aboutsummaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/environment.c4
-rw-r--r--src/parse.c61
-rw-r--r--src/stdlib/text.c37
3 files changed, 42 insertions, 60 deletions
diff --git a/src/environment.c b/src/environment.c
index 21d3db9b..8084758e 100644
--- a/src/environment.c
+++ b/src/environment.c
@@ -327,7 +327,7 @@ env_t *global_env(void)
{"at", "Text$cluster", "func(text:Text, index:Int -> Text)"},
{"by_line", "Text$by_line", "func(text:Text -> func(->Text?))"},
{"by_split", "Text$by_split", "func(text:Text, delimiter='' -> func(->Text?))"},
- {"by_split_any", "Text$by_split_any", "func(text:Text, delimiters=\" $\\t\\r\\n\" -> func(->Text?))"},
+ {"by_split_any", "Text$by_split_any", "func(text:Text, delimiters=' \\t\\r\\n' -> func(->Text?))"},
{"bytes", "Text$utf8_bytes", "func(text:Text -> [Byte])"},
{"caseless_equals", "Text$equal_ignoring_case", "func(a,b:Text, language='C' -> Bool)"},
{"codepoint_names", "Text$codepoint_names", "func(text:Text -> [Text])"},
@@ -351,7 +351,7 @@ env_t *global_env(void)
{"right_pad", "Text$right_pad", "func(text:Text, count:Int, pad=' ', language='C' -> Text)"},
{"slice", "Text$slice", "func(text:Text, from=1, to=-1 -> Text)"},
{"split", "Text$split", "func(text:Text, delimiter='' -> [Text])"},
- {"split_any", "Text$split_any", "func(text:Text, delimiters=\" $\\t\\r\\n\" -> [Text])"},
+ {"split_any", "Text$split_any", "func(text:Text, delimiters=' \\t\\r\\n' -> [Text])"},
{"starts_with", "Text$starts_with", "func(text,prefix:Text -> Bool)"},
{"title", "Text$title", "func(text:Text, language='C' -> Text)"},
{"to", "Text$to", "func(text:Text, last:Int -> Text)"},
diff --git a/src/parse.c b/src/parse.c
index c31bea64..b9a695b9 100644
--- a/src/parse.c
+++ b/src/parse.c
@@ -147,7 +147,7 @@ static PARSER(parse_var);
static PARSER(parse_when);
static PARSER(parse_while);
static PARSER(parse_deserialize);
-static ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open_quote, char close_quote, char open_interp);
+static ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open_quote, char close_quote, char open_interp, bool allow_escapes);
//
// Print a parse error and exit (or use the on_err longjmp)
@@ -255,6 +255,24 @@ static const char *unescape(parse_ctx_t *ctx, const char **out) {
char name[len+1];
memcpy(name, &escape[2], len);
name[len] = '\0';
+
+ if (name[0] == 'U') {
+ for (char *p = &name[1]; *p; p++) {
+ if (!isxdigit(*p)) goto look_up_unicode_name;
+ }
+ // Unicode codepoints by hex
+ char *endptr = NULL;
+ long codepoint = strtol(name+1, &endptr, 16);
+ uint32_t ustr[2] = {codepoint, 0};
+ size_t bufsize = 8;
+ uint8_t buf[bufsize];
+ (void)u32_to_u8(ustr, bufsize, buf, &bufsize);
+ *endpos = escape + 3 + len;
+ return GC_strndup((char*)buf, bufsize);
+ }
+
+ look_up_unicode_name:;
+
uint32_t codepoint = unicode_name_character(name);
if (codepoint == UNINAME_INVALID)
parser_err(ctx, escape, escape + 3 + len,
@@ -265,16 +283,6 @@ static const char *unescape(parse_ctx_t *ctx, const char **out) {
(void)u32_to_u8(&codepoint, 1, (uint8_t*)str, &u8_len);
str[u8_len] = '\0';
return str;
- } else if (escape[1] == 'U' && escape[2]) {
- // Unicode codepoints by hex
- char *endptr = NULL;
- long codepoint = strtol(escape+2, &endptr, 16);
- uint32_t ustr[2] = {codepoint, 0};
- size_t bufsize = 8;
- uint8_t buf[bufsize];
- (void)u32_to_u8(ustr, bufsize, buf, &bufsize);
- *endpos = endptr;
- return GC_strndup((char*)buf, bufsize);
} else if (escape[1] == 'x' && escape[2] && escape[3]) {
// ASCII 2-digit hex
char buf[] = {escape[2], escape[3], 0};
@@ -1187,7 +1195,7 @@ PARSER(parse_bool) {
return NULL;
}
-ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open_quote, char close_quote, char open_interp)
+ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open_quote, char close_quote, char open_interp, bool allow_escapes)
{
const char *pos = *out_pos;
int64_t starting_indent = get_indent(ctx, pos);
@@ -1203,7 +1211,7 @@ ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open
if (chunk) {
ast_t *literal = NewAST(ctx->file, chunk_start, pos, TextLiteral, .cord=chunk);
chunks = new(ast_list_t, .ast=literal, .next=chunks);
- chunk = NULL;
+ chunk = CORD_EMPTY;
}
++pos;
ast_t *interp;
@@ -1212,6 +1220,9 @@ ast_list_t *_parse_text_helper(parse_ctx_t *ctx, const char **out_pos, char open
interp = expect(ctx, interp_start, &pos, parse_term_no_suffix, "I expected an interpolation term here");
chunks = new(ast_list_t, .ast=interp, .next=chunks);
chunk_start = pos;
+ } else if (allow_escapes && *pos == '\\') {
+ const char *c = unescape(ctx, &pos);
+ chunk = CORD_cat(chunk, c);
} else if (!leading_newline && *pos == open_quote && closing[(int)open_quote]) { // Nested pair begin
if (get_indent(ctx, pos) == starting_indent) {
++depth;
@@ -1266,35 +1277,22 @@ PARSER(parse_text) {
const char *start = pos;
const char *lang = NULL;
- // Escape sequence, e.g. \r\n
- if (*pos == '\\') {
- CORD cord = CORD_EMPTY;
- do {
- const char *c = unescape(ctx, &pos);
- cord = CORD_cat(cord, c);
- // cord = CORD_cat_char(cord, c);
- } while (*pos == '\\');
- return NewAST(ctx->file, start, pos, TextLiteral, .cord=cord);
- }
-
char open_quote, close_quote, open_interp = '$';
if (match(&pos, "\"")) { // Double quote
open_quote = '"', close_quote = '"', open_interp = '$';
} else if (match(&pos, "`")) { // Backtick
open_quote = '`', close_quote = '`', open_interp = '$';
} else if (match(&pos, "'")) { // Single quote
- open_quote = '\'', close_quote = '\'', open_interp = '\x03';
+ open_quote = '\'', close_quote = '\'', open_interp = '$';
} else if (match(&pos, "$")) { // Customized strings
lang = get_id(&pos);
// $"..." or $@"...."
static const char *interp_chars = "~!@#$%^&*+=\\?";
- if (match(&pos, "$")) { // Disable interpolation with $
+ if (match(&pos, "$")) { // Disable interpolation with $$
open_interp = '\x03';
} else if (strchr(interp_chars, *pos)) {
open_interp = *pos;
++pos;
- } else if (*pos == '(') {
- open_interp = '@'; // For shell commands
}
static const char *quote_chars = "\"'`|/;([{<";
if (!strchr(quote_chars, *pos))
@@ -1306,7 +1304,8 @@ PARSER(parse_text) {
return NULL;
}
- ast_list_t *chunks = _parse_text_helper(ctx, &pos, open_quote, close_quote, open_interp);
+ bool allow_escapes = (open_quote != '`');
+ ast_list_t *chunks = _parse_text_helper(ctx, &pos, open_quote, close_quote, open_interp, allow_escapes);
return NewAST(ctx->file, start, pos, TextJoin, .lang=lang, .children=chunks);
}
@@ -2277,7 +2276,7 @@ PARSER(parse_inline_c) {
if (!match(&pos, "("))
parser_err(ctx, start, pos, "I expected a '(' here");
chunks = new(ast_list_t, .ast=NewAST(ctx->file, pos, pos, TextLiteral, "({"),
- .next=_parse_text_helper(ctx, &pos, '(', ')', '@'));
+ .next=_parse_text_helper(ctx, &pos, '(', ')', '@', false));
if (type) {
REVERSE_LIST(chunks);
chunks = new(ast_list_t, .ast=NewAST(ctx->file, pos, pos, TextLiteral, "; })"), .next=chunks);
@@ -2286,7 +2285,7 @@ PARSER(parse_inline_c) {
} else {
if (!match(&pos, "{"))
parser_err(ctx, start, pos, "I expected a '{' here");
- chunks = _parse_text_helper(ctx, &pos, '{', '}', '@');
+ chunks = _parse_text_helper(ctx, &pos, '{', '}', '@', false);
}
return NewAST(ctx->file, start, pos, InlineCCode, .chunks=chunks, .type_ast=type);
diff --git a/src/stdlib/text.c b/src/stdlib/text.c
index 621de942..b3e9cebb 100644
--- a/src/stdlib/text.c
