]> git.ipfire.org Git - thirdparty/cups.git/blob - cups/api-httpipp.shtml
Merge changes from CUPS 1.4svn-r7874.
[thirdparty/cups.git] / cups / api-httpipp.shtml
1 <!--
2 "$Id: api-httpipp.shtml 7684 2008-06-23 16:47:38Z mike $"
3
4 HTTP and IPP API introduction for the Common UNIX Printing System (CUPS).
5
6 Copyright 2007-2008 by Apple Inc.
7 Copyright 1997-2006 by Easy Software Products, all rights reserved.
8
9 These coded instructions, statements, and computer programs are the
10 property of Apple Inc. and are protected by Federal copyright
11 law. Distribution and use rights are outlined in the file "LICENSE.txt"
12 which should have been included with this file. If this file is
13 file is missing or damaged, see the license at "http://www.cups.org/".
14 -->
15
16 <h2 class='title'><a name='OVERVIEW'>Overview</a></h2>
17
18 <p>The CUPS HTTP and IPP APIs provide low-level access to the HTTP and IPP
19 protocols and CUPS scheduler. They are typically used by monitoring and
20 administration programs to perform specific functions not supported by the
21 high-level CUPS API functions.</p>
22
23 <p>The HTTP APIs use an opaque structure called
24 <a href='#http_t'><code>http_t</code></a> to manage connections to
25 a particular HTTP or IPP server. The
26 <a href='#httpConnectEncrypt'><code>httpConnectEncrypt</code></a> function is
27 used to create an instance of this structure for a particular server.
28 The constant <code>CUPS_HTTP_DEFAULT</code> can be used with all of the
29 <code>cups</code> functions to refer to the default CUPS server - the functions
30 create a per-thread <a href='#http_t'><code>http_t</code></a> as needed.</p>
31
32 <p>The IPP APIs use two structures for requests (messages sent to the CUPS
33 scheduler) and responses (messages sent back to your application from the
34 scheduler). The <a href='#ipp_t'><code>ipp_t</code></a> structure holds a
35 complete request or response and is allocated using the
36 <a href='#ippNew'><code>ippNew</code></a> or
37 <a href='#ippNewRequest'><code>ippNewRequest</code></a> functions and
38 freed using the <a href='#ippDelete'><code>ippDelete</code></a> function.</p>
39
40 <p>The second structure is called
41 <a href='#ipp_attribute_t'><code>ipp_attribute_t</code></a> and holds a
42 single IPP attribute which consists of a group tag (<code>group_tag</code>), a
43 value type tag (<code>value_tag</code>), the attribute name (<code>name</code>),
44 and 1 or more values (<code>values[]</code>). Attributes are added to an
45 <a href='#ipp_t'><code>ipp_t</code></a> structure using one of the
46 <code>ippAdd</code> functions. For example, use
47 <a href='#ippAddString'><code>ippAddString</code></a> to add a
48 "requesting-user-name" string attribute to a request:</p>
49
50 <pre class='example'>
51 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(IPP_GET_JOBS);
52
53 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "requesting-user-name",
54 NULL, cupsUser());
55 </pre>
56
57 <p>Once you have created an IPP request, use the <code>cups</code>
58 functions to send the request to and read the response from the server.
