12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378237923802381238223832384238523862387238823892390239123922393239423952396239723982399240024012402240324042405240624072408240924102411241224132414241524162417241824192420242124222423242424252426242724282429243024312432243324342435243624372438243924402441244224432444244524462447244824492450245124522453245424552456245724582459246024612462246324642465246624672468246924702471247224732474247524762477247824792480248124822483248424852486248724882489249024912492249324942495249624972498249925002501250225032504250525062507250825092510251125122513251425152516251725182519252025212522252325242525252625272528252925302531253225332534253525362537253825392540254125422543254425452546254725482549255025512552255325542555255625572558255925602561256225632564256525662567256825692570257125722573257425752576257725782579258025812582258325842585258625872588258925902591259225932594259525962597259825992600260126022603260426052606260726082609261026112612261326142615261626172618261926202621262226232624262526262627262826292630263126322633263426352636263726382639264026412642264326442645264626472648264926502651265226532654265526562657265826592660266126622663266426652666266726682669267026712672267326742675267626772678267926802681268226832684268526862687268826892690269126922693269426952696269726982699270027012702270327042705270627072708270927102711271227132714271527162717271827192720272127222723272427252726272727282729273027312732273327342735273627372738273927402741274227432744274527462747274827492750275127522753275427552756275727582759276027612762276327642765276627672768276927702771277227732774277527762777277827792780278127822783278427852786278727882789279027912792279327942795279627972798279928002801280228032804280528062807280828092810281128122813281428152816281728182819282028212822282328242825282628272828282928302831283228332834283528362837283828392840284128422843284428452846284728482849285028512852285328542855285628572858285928602861286228632864286528662867286828692870287128722873287428752876287728782879288028812882288328842885288628872888288928902891289228932894289528962897289828992900290129022903290429052906290729082909291029112912291329142915291629172918291929202921292229232924292529262927292829292930293129322933293429352936293729382939294029412942294329442945294629472948294929502951295229532954295529562957295829592960296129622963296429652966296729682969297029712972297329742975297629772978297929802981298229832984298529862987298829892990299129922993299429952996299729982999300030013002300330043005300630073008300930103011301230133014301530163017301830193020302130223023302430253026302730283029303030313032303330343035303630373038303930403041304230433044304530463047304830493050305130523053305430553056305730583059306030613062306330643065306630673068306930703071307230733074307530763077307830793080308130823083308430853086308730883089309030913092309330943095309630973098309931003101310231033104310531063107310831093110311131123113311431153116311731183119312031213122312331243125312631273128312931303131313231333134313531363137313831393140314131423143314431453146314731483149315031513152315331543155315631573158315931603161316231633164316531663167316831693170317131723173317431753176317731783179318031813182318331843185318631873188318931903191319231933194319531963197319831993200320132023203320432053206320732083209321032113212 |
- /* Copyright (c) 2013, Vsevolod Stakhov
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions are met:
- * * Redistributions of source code must retain the above copyright
- * notice, this list of conditions and the following disclaimer.
- * * Redistributions in binary form must reproduce the above copyright
- * notice, this list of conditions and the following disclaimer in the
- * documentation and/or other materials provided with the distribution.
- *
- * THIS SOFTWARE IS PROVIDED ''AS IS'' AND ANY
- * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
- * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
- * DISCLAIMED. IN NO EVENT SHALL AUTHOR BE LIABLE FOR ANY
- * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
- * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
- * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
- * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
- * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- */
-
- #include <math.h>
- #include "ucl.h"
- #include "ucl_internal.h"
- #include "ucl_chartable.h"
-
- /**
- * @file ucl_parser.c
- * The implementation of ucl parser
- */
-
- struct ucl_parser_saved_state {
- unsigned int line;
- unsigned int column;
- size_t remain;
- const unsigned char *pos;
- };
-
- /**
- * Move up to len characters
- * @param parser
- * @param begin
- * @param len
- * @return new position in chunk
- */
- #define ucl_chunk_skipc(chunk, p) \
- do { \
- if (p == chunk->end) { \
- break; \
- } \
- if (*(p) == '\n') { \
- (chunk)->line ++; \
- (chunk)->column = 0; \
- } \
- else (chunk)->column ++; \
- (p++); \
- (chunk)->pos ++; \
- (chunk)->remain --; \
- } while (0)
-
- static inline void
- ucl_set_err (struct ucl_parser *parser, int code, const char *str, UT_string **err)
- {
- const char *fmt_string, *filename;
- struct ucl_chunk *chunk = parser->chunks;
-
- if (parser->cur_file) {
- filename = parser->cur_file;
- }
- else {
- filename = "<unknown>";
- }
-
- if (chunk->pos < chunk->end) {
- if (isgraph (*chunk->pos)) {
- fmt_string = "error while parsing %s: "
- "line: %d, column: %d - '%s', character: '%c'";
- }
- else {
- fmt_string = "error while parsing %s: "
- "line: %d, column: %d - '%s', character: '0x%02x'";
- }
- ucl_create_err (err, fmt_string,
- filename, chunk->line, chunk->column,
- str, *chunk->pos);
- }
- else {
- ucl_create_err (err, "error while parsing %s: at the end of chunk: %s",
- filename, str);
- }
-
- parser->err_code = code;
- parser->state = UCL_STATE_ERROR;
- }
-
- static void
- ucl_save_comment (struct ucl_parser *parser, const char *begin, size_t len)
- {
- ucl_object_t *nobj;
-
- if (len > 0 && begin != NULL) {
- nobj = ucl_object_fromstring_common (begin, len, 0);
-
- if (parser->last_comment) {
- /* We need to append data to an existing object */
- DL_APPEND (parser->last_comment, nobj);
- }
- else {
- parser->last_comment = nobj;
- }
- }
- }
-
- static void
- ucl_attach_comment (struct ucl_parser *parser, ucl_object_t *obj, bool before)
- {
- if (parser->last_comment) {
- ucl_object_insert_key (parser->comments, parser->last_comment,
- (const char *)&obj, sizeof (void *), true);
-
- if (before) {
- parser->last_comment->flags |= UCL_OBJECT_INHERITED;
- }
-
- parser->last_comment = NULL;
- }
- }
-
- /**
- * Skip all comments from the current pos resolving nested and multiline comments
- * @param parser
- * @return
- */
- static bool
- ucl_skip_comments (struct ucl_parser *parser)
- {
- struct ucl_chunk *chunk = parser->chunks;
- const unsigned char *p, *beg = NULL;
- int comments_nested = 0;
- bool quoted = false;
-
- p = chunk->pos;
-
- start:
- if (chunk->remain > 0 && *p == '#') {
- if (parser->state != UCL_STATE_SCOMMENT &&
- parser->state != UCL_STATE_MCOMMENT) {
- beg = p;
-
- while (p < chunk->end) {
- if (*p == '\n') {
- if (parser->flags & UCL_PARSER_SAVE_COMMENTS) {
- ucl_save_comment (parser, beg, p - beg);
- beg = NULL;
- }
-
- ucl_chunk_skipc (chunk, p);
-
- goto start;
- }
- ucl_chunk_skipc (chunk, p);
- }
- }
- }
- else if (chunk->remain >= 2 && *p == '/') {
- if (p[1] == '*') {
- beg = p;
- ucl_chunk_skipc (chunk, p);
- comments_nested ++;
- ucl_chunk_skipc (chunk, p);
-
- while (p < chunk->end) {
- if (*p == '"' && *(p - 1) != '\\') {
- quoted = !quoted;
- }
-
- if (!quoted) {
- if (*p == '*') {
- ucl_chunk_skipc (chunk, p);
- if (chunk->remain > 0 && *p == '/') {
- comments_nested --;
- if (comments_nested == 0) {
- if (parser->flags & UCL_PARSER_SAVE_COMMENTS) {
- ucl_save_comment (parser, beg, p - beg + 1);
- beg = NULL;
- }
-
- ucl_chunk_skipc (chunk, p);
- goto start;
- }
- }
- ucl_chunk_skipc (chunk, p);
- }
- else if (p[0] == '/' && chunk->remain >= 2 && p[1] == '*') {
- comments_nested ++;
- ucl_chunk_skipc (chunk, p);
- ucl_chunk_skipc (chunk, p);
- continue;
- }
- }
-
- ucl_chunk_skipc (chunk, p);
- }
- if (comments_nested != 0) {
- ucl_set_err (parser, UCL_ENESTED,
- "unfinished multiline comment", &parser->err);
- return false;
- }
- }
- }
-
- if (beg && p > beg && (parser->flags & UCL_PARSER_SAVE_COMMENTS)) {
- ucl_save_comment (parser, beg, p - beg);
- }
-
- return true;
- }
-
- /**
- * Return multiplier for a character
- * @param c multiplier character
- * @param is_bytes if true use 1024 multiplier
- * @return multiplier
- */
- static inline unsigned long
- ucl_lex_num_multiplier (const unsigned char c, bool is_bytes) {
