1# frozen_string_literal: false
2# The HTTPHeader module defines methods for reading and writing
3# HTTP headers.
4#
5# It is used as a mixin by other classes, to provide hash-like
6# access to HTTP header values. Unlike raw hash access, HTTPHeader
7# provides access via case-insensitive keys. It also provides
8# methods for accessing commonly-used HTTP header values in more
9# convenient formats.
10#
11module Net::HTTPHeader
12
13  def initialize_http_header(initheader)
14    @header = {}
15    return unless initheader
16    initheader.each do |key, value|
17      warn "net/http: duplicated HTTP header: #{key}", uplevel: 1 if key?(key) and $VERBOSE
18      if value.nil?
19        warn "net/http: nil HTTP header: #{key}", uplevel: 1 if $VERBOSE
20      else
21        value = value.strip # raise error for invalid byte sequences
22        if value.count("\r\n") > 0
23          raise ArgumentError, "header #{key} has field value #{value.inspect}, this cannot include CR/LF"
24        end
25        @header[key.downcase.to_s] = [value]
26      end
27    end
28  end
29
30  def size   #:nodoc: obsolete
31    @header.size
32  end
33
34  alias length size   #:nodoc: obsolete
35
36  # Returns the header field corresponding to the case-insensitive key.
37  # For example, a key of "Content-Type" might return "text/html"
38  def [](key)
39    a = @header[key.downcase.to_s] or return nil
40    a.join(', ')
41  end
42
43  # Sets the header field corresponding to the case-insensitive key.
44  def []=(key, val)
45    unless val
46      @header.delete key.downcase.to_s
47      return val
48    end
49    set_field(key, val)
50  end
51
52  # [Ruby 1.8.3]
53  # Adds a value to a named header field, instead of replacing its value.
54  # Second argument +val+ must be a String.
55  # See also #[]=, #[] and #get_fields.
56  #
57  #   request.add_field 'X-My-Header', 'a'
58  #   p request['X-My-Header']              #=> "a"
59  #   p request.get_fields('X-My-Header')   #=> ["a"]
60  #   request.add_field 'X-My-Header', 'b'
61  #   p request['X-My-Header']              #=> "a, b"
62  #   p request.get_fields('X-My-Header')   #=> ["a", "b"]
63  #   request.add_field 'X-My-Header', 'c'
64  #   p request['X-My-Header']              #=> "a, b, c"
65  #   p request.get_fields('X-My-Header')   #=> ["a", "b", "c"]
66  #
67  def add_field(key, val)
68    stringified_downcased_key = key.downcase.to_s
69    if @header.key?(stringified_downcased_key)
70      append_field_value(@header[stringified_downcased_key], val)
71    else
72      set_field(key, val)
73    end
74  end
75
76  private def set_field(key, val)
77    case val
78    when Enumerable
79      ary = []
80      append_field_value(ary, val)
81      @header[key.downcase.to_s] = ary
82    else
83      val = val.to_s # for compatibility use to_s instead of to_str
84      if val.b.count("\r\n") > 0
85        raise ArgumentError, 'header field value cannot include CR/LF'
86      end
87      @header[key.downcase.to_s] = [val]
88    end
89  end
90
91  private def append_field_value(ary, val)
92    case val
93    when Enumerable
94      val.each{|x| append_field_value(ary, x)}
95    else
96      val = val.to_s
97      if /[\r\n]/n.match?(val.b)
98        raise ArgumentError, 'header field value cannot include CR/LF'
99      end
100      ary.push val
101    end
102  end
103
104  # [Ruby 1.8.3]
105  # Returns an array of header field strings corresponding to the
106  # case-insensitive +key+.  This method allows you to get duplicated
107  # header fields without any processing.  See also #[].
108  #
109  #   p response.get_fields('Set-Cookie')
110  #     #=> ["session=al98axx; expires=Fri, 31-Dec-1999 23:58:23",
111  #          "query=rubyscript; expires=Fri, 31-Dec-1999 23:58:23"]
112  #   p response['Set-Cookie']
113  #     #=> "session=al98axx; expires=Fri, 31-Dec-1999 23:58:23, query=rubyscript; expires=Fri, 31-Dec-1999 23:58:23"
114  #
115  def get_fields(key)
116    stringified_downcased_key = key.downcase.to_s
117    return nil unless @header[stringified_downcased_key]
118    @header[stringified_downcased_key].dup
119  end
120
121  # Returns the header field corresponding to the case-insensitive key.
