*/ REBSER *Collect_End(REBSER *prior)
/*
** Finish collecting words, and free the Bind_Table for reuse.
**
***********************************************************************/
{
REBVAL *words;
REBINT *binds = WORDS_HEAD(Bind_Table); // GC safe to do here
// Reset binding table (note BUF_WORDS may have expanded):
for (words = BLK_HEAD(BUF_WORDS); NOT_END(words); words++)
binds[VAL_WORD_CANON(words)] = 0;
// If no new words, prior frame:
if (prior && SERIES_TAIL(BUF_WORDS) == SERIES_TAIL(prior)) {
RESET_TAIL(BUF_WORDS); // allow reuse
return FRM_WORD_SERIES(prior);
}
prior = Copy_Series(BUF_WORDS);
RESET_TAIL(BUF_WORDS); // allow reuse
BARE_SERIES(prior); // No GC ever needed for word list
CHECK_BIND_TABLE;
return prior;
}
*/ REBSER *Collect_Block_Words(REBVAL *block, REBVAL *prior, REBCNT modes)
/*
** Collect words from a prior block and new block.
**
***********************************************************************/
{
REBSER *series;
REBCNT start;
REBINT *binds = WORDS_HEAD(Bind_Table); // GC safe to do here
CHECK_BIND_TABLE;
if (SERIES_TAIL(BUF_WORDS)) Crash(RP_WORD_LIST); // still in use
if (prior)
Collect_Simple_Words(prior, BIND_ALL);
start = SERIES_TAIL(BUF_WORDS);
Collect_Simple_Words(block, modes);
// Reset word markers:
for (block = BLK_HEAD(BUF_WORDS); NOT_END(block); block++)
binds[VAL_WORD_CANON(block)] = 0;
series = Copy_Series_Part(BUF_WORDS, start, SERIES_TAIL(BUF_WORDS)-start);
RESET_TAIL(BUF_WORDS); // allow reuse
CHECK_BIND_TABLE;
return series;
}
*/ REBSER *Collect_Words(REBVAL value[], REBVAL prior_value[], REBCNT modes)
/*
** Collect words from a prior block and new block.
**
***********************************************************************/
{
REBSER *series;
REBCNT start;
REBINT *binds = WORDS_HEAD(Bind_Table); // GC safe to do here
CHECK_BIND_TABLE;
if (SERIES_TAIL(BUF_WORDS)) panic Error_0(RE_WORD_LIST); // still in use
if (prior_value)
Collect_Words_Inner_Loop(binds, &prior_value[0], BIND_ALL);
start = SERIES_TAIL(BUF_WORDS);
Collect_Words_Inner_Loop(binds, &value[0], modes);
// Reset word markers:
for (value = BLK_HEAD(BUF_WORDS); NOT_END(value); value++)
binds[VAL_WORD_CANON(value)] = 0;
series = Copy_Array_At_Max_Shallow(
BUF_WORDS, start, SERIES_TAIL(BUF_WORDS) - start
);
RESET_TAIL(BUF_WORDS); // allow reuse
CHECK_BIND_TABLE;
return series;
}
*/ REBYTE *Reset_Buffer(REBSER *buf, REBCNT len)
/*
** Setup to reuse a shared buffer. Expand it if needed.
**
** NOTE:The tail is set to the length position.
**
***********************************************************************/
{
if (!buf) panic Error_0(RE_NO_BUFFER);
RESET_TAIL(buf);
if (SERIES_BIAS(buf)) Reset_Bias(buf);
Expand_Series(buf, 0, len); // sets new tail
return BIN_DATA(buf);
}
*/ void Collect_Words(REBVAL *block, REBFLG modes)