- const struct {
- char c;
- long mult_normal;
- long mult_bytes;
- } multipliers[] = {
- {'m', 1000 * 1000, 1024 * 1024},
- {'k', 1000, 1024},
- {'g', 1000 * 1000 * 1000, 1024 * 1024 * 1024}
- };
- int i;
-
- for (i = 0; i < 3; i ++) {
- if (tolower (c) == multipliers[i].c) {
- if (is_bytes) {
- return multipliers[i].mult_bytes;
- }
- return multipliers[i].mult_normal;
- }
- }
-
- return 1;
- }
-
-
- /**
- * Return multiplier for time scaling
- * @param c
- * @return
- */
- static inline double
- ucl_lex_time_multiplier (const unsigned char c) {
- const struct {
- char c;
- double mult;
- } multipliers[] = {
- {'m', 60},
- {'h', 60 * 60},
- {'d', 60 * 60 * 24},
- {'w', 60 * 60 * 24 * 7},
- {'y', 60 * 60 * 24 * 365}
- };
- int i;
-
- for (i = 0; i < 5; i ++) {
- if (tolower (c) == multipliers[i].c) {
- return multipliers[i].mult;
- }
- }
-
- return 1;
- }
-
- /**
- * Return true if a character is a end of an atom
- * @param c
- * @return
- */
- static inline bool
- ucl_lex_is_atom_end (const unsigned char c)
- {
- return ucl_test_character (c, UCL_CHARACTER_VALUE_END);
- }
-
- static inline bool
- ucl_lex_is_comment (const unsigned char c1, const unsigned char c2)
- {
- if (c1 == '/') {
- if (c2 == '*') {
- return true;
- }
- }
- else if (c1 == '#') {
- return true;
- }
- return false;
- }
-
- /**
- * Check variable found
- * @param parser
- * @param ptr
- * @param remain
- * @param out_len
- * @param strict
- * @param found
- * @return
- */
- static inline const char *
- ucl_check_variable_safe (struct ucl_parser *parser, const char *ptr, size_t remain,
- size_t *out_len, bool strict, bool *found)
- {
- struct ucl_variable *var;
- unsigned char *dst;
- size_t dstlen;
- bool need_free = false;
-
- LL_FOREACH (parser->variables, var) {
- if (strict) {
- if (remain == var->var_len) {
- if (memcmp (ptr, var->var, var->var_len) == 0) {
- *out_len += var->value_len;
- *found = true;
- return (ptr + var->var_len);
- }
- }
- }
- else {
- if (remain >= var->var_len) {
- if (memcmp (ptr, var->var, var->var_len) == 0) {
- *out_len += var->value_len;
- *found = true;
- return (ptr + var->var_len);
- }
- }
- }
- }
-
- /* XXX: can only handle ${VAR} */
- if (!(*found) && parser->var_handler != NULL && strict) {
- /* Call generic handler */
- if (parser->var_handler (ptr, remain, &dst, &dstlen, &need_free,
- parser->var_data)) {
- *found = true;
- if (need_free) {
- free (dst);
- }
- return (ptr + remain);
- }
- }
-
- return ptr;
- }
-
- /**
- * Check for a variable in a given string
- * @param parser
- * @param ptr
- * @param remain
- * @param out_len
- * @param vars_found
- * @return
- */
- static const char *
- ucl_check_variable (struct ucl_parser *parser, const char *ptr,
- size_t remain, size_t *out_len, bool *vars_found)
- {
- const char *p, *end, *ret = ptr;
- bool found = false;
-
- if (*ptr == '{') {
- /* We need to match the variable enclosed in braces */
- p = ptr + 1;
- end = ptr + remain;
- while (p < end) {
- if (*p == '}') {
- ret = ucl_check_variable_safe (parser, ptr + 1, p - ptr - 1,
- out_len, true, &found);
- if (found) {
- /* {} must be excluded actually */
- ret ++;
- if (!*vars_found) {
- *vars_found = true;
- }
- }
- else {
- *out_len += 2;
- }
- break;
- }
- p ++;
- }
- if(p == end) {
- (*out_len) ++;
- }
- }
- else if (*ptr != '$') {
- /* Not count escaped dollar sign */
- ret = ucl_check_variable_safe (parser, ptr, remain, out_len, false, &found);
- if (found && !*vars_found) {
- *vars_found = true;
- }
- if (!found) {
- (*out_len) ++;
- }
- }
- else {
- ret ++;
- (*out_len) ++;
- }
-
- return ret;
- }
-
- /**
- * Expand a single variable
- * @param parser
- * @param ptr
- * @param in_len
- * @param dest
- * @param out_len
- * @return
- */
- static const char *
- ucl_expand_single_variable (struct ucl_parser *parser, const char *ptr,
- size_t in_len, unsigned char **dest, size_t out_len)
- {
- unsigned char *d = *dest, *dst;
- const char *p = ptr + 1, *ret;
- struct ucl_variable *var;
- size_t dstlen;
- bool need_free = false;
- bool found = false;
- bool strict = false;
-
- ret = ptr + 1;
- /* For the $ sign */
- in_len --;
-
- if (*p == '$') {
- *d++ = *p++;
- *dest = d;
- return p;
- }
- else if (*p == '{') {
- p ++;
- in_len --;
- strict = true;
- ret += 2;
- }
-
- LL_FOREACH (parser->variables, var) {
- if (out_len >= var->value_len && in_len >= (var->var_len + (strict ? 1 : 0))) {
- if (memcmp (p, var->var, var->var_len) == 0) {
- if (!strict || p[var->var_len] == '}') {
- memcpy (d, var->value, var->value_len);
- ret += var->var_len;
- d += var->value_len;
- found = true;
- break;
- }
- }
- }
- }
-
- if (!found) {
- if (strict && parser->var_handler != NULL) {
- dstlen = out_len;
-
- if (parser->var_handler (p, in_len, &dst, &dstlen, &need_free,
- parser->var_data)) {
- if (dstlen > out_len) {
- /* We do not have enough space! */
- if (need_free) {
- free (dst);
- }
- }
- else {
- memcpy(d, dst, dstlen);
- ret += in_len;
- d += dstlen;
- found = true;
-
- if (need_free) {
- free(dst);
- }
- }
- }
- }
-
- /* Leave variable as is, in this case we use dest */
- if (!found) {
- if (strict && out_len >= 2) {
- /* Copy '${' */
- memcpy (d, ptr, 2);
- d += 2;
- ret --;
- }
- else {
- memcpy (d, ptr, 1);
- d ++;
- }
- }
- }
-
- *dest = d;
- return ret;
- }
-
- /**
- * Expand variables in string
- * @param parser
- * @param dst
- * @param src
- * @param in_len
- * @return
- */
- static ssize_t
- ucl_expand_variable (struct ucl_parser *parser, unsigned char **dst,
- const char *src, size_t in_len)
- {
- const char *p, *end = src + in_len;
- unsigned char *d, *d_end;
- size_t out_len = 0;
- bool vars_found = false;
-
- if (parser->flags & UCL_PARSER_DISABLE_MACRO) {
- *dst = NULL;
- return in_len;
- }
-
- p = src;
- while (p != end) {
- if (*p == '$' && p + 1 != end) {
- p = ucl_check_variable (parser, p + 1, end - p - 1, &out_len, &vars_found);
- }
- else {
- p ++;
- out_len ++;
- }
- }
-
- if (!vars_found) {
- /* Trivial case */
- *dst = NULL;
- return in_len;
- }
-
- *dst = UCL_ALLOC (out_len + 1);
- if (*dst == NULL) {
- return in_len;
- }
-
- d = *dst;
- d_end = d + out_len;
- p = src;
- while (p != end && d != d_end) {
- if (*p == '$' && p + 1 != end) {
- p = ucl_expand_single_variable (parser, p, end - p, &d, d_end - d);
- }
- else {
- *d++ = *p++;
- }
- }
-
- *d = '\0';
-
- return out_len;
- }
-
- /**
- * Store or copy pointer to the trash stack
- * @param parser parser object
- * @param src src string
- * @param dst destination buffer (trash stack pointer)
- * @param dst_const const destination pointer (e.g. value of object)
- * @param in_len input length
- * @param need_unescape need to unescape source (and copy it)
- * @param need_lowercase need to lowercase value (and copy)
- * @param need_expand need to expand variables (and copy as well)
- * @param unescape_squote unescape single quoted string
- * @return output length (excluding \0 symbol)
- */
- static inline ssize_t
- ucl_copy_or_store_ptr (struct ucl_parser *parser,
- const unsigned char *src, unsigned char **dst,
- const char **dst_const, size_t in_len,
- bool need_unescape, bool need_lowercase, bool need_expand,
- bool unescape_squote)
- {
- ssize_t ret = -1, tret;
- unsigned char *tmp;
-
- if (need_unescape || need_lowercase ||
- (need_expand && parser->variables != NULL) ||
- !(parser->flags & UCL_PARSER_ZEROCOPY)) {
- /* Copy string */
- *dst = UCL_ALLOC (in_len + 1);
- if (*dst == NULL) {
- ucl_set_err (parser, UCL_EINTERNAL, "cannot allocate memory for a string",
- &parser->err);
- return false;
- }
- if (need_lowercase) {
- ret = ucl_strlcpy_tolower (*dst, src, in_len + 1);
- }
- else {
- ret = ucl_strlcpy_unsafe (*dst, src, in_len + 1);
- }
-
- if (need_unescape) {
- if (!unescape_squote) {
- ret = ucl_unescape_json_string (*dst, ret);
- }
- else {
- ret = ucl_unescape_squoted_string (*dst, ret);
- }
- }
-
- if (need_expand) {
- tmp = *dst;
- tret = ret;
- ret = ucl_expand_variable (parser, dst, tmp, ret);
- if (*dst == NULL) {
- /* Nothing to expand */
- *dst = tmp;
- ret = tret;
- }
- else {
- /* Free unexpanded value */
- UCL_FREE (in_len + 1, tmp);
- }
- }
- *dst_const = *dst;
- }
- else {
- *dst_const = src;
- ret = in_len;
- }
-
- return ret;
- }
-
- /**
- * Create and append an object at the specified level
- * @param parser
- * @param is_array
- * @param level
- * @return
- */
- static inline ucl_object_t *
- ucl_parser_add_container (ucl_object_t *obj, struct ucl_parser *parser,
- bool is_array, uint32_t level, bool has_obrace)
- {
- struct ucl_stack *st;
- ucl_object_t *nobj;
-
- if (obj == NULL) {
- nobj = ucl_object_new_full (is_array ? UCL_ARRAY : UCL_OBJECT, parser->chunks->priority);
- if (nobj == NULL) {
- goto enomem0;
- }
- } else {
