| 1 |
/****************************************************************************** |
| 2 |
* |
| 3 |
* Copyright (C) 1997-2022 by Dimitri van Heesch. |
| 4 |
* |
| 5 |
* Permission to use, copy, modify, and distribute this software and its |
| 6 |
* documentation under the terms of the GNU General Public License is hereby |
| 7 |
* granted. No representations are made about the suitability of this software |
| 8 |
* for any purpose. It is provided "as is" without express or implied warranty. |
| 9 |
* See the GNU General Public License for more details. |
| 10 |
* |
| 11 |
* Documents produced by Doxygen are derivative works derived from the |
| 12 |
* input used in their production; they are not affected by this license. |
| 13 |
* |
| 14 |
*/ |
| 15 |
|
| 16 |
// STL includes |
| 17 |
#include <cstdio> |
| 18 |
#include <cstdlib> |
| 19 |
#include <string> |
| 20 |
#include <vector> |
| 21 |
#include <sstream> |
| 22 |
#include <iostream> |
| 23 |
#include <string> |
| 24 |
#include <algorithm> |
| 25 |
#include <filesystem> |
| 26 |
|
| 27 |
// Xapian includes |
| 28 |
#include <xapian.h> |
| 29 |
|
| 30 |
#include "version.h" |
| 31 |
|
| 32 |
|
| 33 |
constexpr Xapian::valueno FIELD_TYPE{ 1 }; |
| 34 |
constexpr Xapian::valueno FIELD_NAME{ 2 }; |
| 35 |
constexpr Xapian::valueno FIELD_ARGS{ 3 }; |
| 36 |
constexpr Xapian::valueno FIELD_TAG{ 4 }; |
| 37 |
constexpr Xapian::valueno FIELD_URL{ 5 }; |
| 38 |
constexpr Xapian::valueno FIELD_KEYW{ 6 }; |
| 39 |
constexpr Xapian::valueno FIELD_DOC{ 7 }; |
| 40 |
|
| 41 |
constexpr int hex2dec(unsigned char c) |
| 42 |
{ |
| 43 |
if (c >= '0' && c <= '9') return c - '0'; |
| 44 |
if (c >= 'a' && c <= 'f') return c - 'a' + 10; |
| 45 |
if (c >= 'A' && c <= 'F') return c - 'A' + 10; |
| 46 |
return -1; |
| 47 |
} |
| 48 |
|
| 49 |
|
| 50 |
bool dirExists(const std::string &dirname) |
| 51 |
{ |
| 52 |
std::error_code ec; |
| 53 |
return std::filesystem::is_directory(dirname, ec); |
| 54 |
} |
| 55 |
|
| 56 |
|
| 57 |
/** decodes a URI encoded string into a normal string. */ |
| 58 |
static std::string uriDecode(const std::string & sSrc) |
| 59 |
{ |
| 60 |
// Note from RFC1630: "Sequences which start with a percent |
| 61 |
// sign but are not followed by two hexadecimal characters |
| 62 |
// (0-9, A-F) are reserved for future extension" |
| 63 |
|
| 64 |
const unsigned char * pSrc = (const unsigned char *)sSrc.c_str(); |
| 65 |
const size_t SRC_LEN = sSrc.length(); |
| 66 |
const unsigned char * const SRC_END = pSrc + SRC_LEN; |
| 67 |
// last decodable '%' |
| 68 |
const unsigned char * const SRC_LAST_DEC = SRC_END - 2; |
| 69 |
|
| 70 |
char * const pStart = new char[SRC_LEN]; |
| 71 |
char * pEnd = pStart; |
| 72 |
|
| 73 |
while (pSrc < SRC_LAST_DEC) |
| 74 |
{ |
| 75 |
if (*pSrc == '%') // replace %2A with corresponding ASCII character |
| 76 |
{ |
| 77 |
int dec1, dec2; |
| 78 |
unsigned char c1=*(pSrc+1); |
| 79 |
unsigned char c2=*(pSrc+2); |
| 80 |
if (-1 != (dec1 = hex2dec(c1)) && -1 != (dec2 = hex2dec(c2))) |
| 81 |
{ |
| 82 |
*pEnd++ = (dec1 << 4) + dec2; |
| 83 |
pSrc += 3; |
| 84 |
continue; |
| 85 |
} |
| 86 |
} |
| 87 |
else if (*pSrc == '+') // replace '+' with space |