/*
** The inner recursive loop used for Collect_Words function below.
**
***********************************************************************/
{
REBINT *binds = WORDS_HEAD(Bind_Table);
REBVAL *word;
REBVAL *value;
for (; NOT_END(block); block++) {
value = block;
//if (modes & BIND_GET && IS_GET_WORD(block)) value = Get_Var(block);
if (ANY_WORD(value)) {
if (!binds[VAL_WORD_CANON(value)]) { // only once per word
if (IS_SET_WORD(value) || modes & BIND_ALL) {
binds[VAL_WORD_CANON(value)] = SERIES_TAIL(BUF_WORDS);
EXPAND_SERIES_TAIL(BUF_WORDS, 1);
word = BLK_LAST(BUF_WORDS);
VAL_SET(word, VAL_TYPE(value));
VAL_SET_OPT(word, OPTS_UNWORD);
VAL_BIND_SYM(word) = VAL_WORD_SYM(value);
// Allow all datatypes (to start):
VAL_BIND_TYPESET(word) = ~((TYPESET(REB_END) | TYPESET(REB_UNSET))); // not END or UNSET
}
} else {
// If word duplicated:
if (modes & BIND_NO_DUP) {
// Reset binding table (note BUF_WORDS may have expanded):
for (word = BLK_HEAD(BUF_WORDS); NOT_END(word); word++)
binds[VAL_WORD_CANON(word)] = 0;
RESET_TAIL(BUF_WORDS); // allow reuse
Trap1(RE_DUP_VARS, value);
}
}
continue;
}
// Recurse into sub-blocks:
if (ANY_EVAL_BLOCK(value) && (modes & BIND_DEEP))
Collect_Words(VAL_BLK_DATA(value), modes);
// In this mode (foreach native), do not allow non-words:
//else if (modes & BIND_GET) Trap_Arg(value);
}
BLK_TERM(BUF_WORDS);
}
*/ static int Read_Dir(REBREQ *dir, REBSER *files)
/*
** Provide option to get file info too.
** Provide option to prepend dir path.
** Provide option to use wildcards.
**
***********************************************************************/
{
REBINT result;
REBCNT len;
REBSER *fname;
REBSER *name;
REBREQ file;
RESET_TAIL(files);
CLEARS(&file);
// Temporary filename storage:
fname = BUF_OS_STR;
file.special.file.path = cast(REBCHR*, Reset_Buffer(fname, MAX_FILE_NAME));
SET_FLAG(dir->modes, RFM_DIR);
dir->common.data = cast(REBYTE*, &file);
while ((result = OS_DO_DEVICE(dir, RDC_READ)) == 0 && !GET_FLAG(dir->flags, RRF_DONE)) {
len = OS_STRLEN(file.special.file.path);
if (GET_FLAG(file.modes, RFM_DIR)) len++;
name = Copy_OS_Str(file.special.file.path, len);
if (GET_FLAG(file.modes, RFM_DIR))
SET_ANY_CHAR(name, name->tail-1, '/');
Val_Init_File(Alloc_Tail_Array(files), name);
}
if (result < 0 && dir->error != -RFE_OPEN_FAIL
&& (
OS_STRCHR(dir->special.file.path, '*')
|| OS_STRCHR(dir->special.file.path, '?')
)
) {
result = 0; // no matches found, but not an error
}
return result;
}
*/ static void Collect_Frame_Inner_Loop(REBINT *binds, REBVAL value[], REBCNT modes)
/*
** The inner recursive loop used for Collect_Frame function below.
**
***********************************************************************/
{
for (; NOT_END(value); value++) {
if (ANY_WORD(value)) {
if (!binds[VAL_WORD_CANON(value)]) { // only once per word
if (IS_SET_WORD(value) || modes & BIND_ALL) {
REBVAL *word;
binds[VAL_WORD_CANON(value)] = SERIES_TAIL(BUF_WORDS);
EXPAND_SERIES_TAIL(BUF_WORDS, 1);
word = BLK_LAST(BUF_WORDS);
Val_Init_Word_Typed(
word,
VAL_TYPE(value),
VAL_WORD_SYM(value),
// Allow all datatypes but END or UNSET (initially):
~((TYPESET(REB_END) | TYPESET(REB_UNSET)))
);
}
} else {
// If word duplicated:
if (modes & BIND_NO_DUP) {
// Reset binding table (note BUF_WORDS may have expanded):
REBVAL *word;
for (word = BLK_HEAD(BUF_WORDS); NOT_END(word); word++)
binds[VAL_WORD_CANON(word)] = 0;
RESET_TAIL(BUF_WORDS); // allow reuse
raise Error_1(RE_DUP_VARS, value);
}
}
continue;
}
// Recurse into sub-blocks:
if (ANY_EVAL_BLOCK(value) && (modes & BIND_DEEP))
Collect_Frame_Inner_Loop(binds, VAL_BLK_DATA(value), modes);
// In this mode (foreach native), do not allow non-words:
//else if (modes & BIND_GET) raise Error_Invalid_Arg(value);
}
BLK_TERM(BUF_WORDS);
}
*/ static int Read_Dir(REBREQ *dir, REBSER *files)