- if (obj->type == (is_array ? UCL_OBJECT : UCL_ARRAY)) {
- /* Bad combination for merge: array and object */
- ucl_set_err (parser, UCL_EMERGE,
- "cannot merge an object with an array",
- &parser->err);
-
- return NULL;
- }
- nobj = obj;
- nobj->type = is_array ? UCL_ARRAY : UCL_OBJECT;
- }
-
- if (!is_array) {
- if (nobj->value.ov == NULL) {
- nobj->value.ov = ucl_hash_create (parser->flags & UCL_PARSER_KEY_LOWERCASE);
- if (nobj->value.ov == NULL) {
- goto enomem1;
- }
- }
- parser->state = UCL_STATE_KEY;
- } else {
- parser->state = UCL_STATE_VALUE;
- }
-
- st = UCL_ALLOC (sizeof (struct ucl_stack));
-
- if (st == NULL) {
- goto enomem1;
- }
-
- st->obj = nobj;
-
- if (level >= UINT16_MAX) {
- ucl_set_err (parser, UCL_ENESTED,
- "objects are nesting too deep (over 65535 limit)",
- &parser->err);
- if (nobj != obj) {
- ucl_object_unref (obj);
- }
-
- UCL_FREE(sizeof (struct ucl_stack), st);
-
- return NULL;
- }
-
-
- st->e.params.level = level;
- st->e.params.line = parser->chunks->line;
- st->chunk = parser->chunks;
-
- if (has_obrace) {
- st->e.params.flags = UCL_STACK_HAS_OBRACE;
- }
- else {
- st->e.params.flags = 0;
- }
-
- LL_PREPEND (parser->stack, st);
- parser->cur_obj = nobj;
-
- return nobj;
- enomem1:
- if (nobj != obj)
- ucl_object_unref (nobj);
- enomem0:
- ucl_set_err (parser, UCL_EINTERNAL, "cannot allocate memory for an object",
- &parser->err);
- return NULL;
- }
-
- int
- ucl_maybe_parse_number (ucl_object_t *obj,
- const char *start, const char *end, const char **pos,
- bool allow_double, bool number_bytes, bool allow_time)
- {
- const char *p = start, *c = start;
- char *endptr;
- bool got_dot = false, got_exp = false, need_double = false,
- is_time = false, valid_start = false, is_hex = false;
- int is_neg = 0;
- double dv = 0;
- int64_t lv = 0;
-
- if (*p == '-') {
- is_neg = 1;
- c ++;
- p ++;
- }
- while (p < end) {
- if (is_hex && isxdigit (*p)) {
- p ++;
- }
- else if (isdigit (*p)) {
- valid_start = true;
- p ++;
- }
- else if (!is_hex && (*p == 'x' || *p == 'X')) {
- is_hex = true;
- allow_double = false;
- c = p + 1;
- p ++;
- }
- else if (allow_double) {
- if (p == c) {
- /* Empty digits sequence, not a number */
- *pos = start;
- return EINVAL;
- }
- else if (*p == '.') {
- if (got_dot) {
- /* Double dots, not a number */
- *pos = start;
- return EINVAL;
- }
- else {
- got_dot = true;
- need_double = true;
- p ++;
- }
- }
- else if (*p == 'e' || *p == 'E') {
- if (got_exp) {
- /* Double exp, not a number */
- *pos = start;
- return EINVAL;
- }
- else {
- got_exp = true;
- need_double = true;
- p ++;
- if (p >= end) {
- *pos = start;
- return EINVAL;
- }
- if (!isdigit (*p) && *p != '+' && *p != '-') {
- /* Wrong exponent sign */
- *pos = start;
- return EINVAL;
- }
- else {
- p ++;
- }
- }
- }
- else {
- /* Got the end of the number, need to check */
- break;
- }
- }
- else if (!allow_double && *p == '.') {
- /* Unexpected dot */
- *pos = start;
- return EINVAL;
- }
- else {
- break;
- }
- }
-
- if (!valid_start || p == c) {
- *pos = start;
- return EINVAL;
- }
-
- char numbuf[128];
-
- if ((size_t)(p - c + 1) >= sizeof(numbuf)) {
- *pos = start;
- return EINVAL;
- }
-
- if (is_neg) {
- numbuf[0] = '-';
- ucl_strlcpy (&numbuf[1], c, p - c + 1);
- }
- else {
- ucl_strlcpy (numbuf, c, p - c + 1);
- }
-
- errno = 0;
- if (need_double) {
- dv = strtod (numbuf, &endptr);
- }
- else {
- if (is_hex) {
- lv = strtoimax (numbuf, &endptr, 16);
- }
- else {
- lv = strtoimax (numbuf, &endptr, 10);
- }
- }
- if (errno == ERANGE) {
- *pos = start;
- return ERANGE;
- }
-
- /* Now check endptr and move it from numbuf to the real ending */
- if (endptr != NULL) {
- long shift = endptr - numbuf - is_neg;
- endptr = (char *)c + shift;
- }
- if (endptr >= end) {
- p = end;
- goto set_obj;
- }
- if (endptr == NULL || ucl_lex_is_atom_end (*endptr) || *endptr == '\0') {
- p = endptr;
- goto set_obj;
- }
-
- if (endptr < end && endptr != start) {
- switch (*p) {
- case 'm':
- case 'M':
- case 'g':
- case 'G':
- case 'k':
- case 'K':
- if (end - p >= 2) {
- if (p[1] == 's' || p[1] == 'S') {
- /* Milliseconds */
- if (!need_double) {
- need_double = true;
- dv = lv;
- }
- is_time = true;
- if (p[0] == 'm' || p[0] == 'M') {
- dv /= 1000.;
- }
- else {
- dv *= ucl_lex_num_multiplier (*p, false);
- }
- p += 2;
- if (end - p > 0 && !ucl_lex_is_atom_end (*p)) {
- *pos = start;
- return EINVAL;
- }
- goto set_obj;
- }
- else if (number_bytes || (p[1] == 'b' || p[1] == 'B')) {
- /* Bytes */
- if (need_double) {
- need_double = false;
- lv = dv;
- }
- lv *= ucl_lex_num_multiplier (*p, true);
- p += 2;
- if (end - p > 0 && !ucl_lex_is_atom_end (*p)) {
- *pos = start;
- return EINVAL;
- }
- goto set_obj;
- }
- else if (ucl_lex_is_atom_end (p[1])) {
- if (need_double) {
- dv *= ucl_lex_num_multiplier (*p, false);
- }
- else {
- lv *= ucl_lex_num_multiplier (*p, number_bytes);
- }
- p ++;
- goto set_obj;
- }
- else if (allow_time && end - p >= 3) {
- if (tolower (p[0]) == 'm' &&
- tolower (p[1]) == 'i' &&
- tolower (p[2]) == 'n') {
- /* Minutes */
- if (!need_double) {
- need_double = true;
- dv = lv;
- }
- is_time = true;
- dv *= 60.;
- p += 3;
- if (end - p > 0 && !ucl_lex_is_atom_end (*p)) {
- *pos = start;
- return EINVAL;
- }
- goto set_obj;
- }
- }
- }
- else {
- if (need_double) {
- dv *= ucl_lex_num_multiplier (*p, false);
- }
- else {
- lv *= ucl_lex_num_multiplier (*p, number_bytes);
- }
- p ++;
- if (end - p > 0 && !ucl_lex_is_atom_end (*p)) {
- *pos = start;
- return EINVAL;
- }
- goto set_obj;
- }
- break;
- case 'S':
- case 's':
- if (allow_time &&
- (p == end - 1 || ucl_lex_is_atom_end (p[1]))) {
- if (!need_double) {
- need_double = true;
- dv = lv;
- }
- p ++;
- is_time = true;
- goto set_obj;
- }
- break;
- case 'h':
- case 'H':
- case 'd':
- case 'D':
- case 'w':
- case 'W':
- case 'Y':
- case 'y':
- if (allow_time &&
- (p == end - 1 || ucl_lex_is_atom_end (p[1]))) {
- if (!need_double) {
- need_double = true;
- dv = lv;
- }
- is_time = true;
- dv *= ucl_lex_time_multiplier (*p);
- p ++;
- goto set_obj;
- }
- break;
- case '\t':
- case ' ':
- while (p < end && ucl_test_character(*p, UCL_CHARACTER_WHITESPACE)) {
- p++;
- }
- if (ucl_lex_is_atom_end(*p))
- goto set_obj;
- break;
- }
- }
- else if (endptr == end) {
- /* Just a number at the end of chunk */
- p = end;
- goto set_obj;
- }
-
- *pos = c;
- return EINVAL;
-
- set_obj:
- if (obj != NULL) {
- if (allow_double && (need_double || is_time)) {
- if (!is_time) {
- obj->type = UCL_FLOAT;
- }
- else {
- obj->type = UCL_TIME;
- }
- obj->value.dv = dv;
- }
- else {
- obj->type = UCL_INT;
- obj->value.iv = lv;
- }
- }
- *pos = p;
- return 0;
- }
-
- /**
- * Parse possible number
- * @param parser
- * @param chunk
- * @param obj
- * @return true if a number has been parsed
- */
- static bool
- ucl_lex_number (struct ucl_parser *parser,
- struct ucl_chunk *chunk, ucl_object_t *obj)
- {
- const unsigned char *pos;
- int ret;
-
- ret = ucl_maybe_parse_number (obj, chunk->pos, chunk->end, (const char **)&pos,
- true, false, ((parser->flags & UCL_PARSER_NO_TIME) == 0));
-
- if (ret == 0) {
- chunk->remain -= pos - chunk->pos;
- chunk->column += pos - chunk->pos;
- chunk->pos = pos;
- return true;
- }
- else if (ret == ERANGE) {
- ucl_set_err (parser, UCL_ESYNTAX, "numeric value out of range",
- &parser->err);
- }
-
- return false;
- }
-
- /**
- * Parse quoted string with possible escapes
- * @param parser
- * @param chunk
- * @param need_unescape
- * @param ucl_escape
- * @param var_expand
- * @return true if a string has been parsed
- */
- static bool
- ucl_lex_json_string (struct ucl_parser *parser,
- struct ucl_chunk *chunk,
- bool *need_unescape,
- bool *ucl_escape,
- bool *var_expand)
- {
- const unsigned char *p = chunk->pos;
- unsigned char c;
- int i;
-
- while (p < chunk->end) {
- c = *p;
- if (c < 0x1F) {
- /* Unmasked control character */
- if (c == '\n') {
- ucl_set_err (parser, UCL_ESYNTAX, "unexpected newline",
- &parser->err);
- }
- else {
- ucl_set_err (parser, UCL_ESYNTAX, "unexpected control character",
- &parser->err);
- }
- return false;
- }
- else if (c == '\\') {
- ucl_chunk_skipc (chunk, p);
- if (p >= chunk->end) {
- ucl_set_err (parser, UCL_ESYNTAX, "unfinished escape character",
- &parser->err);
- return false;
- }
- c = *p;
- if (ucl_test_character (c, UCL_CHARACTER_ESCAPE)) {
- if (c == 'u') {
- ucl_chunk_skipc (chunk, p);
- for (i = 0; i < 4 && p < chunk->end; i ++) {
- if (!isxdigit (*p)) {
- ucl_set_err (parser, UCL_ESYNTAX, "invalid utf escape",
- &parser->err);
- return false;
- }
- ucl_chunk_skipc (chunk, p);
- }
- if (p >= chunk->end) {
- ucl_set_err (parser, UCL_ESYNTAX,
- "unfinished escape character",
- &parser->err);
- return false;
- }
- }
- else {
- ucl_chunk_skipc (chunk, p);
- }
- }