| 88 |
{ |
| 89 |
*pEnd++ = ' '; pSrc++; |
| 90 |
continue; |
| 91 |
} |
| 92 |
*pEnd++ = *pSrc++; |
| 93 |
} |
| 94 |
|
| 95 |
// the last 2- chars |
| 96 |
while (pSrc < SRC_END) *pEnd++ = *pSrc++; |
| 97 |
|
| 98 |
std::string sResult(pStart, pEnd); |
| 99 |
delete [] pStart; |
| 100 |
return sResult; |
| 101 |
} |
| 102 |
|
| 103 |
/** return list of strings that result when splitting \a s using |
| 104 |
* delimiter \a delim |
| 105 |
*/ |
| 106 |
static std::vector<std::string> split(const std::string &s, char delim) |
| 107 |
{ |
| 108 |
std::vector<std::string> elems; |
| 109 |
std::stringstream ss(s); |
| 110 |
std::string item; |
| 111 |
while (getline(ss, item, delim)) elems.push_back(item); |
| 112 |
return elems; |
| 113 |
} |
| 114 |
|
| 115 |
/** Read type T from string \a s */ |
| 116 |
template<class T> |
| 117 |
T fromString(const std::string& s) |
| 118 |
{ |
| 119 |
std::istringstream stream (s); |
| 120 |
T t; |
| 121 |
stream >> t; |
| 122 |
return t; |
| 123 |
} |
| 124 |
|
| 125 |
/** Class that holds the starting position of a word */ |
| 126 |
struct WordPosition |
| 127 |
{ |
| 128 |
WordPosition(size_t s,size_t i) : start(s), index(i) {} |
| 129 |
size_t start; |
| 130 |
size_t index; |
| 131 |
}; |
| 132 |
|
| 133 |
/** Class representing the '<' operator for WordPosition objects based on position. */ |
| 134 |
struct WordPosition_less |
| 135 |
{ |
| 136 |
bool operator()(const WordPosition &p1,const WordPosition &p2) const |
| 137 |
{ |
| 138 |
return p1.start<p2.start; |
| 139 |
} |
| 140 |
}; |
| 141 |
|
| 142 |
/** Class that holds a text fragment */ |
| 143 |
struct Fragment |
| 144 |
{ |
| 145 |
Fragment(const std::string &t,int occ) : text(t), occurrences(occ) {} |
| 146 |
std::string text; |
| 147 |
int occurrences; |
| 148 |
}; |
| 149 |
|
| 150 |
/** Class representing the '>' operator for Fragment objects based on occurrence. */ |
| 151 |
struct Fragment_greater |
| 152 |
{ |
| 153 |
bool operator()(const Fragment &p1,const Fragment &p2) const |
| 154 |
{ |
| 155 |
return p1.occurrences>p2.occurrences; |
| 156 |
} |
| 157 |
}; |
| 158 |
|
| 159 |
/** Class representing a range within a string */ |
| 160 |
struct Range |
| 161 |
{ |
| 162 |
Range(size_t s,size_t e) : start(s), end(e) {} |
| 163 |
size_t start; |
| 164 |
size_t end; |
| 165 |
}; |
| 166 |
|
| 167 |
/** Returns true if [start..start+len] is inside one of the \a ranges. */ |
| 168 |
static bool insideRange(const std::vector<Range> &ranges,size_t start,size_t len) |
| 169 |
{ |
| 170 |
for (std::vector<Range>::const_iterator it = ranges.begin(); |
| 171 |
it!=ranges.end(); ++it |
| 172 |
) |
| 173 |
{ |
| 174 |
Range r = *it; |
| 175 |
if (start>=r.start && start+len<r.end) |
| 176 |
{ |
| 177 |
return true; |
| 178 |
} |
| 179 |
} |
| 180 |
return false; |
| 181 |
} |
| 182 |
|
| 183 |
/** Returns a list of text \a fragments from \a s containing one or |
| 184 |
* more \a words. The list is sorted according to the |
| 185 |
* number of occurrences of words within the fragment. |
| 186 |
*/ |
| 187 |
static void highlighter(const std::string &s, |
| 188 |
const std::vector<std::string> &words, |
| 189 |
