_new_empty_String(str), str_new_empty_String(str));
}

Returns a 3-element array of substrings of self.

Matches a pattern against self, scanning from the beginning. The pattern is:

• string_or_regexp itself, if it is a Regexp.

• Regexp.quote(string_or_regexp), if string_or_regexp is a string.

If the pattern is matched, returns pre-match, first-match, post-match:

'hello'.partition('l')      
'hello'.partition('ll')     
'hello'.partition('h')      
'hello'.partition('o')      
'hello'.partition(/l+/)     
'hello'.partition('')       
'тест'.partition('т')       
'こんにちは'.partition('に')  

If the pattern is not matched, returns a copy of self and two empty strings:

'hello'.partition('x') 

Related: String#rpartition, String#split.

static VALUE
rb_str_prepend_multi(int argc, VALUE *argv, VALUE str)
{
    str_modifiable(str);

    if (argc == 1) {
        rb_str_update(str, 0L, 0L, argv[0]);
    }
    else if (argc > 1) {
        int i;
        VALUE arg_str = rb_str_tmp_new(0);
        rb_enc_copy(arg_str, str);
        for (i = 0; i < argc; i++) {
            rb_str_append(arg_str, argv[i]);
        }
        rb_str_update(str, 0L, 0L, arg_str);
    }

    return str;
}

Prepends each string in other_strings to self and returns self:

s = 'foo'
s.prepend('bar', 'baz') 
s                       

Related: String#concat.

Replaces the contents of self with the contents of other_string:

s = 'foo'        
s.replace('bar') 

static VALUE
rb_str_reverse(VALUE str)
{
    rb_encoding *enc;
    VALUE rev;
    char *s, *e, *p;
    int cr;

    if (RSTRING_LEN(str) <= 1) return str_duplicate(rb_cString, str);
    enc = STR_ENC_GET(str);
    rev = rb_str_new(0, RSTRING_LEN(str));
    s = RSTRING_PTR(str); e = RSTRING_END(str);
    p = RSTRING_END(rev);
    cr = ENC_CODERANGE(str);

    if (RSTRING_LEN(str) > 1) {
        if (single_byte_optimizable(str)) {
            while (s < e) {
                *--p = *s++;
            }
        }
        else if (cr == ENC_CODERANGE_VALID) {
            while (s < e) {
                int clen = rb_enc_fast_mbclen(s, e, enc);

                p -= clen;
                memcpy(p, s, clen);
                s += clen;
            }
        }
        else {
            cr = rb_enc_asciicompat(enc) ?
                ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
            while (s < e) {
                int clen = rb_enc_mbclen(s, e, enc);

                if (clen > 1 || (*s & 0x80)) cr = ENC_CODERANGE_UNKNOWN;
                p -= clen;
                memcpy(p, s, clen);
                s += clen;
            }
        }
    }
    STR_SET_LEN(rev, RSTRING_LEN(str));
    str_enc_copy_direct(rev, str);
    ENC_CODERANGE_SET(rev, cr);

    return rev;
}

Returns a new string with the characters from self in reverse order.

'stressed'.reverse 

static VALUE
rb_str_reverse_bang(VALUE str)
{
    if (RSTRING_LEN(str) > 1) {
        if (single_byte_optimizable(str)) {
            char *s, *e, c;

            str_modify_keep_cr(str);
            s = RSTRING_PTR(str);
            e = RSTRING_END(str) - 1;
            while (s < e) {
                c = *s;
                *s++ = *e;
                *e-- = c;
            }
        }
        else {
            str_shared_replace(str, rb_str_reverse(str));
        }
    }
    else {
        str_modify_keep_cr(str);
    }
    return str;
}

Returns self with its characters reversed:

s = 'stressed'
s.reverse! 
s          

static VALUE
rb_str_rindex_m(int argc, VALUE *argv, VALUE str)
{
    VALUE sub;
    VALUE initpos;
    rb_encoding *enc = STR_ENC_GET(str);
    long pos, len = str_strlen(str, enc); /* str's enc */

    if (rb_scan_args(argc, argv, "11", &sub, &initpos) == 2) {
        pos = NUM2LONG(initpos);
        if (pos < 0 && (pos += len) < 0) {
            if (RB_TYPE_P(sub, T_REGEXP)) {
                rb_backref_set(Qnil);
            }
            return Qnil;
        }
        if (pos > len) pos = len;
    }
    else {
        pos = len;
    }

    if (RB_TYPE_P(sub, T_REGEXP)) {
        /* enc = rb_enc_check(str, sub); */
        pos = str_offset(RSTRING_PTR(str), RSTRING_END(str), pos,
                         enc, single_byte_optimizable(str));

        if (rb_reg_search(sub, str, pos, 1) >= 0) {
            VALUE match = rb_backref_get();
            struct re_registers *regs = RMATCH_REGS(match);
            pos = rb_str_sublen(str, BEG(0));
            return LONG2NUM(pos);
        }
    }
    else {
        StringValue(sub);
        pos = rb_str_rindex(str, sub, pos);
        if (pos >= 0) {
            pos = rb_str_sublen(str, pos);
            return LONG2NUM(pos);
        }
    }
    return Qnil;
}

Returns the Integer index of the last occurrence of the given substring, or nil if none found:

'foo'.rindex('f') 
'foo'.rindex('o') 
'foo'.rindex('oo') 
'foo'.rindex('ooo') 

Returns the Integer index of the last match for the given Regexp regexp, or nil if none found:

'foo'.rindex(/f/) 
'foo'.rindex(/o/) 
'foo'.rindex(/oo/) 
'foo'.rindex(/ooo/) 

The last match means starting at the possible last position, not the last of longest matches.

'foo'.rindex(/o+/) 
$~ 

To get the last longest match, needs to combine with negative lookbehind.

'foo'.rindex(/(?<!o)o+/) 
$~ 

Or String#index with negative lookforward.

'foo'.index(/o+(?!.*o)/) 
$~ 

Integer argument offset, if given and non-negative, specifies the maximum starting position in the string to end the search:

'foo'.rindex('o', 0) 
'foo'.rindex('o', 1) 
'foo'.rindex('o', 2) 
'foo'.rindex('o', 3) 

If offset is a negative Integer, the maximum starting position in the string to end the search is the sum of the string’s length and offset:

'foo'.rindex('o', -1) 
'foo'.rindex('o', -2) 
'foo'.rindex('o', -3) 
'foo'.rindex('o', -4) 

Related: String#index.

static VALUE
rb_str_rjust(int argc, VALUE *argv, VALUE str)
{
    return rb_str_justify(argc, argv, str, 'r');
}

Returns a right-justified copy of self.