/*
** Provide option to get file info too.
** Provide option to prepend dir path.
** Provide option to use wildcards.
**
***********************************************************************/
{
REBINT result;
REBCNT len;
REBSER *fname;
REBSER *name;
REBREQ file;
RESET_TAIL(files);
CLEARS(&file);
// Temporary filename storage:
fname = BUF_OS_STR;
file.file.path = (REBCHR*)Reset_Buffer(fname, MAX_FILE_NAME);
SET_FLAG(dir->modes, RFM_DIR);
dir->data = (REBYTE*)(&file);
while ((result = OS_DO_DEVICE(dir, RDC_READ)) == 0 && !GET_FLAG(dir->flags, RRF_DONE)) {
len = LEN_STR(file.file.path);
if (GET_FLAG(file.modes, RFM_DIR)) len++;
name = Copy_OS_Str(file.file.path, len);
if (GET_FLAG(file.modes, RFM_DIR))
SET_ANY_CHAR(name, name->tail-1, '/');
Set_Series(REB_FILE, Append_Value(files), name);
}
if (result < 0 && dir->error != -RFE_OPEN_FAIL
&& (FIND_CHR(dir->file.path, '*') || FIND_CHR(dir->file.path, '?')))
result = 0; // no matches found, but not an error
return result;
}
*/ int Encode_UTF8_Line(REBSER *dst, REBSER *src, REBCNT idx)
/*
** Encode a unicode source buffer into a binary line of UTF8.
** Include the LF terminator in the result.
** Return the length of the line buffer.
**
***********************************************************************/
{
REBUNI *up = UNI_HEAD(src);
REBCNT len = SERIES_TAIL(src);
REBCNT tail;
REBUNI c;
REBINT n;
REBYTE buf[8];
tail = RESET_TAIL(dst);
while (idx < len) {
if ((c = up[idx]) < 0x80) {
EXPAND_SERIES_TAIL(dst, 1);
BIN_HEAD(dst)[tail++] = (REBYTE)c;
}
else {
n = Encode_UTF8_Char(buf, c);
EXPAND_SERIES_TAIL(dst, n);
memcpy(BIN_SKIP(dst, tail), buf, n);
tail += n;
}
idx++;
if (c == LF) break;
}
BIN_HEAD(dst)[tail] = 0;
SERIES_TAIL(dst) = tail;
return idx;
}
*/ void Resolve_Context(REBSER *target, REBSER *source, REBVAL *only_words, REBFLG all, REBFLG expand)
/*
** Only_words can be a block of words or an index in the target
** (for new words).
**
***********************************************************************/
{
REBINT *binds = WORDS_HEAD(Bind_Table); // GC safe to do here
REBVAL *words;
REBVAL *vals;
REBINT n;
REBINT m;
REBCNT i = 0;
CHECK_BIND_TABLE;
if (IS_PROTECT_SERIES(target)) Trap0(RE_PROTECTED);
if (IS_INTEGER(only_words)) { // Must be: 0 < i <= tail
i = VAL_INT32(only_words); // never <= 0
if (i == 0) i = 1;
if (i >= target->tail) return;
}
Collect_Start(BIND_NO_SELF); // DO NOT TRAP IN THIS SECTION
n = 0;
// If limited resolve, tag the word ids that need to be copied:
if (i) {
// Only the new words of the target:
for (words = FRM_WORD(target, i); NOT_END(words); words++)
binds[VAL_BIND_CANON(words)] = -1;
n = SERIES_TAIL(target) - 1;
}
else if (IS_BLOCK(only_words)) {
// Limit exports to only these words:
for (words = VAL_BLK_DATA(only_words); NOT_END(words); words++) {
if (IS_WORD(words) || IS_SET_WORD(words)) {
binds[VAL_WORD_CANON(words)] = -1;
n++;
}
}
}
// Expand target as needed:
if (expand && n > 0) {
// Determine how many new words to add:
for (words = FRM_WORD(target, 1); NOT_END(words); words++)
if (binds[VAL_BIND_CANON(words)]) n--;
// Expand frame by the amount required:
if (n > 0) Expand_Frame(target, n, 0);
else expand = 0;
}
// Maps a word to its value index in the source context.