59 For example, the <a href='#cupsDoRequest'><code>cupsDoRequest</code></a>
60 function can be used for simple query operations that do not involve files:</p>
61
62 <pre class='example'>
63 #include &lt;cups/cups.h&gt;
64
65
66 <a href='#ipp_t'>ipp_t</a> *<a name='get_jobs'>get_jobs</a>(void)
67 {
68 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(IPP_GET_JOBS);
69
70 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "requesting-user-name",
71 NULL, cupsUser());
72
73 return (<a href='#cupsDoRequest'>cupsDoRequest</a>(CUPS_HTTP_DEFAULT, request, "/"));
74 }
75 </pre>
76
77 <p>The <a href='#cupsDoRequest'><code>cupsDoRequest</code></a> function frees
78 the request structure and returns an IPP response structure or NULL pointer if
79 the request could not be sent to the server. Once you have a response from
80 the server, you can either use the
81 <a href='#ippFindAttribute'><code>ippFindAttribute</code></a> and
82 <a href='#ippFindNextAttribute'><code>ippFindNextAttribute</code></a> functions
83 to find specific attributes, for example:</p>
84
85 <pre class='example'>
86 <a href='#ipp_t'>ipp_t</a> *response;
87 <a href='#ipp_attribute_t'>ipp_attribute_t</a> *attr;
88
89 attr = <a href='#ippFindAttribute'>ippFindAttribute</a>(response, "printer-state", IPP_TAG_ENUM);
90 </pre>
91
92 <p>You can also walk the list of attributes with a simple <code>for</code> loop
93 like this:</p>
94
95 <pre class='example'>
96 <a href='#ipp_t'>ipp_t</a> *response;
97 <a href='#ipp_attribute_t'>ipp_attribute_t</a> *attr;
98
99 for (attr = response->attrs; attr != NULL; attr = attr->next)
100 if (attr->name == NULL)
101 puts("--SEPARATOR--");
102 else
103 puts(attr->name);
104 </pre>
105
106 <p>The <code>for</code> loop approach is normally used when collecting
107 attributes for multiple objects (jobs, printers, etc.) in a response. Attributes
108 with <code>NULL</code> names indicate a separator between the attributes of
109 each object. For example, the following code will list the jobs returned from
110 our previous <a href='#get_jobs'><code>get_jobs</code></a> example code:</p>
111
112 <pre class='example'>
113 <a href='#ipp_t'>ipp_t</a> *response = <a href='#get_jobs'>get_jobs</a>();
114
115 if (response != NULL)
116 {
117 <a href='#ipp_attribute_t'>ipp_attribute_t</a> *attr;
118 int job_id = 0;
119 char *job_name = NULL;
120 char *job_originating_user_name = NULL;
121
122 puts("Job ID Owner Title");
123 puts("------ ---------------- ---------------------------------");
124
125 for (attr = response->attrs; attr != NULL; attr = attr->next)
126 {
127 /* Attributes without names are separators between jobs */
128 if (attr->name == NULL)
129 {
130 if (job_id > 0 &amp;&amp; job_name != NULL &amp;&amp; job_originating_user_name != NULL)
131 printf("%5d %-16s %s\n", job_id, job_originating_user_name, job_name);
132
133 job_id = 0;
134 job_name = NULL;
135 job_originating_user_name = NULL;
136 continue;
137 }
138 else if (!strcmp(attr->name, "job-id") &amp;&amp; attr->value_tag == IPP_TAG_INTEGER)
139 job_id = attr->values[0].integer;
140 else if (!strcmp(attr->name, "job-name") &amp;&amp; attr->value_tag == IPP_TAG_NAME)
141 job_name = attr->values[0].string.text;
142 else if (!strcmp(attr->name, "job-originating-user-name") &amp;&amp;
143 attr->value_tag == IPP_TAG_NAME)
144 job_originating_user_name = attr->values[0].string.text;
145 }
146
147 if (job_id > 0 &amp;&amp; job_name != NULL &amp;&amp; job_originating_user_name != NULL)
148 printf("%5d %-16s %s\n", job_id, job_originating_user_name, job_name);
149 }
150 </pre>
151
