The R Project SVN R-packages

Rev

Rev 8115 | Blame | Compare with Previous | Last modification | View Log | Download | RSS feed 

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
<title></title><link rel="stylesheet" type="text/css" href="../OmegaTech.css"></link><meta name="generator" content="DocBook XSL Stylesheets V1.76.1"></meta><meta name="keywords" content="HTTP, RCurl, R"></meta></head><body class="yui-skin-sam"><div class="article"><div class="titlepage"><div><div><div class="author"><h3 class="author"><span class="firstname">Duncan</span> <span class="surname">Temple Lang</span></h3><div class="affiliation"><span class="orgname">Department of Statistics, UC Davis<br></br></span></div></div></div></div><hr></hr></div><div class="section" title="Overview"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723591729"></a>Overview</h2></div></div></div><p></p>
The <a class="ulink" href="https://www.omegahat.net/RCurl/" target="_top">RCurl package</a> provides
high-level facilities in R to communicate with HTTP servers.  Simply,
it allows us to download URLs, submit forms in different ways, and
generally compose HTTP requests.  It supports HTTPS, the secure HTTP;
handles authentication using passwords; and can use FTP to download
files.  It also handles escaping characters in requests, binary data,
and file uploads.  Users can override or  provide additional headers in
the HTTP request in order to customize the communication.  The
response from the HTTP server is processed as a stream and chunk
encoding automatically handled.  While the default mechanism simply
returns the text from a request, one can specify S functions to
process the response as it is received, redirecting it or processing
in an application specific manner.

<p></p>

All of this could be written in R.  We could use socket connections to
write requests to HTTP servers and receive the response.  To support
HTTPS, we would have to add SSL connections.  To get the same behavior
as RCurl, we would have to implement the HTTP protocol.  This involves
writing the HTTP headers correctly and flexibly, escaping characters,
providing authentication, following redirection commands, handling
responses by decomposing "chunked" content, binary files, etc.  To
submit forms, we have to compose the body of the request by computing
boundary strings and creating the "Content-Disposition" elements.  In
short, there is a lot of work to do and significant potential for
error.  Rather than doing this in R, RCurl uses an existing
implementation that is provided in a widely used C library - 
<a class="ulink" href="https://curl.se/" target="_top">libcurl</a>.
This has several benefits:
<div class="itemizedlist"><ul class="itemizedlist" type="disc"><li class="listitem"> libcurl is well tested and very portable. As a result it
is available on many platforms and bugs have been identified and
fixed.
</li><li class="listitem">
 Many people and applications use libcurl which means that it has
support for common features that are used in various contexts.
</li><li class="listitem">
libcurl is in C and so is fast.
</li></ul></div>

On the negative side, it is hard to extend libcurl.  We can only hope
that the hooks to customize requests are adequate for our needs.  We
can expect that others have run into these extensibility issues and
these needs have been fed back to the designers of libcurl.

<p></p>
R already provides  its own way to download URIs.
Functions like <pre xmlns="" class="rfunction">download.url</pre>
<br xmlns="">
and connection constructors such as <pre xmlns="" class="rfunction">url</pre>
<br xmlns="">.


<p></p> Since the libcurl library provides opaque data structures for
implementing requests and connections, we cannot easily adapt it to
fit our special needs in R.  Specifically, it is harder to merge it
with our view of connections in S (R and S-Plus).  Similarly, it is
harder to integrate it with the event-loop in R so that we can listen
for input pending and essentially put the connection <span class="emphasis"><em>in the
background</em></span>.  The support for threads in libcurl might make this
easier if/when we support threads in R.

</div><div class="section" title="A Quick Tour"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723591796"></a>A Quick Tour</h2></div></div></div>

This document aims to provide a basic overview of the RCurl package.
It doesn't try to provide all the details.  The R function help files
and the libcurl documentation have all the relevant information.
Since the package is an interface to libcurl, it is important to use
the documentation for it regarding features, options, etc.  

<!-- You can consult the <a class="ulink" href="https://curl.se/docs"
target="_top">libcurl documentation</a> and <a class="ulink"
href="http://curl.planetmirror.com/libcurl/c/example.html"
target="_top">libcurl examples</a> (in C code). -->

<p></p>
The RCurl package provides three primary high-level entry points.
These allow us to fetch a URL and submit forms.  The functions are
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">, <pre xmlns="" class="rfunction">getForm</pre>
<br xmlns=""> and
<pre xmlns="" class="rfunction">postForm</pre>
<br xmlns="">.  The first is relatively
straightforward, given the name; it allows us to fetch the contents of
a URI.  The other two functions provide ways to submit a form using
the GET or POST methods.  These are quite different internally, but
for users, both require a set of name-value pairs giving the
parameters for the form submission.  The difference is in how the form
is submitted and the POST method allows us to submit/upload files,
binary content, etc.

<p></p>
Let us look at the <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">
function.
At it simplest, this is just like 
the <pre xmlns="" class="rfunction">download.url</pre>
<br xmlns="">
function in the standard R.
We can fetch a URI with the command something like
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl/index.html")
</pre>
<br xmlns="">
The idea is that we specify the URI.
There are several other arguments to this function,
but for the most part we don't need them.

<p></p>
We can use HTTPS to fetch URIs securely.
For example, 
<pre xmlns="" class="S">
getURL("https://sourceforge.net")
</pre>
<br xmlns="">
This is already more than we can do with the 
regular connections or built-in <pre xmlns="" class="rfunction">download.url</pre>
<br xmlns="">
in R.  (Using an external program allows HTTPS access.)

<p></p> There are three different sets of arguments for the
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> function.  One is named
<i xmlns="" class="rarg">curl</i> and we will cover this in section <a class="xref" href="#CURLHandle" title="CURL Handles">the section called &#8220;CURL Handles&#8221;</a>.  This is merely a way to cumulate requests on a
single connection with shared options.

<p></p>
The  <i xmlns="" class="rarg">write</i> function is again rather specialized.
It allows us to specify an R function that is called each time
libcurl has some text as part of the HTTP response.
It hands this text (as a sequence of bytes) to the function
so that it can process it in whatever way it deems fit.
This corresponds to the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">writefunction</font> option 
for the libcurl operation described next. We have it as an explicit
argument simply because we need to use it to get the return value
in a single action as the default behavior.