// Done by marking all source words (in bind table):
words = FRM_WORDS(source)+1;
for (n = 1; NOT_END(words); n++, words++) {
if (IS_NONE(only_words) || binds[VAL_BIND_CANON(words)])
binds[VAL_WORD_CANON(words)] = n;
}
// Foreach word in target, copy the correct value from source:
n = i ? i : 1;
vals = FRM_VALUE(target, n);
for (words = FRM_WORD(target, n); NOT_END(words); words++, vals++) {
if ((m = binds[VAL_BIND_CANON(words)])) {
binds[VAL_BIND_CANON(words)] = 0; // mark it as set
if (!VAL_PROTECTED(words) && (all || IS_UNSET(vals))) {
if (m < 0) SET_UNSET(vals); // no value in source context
else *vals = *FRM_VALUE(source, m);
//Debug_Num("type:", VAL_TYPE(vals));
//Debug_Str(Get_Word_Name(words));
}
}
}
// Add any new words and values:
if (expand) {
REBVAL *val;
words = FRM_WORDS(source)+1;
for (n = 1; NOT_END(words); n++, words++) {
if (binds[VAL_BIND_CANON(words)]) {
// Note: no protect check is needed here
binds[VAL_BIND_CANON(words)] = 0;
val = Append_Frame(target, 0, VAL_BIND_SYM(words));
*val = *FRM_VALUE(source, n);
}
}
}
else {
// Reset bind table (do not use Collect_End):
if (i) {
for (words = FRM_WORD(target, i); NOT_END(words); words++)
binds[VAL_BIND_CANON(words)] = 0;
}
else if (IS_BLOCK(only_words)) {
for (words = VAL_BLK_DATA(only_words); NOT_END(words); words++) {
if (IS_WORD(words) || IS_SET_WORD(words)) binds[VAL_WORD_CANON(words)] = 0;
}
}
else {
for (words = FRM_WORDS(source)+1; NOT_END(words); words++)
binds[VAL_BIND_CANON(words)] = 0;
}
}
CHECK_BIND_TABLE;
RESET_TAIL(BUF_WORDS); // allow reuse, trapping ok now
}
//
// Make_Set_Operation_Series: C
//
// Do set operations on a series. Case-sensitive if `cased` is TRUE.
// `skip` is the record size.
//
static REBSER *Make_Set_Operation_Series(const REBVAL *val1, const REBVAL *val2, REBCNT flags, REBCNT cased, REBCNT skip)
{
REBSER *buffer; // buffer for building the return series
REBCNT i;
REBINT h = TRUE;
REBFLG first_pass = TRUE; // are we in the first pass over the series?
REBSER *out_ser;
// This routine should only be called with SERIES! values
assert(ANY_SERIES(val1));
if (val2) {
assert(ANY_SERIES(val2));
if (ANY_ARRAY(val1)) {
if (!ANY_ARRAY(val2))
fail (Error_Unexpected_Type(VAL_TYPE(val1), VAL_TYPE(val2)));
// As long as they're both arrays, we're willing to do:
//
// >> union quote (a b c) 'b/d/e
// (a b c d e)
//
// The type of the result will match the first value.