152 <h3><a name='CREATING_URI_STRINGS'>Creating URI Strings</a></h3>
153
154 <p>To ensure proper encoding, the
155 <a href='#httpAssembleURIf'><code>httpAssembleURIf</code></a> function must be
156 used to format a "printer-uri" string for all printer-based requests:</p>
157
158 <pre class='example'>
159 const char *name = "Foo";
160 char uri[1024];
161 <a href='#ipp_t'>ipp_t</a> *request;
162
163 <a href='#httpAssembleURIf'>httpAssembleURIf</a>(HTTP_URI_CODING_ALL, uri, sizeof(uri), "ipp", NULL, cupsServer(),
164 ippPort(), "/printers/%s", name);
165 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_URI, "printer-uri", NULL, uri);
166 </pre>
167
168 <h3><a name='SENDING_REQUESTS_WITH_FILES'>Sending Requests with Files</a></h3>
169
170 <p>The <a href='#cupsDoFileRequest'><code>cupsDoFileRequest</code></a> and
171 <a href='#cupsDoIORequest'><code>cupsDoIORequest</code></a> functions are
172 used for requests involving files. The
173 <a href='#cupsDoFileRequest'><code>cupsDoFileRequest</code></a> function
174 attaches the named file to a request and is typically used when sending a print
175 file or changing a printer's PPD file:</p>
176
177 <pre class='example'>
178 const char *filename = "/usr/share/cups/data/testprint.ps";
179 const char *name = "Foo";
180 char uri[1024];
181 char resource[1024];
182 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(IPP_PRINT_JOB);
183 <a href='#ipp_t'>ipp_t</a> *response;
184
185 /* Use httpAssembleURIf for the printer-uri string */
186 <a href='#httpAssembleURIf'>httpAssembleURIf</a>(HTTP_URI_CODING_ALL, uri, sizeof(uri), "ipp", NULL, cupsServer(),
187 ippPort(), "/printers/%s", name);
188 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_URI, "printer-uri", NULL, uri);
189 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "requesting-user-name",
190 NULL, cupsUser());
191 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "job-name",
192 NULL, "testprint.ps");
193
194 /* Use snprintf for the resource path */
195 snprintf(resource, sizeof(resource), "/printers/%s", name);
196
197 response = <a href='#cupsDoFileRequest'>cupsDoFileRequest</a>(CUPS_HTTP_DEFAULT, request, resource, filename);
198 </pre>
199
200 <p>The <a href='#cupsDoIORequest'><code>cupsDoIORequest</code></a> function
201 optionally attaches a file to the request and optionally saves a file in the
202 response from the server. It is used when using a pipe for the request
203 attachment or when using a request that returns a file, currently only
204 <code>CUPS_GET_DOCUMENT</code> and <code>CUPS_GET_PPD</code>. For example,
205 the following code will download the PPD file for the sample HP LaserJet
206 printer driver:</p>
207
208 <pre class='example'>
209 char tempfile[1024];
210 int tempfd;
211 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(CUPS_GET_PPD);
212 <a href='#ipp_t'>ipp_t</a> *response;
213
214 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "ppd-name",
215 NULL, "laserjet.ppd");
216
217 tempfd = cupsTempFd(tempfile, sizeof(tempfile));
218
219 response = <a href='#cupsDoIORequest'>cupsDoIORequest</a>(CUPS_HTTP_DEFAULT, request, "/", -1, tempfd);
220 </pre>
221
222 <p>The example passes <code>-1</code> for the input file descriptor to specify
223 that no file is to be attached to the request. The PPD file attached to the
224 response is written to the temporary file descriptor we created using the
225 <code>cupsTempFd</code> function.</p>
226
227 <h3><a name='ASYNCHRONOUS_REQUEST_PROCESSING'>Asynchronous Request Processing</a></h3>
228
229 <p>The <a href='#cupsSendRequest'><code>cupsSendRequest</code></a> and
