/** * @file llsdserialize.cpp * @author Phoenix * @date 2006-03-05 * @brief Implementation of LLSD parsers and formatters * * $LicenseInfo:firstyear=2006&license=viewerlgpl$ * Second Life Viewer Source Code * Copyright (C) 2010, Linden Research, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; * version 2.1 of the License only. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this library; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * * Linden Research, Inc., 945 Battery Street, San Francisco, CA 94111 USA * $/LicenseInfo$ */ #include "linden_common.h" #include "llsdserialize.h" #include "llpointer.h" #include "llstreamtools.h" // for fullread #include #include "apr_base64.h" #ifdef LL_USESYSTEMLIBS # include #else # include "zlib-ng/zlib.h" // for davep's dirty little zip functions #endif #if !LL_WINDOWS #include // htonl & ntohl #endif #include "lldate.h" #include "llmemorystream.h" #include "llsd.h" #include "llstring.h" #include "lluri.h" // File constants static const size_t MAX_HDR_LEN = 20; static const S32 UNZIP_LLSD_MAX_DEPTH = 96; static const char LEGACY_NON_HEADER[] = ""; const std::string LLSD_BINARY_HEADER("LLSD/Binary"); const std::string LLSD_XML_HEADER("LLSD/XML"); const std::string LLSD_NOTATION_HEADER("llsd/notation"); //used to deflate a gzipped asset (currently used for navmeshes) #define windowBits 15 #define ENABLE_ZLIB_GZIP 32 // If we published this in llsdserialize.h, we could use it in the // implementation of LLSDOStreamer's operator<<(). template void format_using(const LLSD& data, std::ostream& ostr, LLSDFormatter::EFormatterOptions options=LLSDFormatter::OPTIONS_PRETTY_BINARY) { LLPointer f{ new Formatter }; f->format(data, ostr, options); } template S32 parse_using(std::istream& istr, LLSD& data, size_t max_bytes, S32 max_depth=-1) { LLPointer p{ new Parser }; return p->parse(istr, data, max_bytes, max_depth); } /** * LLSDSerialize */ // static void LLSDSerialize::serialize(const LLSD& sd, std::ostream& str, ELLSD_Serialize type, LLSDFormatter::EFormatterOptions options) { LLPointer f = NULL; switch (type) { case LLSD_BINARY: str << "\n"; f = new LLSDBinaryFormatter; break; case LLSD_XML: str << "\n"; f = new LLSDXMLFormatter; break; case LLSD_NOTATION: str << "\n"; f = new LLSDNotationFormatter; break; default: LL_WARNS() << "serialize request for unknown ELLSD_Serialize" << LL_ENDL; } if (f.notNull()) { f->format(sd, str, options); } } // static bool LLSDSerialize::deserialize(LLSD& sd, std::istream& str, llssize max_bytes) { char hdr_buf[MAX_HDR_LEN + 1] = ""; /* Flawfinder: ignore */ bool fail_if_not_legacy = false; /* * Get the first line before anything. Don't read more than max_bytes: * this get() overload reads no more than (count-1) bytes into the * specified buffer. In the usual case when max_bytes exceeds * sizeof(hdr_buf), get() will read no more than sizeof(hdr_buf)-2. */ str.get(hdr_buf, llmin(max_bytes+1, sizeof(hdr_buf)-1), '\n'); auto inbuf = str.gcount(); // https://en.cppreference.com/w/cpp/io/basic_istream/get // When the get() above sees the specified delimiter '\n', it stops there // without pulling it from the stream. If it turns out that the stream // does NOT contain a header, and the content includes meaningful '\n', // it's important to pull that into hdr_buf too. if (inbuf < max_bytes && str.get(hdr_buf[inbuf])) { // got the delimiting '\n' ++inbuf; // None of the following requires that hdr_buf contain a final '\0' // byte. We could store one if needed, since even the incremented // inbuf won't exceed sizeof(hdr_buf)-1, but there's no need. } std::string header{ hdr_buf, inbuf }; if (str.fail()) { str.clear(); fail_if_not_legacy = true; } if (!strncasecmp(LEGACY_NON_HEADER, hdr_buf, strlen(LEGACY_NON_HEADER))) /* Flawfinder: ignore */ { // Create a LLSD XML parser, and parse the first chunk read above. LLSDXMLParser x; x.parsePart(hdr_buf, inbuf); // Parse the first part that was already read auto parsed = x.parse(str, sd, max_bytes - inbuf); // Parse the rest of it // Formally we should probably check (parsed != PARSE_FAILURE && // parsed > 0), but since PARSE_FAILURE is -1, this suffices. return (parsed > 0); } if (fail_if_not_legacy) { LL_WARNS() << "deserialize LLSD parse failure" << LL_ENDL; return false; } /* * Remove the newline chars */ std::string::size_type lastchar = header.find_last_not_of("\r\n"); if (lastchar != std::string::npos) { // It's important that find_last_not_of() returns size_type, which is // why lastchar explicitly declares the type above. erase(size_type) // erases from that offset to the end of the string, whereas // erase(iterator) erases only a single character. header.erase(lastchar+1); } // trim off the header syntax auto start = header.find_first_not_of("(str, sd, max_bytes-inbuf) > 0); } else if (0 == LLStringUtil::compareInsensitive(header, LLSD_XML_HEADER)) { return (parse_using(str, sd, max_bytes-inbuf) > 0); } else if (0 == LLStringUtil::compareInsensitive(header, LLSD_NOTATION_HEADER)) { return (parse_using(str, sd, max_bytes-inbuf) > 0); } else // no header we recognize { LLPointer p; if (inbuf && hdr_buf[0] == '<') { // looks like XML LL_DEBUGS() << "deserialize request with no header, assuming XML" << LL_ENDL; p = new LLSDXMLParser; } else { // assume notation LL_DEBUGS() << "deserialize request with no header, assuming notation" << LL_ENDL; p = new LLSDNotationParser; } // Since we've already read 'inbuf' bytes into 'hdr_buf', prepend that // data to whatever remains in 'str'. LLMemoryStreamBuf already(reinterpret_cast(hdr_buf), inbuf); cat_streambuf prebuff(&already, str.rdbuf()); std::istream prepend(&prebuff); #if 1 return (p->parse(prepend, sd, max_bytes) > 0); #else // debugging the reconstituted 'prepend' stream // allocate a buffer that we hope is big enough for the whole thing std::vector wholemsg((max_bytes == size_t(SIZE_UNLIMITED))? 