aboutsummaryrefslogtreecommitdiff
path: root/docs/LIBCURL-STRUCTS
blob: 11dee85398b395e1a8f8cff72b5db7af644d29e5 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
                                  _   _ ____  _
                              ___| | | |  _ \| |
                             / __| | | | |_) | |
                            | (__| |_| |  _ <| |___
                             \___|\___/|_| \_\_____|

Structs in libcurl

This document should cover 7.32.0 pretty accurately, but will make sense even
for older and later versions as things don't change drastically that often.

 1. The main structs in libcurl
  1.1 SessionHandle
  1.2 connectdata
  1.3 Curl_multi
  1.4 Curl_handler
  1.5 conncache
  1.6 Curl_share
  1.7 CookieInfo

==============================================================================

1. The main structs in libcurl

  1.1 SessionHandle

  The SessionHandle handle struct is the one returned to the outside in the
  external API as a "CURL *". This is usually known as an easy handle in API
  documentations and examples.

  Information and state that is related to the actual connection is in the
  'connectdata' struct. When a transfer is about to be made, libcurl will
  either create a new connection or re-use an existing one. The particular
  connectdata that is used by this handle is pointed out by
  SessionHandle->easy_conn.

  Data and information that regard this particular single transfer is put in
  the SingleRequest sub-struct.

  When the SessionHandle struct is added to a multi handle, as it must be in
  order to do any transfer, the ->multi member will point to the Curl_multi
  struct it belongs to. The ->prev and ->next members will then be used by the
  multi code to keep a linked list of SessionHandle structs that are added to
  that same multi handle. libcurl always uses multi so ->multi *will* point to
  a Curl_multi when a transfer is in progress.

  ->mstate is the multi state of this particular SessionHandle. When
  multi_runsingle() is called, it will act on this handle according to which
  state it is in. The mstate is also what tells which sockets to return for a
  specific SessionHandle when curl_multi_fdset() is called etc.

  The libcurl source code generally use the name 'data' for the variable that
  points to the SessionHandle.

  When doing multiplexed HTTP/2 transfers, each SessionHandle is associated
  with an individual stream, sharing the same connectdata struct. Multiplexing
  makes it even more important to keep things associated with the right thing!

  1.2 connectdata

  A general idea in libcurl is to keep connections around in a connection
  "cache" after they have been used in case they will be used again and then
  re-use an existing one instead of creating a new as it creates a significant
  performance boost.

  Each 'connectdata' identifies a single physical connection to a server. If
  the connection can't be kept alive, the connection will be closed after use
  and then this struct can be removed from the cache and freed.

  Thus, the same SessionHandle can be used multiple times and each time select
  another connectdata struct to use for the connection. Keep this in mind, as
  it is then important to consider if options or choices are based on the
  connection or the SessionHandle.

  Functions in libcurl will assume that connectdata->data points to the
  SessionHandle that uses this connection (for the moment).

  As a special complexity, some protocols supported by libcurl require a
  special disconnect procedure that is more than just shutting down the
  socket. It can involve sending one or more commands to the server before
  doing so. Since connections are kept in the connection cache after use, the
  original SessionHandle may no longer be around when the time comes to shut
  down a particular connection. For this purpose, libcurl holds a special
  dummy 'closure_handle' SessionHandle in the Curl_multi struct to 

  FTP uses two TCP connections for a typical transfer but it keeps both in
  this single struct and thus can be considered a single connection for most
  internal concerns.

  The libcurl source code generally use the name 'conn' for the variable that
  points to the connectdata.


  1.3 Curl_multi

  Internally, the easy interface is implemented as a wrapper around multi
  interface functions. This makes everything multi interface.

  Curl_multi is the multi handle struct exposed as "CURLM *" in external APIs.

  This struct holds a list of SessionHandle structs that have been added to
  this handle with curl_multi_add_handle(). The start of the list is ->easyp
  and ->num_easy is a counter of added SessionHandles.

  ->msglist is a linked list of messages to send back when
  curl_multi_info_read() is called. Basically a node is added to that list
  when an individual SessionHandle's transfer has completed.

  ->hostcache points to the name cache. It is a hash table for looking up name
  to IP. The nodes have a limited life time in there and this cache is meant
  to reduce the time for when the same name is wanted within a short period of
  time.

  ->timetree points to a tree of SessionHandles, sorted by the remaining time
  until it should be checked - normally some sort of timeout. Each
  SessionHandle has one node in the tree.

  ->sockhash is a hash table to allow fast lookups of socket descriptor to
  which SessionHandle that uses that descriptor. This is necessary for the
  multi_socket API.