230 <a href='#cupsGetResponse'><code>cupsGetResponse</code></a> support
231 asynchronous communications with the server. Unlike the other request
232 functions, the IPP request is not automatically freed, so remember to
233 free your request with the <a href='#ippDelete'><code>ippDelete</code></a>
234 function.</p>
235
236 <p>File data is attached to the request using the
237 <a href='#cupsWriteRequestData'><code>cupsWriteRequestData</code></a>
238 function, while file data returned from the server is read using the
239 <a href='#cupsReadResponseData'><code>cupsReadResponseData</code></a>
240 function. We can rewrite the previous <code>CUPS_GET_PPD</code> example
241 to use the asynchronous functions quite easily:</p>
242
243 <pre class='example'>
244 char tempfile[1024];
245 int tempfd;
246 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(CUPS_GET_PPD);
247 <a href='#ipp_t'>ipp_t</a> *response;
248
249 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "ppd-name",
250 NULL, "laserjet.ppd");
251
252 tempfd = cupsTempFd(tempfile, sizeof(tempfile));
253
254 if (<a href='#cupsSendRequest'>cupsSendRequest</a>(CUPS_HTTP_DEFAULT, request, "/") == HTTP_CONTINUE)
255 {
256 response = <a href='#cupsGetResponse'>cupsGetResponse</a>(CUPS_HTTP_DEFAULT, "/");
257
258 if (response != NULL)
259 {
260 ssize_t bytes;
261 char buffer[8192];
262
263 while ((bytes = <a href='#cupsReadResponseData'>cupsReadResponseData</a>(CUPS_HTTP_DEFAULT, buffer, sizeof(buffer))) > 0)
264 write(tempfd, buffer, bytes);
265 }
266 }
267
268 /* Free the request! */
269 <a href='#ippDelete'>ippDelete</a>(request);
270 </pre>
271
272 <p>The <a href='#cupsSendRequest'><code>cupsSendRequest</code></a> function
273 returns the initial HTTP request status, typically either
274 <code>HTTP_CONTINUE</code> or <code>HTTP_UNAUTHORIZED</code>. The latter status
275 is returned when the request requires authentication of some sort. The
276 <a href='#cupsDoAuthentication'><code>cupsDoAuthentication</code></a> function
277 must be called when your see <code>HTTP_UNAUTHORIZED</code> and the request
278 re-sent. We can add authentication support to our example code by using a
279 <code>do ... while</code> loop:</p>
280
281 <pre class='example'>
282 char tempfile[1024];
283 int tempfd;
284 <a href='#ipp_t'>ipp_t</a> *request = <a href='#ippNewRequest'>ippNewRequest</a>(CUPS_GET_PPD);
285 <a href='#ipp_t'>ipp_t</a> *response;
286 http_status_t status;
287
288 <a href='#ippAddString'>ippAddString</a>(request, IPP_TAG_OPERATION, IPP_TAG_NAME, "ppd-name",
289 NULL, "laserjet.ppd");
290
291 tempfd = cupsTempFd(tempfile, sizeof(tempfile));
292
293 /* Loop for authentication */
294 do
295 {
296 status = <a href='#cupsSendRequest'>cupsSendRequest</a>(CUPS_HTTP_DEFAULT, request, "/");
297
298 if (status == HTTP_UNAUTHORIZED)
299 {
300 /* Try to authenticate, break out of the loop if that fails */
301 if (<a href='#cupsDoAuthentication'>cupsDoAuthentication</a>(CUPS_HTTP_DEFAULT, "POST", "/"))
302 break;
303 }
304 }
305 while (status != HTTP_CONTINUE &amp;&amp; status != HTTP_UNAUTHORIZED);
306
307 if (status == HTTP_CONTINUE)
308 {
309 response = <a href='#cupsGetResponse'>cupsGetResponse</a>(CUPS_HTTP_DEFAULT, "/");
310
311 if (response != NULL)
312 {
313 ssize_t bytes;
314 char buffer[8192];
315
316 while ((bytes = <a href='#cupsReadResponseData'>cupsReadResponseData</a>(CUPS_HTTP_DEFAULT, buffer, sizeof(buffer))) > 0)
317 write(tempfd, buffer, bytes);
318 }
319 }
320
321 /* Free the request! */
322 <a href='#ippDelete'>ippDelete</a>(request);
323 </pre>