If integer argument size is greater than the size (in characters) of self, returns a new string of length size that is a copy of self, right justified and padded on the left with pad_string:

'hello'.rjust(10)       
'hello  '.rjust(10)     
'hello'.rjust(10, 'ab') 
'тест'.rjust(10)        
'こんにちは'.rjust(10)    

If size is not greater than the size of self, returns a copy of self:

'hello'.rjust(5, 'ab')  
'hello'.rjust(1, 'ab')  

Related: String#ljust, String#center.

static VALUE
rb_str_rpartition(VALUE str, VALUE sep)
{
    long pos = RSTRING_LEN(str);

    sep = get_pat_quoted(sep, 0);
    if (RB_TYPE_P(sep, T_REGEXP)) {
        if (rb_reg_search(sep, str, pos, 1) < 0) {
            goto failed;
        }
        VALUE match = rb_backref_get();
        struct re_registers *regs = RMATCH_REGS(match);

        pos = BEG(0);
        sep = rb_str_subseq(str, pos, END(0) - pos);
    }
    else {
        pos = rb_str_sublen(str, pos);
        pos = rb_str_rindex(str, sep, pos);
        if (pos < 0) {
            goto failed;
        }
    }

    return rb_ary_new3(3, rb_str_subseq(str, 0, pos),
                          sep,
                          rb_str_subseq(str, pos+RSTRING_LEN(sep),
                                        RSTRING_LEN(str)-pos-RSTRING_LEN(sep)));
  failed:
    return rb_ary_new3(3, str_new_empty_String(str), str_new_empty_String(str), str_duplicate(rb_cString, str));
}

Returns a 3-element array of substrings of self.

Matches a pattern against self, scanning backwards from the end. The pattern is:

• string_or_regexp itself, if it is a Regexp.

• Regexp.quote(string_or_regexp), if string_or_regexp is a string.

If the pattern is matched, returns pre-match, last-match, post-match:

'hello'.rpartition('l')      
'hello'.rpartition('ll')     
'hello'.rpartition('h')      
'hello'.rpartition('o')      
'hello'.rpartition(/l+/)     
'hello'.rpartition('')       
'тест'.rpartition('т')       
'こんにちは'.rpartition('に')  

If the pattern is not matched, returns two empty strings and a copy of self:

'hello'.rpartition('x') 

Related: String#partition, String#split.

static VALUE
rb_str_rstrip(VALUE str)
{
    rb_encoding *enc;
    char *start;
    long olen, roffset;

    enc = STR_ENC_GET(str);
    RSTRING_GETMEM(str, start, olen);
    roffset = rstrip_offset(str, start, start+olen, enc);

    if (roffset <= 0) return str_duplicate(rb_cString, str);
    return rb_str_subseq(str, 0, olen-roffset);
}

Returns a copy of the receiver with trailing whitespace removed; see Whitespace in Strings:

whitespace = "\x00\t\n\v\f\r "
s = whitespace + 'abc' + whitespace
s        
s.rstrip 

Related: String#lstrip, String#strip.

static VALUE
rb_str_rstrip_bang(VALUE str)
{
    rb_encoding *enc;
    char *start;
    long olen, roffset;

    str_modify_keep_cr(str);
    enc = STR_ENC_GET(str);
    RSTRING_GETMEM(str, start, olen);
    roffset = rstrip_offset(str, start, start+olen, enc);
    if (roffset > 0) {
        long len = olen - roffset;

        STR_SET_LEN(str, len);
        TERM_FILL(start+len, rb_enc_mbminlen(enc));
        return str;
    }
    return Qnil;
}

Like String#rstrip, except that any modifications are made in self; returns self if any modification are made, nil otherwise.

Related: String#lstrip!, String#strip!.

static VALUE
rb_str_scan(VALUE str, VALUE pat)
{
    VALUE result;
    long start = 0;
    long last = -1, prev = 0;
    char *p = RSTRING_PTR(str); long len = RSTRING_LEN(str);

    pat = get_pat_quoted(pat, 1);
    mustnot_broken(str);
    if (!rb_block_given_p()) {
        VALUE ary = rb_ary_new();

        while (!NIL_P(result = scan_once(str, pat, &start, 0))) {
            last = prev;
            prev = start;
            rb_ary_push(ary, result);
        }
        if (last >= 0) rb_pat_search(pat, str, last, 1);
        else rb_backref_set(Qnil);
        return ary;
    }

    while (!NIL_P(result = scan_once(str, pat, &start, 1))) {
        last = prev;
        prev = start;
        rb_yield(result);
        str_mod_check(str, p, len);
    }
    if (last >= 0) rb_pat_search(pat, str, last, 1);
    return str;
}

Matches a pattern against self; the pattern is:

• string_or_regexp itself, if it is a Regexp.

• Regexp.quote(string_or_regexp), if string_or_regexp is a string.

Iterates through self, generating a collection of matching results:

• If the pattern contains no groups, each result is the matched string, $&.

• If the pattern contains groups, each result is an array containing one entry per group.

With no block given, returns an array of the results:

s = 'cruel world'
s.scan(/\w+/)      
s.scan(/.../)      
s.scan(/(...)/)    
s.scan(/(..)(..)/) 

With a block given, calls the block with each result; returns self:

s.scan(/\w+/) {|w| print "<<#{w}>> " }
print "\n"
s.scan(/(.)(.)/) {|x,y| print y, x }
print "\n"

Output:

<<cruel>> <<world>>
rceu lowlr

static VALUE
str_scrub(int argc, VALUE *argv, VALUE str)
{
    VALUE repl = argc ? (rb_check_arity(argc, 0, 1), argv[0]) : Qnil;
    VALUE new = rb_str_scrub(str, repl);
    return NIL_P(new) ? str_duplicate(rb_cString, str): new;
}

Returns a copy of self with each invalid byte sequence replaced by the given replacement_string.

With no block given and no argument, replaces each invalid sequence with the default replacement string ("�" for a Unicode encoding, '?' otherwise):

s = "foo\x81\x81bar"
s.scrub 

With no block given and argument replacement_string given, replaces each invalid sequence with that string:

"foo\x81\x81bar".scrub('xyzzy') 

With a block given, replaces each invalid sequence with the value of the block:

"foo\x81\x81bar".scrub {|bytes| p bytes; 'XYZZY' }

Output:

"\x81"
"\x81"

static VALUE
str_scrub_bang(int argc, VALUE *argv, VALUE str)
{
    VALUE repl = argc ? (rb_check_arity(argc, 0, 1), argv[0]) : Qnil;
    VALUE new = rb_str_scrub(str, repl);
    if (!NIL_P(new)) rb_str_replace(str, new);
    return str;
}

Like String#scrub, except that any replacements are made in self.

VALUE
rb_str_setbyte(VALUE str, VALUE index, VALUE value)
{
    long pos = NUM2LONG(index);
    long len = RSTRING_LEN(str);
    char *ptr, *head, *left = 0;
    rb_encoding *enc;
    int cr = ENC_CODERANGE_UNKNOWN, width, nlen;

    if (pos < -len || len <= pos)
        rb_raise(rb_eIndexError, "index %ld out of string", pos);
    if (pos < 0)
        pos += len;

    VALUE v = rb_to_int(value);
    VALUE w = rb_int_and(v, INT2FIX(0xff));
    char byte = (char)(NUM2INT(w) & 0xFF);

    if (!str_independent(str))
        str_make_independent(str);
    enc = STR_ENC_GET(str);
    head = RSTRING_PTR(str);
    ptr = &head[pos];
    if (!STR_EMBED_P(str)) {
        cr = ENC_CODERANGE(str);
        switch (cr) {
          case ENC_CODERANGE_7BIT:
            left = ptr;
            *ptr = byte;
            if (ISASCII(byte)) goto end;
            nlen = rb_enc_precise_mbclen(left, head+len, enc);
            if (!MBCLEN_CHARFOUND_P(nlen))
                ENC_CODERANGE_SET(str, ENC_CODERANGE_BROKEN);
            else
                ENC_CODERANGE_SET(str, ENC_CODERANGE_VALID);
            goto end;
          case ENC_CODERANGE_VALID:
            left = rb_enc_left_char_head(head, ptr, head+len, enc);
            width = rb_enc_precise_mbclen(left, head+len, enc);
            *ptr = byte;
            nlen = rb_enc_precise_mbclen(left, head+len, enc);
            if (!MBCLEN_CHARFOUND_P(nlen))
                ENC_CODERANGE_SET(str, ENC_CODERANGE_BROKEN);
            else if (MBCLEN_CHARFOUND_LEN(nlen) != width || ISASCII(byte))
                ENC_CODERANGE_CLEAR(str);
            goto end;
        }
    }
    ENC_CODERANGE_CLEAR(str);
    *ptr = byte;

  end:
    return value;
}

Sets the byte at zero-based index to integer; returns integer:

s = 'abcde'      
s.setbyte(0, 98) 
s                

Related: String#getbyte.

def shellescape
  Shellwords.escape(self)
end

Escapes str so that it can be safely used in a Bourne shell command line.