1024 : max_bytes); prepend.read(wholemsg.data(), std::min(max_bytes, wholemsg.size())); LLMemoryStream replay(reinterpret_cast(wholemsg.data()), prepend.gcount()); auto success{ p->parse(replay, sd, prepend.gcount()) > 0 }; { LL_DEBUGS() << (success? "parsed: $$" : "failed: '") << std::string(wholemsg.data(), llmin(prepend.gcount(), 100)) << "$$" << LL_ENDL; } return success; #endif } } /** * Endian handlers */ #if LL_BIG_ENDIAN U64 ll_htonll(U64 hostlonglong) { return hostlonglong; } U64 ll_ntohll(U64 netlonglong) { return netlonglong; } F64 ll_htond(F64 hostlonglong) { return hostlonglong; } F64 ll_ntohd(F64 netlonglong) { return netlonglong; } #else // I read some comments one a indicating that doing an integer add // here would be faster than a bitwise or. For now, the or has // programmer clarity, since the intended outcome matches the // operation. U64 ll_htonll(U64 hostlonglong) { return ((U64)(htonl((U32)((hostlonglong >> 32) & 0xFFFFFFFF))) | ((U64)(htonl((U32)(hostlonglong & 0xFFFFFFFF))) << 32)); } U64 ll_ntohll(U64 netlonglong) { return ((U64)(ntohl((U32)((netlonglong >> 32) & 0xFFFFFFFF))) | ((U64)(ntohl((U32)(netlonglong & 0xFFFFFFFF))) << 32)); } union LLEndianSwapper { F64 d; U64 i; }; F64 ll_htond(F64 hostdouble) { LLEndianSwapper tmp; tmp.d = hostdouble; tmp.i = ll_htonll(tmp.i); return tmp.d; } F64 ll_ntohd(F64 netdouble) { LLEndianSwapper tmp; tmp.d = netdouble; tmp.i = ll_ntohll(tmp.i); return tmp.d; } #endif /** * Local functions. */ /** * @brief Figure out what kind of string it is (raw or delimited) and handoff. * * @param istr The stream to read from. * @param value [out] The string which was found. * @param max_bytes The maximum possible length of the string. Passing in * a negative value will skip this check. * @return Returns number of bytes read off of the stream. Returns * PARSE_FAILURE (-1) on failure. */ llssize deserialize_string(std::istream& istr, std::string& value, llssize max_bytes); /** * @brief Parse a delimited string. * * @param istr The stream to read from, with the delimiter already popped. * @param value [out] The string which was found. * @param d The delimiter to use. * @return Returns number of bytes read off of the stream. Returns * PARSE_FAILURE (-1) on failure. */ llssize deserialize_string_delim(std::istream& istr, std::string& value, char d); /** * @brief Read a raw string off the stream. * * @param istr The stream to read from, with the (len) parameter * leading the stream. * @param value [out] The string which was found. * @param d The delimiter to use. * @param max_bytes The maximum possible length of the string. Passing in * a negative value will skip this check. * @return Returns number of bytes read off of the stream. Returns * PARSE_FAILURE (-1) on failure. */ llssize deserialize_string_raw( std::istream& istr, std::string& value, llssize max_bytes); /** * @brief helper method for dealing with the different notation boolean format. * * @param istr The stream to read from with the leading character stripped. * @param data [out] the result of the parse. * @param compare The string to compare the boolean against * @param vale The value to assign to data if the parse succeeds. * @return Returns number of bytes read off of the stream. Returns * PARSE_FAILURE (-1) on failure. */ llssize deserialize_boolean( std::istream& istr, LLSD& data, const std::string& compare, bool value); /** * @brief Do notation escaping of a string to an ostream. * * @param value The string to escape and serialize * @param str The stream to serialize to. */ void serialize_string(const std::string& value, std::ostream& str); /** * Local constants. */ static const std::string NOTATION_TRUE_SERIAL("true"); static const std::string NOTATION_FALSE_SERIAL("false"); static const char BINARY_TRUE_SERIAL = '1'; static const char BINARY_FALSE_SERIAL = '0'; /** * LLSDParser */ LLSDParser::LLSDParser() : mCheckLimits(true), mMaxBytesLeft(0), mParseLines(false) { } // virtual LLSDParser::~LLSDParser() { } S32 LLSDParser::parse(std::istream& istr, LLSD& data, llssize max_bytes, S32 max_depth) { mCheckLimits = (LLSDSerialize::SIZE_UNLIMITED == max_bytes) ? false : true; mMaxBytesLeft = max_bytes; return doParse(istr, data, max_depth); } // Parse using routine to get() lines, faster than parse() S32 LLSDParser::parseLines(std::istream& istr, LLSD& data) { mCheckLimits = false; mParseLines = true; return doParse(istr, data); } int LLSDParser::get(std::istream& istr) const { if(mCheckLimits) --mMaxBytesLeft; return istr.get(); } std::istream& LLSDParser::get( std::istream& istr, char* s, std::streamsize n, char delim) const { istr.get(s, n, delim); if(mCheckLimits) mMaxBytesLeft -= istr.gcount(); return istr; } std::istream& LLSDParser::get( std::istream& istr, std::streambuf& sb, char delim) const { istr.get(sb, delim); if(mCheckLimits) mMaxBytesLeft -= istr.gcount(); return istr; } std::istream& LLSDParser::ignore(std::istream& istr) const { istr.ignore(); if(mCheckLimits) --mMaxBytesLeft; return istr; } std::istream& LLSDParser::putback(std::istream& istr, char c) const { istr.putback(c); if(mCheckLimits) ++mMaxBytesLeft; return istr; } std::istream& LLSDParser::read( std::istream& istr, char* s, std::streamsize n) const { istr.read(s, n); if(mCheckLimits) mMaxBytesLeft -= istr.gcount(); return istr; } void LLSDParser::account(llssize bytes) const { if(mCheckLimits) mMaxBytesLeft -= bytes; } /** * LLSDNotationParser */ LLSDNotationParser::LLSDNotationParser() { } // virtual LLSDNotationParser::~LLSDNotationParser() { } // virtual S32 LLSDNotationParser::doParse(std::istream& istr, LLSD& data, S32 max_depth) const { // map: { string:object, string:object } // array: [ object, object, object ] // undef: ! // boolean: true | false | 1 | 0 | T | F | t | f | TRUE | FALSE // integer: i#### // real: r#### // uuid: u#### // string: "g'day" | 'have a "nice" day' | s(size)"raw data" // uri: l"escaped" // date: d"YYYY-MM-DDTHH:MM:SS.FFZ" // binary: b##"ff3120ab1" | b(size)"raw data" char c; c = istr.peek(); if (max_depth == 0) { return PARSE_FAILURE; } while(isspace(c)) { // pop the whitespace. c = get(istr); c = istr.peek(); continue; } if(!istr.good()) { return 0; } S32 parse_count = 1; switch(c) { case '{': { S32 child_count = parseMap(istr, data, max_depth - 1); if((child_count == PARSE_FAILURE) || data.isUndefined()) { parse_count = PARSE_FAILURE; } else { parse_count += child_count; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading map." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case '[': { S32 child_count = parseArray(istr, data, max_depth - 1); if((child_count == PARSE_FAILURE) || data.isUndefined()) { parse_count = PARSE_FAILURE; } else { parse_count += child_count; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading array." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case '!': c = get(istr); data.clear(); break; case '0': c = get(istr); data = false; break; case 'F': case 'f': ignore(istr); c = istr.peek(); if(isalpha(c)) { auto cnt = deserialize_boolean( istr, data, NOTATION_FALSE_SERIAL, false); if(PARSE_FAILURE == cnt) parse_count = cnt; else account(cnt); } else { data = false; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading boolean." << LL_ENDL; parse_count = PARSE_FAILURE; } break; case '1': c = get(istr); data = true; break; case 'T': case 't': ignore(istr); c = istr.peek(); if(isalpha(c)) { auto cnt = deserialize_boolean(istr,data,NOTATION_TRUE_SERIAL,true); if(PARSE_FAILURE == cnt) parse_count = cnt; else account(cnt); } else { data = true; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading boolean." << LL_ENDL; parse_count = PARSE_FAILURE; } break; case 'i': { c = get(istr); S32 integer = 0; istr >> integer; data = integer; if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading integer." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'r': { c = get(istr); F64 real = 0.0; istr >> real; data = real; if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading real." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'u': { c = get(istr); LLUUID id; istr >> id; data = id; if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading uuid." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case '\"': case '\'': case 's': if(!parseString(istr, data)) { parse_count = PARSE_FAILURE; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading string." << LL_ENDL; parse_count = PARSE_FAILURE; } break; case 'l': { c = get(istr); // pop the 'l' c = get(istr); // pop the delimiter std::string str; auto cnt = deserialize_string_delim(istr, str, c); if(PARSE_FAILURE == cnt) { parse_count = PARSE_FAILURE; } else { data = LLURI(str); account(cnt); } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading link." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'd': { c = get(istr); // pop the 'd' c = get(istr); // pop the delimiter std::string str; auto cnt = deserialize_string_delim(istr, str, c); if(PARSE_FAILURE == cnt) { parse_count = PARSE_FAILURE; } else { data = LLDate(str); account(cnt); } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading date." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'b': if(!parseBinary(istr, data)) { parse_count = PARSE_FAILURE; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading data." << LL_ENDL; parse_count = PARSE_FAILURE; } break; default: parse_count = PARSE_FAILURE; LL_INFOS() << "Unrecognized character while parsing: int(" << int(c) << ")" << LL_ENDL; break; } if(PARSE_FAILURE == parse_count) { data.clear(); } return parse_count; } S32 LLSDNotationParser::parseMap(std::istream& istr, LLSD& map, S32 max_depth) const { // map: { string:object, string:object } map = LLSD::emptyMap(); S32 parse_count = 0; char c = get(istr); if(c == '{') { // eat commas, white bool found_name = false; std::string name; c = get(istr); while(c != '}' && istr.good()) { if(!found_name) { if((c == '\"') || (c == '\'') || (c == 's')) { putback(istr, c); found_name = true; auto count = deserialize_string(istr, name, mMaxBytesLeft); if(PARSE_FAILURE == count) return PARSE_FAILURE; account(count); } c = get(istr); } else { if(isspace(c) || (c == ':')) { c = get(istr); continue; } putback(istr, c); LLSD child; S32 count = doParse(istr, child, max_depth); if(count > 0) { // There must be a value for every key, thus // child_count must be greater than 0. parse_count += count; map.insert(name, child); } else { return PARSE_FAILURE; } found_name = false; c = get(istr); } } if(c != '}') { map.clear(); return PARSE_FAILURE; } } return parse_count; } S32 LLSDNotationParser::parseArray(std::istream& istr, LLSD& array, S32 max_depth) const { // array: [ object, object, object ] array = LLSD::emptyArray(); S32 parse_count = 0; char c = get(istr); if(c == '[') { // eat commas, white c = get(istr); while((c != ']') && istr.good()) { LLSD child; if(isspace(c) || (c == ',')) { c = get(istr); continue; } putback(istr, c); S32 count = doParse(istr, child, max_depth); if(PARSE_FAILURE == count) { return PARSE_FAILURE; } else { parse_count += count; array.append(child); } c = get(istr); } if(c != ']') { return PARSE_FAILURE; } } return parse_count; } bool LLSDNotationParser::parseString(std::istream& istr, LLSD& data) const { std::string value; auto count = deserialize_string(istr, value, mMaxBytesLeft); if(PARSE_FAILURE == count) return false; account(count); data = value; return true; } bool LLSDNotationParser::parseBinary(std::istream& istr, LLSD& data) const { // binary: b##"ff3120ab1" // or: b(len)"..." // I want to manually control those values here to make sure the // parser doesn't break when someone changes a constant somewhere // else. const U32 BINARY_BUFFER_SIZE = 256; const U32 STREAM_GET_COUNT = 255; // need to read the base out. char buf[BINARY_BUFFER_SIZE]; /* Flawfinder: ignore */ get(istr, buf, STREAM_GET_COUNT, '"'); char c = get(istr); if(c != '"') return false; if(0 == strncmp("b(", buf, 2)) { // We probably have a valid raw binary stream. determine // the size, and read it. auto len = strtol(buf + 2, NULL, 0); if(mCheckLimits && (len > mMaxBytesLeft)) return false; std::vector value; if(len) { value.resize(len); account(fullread(istr, (char *)&value[0], len)); } c = get(istr); // strip off the trailing double-quote data = value; } else if(0 == strncmp("b64", buf, 3)) { // *FIX: A bit inefficient, but works for now. To make the // format better, I would need to add a hint into the // serialization format that indicated how long it was. std::stringstream coded_stream; get(istr, *(coded_stream.rdbuf()), '\"'); c = get(istr); std::string encoded(coded_stream.str()); S32 len = apr_base64_decode_len(encoded.c_str()); std::vector value; if(len) { value.resize(len); len = apr_base64_decode_binary(&value[0], encoded.c_str()); value.resize(len); } data = value; } else if(0 == strncmp("b16", buf, 3)) { // yay, base 16. We pop the next character which is either a // double quote or base 16 data. If it's a double quote, we're // done parsing. If it's not, put the data back, and read the // stream until the next double quote. char* read; /*Flawfinder: ignore*/ U8 byte; U8 byte_buffer[BINARY_BUFFER_SIZE]; U8* write; std::vector value; c = get(istr); while(c != '"') { putback(istr, c); read = buf; write = byte_buffer; get(istr, buf, STREAM_GET_COUNT, '"'); c = get(istr); while(*read != '\0') /*Flawfinder: ignore*/ { byte = hex_as_nybble(*read++); byte = byte << 4; byte |= hex_as_nybble(*read++); *write++ = byte; } // copy the data out of the byte buffer value.insert(value.end(), byte_buffer, write); } data = value; } else { return false; } return true; } /** * LLSDBinaryParser */ LLSDBinaryParser::LLSDBinaryParser() { } // virtual LLSDBinaryParser::~LLSDBinaryParser() { } // virtual S32 LLSDBinaryParser::doParse(std::istream& istr, LLSD& data, S32 max_depth) const { /** * Undefined: '!'