  ->conn_cache points to the connection cache. It keeps track of all
  connections that are kept after use. The cache has a maximum size.

  ->closure_handle is described in the 'connectdata' section.

  The libcurl source code generally use the name 'multi' for the variable that
  points to the Curl_multi struct.


  1.4 Curl_handler

  Each unique protocol that is supported by libcurl needs to provide at least
  one Curl_handler struct. It defines what the protocol is called and what
  functions the main code should call to deal with protocol specific issues.
  In general, there's a source file named [protocol].c in which there's a
  "struct Curl_handler Curl_handler_[protocol]" declared. In url.c there's
  then the main array with all individual Curl_handler structs pointed to from
  a single array which is scanned through when a URL is given to libcurl to
  work with.

  ->scheme is the URL scheme name, usually spelled out in uppercase. That's
  "HTTP" or "FTP" etc. SSL versions of the protcol need its own Curl_handler
  setup so HTTPS separate from HTTP.

  ->setup_connection is called to allow the protocol code to allocate protocol
  specific data that then gets associated with that SessionHandle for the rest
  of this transfer. It gets freed again at the end of the transfer. It will be
  called before the 'connectdata' for the transfer has been selected/created.
  Most protocols will allocate its private 'struct [PROTOCOL]' here and assign
  SessionHandle->req.protop to point to it.

  ->connect_it allows a protocol to do some specific actions after the TCP
  connect is done, that can still be considered part of the connection phase.

  Some protocols will alter the connectdata->recv[] and connectdata->send[]
  function pointers in this function.

  ->connecting is similarly a function that keeps getting called as long as the
  protocol considers itself still in the connecting phase.

  ->do_it is the function called to issue the transfer request. What we call
  the DO action internally. If the DO is not enough and things need to be kept
  getting done for the entire DO sequence to complete, ->doing is then usually
  also provided. Each protocol that needs to do multiple commands or similar
  for do/doing need to implement their own state machines (see SCP, SFTP,
  FTP). Some protocols (only FTP and only due to historical reasons) has a
  separate piece of the DO state called DO_MORE.

  ->doing keeps getting called while issuing the transfer request command(s)

  ->done gets called when the transfer is complete and DONE. That's after the
  main data has been transferred.

  ->do_more gets called during the DO_MORE state. The FTP protocol uses this
  state when setting up the second connection.

  ->proto_getsock
  ->doing_getsock
  ->domore_getsock
  ->perform_getsock
  Functions that return socket information. Which socket(s) to wait for which
  action(s) during the particular multi state.

  ->disconnect is called immediately before the TCP connection is shutdown.

  ->readwrite gets called during transfer to allow the protocol to do extra
  reads/writes

  ->defport is the default report TCP or UDP port this protocol uses

  ->protocol is one or more bits in the CURLPROTO_* set. The SSL versions have
  their "base" protocol set and then the SSL variation. Like "HTTP|HTTPS".

  ->flags is a bitmask with additional information about the protocol that will
  make it get treated differently by the generic engine:

    PROTOPT_SSL - will make it connect and negotiate SSL

    PROTOPT_DUAL - this protocol uses two connections

    PROTOPT_CLOSEACTION - this protocol has actions to do before closing the
    connection. This flag is no longer used by code, yet still set for a bunch
    protocol handlers.
  
    PROTOPT_DIRLOCK - "direction lock". The SSH protocols set this bit to
    limit which "direction" of socket actions that the main engine will
    concern itself about.

    PROTOPT_NONETWORK - a protocol that doesn't use network (read file:)

    PROTOPT_NEEDSPWD - this protocol needs a password and will use a default
    one unless one is provided

    PROTOPT_NOURLQUERY - this protocol can't handle a query part on the URL
    (?foo=bar)


  1.5 conncache

  Is a hash table with connections for later re-use. Each SessionHandle has
  a pointer to its connection cache. Each multi handle sets up a connection
  cache that all added SessionHandles share by default.


  1.6 Curl_share
  
  The libcurl share API allocates a Curl_share struct, exposed to the external
  API as "CURLSH *".

  The idea is that the struct can have a set of own versions of caches and
  pools and then by providing this struct in the CURLOPT_SHARE option, those
  specific SessionHandles will use the caches/pools that this share handle
  holds.

  Then individual SessionHandle structs can be made to share specific things
  that they otherwise wouldn't, such as cookies.

  The Curl_share struct can currently hold cookies, DNS cache and the SSL
  session cache.

  
  1.7 CookieInfo

  This is the main cookie struct. It holds all known cookies and related
  information. Each SessionHandle has its own private CookieInfo even when
  they are added to a multi handle. They can be made to share cookies by using
  the share API.