See Shellwords.shellescape for details.

def shellsplit
  Shellwords.split(self)
end

Splits str into an array of tokens in the same way the UNIX Bourne shell does.

See Shellwords.shellsplit for details.

static VALUE
rb_str_slice_bang(int argc, VALUE *argv, VALUE str)
{
    VALUE result = Qnil;
    VALUE indx;
    long beg, len = 1;
    char *p;

    rb_check_arity(argc, 1, 2);
    str_modify_keep_cr(str);
    indx = argv[0];
    if (RB_TYPE_P(indx, T_REGEXP)) {
        if (rb_reg_search(indx, str, 0, 0) < 0) return Qnil;
        VALUE match = rb_backref_get();
        struct re_registers *regs = RMATCH_REGS(match);
        int nth = 0;
        if (argc > 1 && (nth = rb_reg_backref_number(match, argv[1])) < 0) {
            if ((nth += regs->num_regs) <= 0) return Qnil;
        }
        else if (nth >= regs->num_regs) return Qnil;
        beg = BEG(nth);
        len = END(nth) - beg;
        goto subseq;
    }
    else if (argc == 2) {
        beg = NUM2LONG(indx);
        len = NUM2LONG(argv[1]);
        goto num_index;
    }
    else if (FIXNUM_P(indx)) {
        beg = FIX2LONG(indx);
        if (!(p = rb_str_subpos(str, beg, &len))) return Qnil;
        if (!len) return Qnil;
        beg = p - RSTRING_PTR(str);
        goto subseq;
    }
    else if (RB_TYPE_P(indx, T_STRING)) {
        beg = rb_str_index(str, indx, 0);
        if (beg == -1) return Qnil;
        len = RSTRING_LEN(indx);
        result = str_duplicate(rb_cString, indx);
        goto squash;
    }
    else {
        switch (rb_range_beg_len(indx, &beg, &len, str_strlen(str, NULL), 0)) {
          case Qnil:
            return Qnil;
          case Qfalse:
            beg = NUM2LONG(indx);
            if (!(p = rb_str_subpos(str, beg, &len))) return Qnil;
            if (!len) return Qnil;
            beg = p - RSTRING_PTR(str);
            goto subseq;
          default:
            goto num_index;
        }
    }

  num_index:
    if (!(p = rb_str_subpos(str, beg, &len))) return Qnil;
    beg = p - RSTRING_PTR(str);

  subseq:
    result = rb_str_new(RSTRING_PTR(str)+beg, len);
    rb_enc_cr_str_copy_for_substr(result, str);

  squash:
    if (len > 0) {
        if (beg == 0) {
            rb_str_drop_bytes(str, len);
        }
        else {
            char *sptr = RSTRING_PTR(str);
            long slen = RSTRING_LEN(str);
            if (beg + len > slen) /* pathological check */
                len = slen - beg;
            memmove(sptr + beg,
                    sptr + beg + len,
                    slen - (beg + len));
            slen -= len;
            STR_SET_LEN(str, slen);
            TERM_FILL(&sptr[slen], TERM_LEN(str));
        }
    }
    return result;
}

Removes and returns the substring of self specified by the arguments. See String Slices.

A few examples:

string = "This is a string"
string.slice!(2)        
string.slice!(3..6)     
string.slice!(/s.*t/)   
string.slice!("r")      
string                  

static VALUE
rb_str_split_m(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    VALUE spat;
    VALUE limit;
    split_type_t split_type;
    long beg, end, i = 0, empty_count = -1;
    int lim = 0;
    VALUE result, tmp;

    result = rb_block_given_p() ? Qfalse : Qnil;
    if (rb_scan_args(argc, argv, "02", &spat, &limit) == 2) {
        lim = NUM2INT(limit);
        if (lim <= 0) limit = Qnil;
        else if (lim == 1) {
            if (RSTRING_LEN(str) == 0)
                return result ? rb_ary_new2(0) : str;
            tmp = str_duplicate(rb_cString, str);
            if (!result) {
                rb_yield(tmp);
                return str;
            }
            return rb_ary_new3(1, tmp);
        }
        i = 1;
    }
    if (NIL_P(limit) && !lim) empty_count = 0;

    enc = STR_ENC_GET(str);
    split_type = SPLIT_TYPE_REGEXP;
    if (!NIL_P(spat)) {
        spat = get_pat_quoted(spat, 0);
    }
    else if (NIL_P(spat = rb_fs)) {
        split_type = SPLIT_TYPE_AWK;
    }
    else if (!(spat = rb_fs_check(spat))) {
        rb_raise(rb_eTypeError, "value of $; must be String or Regexp");
    }
    else {
        rb_category_warn(RB_WARN_CATEGORY_DEPRECATED, "$; is set to non-nil value");
    }
    if (split_type != SPLIT_TYPE_AWK) {
        switch (BUILTIN_TYPE(spat)) {
          case T_REGEXP:
            rb_reg_options(spat); /* check if uninitialized */
            tmp = RREGEXP_SRC(spat);
            split_type = literal_split_pattern(tmp, SPLIT_TYPE_REGEXP);
            if (split_type == SPLIT_TYPE_AWK) {
                spat = tmp;
                split_type = SPLIT_TYPE_STRING;
            }
            break;

          case T_STRING:
            mustnot_broken(spat);
            split_type = literal_split_pattern(spat, SPLIT_TYPE_STRING);
            break;

          default:
            UNREACHABLE_RETURN(Qnil);
        }
    }

#define SPLIT_STR(beg, len) (empty_count = split_string(result, str, beg, len, empty_count))

    beg = 0;
    char *ptr = RSTRING_PTR(str);
    char *eptr = RSTRING_END(str);
    if (split_type == SPLIT_TYPE_AWK) {
        char *bptr = ptr;
        int skip = 1;
        unsigned int c;

        if (result) result = rb_ary_new();
        end = beg;
        if (is_ascii_string(str)) {
            while (ptr < eptr) {
                c = (unsigned char)*ptr++;
                if (skip) {
                    if (ascii_isspace(c)) {
                        beg = ptr - bptr;
                    }
                    else {
                        end = ptr - bptr;
                        skip = 0;
                        if (!NIL_P(limit) && lim <= i) break;
                    }
                }
                else if (ascii_isspace(c)) {
                    SPLIT_STR(beg, end-beg);
                    skip = 1;
                    beg = ptr - bptr;
                    if (!NIL_P(limit)) ++i;
                }
                else {
                    end = ptr - bptr;
                }
            }
        }
        else {
            while (ptr < eptr) {
                int n;