- *need_unescape = true;
- *ucl_escape = true;
- continue;
- }
- else if (c == '"') {
- ucl_chunk_skipc (chunk, p);
- return true;
- }
- else if (ucl_test_character (c, UCL_CHARACTER_UCL_UNSAFE)) {
- *ucl_escape = true;
- }
- else if (c == '$') {
- *var_expand = true;
- }
- ucl_chunk_skipc (chunk, p);
- }
-
- ucl_set_err (parser, UCL_ESYNTAX,
- "no quote at the end of json string",
- &parser->err);
- return false;
- }
-
- /**
- * Process single quoted string
- * @param parser
- * @param chunk
- * @param need_unescape
- * @return
- */
- static bool
- ucl_lex_squoted_string (struct ucl_parser *parser,
- struct ucl_chunk *chunk, bool *need_unescape)
- {
- const unsigned char *p = chunk->pos;
- unsigned char c;
-
- while (p < chunk->end) {
- c = *p;
- if (c == '\\') {
- ucl_chunk_skipc (chunk, p);
-
- if (p >= chunk->end) {
- ucl_set_err (parser, UCL_ESYNTAX,
- "unfinished escape character",
- &parser->err);
- return false;
- }
- else {
- ucl_chunk_skipc (chunk, p);
- }
-
- *need_unescape = true;
- continue;
- }
- else if (c == '\'') {
- ucl_chunk_skipc (chunk, p);
- return true;
- }
-
- ucl_chunk_skipc (chunk, p);
- }
-
- ucl_set_err (parser, UCL_ESYNTAX,
- "no quote at the end of single quoted string",
- &parser->err);
- return false;
- }
-
- static void
- ucl_parser_append_elt (struct ucl_parser *parser, ucl_hash_t *cont,
- ucl_object_t *top,
- ucl_object_t *elt)
- {
- ucl_object_t *nobj;
-
- if ((parser->flags & UCL_PARSER_NO_IMPLICIT_ARRAYS) == 0) {
- /* Implicit array */
- top->flags |= UCL_OBJECT_MULTIVALUE;
- DL_APPEND (top, elt);
- parser->stack->obj->len ++;
- }
- else {
- if ((top->flags & UCL_OBJECT_MULTIVALUE) != 0) {
- /* Just add to the explicit array */
- ucl_array_append (top, elt);
- }
- else {
- /* Convert to an array */
- nobj = ucl_object_typed_new (UCL_ARRAY);
- nobj->key = top->key;
- nobj->keylen = top->keylen;
- nobj->flags |= UCL_OBJECT_MULTIVALUE;
- ucl_array_append (nobj, top);
- ucl_array_append (nobj, elt);
- ucl_hash_replace (cont, top, nobj);
- }
- }
- }
-
- bool
- ucl_parser_process_object_element (struct ucl_parser *parser, ucl_object_t *nobj)
- {
- ucl_hash_t *container;
- ucl_object_t *tobj = NULL, *cur;
- char errmsg[256];
-
- container = parser->stack->obj->value.ov;
-
- DL_FOREACH (parser->stack->obj, cur) {
- tobj = __DECONST (ucl_object_t *, ucl_hash_search_obj (cur->value.ov, nobj));
-
- if (tobj != NULL) {
- break;
- }
- }
-
-
- if (tobj == NULL) {
- container = ucl_hash_insert_object (container, nobj,
- parser->flags & UCL_PARSER_KEY_LOWERCASE);
- if (container == NULL) {
- return false;
- }
- nobj->prev = nobj;
- nobj->next = NULL;
- parser->stack->obj->len ++;
- }
- else {
- unsigned priold = ucl_object_get_priority (tobj),
- prinew = ucl_object_get_priority (nobj);
- switch (parser->chunks->strategy) {
-
- case UCL_DUPLICATE_APPEND:
- /*
- * The logic here is the following:
- *
- * - if we have two objects with the same priority, then we form an
- * implicit or explicit array
- * - if a new object has bigger priority, then we overwrite an old one
- * - if a new object has lower priority, then we ignore it
- */
- /* Special case for inherited objects */
- if (tobj->flags & UCL_OBJECT_INHERITED) {
- prinew = priold + 1;
- }
-
- if (priold == prinew) {
- ucl_parser_append_elt (parser, container, tobj, nobj);
- }
- else if (priold > prinew) {
- /*
- * We add this new object to a list of trash objects just to ensure
- * that it won't come to any real object
- * XXX: rather inefficient approach
- */
- DL_APPEND (parser->trash_objs, nobj);
- }
- else {
- ucl_hash_replace (container, tobj, nobj);
- ucl_object_unref (tobj);
- }
-
- break;
-
- case UCL_DUPLICATE_REWRITE:
- /* We just rewrite old values regardless of priority */
- ucl_hash_replace (container, tobj, nobj);
- ucl_object_unref (tobj);
-
- break;
-
- case UCL_DUPLICATE_ERROR:
- snprintf(errmsg, sizeof(errmsg),
- "duplicate element for key '%s' found",
- nobj->key);
- ucl_set_err (parser, UCL_EMERGE, errmsg, &parser->err);
- return false;
-
- case UCL_DUPLICATE_MERGE:
- /*
- * Here we do have some old object so we just push it on top of objects stack
- * Check priority and then perform the merge on the remaining objects
- */
- if (tobj->type == UCL_OBJECT || tobj->type == UCL_ARRAY) {
- ucl_object_unref (nobj);
- nobj = tobj;
- }
- else if (priold == prinew) {
- ucl_parser_append_elt (parser, container, tobj, nobj);
- }
- else if (priold > prinew) {
- /*
- * We add this new object to a list of trash objects just to ensure
- * that it won't come to any real object
- * XXX: rather inefficient approach
- */
- DL_APPEND (parser->trash_objs, nobj);
- }
- else {
- ucl_hash_replace (container, tobj, nobj);
- ucl_object_unref (tobj);
- }
- break;
- }
- }
-
- parser->stack->obj->value.ov = container;
- parser->cur_obj = nobj;
- ucl_attach_comment (parser, nobj, false);
-
- return true;
- }
-
- /**
- * Parse a key in an object
- * @param parser
- * @param chunk
- * @param next_key
- * @param end_of_object
- * @return true if a key has been parsed
- */
- static bool
- ucl_parse_key (struct ucl_parser *parser, struct ucl_chunk *chunk,
- bool *next_key, bool *end_of_object)
- {
- const unsigned char *p, *c = NULL, *end, *t;
- const char *key = NULL;
- bool got_quote = false, got_eq = false, got_semicolon = false,
- need_unescape = false, ucl_escape = false, var_expand = false,
- got_content = false, got_sep = false;
- ucl_object_t *nobj;
- ssize_t keylen;
-
- p = chunk->pos;
-
- if (*p == '.') {
- /* It is macro actually */
- if (!(parser->flags & UCL_PARSER_DISABLE_MACRO)) {
- ucl_chunk_skipc (chunk, p);
- }
-
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_MACRO_NAME;
- *end_of_object = false;
- return true;
- }
- while (p < chunk->end) {
- /*
- * A key must start with alpha, number, '/' or '_' and end with space character
- */
- if (c == NULL) {
- if (chunk->remain >= 2 && ucl_lex_is_comment (p[0], p[1])) {
- if (!ucl_skip_comments (parser)) {
- return false;
- }
- p = chunk->pos;
- }
- else if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- ucl_chunk_skipc (chunk, p);
- }
- else if (ucl_test_character (*p, UCL_CHARACTER_KEY_START)) {
- /* The first symbol */
- c = p;
- ucl_chunk_skipc (chunk, p);
- got_content = true;
- }
- else if (*p == '"') {
- /* JSON style key */
- c = p + 1;
- got_quote = true;
- got_content = true;
- ucl_chunk_skipc (chunk, p);
- }
- else if (*p == '}') {
- /* We have actually end of an object */
- *end_of_object = true;
- return true;
- }
- else if (*p == '.') {
- ucl_chunk_skipc (chunk, p);
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_MACRO_NAME;
- return true;
- }
- else {
- /* Invalid identifier */
- ucl_set_err (parser, UCL_ESYNTAX, "key must begin with a letter",
- &parser->err);
- return false;
- }
- }
- else {
- /* Parse the body of a key */
- if (!got_quote) {
- if (ucl_test_character (*p, UCL_CHARACTER_KEY)) {
- got_content = true;
- ucl_chunk_skipc (chunk, p);
- }
- else if (ucl_test_character (*p, UCL_CHARACTER_KEY_SEP)) {
- end = p;
- break;
- }
- else {
- ucl_set_err (parser, UCL_ESYNTAX, "invalid character in a key",
- &parser->err);
- return false;
- }
- }
- else {
- /* We need to parse json like quoted string */
- if (!ucl_lex_json_string (parser, chunk, &need_unescape, &ucl_escape, &var_expand)) {
- return false;
- }
- /* Always escape keys obtained via json */
- end = chunk->pos - 1;
- p = chunk->pos;
- break;
- }
- }
- }
-
- if (p >= chunk->end && got_content) {
- ucl_set_err (parser, UCL_ESYNTAX, "unfinished key", &parser->err);
- return false;
- }
- else if (!got_content) {
- return true;
- }
- *end_of_object = false;
- /* We are now at the end of the key, need to parse the rest */
- while (p < chunk->end) {
- if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE)) {
- ucl_chunk_skipc (chunk, p);
- }
- else if (*p == '=') {
- if (!got_eq && !got_semicolon) {
- ucl_chunk_skipc (chunk, p);
- got_eq = true;
- }
- else {
- ucl_set_err (parser, UCL_ESYNTAX, "unexpected '=' character",
- &parser->err);
- return false;
- }
- }
- else if (*p == ':') {
- if (!got_eq && !got_semicolon) {
- ucl_chunk_skipc (chunk, p);
- got_semicolon = true;
- }
- else {
- ucl_set_err (parser, UCL_ESYNTAX, "unexpected ':' character",
- &parser->err);
- return false;
- }
- }
- else if (chunk->remain >= 2 && ucl_lex_is_comment (p[0], p[1])) {
- /* Check for comment */
- if (!ucl_skip_comments (parser)) {
- return false;
- }
- p = chunk->pos;
- }
- else {
- /* Start value */
- break;
- }
- }
-
- if (p >= chunk->end && got_content) {
- ucl_set_err (parser, UCL_ESYNTAX, "unfinished key", &parser->err);
- return false;
- }
-
- got_sep = got_semicolon || got_eq;
-
- if (!got_sep) {
- /*
- * Maybe we have more keys nested, so search for termination character.