}
else if (!IS_BINARY(val1)) {
// We will similarly do any two ANY-STRING! types:
//
// >> union <abc> "bde"
// <abcde>
if (IS_BINARY(val2))
fail (Error_Unexpected_Type(VAL_TYPE(val1), VAL_TYPE(val2)));
}
else {
// Binaries only operate with other binaries
if (!IS_BINARY(val2))
fail (Error_Unexpected_Type(VAL_TYPE(val1), VAL_TYPE(val2)));
}
}
// Calculate i as length of result block.
i = VAL_LEN(val1);
if (flags & SOP_FLAG_BOTH) i += VAL_LEN(val2);
if (ANY_ARRAY(val1)) {
REBSER *hser = 0; // hash table for series
REBSER *hret; // hash table for return series
buffer = BUF_EMIT; // use preallocated shared block
Resize_Series(buffer, i);
hret = Make_Hash_Sequence(i); // allocated
// Optimization note: !!
// This code could be optimized for small blocks by not hashing them
// and extending Find_Key to do a FIND on the value itself w/o the hash.
do {
REBSER *ser = VAL_SERIES(val1); // val1 and val2 swapped 2nd pass!
// Check what is in series1 but not in series2:
if (flags & SOP_FLAG_CHECK)
hser = Hash_Block(val2, cased);
// Iterate over first series:
i = VAL_INDEX(val1);
for (; i < SERIES_TAIL(ser); i += skip) {
REBVAL *item = BLK_SKIP(ser, i);
if (flags & SOP_FLAG_CHECK) {
h = Find_Key(VAL_SERIES(val2), hser, item, skip, cased, 1);
h = (h >= 0);
if (flags & SOP_FLAG_INVERT) h = !h;
}
if (h) Find_Key(buffer, hret, item, skip, cased, 2);
}
if (flags & SOP_FLAG_CHECK)
Free_Series(hser);
if (!first_pass) break;
first_pass = FALSE;
// Iterate over second series?
if ((i = ((flags & SOP_FLAG_BOTH) != 0))) {
const REBVAL *temp = val1;
val1 = val2;
val2 = temp;
}
} while (i);
if (hret)
Free_Series(hret);
out_ser = Copy_Array_Shallow(buffer);
RESET_TAIL(buffer); // required - allow reuse
}
else {
if (IS_BINARY(val1)) {
// All binaries use "case-sensitive" comparison (e.g. each byte
// is treated distinctly)
cased = TRUE;
}
buffer = BUF_MOLD;
Reset_Buffer(buffer, i);
RESET_TAIL(buffer);
do {
REBSER *ser = VAL_SERIES(val1); // val1 and val2 swapped 2nd pass!
REBUNI uc;
// Iterate over first series:
i = VAL_INDEX(val1);
for (; i < SERIES_TAIL(ser); i += skip) {
uc = GET_ANY_CHAR(ser, i);
if (flags & SOP_FLAG_CHECK) {
h = (NOT_FOUND != Find_Str_Char(
VAL_SERIES(val2),
0,
VAL_INDEX(val2),
VAL_TAIL(val2),
skip,
uc,
cased ? AM_FIND_CASE : 0
));
if (flags & SOP_FLAG_INVERT) h = !h;
}
if (!h) continue;
if (
NOT_FOUND == Find_Str_Char(
buffer,
0,
0,
SERIES_TAIL(buffer),
skip,
uc,
cased ? AM_FIND_CASE : 0
)
) {
Append_String(buffer, ser, i, skip);
}
}
if (!first_pass) break;
first_pass = FALSE;
// Iterate over second series?
if ((i = ((flags & SOP_FLAG_BOTH) != 0))) {
const REBVAL *temp = val1;
val1 = val2;
val2 = temp;
}
} while (i);
out_ser = Copy_String(buffer, 0, -1);
}
return out_ser;
}
*/ static REBINT Do_Set_Operation(struct Reb_Call *call_, REBCNT flags)