                c = rb_enc_codepoint_len(ptr, eptr, &n, enc);
                ptr += n;
                if (skip) {
                    if (rb_isspace(c)) {
                        beg = ptr - bptr;
                    }
                    else {
                        end = ptr - bptr;
                        skip = 0;
                        if (!NIL_P(limit) && lim <= i) break;
                    }
                }
                else if (rb_isspace(c)) {
                    SPLIT_STR(beg, end-beg);
                    skip = 1;
                    beg = ptr - bptr;
                    if (!NIL_P(limit)) ++i;
                }
                else {
                    end = ptr - bptr;
                }
            }
        }
    }
    else if (split_type == SPLIT_TYPE_STRING) {
        char *str_start = ptr;
        char *substr_start = ptr;
        char *sptr = RSTRING_PTR(spat);
        long slen = RSTRING_LEN(spat);

        if (result) result = rb_ary_new();
        mustnot_broken(str);
        enc = rb_enc_check(str, spat);
        while (ptr < eptr &&
               (end = rb_memsearch(sptr, slen, ptr, eptr - ptr, enc)) >= 0) {
            /* Check we are at the start of a char */
            char *t = rb_enc_right_char_head(ptr, ptr + end, eptr, enc);
            if (t != ptr + end) {
                ptr = t;
                continue;
            }
            SPLIT_STR(substr_start - str_start, (ptr+end) - substr_start);
            ptr += end + slen;
            substr_start = ptr;
            if (!NIL_P(limit) && lim <= ++i) break;
        }
        beg = ptr - str_start;
    }
    else if (split_type == SPLIT_TYPE_CHARS) {
        char *str_start = ptr;
        int n;

        if (result) result = rb_ary_new_capa(RSTRING_LEN(str));
        mustnot_broken(str);
        enc = rb_enc_get(str);
        while (ptr < eptr &&
               (n = rb_enc_precise_mbclen(ptr, eptr, enc)) > 0) {
            SPLIT_STR(ptr - str_start, n);
            ptr += n;
            if (!NIL_P(limit) && lim <= ++i) break;
        }
        beg = ptr - str_start;
    }
    else {
        if (result) result = rb_ary_new();
        long len = RSTRING_LEN(str);
        long start = beg;
        long idx;
        int last_null = 0;
        struct re_registers *regs;
        VALUE match = 0;

        for (; rb_reg_search(spat, str, start, 0) >= 0;
             (match ? (rb_match_unbusy(match), rb_backref_set(match)) : (void)0)) {
            match = rb_backref_get();
            if (!result) rb_match_busy(match);
            regs = RMATCH_REGS(match);
            end = BEG(0);
            if (start == end && BEG(0) == END(0)) {
                if (!ptr) {
                    SPLIT_STR(0, 0);
                    break;
                }
                else if (last_null == 1) {
                    SPLIT_STR(beg, rb_enc_fast_mbclen(ptr+beg, eptr, enc));
                    beg = start;
                }
                else {
                    if (start == len)
                        start++;
                    else
                        start += rb_enc_fast_mbclen(ptr+start,eptr,enc);
                    last_null = 1;
                    continue;
                }
            }
            else {
                SPLIT_STR(beg, end-beg);
                beg = start = END(0);
            }
            last_null = 0;

            for (idx=1; idx < regs->num_regs; idx++) {
                if (BEG(idx) == -1) continue;
                SPLIT_STR(BEG(idx), END(idx)-BEG(idx));
            }
            if (!NIL_P(limit) && lim <= ++i) break;
        }
        if (match) rb_match_unbusy(match);
    }
    if (RSTRING_LEN(str) > 0 && (!NIL_P(limit) || RSTRING_LEN(str) > beg || lim < 0)) {
        SPLIT_STR(beg, RSTRING_LEN(str)-beg);
    }

    return result ? result : str;
}

Returns an array of substrings of self that are the result of splitting self at each occurrence of the given field separator field_sep.

When field_sep is $;:

• If $; is nil (its default value), the split occurs just as if field_sep were given as a space character (see below).

• If $; is a string, the split occurs just as if field_sep were given as that string (see below).

When field_sep is ' ' and limit is 0 (its default value), the split occurs at each sequence of whitespace:

'abc def ghi'.split(' ')         => ["abc", "def", "ghi"]
"abc \n\tdef\t\n  ghi".split(' ') 
'abc  def   ghi'.split(' ')      => ["abc", "def", "ghi"]
''.split(' ')                    => []

When field_sep is a string different from ' ' and limit is 0, the split occurs at each occurrence of field_sep; trailing empty substrings are not returned:

'abracadabra'.split('ab')  => ["", "racad", "ra"]
'aaabcdaaa'.split('a')     => ["", "", "", "bcd"]
''.split('a')              => []
'3.14159'.split('1')       => ["3.", "4", "59"]
'!@#$%^$&*($)_+'.split('$') 
'тест'.split('т')          => ["", "ес"]
'こんにちは'.split('に')     => ["こん", "ちは"]

When field_sep is a Regexp and limit is 0, the split occurs at each occurrence of a match; trailing empty substrings are not returned:

'abracadabra'.split(/ab/) 
'aaabcdaaa'.split(/a/)   => ["", "", "", "bcd"]
'aaabcdaaa'.split(//)    => ["a", "a", "a", "b", "c", "d", "a", "a", "a"]
'1 + 1 == 2'.split(/\W+/) 

If the Regexp contains groups, their matches are also included in the returned array:

'1:2:3'.split(/(:)()()/, 2) 

As seen above, if limit is 0, trailing empty substrings are not returned:

'aaabcdaaa'.split('a')   => ["", "", "", "bcd"]

If limit is positive integer n, no more than n - 1- splits occur, so that at most n substrings are returned, and trailing empty substrings are included:

'aaabcdaaa'.split('a', 1) 
'aaabcdaaa'.split('a', 2) 
'aaabcdaaa'.split('a', 5) 
'aaabcdaaa'.split('a', 7) 
'aaabcdaaa'.split('a', 8) 

Note that if field_sep is a Regexp containing groups, their matches are in the returned array, but do not count toward the limit.

If limit is negative, it behaves the same as if limit was zero, meaning that there is no limit, and trailing empty substrings are included:

'aaabcdaaa'.split('a', -1) 

If a block is given, it is called with each substring:

'abc def ghi'.split(' ') {|substring| p substring }

Output:

"abc"
"def"
"ghi"

Related: String#partition, String#rpartition.

static VALUE
rb_str_squeeze(int argc, VALUE *argv, VALUE str)
{
    str = str_duplicate(rb_cString, str);
    rb_str_squeeze_bang(argc, argv, str);
    return str;
}

Returns a copy of self with characters specified by selectors “squeezed” (see Multiple Character Selectors):

“Squeezed” means that each multiple-character run of a selected character is squeezed down to a single character; with no arguments given, squeezes all characters:

"yellow moon".squeeze                  
"  now   is  the".squeeze(" ")         
"putters shoot balls".squeeze("m-z")   

static VALUE
rb_str_squeeze_bang(int argc, VALUE *argv, VALUE str)
{
    char squeez[TR_TABLE_SIZE];
    rb_encoding *enc = 0;
    VALUE del = 0, nodel = 0;
    unsigned char *s, *send, *t;
    int i, modify = 0;
    int ascompat, singlebyte = single_byte_optimizable(str);
    unsigned int save;

    if (argc == 0) {
        enc = STR_ENC_GET(str);
    }
    else {
        for (i=0; i<argc; i++) {
            VALUE s = argv[i];

