1 | #include "listsetsaction.h"
|
---|
2 |
|
---|
3 | #if defined(GSDL_USE_STL_H)
|
---|
4 | #include <fstream.h>
|
---|
5 | #else
|
---|
6 | #include <fstream>
|
---|
7 | #endif
|
---|
8 |
|
---|
9 | #include "recptprototools.h"
|
---|
10 | #include "oaitools.h"
|
---|
11 |
|
---|
12 | bool listsetsaction::validateAction(recptproto *protocol, oaiargs ¶ms)
|
---|
13 | {
|
---|
14 | // ----------------------------------------------------------------------------
|
---|
15 | // 1. Check for invalid arguments
|
---|
16 | // ----------------------------------------------------------------------------
|
---|
17 | bool invalid_argument_supplied = false;
|
---|
18 | text_tmap::const_iterator param_iterator = params.begin();
|
---|
19 | while (param_iterator != params.end())
|
---|
20 | {
|
---|
21 | // Check for arguments that aren't valid for this action
|
---|
22 | if (param_iterator->first != "verb" &&
|
---|
23 | param_iterator->first != "resumptionToken")
|
---|
24 | {
|
---|
25 | // We've found an invalid argument
|
---|
26 | invalid_argument_supplied = true;
|
---|
27 |
|
---|
28 | // Delete the invalid argument from the list so it doesn't end up in the <request> tag that is returned
|
---|
29 | params.erase(param_iterator->first);
|
---|
30 | }
|
---|
31 |
|
---|
32 | param_iterator++;
|
---|
33 | }
|
---|
34 |
|
---|
35 | // If we found an invalid argument it's an error, so don't go any further
|
---|
36 | if (invalid_argument_supplied)
|
---|
37 | {
|
---|
38 | this->errorType = "badArgument";
|
---|
39 | return false;
|
---|
40 | }
|
---|
41 |
|
---|
42 | // ----------------------------------------------------------------------------
|
---|
43 | // 2. Handle any exclusive arguments
|
---|
44 | // ----------------------------------------------------------------------------
|
---|
45 |
|
---|
46 | // The resumptionToken argument is exclusive
|
---|
47 | if (params["resumptionToken"] != "")
|
---|
48 | {
|
---|
49 | // This argument is exclusive, so no other arguments are allowed (except "verb" of course)
|
---|
50 | if (params.getSize() != 2)
|
---|
51 | {
|
---|
52 | this->errorType = "badArgument";
|
---|
53 | return false;
|
---|
54 | }
|
---|
55 |
|
---|
56 | // Check the resumption token is valid
|
---|
57 | ResumptionToken token(params["resumptionToken"]);
|
---|
58 | if (true) // TO DO: Fix this (the token.isValid() function is useless for ListSets)
|
---|
59 | {
|
---|
60 | // Everything is fine, and we don't continue further because this is an exclusive argument
|
---|
61 | this->errorType = "";
|
---|
62 | return true;
|
---|
63 | }
|
---|
64 | else
|
---|
65 | {
|
---|
66 | // There was an error with the resumption token
|
---|
67 | this->errorType = "badResumptionToken";
|
---|
68 | return false;
|
---|
69 | }
|
---|
70 | }
|
---|
71 |
|
---|
72 | // ----------------------------------------------------------------------------
|
---|
73 | // 3. Handle any required arguments
|
---|
74 | // ----------------------------------------------------------------------------
|
---|
75 |
|
---|
76 | // None!
|
---|
77 |
|
---|
78 | // ----------------------------------------------------------------------------
|
---|
79 | // 4. Check any remaining arguments
|
---|
80 | // ----------------------------------------------------------------------------
|
---|
81 |
|
---|
82 | // None!