<p></p> The third set of arguments is the most general and is handled
by the ... in the <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> function.  With
this, one can specify name-value pairs governing the actual request.
There are numerous possible settings that one can specify.  The basic
idea is that one can set options provided by the
<i xmlns="">curl_easy_setopt</i> routine.  These allow us to
set parameters for many different aspects of the request.  For
example, we can specify additional headers for the HTTP request, or
include a password for the Web site.
The set of possible options can be determined
via the function 
<pre xmlns="" class="rfunction">getCurlOptionConstants</pre>
<br xmlns="">.
and the set of names for the different options
can be found via the command
<pre xmlns="" class="S">
names(getCurlOptionsConstants())
</pre>
<br xmlns="">
This is a collection of names of options that are understood
by many of the functions in the  RCurl package.

<p></p>
At present, there are 113 possible options.
<pre xmlns="" class="routput">
sort(names(getCurlOptionsConstants()))
  [1] "autoreferer"             "buffersize"             
  [3] "cainfo"                  "capath"                 
  [5] "closepolicy"             "connecttimeout"         
  [7] "cookie"                  "cookiefile"             
  [9] "cookiejar"               "cookiesession"          
 [11] "crlf"                    "customrequest"          
 [13] "debugdata"               "debugfunction"          
 [15] "dns.cache.timeout"       "dns.use.global.cache"   
 [17] "egdsocket"               "encoding"               
 [19] "errorbuffer"             "failonerror"            
 [21] "file"                    "filetime"               
 [23] "followlocation"          "forbid.reuse"           
 [25] "fresh.connect"           "ftp.create.missing.dirs"
 [27] "ftp.response.timeout"    "ftp.ssl"                
 [29] "ftp.use.eprt"            "ftp.use.epsv"           
 [31] "ftpappend"               "ftplistonly"            
 [33] "ftpport"                 "header"                 
 [35] "headerfunction"          "http.version"           
 [37] "http200aliases"          "httpauth"               
 [39] "httpget"                 "httpheader"             
 [41] "httppost"                "httpproxytunnel"        
 [43] "infile"                  "infilesize"             
 [45] "infilesize.large"        "interface"              
 [47] "ipresolve"               "krb4level"              
 [49] "low.speed.limit"         "low.speed.time"         
 [51] "maxconnects"             "maxfilesize"            
 [53] "maxfilesize.large"       "maxredirs"              
 [55] "netrc"                   "netrc.file"             
 [57] "nobody"                  "noprogress"             
 [59] "nosignal"                "port"                   
 [61] "post"                    "postfields"             
 [63] "postfieldsize"           "postfieldsize.large"    
 [65] "postquote"               "prequote"               
 [67] "private"                 "progressdata"           
 [69] "progressfunction"        "proxy"                  
 [71] "proxyauth"               "proxyport"              
 [73] "proxytype"               "proxyuserpwd"           
 [75] "put"                     "quote"                  
 [77] "random.file"             "range"                  
 [79] "readfunction"            "referer"                
 [81] "resume.from"             "resume.from.large"      
 [83] "share"                   "ssl.cipher.list"        
 [85] "ssl.ctx.data"            "ssl.ctx.function"       
 [87] "ssl.verifyhost"          "ssl.verifypeer"         
 [89] "sslcert"                 "sslcertpasswd"          
 [91] "sslcerttype"             "sslengine"              
 [93] "sslengine.default"       "sslkey"                 
 [95] "sslkeypasswd"            "sslkeytype"             
 [97] "sslversion"              "stderr"                 
 [99] "tcp.nodelay"             "telnetoptions"          
[101] "timecondition"           "timeout"                
[103] "timevalue"               "transfertext"           
[105] "unrestricted.auth"       "upload"                 
[107] "url"                     "useragent"              
[109] "userpwd"                 "verbose"                
[111] "writefunction"           "writeheader"            
[113] "writeinfo"              
</pre>
Each of these and what it controls is described in the libcurl man(ual) page
for <i xmlns="">curl_easy_setopt</i>
and that is the authoritative documentation.
Anything we provide here is merely repetition or additional
explanation.

<p></p>
The names of the options require a slight explanation.  These
correspond to symbolic names in the C code of libcurl.  For example,
the option <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">url</font> in R corresponds to
<i xmlns="" xmlns:c="http://www.C.org" xmlns:cpp="http://www.cplusplus.org" class="cenumValue">CURLOPT_URL</i> in C.  Firstly, uppercase
letters are annoying to type and read, so we have mapped them to lower
case letters in R.  We have also removed the prefix "CURLOPT_" since
we know the context in which they option names are being used.  And
lastly, any option names that have a _ (after we have removed the
CURLOPT_ prefix) are changed to replace the '_' with a '.'  so we can
type them in R without having to quote them.  For example, combining
these three rules, "CURLOPT_URL" becomes <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">url</font> and
<i xmlns="" xmlns:c="http://www.C.org" xmlns:cpp="http://www.cplusplus.org" class="cenumValue">CURLOPT_NETRC_FILE</i> becomes <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">netrc.file</font>.
That is the mapping scheme.

<p></p>
The code that handles options in RCurl automatically 
maps the user's inputs to lower case. This means
that you can use any mixture of upper-case
that makes your code more readable to you and others.
For example, we might
write
<code xmlns="" class="Sexpression">
 writeFunction = basicTextGatherer()
</code>
or
<code xmlns="" class="Sexpression">
  HTTPHeader = c(Accept="text/html")
</code><p></p>
We specify one or more options by using the names.  To make
interactive use easier, we perform partial matching on the names
relative to the set of know names.  So, for example, we could specify
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl/testPassword",
        verbose = TRUE)
</pre>
<br xmlns="">
or, more succinctly,
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl/testPassword",
        v = TRUE)
</pre>
<br xmlns="">
Obviously, the first is more readable and less ambiguous.
Please use the full form when writing
"software". But you might use the abbreviated form when
working interactively.