            StringValue(s);
            enc = rb_enc_check(str, s);
            if (singlebyte && !single_byte_optimizable(s))
                singlebyte = 0;
            tr_setup_table(s, squeez, i==0, &del, &nodel, enc);
        }
    }

    str_modify_keep_cr(str);
    s = t = (unsigned char *)RSTRING_PTR(str);
    if (!s || RSTRING_LEN(str) == 0) return Qnil;
    send = (unsigned char *)RSTRING_END(str);
    save = -1;
    ascompat = rb_enc_asciicompat(enc);

    if (singlebyte) {
        while (s < send) {
            unsigned int c = *s++;
            if (c != save || (argc > 0 && !squeez[c])) {
                *t++ = save = c;
            }
        }
    }
    else {
        while (s < send) {
            unsigned int c;
            int clen;

            if (ascompat && (c = *s) < 0x80) {
                if (c != save || (argc > 0 && !squeez[c])) {
                    *t++ = save = c;
                }
                s++;
            }
            else {
                c = rb_enc_codepoint_len((char *)s, (char *)send, &clen, enc);

                if (c != save || (argc > 0 && !tr_find(c, squeez, del, nodel))) {
                    if (t != s) rb_enc_mbcput(c, t, enc);
                    save = c;
                    t += clen;
                }
                s += clen;
            }
        }
    }

    TERM_FILL((char *)t, TERM_LEN(str));
    if ((char *)t - RSTRING_PTR(str) != RSTRING_LEN(str)) {
        STR_SET_LEN(str, (char *)t - RSTRING_PTR(str));
        modify = 1;
    }

    if (modify) return str;
    return Qnil;
}

Like String#squeeze, but modifies self in place. Returns self if any changes were made, nil otherwise.

static VALUE
rb_str_start_with(int argc, VALUE *argv, VALUE str)
{
    int i;

    for (i=0; i<argc; i++) {
        VALUE tmp = argv[i];
        if (RB_TYPE_P(tmp, T_REGEXP)) {
            if (rb_reg_start_with_p(tmp, str))
                return Qtrue;
        }
        else {
            const char *p, *s, *e;
            long slen, tlen;
            rb_encoding *enc;

            StringValue(tmp);
            enc = rb_enc_check(str, tmp);
            if ((tlen = RSTRING_LEN(tmp)) == 0) return Qtrue;
            if ((slen = RSTRING_LEN(str)) < tlen) continue;
            p = RSTRING_PTR(str);
            e = p + slen;
            s = p + tlen;
            if (!at_char_right_boundary(p, s, e, enc))
                continue;
            if (memcmp(p, RSTRING_PTR(tmp), tlen) == 0)
                return Qtrue;
        }
    }
    return Qfalse;
}

Returns whether self starts with any of the given string_or_regexp.

Matches patterns against the beginning of self. For each given string_or_regexp, the pattern is:

• string_or_regexp itself, if it is a Regexp.

• Regexp.quote(string_or_regexp), if string_or_regexp is a string.

Returns true if any pattern matches the beginning, false otherwise:

'hello'.start_with?('hell')               
'hello'.start_with?(/H/i)                 
'hello'.start_with?('heaven', 'hell')     
'hello'.start_with?('heaven', 'paradise') 
'тест'.start_with?('т')                   
'こんにちは'.start_with?('こ')              

Related: String#end_with?.

static VALUE
rb_str_strip(VALUE str)
{
    char *start;
    long olen, loffset, roffset;
    rb_encoding *enc = STR_ENC_GET(str);

    RSTRING_GETMEM(str, start, olen);
    loffset = lstrip_offset(str, start, start+olen, enc);
    roffset = rstrip_offset(str, start+loffset, start+olen, enc);

    if (loffset <= 0 && roffset <= 0) return str_duplicate(rb_cString, str);
    return rb_str_subseq(str, loffset, olen-loffset-roffset);
}

Returns a copy of the receiver with leading and trailing whitespace removed; see Whitespace in Strings:

whitespace = "\x00\t\n\v\f\r "
s = whitespace + 'abc' + whitespace
s       
s.strip 

Related: String#lstrip, String#rstrip.

static VALUE
rb_str_strip_bang(VALUE str)
{
    char *start;
    long olen, loffset, roffset;
    rb_encoding *enc;

    str_modify_keep_cr(str);
    enc = STR_ENC_GET(str);
    RSTRING_GETMEM(str, start, olen);
    loffset = lstrip_offset(str, start, start+olen, enc);
    roffset = rstrip_offset(str, start+loffset, start+olen, enc);

    if (loffset > 0 || roffset > 0) {
        long len = olen-roffset;
        if (loffset > 0) {
            len -= loffset;
            memmove(start, start + loffset, len);
        }
        STR_SET_LEN(str, len);
        TERM_FILL(start+len, rb_enc_mbminlen(enc));
        return str;
    }
    return Qnil;
}

Like String#strip, except that any modifications are made in self; returns self if any modification are made, nil otherwise.

Related: String#lstrip!, String#strip!.

static VALUE
rb_str_sub(int argc, VALUE *argv, VALUE str)
{
    str = str_duplicate(rb_cString, str);
    rb_str_sub_bang(argc, argv, str);
    return str;
}

Returns a copy of self with only the first occurrence (not all occurrences) of the given pattern replaced.

See Substitution Methods.

Related: String#sub!, String#gsub, String#gsub!.

static VALUE
rb_str_sub_bang(int argc, VALUE *argv, VALUE str)
{
    VALUE pat, repl, hash = Qnil;
    int iter = 0;
    long plen;
    int min_arity = rb_block_given_p() ? 1 : 2;
    long beg;

    rb_check_arity(argc, min_arity, 2);
    if (argc == 1) {
        iter = 1;
    }
    else {
        repl = argv[1];
        hash = rb_check_hash_type(argv[1]);
        if (NIL_P(hash)) {
            StringValue(repl);
        }
    }

    pat = get_pat_quoted(argv[0], 1);

    str_modifiable(str);
    beg = rb_pat_search(pat, str, 0, 1);
    if (beg >= 0) {
        rb_encoding *enc;
        int cr = ENC_CODERANGE(str);
        long beg0, end0;
        VALUE match, match0 = Qnil;
        struct re_registers *regs;
        char *p, *rp;
        long len, rlen;

        match = rb_backref_get();
        regs = RMATCH_REGS(match);
        if (RB_TYPE_P(pat, T_STRING)) {
            beg0 = beg;
            end0 = beg0 + RSTRING_LEN(pat);
            match0 = pat;
        }
        else {
            beg0 = BEG(0);
            end0 = END(0);
            if (iter) match0 = rb_reg_nth_match(0, match);
        }

        if (iter || !NIL_P(hash)) {
            p = RSTRING_PTR(str); len = RSTRING_LEN(str);

            if (iter) {
                repl = rb_obj_as_string(rb_yield(match0));
            }
            else {
                repl = rb_hash_aref(hash, rb_str_subseq(str, beg0, end0 - beg0));
                repl = rb_obj_as_string(repl);
            }
            str_mod_check(str, p, len);
            rb_check_frozen(str);
        }
        else {
            repl = rb_reg_regsub(repl, str, regs, RB_TYPE_P(pat, T_STRING) ? Qnil : pat);
        }

