1 | /** |
---|
2 | * Licensed to the Apache Software Foundation (ASF) under one |
---|
3 | * or more contributor license agreements. See the NOTICE file |
---|
4 | * distributed with this work for additional information |
---|
5 | * regarding copyright ownership. The ASF licenses this file |
---|
6 | * to you under the Apache License, Version 2.0 (the |
---|
7 | * "License"); you may not use this file except in compliance |
---|
8 | * with the License. You may obtain a copy of the License at |
---|
9 | * |
---|
10 | * http://www.apache.org/licenses/LICENSE-2.0 |
---|
11 | * |
---|
12 | * Unless required by applicable law or agreed to in writing, software |
---|
13 | * distributed under the License is distributed on an "AS IS" BASIS, |
---|
14 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
---|
15 | * See the License for the specific language governing permissions and |
---|
16 | * limitations under the License. |
---|
17 | */ |
---|
18 | |
---|
19 | #include "xmlarchive.hh" |
---|
20 | #include <stdlib.h> |
---|
21 | |
---|
22 | using namespace hadoop; |
---|
23 | |
---|
24 | void hadoop::MySAXHandler::startElement(const XMLCh* const name, AttributeList& attr) |
---|
25 | { |
---|
26 | charsValid = false; |
---|
27 | char* qname = XMLString::transcode(name); |
---|
28 | if(std::string("boolean") == qname || |
---|
29 | std::string("ex:i1") == qname || |
---|
30 | std::string("i4") == qname || |
---|
31 | std::string("int") == qname || |
---|
32 | std::string("ex:i8") == qname || |
---|
33 | std::string("ex:float") == qname || |
---|
34 | std::string("double") == qname || |
---|
35 | std::string("string") == qname) { |
---|
36 | std::string s(qname); |
---|
37 | Value v(s); |
---|
38 | vlist.push_back(v); |
---|
39 | charsValid = true; |
---|
40 | } else if(std::string("struct") == qname || |
---|
41 | std::string("array") == qname) { |
---|
42 | std::string s(qname); |
---|
43 | Value v(s); |
---|
44 | vlist.push_back(v); |
---|
45 | } |
---|
46 | XMLString::release(&qname); |
---|
47 | } |
---|
48 | |
---|
49 | void hadoop::MySAXHandler::endElement(const XMLCh* const name) |
---|
50 | { |
---|
51 | charsValid = false; |
---|
52 | char* qname = XMLString::transcode(name); |
---|
53 | if(std::string("struct") == qname || |
---|
54 | std::string("array") == qname) { |
---|
55 | std::string s = "/"; |
---|
56 | Value v(s + qname); |
---|
57 | vlist.push_back(v); |
---|
58 | } |
---|
59 | XMLString::release(&qname); |
---|
60 | } |
---|
61 | |
---|
62 | void hadoop::MySAXHandler::characters(const XMLCh* const buf, const unsigned int len) |
---|
63 | { |
---|
64 | if (charsValid) { |
---|
65 | char *cstr = XMLString::transcode(buf); |
---|
66 | Value& v = vlist.back(); |
---|
67 | v.addChars(cstr, strlen(cstr)); |
---|
68 | XMLString::release(&cstr); |
---|
69 | } |
---|
70 | } |
---|
71 | |
---|
72 | static char hexchars[] = {'0', '1', '2', '3', '4', '5', '6', '7', '8', '9', |
---|
73 | 'A', 'B', 'C', 'D', 'E', 'F' }; |
---|
74 | |
---|
75 | static std::string toXMLString(std::string s) |
---|
76 | { |
---|
77 | std::string r; |
---|
78 | size_t len = s.length(); |
---|
79 | size_t i; |
---|
80 | const char* data = s.data(); |
---|
81 | for (i=0; i<len; i++, data++) { |
---|
82 | char ch = *data; |
---|
83 | if (ch == '<') { |
---|
84 | r.append("<"); |
---|
85 | } else if (ch == '&') { |
---|
86 | r.append("&"); |
---|
87 | } else if (ch == '%') { |
---|
88 | r.append("%0025"); |
---|
89 | } else if (ch < 0x20) { |
---|
90 | uint8_t* pb = (uint8_t*) &ch; |
---|
91 | char ch1 = hexchars[*pb/16]; |
---|
92 | char ch2 = hexchars[*pb%16]; |
---|
93 | r.push_back('%'); |
---|
94 | r.push_back('0'); |
---|