<p></p>
Each option expects a certain type of value from R.
For example, the following options
expect a number or logical value.
<pre xmlns="" class="routput">
 [1] "autoreferer"             "buffersize"             
 [3] "closepolicy"             "connecttimeout"         
 [5] "cookiesession"           "crlf"                   
 [7] "dns.cache.timeout"       "dns.use.global.cache"   
 [9] "failonerror"             "followlocation"         
[11] "forbid.reuse"            "fresh.connect"          
[13] "ftp.create.missing.dirs" "ftp.response.timeout"   
[15] "ftp.ssl"                 "ftp.use.eprt"           
[17] "ftp.use.epsv"            "ftpappend"              
[19] "ftplistonly"             "header"                 
[21] "http.version"            "httpauth"               
[23] "httpget"                 "httpproxytunnel"        
[25] "infilesize"              "ipresolve"              
[27] "low.speed.limit"         "low.speed.time"         
[29] "maxconnects"             "maxfilesize"            
[31] "maxredirs"               "netrc"                  
[33] "nobody"                  "noprogress"             
[35] "nosignal"                "port"                   
[37] "post"                    "postfieldsize"          
[39] "proxyauth"               "proxyport"              
[41] "proxytype"               "put"                    
[43] "resume.from"             "ssl.verifyhost"         
[45] "ssl.verifypeer"          "sslengine.default"      
[47] "sslversion"              "tcp.nodelay"            
[49] "timecondition"           "timeout"                
[51] "timevalue"               "transfertext"           
[53] "unrestricted.auth"       "upload"                 
[55] "verbose"                
</pre>
The <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">connecttimeout</font> gives the maximum number 
of seconds the connection should take before
raising an error, so this is a number.
The <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">header</font> option, on the other hand, 
is merely a flag to indicate whether header information
from the response should be included.
So this can be a logical value (or a number that is
0 to say FALSE or non-zero for TRUE.) 
At present, all numbers passed from R are converted to 
<i xmlns="" class="$class">long</i> when used in libcurl. 

<p></p>
Many options are specified as strings.
For example, we can specify  
the user password for a URI as
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl/testPassword/index.html", userpwd = "bob:duncantl", verbose = TRUE)
</pre>
<br xmlns="">
Note that we also turned on the "verbose" option so that we can see what libcurl is doing.
This is extremely convenient when trying to understand why things aren't
working (or are working in a particular way!).

<p></p>
Another example of using strings is to
specify a <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">referer</font> URI and a user-agent.
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl/index.html", useragent="RCurl", referer="https://www.omegahat.net")
</pre>
<br xmlns="">
(Again, you might want to turn on the "verbose" option
to see what libcurl  is doing with this information.)

<p></p>
The libcurl facilities allow us to not only set our own values for
fields used in the HTTP request header (such as the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">referer</font> or
user-agent), but it also allows us to set an entire collection of new
fields or replacements for any existing field.  We do this in R using
the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">httpheader</font> option for libcurl and we specify a value which is a
named character vector.
For example, suppose we want to provide a value
for the Accept field and add a new field named,
say, Made-up-field.
We could do this in the request as
<pre xmlns="" class="S">
getURL("https://www.omegahat.net/RCurl", httpheader = c(Accept="text/html", 'Made-up-field' = "bob"))
</pre>
<br xmlns="">
If you turn on the verbose option again for this request, you will see these
fields being set.
<pre xmlns="" class="routput">
&gt; getURL("https://www.omegahat.net", httpheader = c(Accept="text/html", 'Made-up-field' = "bob"), verbose = TRUE)
* About to connect() to www.omegahat.net port 80
* Connected to www.omegahat.net (169.237.46.32) port 80
&gt; GET / HTTP/1.1
Host: www.omegahat.net
Pragma: no-cache
Accept: text/html
Made-up-field: bob

</pre>
(Note that not all servers will tolerate setting header fields arbitrarily
and may return an error.)

<p></p>
The key thing to note is that headers are specified as name-value
pairs in a character vector.  R takes these and pastes the name and
value together and passes the resulting character vector to libcurl.
So while it is convenient to express the headers as 
<pre xmlns="" class="S">
 c(name = "value", name = "value")
</pre>
<br xmlns="">
if you already have the data in the form
<pre xmlns="" class="S">
 c("name: value", "name: value")
</pre>
<br xmlns="">
you can use that directly.


<p></p>

Some of the libcurl options expect a C routine.  For example, when
libcurl is receiving the response from the HTTP server, it will call
the C routine specified via the option <i xmlns="" xmlns:c="http://www.C.org" xmlns:cpp="http://www.cplusplus.org" class="cenumValue">CURLOPT_WRITEFUNCTION</i> each time
it has a full buffer of bytes.  While it is possible for us to be able
to specify a C routine from R (using
<pre xmlns="" class="rfunction">getNativeSymbolInfo</pre>
<br xmlns="">), we currently don't
support this.  Instead, it is more natural to specify an R function
which is to be called when appropriate.  And this is indeed how we do
things in RCurl.  One can specify a function for the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">writefunction</font><font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">writeheader</font> and <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">debugfunction</font> options.  (We can add support for
the others such as <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">readfunction</font>.) To use these is quite simple.  We
expect an R function that takes a single argument which is the
character of bytes to process.  The function can do what it wants with
this argument.  Typically, it will accumulate it in a persistent
variable (e.g. using closures) or process it on-the-fly such as adding
to a plot, passing it to an HTML parser, ....  

<p></p>
The function <pre xmlns="" class="rfunction">basicTextGatherer</pre>
<br xmlns=""> is an example
of the idea and this mechanism is used in
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">.  Suppose, for some reason, we wanted
to read the header information that was returned by HTTP server in the
response to our request.  (This has interesting things like cookies,
content type, etc.  that libcurl uses internally, but we may also want
to process.)  Then we would firstly use the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">header</font> option to turn on
the libcurl facility to report the response header information.
If we just do this, the header information will be included in
the text that <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> returns.
This is fine, but we will have to separate it out
by finding the first line, etc.
Instead, it is easier to ask libcurl to hand the header
information to use separate from the text/body of the response.
We can do this by creating a callback function via the
<pre xmlns="" class="rfunction">basicTextGatherer</pre>
<br xmlns=""> function.
<pre xmlns="" class="S">
 h = basicTextGatherer()
 txt = getURL("https://www.omegahat.net/RCurl", header = TRUE, headerfunction = h$update)
</pre>
<br xmlns="">
All we have done is create a collection of functions (stored in
<b xmlns="" class="$" title="">h</b>) and passed the update callback to libcurl.  Each
time libcurl receives more of the headers, it calls this function with
the header text.  It may call this just once or several times. This
depends on how large the header information is, how libcurl buffers
the information, etc.

