static String*
S_unescape(QueryParser *self, String *orig, CharBuf *buf) {
StringIterator *iter = Str_Top(orig);
int32_t code_point;
UNUSED_VAR(self);
CB_Set_Size(buf, 0);
CB_Grow(buf, Str_Get_Size(orig) + 4);
while (STRITER_DONE != (code_point = StrIter_Next(iter))) {
if (code_point == '\\') {
int32_t next_code_point = StrIter_Next(iter);
if (next_code_point == ':'
|| next_code_point == '"'
|| next_code_point == '\\'
) {
CB_Cat_Char(buf, next_code_point);
}
else {
CB_Cat_Char(buf, code_point);
if (next_code_point != STRITER_DONE) {
CB_Cat_Char(buf, next_code_point);
}
}
}
else {
CB_Cat_Char(buf, code_point);
}
}
DECREF(iter);
return CB_To_String(buf);
}
Hash*
Hash_deserialize(Hash *self, InStream *instream) {
uint32_t size = InStream_Read_C32(instream);
uint32_t num_charbufs = InStream_Read_C32(instream);
uint32_t num_other = size - num_charbufs;
CharBuf *key = num_charbufs ? CB_new(0) : NULL;
Hash_init(self, size);
// Read key-value pairs with CharBuf keys.
while (num_charbufs--) {
uint32_t len = InStream_Read_C32(instream);
char *key_buf = CB_Grow(key, len);
InStream_Read_Bytes(instream, key_buf, len);
key_buf[len] = '\0';
CB_Set_Size(key, len);
Hash_Store(self, (Obj*)key, THAW(instream));
}
DECREF(key);
// Read remaining key/value pairs.
while (num_other--) {
Obj *k = THAW(instream);
Hash_Store(self, k, THAW(instream));
DECREF(k);
}
return self;
}
Obj*
TextSortCache_value(TextSortCache *self, int32_t ord, Obj *blank)
{
if (ord == self->null_ord) {
return NULL;
}
InStream_Seek(self->ix_in, ord * sizeof(int64_t));
int64_t offset = InStream_Read_I64(self->ix_in);
if (offset == NULL_SENTINEL) {
return NULL;
}
else {
uint32_t next_ord = ord + 1;
int64_t next_offset;
while (1) {
InStream_Seek(self->ix_in, next_ord * sizeof(int64_t));
next_offset = InStream_Read_I64(self->ix_in);
if (next_offset != NULL_SENTINEL) { break; }
next_ord++;
}
// Read character data into CharBuf.
CERTIFY(blank, CHARBUF);
int64_t len = next_offset - offset;
char *ptr = CB_Grow((CharBuf*)blank, (size_t)len);
InStream_Seek(self->dat_in, offset);
InStream_Read_Bytes(self->dat_in, ptr, (size_t)len);
ptr[len] = '\0';
CB_Set_Size((CharBuf*)blank, (size_t)len);
}
return blank;
}
void
TextTermStepper_read_key_frame(TextTermStepper *self, InStream *instream)
{
const uint32_t text_len = InStream_Read_C32(instream);
CharBuf *value;
char *ptr;
// Allocate space.
if (self->value == NULL) {
self->value = (Obj*)CB_new(text_len);
}
value = (CharBuf*)self->value;
ptr = CB_Grow(value, text_len);
// Set the value text.
InStream_Read_Bytes(instream, ptr, text_len);
CB_Set_Size(value, text_len);
if (!StrHelp_utf8_valid(ptr, text_len)) {
THROW(ERR, "Invalid UTF-8 sequence in '%o' at byte %i64",
InStream_Get_Filename(instream),
InStream_Tell(instream) - text_len);
}
// Null-terminate.
ptr[text_len] = '\0';
}
void
TextTermStepper_Read_Delta_IMP(TextTermStepper *self, InStream *instream) {
TextTermStepperIVARS *const ivars = TextTermStepper_IVARS(self);
const uint32_t text_overlap = InStream_Read_C32(instream);
const uint32_t finish_chars_len = InStream_Read_C32(instream);
const uint32_t total_text_len = text_overlap + finish_chars_len;
// Allocate space.