* Boolean: '1' for true '0' for false
* Integer: 'i' + 4 bytes network byte order
* Real: 'r' + 8 bytes IEEE double
* UUID: 'u' + 16 byte unsigned integer
* String: 's' + 4 byte integer size + string
* strings also secretly support the notation format * Date: 'd' + 8 byte IEEE double for seconds since epoch
* URI: 'l' + 4 byte integer size + string uri
* Binary: 'b' + 4 byte integer size + binary data
* Array: '[' + 4 byte integer size + all values + ']'
* Map: '{' + 4 byte integer size every(key + value) + '}'
* map keys are serialized as s + 4 byte integer size + string or in the * notation format. */ char c; c = get(istr); if(!istr.good()) { return 0; } if (max_depth == 0) { return PARSE_FAILURE; } S32 parse_count = 1; switch(c) { case '{': { S32 child_count = parseMap(istr, data, max_depth - 1); if((child_count == PARSE_FAILURE) || data.isUndefined()) { parse_count = PARSE_FAILURE; } else { parse_count += child_count; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary map." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case '[': { S32 child_count = parseArray(istr, data, max_depth - 1); if((child_count == PARSE_FAILURE) || data.isUndefined()) { parse_count = PARSE_FAILURE; } else { parse_count += child_count; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary array." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case '!': data.clear(); break; case '0': data = false; break; case '1': data = true; break; case 'i': { U32 value_nbo = 0; read(istr, (char*)&value_nbo, sizeof(U32)); /*Flawfinder: ignore*/ data = (S32)ntohl(value_nbo); if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary integer." << LL_ENDL; } break; } case 'r': { F64 real_nbo = 0.0; read(istr, (char*)&real_nbo, sizeof(F64)); /*Flawfinder: ignore*/ data = ll_ntohd(real_nbo); if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary real." << LL_ENDL; } break; } case 'u': { LLUUID id; read(istr, (char*)(&id.mData), UUID_BYTES); /*Flawfinder: ignore*/ data = id; if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary uuid." << LL_ENDL; } break; } case '\'': case '"': { std::string value; auto cnt = deserialize_string_delim(istr, value, c); if(PARSE_FAILURE == cnt) { parse_count = PARSE_FAILURE; } else { data = value; account(cnt); } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary (notation-style) string." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 's': { std::string value; if(parseString(istr, value)) { data = value; } else { parse_count = PARSE_FAILURE; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary string." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'l': { std::string value; if(parseString(istr, value)) { data = LLURI(value); } else { parse_count = PARSE_FAILURE; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary link." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'd': { F64 real = 0.0; read(istr, (char*)&real, sizeof(F64)); /*Flawfinder: ignore*/ data = LLDate(real); if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary date." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } case 'b': { // We probably have a valid raw binary stream. determine // the size, and read it. U32 size_nbo = 0; read(istr, (char*)&size_nbo, sizeof(U32)); /*Flawfinder: ignore*/ S32 size = (S32)ntohl(size_nbo); if(mCheckLimits && (size > mMaxBytesLeft)) { parse_count = PARSE_FAILURE; } else { std::vector value; if(size > 0) { value.resize(size); account(fullread(istr, (char*)&value[0], size)); } data = value; } if(istr.fail()) { LL_INFOS() << "STREAM FAILURE reading binary." << LL_ENDL; parse_count = PARSE_FAILURE; } break; } default: parse_count = PARSE_FAILURE; LL_INFOS() << "Unrecognized character while parsing: int(" << int(c) << ")" << LL_ENDL; break; } if(PARSE_FAILURE == parse_count) { data.clear(); } return parse_count; } S32 LLSDBinaryParser::parseMap(std::istream& istr, LLSD& map, S32 max_depth) const { map = LLSD::emptyMap(); U32 value_nbo = 0; read(istr, (char*)&value_nbo, sizeof(U32)); /*Flawfinder: ignore*/ S32 size = (S32)ntohl(value_nbo); S32 parse_count = 0; S32 count = 0; char c = get(istr); while(c != '}' && (count < size) && istr.good()) { std::string name; switch(c) { case 'k': if(!parseString(istr, name)) { return PARSE_FAILURE; } break; case '\'': case '"': { auto cnt = deserialize_string_delim(istr, name, c); if(PARSE_FAILURE == cnt) return PARSE_FAILURE; account(cnt); break; } } LLSD child; S32 child_count = doParse(istr, child, max_depth); if(child_count > 0) { // There must be a value for every key, thus child_count // must be greater than 0. parse_count += child_count; map.insert(name, child); } else { return PARSE_FAILURE; } ++count; c = get(istr); } if((c != '}') || (count < size)) { // Make sure it is correctly terminated and we parsed as many // as were said to be there. return PARSE_FAILURE; } return parse_count; } S32 LLSDBinaryParser::parseArray(std::istream& istr, LLSD& array, S32 max_depth) const { array = LLSD::emptyArray(); U32 value_nbo = 0; read(istr, (char*)&value_nbo, sizeof(U32)); /*Flawfinder: ignore*/ S32 size = (S32)ntohl(value_nbo); // *FIX: This would be a good place to reserve some space in the // array... S32 parse_count = 0; S32 count = 0; char c = istr.peek(); while((c != ']') && (count < size) && istr.good()) { LLSD child; S32 child_count = doParse(istr, child, max_depth); if(PARSE_FAILURE == child_count) { return PARSE_FAILURE; } if(child_count) { parse_count += child_count; array.append(child); } ++count; c = istr.peek(); } c = get(istr); if((c != ']') || (count < size)) { // Make sure it is correctly terminated and we parsed as many // as were said to be there. return PARSE_FAILURE; } return parse_count; } bool LLSDBinaryParser::parseString( std::istream& istr, std::string& value) const { // *FIX: This is memory inefficient. U32 value_nbo = 0; read(istr, (char*)&value_nbo, sizeof(U32)); /*Flawfinder: ignore*/ S32 size = (S32)ntohl(value_nbo); if(mCheckLimits && (size > mMaxBytesLeft)) return false; if(size < 0) return false; std::vector buf; if(size) { buf.resize(size); account(fullread(istr, &buf[0], size)); value.assign(buf.begin(), buf.end()); } return true; } /** * LLSDFormatter */ LLSDFormatter::LLSDFormatter(bool boolAlpha, const std::string& realFmt, EFormatterOptions options): mOptions(options) { boolalpha(boolAlpha); realFormat(realFmt); } // virtual LLSDFormatter::~LLSDFormatter() { } void LLSDFormatter::boolalpha(bool alpha) { mBoolAlpha = alpha; } void LLSDFormatter::realFormat(const std::string& format) { mRealFormat = format; } S32 LLSDFormatter::format(const LLSD& data, std::ostream& ostr) const { // pass options captured by constructor return format(data, ostr, mOptions); } S32 LLSDFormatter::format(const LLSD& data, std::ostream& ostr, EFormatterOptions options) const { return format_impl(data, ostr, options, 0); } void LLSDFormatter::formatReal(LLSD::Real real, std::ostream& ostr) const { std::string buffer = llformat(mRealFormat.c_str(), real); ostr << buffer; } /** * LLSDNotationFormatter */ LLSDNotationFormatter::LLSDNotationFormatter(bool boolAlpha, const std::string& realFormat, EFormatterOptions options): LLSDFormatter(boolAlpha, realFormat, options) { } // virtual LLSDNotationFormatter::~LLSDNotationFormatter() { } // static std::string LLSDNotationFormatter::escapeString(const std::string& in) { std::ostringstream ostr; serialize_string(in, ostr); return ostr.str(); } S32 LLSDNotationFormatter::format_impl(const LLSD& data, std::ostream& ostr, EFormatterOptions options, U32 level) const { S32 format_count = 1; std::string pre; std::string post; if (options & LLSDFormatter::OPTIONS_PRETTY) { for (U32 i = 0; i < level; i++) { pre += " "; } post = "\n"; } switch(data.type()) { case LLSD::TypeMap: { if (0 != level) ostr << post << pre; ostr << "{"; std::string inner_pre; if (options & LLSDFormatter::OPTIONS_PRETTY) { inner_pre = pre + " "; } bool need_comma = false; LLSD::map_const_iterator iter = data.beginMap(); LLSD::map_const_iterator end = data.endMap(); for(; iter != end; ++iter) { if(need_comma) ostr << ","; need_comma = true; ostr << post << inner_pre << '\''; serialize_string((*iter).first, ostr); ostr << "':"; format_count += format_impl((*iter).second, ostr, options, level + 2); } ostr << post << pre << "}"; break; } case LLSD::TypeArray: { ostr << post << pre << "["; bool need_comma = false; LLSD::array_const_iterator iter = data.beginArray(); LLSD::array_const_iterator end = data.endArray(); for(; iter != end; ++iter) { if(need_comma) ostr << ","; need_comma = true; format_count += format_impl(*iter, ostr, options, level + 1); } ostr << "]"; break; } case LLSD::TypeUndefined: ostr << "!"; break; case LLSD::TypeBoolean: if(mBoolAlpha || #if( LL_WINDOWS || __GNUC__ > 2) (ostr.flags() & std::ios::boolalpha) #else (ostr.flags() & 0x0100) #endif ) { ostr << (data.asBoolean() ? NOTATION_TRUE_SERIAL : NOTATION_FALSE_SERIAL); } else { ostr << (data.asBoolean() ? 1 : 0); } break; case LLSD::TypeInteger: ostr << "i" << data.asInteger(); break; case LLSD::TypeReal: ostr << "r"; if(mRealFormat.empty()) { ostr << data.asReal(); } else { formatReal(data.asReal(), ostr); } break; case LLSD::TypeUUID: ostr << "u" << data.asUUID(); break; case LLSD::TypeString: ostr << '\''; serialize_string(data.asStringRef(), ostr); ostr << '\''; break; case LLSD::TypeDate: ostr << "d\"" << data.asDate() << "\""; break; case LLSD::TypeURI: ostr << "l\""; serialize_string(data.asString(), ostr); ostr << "\""; break; case LLSD::TypeBinary: { // *FIX: memory inefficient. const std::vector& buffer = data.asBinary(); if (options & LLSDFormatter::OPTIONS_PRETTY_BINARY) { ostr << "b16\""; if (! buffer.empty()) { std::ios_base::fmtflags old_flags = ostr.flags(); ostr.setf( std::ios::hex, std::ios::basefield ); // It shouldn't strictly matter whether the emitted hex digits // are uppercase; LLSDNotationParser handles either; but as of // 2020-05-13, Python's llbase.llsd requires uppercase hex. ostr << std::uppercase; auto oldfill(ostr.fill('0')); auto oldwidth(ostr.width()); for (size_t i = 0; i < buffer.size(); i++) { // have to restate setw() before every conversion ostr << std::setw(2) << (int) buffer[i]; } ostr.width(oldwidth); ostr.fill(oldfill); ostr.flags(old_flags); } } else // ! OPTIONS_PRETTY_BINARY { ostr << "b(" << buffer.size() << ")\""; if (! buffer.empty()) { ostr.write((const char*)&buffer[0], buffer.size()); } } ostr << "\""; break; } default: // *NOTE: This should never happen. ostr << "!"; break; } return format_count; } /** * LLSDBinaryFormatter */ LLSDBinaryFormatter::LLSDBinaryFormatter(bool boolAlpha, const std::string& realFormat, EFormatterOptions