        enc = rb_enc_compatible(str, repl);
        if (!enc) {
            rb_encoding *str_enc = STR_ENC_GET(str);
            p = RSTRING_PTR(str); len = RSTRING_LEN(str);
            if (coderange_scan(p, beg0, str_enc) != ENC_CODERANGE_7BIT ||
                coderange_scan(p+end0, len-end0, str_enc) != ENC_CODERANGE_7BIT) {
                rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s",
                         rb_enc_inspect_name(str_enc),
                         rb_enc_inspect_name(STR_ENC_GET(repl)));
            }
            enc = STR_ENC_GET(repl);
        }
        rb_str_modify(str);
        rb_enc_associate(str, enc);
        if (ENC_CODERANGE_UNKNOWN < cr && cr < ENC_CODERANGE_BROKEN) {
            int cr2 = ENC_CODERANGE(repl);
            if (cr2 == ENC_CODERANGE_BROKEN ||
                (cr == ENC_CODERANGE_VALID && cr2 == ENC_CODERANGE_7BIT))
                cr = ENC_CODERANGE_UNKNOWN;
            else
                cr = cr2;
        }
        plen = end0 - beg0;
        rlen = RSTRING_LEN(repl);
        len = RSTRING_LEN(str);
        if (rlen > plen) {
            RESIZE_CAPA(str, len + rlen - plen);
        }
        p = RSTRING_PTR(str);
        if (rlen != plen) {
            memmove(p + beg0 + rlen, p + beg0 + plen, len - beg0 - plen);
        }
        rp = RSTRING_PTR(repl);
        memmove(p + beg0, rp, rlen);
        len += rlen - plen;
        STR_SET_LEN(str, len);
        TERM_FILL(&RSTRING_PTR(str)[len], TERM_LEN(str));
        ENC_CODERANGE_SET(str, cr);

        RB_GC_GUARD(match);

        return str;
    }
    return Qnil;
}

Replaces the first occurrence (not all occurrences) of the given pattern on self; returns self if a replacement occurred, nil otherwise.

See Substitution Methods.

Related: String#sub, String#gsub, String#gsub!.

VALUE
rb_str_succ(VALUE orig)
{
    VALUE str;
    str = rb_str_new(RSTRING_PTR(orig), RSTRING_LEN(orig));
    rb_enc_cr_str_copy_for_substr(str, orig);
    return str_succ(str);
}

Returns the successor to self. The successor is calculated by incrementing characters.

The first character to be incremented is the rightmost alphanumeric: or, if no alphanumerics, the rightmost character:

'THX1138'.succ 
'<<koala>>'.succ 
'***'.succ 

The successor to a digit is another digit, “carrying” to the next-left character for a “rollover” from 9 to 0, and prepending another digit if necessary:

'00'.succ 
'09'.succ 
'99'.succ 

The successor to a letter is another letter of the same case, carrying to the next-left character for a rollover, and prepending another same-case letter if necessary:

'aa'.succ 
'az'.succ 
'zz'.succ 
'AA'.succ 
'AZ'.succ 
'ZZ'.succ 

The successor to a non-alphanumeric character is the next character in the underlying character set’s collating sequence, carrying to the next-left character for a rollover, and prepending another character if necessary:

s = 0.chr * 3
s 
s.succ 
s = 255.chr * 3
s 
s.succ 

Carrying can occur between and among mixtures of alphanumeric characters:

s = 'zz99zz99'
s.succ 
s = '99zz99zz'
s.succ 

The successor to an empty String is a new empty String:

''.succ 

static VALUE
rb_str_succ_bang(VALUE str)
{
    rb_str_modify(str);
    str_succ(str);
    return str;
}

Equivalent to String#succ, but modifies self in place; returns self.

static VALUE
rb_str_sum(int argc, VALUE *argv, VALUE str)
{
    int bits = 16;
    char *ptr, *p, *pend;
    long len;
    VALUE sum = INT2FIX(0);
    unsigned long sum0 = 0;

    if (rb_check_arity(argc, 0, 1) && (bits = NUM2INT(argv[0])) < 0) {
        bits = 0;
    }
    ptr = p = RSTRING_PTR(str);
    len = RSTRING_LEN(str);
    pend = p + len;

    while (p < pend) {
        if (FIXNUM_MAX - UCHAR_MAX < sum0) {
            sum = rb_funcall(sum, '+', 1, LONG2FIX(sum0));
            str_mod_check(str, ptr, len);
            sum0 = 0;
        }
        sum0 += (unsigned char)*p;
        p++;
    }

    if (bits == 0) {
        if (sum0) {
            sum = rb_funcall(sum, '+', 1, LONG2FIX(sum0));
        }
    }
    else {
        if (sum == INT2FIX(0)) {
            if (bits < (int)sizeof(long)*CHAR_BIT) {
                sum0 &= (((unsigned long)1)<<bits)-1;
            }
            sum = LONG2FIX(sum0);
        }
        else {
            VALUE mod;

            if (sum0) {
                sum = rb_funcall(sum, '+', 1, LONG2FIX(sum0));
            }

            mod = rb_funcall(INT2FIX(1), idLTLT, 1, INT2FIX(bits));
            mod = rb_funcall(mod, '-', 1, INT2FIX(1));
            sum = rb_funcall(sum, '&', 1, mod);
        }
    }
    return sum;
}

Returns a basic n-bit checksum of the characters in self; the checksum is the sum of the binary value of each byte in self, modulo 2**n - 1:

'hello'.sum     
'hello'.sum(4)  
'hello'.sum(64) 
'тест'.sum      
'こんにちは'.sum  

This is not a particularly strong checksum.

static VALUE
rb_str_swapcase(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE;
    VALUE ret;

    flags = check_case_options(argc, argv, flags);
    enc = str_true_enc(str);
    if (RSTRING_LEN(str) == 0 || !RSTRING_PTR(str)) return str_duplicate(rb_cString, str);
    if (flags&ONIGENC_CASE_ASCII_ONLY) {
        ret = rb_str_new(0, RSTRING_LEN(str));
        rb_str_ascii_casemap(str, ret, &flags, enc);
    }
    else {
        ret = rb_str_casemap(str, &flags, enc);
    }
    return ret;
}

Returns a string containing the characters in self, with cases reversed; each uppercase character is downcased; each lowercase character is upcased:

s = 'Hello World!' 
s.swapcase         

The casing may be affected by the given options; see Case Mapping.

Related: String#swapcase!.

static VALUE
rb_str_swapcase_bang(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE;

    flags = check_case_options(argc, argv, flags);
    str_modify_keep_cr(str);
    enc = str_true_enc(str);
    if (flags&ONIGENC_CASE_ASCII_ONLY)
        rb_str_ascii_casemap(str, str, &flags, enc);
    else
        str_shared_replace(str, rb_str_casemap(str, &flags, enc));

    if (ONIGENC_CASE_MODIFIED&flags) return str;
    return Qnil;
}

Upcases each lowercase character in self; downcases uppercase character; returns self if any changes were made, nil otherwise:

s = 'Hello World!' 
s.swapcase!        
s                  
''.swapcase!       

The casing may be affected by the given options; see Case Mapping.