<p></p>
Having called <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">, we have the text
from the URI. The header information is available from
<b xmlns="" class="$" title="">h</b>, specifically its <b xmlns="" class="$" title="">value</b>
function element.
<pre xmlns="" class="S">
 h$value()
</pre>
<br xmlns=""><p></p>

The <pre xmlns="" class="rfunction">debugGatherer</pre>
<br xmlns=""> is another example of a
callback that can be used with libcurl.  If we set the "verbose"
option to <i xmlns=""><code>TRUE</code></i>, libcurl will provide a lot of information about
its actions.  By default, these will be written on the console
(e.g. stderr).  In some cases, we would not want these to be on the
screen but instead, for example, displayed in a GUI or stored in a
variable for closer examination.  We can do this by providing a
callback function for the debugging output via the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">debugfunction</font>
option for libcurl.
The <pre xmlns="" class="rfunction">debugGatherer</pre>
<br xmlns=""> is a simple
one that merely cumulates its inputs in different
categories and makes them available via
the <b xmlns="" class="$" title="">value</b> function.
The setup is easy:
<pre xmlns="" class="S">
  d = debugGatherer()
  x = getURL("https://www.omegahat.net/RCurl", debugfunction=d$update, verbose = TRUE)
</pre>
<br xmlns="">
At the end of the request, again we have the text from the URI in
<b xmlns="" class="$" title="">x</b>, but we also have the debugging information.
libcurl has called our <b xmlns="" class="$" title="">update</b> function
each time it has some information (either from the
HTTP server or from its own internal dialog).
<pre xmlns="" class="routput">
(R) names(d$value())
[1] "text"      "headerIn"  "headerOut" "dataIn"    "dataOut"  
</pre>
The headerIn and headerOut fields report the
text of the header  for the response from the Web server
and for our request respectively.
Similarly, the dataIn and dataOut fields give
the body of the response and request.
And the text is just messages from libcurl.


<p></p>
We should note that not all options are (currently)) meaningful in R.
For example, it is not <span class="emphasis"><em>currently</em></span> possible to redirect
standard error for libcurl to a different <i xmlns="" class="$class">FILE*</i> via
the "stderr" option.  (In the future, we may be able to specify an R
function for writing errors from libcurl, but we have not put that in
yet.)


<div class="section" title="Forms"><div class="titlepage"><div><div><h3 class="title"><a id="id1170723592176"></a>Forms</h3></div></div></div> The RCurl package provides many additional
mechanisms for downloading URIs that R does not currently have
built-in. But perhaps the most pressing reason for developing the
RCurl package was the need to submit forms.  The <a class="xref" href="#bib:odbAccess" title="The RHTMLForms package: creating S functions from HTML forms.">odbAccess</a> package is
a package that can read an HTML page with one or more forms and create
an S function for each form that allows S users to submit the form
programmatically rather than requiring interactively browsing the
page, saving the result to a file and then loading it into R.
In order for these functions to work, we need to be able to submit the
contents of the form from S as if it came from a regular browser.
We use RCurl to do this.

<p></p>
There are two mechanisms used for submitting HTML forms: GET and POST.
Both take a set of name-value pairs giving the arguments
to parameterize the call.
The difference between the mechanisms is how these
name-value pairs are delivered to the HTTP server.
The GET method puts the name-value pairs of parameters
at the end of the URI name,
e.g.
<div class="literallayout"><p><br></br>
<br></br>
  https://www.omegahat.net/cgi-bin/form.pl?a=1&amp;b=2<br></br>
<br></br>
</p></div>
The POST method expects the name-value pairs to be
sent as the body of the HTTP request,
each put in its own "paragraph" or stanza.
This is more complicated but supports sending binary data, etc.

<p></p>
Which of the GET and POST mechanism is appropriate is specified with the HTML form itself via the
<font xmlns="" xmlns:html="http://www.w3.org/TR/html401" class="htmlAttribute">action</font> attribute of the
<font xmlns="" xmlns:html="http://www.w3.org/TR/html401" class="htmlTag">FORM</font> itself.  To the user, however, the browser
takes care of figuring out the correct way to deliver the name-value
pairs specified by the user when interacting with the components of
the form.  In RCurl, we don't have access to the original HTML form so
we cannot tell what mechanism to use.  It is up to the caller to
determine whether to use <pre xmlns="" class="rfunction">getForm</pre>
<br xmlns=""> or
<pre xmlns="" class="rfunction">postForm</pre>
<br xmlns=""> depending on the value of the
<font xmlns="" xmlns:html="http://www.w3.org/TR/html401" class="htmlAttribute">action</font> attribute in the original HTML
file.  


<p></p>
After determining whether to use
POST or GET, the interface to the functions is typically the same to the
user.  Essentially, she need only specify the name-value pairs for
each of the form elements.  We do this via a named list or named
character vector. (The list simply allows us to have objects of
different type other than strings!)  We must specify all the fields,
including the hidden fields, if the the processor on the HTTP server
is to make sense of it.  RCurl doesn't try to interpret the name-value
pairs, but just transports them.

<p></p>
Let's look at an example of sending a query to Google
(via HTTP rather than its API).
<pre xmlns="" class="S">
  getForm("http://www.google.com/search", hl="en", lr="", ie="ISO-8859-1",  q="RCurl", btnG="Search")
</pre>
<br xmlns="">
The result is the HTML you would ordinarily see in your browser.
You might use
<pre xmlns="" class="rfunction">htmlTreeParse</pre>
<br xmlns="">
to parse it.
What is important in the example is that we are specifying the required fields
in the query as named arguments to R. 
<pre xmlns="" class="rfunction">getForm</pre>
<br xmlns=""> takes care of bringing them together and constructing
the full URI name. Note that libcurl also handles escaping the 
special characters, e.g. converting a space to %20.
Note that if you wanted to explicitly do this escaping on a string 
rather than having libcurl implicitly do it, you can 
use <pre xmlns="" class="rfunction">curlEscape</pre>
<br xmlns="">.
Similarly, there is a function
<pre xmlns="" class="rfunction">curlUnescape</pre>
<br xmlns=""> to reverse the escaping and make a string "human-readable".

<p></p><pre xmlns="" class="rfunction">postForm</pre>
<br xmlns=""> is almost identical.  Let's submit a POST form
to <code>http://www.speakeasy.org/~cgires/perl_form.cgi</code> (no
longer available).
<!-- <a class="ulink" href="http://www.speakeasy.org/~cgires/perl_form.cgi" target="_top">http://www.speakeasy.org/~cgires/perl_form.cgi</a><pre xmlns="" class="S"> -->
  postForm("http://www.speakeasy.org/~cgires/perl_form.cgi",
            "some_text" = "Duncan",
            "choice" = "Ho",
            "radbut" = "eep",
            "box" = "box1, box2"
          )
</pre>
<br xmlns="">
Here, the form elements are named some_text, choice, radbut, box.  We
have simply provided values for them.  Again, the result is the
regular response from the HTTP server.