/*
** Do set operations on a series.
**
***********************************************************************/
{
REBVAL *val;
REBVAL *val1;
REBVAL *val2 = 0;
REBSER *ser;
REBSER *hser = 0; // hash table for series
REBSER *retser; // return series
REBSER *hret; // hash table for return series
REBCNT i;
REBINT h = TRUE;
REBCNT skip = 1; // record size
REBCNT cased = 0; // case sensitive when TRUE
SET_NONE(D_OUT);
val1 = D_ARG(1);
i = 2;
// Check for second series argument:
if (flags != SET_OP_UNIQUE) {
val2 = D_ARG(i++);
if (VAL_TYPE(val1) != VAL_TYPE(val2))
raise Error_Unexpected_Type(VAL_TYPE(val1), VAL_TYPE(val2));
}
// Refinements /case and /skip N
cased = D_REF(i++); // cased
if (D_REF(i++)) skip = Int32s(D_ARG(i), 1);
switch (VAL_TYPE(val1)) {
case REB_BLOCK:
i = VAL_LEN(val1);
// Setup result block:
if (GET_FLAG(flags, SOP_BOTH)) i += VAL_LEN(val2);
retser = BUF_EMIT; // use preallocated shared block
Resize_Series(retser, i);
hret = Make_Hash_Sequence(i); // allocated
// Optimization note: !!
// This code could be optimized for small blocks by not hashing them
// and extending Find_Key to do a FIND on the value itself w/o the hash.
do {
// Check what is in series1 but not in series2:
if (GET_FLAG(flags, SOP_CHECK))
hser = Hash_Block(val2, cased);
// Iterate over first series:
ser = VAL_SERIES(val1);
i = VAL_INDEX(val1);
for (; val = BLK_SKIP(ser, i), i < SERIES_TAIL(ser); i += skip) {
if (GET_FLAG(flags, SOP_CHECK)) {
h = Find_Key(VAL_SERIES(val2), hser, val, skip, cased, 1) >= 0;
if (GET_FLAG(flags, SOP_INVERT)) h = !h;
}
if (h) Find_Key(retser, hret, val, skip, cased, 2);
}
// Iterate over second series?
if ((i = GET_FLAG(flags, SOP_BOTH))) {
val = val1;
val1 = val2;
val2 = val;
CLR_FLAG(flags, SOP_BOTH);
}
if (GET_FLAG(flags, SOP_CHECK))
Free_Series(hser);
} while (i);
if (hret)
Free_Series(hret);
Val_Init_Block(D_OUT, Copy_Array_Shallow(retser));
RESET_TAIL(retser); // required - allow reuse
break;
case REB_BINARY:
cased = TRUE;
SET_TYPE(D_OUT, REB_BINARY);
case REB_STRING:
i = VAL_LEN(val1);
// Setup result block:
if (GET_FLAG(flags, SOP_BOTH)) i += VAL_LEN(val2);
retser = BUF_MOLD;
Reset_Buffer(retser, i);
RESET_TAIL(retser);
do {
REBUNI uc;
cased = cased ? AM_FIND_CASE : 0;
// Iterate over first series:
ser = VAL_SERIES(val1);
i = VAL_INDEX(val1);
for (; i < SERIES_TAIL(ser); i += skip) {
uc = GET_ANY_CHAR(ser, i);
if (GET_FLAG(flags, SOP_CHECK)) {
h = Find_Str_Char(VAL_SERIES(val2), 0, VAL_INDEX(val2), VAL_TAIL(val2), skip, uc, cased) != NOT_FOUND;
if (GET_FLAG(flags, SOP_INVERT)) h = !h;
}
if (h && (Find_Str_Char(retser, 0, 0, SERIES_TAIL(retser), skip, uc, cased) == NOT_FOUND)) {
Append_String(retser, ser, i, skip);
}
}
// Iterate over second series?