std::vector<Fragment> &fragments) |
| 190 |
{ |
| 191 |
const std::string spanStart="<span class=\"hl\">"; |
| 192 |
const std::string spanEnd="</span>"; |
| 193 |
const std::string dots="..."; |
| 194 |
const size_t fragLen = 60; |
| 195 |
size_t sl=s.length(); |
| 196 |
|
| 197 |
// find positions of words in s |
| 198 |
int j=0; |
| 199 |
std::vector<WordPosition> positions; |
| 200 |
for (std::vector<std::string>::const_iterator it=words.begin(); |
| 201 |
it!=words.end(); |
| 202 |
++it,++j |
| 203 |
) |
| 204 |
{ |
| 205 |
size_t pos=0; |
| 206 |
size_t i; |
| 207 |
std::string word = *it; |
| 208 |
while ((i=s.find(word,pos))!=std::string::npos) |
| 209 |
{ |
| 210 |
positions.push_back(WordPosition(i,j)); |
| 211 |
pos=i+word.length(); |
| 212 |
} |
| 213 |
} |
| 214 |
// sort on position |
| 215 |
std::sort(positions.begin(),positions.end(),WordPosition_less()); |
| 216 |
// get fragments around words |
| 217 |
std::vector<Range> ranges; |
| 218 |
for (std::vector<WordPosition>::const_iterator it=positions.begin(); |
| 219 |
it!=positions.end(); |
| 220 |
++it) |
| 221 |
{ |
| 222 |
WordPosition wp = *it; |
| 223 |
std::string w = words[wp.index]; |
| 224 |
size_t i=wp.start; |
| 225 |
size_t wl=w.length(); |
| 226 |
if (!insideRange(ranges,i,wl)) |
| 227 |
{ |
| 228 |
if (wl>fragLen) |
| 229 |
{ |
| 230 |
fragments.push_back(Fragment(spanStart+w+spanEnd,1)); |
| 231 |
ranges.push_back(Range(i,i+wl)); |
| 232 |
} |
| 233 |
else |
| 234 |
{ |
| 235 |
std::string startFragment,endFragment; |
| 236 |
int bi=static_cast<int>(i)-static_cast<int>((fragLen-wl)/2); |
| 237 |
size_t ei=i+wl+(fragLen-wl)/2; |
| 238 |
int occ=0; |
| 239 |
if (bi<0) { ei-=bi; bi=0; } else startFragment=dots; |
| 240 |
if (ei>sl) { ei=sl; } else endFragment=dots; |
| 241 |
while (bi>0 && !isspace(s[bi])) bi--; // round to start of the word |
| 242 |
while (ei<sl && !isspace(s[ei])) ei++; // round to end of the word |
| 243 |
// highlight any word in s between indexes bi and ei |
| 244 |
std::string fragment=startFragment; |
| 245 |
size_t pos=bi; |
| 246 |
for (std::vector<WordPosition>::const_iterator it2=positions.begin(); |
| 247 |
it2!=positions.end(); |
| 248 |
++it2) |
| 249 |
{ |
| 250 |
WordPosition wp2 = *it2; |
| 251 |
std::string w2 = words[wp2.index]; |
| 252 |
size_t wl2 = w2.length(); |
| 253 |
if (wp2.start>=bi && wp2.start+wl2<=ei) // word is inside the range! |
| 254 |
{ |
| 255 |
fragment+=s.substr(pos,wp2.start-pos)+ |
| 256 |
spanStart+ |
| 257 |
s.substr(wp2.start,wl2)+ |
| 258 |
spanEnd; |
| 259 |
pos=wp2.start+wl2; |
| 260 |
occ++; |
| 261 |
} |
| 262 |
} |
| 263 |
fragment+=s.substr(pos,ei-pos)+endFragment; |
| 264 |
fragments.push_back(Fragment(fragment,occ)); |
| 265 |
ranges.push_back(Range(bi,ei)); |
| 266 |
} |
| 267 |
} |
| 268 |
} |
| 269 |
std::sort(fragments.begin(),fragments.end(),Fragment_greater()); |
| 270 |
} |
| 271 |
|
| 272 |
/** Escapes a string such that is can be included in a JSON structure */ |
| 273 |
static std::string escapeString(const std::string &s) |
| 274 |
{ |
| 275 |
std::stringstream dst; |
| 276 |
for (size_t i=0;i<s.length();i++) |
| 277 |
{ |
| 278 |
char ch = s[i]; |
| 279 |
switch (ch) |
| 280 |
{ |