Related: String#swapcase.

static VALUE
string_to_c(VALUE self)
{
    VALUE num;

    rb_must_asciicompat(self);

    (void)parse_comp(rb_str_fill_terminator(self, 1), FALSE, &num);

    return num;
}

Returns self interpreted as a Complex object; leading whitespace and trailing garbage are ignored:

'9'.to_c                 
'2.5'.to_c               
'2.5/1'.to_c             
'-3/2'.to_c              
'-i'.to_c                
'45i'.to_c               
'3-4i'.to_c              
'-4e2-4e-2i'.to_c        
'-0.0-0.0i'.to_c         
'1/2+3/4i'.to_c          
'1.0@0'.to_c             
"1.0@#{Math::PI/2}".to_c 
"1.0@#{Math::PI}".to_c   

Returns Complex zero if the string cannot be converted:

'ruby'.to_c        

See Kernel#Complex.

static VALUE
rb_str_to_f(VALUE str)
{
    return DBL2NUM(rb_str_to_dbl(str, FALSE));
}

Returns the result of interpreting leading characters in self as a Float:

'3.14159'.to_f  
'1.234e-2'.to_f 

Characters past a leading valid number (in the given base) are ignored:

'3.14 (pi to two places)'.to_f 

Returns zero if there is no leading valid number:

'abcdef'.to_f 

static VALUE
rb_str_to_i(int argc, VALUE *argv, VALUE str)
{
    int base = 10;

    if (rb_check_arity(argc, 0, 1) && (base = NUM2INT(argv[0])) < 0) {
        rb_raise(rb_eArgError, "invalid radix %d", base);
    }
    return rb_str_to_inum(str, base, FALSE);
}

Returns the result of interpreting leading characters in self as an integer in the given base (which must be in (0, 2..36)):

'123456'.to_i     
'123def'.to_i(16) 

With base zero, string object may contain leading characters to specify the actual base:

'123def'.to_i(0)   
'0123def'.to_i(0)  
'0b123def'.to_i(0) 
'0o123def'.to_i(0) 
'0d123def'.to_i(0) 
'0x123def'.to_i(0) 

Characters past a leading valid number (in the given base) are ignored:

'12.345'.to_i   
'12345'.to_i(2) 

Returns zero if there is no leading valid number:

'abcdef'.to_i 
'2'.to_i(2)   

static VALUE
string_to_r(VALUE self)
{
    VALUE num;

    rb_must_asciicompat(self);

    num = parse_rat(RSTRING_PTR(self), RSTRING_END(self), 0, TRUE);

    if (RB_FLOAT_TYPE_P(num) && !FLOAT_ZERO_P(num))
        rb_raise(rb_eFloatDomainError, "Infinity");
    return num;
}

Returns the result of interpreting leading characters in str as a rational. Leading whitespace and extraneous characters past the end of a valid number are ignored. Digit sequences can be separated by an underscore. If there is not a valid number at the start of str, zero is returned. This method never raises an exception.

'  2  '.to_r       
'300/2'.to_r       
'-9.2'.to_r        
'-9.2e2'.to_r      
'1_234_567'.to_r   
'21 June 09'.to_r  
'21/06/09'.to_r    
'BWV 1079'.to_r    

NOTE: “0.3”.to_r isn’t the same as 0.3.to_r. The former is equivalent to “3/10”.to_r, but the latter isn’t so.

"0.3".to_r == 3/10r  
0.3.to_r   == 3/10r  

See also Kernel#Rational.

static VALUE
rb_str_to_s(VALUE str)
{
    if (rb_obj_class(str) != rb_cString) {
        return str_duplicate(rb_cString, str);
    }
    return str;
}

Returns self if self is a String, or self converted to a String if self is a subclass of String.

Returns the Symbol corresponding to str, creating the symbol if it did not previously exist. See Symbol#id2name.

"Koala".intern         
s = 'cat'.to_sym       
s == :cat              
s = '@cat'.to_sym      
s == :@cat             

This can also be used to create symbols that cannot be represented using the :xxx notation.

'cat and dog'.to_sym   

static VALUE
rb_str_tr(VALUE str, VALUE src, VALUE repl)
{
    str = str_duplicate(rb_cString, str);
    tr_trans(str, src, repl, 0);
    return str;
}

Returns a copy of self with each character specified by string selector translated to the corresponding character in string replacements. The correspondence is positional:

• Each occurrence of the first character specified by selector is translated to the first character in replacements.

• Each occurrence of the second character specified by selector is translated to the second character in replacements.

• And so on.

Example:

'hello'.tr('el', 'ip') 

If replacements is shorter than selector, it is implicitly padded with its own last character:

'hello'.tr('aeiou', '-')   
'hello'.tr('aeiou', 'AA-') 

Arguments selector and replacements must be valid character selectors (see Character Selectors), and may use any of its valid forms, including negation, ranges, and escaping:

'hello'.tr('^aeiou', '-') 

'ibm'.tr('b-z', 'a-z') 

'hel^lo'.tr('\^aeiou', '-')     
'i-b-m'.tr('b\-z', 'a-z')       
'foo\\bar'.tr('ab\\', 'XYZ')    

static VALUE
rb_str_tr_bang(VALUE str, VALUE src, VALUE repl)
{
    return tr_trans(str, src, repl, 0);
}

Like String#tr, but modifies self in place. Returns self if any changes were made, nil otherwise.

static VALUE
rb_str_tr_s(VALUE str, VALUE src, VALUE repl)
{
    str = str_duplicate(rb_cString, str);
    tr_trans(str, src, repl, 1);
    return str;
}

Like String#tr, but also squeezes the modified portions of the translated string; returns a new string (translated and squeezed).

'hello'.tr_s('l', 'r')   
'hello'.tr_s('el', '-')  
'hello'.tr_s('el', 'hx') 

Related: String#squeeze.

static VALUE
rb_str_tr_s_bang(VALUE str, VALUE src, VALUE repl)
{
    return tr_trans(str, src, repl, 1);
}

Like String#tr_s, but modifies self in place. Returns self if any changes were made, nil otherwise.

Related: String#squeeze!.

static VALUE
str_undump(VALUE str)
{
    const char *s = RSTRING_PTR(str);
    const char *s_end = RSTRING_END(str);
    rb_encoding *enc = rb_enc_get(str);
    VALUE undumped = rb_enc_str_new(s, 0L, enc);
    bool utf8 = false;
    bool binary = false;
    int w;

    rb_must_asciicompat(str);
    if (rb_str_is_ascii_only_p(str) == Qfalse) {
        rb_raise(rb_eRuntimeError, "non-ASCII character detected");
    }
    if (!str_null_check(str, &w)) {
        rb_raise(rb_eRuntimeError, "string contains null byte");
    }
    if (RSTRING_LEN(str) < 2) goto invalid_format;
    if (*s != '"') goto invalid_format;

    /* strip '"' at the start */
    s++;

    for (;;) {
        if (s >= s_end) {
            rb_raise(rb_eRuntimeError, "unterminated dumped string");
        }

        if (*s == '"') {
            /* epilogue */
            s++;
            if (s == s_end) {
                /* ascii compatible dumped string */
                break;
            }
            else {
                static const char force_encoding_suffix[] = ".force_encoding(\""; /* "\")" */
                static const char dup_suffix[] = ".dup";
                const char *encname;
                int encidx;
                ptrdiff_t size;