122  # Returns the default value +args+, or the result of the block, or
123  # raises an IndexError if there's no header field named +key+
124  # See Hash#fetch
125  def fetch(key, *args, &block)   #:yield: +key+
126    a = @header.fetch(key.downcase.to_s, *args, &block)
127    a.kind_of?(Array) ? a.join(', ') : a
128  end
129
130  # Iterates through the header names and values, passing in the name
131  # and value to the code block supplied.
132  #
133  # Returns an enumerator if no block is given.
134  #
135  # Example:
136  #
137  #     response.header.each_header {|key,value| puts "#{key} = #{value}" }
138  #
139  def each_header   #:yield: +key+, +value+
140    block_given? or return enum_for(__method__) { @header.size }
141    @header.each do |k,va|
142      yield k, va.join(', ')
143    end
144  end
145
146  alias each each_header
147
148  # Iterates through the header names in the header, passing
149  # each header name to the code block.
150  #
151  # Returns an enumerator if no block is given.
152  def each_name(&block)   #:yield: +key+
153    block_given? or return enum_for(__method__) { @header.size }
154    @header.each_key(&block)
155  end
156
157  alias each_key each_name
158
159  # Iterates through the header names in the header, passing
160  # capitalized header names to the code block.
161  #
162  # Note that header names are capitalized systematically;
163  # capitalization may not match that used by the remote HTTP
164  # server in its response.
165  #
166  # Returns an enumerator if no block is given.
167  def each_capitalized_name  #:yield: +key+
168    block_given? or return enum_for(__method__) { @header.size }
169    @header.each_key do |k|
170      yield capitalize(k)
171    end
172  end
173
174  # Iterates through header values, passing each value to the
175  # code block.
176  #
177  # Returns an enumerator if no block is given.
178  def each_value   #:yield: +value+
179    block_given? or return enum_for(__method__) { @header.size }
180    @header.each_value do |va|
181      yield va.join(', ')
182    end
183  end
184
185  # Removes a header field, specified by case-insensitive key.
186  def delete(key)
187    @header.delete(key.downcase.to_s)
188  end
189
190  # true if +key+ header exists.
191  def key?(key)
192    @header.key?(key.downcase.to_s)
193  end
194
195  # Returns a Hash consisting of header names and array of values.
196  # e.g.
197  # {"cache-control" => ["private"],
198  #  "content-type" => ["text/html"],
199  #  "date" => ["Wed, 22 Jun 2005 22:11:50 GMT"]}
200  def to_hash
201    @header.dup
202  end
203
204  # As for #each_header, except the keys are provided in capitalized form.
205  #
206  # Note that header names are capitalized systematically;
207  # capitalization may not match that used by the remote HTTP
208  # server in its response.
209  #
210  # Returns an enumerator if no block is given.
211  def each_capitalized
212    block_given? or return enum_for(__method__) { @header.size }
213    @header.each do |k,v|
214      yield capitalize(k), v.join(', ')
215    end
216  end
217
218  alias canonical_each each_capitalized
219
220  def capitalize(name)
221    name.to_s.split(/-/).map {|s| s.capitalize }.join('-')
222  end
223  private :capitalize
224
225  # Returns an Array of Range objects which represent the Range:
226  # HTTP header field, or +nil+ if there is no such header.
227  def range
228    return nil unless @header['range']
229
230    value = self['Range']
231    # byte-range-set = *( "," OWS ) ( byte-range-spec / suffix-byte-range-spec )
232    #   *( OWS "," [ OWS ( byte-range-spec / suffix-byte-range-spec ) ] )
233    # corrected collected ABNF
234    # http://tools.ietf.org/html/draft-ietf-httpbis-p5-range-19#section-5.4.1
235    # http://tools.ietf.org/html/draft-ietf-httpbis-p5-range-19#appendix-C
236    # http://tools.ietf.org/html/draft-ietf-httpbis-p1-messaging-19#section-3.2.5
237    unless /\Abytes=((?:,[ \t]*)*(?:\d+-\d*|-\d+)(?:[ \t]*,(?:[ \t]*\d+-\d*|-\d+)?)*)\z/ =~ value
238      raise Net::HTTPHeaderSyntaxError, "invalid syntax for byte-ranges-specifier: '#{value}'"
239    end
240
241    byte_range_set = $1
242    result = byte_range_set.split(/,/).map {|spec|
243      m = /(\d+)?\s*-\s*(\d+)?/i.match(spec) or
244              raise Net::HTTPHeaderSyntaxError, "invalid byte-range-spec: '#{spec}'"
245      d1 = m[1].to_i
246      d2 = m[2].to_i
247      if m[1] and m[2]
248        if d1 > d2
249          raise Net::HTTPHeaderSyntaxError, "last-byte-pos MUST greater than or equal to first-byte-pos but '#{spec}'"
250        end
251        d1..d2
252      elsif m[1]
253        d1..-1
254      elsif m[2]
255        -d2..-1
256      else
257        raise Net::HTTPHeaderSyntaxError, 'range is not specified'
258      end
259    }
260    # if result.empty?