95 | r.push_back('0'); |
---|
96 | r.push_back(ch1); |
---|
97 | r.push_back(ch2); |
---|
98 | } else { |
---|
99 | r.push_back(ch); |
---|
100 | } |
---|
101 | } |
---|
102 | return r; |
---|
103 | } |
---|
104 | |
---|
105 | static uint8_t h2b(char ch) { |
---|
106 | if ((ch >= '0') || (ch <= '9')) { |
---|
107 | return ch - '0'; |
---|
108 | } |
---|
109 | if ((ch >= 'a') || (ch <= 'f')) { |
---|
110 | return ch - 'a' + 10; |
---|
111 | } |
---|
112 | if ((ch >= 'A') || (ch <= 'F')) { |
---|
113 | return ch - 'A' + 10; |
---|
114 | } |
---|
115 | return 0; |
---|
116 | } |
---|
117 | |
---|
118 | static std::string fromXMLString(std::string s) |
---|
119 | { |
---|
120 | std::string r; |
---|
121 | size_t len = s.length(); |
---|
122 | size_t i; |
---|
123 | uint8_t* pb = (uint8_t*) s.data(); |
---|
124 | for (i = 0; i < len; i++) { |
---|
125 | uint8_t b = *pb; |
---|
126 | if (b == '%') { |
---|
127 | char *pc = (char*) (pb+1); |
---|
128 | // ignore the first two characters, which are always '0' |
---|
129 | *pc++; |
---|
130 | *pc++;; |
---|
131 | char ch1 = *pc++; |
---|
132 | char ch2 = *pc++; |
---|
133 | pb += 4; |
---|
134 | uint8_t cnv = h2b(ch1)*16 + h2b(ch2); |
---|
135 | pc = (char*) &cnv; |
---|
136 | r.push_back(*pc); |
---|
137 | } else { |
---|
138 | char *pc = (char*) pb; |
---|
139 | r.push_back(*pc); |
---|
140 | } |
---|
141 | pb++; |
---|
142 | } |
---|
143 | return r; |
---|
144 | } |
---|
145 | |
---|
146 | static std::string toXMLBuffer(std::string s, size_t len) |
---|
147 | { |
---|
148 | std::string r; |
---|
149 | size_t i; |
---|
150 | uint8_t* data = (uint8_t*) s.data(); |
---|
151 | for (i=0; i<len; i++, data++) { |
---|
152 | uint8_t b = *data; |
---|
153 | char ch1 = hexchars[b/16]; |
---|
154 | char ch2 = hexchars[b%16]; |
---|
155 | r.push_back(ch1); |
---|
156 | r.push_back(ch2); |
---|
157 | } |
---|
158 | return r; |
---|
159 | } |
---|
160 | |
---|
161 | static std::string fromXMLBuffer(std::string s, size_t& len) |
---|
162 | { |
---|
163 | len = s.length(); |
---|
164 | if (len%2 == 1) { // len is guaranteed to be even |
---|
165 | throw new IOException("Errror deserializing buffer."); |
---|
166 | } |
---|
167 | len = len >> 1; |
---|
168 | std::string t; |
---|
169 | for (size_t idx = 0; idx < len; idx++) { |
---|
170 | char buf[3]; |
---|
171 | buf[0] = s[2*idx]; |
---|
172 | buf[1] = s[2*idx+1]; |
---|
173 | buf[2] = '\0'; |
---|
174 | int i; |
---|
175 | if (1 != sscanf(buf, "%2x", &i)) { |
---|
176 | throw new IOException("Errror deserializing buffer."); |
---|
177 | } |
---|
178 | t.push_back((char) i); |
---|
179 | } |
---|
180 | len = t.length(); |
---|
181 | return t; |
---|
182 | } |
---|
183 | |
---|
184 | void hadoop::IXmlArchive::deserialize(int8_t& t, const char* tag) |
---|
185 | { |
---|
186 | Value v = next(); |
---|
187 | if (v.getType() != "ex:i1") { |
---|
188 | throw new IOException("Error deserializing byte"); |
---|
189 | } |
---|
190 | t = (int8_t) strtol(v.getValue().c_str(), NULL, 10); |
---|
191 | } |
---|
192 | |
---|
193 | void hadoop::IXmlArchive::deserialize(bool& t, const char* tag) |
---|
194 | { |
---|
195 | Value v = next(); |
---|
196 | if (v.getType() != "boolean") { |
---|
197 | throw new IOException("Error deserializing boolean"); |
---|
198 | } |
---|
199 | t = (v.getValue() == "1"); |
---|
200 | } |
---|
201 | |
---|
202 | void hadoop::IXmlArchive::deserialize(int32_t& t, const char* tag) |
---|
203 | { |
---|
204 | Value v = next(); |
---|
205 | if (v.getType() != "i4" && v.getType() != "int") { |
---|
206 | throw new IOException("Error deserializing int"); |
---|
207 | } |
---|
208 | t = (int32_t) strtol(v.getValue().c_str(), NULL, 10); |
