aboutsummaryrefslogtreecommitdiffstats
path: root/doc/src/guide/multipart.asciidoc
blob: 630b2107343666eb75481cf0333b37df71e8f178 (plain) (blame)
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
[[multipart]]
== Multipart requests

Multipart originates from MIME, an Internet standard that
extends the format of emails.

A multipart message is a list of parts. A part contains
headers and a body. The body of the parts may be
of any media type, and contain text or binary data.
It is possible for parts to contain a multipart media
type.

In the context of HTTP, multipart is most often used
with the `multipart/form-data` media type. It is what
browsers use to upload files through HTML forms.

The `multipart/byteranges` is also common. It is the
media type used to send arbitrary bytes from a resource,
enabling clients to resume downloads.

=== Form-data

In the normal case, when a form is submitted, the
browser will use the `application/x-www-form-urlencoded`
content-type. This type is just a list of keys and
values and is therefore not fit for uploading files.

That's where the `multipart/form-data` content-type
comes in. When the form is configured to use this
content-type, the browser will create a multipart
message where each part corresponds to a field on
the form. For files, it also adds some metadata in
the part headers, like the file name.

A form with a text input, a file input and a select
choice box will result in a multipart message with
three parts, one for each field.

The browser does its best to determine the media type
of the files it sends this way, but you should not
rely on it for determining the contents of the file.
Proper investigation of the contents is recommended.

=== Checking for multipart messages

The content-type header indicates the presence of
a multipart message:

[source,erlang]
----
{<<"multipart">>, <<"form-data">>, _}
    = cowboy_req:parse_header(<<"content-type">>, Req).
----

=== Reading a multipart message

Cowboy provides two sets of functions for reading
request bodies as multipart messages.

The `cowboy_req:read_part/1,2` functions return the
next part's headers, if any.

The `cowboy_req:read_part_body/1,2` functions return
the current part's body. For large bodies you may
need to call the function multiple times.

To read a multipart message you need to iterate over
all its parts:

[source,erlang]
----
multipart(Req0) ->
    case cowboy_req:read_part(Req0) of
        {ok, _Headers, Req1} ->
            {ok, _Body, Req} = cowboy_req:read_part_body(Req1),
            multipart(Req);
        {done, Req} ->
            Req
    end.
----

When part bodies are too large, Cowboy will return
a `more` tuple, and allow you to loop until the part
body has been fully read.

The function `cow_multipart:form_data/1` can be used
to quickly obtain information about a part from a
`multipart/form-data` message. The function returns
a `data` or a `file` tuple depending on whether this
is a normal field or a file being uploaded.

The following snippet will use this function and
use different strategies depending on whether the
part is a file:

[source,erlang]
----
multipart(Req0) ->
    case cowboy_req:read_part(Req0) of
        {ok, Headers, Req1} ->
            Req = case cow_multipart:form_data(Headers) of
                {data, _FieldName} ->
                    {ok, _Body, Req2} = cowboy_req:read_part_body(Req1),
                    Req2;
                {file, _FieldName, _Filename, _CType, _CTransferEncoding} ->
                    stream_file(Req1)
            end,
            multipart(Req);
        {done, Req} ->
            Req
    end.

stream_file(Req0) ->
    case cowboy_req:read_part_body(Req0) of
        {ok, _Body, Req} ->
            Req;
        {more, _Body, Req} ->
            stream_file(Req)
    end.
----

Both the part header and body reading functions can take
options that will be given to the request body reading
functions. By default, `cowboy_req:read_part/1` reads
up to 64KB for up to 5 seconds. `cowboy_req:read_part_body/1`
has the same defaults as `cowboy_req:read_body/1`.

To change the defaults for part headers:

[source,erlang]
cowboy_req:read_part(Req, #{length => 128000}).

And for part bodies:

[source,erlang]
cowboy_req:read_part_body(Req, #{length => 1000000, period => 7000}).

=== Skipping unwanted parts

Part bodies do not have to be read. Cowboy will automatically
skip it when you request the next part's body.

The following snippet reads all part headers and skips
all bodies:

[source,erlang]
----
multipart(Req0) ->
    case cowboy_req:part(Req0) of
        {ok, _Headers, Req} ->
            multipart(Req);
        {done, Req} ->
            Req
    end.
----

Similarly, if you start reading the body and it ends up
being too big, you can simply continue with the next part.
Cowboy will automatically skip what remains.

While Cowboy can skip part bodies automatically, the read
rate is not configurable. Depending on your application
you may want to skip manually, in particular if you observe
poor performance while skipping.

You do not have to read all parts either. You can stop
reading as soon as you find the data you need.

// @todo Cover the building of multipart messages.