Class: Hash

Inherits:
Object
  • Object
show all
Defined in:
lib/ext/hash.rb,
lib/ext/translator.rb

Constant Summary collapse

CAMA_HTML_ATTR_NAME =

Attribute names are emitted verbatim, so they are restricted to a conservative subset of what HTML allows rather than escaped: escaping cannot help here, because the characters that turn one name into two (whitespace, =, /, quotes) are not HTML metacharacters. Covers the forms actually used - id, class, data-*, aria-*, xml:lang.

/\A[a-zA-Z_:][-a-zA-Z0-9_:.]*\z/

Instance Method Summary collapse

Instance Method Details

#find_by(val, attr = 'id') ⇒ Object

used for hash of objects



47
48
49
50
51
52
 
# File 'lib/ext/hash.rb', line 47

def find_by(val, attr = 'id')
  each_value do |p|
    return p if p[attr].to_s == val.to_s
  end
  nil
end

#to_attr_format(split = ' ') ⇒ Object

convert hash to string like class="class val" name='name val'

Values are escaped for the HTML attribute context, so no value can close its own attribute and introduce another. CGI.escapeHTML is used rather than ERB::Util.html_escape deliberately: the latter is a no-op on an html_safe value, which would let a caller pass a string whose quote survives into the output. Callers of this method are building attribute values, never markup, so unconditional escaping is the correct contract.

Pairs whose key is not a valid attribute name are dropped. A key such as x onfocus=alert(1) y would otherwise render as three attributes, so there is no safe way to emit it; anything with a space in it was already producing malformed markup before this guard existed.



21
22
23
24
25
26
27
28
29
 
# File 'lib/ext/hash.rb', line 21

def to_attr_format(split = ' ')
  res = []
  each do |key, value|
    next unless key.to_s.match?(CAMA_HTML_ATTR_NAME)

    res << "#{key} = \"#{CGI.escapeHTML(value.to_s)}\""
  end
  res.join(split)
end

#to_attr_url_format ⇒ Object

convert hash to attributes for url_path

Emits a Ruby fragment (:key => "value") for code generation, not HTML, so it must NOT use the HTML escaper above — entity-encoding would corrupt the generated code. inspect produces a complete, correctly escaped double-quoted literal; the previous hand-rolled escape handled the quote but not the backslash, so a value like a\b became a backspace and a\"b produced a fragment that would not parse.



38
39
40
41
42
43
44
 
# File 'lib/ext/hash.rb', line 38

def to_attr_url_format
  res = []
  each do |key, value|
    res << ":#{key} => #{value.to_s.inspect}"
  end
  res.join ','
end

#to_sym ⇒ Object 



54
55
56
 
# File 'lib/ext/hash.rb', line 54

def to_sym
  symbolize(self)
end

#to_translate ⇒ Object

convert hash to translation string structure sample: "hola mundo", en: "Hello World" ==> Hola MundoHello World



74
75
76
77
78
79
80
 
# File 'lib/ext/translator.rb', line 74

def to_translate
  res = []
  each do |key, val|
    res << "<!--:#{key}-->#{val}<!--:-->"
  end
  res.join('')
end