---|
209 | } |
---|
210 | |
---|
211 | void hadoop::IXmlArchive::deserialize(int64_t& t, const char* tag) |
---|
212 | { |
---|
213 | Value v = next(); |
---|
214 | if (v.getType() != "ex:i8") { |
---|
215 | throw new IOException("Error deserializing long"); |
---|
216 | } |
---|
217 | t = strtoll(v.getValue().c_str(), NULL, 10); |
---|
218 | } |
---|
219 | |
---|
220 | void hadoop::IXmlArchive::deserialize(float& t, const char* tag) |
---|
221 | { |
---|
222 | Value v = next(); |
---|
223 | if (v.getType() != "ex:float") { |
---|
224 | throw new IOException("Error deserializing float"); |
---|
225 | } |
---|
226 | t = strtof(v.getValue().c_str(), NULL); |
---|
227 | } |
---|
228 | |
---|
229 | void hadoop::IXmlArchive::deserialize(double& t, const char* tag) |
---|
230 | { |
---|
231 | Value v = next(); |
---|
232 | if (v.getType() != "double") { |
---|
233 | throw new IOException("Error deserializing double"); |
---|
234 | } |
---|
235 | t = strtod(v.getValue().c_str(), NULL); |
---|
236 | } |
---|
237 | |
---|
238 | void hadoop::IXmlArchive::deserialize(std::string& t, const char* tag) |
---|
239 | { |
---|
240 | Value v = next(); |
---|
241 | if (v.getType() != "string") { |
---|
242 | throw new IOException("Error deserializing string"); |
---|
243 | } |
---|
244 | t = fromXMLString(v.getValue()); |
---|
245 | } |
---|
246 | |
---|
247 | void hadoop::IXmlArchive::deserialize(std::string& t, size_t& len, const char* tag) |
---|
248 | { |
---|
249 | Value v = next(); |
---|
250 | if (v.getType() != "string") { |
---|
251 | throw new IOException("Error deserializing buffer"); |
---|
252 | } |
---|
253 | t = fromXMLBuffer(v.getValue(), len); |
---|
254 | } |
---|
255 | |
---|
256 | void hadoop::IXmlArchive::startRecord(Record& s, const char* tag) |
---|
257 | { |
---|
258 | Value v = next(); |
---|
259 | if (v.getType() != "struct") { |
---|
260 | throw new IOException("Error deserializing record"); |
---|
261 | } |
---|
262 | } |
---|
263 | |
---|
264 | void hadoop::IXmlArchive::endRecord(Record& s, const char* tag) |
---|
265 | { |
---|
266 | Value v = next(); |
---|
267 | if (v.getType() != "/struct") { |
---|
268 | throw new IOException("Error deserializing record"); |
---|
269 | } |
---|
270 | } |
---|
271 | |
---|
272 | Index* hadoop::IXmlArchive::startVector(const char* tag) |
---|
273 | { |
---|
274 | Value v = next(); |
---|
275 | if (v.getType() != "array") { |
---|
276 | throw new IOException("Error deserializing vector"); |
---|
277 | } |
---|
278 | return new XmlIndex(vlist, vidx); |
---|
279 | } |
---|
280 | |
---|
281 | void hadoop::IXmlArchive::endVector(Index* idx, const char* tag) |
---|
282 | { |
---|
283 | Value v = next(); |
---|
284 | if (v.getType() != "/array") { |
---|
285 | throw new IOException("Error deserializing vector"); |
---|
286 | } |
---|
287 | delete idx; |
---|
288 | } |
---|
289 | |
---|
290 | Index* hadoop::IXmlArchive::startMap(const char* tag) |
---|
291 | { |
---|
292 | Value v = next(); |
---|
293 | if (v.getType() != "array") { |
---|
294 | throw new IOException("Error deserializing map"); |
---|
295 | } |
---|
296 | return new XmlIndex(vlist, vidx); |
---|
297 | } |
---|
298 | |
---|
299 | void hadoop::IXmlArchive::endMap(Index* idx, const char* tag) |
---|
300 | { |
---|
301 | Value v = next(); |
---|
302 | if (v.getType() != "/array") { |
---|
303 | throw new IOException("Error deserializing map"); |
---|
304 | } |
---|
305 | delete idx; |
---|
306 | } |
---|
307 | |
---|
308 | void hadoop::OXmlArchive::serialize(int8_t t, const char* tag) |
---|
309 | { |
---|
310 | printBeginEnvelope(tag); |
---|
311 | p("<ex:i1>"); |
---|
312 | char sval[5]; |
---|
313 | sprintf(sval, "%d", t); |