- * Possible choices:
- * 1) key1 key2 ... keyN [:=] value <- we treat that as error
- * 2) key1 ... keyN {} or [] <- we treat that as nested objects
- * 3) key1 value[;,\n] <- we treat that as linear object
- */
- t = p;
- *next_key = false;
- while (ucl_test_character (*t, UCL_CHARACTER_WHITESPACE)) {
- t ++;
- }
- /* Check first non-space character after a key */
- if (*t != '{' && *t != '[') {
- while (t < chunk->end) {
- if (*t == ',' || *t == ';' || *t == '\n' || *t == '\r') {
- break;
- }
- else if (*t == '{' || *t == '[') {
- *next_key = true;
- break;
- }
- t ++;
- }
- }
- }
-
- /* Create a new object */
- nobj = ucl_object_new_full (UCL_NULL, parser->chunks->priority);
- if (nobj == NULL) {
- return false;
- }
- keylen = ucl_copy_or_store_ptr (parser, c, &nobj->trash_stack[UCL_TRASH_KEY],
- &key, end - c, need_unescape, parser->flags & UCL_PARSER_KEY_LOWERCASE,
- false, false);
- if (keylen == -1) {
- ucl_object_unref (nobj);
- return false;
- }
- else if (keylen == 0) {
- ucl_set_err (parser, UCL_ESYNTAX, "empty keys are not allowed", &parser->err);
- ucl_object_unref (nobj);
- return false;
- }
-
- nobj->key = key;
- nobj->keylen = keylen;
-
- if (!ucl_parser_process_object_element (parser, nobj)) {
- return false;
- }
-
- if (ucl_escape) {
- nobj->flags |= UCL_OBJECT_NEED_KEY_ESCAPE;
- }
-
-
- return true;
- }
-
- /**
- * Parse a cl string
- * @param parser
- * @param chunk
- * @param var_expand
- * @param need_unescape
- * @return true if a key has been parsed
- */
- static bool
- ucl_parse_string_value (struct ucl_parser *parser,
- struct ucl_chunk *chunk, bool *var_expand, bool *need_unescape)
- {
- const unsigned char *p;
- enum {
- UCL_BRACE_ROUND = 0,
- UCL_BRACE_SQUARE,
- UCL_BRACE_FIGURE
- };
- int braces[3][2] = {{0, 0}, {0, 0}, {0, 0}};
-
- p = chunk->pos;
-
- while (p < chunk->end) {
-
- /* Skip pairs of figure braces */
- if (*p == '{') {
- braces[UCL_BRACE_FIGURE][0] ++;
- }
- else if (*p == '}') {
- braces[UCL_BRACE_FIGURE][1] ++;
- if (braces[UCL_BRACE_FIGURE][1] <= braces[UCL_BRACE_FIGURE][0]) {
- /* This is not a termination symbol, continue */
- ucl_chunk_skipc (chunk, p);
- continue;
- }
- }
- /* Skip pairs of square braces */
- else if (*p == '[') {
- braces[UCL_BRACE_SQUARE][0] ++;
- }
- else if (*p == ']') {
- braces[UCL_BRACE_SQUARE][1] ++;
- if (braces[UCL_BRACE_SQUARE][1] <= braces[UCL_BRACE_SQUARE][0]) {
- /* This is not a termination symbol, continue */
- ucl_chunk_skipc (chunk, p);
- continue;
- }
- }
- else if (*p == '$') {
- *var_expand = true;
- }
- else if (*p == '\\') {
- *need_unescape = true;
- ucl_chunk_skipc (chunk, p);
- if (p < chunk->end) {
- ucl_chunk_skipc (chunk, p);
- }
- continue;
- }
-
- if (ucl_lex_is_atom_end (*p) || (chunk->remain >= 2 && ucl_lex_is_comment (p[0], p[1]))) {
- break;
- }
- ucl_chunk_skipc (chunk, p);
- }
-
- return true;
- }
-
- /**
- * Parse multiline string ending with \n{term}\n
- * @param parser
- * @param chunk
- * @param term
- * @param term_len
- * @param beg
- * @param var_expand
- * @return size of multiline string or 0 in case of error
- */
- static int
- ucl_parse_multiline_string (struct ucl_parser *parser,
- struct ucl_chunk *chunk, const unsigned char *term,
- int term_len, unsigned char const **beg,
- bool *var_expand)
- {
- const unsigned char *p, *c, *tend;
- bool newline = false;
- int len = 0;
-
- p = chunk->pos;
-
- c = p;
-
- while (p < chunk->end) {
- if (newline) {
- if (chunk->end - p < term_len) {
- return 0;
- }
- else if (memcmp (p, term, term_len) == 0) {
- tend = p + term_len;
- if (*tend != '\n' && *tend != ';' && *tend != ',') {
- /* Incomplete terminator */
- ucl_chunk_skipc (chunk, p);
- continue;
- }
- len = p - c;
- chunk->remain -= term_len;
- chunk->pos = p + term_len;
- chunk->column = term_len;
- *beg = c;
- break;
- }
- }
- if (*p == '\n') {
- newline = true;
- }
- else {
- if (*p == '$') {
- *var_expand = true;
- }
- newline = false;
- }
- ucl_chunk_skipc (chunk, p);
- }
-
- return len;
- }
-
- static inline ucl_object_t*
- ucl_parser_get_container (struct ucl_parser *parser)
- {
- ucl_object_t *t, *obj = NULL;
-
- if (parser == NULL || parser->stack == NULL || parser->stack->obj == NULL) {
- return NULL;
- }
-
- if (parser->stack->obj->type == UCL_ARRAY) {
- /* Object must be allocated */
- obj = ucl_object_new_full (UCL_NULL, parser->chunks->priority);
- t = parser->stack->obj;
-
- if (!ucl_array_append (t, obj)) {
- ucl_object_unref (obj);
- return NULL;
- }
-
- parser->cur_obj = obj;
- ucl_attach_comment (parser, obj, false);
- }
- else {
- /* Object has been already allocated */
- obj = parser->cur_obj;
- }
-
- return obj;
- }
-
- /**
- * Handle value data
- * @param parser
- * @param chunk
- * @return
- */
- static bool
- ucl_parse_value (struct ucl_parser *parser, struct ucl_chunk *chunk)
- {
- const unsigned char *p, *c;
- ucl_object_t *obj = NULL;
- unsigned int stripped_spaces;
- ssize_t str_len;
- bool need_unescape = false, ucl_escape = false, var_expand = false;
-
- p = chunk->pos;
-
- /* Skip any spaces and comments */
- if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE) ||
- (chunk->remain >= 2 && ucl_lex_is_comment (p[0], p[1]))) {
- while (p < chunk->end && ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- ucl_chunk_skipc (chunk, p);
- }
- if (!ucl_skip_comments (parser)) {
- return false;
- }
- p = chunk->pos;
- }
-
- while (p < chunk->end) {
- c = p;
- switch (*p) {
- case '"':
- ucl_chunk_skipc (chunk, p);
-
- if (!ucl_lex_json_string (parser, chunk, &need_unescape, &ucl_escape,
- &var_expand)) {
- return false;
- }
-
- obj = ucl_parser_get_container (parser);
- if (!obj) {
- return false;
- }
-
- str_len = chunk->pos - c - 2;
- obj->type = UCL_STRING;
- if ((str_len = ucl_copy_or_store_ptr (parser, c + 1,
- &obj->trash_stack[UCL_TRASH_VALUE],
- &obj->value.sv, str_len, need_unescape, false,
- var_expand, false)) == -1) {
- return false;
- }
-
- obj->len = str_len;
- parser->state = UCL_STATE_AFTER_VALUE;
-
- return true;
- break;
- case '\'':
- ucl_chunk_skipc (chunk, p);
-
- if (!ucl_lex_squoted_string (parser, chunk, &need_unescape)) {
- return false;
- }
-
- obj = ucl_parser_get_container (parser);
- if (!obj) {
- return false;
- }
-
- str_len = chunk->pos - c - 2;
- obj->type = UCL_STRING;
- obj->flags |= UCL_OBJECT_SQUOTED;
-
- if ((str_len = ucl_copy_or_store_ptr (parser, c + 1,
- &obj->trash_stack[UCL_TRASH_VALUE],
- &obj->value.sv, str_len, need_unescape, false,
- var_expand, true)) == -1) {
- return false;
- }
-
- obj->len = str_len;
-
- parser->state = UCL_STATE_AFTER_VALUE;
-
- return true;
- break;
- case '{':
- obj = ucl_parser_get_container (parser);
- if (obj == NULL) {
- return false;
- }
- /* We have a new object */
- if (parser->stack) {
- obj = ucl_parser_add_container (obj, parser, false,
- parser->stack->e.params.level, true);
- }
- else {
- return false;
- }
- if (obj == NULL) {
- return false;
- }
-
- ucl_chunk_skipc (chunk, p);
-
- return true;
- break;
- case '[':
- obj = ucl_parser_get_container (parser);
- if (obj == NULL) {
- return false;
- }
- /* We have a new array */
- if (parser->stack) {
- obj = ucl_parser_add_container (obj, parser, true,
- parser->stack->e.params.level, true);
- }
- else {
- return false;
- }
-
- if (obj == NULL) {
- return false;
- }
-
- ucl_chunk_skipc (chunk, p);
-
- return true;
- break;
- case ']':
- /* We have the array ending */
- if (parser->stack && parser->stack->obj->type == UCL_ARRAY) {
- parser->state = UCL_STATE_AFTER_VALUE;
- return true;
- }
- else {
- goto parse_string;
- }
- break;
- case '<':
- obj = ucl_parser_get_container (parser);
- /* We have something like multiline value, which must be <<[A-Z]+\n */
- if (chunk->end - p > 3) {
- if (memcmp (p, "<<", 2) == 0) {
- p += 2;
- /* We allow only uppercase characters in multiline definitions */
- while (p < chunk->end && *p >= 'A' && *p <= 'Z') {
- p ++;
- }
- if(p == chunk->end) {
- ucl_set_err (parser, UCL_ESYNTAX,
- "unterminated multiline value", &parser->err);
- return false;
- }
- if (*p =='\n') {
- /* Set chunk positions and start multiline parsing */
- chunk->remain -= p - c + 1;
- c += 2;
- chunk->pos = p + 1;
- chunk->column = 0;
- chunk->line ++;
- if ((str_len = ucl_parse_multiline_string (parser, chunk, c,
- p - c, &c, &var_expand)) == 0) {
- ucl_set_err (parser, UCL_ESYNTAX,
- "unterminated multiline value", &parser->err);
- return false;
- }
-
- obj->type = UCL_STRING;
- obj->flags |= UCL_OBJECT_MULTILINE;
- if ((str_len = ucl_copy_or_store_ptr (parser, c,
- &obj->trash_stack[UCL_TRASH_VALUE],
- &obj->value.sv, str_len - 1, false,
- false, var_expand, false)) == -1) {
- return false;
- }
- obj->len = str_len;
-
- parser->state = UCL_STATE_AFTER_VALUE;
-
- return true;
- }
- }
- }
- /* Fallback to ordinary strings */
- /* FALLTHRU */
- default:
- parse_string:
- if (obj == NULL) {
- obj = ucl_parser_get_container (parser);
- }
-
- /* Parse atom */
- if (ucl_test_character (*p, UCL_CHARACTER_VALUE_DIGIT_START)) {
- if (!ucl_lex_number (parser, chunk, obj)) {
- if (parser->state == UCL_STATE_ERROR) {
- return false;
- }
- }
- else {
- parser->state = UCL_STATE_AFTER_VALUE;
- return true;
- }
- /* Fallback to normal string */
- }
-
- if (!ucl_parse_string_value (parser, chunk, &var_expand,
- &need_unescape)) {
- return false;
- }
- /* Cut trailing spaces */
- stripped_spaces = 0;
- while (ucl_test_character (*(chunk->pos - 1 - stripped_spaces),