if ((i = GET_FLAG(flags, SOP_BOTH))) {
val = val1;
val1 = val2;
val2 = val;
CLR_FLAG(flags, SOP_BOTH);
}
} while (i);
ser = Copy_String(retser, 0, -1);
if (IS_BINARY(D_OUT))
Val_Init_Binary(D_OUT, ser);
else
Val_Init_String(D_OUT, ser);
break;
case REB_BITSET:
switch (flags) {
case SET_OP_UNIQUE:
return R_ARG1;
case SET_OP_UNION:
i = A_OR;
break;
case SET_OP_INTERSECT:
i = A_AND;
break;
case SET_OP_DIFFERENCE:
i = A_XOR;
break;
case SET_OP_EXCLUDE:
i = 0; // special case
break;
}
ser = Xandor_Binary(i, val1, val2);
Val_Init_Bitset(D_OUT, ser);
break;
case REB_TYPESET:
switch (flags) {
case SET_OP_UNIQUE:
break;
case SET_OP_UNION:
VAL_TYPESET(val1) |= VAL_TYPESET(val2);
break;
case SET_OP_INTERSECT:
VAL_TYPESET(val1) &= VAL_TYPESET(val2);
break;
case SET_OP_DIFFERENCE:
VAL_TYPESET(val1) ^= VAL_TYPESET(val2);
break;
case SET_OP_EXCLUDE:
VAL_TYPESET(val1) &= ~VAL_TYPESET(val2);
break;
}
return R_ARG1;
default:
raise Error_Invalid_Arg(val1);
}
return R_OUT;
}
*/ REBYTE *Scan_Item(REBYTE *src, REBYTE *end, REBUNI term, REBYTE *invalid)
/*
** Scan as UTF8 an item like a file or URL.
**
** Returns continuation point or zero for error.
**
** Put result into the MOLD_BUF as uni-chars.
**
***********************************************************************/
{
REBUNI c;
REBSER *buf;
buf = BUF_MOLD;
RESET_TAIL(buf);
while (src < end && *src != term) {
c = *src;
// End of stream?
if (c == 0) break;
// If no term, then any white will terminate:
if (!term && IS_WHITE(c)) break;
// Ctrl chars are invalid:
if (c < ' ') return 0; // invalid char
if (c == '\\') c = '/';
// Accept %xx encoded char:
else if (c == '%') {
if (!Scan_Hex2(src+1, &c, FALSE)) return 0;
src += 2;
}
// Accept ^X encoded char:
else if (c == '^') {
if (src+1 == end) return 0; // nothing follows ^
c = Scan_Char(&src);
if (!term && IS_WHITE(c)) break;
src--;
}
// Accept UTF8 encoded char:
else if (c >= 0x80) {
c = Decode_UTF8_Char(&src, 0); // zero on error
if (c == 0) return 0;
}
// Is char as literal valid? (e.g. () [] etc.)
else if (invalid && strchr(invalid, c)) return 0;
src++;
*UNI_SKIP(buf, buf->tail) = c; // not affected by Extend_Series
if (++(buf->tail) >= SERIES_REST(buf)) Extend_Series(buf, 1);
}
if (*src && *src == term) src++;
UNI_TERM(buf);
return src;
}
*/ REBYTE *Scan_Quote(REBYTE *src, SCAN_STATE *scan_state)
/*
** Scan a quoted string, handling all the escape characters.
**
** The result will be put into the temporary MOLD_BUF unistring.
**
***********************************************************************/
{
REBINT nest = 0;
REBUNI term;
REBINT chr;
REBCNT lines = 0;
REBSER *buf = BUF_MOLD;
RESET_TAIL(buf);
term = (*src++ == '{') ? '}' : '"'; // pick termination
while (*src != term || nest > 0) {
chr = *src;
switch (chr) {
case 0:
return 0; // Scan_state shows error location.
case '^':
chr = Scan_Char(&src);
if (chr == -1) return 0;
src--;
break;
case '{':
if (term != '"') nest++;
break;
case '}':
if (term != '"' && nest > 0) nest--;
break;
case CR:
if (src[1] == LF) src++;
// fall thru
case LF:
if (term == '"') return 0;
lines++;
chr = LF;
break;
default:
if (chr >= 0x80) {
chr = Decode_UTF8_Char(&src, 0); // zero on error
if (chr == 0) return 0;
}
}
src++;
*UNI_SKIP(buf, buf->tail) = chr;
if (++(buf->tail) >= SERIES_REST(buf)) Extend_Series(buf, 1);
}
src++; // Skip ending quote or brace.
if (scan_state) scan_state->line_count += lines;
UNI_TERM(buf);
return src;
}