options): LLSDFormatter(boolAlpha, realFormat, options) { } // virtual LLSDBinaryFormatter::~LLSDBinaryFormatter() { } // virtual S32 LLSDBinaryFormatter::format_impl(const LLSD& data, std::ostream& ostr, EFormatterOptions options, U32 level) const { S32 format_count = 1; switch(data.type()) { case LLSD::TypeMap: { ostr.put('{'); U32 size_nbo = htonl(data.size()); ostr.write((const char*)(&size_nbo), sizeof(U32)); LLSD::map_const_iterator iter = data.beginMap(); LLSD::map_const_iterator end = data.endMap(); for(; iter != end; ++iter) { ostr.put('k'); formatString((*iter).first, ostr); format_count += format_impl((*iter).second, ostr, options, level+1); } ostr.put('}'); break; } case LLSD::TypeArray: { ostr.put('['); U32 size_nbo = htonl(data.size()); ostr.write((const char*)(&size_nbo), sizeof(U32)); LLSD::array_const_iterator iter = data.beginArray(); LLSD::array_const_iterator end = data.endArray(); for(; iter != end; ++iter) { format_count += format_impl(*iter, ostr, options, level+1); } ostr.put(']'); break; } case LLSD::TypeUndefined: ostr.put('!'); break; case LLSD::TypeBoolean: if(data.asBoolean()) ostr.put(BINARY_TRUE_SERIAL); else ostr.put(BINARY_FALSE_SERIAL); break; case LLSD::TypeInteger: { ostr.put('i'); U32 value_nbo = htonl(data.asInteger()); ostr.write((const char*)(&value_nbo), sizeof(U32)); break; } case LLSD::TypeReal: { ostr.put('r'); F64 value_nbo = ll_htond(data.asReal()); ostr.write((const char*)(&value_nbo), sizeof(F64)); break; } case LLSD::TypeUUID: { ostr.put('u'); LLUUID temp = data.asUUID(); ostr.write((const char*)(&(temp.mData)), UUID_BYTES); break; } case LLSD::TypeString: ostr.put('s'); formatString(data.asStringRef(), ostr); break; case LLSD::TypeDate: { ostr.put('d'); F64 value = data.asReal(); ostr.write((const char*)(&value), sizeof(F64)); break; } case LLSD::TypeURI: ostr.put('l'); formatString(data.asString(), ostr); break; case LLSD::TypeBinary: { ostr.put('b'); const std::vector& buffer = data.asBinary(); U32 size_nbo = htonl(buffer.size()); ostr.write((const char*)(&size_nbo), sizeof(U32)); if(buffer.size()) ostr.write((const char*)&buffer[0], buffer.size()); break; } default: // *NOTE: This should never happen. ostr.put('!'); break; } return format_count; } void LLSDBinaryFormatter::formatString( const std::string& string, std::ostream& ostr) const { U32 size_nbo = htonl(string.size()); ostr.write((const char*)(&size_nbo), sizeof(U32)); ostr.write(string.c_str(), string.size()); } /** * local functions */ llssize deserialize_string(std::istream& istr, std::string& value, llssize max_bytes) { int c = istr.get(); if(istr.fail()) { // No data in stream, bail out but mention the character we // grabbed. return LLSDParser::PARSE_FAILURE; } llssize rv = LLSDParser::PARSE_FAILURE; switch(c) { case '\'': case '"': rv = deserialize_string_delim(istr, value, c); break; case 's': // technically, less than max_bytes, but this is just meant to // catch egregious protocol errors. parse errors will be // caught in the case of incorrect counts. rv = deserialize_string_raw(istr, value, max_bytes); break; default: break; } if(LLSDParser::PARSE_FAILURE == rv) return rv; return rv + 1; // account for the character grabbed at the top. } llssize deserialize_string_delim( std::istream& istr, std::string& value, char delim) { std::ostringstream write_buffer; bool found_escape = false; bool found_hex = false; bool found_digit = false; U8 byte = 0; llssize count = 0; while (true) { int next_byte = istr.get(); ++count; if(istr.fail()) { // If our stream is empty, break out value = write_buffer.str(); return LLSDParser::PARSE_FAILURE; } char next_char = (char)next_byte; // Now that we know it's not EOF if(found_escape) { // next character(s) is a special sequence. if(found_hex) { if(found_digit) { found_digit = false; found_hex = false; found_escape = false; byte = byte << 4; byte |= hex_as_nybble(next_char); write_buffer << byte; byte = 0; } else { // next character is the first nybble of // found_digit = true; byte = hex_as_nybble(next_char); } } else if(next_char == 'x') { found_hex = true; } else { switch(next_char) { case 'a': write_buffer << '\a'; break; case 'b': write_buffer << '\b'; break; case 'f': write_buffer << '\f'; break; case 'n': write_buffer << '\n'; break; case 'r': write_buffer << '\r'; break; case 't': write_buffer << '\t'; break; case 'v': write_buffer << '\v'; break; default: write_buffer << next_char; break; } found_escape = false; } } else if(next_char == '\\') { found_escape = true; } else if(next_char == delim) { break; } else { write_buffer << next_char; } } value = write_buffer.str(); return count; } llssize deserialize_string_raw( std::istream& istr, std::string& value, llssize max_bytes) { llssize count = 0; const S32 BUF_LEN = 20; char buf[BUF_LEN]; /* Flawfinder: ignore */ istr.get(buf, BUF_LEN - 1, ')'); count += istr.gcount(); int c = istr.get(); c = istr.get(); count += 2; if(((c == '"') || (c == '\'')) && (buf[0] == '(')) { // We probably have a valid raw string. determine // the size, and read it. // *FIX: This is memory inefficient. auto len = strtol(buf + 1, NULL, 0); if((max_bytes>0)&&(len>max_bytes)) return LLSDParser::PARSE_FAILURE; std::vector buf; if(len) { buf.resize(len); count += fullread(istr, (char *)&buf[0], len); value.assign(buf.begin(), buf.end()); } c = istr.get(); ++count; if(!