- UCL_CHARACTER_WHITESPACE)) {
- stripped_spaces ++;
- }
- str_len = chunk->pos - c - stripped_spaces;
- if (str_len <= 0) {
- ucl_set_err (parser, UCL_ESYNTAX, "string value must not be empty",
- &parser->err);
- return false;
- }
- else if (str_len == 4 && memcmp (c, "null", 4) == 0) {
- obj->len = 0;
- obj->type = UCL_NULL;
- }
- else if (str_len == 3 && memcmp (c, "nan", 3) == 0) {
- obj->len = 0;
- obj->type = UCL_FLOAT;
- obj->value.dv = NAN;
- }
- else if (str_len == 3 && memcmp (c, "inf", 3) == 0) {
- obj->len = 0;
- obj->type = UCL_FLOAT;
- obj->value.dv = INFINITY;
- }
- else if (!ucl_maybe_parse_boolean (obj, c, str_len)) {
- obj->type = UCL_STRING;
- if ((str_len = ucl_copy_or_store_ptr (parser, c,
- &obj->trash_stack[UCL_TRASH_VALUE],
- &obj->value.sv, str_len, need_unescape,
- false, var_expand, false)) == -1) {
- return false;
- }
- obj->len = str_len;
- }
-
- parser->state = UCL_STATE_AFTER_VALUE;
-
- return true;
- break;
- }
- }
-
- return true;
- }
-
- /**
- * Handle after value data
- * @param parser
- * @param chunk
- * @return
- */
- static bool
- ucl_parse_after_value (struct ucl_parser *parser, struct ucl_chunk *chunk)
- {
- const unsigned char *p;
- bool got_sep = false;
- struct ucl_stack *st;
-
- p = chunk->pos;
-
- while (p < chunk->end) {
- if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE)) {
- /* Skip whitespaces */
- ucl_chunk_skipc (chunk, p);
- }
- else if (chunk->remain >= 2 && ucl_lex_is_comment (p[0], p[1])) {
- /* Skip comment */
- if (!ucl_skip_comments (parser)) {
- return false;
- }
- /* Treat comment as a separator */
- got_sep = true;
- p = chunk->pos;
- }
- else if (ucl_test_character (*p, UCL_CHARACTER_VALUE_END)) {
- if (*p == '}' || *p == ']') {
- if (parser->stack == NULL) {
- ucl_set_err (parser, UCL_ESYNTAX,
- "end of array or object detected without corresponding start",
- &parser->err);
- return false;
- }
- if ((*p == '}' && parser->stack->obj->type == UCL_OBJECT) ||
- (*p == ']' && parser->stack->obj->type == UCL_ARRAY)) {
-
- /* Pop all nested objects from a stack */
- st = parser->stack;
-
- if (!(st->e.params.flags & UCL_STACK_HAS_OBRACE)) {
- parser->err_code = UCL_EUNPAIRED;
- ucl_create_err (&parser->err,
- "%s:%d object closed with } is not opened with { at line %d",
- chunk->fname ? chunk->fname : "memory",
- parser->chunks->line, st->e.params.line);
-
- return false;
- }
-
- parser->stack = st->next;
- UCL_FREE (sizeof (struct ucl_stack), st);
-
- if (parser->cur_obj) {
- ucl_attach_comment (parser, parser->cur_obj, true);
- }
-
- while (parser->stack != NULL) {
- st = parser->stack;
-
- if (st->next == NULL) {
- break;
- }
- else if (st->next->e.params.level == st->e.params.level) {
- break;
- }
-
-
- parser->stack = st->next;
- parser->cur_obj = st->obj;
- UCL_FREE (sizeof (struct ucl_stack), st);
- }
- }
- else {
- ucl_set_err (parser, UCL_ESYNTAX,
- "unexpected terminating symbol detected",
- &parser->err);
- return false;
- }
-
- if (parser->stack == NULL) {
- /* Ignore everything after a top object */
- return true;
- }
- else {
- ucl_chunk_skipc (chunk, p);
- }
- got_sep = true;
- }
- else {
- /* Got a separator */
- got_sep = true;
- ucl_chunk_skipc (chunk, p);
- }
- }
- else {
- /* Anything else */
- if (!got_sep) {
- ucl_set_err (parser, UCL_ESYNTAX, "delimiter is missing",
- &parser->err);
- return false;
- }
- return true;
- }
- }
-
- return true;
- }
-
- static bool
- ucl_skip_macro_as_comment (struct ucl_parser *parser,
- struct ucl_chunk *chunk)
- {
- const unsigned char *p, *c;
- enum {
- macro_skip_start = 0,
- macro_has_symbols,
- macro_has_obrace,
- macro_has_quote,
- macro_has_backslash,
- macro_has_sqbrace,
- macro_save
- } state = macro_skip_start, prev_state = macro_skip_start;
-
- p = chunk->pos;
- c = chunk->pos;
-
- while (p < chunk->end) {
- switch (state) {
- case macro_skip_start:
- if (!ucl_test_character (*p, UCL_CHARACTER_WHITESPACE)) {
- state = macro_has_symbols;
- }
- else if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- state = macro_save;
- continue;
- }
-
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_has_symbols:
- if (*p == '{') {
- state = macro_has_sqbrace;
- }
- else if (*p == '(') {
- state = macro_has_obrace;
- }
- else if (*p == '"') {
- state = macro_has_quote;
- }
- else if (*p == '\n') {
- state = macro_save;
- continue;
- }
-
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_has_obrace:
- if (*p == '\\') {
- prev_state = state;
- state = macro_has_backslash;
- }
- else if (*p == ')') {
- state = macro_has_symbols;
- }
-
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_has_sqbrace:
- if (*p == '\\') {
- prev_state = state;
- state = macro_has_backslash;
- }
- else if (*p == '}') {
- state = macro_save;
- }
-
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_has_quote:
- if (*p == '\\') {
- prev_state = state;
- state = macro_has_backslash;
- }
- else if (*p == '"') {
- state = macro_save;
- }
-
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_has_backslash:
- state = prev_state;
- ucl_chunk_skipc (chunk, p);
- break;
-
- case macro_save:
- if (parser->flags & UCL_PARSER_SAVE_COMMENTS) {
- ucl_save_comment (parser, c, p - c);
- }
-
- return true;
- }
- }
-
- return false;
- }
-
- /**
- * Handle macro data
- * @param parser
- * @param chunk
- * @param marco
- * @param macro_start
- * @param macro_len
- * @return
- */
- static bool
- ucl_parse_macro_value (struct ucl_parser *parser,
- struct ucl_chunk *chunk, struct ucl_macro *macro,
- unsigned char const **macro_start, size_t *macro_len)
- {
- const unsigned char *p, *c;
- bool need_unescape = false, ucl_escape = false, var_expand = false;
-
- p = chunk->pos;
-
- switch (*p) {
- case '"':
- /* We have macro value encoded in quotes */
- c = p;
- ucl_chunk_skipc (chunk, p);
- if (!ucl_lex_json_string (parser, chunk, &need_unescape, &ucl_escape, &var_expand)) {
- return false;
- }
-
- *macro_start = c + 1;
- *macro_len = chunk->pos - c - 2;
- p = chunk->pos;
- break;
- case '{':
- /* We got a multiline macro body */
- ucl_chunk_skipc (chunk, p);
- /* Skip spaces at the beginning */
- while (p < chunk->end) {
- if (ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- ucl_chunk_skipc (chunk, p);
- }
- else {
- break;
- }
- }
- c = p;
- while (p < chunk->end) {
- if (*p == '}') {
- break;
- }
- ucl_chunk_skipc (chunk, p);
- }
- *macro_start = c;
- *macro_len = p - c;
- ucl_chunk_skipc (chunk, p);
- break;
- default:
- /* Macro is not enclosed in quotes or braces */
- c = p;
- while (p < chunk->end) {
- if (ucl_lex_is_atom_end (*p)) {
- break;
- }
- ucl_chunk_skipc (chunk, p);
- }
- *macro_start = c;
- *macro_len = p - c;
- break;
- }
-
- /* We are at the end of a macro */
- /* Skip ';' and space characters and return to previous state */
- while (p < chunk->end) {
- if (!ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE) && *p != ';') {
- break;
- }
- ucl_chunk_skipc (chunk, p);
- }
- return true;
- }
-
- /**
- * Parse macro arguments as UCL object
- * @param parser parser structure
- * @param chunk the current data chunk
- * @return
- */
- static ucl_object_t *
- ucl_parse_macro_arguments (struct ucl_parser *parser,
- struct ucl_chunk *chunk)
- {
- ucl_object_t *res = NULL;
- struct ucl_parser *params_parser;
- int obraces = 1, ebraces = 0, state = 0;
- const unsigned char *p, *c;
- size_t args_len = 0;
- struct ucl_parser_saved_state saved;
-
- saved.column = chunk->column;
- saved.line = chunk->line;
- saved.pos = chunk->pos;
- saved.remain = chunk->remain;
- p = chunk->pos;
-
- if (*p != '(' || chunk->remain < 2) {
- return NULL;
- }
-
- /* Set begin and start */
- ucl_chunk_skipc (chunk, p);
- c = p;
-
- while ((p) < (chunk)->end) {
- switch (state) {
- case 0:
- /* Parse symbols and check for '(', ')' and '"' */
- if (*p == '(') {
- obraces ++;
- }
- else if (*p == ')') {
- ebraces ++;
- }
- else if (*p == '"') {
- state = 1;
- }
- /* Check pairing */
- if (obraces == ebraces) {
- state = 99;
- }
- else {
- args_len ++;
- }
- /* Check overflow */
- if (chunk->remain == 0) {
- goto restore_chunk;
- }
- ucl_chunk_skipc (chunk, p);
- break;
- case 1:
- /* We have quote character, so skip all but quotes */
- if (*p == '"' && *(p - 1) != '\\') {
- state = 0;
- }
- if (chunk->remain == 0) {
- goto restore_chunk;
- }
- args_len ++;
- ucl_chunk_skipc (chunk, p);
- break;
- case 99:
- /*
- * We have read the full body of arguments, so we need to parse and set
- * object from that
- */
- params_parser = ucl_parser_new (parser->flags);
- if (!ucl_parser_add_chunk (params_parser, c, args_len)) {
- ucl_set_err (parser, UCL_ESYNTAX, "macro arguments parsing error",
- &parser->err);
- }
- else {
- res = ucl_parser_get_object (params_parser);
- }
- ucl_parser_free (params_parser);
-
- return res;
-
- break;
- }
- }
-
- return res;
-
- restore_chunk:
- chunk->column = saved.column;
- chunk->line = saved.line;
- chunk->pos = saved.pos;
- chunk->remain = saved.remain;
-
- return NULL;
- }
-
- #define SKIP_SPACES_COMMENTS(parser, chunk, p) do { \
- while ((p) < (chunk)->end) { \
- if (!ucl_test_character (*(p), UCL_CHARACTER_WHITESPACE_UNSAFE)) { \
- if ((chunk)->remain >= 2 && ucl_lex_is_comment ((p)[0], (p)[1])) { \
- if (!ucl_skip_comments (parser)) { \
- return false; \
- } \
- p = (chunk)->pos; \
- } \
- break; \
- } \
- ucl_chunk_skipc (chunk, p); \
- } \
- } while(0)
-
- /**