CharBuf *charbuf = (CharBuf*)ivars->value;
char *ptr = CB_Grow(charbuf, total_text_len);
// Set the value text.
InStream_Read_Bytes(instream, ptr + text_overlap, finish_chars_len);
CB_Set_Size(charbuf, total_text_len);
if (!StrHelp_utf8_valid(ptr, total_text_len)) {
THROW(ERR, "Invalid UTF-8 sequence in '%o' at byte %i64",
InStream_Get_Filename(instream),
InStream_Tell(instream) - finish_chars_len);
}
// Null-terminate.
ptr[total_text_len] = '\0';
// Invalidate string.
DECREF(ivars->string);
ivars->string = NULL;
}
static String*
S_encode_entities(String *text, CharBuf *buf) {
StringIterator *iter = Str_Top(text);
size_t space = 0;
const int MAX_ENTITY_BYTES = 9; // &#dddddd;
// Scan first so that we only allocate once.
int32_t code_point;
while (STRITER_DONE != (code_point = StrIter_Next(iter))) {
if (code_point > 127
|| (!isgraph(code_point) && !isspace(code_point))
|| code_point == '<'
|| code_point == '>'
|| code_point == '&'
|| code_point == '"'
) {
space += MAX_ENTITY_BYTES;
}
else {
space += 1;
}
}
CB_Grow(buf, space);
CB_Set_Size(buf, 0);
DECREF(iter);
iter = Str_Top(text);
while (STRITER_DONE != (code_point = StrIter_Next(iter))) {
if (code_point > 127
|| (!isgraph(code_point) && !isspace(code_point))
) {
CB_catf(buf, "&#%u32;", code_point);
}
else if (code_point == '<') {
CB_Cat_Trusted_Utf8(buf, "<", 4);
}
else if (code_point == '>') {
CB_Cat_Trusted_Utf8(buf, ">", 4);
}
else if (code_point == '&') {
CB_Cat_Trusted_Utf8(buf, "&", 5);
}
else if (code_point == '"') {
CB_Cat_Trusted_Utf8(buf, """, 6);
}
else {
CB_Cat_Char(buf, code_point);
}
}
DECREF(iter);
return CB_To_String(buf);
}
static void
S_vcat_mess(CharBuf *message, const char *file, int line, const char *func,
const char *pattern, va_list args)
{
size_t guess_len = strlen(file)
+ func ? strlen(func) : 0
+ strlen(pattern)
+ 30;
CB_Grow(message, guess_len);
CB_VCatF(message, pattern, args);
if (func != NULL)
CB_catf(message, " at %s:%i32 %s ", file, (i32_t)line, func);
else
CB_catf(message, " at %s:%i32", file, (i32_t)line);
}
void
StrHelp_add_indent(CharBuf *charbuf, size_t amount)
{
u32_t num_margins = 1;
size_t new_size;
char *limit = CBEND(charbuf);
char *source = charbuf->ptr;
char *dest;
/* Add a margin for every newline. */
for ( ; source < limit; source++) {
if (*source == '\n')
num_margins++;
}
/* Make space for margins. */
new_size = CB_Get_Size(charbuf) + (num_margins * amount);
CB_Grow(charbuf, new_size);
source = CBEND(charbuf);
CB_Set_Size(charbuf, new_size);
dest = CBEND(charbuf);
*dest-- = '\0';
source--;
while (source >= charbuf->ptr) {
if (*source == '\n') {
int i = amount;
while (i--) {
*dest-- = ' ';
}
}
*dest-- = *source--;
}
while (dest >= charbuf->ptr) {
*dest-- = ' ';
}
}
static void
S_grow(CharBuf *self, size_t size) {
if (size >= self->cap) {
CB_Grow(self, size);
}
}