---|
314 | p(sval); |
---|
315 | p("</ex:i1>"); |
---|
316 | printEndEnvelope(tag); |
---|
317 | } |
---|
318 | |
---|
319 | void hadoop::OXmlArchive::serialize(bool t, const char* tag) |
---|
320 | { |
---|
321 | printBeginEnvelope(tag); |
---|
322 | p("<boolean>"); |
---|
323 | p(t ? "1" : "0"); |
---|
324 | p("</boolean>"); |
---|
325 | printEndEnvelope(tag); |
---|
326 | } |
---|
327 | |
---|
328 | void hadoop::OXmlArchive::serialize(int32_t t, const char* tag) |
---|
329 | { |
---|
330 | printBeginEnvelope(tag); |
---|
331 | p("<i4>"); |
---|
332 | char sval[128]; |
---|
333 | sprintf(sval, "%d", t); |
---|
334 | p(sval); |
---|
335 | p("</i4>"); |
---|
336 | printEndEnvelope(tag); |
---|
337 | } |
---|
338 | |
---|
339 | void hadoop::OXmlArchive::serialize(int64_t t, const char* tag) |
---|
340 | { |
---|
341 | printBeginEnvelope(tag); |
---|
342 | p("<ex:i8>"); |
---|
343 | char sval[128]; |
---|
344 | sprintf(sval, "%lld", t); |
---|
345 | p(sval); |
---|
346 | p("</ex:i8>"); |
---|
347 | printEndEnvelope(tag); |
---|
348 | |
---|
349 | } |
---|
350 | |
---|
351 | void hadoop::OXmlArchive::serialize(float t, const char* tag) |
---|
352 | { |
---|
353 | printBeginEnvelope(tag); |
---|
354 | p("<ex:float>"); |
---|
355 | char sval[128]; |
---|
356 | sprintf(sval, "%f", t); |
---|
357 | p(sval); |
---|
358 | p("</ex:float>"); |
---|
359 | printEndEnvelope(tag); |
---|
360 | } |
---|
361 | |
---|
362 | void hadoop::OXmlArchive::serialize(double t, const char* tag) |
---|
363 | { |
---|
364 | printBeginEnvelope(tag); |
---|
365 | p("<double>"); |
---|
366 | char sval[128]; |
---|
367 | sprintf(sval, "%lf", t); |
---|
368 | p(sval); |
---|
369 | p("</double>"); |
---|
370 | printEndEnvelope(tag); |
---|
371 | } |
---|
372 | |
---|
373 | void hadoop::OXmlArchive::serialize(const std::string& t, const char* tag) |
---|
374 | { |
---|
375 | printBeginEnvelope(tag); |
---|
376 | p("<string>"); |
---|
377 | std::string s = toXMLString(t); |
---|
378 | stream.write(s.data(), s.length()); |
---|
379 | p("</string>"); |
---|
380 | printEndEnvelope(tag); |
---|
381 | } |
---|
382 | |
---|
383 | void hadoop::OXmlArchive::serialize(const std::string& t, size_t len, const char* tag) |
---|
384 | { |
---|
385 | printBeginEnvelope(tag); |
---|
386 | p("<string>"); |
---|
387 | std::string s = toXMLBuffer(t, len); |
---|
388 | stream.write(s.data(), s.length()); |
---|
389 | p("</string>"); |
---|
390 | printEndEnvelope(tag); |
---|
391 | } |
---|
392 | |
---|
393 | void hadoop::OXmlArchive::startRecord(const Record& s, const char* tag) |
---|
394 | { |
---|
395 | insideRecord(tag); |
---|
396 | p("<struct>\n"); |
---|
397 | } |
---|
398 | |
---|
399 | void hadoop::OXmlArchive::endRecord(const Record& s, const char* tag) |
---|
400 | { |
---|
401 | p("</struct>\n"); |
---|
402 | outsideRecord(tag); |
---|
403 | } |
---|
404 | |
---|
405 | void hadoop::OXmlArchive::startVector(size_t len, const char* tag) |
---|
406 | { |
---|
407 | insideVector(tag); |
---|
408 | p("<array>\n"); |
---|
409 | } |
---|
410 | |
---|
411 | void hadoop::OXmlArchive::endVector(size_t len, const char* tag) |
---|
412 | { |
---|
413 | p("</array>\n"); |
---|
414 | outsideVector(tag); |
---|
415 | } |
---|
416 | |
---|
417 | void hadoop::OXmlArchive::startMap(size_t len, const char* tag) |
---|
418 | { |
---|
419 | insideMap(tag); |
---|
420 | p("<array>\n"); |
---|
421 | } |
---|
422 | |
---|
423 | void hadoop::OXmlArchive::endMap(size_t len, const char* tag) |
---|
424 | { |
---|
425 | p("</array>\n"); |
---|
426 | outsideMap(tag); |
---|
427 | } |
---|
428 | |
---|
429 | hadoop::OXmlArchive::~OXmlArchive() |
---|
430 | { |
---|
431 | } |
---|