- * Handle the main states of rcl parser
- * @param parser parser structure
- * @return true if chunk has been parsed and false in case of error
- */
- static bool
- ucl_state_machine (struct ucl_parser *parser)
- {
- ucl_object_t *obj, *macro_args;
- struct ucl_chunk *chunk = parser->chunks;
- const unsigned char *p, *c = NULL, *macro_start = NULL;
- unsigned char *macro_escaped;
- size_t macro_len = 0;
- struct ucl_macro *macro = NULL;
- bool next_key = false, end_of_object = false, ret;
-
- if (parser->top_obj == NULL) {
- parser->state = UCL_STATE_INIT;
- }
-
- p = chunk->pos;
- while (chunk->pos < chunk->end) {
- switch (parser->state) {
- case UCL_STATE_INIT:
- /*
- * At the init state we can either go to the parse array or object
- * if we got [ or { correspondingly or can just treat new data as
- * a key of newly created object
- */
- if (!ucl_skip_comments (parser)) {
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- else {
- bool seen_obrace = false;
-
- /* Skip any spaces */
- while (p < chunk->end && ucl_test_character (*p,
- UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- ucl_chunk_skipc (chunk, p);
- }
-
- p = chunk->pos;
-
- if (p < chunk->end) {
- if (*p == '[') {
- parser->state = UCL_STATE_VALUE;
- ucl_chunk_skipc (chunk, p);
- seen_obrace = true;
- }
- else {
-
- if (*p == '{') {
- ucl_chunk_skipc (chunk, p);
- parser->state = UCL_STATE_KEY_OBRACE;
- seen_obrace = true;
- }
- else {
- parser->state = UCL_STATE_KEY;
- }
- }
- }
-
- if (parser->top_obj == NULL) {
- if (parser->state == UCL_STATE_VALUE) {
- obj = ucl_parser_add_container (NULL, parser, true, 0,
- seen_obrace);
- }
- else {
- obj = ucl_parser_add_container (NULL, parser, false, 0,
- seen_obrace);
- }
-
- if (obj == NULL) {
- return false;
- }
-
- parser->top_obj = obj;
- parser->cur_obj = obj;
- }
-
- }
- break;
- case UCL_STATE_KEY:
- case UCL_STATE_KEY_OBRACE:
- /* Skip any spaces */
- while (p < chunk->end && ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE)) {
- ucl_chunk_skipc (chunk, p);
- }
- if (p == chunk->end || *p == '}') {
- /* We have the end of an object */
- parser->state = UCL_STATE_AFTER_VALUE;
- continue;
- }
- if (parser->stack == NULL) {
- /* No objects are on stack, but we want to parse a key */
- ucl_set_err (parser, UCL_ESYNTAX, "top object is finished but the parser "
- "expects a key", &parser->err);
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- if (!ucl_parse_key (parser, chunk, &next_key, &end_of_object)) {
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
-
- if (end_of_object) {
- p = chunk->pos;
- parser->state = UCL_STATE_AFTER_VALUE;
- continue;
- }
- else if (parser->state != UCL_STATE_MACRO_NAME) {
- if (next_key && parser->stack->obj->type == UCL_OBJECT) {
- /* Parse more keys and nest objects accordingly */
- obj = ucl_parser_add_container (parser->cur_obj,
- parser,
- false,
- parser->stack->e.params.level + 1,
- parser->state == UCL_STATE_KEY_OBRACE);
- if (obj == NULL) {
- return false;
- }
- }
- else {
- parser->state = UCL_STATE_VALUE;
- }
- }
- else {
- c = chunk->pos;
- }
- p = chunk->pos;
- break;
- case UCL_STATE_VALUE:
- /* We need to check what we do have */
- if (!parser->cur_obj || !ucl_parse_value (parser, chunk)) {
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- /* State is set in ucl_parse_value call */
- p = chunk->pos;
- break;
- case UCL_STATE_AFTER_VALUE:
- if (!ucl_parse_after_value (parser, chunk)) {
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
-
- if (parser->stack != NULL) {
- if (parser->stack->obj->type == UCL_OBJECT) {
- parser->state = UCL_STATE_KEY;
- }
- else {
- /* Array */
- parser->state = UCL_STATE_VALUE;
- }
- }
- else {
- /* Skip everything at the end */
- return true;
- }
-
- p = chunk->pos;
- break;
- case UCL_STATE_MACRO_NAME:
- if (parser->flags & UCL_PARSER_DISABLE_MACRO) {
- if (!ucl_skip_macro_as_comment (parser, chunk)) {
- /* We have invalid macro */
- ucl_create_err (&parser->err,
- "error at %s:%d at column %d: invalid macro",
- chunk->fname ? chunk->fname : "memory",
- chunk->line,
- chunk->column);
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- else {
- p = chunk->pos;
- parser->state = parser->prev_state;
- }
- }
- else {
- if (!ucl_test_character (*p, UCL_CHARACTER_WHITESPACE_UNSAFE) &&
- *p != '(') {
- ucl_chunk_skipc (chunk, p);
- }
- else {
- if (c != NULL && p - c > 0) {
- /* We got macro name */
- macro_len = (size_t) (p - c);
- HASH_FIND (hh, parser->macroes, c, macro_len, macro);
- if (macro == NULL) {
- ucl_create_err (&parser->err,
- "error at %s:%d at column %d: "
- "unknown macro: '%.*s', character: '%c'",
- chunk->fname ? chunk->fname : "memory",
- chunk->line,
- chunk->column,
- (int) (p - c),
- c,
- *chunk->pos);
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- /* Now we need to skip all spaces */
- SKIP_SPACES_COMMENTS(parser, chunk, p);
- parser->state = UCL_STATE_MACRO;
- }
- else {
- /* We have invalid macro name */
- ucl_create_err (&parser->err,
- "error at %s:%d at column %d: invalid macro name",
- chunk->fname ? chunk->fname : "memory",
- chunk->line,
- chunk->column);
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- }
- }
- break;
- case UCL_STATE_MACRO:
- if (*chunk->pos == '(') {
- macro_args = ucl_parse_macro_arguments (parser, chunk);
- p = chunk->pos;
- if (macro_args) {
- SKIP_SPACES_COMMENTS(parser, chunk, p);
- }
- }
- else {
- macro_args = NULL;
- }
- if (!ucl_parse_macro_value (parser, chunk, macro,
- ¯o_start, ¯o_len)) {
- parser->prev_state = parser->state;
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- macro_len = ucl_expand_variable (parser, ¯o_escaped,
- macro_start, macro_len);
- parser->state = parser->prev_state;
-
- if (macro_escaped == NULL && macro != NULL) {
- if (macro->is_context) {
- ret = macro->h.context_handler (macro_start, macro_len,
- macro_args,
- parser->top_obj,
- macro->ud);
- }
- else {
- ret = macro->h.handler (macro_start, macro_len, macro_args,
- macro->ud);
- }
- }
- else if (macro != NULL) {
- if (macro->is_context) {
- ret = macro->h.context_handler (macro_escaped, macro_len,
- macro_args,
- parser->top_obj,
- macro->ud);
- }
- else {
- ret = macro->h.handler (macro_escaped, macro_len, macro_args,
- macro->ud);
- }
-
- UCL_FREE (macro_len + 1, macro_escaped);
- }
- else {
- ret = false;
- ucl_set_err (parser, UCL_EINTERNAL,
- "internal error: parser has macro undefined", &parser->err);
- }
-
- /*
- * Chunk can be modified within macro handler
- */
- chunk = parser->chunks;
- p = chunk->pos;
-
- if (macro_args) {
- ucl_object_unref (macro_args);
- }
-
- if (!ret) {
- return false;
- }
- break;
- default:
- ucl_set_err (parser, UCL_EINTERNAL,
- "internal error: parser is in an unknown state", &parser->err);
- parser->state = UCL_STATE_ERROR;
- return false;
- }
- }
-
- if (parser->last_comment) {
- if (parser->cur_obj) {
- ucl_attach_comment (parser, parser->cur_obj, true);
- }
- else if (parser->stack && parser->stack->obj) {
- ucl_attach_comment (parser, parser->stack->obj, true);
- }
- else if (parser->top_obj) {
- ucl_attach_comment (parser, parser->top_obj, true);
- }
- else {
- ucl_object_unref (parser->last_comment);
- }
- }
-
- if (parser->stack != NULL && parser->state != UCL_STATE_ERROR) {
- struct ucl_stack *st;
- bool has_error = false;
-
- LL_FOREACH (parser->stack, st) {
- if (st->chunk != parser->chunks) {
- break; /* Not our chunk, give up */
- }
- if (st->e.params.flags & UCL_STACK_HAS_OBRACE) {
- if (parser->err == NULL) {
- utstring_new (parser->err);
- }
-
- utstring_printf (parser->err, "%s:%d unmatched open brace at %d; ",
- chunk->fname ? chunk->fname : "memory",
- parser->chunks->line,
- st->e.params.line);
-
- has_error = true;
- }
- }
-
- if (has_error) {
- parser->err_code = UCL_EUNPAIRED;
-
- return false;
- }
- }
-
- return true;
- }
-
- #define UPRM_SAFE(fn, a, b, c, el) do { \
- if (!fn(a, b, c, a)) \
- goto el; \
- } while (0)
-
- struct ucl_parser*
- ucl_parser_new (int flags)
- {
- struct ucl_parser *parser;
-
- parser = UCL_ALLOC (sizeof (struct ucl_parser));
- if (parser == NULL) {
- return NULL;
- }
-
- memset (parser, 0, sizeof (struct ucl_parser));
-
- UPRM_SAFE(ucl_parser_register_macro, parser, "include", ucl_include_handler, e0);
- UPRM_SAFE(ucl_parser_register_macro, parser, "try_include", ucl_try_include_handler, e0);
- UPRM_SAFE(ucl_parser_register_macro, parser, "includes", ucl_includes_handler, e0);
- UPRM_SAFE(ucl_parser_register_macro, parser, "priority", ucl_priority_handler, e0);
- UPRM_SAFE(ucl_parser_register_macro, parser, "load", ucl_load_handler, e0);
- UPRM_SAFE(ucl_parser_register_context_macro, parser, "inherit", ucl_inherit_handler, e0);
-
- parser->flags = flags;
- parser->includepaths = NULL;
-
- if (flags & UCL_PARSER_SAVE_COMMENTS) {
- parser->comments = ucl_object_typed_new (UCL_OBJECT);
- }
-
- if (!(flags & UCL_PARSER_NO_FILEVARS)) {
- /* Initial assumption about filevars */
- ucl_parser_set_filevars (parser, NULL, false);
- }
-
- return parser;
- e0:
- ucl_parser_free(parser);
- return NULL;
- }
-
- bool
- ucl_parser_set_default_priority (struct ucl_parser *parser, unsigned prio)
- {
- if (parser == NULL) {
- return false;
- }
-
- parser->default_priority = prio;
-
- return true;
- }
-
- int
- ucl_parser_get_default_priority (struct ucl_parser *parser)
- {
- if (parser == NULL) {
- return -1;
- }
-
- return parser->default_priority;
- }
-
- bool