261    # byte-range-set must include at least one byte-range-spec or suffix-byte-range-spec
262    # but above regexp already denies it.
263    if result.size == 1 && result[0].begin == 0 && result[0].end == -1
264      raise Net::HTTPHeaderSyntaxError, 'only one suffix-byte-range-spec with zero suffix-length'
265    end
266    result
267  end
268
269  # Sets the HTTP Range: header.
270  # Accepts either a Range object as a single argument,
271  # or a beginning index and a length from that index.
272  # Example:
273  #
274  #   req.range = (0..1023)
275  #   req.set_range 0, 1023
276  #
277  def set_range(r, e = nil)
278    unless r
279      @header.delete 'range'
280      return r
281    end
282    r = (r...r+e) if e
283    case r
284    when Numeric
285      n = r.to_i
286      rangestr = (n > 0 ? "0-#{n-1}" : "-#{-n}")
287    when Range
288      first = r.first
289      last = r.end
290      last -= 1 if r.exclude_end?
291      if last == -1
292        rangestr = (first > 0 ? "#{first}-" : "-#{-first}")
293      else
294        raise Net::HTTPHeaderSyntaxError, 'range.first is negative' if first < 0
295        raise Net::HTTPHeaderSyntaxError, 'range.last is negative' if last < 0
296        raise Net::HTTPHeaderSyntaxError, 'must be .first < .last' if first > last
297        rangestr = "#{first}-#{last}"
298      end
299    else
300      raise TypeError, 'Range/Integer is required'
301    end
302    @header['range'] = ["bytes=#{rangestr}"]
303    r
304  end
305
306  alias range= set_range
307
308  # Returns an Integer object which represents the HTTP Content-Length:
309  # header field, or +nil+ if that field was not provided.
310  def content_length
311    return nil unless key?('Content-Length')
312    len = self['Content-Length'].slice(/\d+/) or
313        raise Net::HTTPHeaderSyntaxError, 'wrong Content-Length format'
314    len.to_i
315  end
316
317  def content_length=(len)
318    unless len
319      @header.delete 'content-length'
320      return nil
321    end
322    @header['content-length'] = [len.to_i.to_s]
323  end
324
325  # Returns "true" if the "transfer-encoding" header is present and
326  # set to "chunked".  This is an HTTP/1.1 feature, allowing the
327  # the content to be sent in "chunks" without at the outset
328  # stating the entire content length.
329  def chunked?
330    return false unless @header['transfer-encoding']
331    field = self['Transfer-Encoding']
332    (/(?:\A|[^\-\w])chunked(?![\-\w])/i =~ field) ? true : false
333  end
334
335  # Returns a Range object which represents the value of the Content-Range:
336  # header field.
337  # For a partial entity body, this indicates where this fragment
338  # fits inside the full entity body, as range of byte offsets.
339  def content_range
340    return nil unless @header['content-range']
341    m = %r<bytes\s+(\d+)-(\d+)/(\d+|\*)>i.match(self['Content-Range']) or
342        raise Net::HTTPHeaderSyntaxError, 'wrong Content-Range format'
343    m[1].to_i .. m[2].to_i
344  end
345
346  # The length of the range represented in Content-Range: header.
347  def range_length
348    r = content_range() or return nil
349    r.end - r.begin + 1
350  end
351
352  # Returns a content type string such as "text/html".
353  # This method returns nil if Content-Type: header field does not exist.
354  def content_type
355    return nil unless main_type()
356    if sub_type()
357    then "#{main_type()}/#{sub_type()}"
358    else main_type()
359    end
360  end
361
362  # Returns a content type string such as "text".
363  # This method returns nil if Content-Type: header field does not exist.
364  def main_type
365    return nil unless @header['content-type']
366    self['Content-Type'].split(';').first.to_s.split('/')[0].to_s.strip
367  end
368
369  # Returns a content type string such as "html".
370  # This method returns nil if Content-Type: header field does not exist
371  # or sub-type is not given (e.g. "Content-Type: text").