((c == '"') || (c == '\''))) { return LLSDParser::PARSE_FAILURE; } } else { return LLSDParser::PARSE_FAILURE; } return count; } static const char* NOTATION_STRING_CHARACTERS[256] = { "\\x00", // 0 "\\x01", // 1 "\\x02", // 2 "\\x03", // 3 "\\x04", // 4 "\\x05", // 5 "\\x06", // 6 "\\a", // 7 "\\b", // 8 "\\t", // 9 "\\n", // 10 "\\v", // 11 "\\f", // 12 "\\r", // 13 "\\x0e", // 14 "\\x0f", // 15 "\\x10", // 16 "\\x11", // 17 "\\x12", // 18 "\\x13", // 19 "\\x14", // 20 "\\x15", // 21 "\\x16", // 22 "\\x17", // 23 "\\x18", // 24 "\\x19", // 25 "\\x1a", // 26 "\\x1b", // 27 "\\x1c", // 28 "\\x1d", // 29 "\\x1e", // 30 "\\x1f", // 31 " ", // 32 "!", // 33 "\"", // 34 "#", // 35 "$", // 36 "%", // 37 "&", // 38 "\\'", // 39 "(", // 40 ")", // 41 "*", // 42 "+", // 43 ",", // 44 "-", // 45 ".", // 46 "/", // 47 "0", // 48 "1", // 49 "2", // 50 "3", // 51 "4", // 52 "5", // 53 "6", // 54 "7", // 55 "8", // 56 "9", // 57 ":", // 58 ";", // 59 "<", // 60 "=", // 61 ">", // 62 "?", // 63 "@", // 64 "A", // 65 "B", // 66 "C", // 67 "D", // 68 "E", // 69 "F", // 70 "G", // 71 "H", // 72 "I", // 73 "J", // 74 "K", // 75 "L", // 76 "M", // 77 "N", // 78 "O", // 79 "P", // 80 "Q", // 81 "R", // 82 "S", // 83 "T", // 84 "U", // 85 "V", // 86 "W", // 87 "X", // 88 "Y", // 89 "Z", // 90 "[", // 91 "\\\\", // 92 "]", // 93 "^", // 94 "_", // 95 "`", // 96 "a", // 97 "b", // 98 "c", // 99 "d", // 100 "e", // 101 "f", // 102 "g", // 103 "h", // 104 "i", // 105 "j", // 106 "k", // 107 "l", // 108 "m", // 109 "n", // 110 "o", // 111 "p", // 112 "q", // 113 "r", // 114 "s", // 115 "t", // 116 "u", // 117 "v", // 118 "w", // 119 "x", // 120 "y", // 121 "z", // 122 "{", // 123 "|", // 124 "}", // 125 "~", // 126 "\\x7f", // 127 "\\x80", // 128 "\\x81", // 129 "\\x82", // 130 "\\x83", // 131 "\\x84", // 132 "\\x85", // 133 "\\x86", // 134 "\\x87", // 135 "\\x88", // 136 "\\x89", // 137 "\\x8a", // 138 "\\x8b", // 139 "\\x8c", // 140 "\\x8d", // 141 "\\x8e", // 142 "\\x8f", // 143 "\\x90", // 144 "\\x91", // 145 "\\x92", // 146 "\\x93", // 147 "\\x94", // 148 "\\x95", // 149 "\\x96", // 150 "\\x97", // 151 "\\x98", // 152 "\\x99", // 153 "\\x9a", // 154 "\\x9b", // 155 "\\x9c", // 156 "\\x9d", // 157 "\\x9e", // 158 "\\x9f", // 159 "\\xa0", // 160 "\\xa1", // 161 "\\xa2", // 162 "\\xa3", // 163 "\\xa4", // 164 "\\xa5", // 165 "\\xa6", // 166 "\\xa7", // 167 "\\xa8", // 168 "\\xa9", // 169 "\\xaa", // 170 "\\xab", // 171 "\\xac", // 172 "\\xad", // 173 "\\xae", // 174 "\\xaf", // 175 "\\xb0", // 176 "\\xb1", // 177 "\\xb2", // 178 "\\xb3", // 179 "\\xb4", // 180 "\\xb5", // 181 "\\xb6", // 182 "\\xb7", // 183 "\\xb8", // 184 "\\xb9", // 185 "\\xba", // 186 "\\xbb", // 187 "\\xbc", // 188 "\\xbd", // 189 "\\xbe", // 190 "\\xbf", // 191 "\\xc0", // 192 "\\xc1", // 193 "\\xc2", // 194 "\\xc3", // 195 "\\xc4", // 196 "\\xc5", // 197 "\\xc6", // 198 "\\xc7", // 199 "\\xc8", // 200 "\\xc9", // 201 "\\xca", // 202 "\\xcb", // 203 "\\xcc", // 204 "\\xcd", // 205 "\\xce", // 206 "\\xcf", // 207 "\\xd0", // 208 "\\xd1", // 209 "\\xd2", // 210 "\\xd3", // 211 "\\xd4", // 212 "\\xd5", // 213 "\\xd6", // 214 "\\xd7", // 215 "\\xd8", // 216 "\\xd9", // 217 "\\xda", // 218 "\\xdb", // 219 "\\xdc", // 220 "\\xdd", // 221 "\\xde", // 222 "\\xdf", // 223 "\\xe0", // 224 "\\xe1", // 225 "\\xe2", // 226 "\\xe3", // 227 "\\xe4", // 228 "\\xe5", // 229 "\\xe6", // 230 "\\xe7", // 231 "\\xe8", // 232 "\\xe9", // 233 "\\xea", // 234 "\\xeb", // 235 "\\xec", // 236 "\\xed", // 237 "\\xee", // 238 "\\xef", // 239 "\\xf0", // 240 "\\xf1", // 241 "\\xf2", // 242 "\\xf3", // 243 "\\xf4", // 244 "\\xf5", // 245 "\\xf6", // 246 "\\xf7", // 247 "\\xf8", // 248 "\\xf9", // 249 "\\xfa", // 250 "\\xfb", // 251 "\\xfc", // 252 "\\xfd", // 253 "\\xfe", // 254 "\\xff" // 255 }; void serialize_string(const std::string& value, std::ostream& str) { std::string::const_iterator it = value.begin(); std::string::const_iterator end = value.end(); U8 c; for(; it != end; ++it) { c = (U8)(*it); str << NOTATION_STRING_CHARACTERS[c]; } } llssize deserialize_boolean( std::istream& istr, LLSD& data, const std::string& compare, bool value) { // // this method is a little goofy, because it gets the stream at // the point where the t or f has already been // consumed. Basically, parse for a patch to the string passed in // starting at index 1. If it's a match: // * assign data to value // * return the number of bytes read // otherwise: // * set data to LLSD::null // * return LLSDParser::PARSE_FAILURE (-1) // llssize bytes_read = 0; std::string::size_type ii = 0; char c = istr.peek(); while((++ii < compare.size()) && (tolower(c) == (int)compare[ii]) && istr.good()) { istr.ignore(); ++bytes_read; c = istr.peek(); } if(compare.size() != ii) { data.clear(); return LLSDParser::PARSE_FAILURE; } data = value; return bytes_read; } std::ostream& operator<<(std::ostream& s, const LLSD& llsd) { s << LLSDNotationStreamer(llsd); return s; } //dirty little zippers -- yell at davep if these are horrid //return a string containing gzipped bytes of binary serialized LLSD // VERY inefficient -- creates several copies of LLSD block in memory std::string zip_llsd(LLSD& data) { std::stringstream llsd_strm; LLSDSerialize::toBinary(data, llsd_strm); const U32 CHUNK = 65536; z_stream strm; strm.zalloc = Z_NULL; strm.zfree = Z_NULL; strm.opaque = Z_NULL; S32 ret = deflateInit(&strm, Z_BEST_COMPRESSION); if (ret != Z_OK) { LL_WARNS() << "Failed to compress LLSD block." << LL_ENDL; return std::string(); } std::string source = llsd_strm.str(); U8 out[CHUNK]; strm.avail_in = narrow(source.size()); strm.next_in = (U8*) source.data(); U8* output = NULL; U32 cur_size = 0; U32 have = 0; do { strm.avail_out = CHUNK; strm.next_out = out; ret = deflate(&strm, Z_FINISH); if (ret == Z_OK || ret == Z_STREAM_END) { //copy result into output if (strm.avail_out >= CHUNK) { free(output); LL_WARNS() << "Failed to compress LLSD block." << LL_ENDL; return std::string(); } have = CHUNK-strm.avail_out; U8* new_output = (U8*) realloc(output, cur_size+have); if (new_output == NULL) { LL_WARNS() << "Failed to compress LLSD block: can't reallocate memory, current size: " << cur_size << " bytes; requested " << cur_size + have << " bytes." << LL_ENDL; deflateEnd(&strm); if (output) { free(output); } return std::string(); } output = new_output; memcpy(output+cur_size, out, have); cur_size += have; } else { free(output); LL_WARNS() << "Failed to compress LLSD block." << LL_ENDL; return std::string(); } } while (ret == Z_OK); std::string::size_type size = cur_size; std::string result((char*) output, size); deflateEnd(&strm); free(output); return result; } //decompress a block of LLSD from provided istream // not very efficient -- creats a copy of decompressed LLSD block in memory // and deserializes from that copy using LLSDSerialize LLUZipHelper::EZipRresult LLUZipHelper::unzip_llsd(LLSD& data, std::istream& is, S32 size) { U8* result = NULL; llssize cur_size = 0; z_stream strm; const U32 CHUNK = 65536; U8 *in = new(std::nothrow) U8[size]; if (!in) { return ZR_MEM_ERROR; } is.read((char*) in, size); U8 out[CHUNK]; strm.zalloc = Z_NULL; strm.zfree = Z_NULL; strm.opaque = Z_NULL; strm.avail_in = size; strm.next_in = in; S32 ret = inflateInit(&strm); do { strm.avail_out = CHUNK; strm.next_out = out; ret = inflate(&strm, Z_NO_FLUSH); if (ret == Z_STREAM_ERROR) { inflateEnd(&strm); free(result); delete [] in; return ZR_DATA_ERROR; } switch (ret) { case Z_NEED_DICT: ret = Z_DATA_ERROR; case Z_DATA_ERROR: case Z_MEM_ERROR: inflateEnd(&strm); free(result); delete [] in; return ZR_MEM_ERROR; break; } U32 have = CHUNK-strm.avail_out; U8* new_result = (U8*)realloc(result, cur_size + have); if (new_result == NULL) { inflateEnd(&strm); if (result) { free(result); } delete[] in; return ZR_MEM_ERROR; } result = new_result; memcpy(result+cur_size, out, have); cur_size += have; } while (ret == Z_OK); inflateEnd(&strm); delete [] in; if (ret != Z_STREAM_END) { free(result); return ZR_DATA_ERROR; } //result now points to the decompressed LLSD block { std::istringstream istr; // Since we are using this for meshes, data we are dealing with tend to be large. // So string can potentially fail to allocate, make sure this won't cause problems try { std::string res_str((char*)result, cur_size); std::string deprecated_header(""); if (res_str.substr(0, deprecated_header.size()) == deprecated_header) { res_str = res_str.substr(deprecated_header.size() + 1, cur_size); } cur_size = res_str.size(); istr.str(res_str); } #ifdef LL_WINDOWS catch (std::length_error) { free(result); return ZR_SIZE_ERROR; } #endif catch (std::bad_alloc&) { free(result); return ZR_MEM_ERROR; } if (!LLSDSerialize::fromBinary(data, istr, cur_size, UNZIP_LLSD_MAX_DEPTH)) { free(result); return ZR_PARSE_ERROR; } } free(result); return ZR_OK; } //This unzip function will only work with a gzip header and trailer - while the contents //of the actual compressed data is the same for either format (gzip vs zlib ), the headers //and trailers are different for the formats. U8* unzip_llsdNavMesh( bool& valid, size_t& outsize, std::istream& is, S32 size ) { if (size == 0) { LL_WARNS() << "No data to unzip." << LL_ENDL; return NULL; } U8* result = NULL; U32 cur_size = 0; z_stream strm; const U32 CHUNK = 0x4000; U8 *in = new(std::nothrow) U8[size]; if (in == NULL) { LL_WARNS() << "Memory allocation failure." << LL_ENDL; return NULL; } is.read((char*) in, size); U8 out[CHUNK]; strm.zalloc = Z_NULL; strm.zfree = Z_NULL; strm.opaque = Z_NULL; strm.avail_in = size; strm.next_in = in; S32 ret = inflateInit2(&strm, windowBits | ENABLE_ZLIB_GZIP ); do { strm.avail_out = CHUNK; strm.next_out = out; ret = inflate(&strm, Z_NO_FLUSH); if (ret == Z_STREAM_ERROR) { inflateEnd(&strm); free(result); delete [] in; valid = false; } switch (ret) { case Z_NEED_DICT: ret = Z_DATA_ERROR; case Z_DATA_ERROR: case Z_MEM_ERROR: inflateEnd(&strm); free(result); delete [] in; valid = false; break; } U32 have = CHUNK-strm.avail_out; U8* new_result = (U8*) realloc(result, cur_size + have); if (new_result == NULL) { LL_WARNS() << "Failed to unzip LLSD NavMesh block: can't reallocate memory, current size: " << cur_size << " bytes; requested " << cur_size + have << " bytes; total syze: ." << size << " bytes." << LL_ENDL; inflateEnd(&strm); if (result) { free(result); } delete[] in; valid = false; return NULL; } result = new_result; memcpy(result+cur_size, out, have); cur_size += have; } while (ret == Z_OK); inflateEnd(&strm); delete [] in; if (ret != Z_STREAM_END) { free(result); valid = false; return NULL; } //result now points to the decompressed LLSD block { outsize= cur_size; valid = true; } return result; }