| 281 |
case '\"': dst << "\\\""; break; |
| 282 |
default: dst << ch; break; |
| 283 |
} |
| 284 |
} |
| 285 |
return dst.str(); |
| 286 |
} |
| 287 |
|
| 288 |
static void showError(const std::string &callback,const std::string &error) |
| 289 |
{ |
| 290 |
std::cout << callback << "({\"error\":\"" << error << "\"})"; |
| 291 |
exit(0); |
| 292 |
} |
| 293 |
|
| 294 |
static void usage(const char *name, int exitVal = 1) |
| 295 |
{ |
| 296 |
std::cerr << "Usage: " << name << "[query_string]" << std::endl; |
| 297 |
std::cerr << " " << "alternatively the query string can be given by the environment variable QUERY_STRING" << std::endl; |
| 298 |
exit(exitVal); |
| 299 |
} |
| 300 |
|
| 301 |
/** Main routine */ |
| 302 |
int main(int argc,char **argv) |
| 303 |
{ |
| 304 |
// process inputs that were passed to us via QUERY_STRING |
| 305 |
std::string callback; |
| 306 |
try |
| 307 |
{ |
| 308 |
std::string queryString; |
| 309 |
if (argc == 1) |
| 310 |
{ |
| 311 |
const char *queryEnv = getenv("QUERY_STRING"); |
| 312 |
if (queryEnv) |
| 313 |
{ |
| 314 |
queryString = queryEnv; |
| 315 |
} |
| 316 |
else |
| 317 |
{ |
| 318 |
usage(argv[0]); |
| 319 |
} |
| 320 |
} |
| 321 |
else if (argc == 2) |
| 322 |
{ |
| 323 |
if (std::string(argv[1])=="-h" || std::string(argv[1])=="--help") |
| 324 |
{ |
| 325 |
usage(argv[0],0); |
| 326 |
} |
| 327 |
else if (std::string(argv[1])=="-v" || std::string(argv[1])=="--version") |
| 328 |
{ |
| 329 |
std::cerr << argv[0] << " version: " << getFullVersion() << std::endl; |
| 330 |
exit(0); |
| 331 |
} |
| 332 |
else |
| 333 |
{ |
| 334 |
queryString = argv[1]; |
| 335 |
} |
| 336 |
} |
| 337 |
else |
| 338 |
{ |
| 339 |
usage(argv[0]); |
| 340 |
} |
| 341 |
|
| 342 |
std::cout << "Content-Type:application/javascript;charset=utf-8\r\n\n"; |
| 343 |
// parse query string |
| 344 |
std::vector<std::string> parts = split(queryString,'&'); |
| 345 |
std::string searchFor; |
| 346 |
int num=1,page=0; |
| 347 |
for (std::vector<std::string>::const_iterator it=parts.begin();it!=parts.end();++it) |
| 348 |
{ |
| 349 |
std::vector<std::string> kv = split(*it,'='); |
| 350 |
if (kv.size()==2) |
| 351 |
{ |
| 352 |
std::string val = uriDecode(kv[1]); |
| 353 |
if (kv[0]=="q") searchFor = val; |
| 354 |
else if (kv[0]=="n") num = fromString<int>(val); |
| 355 |
else if (kv[0]=="p") page = fromString<int>(val); |
| 356 |
else if (kv[0]=="cb") callback = val; |
| 357 |
} |
| 358 |
} |
| 359 |
|
| 360 |
std::string indexDir = "doxysearch.db"; |
| 361 |
|
| 362 |
if (queryString=="test") // user test |
| 363 |
{ |
| 364 |
bool dbOk = dirExists(indexDir); |
| 365 |
if (dbOk) |
| 366 |
{ |
| 367 |
std::cout << "Test successful."; |
| 368 |
} |
| 369 |
else |
| 370 |
{ |
| 371 |
std::cout << "Test failed: cannot find search index " << indexDir; |
| 372 |
} |
| 373 |
exit(0); |
| 374 |
} |
| 375 |
|
| 376 |
// create query |
| 377 |
Xapian::Database db(indexDir); |
| 378 |
Xapian::Enquire enquire(db); |
| 379 |
|
| 380 |
std::vector<std::string> words = split(searchFor,' '); |
| 381 |
Xapian::QueryParser parser; |
| 382 |
parser.set_database(db); |
| 383 |
parser.set_default_op(Xapian::Query::OP_AND); |
| 384 |