|
---|
83 |
|
---|
84 | // If we've reached here everything must be fine
|
---|
85 | this->errorType = "";
|
---|
86 | return true;
|
---|
87 | }
|
---|
88 |
|
---|
89 | bool listsetsaction::output_content(ostream &output, recptproto *protocol, text_tset &collections, oaiargs ¶ms)
|
---|
90 | {
|
---|
91 | text_tset::iterator here = collections.begin();
|
---|
92 | text_tset::iterator end = collections.end();
|
---|
93 | while (here != end) {
|
---|
94 | text_t collect = *here;
|
---|
95 | this->output_content(output, protocol, params);
|
---|
96 | ++here;
|
---|
97 | }
|
---|
98 | return true;
|
---|
99 | }
|
---|
100 |
|
---|
101 | bool listsetsaction::output_content(ostream &output, recptproto *protocol, oaiargs ¶ms)
|
---|
102 | {
|
---|
103 | // output the total list of classifier points
|
---|
104 |
|
---|
105 | // variables required
|
---|
106 | text_t browseOID = "browse";
|
---|
107 | FilterResponse_t response;
|
---|
108 | comerror_t err;
|
---|
109 | text_tarray & collections = this->configuration->getCollectionsList();
|
---|
110 | text_tset metadata;
|
---|
111 | ofstream logout("oai.log", ios::app);
|
---|
112 |
|
---|
113 | // get a list of the collections available
|
---|
114 | // protocol->get_collection_list(collections, err, output);
|
---|
115 | if (collections.size() == 0) {
|
---|
116 | logout << "Found *no* OAI collections - check main.cfg for oaicollection items and read the OAI documentation.\n";
|
---|
117 | }
|
---|
118 |
|
---|
119 | // check resumption token
|
---|
120 | int startSet = 0;
|
---|
121 | if (params["resumptionToken"] != "") {
|
---|
122 | ResumptionToken token(params["resumptionToken"]);
|
---|
123 | startSet = token.getPosition() - 1; // first document is said to be 1..
|
---|
124 | }
|
---|
125 | this->replyToken = NULL;
|
---|
126 |
|
---|
127 | this->setNumber = 0;
|
---|
128 | this->setsOutput = 0;
|
---|
129 | for(int current_col = 0; current_col < collections.size(); ++current_col) {
|
---|
130 | // output the collection as a set, first, then its children
|
---|
131 | text_t gsdlCollect = collections[current_col];
|
---|
132 |
|
---|
133 | if (this->setsOutput == this->configuration->resumeAfter())
|
---|
134 | {
|
---|
135 | this->replyToken = new ResumptionToken("", "", "");
|
---|
136 | this->replyToken->setPosition("", this->setNumber+1);
|
---|
137 | break;
|
---|
138 | }
|
---|
139 |
|
---|
140 | if (this->setNumber >= startSet)
|
---|
141 | {
|
---|
142 | output << " <set>" << endl;
|
---|
143 | output << " <setSpec>" << gsdlCollect << "</setSpec>" << endl;;
|
---|
144 | output << " <setName>" << gsdlCollect << "</setName>" << endl;
|
---|
145 | output << " </set>" << endl;
|
---|
146 | this->setsOutput++;
|
---|
147 | }
|
---|
148 | setNumber++;
|
---|
149 |
|
---|
150 | // get all the children of the (relevant) classifier data structures
|
---|
151 | get_children(browseOID, gsdlCollect, "", metadata, false, protocol, response, logout);
|
---|
152 | // and send them to the "recurse_content" list
|
---|
153 | for (int c = 0; c < response.numDocs; ++c) {
|
---|
154 | this->recurse_content(output, protocol, gsdlCollect, response.docInfo[c].OID, gsdlCollect, startSet);
|
---|
155 | }
|
---|
156 | }
|
---|
157 |
|
---|
158 | // do a resumption token if required; errors cancel a token...