<p></p>
Sometimes we already have the arguments in a list.  It is slightly
more complex then to pass them to the function via the <b xmlns="">...</b>
argument.  The two form submission functions in RCurl
(<pre xmlns="" class="rfunction">getForm</pre>
<br xmlns=""> and
<pre xmlns="" class="rfunction">postForm</pre>
<br xmlns="">) also accept the name-value
arguments via the <b xmlns="">...</b> parameter.  This arises in programmatic
access to the functions rather than interactive use.

<p></p>

Since we use <b xmlns="">...</b> for the name-value pairs of the form, we cannot
specify the libcurl options (unambiguously) in this way and we require
than any such options to control the HTTP request at the libcurl-level
be passed via the <i xmlns="" class="rarg">.opts</i> parameter.  RCurl and libcurl
construct the HTTP request and after that, the request is just like a
regular URI download. All of the usual techniques for reading the
response, its header, etc.  work.


</div></div><div class="section" title="CURL Handles"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="CURLHandle"></a>CURL Handles</h2></div></div></div>

The functions we have presented above are the high-level entry points
that allow R users to make the common-style HTTP requests.  The RCurl
package is capable of more however. It provides access to the basic
libcurl primitives which one can use to compose more complicated and
non-standard HTTP requests.  For the most part, one merely specifies
libcurl options by name to the different functions and these take
effect for that call.  An alternative model (used more in C code) is
that we first create a libcurl object to represent the HTTP request,
then we customize it by setting options and then we invoke the
request.  This is far more involved than we need in R. There is a
simplicity about the <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> function that
removes the need to know about the internal C structure representing
the call.  However, there are occasions when it is useful to know
about this and exploit it.  Specifically, one can create an instance
of this libcurl "handle" and use it in several requests.  This has the
advantage that we do not have to set the options in each call, but
rather can do this just once.  This saves a marginal amount of time in
R by reducing the computations, but it will be essentially negligible
relative to the network latency involved in the request itself.  What
is more important is that if the sequence of requests are to the same
server, the libcurl engine can maintain the connection to the server
and avoid having to reestablish it each time. This handshaking is
quite expensive, so reusing the "handle" in such situations can yield
non-trivial performance gains.  It is also even possible to "pipeline"
requests by sending multiple requests before getting the answer back
for the first one. This again can improve performance.

<p></p>
Now that we both know about the internal libcurl structures and know
why we might be interested in reusing them across requests, the
question remains how do we do this.  It is quite easy.  Each of the
"action" functions in the package (i.e. that work with libcurl
directly) have a parameter named <i xmlns="" class="rarg">curl</i>.  For each of
these functions, the default value is
<pre xmlns="" class="rfunction">getCurlHandle</pre>
<br xmlns=""> and what this means is that, if
no value is given for <i xmlns="" class="rarg">curl</i>, a new handle is created for
the duration of this call.
So it is easy for us to create such a handle before calling
one of these functions and then pass that as the value 
for <i xmlns="" class="rarg">curl</i>.
For example, we can make two requests to the
www.omegahat.net site using the same handle
as follows:
<pre xmlns="" class="S">
 handle = getCurlHandle()
 a = getURL("https://www.omegahat.net/RCurl", curl = handle)
 b = getURL("https://www.omegahat.net/", curl = handle)
</pre>
<br xmlns=""><p></p>
It is important to remember that if we set any options in any of the
calls, these will be set in the libcurl handle and these will persist
across requests unless they are reset.  For example, if we had set the
<code xmlns="" class="Sexpression">header=TRUE</code> option in the first call
above, it would remain set for the second call. This can be sometimes
inconvenient. In such cases, either use separate libcurl handles, or
reset the options.


<p></p>
The function 
<pre xmlns="" class="rfunction">dupCurlHandle</pre>
<br xmlns=""> allows us to
create a new libcurl handle that is an exact copy of the
existing one.  This allows us to quickly reuse
existing settings without having them affect
other requests.
(The data in the option values are not copied).
See <i xmlns="">curl_easy_duphandle</i>.

<p></p>
By reusing libcurl handles, we avoid reallocating
a new one and potentially benefit from improved connectivity.
One downside, however, when reusing handles is that the options we set in
R need to be copied as C data since they will persist across
R function calls in the libcurl handle itself.
As a result, there are additional computations needed.
Again, this is negligible in almost all cases
and will be dominated by the network speed.


<p></p>
libcurl doesn't have any explicit function for fetching a
URL. Instead, it uses a powerful but simple interface which involves
merely setting the options in the libcurl handle as desired and then
invoking the request.  So one just prepares the request and forces it
to be sent.  This is done via the <pre xmlns="" class="rfunction">curlPerform</pre>
<br xmlns="">
function in R. This is how <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> is actually
implemented.

</div><div class="section" title="Statistics on the Request"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592381"></a>Statistics on the Request</h2></div></div></div>
S is a statistical programming language and environment
so why not gather data when we can.
The function
<pre xmlns="" class="rfunction">getCurlInfo</pre>
<br xmlns="">
allows us to find out information about the
last request made for a given libcurl handle.
(Of course, this assumes we have explicitly 
created the handle rather than used the default
value for <i xmlns="" class="rarg">curl</i> and so lost it.)
<pre xmlns="" class="S">
h = getCurlHandle()
getURL("https://www.omegahat.net", curl = h)
names(getCurlInfo(h))
</pre>
<br xmlns="">
The names of the resulting elements are
<pre xmlns="" class="routput">
 [1] "effective.url"           "response.code"          
 [3] "total.time"              "namelookup.time"        
 [5] "connect.time"            "pretransfer.time"       
 [7] "size.upload"             "size.download"          
 [9] "speed.download"          "speed.upload"           
[11] "header.size"             "request.size"           
[13] "ssl.verifyresult"        "filetime"               
[15] "content.length.download" "content.length.upload"  
[17] "starttransfer.time"      "content.type"           
[19] "redirect.time"           "redirect.count"         
[21] "private"                 "http.connectcode"       
[23] "httpauth.avail"          "proxyauth.avail"        
</pre>
These provide us the actual name of the URI downloaded after
redirections, etc.; information about the transfer speed, etc.; etc.
See the man page for <i xmlns="">curl_easy_getinfo</i>.