                /* check separately for strings dumped by older versions */
                size = sizeof(dup_suffix) - 1;
                if (s_end - s > size && memcmp(s, dup_suffix, size) == 0) s += size;

                size = sizeof(force_encoding_suffix) - 1;
                if (s_end - s <= size) goto invalid_format;
                if (memcmp(s, force_encoding_suffix, size) != 0) goto invalid_format;
                s += size;

                if (utf8) {
                    rb_raise(rb_eRuntimeError, "dumped string contained Unicode escape but used force_encoding");
                }

                encname = s;
                s = memchr(s, '"', s_end-s);
                size = s - encname;
                if (!s) goto invalid_format;
                if (s_end - s != 2) goto invalid_format;
                if (s[0] != '"' || s[1] != ')') goto invalid_format;

                encidx = rb_enc_find_index2(encname, (long)size);
                if (encidx < 0) {
                    rb_raise(rb_eRuntimeError, "dumped string has unknown encoding name");
                }
                rb_enc_associate_index(undumped, encidx);
            }
            break;
        }

        if (*s == '\\') {
            s++;
            if (s >= s_end) {
                rb_raise(rb_eRuntimeError, "invalid escape");
            }
            undump_after_backslash(undumped, &s, s_end, &enc, &utf8, &binary);
        }
        else {
            rb_str_cat(undumped, s++, 1);
        }
    }

    RB_GC_GUARD(str);

    return undumped;
invalid_format:
    rb_raise(rb_eRuntimeError, "invalid dumped string; not wrapped with '\"' nor '\"...\".force_encoding(\"...\")' form");
}

Returns an unescaped version of self:

s_orig = "\f\x00\xff\\\""    
s_dumped = s_orig.dump       
s_undumped = s_dumped.undump 
s_undumped == s_orig         

Related: String#dump (inverse of String#undump).

static VALUE
rb_str_unicode_normalize(int argc, VALUE *argv, VALUE str)
{
    return unicode_normalize_common(argc, argv, str, id_normalize);
}

Returns a copy of self with Unicode normalization applied.

Argument form must be one of the following symbols (see Unicode normalization forms):

• :nfc: Canonical decomposition, followed by canonical composition.

• :nfd: Canonical decomposition.

• :nfkc: Compatibility decomposition, followed by canonical composition.

• :nfkd: Compatibility decomposition.

The encoding of self must be one of:

• Encoding::UTF_8

• Encoding::UTF_16BE

• Encoding::UTF_16LE

• Encoding::UTF_32BE

• Encoding::UTF_32LE

• Encoding::GB18030

• Encoding::UCS_2BE

• Encoding::UCS_4BE

Examples:

"a\u0300".unicode_normalize      
"\u00E0".unicode_normalize(:nfd) 

Related: String#unicode_normalize!, String#unicode_normalized?.

static VALUE
rb_str_unicode_normalize_bang(int argc, VALUE *argv, VALUE str)
{
    return rb_str_replace(str, unicode_normalize_common(argc, argv, str, id_normalize));
}

Like String#unicode_normalize, except that the normalization is performed on self.

Related String#unicode_normalized?.

static VALUE
rb_str_unicode_normalized_p(int argc, VALUE *argv, VALUE str)
{
    return unicode_normalize_common(argc, argv, str, id_normalized_p);
}

Returns true if self is in the given form of Unicode normalization, false otherwise. The form must be one of :nfc, :nfd, :nfkc, or :nfkd.

Examples:

"a\u0300".unicode_normalized?       
"a\u0300".unicode_normalized?(:nfd) 
"\u00E0".unicode_normalized?        
"\u00E0".unicode_normalized?(:nfd)  

Raises an exception if self is not in a Unicode encoding:

s = "\xE0".force_encoding('ISO-8859-1')
s.unicode_normalized? 

Related: String#unicode_normalize, String#unicode_normalize!.

def unpack(fmt, offset: 0)
  Primitive.attr! :use_block
  Primitive.pack_unpack(fmt, offset)
end

Extracts data from self.

If block is not given, forming objects that become the elements of a new array, and returns that array. Otherwise, yields each object.

See Packed Data.

def unpack1(fmt, offset: 0)
  Primitive.pack_unpack1(fmt, offset)
end

Like String#unpack, but unpacks and returns only the first extracted object. See Packed Data.

static VALUE
rb_str_upcase(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE;
    VALUE ret;

    flags = check_case_options(argc, argv, flags);
    enc = str_true_enc(str);
    if (case_option_single_p(flags, enc, str)) {
        ret = rb_str_new(RSTRING_PTR(str), RSTRING_LEN(str));
        str_enc_copy_direct(ret, str);
        upcase_single(ret);
    }
    else if (flags&ONIGENC_CASE_ASCII_ONLY) {
        ret = rb_str_new(0, RSTRING_LEN(str));
        rb_str_ascii_casemap(str, ret, &flags, enc);
    }
    else {
        ret = rb_str_casemap(str, &flags, enc);
    }

    return ret;
}

Returns a string containing the upcased characters in self:

s = 'Hello World!' 
s.upcase           

The casing may be affected by the given options; see Case Mapping.

Related: String#upcase!, String#downcase, String#downcase!.

static VALUE
rb_str_upcase_bang(int argc, VALUE *argv, VALUE str)
{
    rb_encoding *enc;
    OnigCaseFoldType flags = ONIGENC_CASE_UPCASE;

    flags = check_case_options(argc, argv, flags);
    str_modify_keep_cr(str);
    enc = str_true_enc(str);
    if (case_option_single_p(flags, enc, str)) {
        if (upcase_single(str))
            flags |= ONIGENC_CASE_MODIFIED;
    }
    else if (flags&ONIGENC_CASE_ASCII_ONLY)
        rb_str_ascii_casemap(str, str, &flags, enc);
    else
        str_shared_replace(str, rb_str_casemap(str, &flags, enc));

    if (ONIGENC_CASE_MODIFIED&flags) return str;
    return Qnil;
}

Upcases the characters in self; returns self if any changes were made, nil otherwise:

s = 'Hello World!' 
s.upcase!          
s                  
s.upcase!          

The casing may be affected by the given options; see Case Mapping.

Related: String#upcase, String#downcase, String#downcase!.

static VALUE
rb_str_upto(int argc, VALUE *argv, VALUE beg)
{
    VALUE end, exclusive;

    rb_scan_args(argc, argv, "11", &end, &exclusive);
    RETURN_ENUMERATOR(beg, argc, argv);
    return rb_str_upto_each(beg, end, RTEST(exclusive), str_upto_i, Qnil);
}

With a block given, calls the block with each String value returned by successive calls to String#succ; the first value is self, the next is self.succ, and so on; the sequence terminates when value other_string is reached; returns self:

'a8'.upto('b6') {|s| print s, ' ' } 

Output:

a8 a9 b0 b1 b2 b3 b4 b5 b6

If argument exclusive is given as a truthy object, the last value is omitted:

'a8'.upto('b6', true) {|s| print s, ' ' } 

Output:

a8 a9 b0 b1 b2 b3 b4 b5

If other_string would not be reached, does not call the block:

'25'.upto('5') {|s| fail s }
'aa'.upto('a') {|s| fail s }

With no block given, returns a new Enumerator:

'a8'.upto('b6') 

static VALUE
rb_str_valid_encoding_p(VALUE str)
{
    int cr = rb_enc_str_coderange(str);

    return RBOOL(cr != ENC_CODERANGE_BROKEN);
}

Returns true if self is encoded correctly, false otherwise:

"\xc2\xa1".force_encoding("UTF-8").valid_encoding? 
"\xc2".force_encoding("UTF-8").valid_encoding?     
"\x80".force_encoding("UTF-8").valid_encoding?     

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4