parser.set_stemming_strategy(Xapian::QueryParser::STEM_ALL); |
| 385 |
Xapian::termcount max_expansion=100; |
| 386 |
#if (XAPIAN_MAJOR_VERSION==1) && (XAPIAN_MINOR_VERSION==2) |
| 387 |
parser.set_max_wildcard_expansion(max_expansion); |
| 388 |
#else |
| 389 |
parser.set_max_expansion(max_expansion,Xapian::Query::WILDCARD_LIMIT_MOST_FREQUENT); |
| 390 |
#endif |
| 391 |
Xapian::Query query=parser.parse_query(searchFor, |
| 392 |
Xapian::QueryParser::FLAG_DEFAULT | |
| 393 |
Xapian::QueryParser::FLAG_WILDCARD | |
| 394 |
Xapian::QueryParser::FLAG_PHRASE | |
| 395 |
Xapian::QueryParser::FLAG_PARTIAL |
| 396 |
); |
| 397 |
enquire.set_query(query); |
| 398 |
|
| 399 |
// get results |
| 400 |
Xapian::MSet matches = enquire.get_mset(page*num,num); |
| 401 |
unsigned int hits = matches.get_matches_estimated(); |
| 402 |
unsigned int offset = page*num; |
| 403 |
unsigned int pages = num>0 ? (hits+num-1)/num : 0; |
| 404 |
if (offset>hits) offset=hits; |
| 405 |
if (offset+num>hits) num=hits-offset; |
| 406 |
|
| 407 |
// write results as JSONP |
| 408 |
std::cout << callback.c_str() << "("; |
| 409 |
std::cout << "{" << std::endl |
| 410 |
<< " \"hits\":" << hits << "," << std::endl |
| 411 |
<< " \"first\":" << offset << "," << std::endl |
| 412 |
<< " \"count\":" << num << "," << std::endl |
| 413 |
<< " \"page\":" << page << "," << std::endl |
| 414 |
<< " \"pages\":" << pages << "," << std::endl |
| 415 |
<< " \"query\": \"" << escapeString(searchFor) << "\"," << std::endl |
| 416 |
<< " \"items\":[" << std::endl; |
| 417 |
// foreach search result |
| 418 |
unsigned int o = offset; |
| 419 |
for (Xapian::MSetIterator i = matches.begin(); i != matches.end(); ++i,++o) |
| 420 |
{ |
| 421 |
std::vector<Fragment> hl; |
| 422 |
Xapian::Document doc = i.get_document(); |
| 423 |
highlighter(doc.get_value(FIELD_DOC),words,hl); |
| 424 |
std::cout << " {\"type\": \"" << doc.get_value(FIELD_TYPE) << "\"," << std::endl |
| 425 |
<< " \"name\": \"" << doc.get_value(FIELD_NAME) << escapeString(doc.get_value(FIELD_ARGS)) << "\"," << std::endl |
| 426 |
<< " \"tag\": \"" << doc.get_value(FIELD_TAG) << "\"," << std::endl |
| 427 |
<< " \"url\": \"" << doc.get_value(FIELD_URL) << "\"," << std::endl; |
| 428 |
std::cout << " \"fragments\":[" << std::endl; |
| 429 |
int c=0; |
| 430 |
bool first=true; |
| 431 |
for (std::vector<Fragment>::const_iterator it = hl.begin();it!=hl.end() && c<3;++it,++c) |
| 432 |
{ |
| 433 |
if (!first) std::cout << "," << std::endl; |
| 434 |
std::cout << " \"" << escapeString((*it).text) << "\""; |
| 435 |
first=false; |
| 436 |
} |
| 437 |
if (!first) std::cout << std::endl; |
| 438 |
std::cout << " ]" << std::endl; |
| 439 |
std::cout << " }"; |
| 440 |
if (o<offset+num-1) std::cout << ","; |
| 441 |
std::cout << std::endl; |
| 442 |
} |
| 443 |
std::cout << " ]" << std::endl << "})" << std::endl; |
| 444 |
} |
| 445 |
catch (const Xapian::Error &e) // Xapian exception |
| 446 |
{ |
| 447 |
showError(callback,e.get_description()); |
| 448 |
} |
| 449 |
catch (...) // Any other exception |
| 450 |
{ |
| 451 |
showError(callback,"Unknown Exception!"); |
| 452 |
exit(1); |
| 453 |
} |
| 454 |
return 0; |
| 455 |
} |