</div><div class="section" title="libcurl Version Information"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592415"></a>libcurl Version Information</h2></div></div></div>
The RCurl package provides a way to obtain
reflectance information about libcurl itself.
The function <pre xmlns="" class="rfunction">curlVersion</pre>
<br xmlns="">
returns the contents of the
<b xmlns="" xmlns:c="http://www.C.org" xmlns:cpp="http://www.cplusplus.org" class="cstruct">curl_version_info_data</b> structure.

<p></p>
For my  installation,
the return value from <pre xmlns="" class="rfunction">curlVersion</pre>
<br xmlns=""> is
<pre xmlns="" class="routput">
$age
[1] 2

$version
[1] "7.12.0"

$vesion_num
[1] 461824

$host
[1] "powerpc-apple-darwin7.4.0"

$features
     ipv6       ssl      libz      ntlm largefile 
        1         4         8        16       512 

$ssl_version
[1] " OpenSSL/0.9.7b"

$ssl_version_num
[1] 9465903

$libz_version
[1] "1.2.1"

$protocols
[1] "ftp"    "gopher" "telnet" "dict"   "ldap"   "http"   "file"   "https" 
[9] "ftps"  

$ares
[1] ""

$ares_num
[1] 0

$libidn
[1] ""
</pre>

The help page for the R function
explains the fields which are hopefully
clear from the names.
The only ones that might be obscure are
<b xmlns="" class="$" title="">ares</b> and <b xmlns="" class="$" title="">libidn</b>.
<b xmlns="" class="$" title="">ares</b> refers to asynchronous domain name server (DNS) lookup
for resolving the IP address (e.g. 128.41.12.2) corresponding to a machine name (e.g. www.omegahat.net).
"GNU Libidn is an implementation of the Stringprep, Punycode and IDNA specifications defined by the IETF Internationalized Domain Names (IDN)"
(taken from <a class="ulink" href="http://www.gnu.org/software/libidn/" target="_top">http://www.gnu.org/software/libidn/</a>).

</div><div class="section" title="Initialization"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592460"></a>Initialization</h2></div></div></div> As with most C libraries, one must
typically initialize it to get the basic run-time structure
established.  Fortunately, when we call any of the R functions, this
is taken care of implicitly; libcurl handles this when we create a new
handle.  However, sometimes it is important to explicitly specify
options to control how the library is initialized.  RCurl provides a
way to do this via the <pre xmlns="" class="rfunction">curlGlobalInit</pre>
<br xmlns="">
function.  The only argument is a flag that indicates what features to
initialize.  For the most part, the defaults work best and we can
leave libcurl to perform this initialization. However, we may need to
be careful that we are not re-initializing a setting that R (or
generally the host application) has already set.  This may happen on
Windows as libcurl initialize the Win32 socket library.
We can avoid this, if necessary, but telling libcurl to initialize
only the SSL facilities.

<p></p>
The argument to <pre xmlns="" class="rfunction">curlGlobalInit</pre>
<br xmlns=""> is typically a
character vector of names of features to turn on.  The possible names
can be obtained from <b xmlns="" class="$" title="">CurlGlobalBits</b> which is a named
integer vector:
<pre xmlns="" class="routput">
 none   ssl win32   all 
    0     1     2     3 
attr(,"class")
[1] "CurlGlobalBits" "BitIndicator"  
</pre>
We would call <pre xmlns="" class="rfunction">curlGlobalInit</pre>
<br xmlns=""> as
<pre xmlns="" class="S">
curlGlobalInit(c("ssl", "win32"))
</pre>
<br xmlns="">
or
<pre xmlns="" class="S">
curlGlobalInit(c("ssl"))
</pre>
<br xmlns="">
to activate both SSL and Win32 sockets,
or just SSL respectively.

<p></p>
One can specify integer values directly, but this is
less readable to others (or yourself in a few weeks!).
The names are converted and combined to a flag using
<pre xmlns="" class="rfunction">setBitIndicators</pre>
<br xmlns="">.

</div><div class="section" title="Options"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592509"></a>Options</h2></div></div></div>

 It is hopefully clear that it is the libcurl options that make this
 interface work and allow us to make interesting queries. From
 specifying the URI to how to read the text, to providing passwords,
 it is the options that are critical.  For the most part, these
 options are passed by name to functions in RCurl via the <b xmlns="">...</b>
 mechanism in R and the <i xmlns="" class="rarg">.opts</i> argument.
These two collections of arguments are merged, with those 
in <b xmlns="">...</b> overriding corresponding ones in the
<i xmlns="" class="rarg">.opts</i> object.
<p></p>
  Why do we have the <i xmlns="" class="rarg">.opts</i> argument?
The reason is similar to the <i xmlns="" class="rarg">.params</i>
in the form functions: often we have the options
in a list and it is not as convenient to use
the <b xmlns="">...</b> approach.
Having both allows the caller/programmer to use
whichever is most convenient.
<p></p>

One case in which the <i xmlns="" class="rarg">.opts</i> argument is useful is if we
want to prepare a set of options that are to be used in all (or a set
of) calls.  We can combine these arguments into a list just once and
then pass them to each HTTP request easily by simply using that
variable.  Since we merge the values in <b xmlns="">...</b> and
<i xmlns="" class="rarg">.opts</i>, this works nicely.

<p></p>
To create such a list of options, we can use the function
<pre xmlns="" class="rfunction">curlOpts</pre>
<br xmlns="">.  This creates an S3-style object
with class <i xmlns="">CURLOptions</i>.  This function never
involves libcurl, but sorts out the names of the options by using
partial matching (via the <pre xmlns="" class="rfunction">mapCurlOptNames</pre>
<br xmlns="">
function) and returns an R object with the options as name-value pairs in a list.
The fact that this is a class means that if we access any elements, the
full names are used, even when we set an element. This means that the
names are kept resolved as we use it in R and correspond unambiguously
to real libcurl options.

<p></p>
We can use this function something like the following.
<pre xmlns="" class="S">
opts = curlOptions(header = TRUE, userpwd = "bob:duncantl", netrc = TRUE)
getURL("https://www.omegahat.net/RCurl/testPassword/index.html", verbose = TRUE, .opts = opts)
</pre>
<br xmlns="">
Here we create the options ahead of time and use them in a call while
specifying additional options (i.e. "verbose").