+++ b/src/stdlib/text.c
@@ -1352,12 +1352,9 @@ public Text_t Text$quoted(Text_t text, bool colorize, Text_t quotation_mark)
int32_t quote_char = Text$get_grapheme(quotation_mark, 0);
#define add_escaped(str) ({ if (colorize) ret = concat2_assuming_safe(ret, Text("\x1b[34;1m")); \
- if (!just_escaped) ret = concat2_assuming_safe(ret, Text("$")); \
ret = concat2_assuming_safe(ret, Text("\\" str)); \
- just_escaped = true; \
if (colorize) ret = concat2_assuming_safe(ret, Text("\x1b[0;35m")); })
TextIter_t state = NEW_TEXT_ITER_STATE(text);
- bool just_escaped = false;
// TODO: optimize for spans of non-escaped text
for (int64_t i = 0; i < text.length; i++) {
int32_t g = Text$get_grapheme_fast(&state, i);
@@ -1371,21 +1368,11 @@ public Text_t Text$quoted(Text_t text, bool colorize, Text_t quotation_mark)
case '\t': add_escaped("t"); break;
case '\v': add_escaped("v"); break;
case '\\': {
- if (just_escaped) {
- add_escaped("\\");
- } else {
- ret = concat2_assuming_safe(ret, Text("\\"));
- just_escaped = false;
- }
+ add_escaped("\\");
break;
}
case '$': {
- if (quote_char == '\'') {
- ret = concat2_assuming_safe(ret, Text("$"));
- just_escaped = false;
- } else {
- add_escaped("$");
- }
+ add_escaped("$");
break;
}
case '\x00' ... '\x06': case '\x0E' ... '\x1A':
@@ -1397,7 +1384,6 @@ public Text_t Text$quoted(Text_t text, bool colorize, Text_t quotation_mark)
ret = concat2_assuming_safe(ret, Text$from_strn(tmp, 2));
if (colorize)
ret = concat2_assuming_safe(ret, Text("\x1b[0;35m"));
- just_escaped = true;
break;
}
default: {
@@ -1405,7 +1391,6 @@ public Text_t Text$quoted(Text_t text, bool colorize, Text_t quotation_mark)
ret = concat2_assuming_safe(ret, quotation_mark);
} else {
ret = concat2_assuming_safe(ret, Text$slice(text, I(i+1), I(i+1)));
- just_escaped = false;
}
break;
}
@@ -1427,14 +1412,12 @@ public Text_t Text$as_text(const void *vtext, bool colorize, const TypeInfo_t *i
Text_t text = *(Text_t*)vtext;
// Figure out the best quotation mark to use:
- bool has_dollar = false, has_double_quote = false, has_backtick = false,
+ bool has_double_quote = false, has_backtick = false,
has_single_quote = false, needs_escapes = false;
TextIter_t state = NEW_TEXT_ITER_STATE(text);
for (int64_t i = 0; i < text.length; i++) {
int32_t g = Text$get_grapheme_fast(&state, i);
- if (g == '$') {
- has_dollar = true;
- } else if (g == '"') {
+ if (g == '"') {
has_double_quote = true;
} else if (g == '`') {
has_backtick = true;
@@ -1444,15 +1427,15 @@ public Text_t Text$as_text(const void *vtext, bool colorize, const TypeInfo_t *i
}
Text_t quote;
- // If there's dollar signs and/or double quotes in the string, it would
- // be nice to avoid needing to escape them by using single quotes, but
- // only if we don't have single quotes or need to escape anything else
- // (because single quotes don't have interpolation):
- if ((has_dollar || has_double_quote) && !has_single_quote && !needs_escapes)
+ // If there's double quotes in the string, it would be nice to avoid
+ // needing to escape them by using single quotes, but only if we don't have
+ // single quotes or need to escape anything else (because single quotes
+ // don't have interpolation):
+ if (has_double_quote && !has_single_quote)
quote = Text("'");
// If there is a double quote, but no backtick, we can save a bit of
// escaping by using backtick instead of double quote:
- else if (has_double_quote && !has_backtick)
+ else if (has_double_quote && has_single_quote && !has_backtick && !needs_escapes)
quote = Text("`");
// Otherwise fall back to double quotes as the default quoting style:
else
generated by cgit v1.2.3 (git 2.25.1) at 2026-03-01 08:13:46 +0000