- ucl_parser_register_macro (struct ucl_parser *parser, const char *macro,
- ucl_macro_handler handler, void* ud)
- {
- struct ucl_macro *new;
-
- if (macro == NULL || handler == NULL) {
- return false;
- }
-
- new = UCL_ALLOC (sizeof (struct ucl_macro));
- if (new == NULL) {
- return false;
- }
-
- memset (new, 0, sizeof (struct ucl_macro));
- new->h.handler = handler;
- new->name = strdup (macro);
- if (new->name == NULL) {
- UCL_FREE (sizeof (struct ucl_macro), new);
- return false;
- }
- new->ud = ud;
- HASH_ADD_KEYPTR (hh, parser->macroes, new->name, strlen (new->name), new);
- return true;
- }
-
- bool
- ucl_parser_register_context_macro (struct ucl_parser *parser, const char *macro,
- ucl_context_macro_handler handler, void* ud)
- {
- struct ucl_macro *new;
-
- if (macro == NULL || handler == NULL) {
- return false;
- }
-
- new = UCL_ALLOC (sizeof (struct ucl_macro));
- if (new == NULL) {
- return false;
- }
-
- memset (new, 0, sizeof (struct ucl_macro));
- new->h.context_handler = handler;
- new->name = strdup (macro);
- if (new->name == NULL) {
- UCL_FREE (sizeof (struct ucl_macro), new);
- return false;
- }
- new->ud = ud;
- new->is_context = true;
- HASH_ADD_KEYPTR (hh, parser->macroes, new->name, strlen (new->name), new);
- return true;
- }
-
- void
- ucl_parser_register_variable (struct ucl_parser *parser, const char *var,
- const char *value)
- {
- struct ucl_variable *new = NULL, *cur;
-
- if (var == NULL) {
- return;
- }
-
- /* Find whether a variable already exists */
- LL_FOREACH (parser->variables, cur) {
- if (strcmp (cur->var, var) == 0) {
- new = cur;
- break;
- }
- }
-
- if (value == NULL) {
-
- if (new != NULL) {
- /* Remove variable */
- DL_DELETE (parser->variables, new);
- free (new->var);
- free (new->value);
- UCL_FREE (sizeof (struct ucl_variable), new);
- }
- else {
- /* Do nothing */
- return;
- }
- }
- else {
- if (new == NULL) {
- new = UCL_ALLOC (sizeof (struct ucl_variable));
- if (new == NULL) {
- return;
- }
- memset (new, 0, sizeof (struct ucl_variable));
- new->var = strdup (var);
- new->var_len = strlen (var);
- new->value = strdup (value);
- new->value_len = strlen (value);
-
- DL_APPEND (parser->variables, new);
- }
- else {
- free (new->value);
- new->value = strdup (value);
- new->value_len = strlen (value);
- }
- }
- }
-
- void
- ucl_parser_set_variables_handler (struct ucl_parser *parser,
- ucl_variable_handler handler, void *ud)
- {
- parser->var_handler = handler;
- parser->var_data = ud;
- }
-
- bool
- ucl_parser_add_chunk_full (struct ucl_parser *parser, const unsigned char *data,
- size_t len, unsigned priority, enum ucl_duplicate_strategy strat,
- enum ucl_parse_type parse_type)
- {
- struct ucl_chunk *chunk;
- struct ucl_parser_special_handler *special_handler;
-
- if (parser == NULL) {
- return false;
- }
-
- if (data == NULL && len != 0) {
- ucl_create_err (&parser->err, "invalid chunk added");
- return false;
- }
-
- if (parser->state != UCL_STATE_ERROR) {
- chunk = UCL_ALLOC (sizeof (struct ucl_chunk));
- if (chunk == NULL) {
- ucl_create_err (&parser->err, "cannot allocate chunk structure");
- return false;
- }
-
- memset (chunk, 0, sizeof (*chunk));
-
- /* Apply all matching handlers from the first to the last */
- LL_FOREACH (parser->special_handlers, special_handler) {
- if ((special_handler->flags & UCL_SPECIAL_HANDLER_PREPROCESS_ALL) ||
- (len >= special_handler->magic_len &&
- memcmp (data, special_handler->magic, special_handler->magic_len) == 0)) {
- unsigned char *ndata = NULL;
- size_t nlen = 0;
-
- if (!special_handler->handler (parser, data, len, &ndata, &nlen,
- special_handler->user_data)) {
- UCL_FREE(sizeof (struct ucl_chunk), chunk);
- ucl_create_err (&parser->err, "call for external handler failed");
-
- return false;
- }
-
- struct ucl_parser_special_handler_chain *nchain;
- nchain = UCL_ALLOC (sizeof (*nchain));
- nchain->begin = ndata;
- nchain->len = nlen;
- nchain->special_handler = special_handler;
-
- /* Free order is reversed */
- LL_PREPEND (chunk->special_handlers, nchain);
-
- data = ndata;
- len = nlen;
- }
- }
-
- if (parse_type == UCL_PARSE_AUTO && len > 0) {
- /* We need to detect parse type by the first symbol */
- if ((*data & 0x80) == 0x80) {
- parse_type = UCL_PARSE_MSGPACK;
- }
- else if (*data == '(') {
- parse_type = UCL_PARSE_CSEXP;
- }
- else {
- parse_type = UCL_PARSE_UCL;
- }
- }
-
- chunk->begin = data;
- chunk->remain = len;
- chunk->pos = chunk->begin;
- chunk->end = chunk->begin + len;
- chunk->line = 1;
- chunk->column = 0;
- chunk->priority = priority;
- chunk->strategy = strat;
- chunk->parse_type = parse_type;
-
- if (parser->cur_file) {
- chunk->fname = strdup (parser->cur_file);
- }
-
- LL_PREPEND (parser->chunks, chunk);
- parser->recursion ++;
-
- if (parser->recursion > UCL_MAX_RECURSION) {
- ucl_create_err (&parser->err, "maximum include nesting limit is reached: %d",
- parser->recursion);
- return false;
- }
-
- if (len > 0) {
- /* Need to parse something */
- switch (parse_type) {
- default:
- case UCL_PARSE_UCL:
- return ucl_state_machine (parser);
- case UCL_PARSE_MSGPACK:
- return ucl_parse_msgpack (parser);
- case UCL_PARSE_CSEXP:
- return ucl_parse_csexp (parser);
- }
- }
- else {
- /* Just add empty chunk and go forward */
- if (parser->top_obj == NULL) {
- /*
- * In case of empty object, create one to indicate that we've
- * read something
- */
- parser->top_obj = ucl_object_new_full (UCL_OBJECT, priority);
- }
-
- return true;
- }
- }
-
- ucl_create_err (&parser->err, "a parser is in an invalid state");
-
- return false;
- }
-
- bool
- ucl_parser_add_chunk_priority (struct ucl_parser *parser,
- const unsigned char *data, size_t len, unsigned priority)
- {
- /* We dereference parser, so this check is essential */
- if (parser == NULL) {
- return false;
- }
-
- return ucl_parser_add_chunk_full (parser, data, len,
- priority, UCL_DUPLICATE_APPEND, UCL_PARSE_UCL);
- }
-
- bool
- ucl_parser_add_chunk (struct ucl_parser *parser, const unsigned char *data,
- size_t len)
- {
- if (parser == NULL) {
- return false;
- }
-
- return ucl_parser_add_chunk_full (parser, data, len,
- parser->default_priority, UCL_DUPLICATE_APPEND, UCL_PARSE_UCL);
- }
-
- bool
- ucl_parser_insert_chunk (struct ucl_parser *parser, const unsigned char *data,
- size_t len)
- {
- if (parser == NULL || parser->top_obj == NULL) {
- return false;
- }
-
- bool res;
- struct ucl_chunk *chunk;
-
- int state = parser->state;
- parser->state = UCL_STATE_INIT;
-
- /* Prevent inserted chunks from unintentionally closing the current object */
- if (parser->stack != NULL && parser->stack->next != NULL) {
- parser->stack->e.params.level = parser->stack->next->e.params.level;
- }
-
- res = ucl_parser_add_chunk_full (parser, data, len, parser->chunks->priority,
- parser->chunks->strategy, parser->chunks->parse_type);
-
- /* Remove chunk from the stack */
- chunk = parser->chunks;
- if (chunk != NULL) {
- parser->chunks = chunk->next;
- ucl_chunk_free (chunk);
- parser->recursion --;
- }
-
- parser->state = state;
-
- return res;
- }
-
- bool
- ucl_parser_add_string_priority (struct ucl_parser *parser, const char *data,
- size_t len, unsigned priority)
- {
- if (data == NULL) {
- ucl_create_err (&parser->err, "invalid string added");
- return false;
- }
- if (len == 0) {
- len = strlen (data);
- }
-
- return ucl_parser_add_chunk_priority (parser,
- (const unsigned char *)data, len, priority);
- }
-
- bool
- ucl_parser_add_string (struct ucl_parser *parser, const char *data,
- size_t len)
- {
- if (parser == NULL) {
- return false;
- }
-
- return ucl_parser_add_string_priority (parser,
- (const unsigned char *)data, len, parser->default_priority);
- }
-
- bool
- ucl_set_include_path (struct ucl_parser *parser, ucl_object_t *paths)
- {
- if (parser == NULL || paths == NULL) {
- return false;
- }
-
- if (parser->includepaths == NULL) {
- parser->includepaths = ucl_object_copy (paths);
- }
- else {
- ucl_object_unref (parser->includepaths);
- parser->includepaths = ucl_object_copy (paths);
- }
-
- if (parser->includepaths == NULL) {
- return false;
- }
-
- return true;
- }
-
- unsigned char ucl_parser_chunk_peek (struct ucl_parser *parser)
- {
- if (parser == NULL || parser->chunks == NULL || parser->chunks->pos == NULL || parser->chunks->end == NULL ||
- parser->chunks->pos == parser->chunks->end) {
- return 0;
- }
-
- return( *parser->chunks->pos );
- }
-
- bool ucl_parser_chunk_skip (struct ucl_parser *parser)
- {
- if (parser == NULL || parser->chunks == NULL || parser->chunks->pos == NULL || parser->chunks->end == NULL ||
- parser->chunks->pos == parser->chunks->end) {
- return false;
- }
-
- const unsigned char *p = parser->chunks->pos;
- ucl_chunk_skipc( parser->chunks, p );
- if( parser->chunks->pos != NULL ) return true;
- return false;
- }
-
- ucl_object_t*
- ucl_parser_get_current_stack_object (struct ucl_parser *parser, unsigned int depth)
- {
- ucl_object_t *obj;
-
- if (parser == NULL || parser->stack == NULL) {
- return NULL;
- }
-
- struct ucl_stack *stack = parser->stack;
- if(stack == NULL || stack->obj == NULL || ucl_object_type (stack->obj) != UCL_OBJECT)
- {
- return NULL;
- }
-
- for( unsigned int i = 0; i < depth; ++i )
- {
- stack = stack->next;
- if(stack == NULL || stack->obj == NULL || ucl_object_type (stack->obj) != UCL_OBJECT)
- {
- return NULL;
- }
- }
-
- obj = ucl_object_ref (stack->obj);
- return obj;
- }
|