<p></p>
Some readers will have noticed that we could achieve the same effect
of having a set of fixed options that are used in a collection of
calls by reusing a libcurl handle.  We could create the handle, set
the common options, and then use that handle in the set of calls.
This is indeed a natural and often good way to do things.
The following code does what we want.
<pre xmlns="" class="S">
h = getCurlHandle(header = TRUE, userpwd = "bob:duncantl", netrc = TRUE)
getURL("https://www.omegahat.net/RCurl/testPassword/index.html", verbose = TRUE, curl = h)
</pre>
<br xmlns="">
The first line creates a new handle and fills in the three
"persistent" options.  These are in the handle itself, not in R at
this stage.  Now, when we perform the request via
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">, we specify this libcurl handle and
provide the "verbose" option.

<p></p>
The function <pre xmlns="" class="rfunction">curlSetOpt</pre>
<br xmlns=""> is used implicitly in
the code above and this actually sets the option-values in a libcurl
handle.  It can also be used to simply resolve them.

</div><div class="section" title="Example Application"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592603"></a>Example Application</h2></div></div></div>

In this section, we will outline a more complex use of the RCurl
facilities.  Specifically, we will use it to send SOAP <a class="xref" href="#bib:SOAP" title="Programming Web Services with SOAP">SOAP</a> requests.


SOAP uses HTTP to send XML content that encodes a method invocation.
We have to add the appropriate fields to the HTTP header to identify 
the SOAP call and then insert the XML that defines the method call
in the body of the request.
Using our favorite client SOAP facility
(e.g. SOAP::Lite, <a class="ulink" href="https://www.omegahat.net/SSOAP/" target="_top">SSOAP</a>), 
we can send the SOAP request that performs the HTTP request. 
We can find out what the actual HTTP request looks like using
facilities in those tools or actually sniffing the packets as they go across the wire using
tcpdump or ethereal or some such tool.
This is what we see.
The HTTP header in the request is
<div class="literallayout"><p><br></br>
<br></br>
POST /hibye.cgi HTTP/1.1<br></br>
Connection: close<br></br>
Accept: text/xml<br></br>
Accept: multipart/*<br></br>
Host: services.soaplite.com<br></br>
User-Agent: SOAP::Lite/Perl/0.55<br></br>
Content-Length: 450<br></br>
Content-Type: text/xml; charset=utf-8<br></br>
SOAPAction: "http://www.soaplite.com/Demo#hi"<br></br>
<br></br>
</p></div>
The body of the request is
<div class="literallayout"><p><br></br>
<br></br>
&lt;?xml version="1.0" encoding="UTF-8"?&gt;<br></br>
&lt;SOAP-ENV:Envelope SOAP-ENV:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/" <br></br>
                   xmlns:SOAP-ENV="http://schemas.xmlsoap.org/soap/envelope/" <br></br>
                   xmlns:xsd="http://www.w3.org/1999/XMLSchema" <br></br>
                   xmlns:SOAP-ENC="http://schemas.xmlsoap.org/soap/encoding/" <br></br>
                   xmlns:xsi="http://www.w3.org/1999/XMLSchema-instance"&gt;<br></br>
  &lt;SOAP-ENV:Body&gt;<br></br>
       &lt;namesp1:hi xmlns:namesp1="http://www.soaplite.com/Demo"/&gt;<br></br>
  &lt;/SOAP-ENV:Body&gt;<br></br>
&lt;/SOAP-ENV:Envelope&gt;<br></br>
<br></br>
</p></div>

So we need to add fields to the HTTP header.
Specifically, we need 
<div class="literallayout"><p><br></br>
Accept: text/xml<br></br>
Accept: multipart/*<br></br>
SOAPAction: "http://www.soaplite.com/Demo#hi"<br></br>
Content-Type: text/xml; charset=utf-8<br></br>
</p></div>
libcurl should take care of the Content-Length field.

The body is specified for the HTTP request using the <font xmlns="" xmlns:curl="https://curl.se/" color="red" class="curlOpt">postfields</font> option.
To do this using the <i xmlns=""><a href="https://cran.r-project.org/package=RCurl">RCurl</a></i> package, we use
the following code.
<pre xmlns="" class="S">

body = '&lt;?xml version="1.0" encoding="UTF-8"?&gt;\
&lt;SOAP-ENV:Envelope SOAP-ENV:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/" \
                   xmlns:SOAP-ENV="http://schemas.xmlsoap.org/soap/envelope/" \
                   xmlns:xsd="http://www.w3.org/1999/XMLSchema" \
                   xmlns:SOAP-ENC="http://schemas.xmlsoap.org/soap/encoding/" \
                   xmlns:xsi="http://www.w3.org/1999/XMLSchema-instance"&gt;\
  &lt;SOAP-ENV:Body&gt;\
       &lt;namesp1:hi xmlns:namesp1="http://www.soaplite.com/Demo"/&gt;\
  &lt;/SOAP-ENV:Body&gt;\
&lt;/SOAP-ENV:Envelope&gt;\n'

curlPerform(url="http://services.soaplite.com/hibye.cgi",
            httpheader=c(Accept="text/xml", Accept="multipart/*", SOAPAction='"http://www.soaplite.com/Demo#hi"',
                         'Content-Type' = "text/xml; charset=utf-8"),
            postfields=body,
            verbose = TRUE
           )

</pre>
<br xmlns="">
Note that this similar to calling <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> and we have used it to
illustrate how we can use <pre xmlns="" class="rfunction">curlPerform</pre>
<br xmlns="">
directly. The only difference is that the result is printed to the
console, not returned to us as a character vector.  This is a problem
when we really want to process the response.  So for that, we would
simply replace the call to <pre xmlns="" class="rfunction">curlPerform</pre>
<br xmlns=""> with
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">:
<pre xmlns="" class="S">
curlPerform(url="http://services.soaplite.com/hibye.cgi",
            httpheader=c(Accept="text/xml", Accept="multipart/*", SOAPAction='"http://www.soaplite.com/Demo#hi"',
                         'Content-Type' = "text/xml; charset=utf-8"),
            postfields=body,
            verbose = TRUE
           )
</pre>
<br xmlns=""></div><div class="section" title="Additional Notes"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="id1170723592682"></a>Additional Notes</h2></div></div></div>
These were written before the package was finished.
They are left here, but may not be helpful.


<div class="section" title="Providing Text Handlers"><div class="titlepage"><div><div><h3 class="title"><a id="id1170723592688"></a>Providing Text Handlers</h3></div></div></div> 

One can provide a different text handler to consume/process the text
that is received by the libcurl engine from the HTTP response.  We can
do this with either an S function or with a C routine.  At the lowest
level, there is a C routine, but R users will typically be most
comfortable with a higher-level function.  Functions are more robust,
and also provide a more obvious way of maintaining state.  We use
closures and environments in R to endow a function with its own local
variables that persist across calls.