372  def sub_type
373    return nil unless @header['content-type']
374    _, sub = *self['Content-Type'].split(';').first.to_s.split('/')
375    return nil unless sub
376    sub.strip
377  end
378
379  # Any parameters specified for the content type, returned as a Hash.
380  # For example, a header of Content-Type: text/html; charset=EUC-JP
381  # would result in type_params returning {'charset' => 'EUC-JP'}
382  def type_params
383    result = {}
384    list = self['Content-Type'].to_s.split(';')
385    list.shift
386    list.each do |param|
387      k, v = *param.split('=', 2)
388      result[k.strip] = v.strip
389    end
390    result
391  end
392
393  # Sets the content type in an HTTP header.
394  # The +type+ should be a full HTTP content type, e.g. "text/html".
395  # The +params+ are an optional Hash of parameters to add after the
396  # content type, e.g. {'charset' => 'iso-8859-1'}
397  def set_content_type(type, params = {})
398    @header['content-type'] = [type + params.map{|k,v|"; #{k}=#{v}"}.join('')]
399  end
400
401  alias content_type= set_content_type
402
403  # Set header fields and a body from HTML form data.
404  # +params+ should be an Array of Arrays or
405  # a Hash containing HTML form data.
406  # Optional argument +sep+ means data record separator.
407  #
408  # Values are URL encoded as necessary and the content-type is set to
409  # application/x-www-form-urlencoded
410  #
411  # Example:
412  #    http.form_data = {"q" => "ruby", "lang" => "en"}
413  #    http.form_data = {"q" => ["ruby", "perl"], "lang" => "en"}
414  #    http.set_form_data({"q" => "ruby", "lang" => "en"}, ';')
415  #
416  def set_form_data(params, sep = '&')
417    query = URI.encode_www_form(params)
418    query.gsub!(/&/, sep) if sep != '&'
419    self.body = query
420    self.content_type = 'application/x-www-form-urlencoded'
421  end
422
423  alias form_data= set_form_data
424
425  # Set an HTML form data set.
426  # +params+ is the form data set; it is an Array of Arrays or a Hash
427  # +enctype is the type to encode the form data set.
428  # It is application/x-www-form-urlencoded or multipart/form-data.
429  # +formopt+ is an optional hash to specify the detail.
430  #
431  # boundary:: the boundary of the multipart message
432  # charset::  the charset of the message. All names and the values of
433  #            non-file fields are encoded as the charset.
434  #
435  # Each item of params is an array and contains following items:
436  # +name+::  the name of the field
437  # +value+:: the value of the field, it should be a String or a File
438  # +opt+::   an optional hash to specify additional information
439  #
440  # Each item is a file field or a normal field.
441  # If +value+ is a File object or the +opt+ have a filename key,
442  # the item is treated as a file field.
443  #
444  # If Transfer-Encoding is set as chunked, this send the request in
445  # chunked encoding. Because chunked encoding is HTTP/1.1 feature,
446  # you must confirm the server to support HTTP/1.1 before sending it.
447  #
448  # Example:
449  #    http.set_form([["q", "ruby"], ["lang", "en"]])
450  #
451  # See also RFC 2388, RFC 2616, HTML 4.01, and HTML5
452  #
453  def set_form(params, enctype='application/x-www-form-urlencoded', formopt={})
454    @body_data = params
455    @body = nil
456    @body_stream = nil
457    @form_option = formopt
458    case enctype
459    when /\Aapplication\/x-www-form-urlencoded\z/i,
460      /\Amultipart\/form-data\z/i
461      self.content_type = enctype
462    else
463      raise ArgumentError, "invalid enctype: #{enctype}"
464    end
465  end
466
467  # Set the Authorization: header for "Basic" authorization.
468  def basic_auth(account, password)
469    @header['authorization'] = [basic_encode(account, password)]
470  end
471
472  # Set Proxy-Authorization: header for "Basic" authorization.
473  def proxy_basic_auth(account, password)
474    @header['proxy-authorization'] = [basic_encode(account, password)]
475  end
476
477  def basic_encode(account, password)
478    'Basic ' + ["#{account}:#{password}"].pack('m0')
479  end
480  private :basic_encode
481
482  def connection_close?
483    token = /(?:\A|,)\s*close\s*(?:\z|,)/i
484    @header['connection']&.grep(token) {return true}
485    @header['proxy-connection']&.grep(token) {return true}
486    false
487  end
488
489  def connection_keep_alive?
490    token = /(?:\A|,)\s*keep-alive\s*(?:\z|,)/i
491    @header['connection']&.grep(token) {return true}
492    @header['proxy-connection']&.grep(token) {return true}
493    false
494  end
495
496end
497