|
---|
159 | if (this->replyToken != NULL && this->errorType == "") {
|
---|
160 | // Don't add any whitespace around the resumption token as it can confuse harvesters/validators
|
---|
161 | output << " <resumptionToken>" << this->replyToken->getToken() << "</resumptionToken>" << endl;
|
---|
162 | }
|
---|
163 |
|
---|
164 | return true;
|
---|
165 | }
|
---|
166 |
|
---|
167 | void listsetsaction::recurse_content(ostream &output, recptproto *protocol, text_t &collection,
|
---|
168 | const text_t &classifier, text_t setHierarchy, int startSet)
|
---|
169 | {
|
---|
170 | // metadata for this call
|
---|
171 | FilterResponse_t response;
|
---|
172 | text_tset metadata;
|
---|
173 | ofstream logout("oai.log", ios::app);
|
---|
174 |
|
---|
175 | if (this->setsOutput == this->configuration->resumeAfter())
|
---|
176 | {
|
---|
177 | this->replyToken = new ResumptionToken("", "", "");
|
---|
178 | this->replyToken->setPosition("", this->setNumber+1);
|
---|
179 | return;
|
---|
180 | }
|
---|
181 |
|
---|
182 | metadata.insert("contains");
|
---|
183 | metadata.insert("Title");
|
---|
184 | metadata.insert("supportsmemberof");
|
---|
185 |
|
---|
186 | // get the document information
|
---|
187 | if (!get_info(classifier, collection, "", metadata, false, protocol, response, logout)) {
|
---|
188 | //cerr << "recurse content: Bad identifier or protocol " << classifier << endl;
|
---|
189 | return;
|
---|
190 | }
|
---|
191 |
|
---|
192 | // check for top-level classifiers, check if the set name includes a '.'; if
|
---|
193 | // not, it is a top-level classifier: check for memberof support. Those without
|
---|
194 | // memberof support will not be supported on OAI
|
---|
195 | if (findchar(classifier.begin(), classifier.end(), '.') == classifier.end()) {
|
---|
196 | if (response.docInfo[0].metadata["supportsmemberof"].values.size() > 0) {
|
---|
197 | text_t memberOf = response.docInfo[0].metadata["supportsmemberof"].values[0];
|
---|
198 | if (memberOf != "true") {
|
---|
199 | return;
|
---|
200 | }
|
---|
201 | }
|
---|
202 | else {
|
---|
203 | return;
|
---|
204 | }
|
---|
205 | }
|
---|
206 |
|
---|
207 | MetadataInfo_tmap::iterator here = response.docInfo[0].metadata.begin();
|
---|
208 | MetadataInfo_tmap::iterator end = response.docInfo[0].metadata.end();
|
---|
209 | text_t title;
|
---|
210 |
|
---|
211 | while (here != end)
|
---|
212 | {
|
---|
213 | // Each set should only have one title - hence we only output one title here
|
---|
214 | // (it is a set title, not a collection)
|
---|
215 | if (here->first == "Title" && here->second.values.size() > 0) {
|
---|
216 | title = here->second.values[0];
|
---|
217 | }
|
---|
218 |
|
---|
219 | ++here;
|
---|
220 | }
|
---|
221 |
|
---|
222 | // output the xml for this set; use the classifier id for the name
|
---|
223 | // if the title is blank
|
---|
224 | // curSet holds the colon-separated sequence of parent sets of the current set
|
---|
225 | text_t curSet;
|
---|
226 | if (this->setNumber >= startSet)
|
---|
227 | {
|
---|
228 | output << " <set>" << endl;
|
---|
229 | text_t oai_classifier = classifier;
|
---|
230 | oaiclassifier::toOAI(collection, oai_classifier);
|
---|
231 | output << " <setSpec>" << oai_classifier << "</setSpec>" << endl;
|
---|
232 | output << " <setName>";
|
---|
233 | if (!title.empty()) {
|
---|
234 | curSet = setHierarchy + ":" + title;
|
---|
235 | }
|
---|
236 | else {
|
---|
237 | curSet = classifier; // Pretty much never gets here (shouldn't, at least)
|
---|
238 | }
|
---|
239 | output << curSet;
|
---|
240 | output << "</setName>" << endl;
|
---|
241 | output << " </set>" << endl;
|
---|
242 | this->setsOutput++;
|
---|
243 | }
|
---|
244 | this->setNumber++;
|
---|
245 |
|
---|
246 | // get the children of this classifier and iterate them
|
---|
247 | get_children(classifier, collection, "", metadata, false, protocol, response, logout);
|
---|
248 | for (int c = 0; c < response.numDocs; ++c) {
|
---|
249 | text_t child = response.docInfo[c].OID;
|
---|
250 |
|
---|
251 | if (child == classifier)
|
---|
252 | continue;
|
---|
253 |
|
---|
254 | // check for non classifier items and exclude them
|
---|
255 | text_t childHead;
|
---|
256 | text_t::const_iterator start = child.begin();
|
---|
257 | text_t::const_iterator here = child.begin();
|
---|
258 | here += 2;
|
---|
259 | childHead = substr(start, here);
|
---|
260 |
|
---|
261 | if (childHead != "CL")
|
---|
262 | continue;
|
---|
263 |
|
---|
264 | // Recurse for "proper" classifier children. Pass curSet, the colon-separated list of
|
---|
265 | // parent sets. curSet is pass-by-value, so that as we step out of recursion we remember
|
---|
266 | // old set hierarchies.
|
---|
267 | this->recurse_content(output, protocol, collection, child, curSet, startSet);
|
---|
268 | }
|
---|
269 |
|
---|
270 | return;
|
---|
271 | }
|
---|