<p></p>

One can envisage a scenario in which the text handler would want to be
able to access the CURL handle that was being used in the request. For
example, it might use this to find the base URL, to determine the
actual host, or identify settings that are in effect, or simply reuse
the handle (by duplicating it).  Rather than make the curl object
explicitly available, one can initialize it separately ahead of the
call to <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns="">, and then make it available to the R
function as a variable in the function's environment.


A more realistic example is the following.  Suppose we want to parse
HTML files and follow the links within those files to find their
links. This is a spider or 'bot.  We can fetch the entire document via
<pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> and then parse it (using
<pre xmlns="" class="rfunction">htmlTreeParse</pre>
<br xmlns="">).  Alternatively, we can use
<pre xmlns="" class="rfunction">htmlTreeParse</pre>
<br xmlns=""> and provide it with a connection from
which the HTML/XML parser requests content as it is needed.  We can
then setup this connection from which the XML parser reads by having
it be supplied by the <pre xmlns="" class="rfunction">getURL</pre>
<br xmlns=""> text gatherer.  We can parse a document
and collect the names of all the links to which it refers and then
process each of them.  An alternative is to process an individual link
when it is discovered. There are trade-offs between the two
approaches. However, it is good to be able to do both.

<p></p> 

In order to do this, we might want to have access to the CURL
handle within the XML parser.  When we handle an HREF element
(i.e. &lt;a href=&gt;), we would then duplicate the handle and start
another HTML parser.


<p></p>


Obviously, this example is also somewhat contrived as the XML/HTML
parsing facilities have their own HTTP facilities.  However, they do
not understand all the finer points of HTTP such as SSL, FTP, passwords,
etc.
<div class="note" title="Note" style="margin-left: 0.5in; margin-right: 0.5in;"><table border="0" summary="Note"><tr><td rowspan="2" align="center" valign="top" width="25"><img alt="[Note]" src="http://docbook.sourceforge.net/release/xsl/current/images/note.svg"></img></td><th align="left">Note</th></tr><tr><td align="left" valign="top">This is not a very compelling example anymore!</td></tr></table></div></div><div class="section" title="Alternatives"><div class="titlepage"><div><div><h3 class="title"><a id="id1170723592740"></a>Alternatives</h3></div></div></div><p>
 Using libcurl is by no means the only
approach to getting HTTP access in R.  Firstly, we have HTTP access in
R via the facilities incorporated from libxml (nanohttp and nanoftp).
These are, as the names suggest, basic implementations of the protocols
and do not provide all the bells and whistles we might need generally.
Also, they are not customizable from within R. Specifically, we cannot
add header fields, handle binary data, set the body of the request, etc.
</p><p>
We can use R's socket connections and implement the details of HTTP
ourselves.  There is a great deal of work in this as we have discussed
before. Also, we currently don't have secure sockets (i.e. using SSL)
in R<sup>[<a id="id1170723592757" href="#ftn.id1170723592757" class="footnote">1</a>]</sup> I initially started using this approach
so that I could discover the nuances of HTTP.  It quickly gets
overwhelming to handle all the details.  It is more tedious than
technically challenging, especially when others have done it already
in C libraries and done it well.  The code that I have is in an
unreleased package named <i xmlns="">httpClient</i>. If anyone is interested, please
contact me.  Using R's sockets is also used in the <a class="ulink" href="https://cran.r-project.org/package=httpRequest" target="_top">httpRequest</a>
package on CRAN.  This allows submitting forms and retrieving URIs.
It is useful and, as the authors state, a "basic HTTP request"
implementation.  It doesn't escape characters, handle chunked
responses, do redirects, support SSL, etc.  It is flexible but leaves
a lot to the user to do to setup the request and process the response.
RCurl inherits many, many good features for "free" from libcurl.
</p><p>
libcurl is not the only C-level library that we could have used.
Alternative libraries include <a class="ulink" href="https://dev.w3.org/libwww/Library//" target="_top">libwww</a> from the W3 group.  We
<span class="emphasis"><em>may</em></span> find that that is more suitable, but libcurl
will definitely suffice for the present.
<!-- </p></div><div class="section" title="Notes"><div class="titlepage"><div><div><h3 class="title"><a id="id1170723592792"></a>Notes</h3></div></div></div><p> -->
<!-- libcurl can use <a class="ulink" href="ftp://athena-dist.mit.edu/pub/ATHENA/ares/" target="_top">ares</a> for -->
<!-- asynchronous DNS resolution. -->
</p></div></div><div class="section"><div class="titlepage"></div><div class="bibliography" title="Bibliography"><div class="titlepage"><div><div><h3 class="title"><a id="id1170723592811"></a>Bibliography</h3></div></div></div><div class="biblioentry" title="The RHTMLForms package: creating S functions from HTML forms."><a id="bib:odbAccess"></a><p>[1] <span class="authorgroup"><span class="firstname">Duncan</span> <span class="surname">Temple Lang</span>, <span class="firstname">Sandrine</span> <span class="surname">Dudoit</span>, and <span class="firstname">Sunduz</span> <span class="surname">Keles</span>. </span><span class="title"><em>The RHTMLForms package: creating S functions from HTML forms.</em>. </span><a class="ulink" href="https://www.omegahat.net/RHTMLForms/" target="_top">RHTMLForms</a></p></div><div class="biblioentry" title="Programming Web Services with SOAP"><a id="bib:SOAP"></a><p>[2] <span class="authorgroup"><span class="firstname">James</span> <span class="surname">Snell</span>, <span class="firstname">Doug</span> <span class="surname">Tidwell</span>, and <span class="firstname">Pavel</span> <span class="surname">Kulchenko</span>. </span><span class="title"><em>Programming Web Services with SOAP</em>. </span><a class="ulink" href="http://www.oreilly.com/catalog/progwebsoap/" target="_top">O'Reilly</a></p></div></div></div><div class="footnotes"><br></br><hr width="100" align="left"></hr><div class="footnote"><p><sup>[<a id="ftn.id1170723592757" href="#id1170723592757" class="para">1</a>] </sup>I have a local version (not with SSL) but they are not
connections since the connection data structure is not exposed in the
R API, yet!</p></div></div></div></body></html>