Skip to content

Reverse sync from upstream #3722

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
kddnewton merged 1 commit into from
Nov 16, 2025
Merged

Reverse sync from upstream #3722

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions config.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -217,8 +217,10 @@ errors:
- PARAMETER_UNEXPECTED_FWD
- PARAMETER_UNEXPECTED_NO_KW
- PARAMETER_WILD_LOOSE_COMMA
- PATTERN_ALTERNATIVE_AFTER_CAPTURE
- PATTERN_ARRAY_MULTIPLE_RESTS
- PATTERN_CAPTURE_DUPLICATE
- PATTERN_CAPTURE_IN_ALTERNATIVE
- PATTERN_EXPRESSION_AFTER_BRACKET
- PATTERN_EXPRESSION_AFTER_COMMA
- PATTERN_EXPRESSION_AFTER_HROCKET
Expand Down
74 changes: 48 additions & 26 deletions src/prism.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16947,6 +16947,16 @@ parse_strings(pm_parser_t *parser, pm_node_t *current, bool accepts_label, uint1
#define PM_PARSE_PATTERN_TOP 1
#define PM_PARSE_PATTERN_MULTI 2

/** Information used to track the state of captures in patterns. */
typedef struct {
/** Whether we're currently parsing an alternative pattern. This is used to
* disallow captures in alternative patterns. */
bool in_alternative_pattern;
/** Whether we've seen a capture in this pattern. This is used to disallow
* captures in alternative patterns. */
bool capture_in_pattern;
} pm_pattern_capturing_t;

static pm_node_t *
parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flags, pm_diagnostic_id_t diag_id, uint16_t depth);

Expand All @@ -16956,13 +16966,16 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
* an error to the parser.
*/
static void
parse_pattern_capture(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_constant_id_t capture, const pm_location_t *location) {
parse_pattern_capture(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_constant_id_t capture, const pm_location_t *location, pm_pattern_capturing_t *capturing) {
// Skip this capture if it starts with an underscore.
if (*location->start == '_') return;

if (pm_constant_id_list_includes(captures, capture)) {
pm_parser_err(parser, location->start, location->end, PM_ERR_PATTERN_CAPTURE_DUPLICATE);
} else if (capturing->in_alternative_pattern && parser->version >= PM_OPTIONS_VERSION_CRUBY_3_5) {
pm_parser_err(parser, location->start, location->end, PM_ERR_PATTERN_CAPTURE_IN_ALTERNATIVE);
} else {
capturing->capture_in_pattern = true;
pm_constant_id_list_append(captures, capture);
}
}
Expand Down Expand Up @@ -17091,7 +17104,7 @@ parse_pattern_constant_path(pm_parser_t *parser, pm_constant_id_list_t *captures
* Parse a rest pattern.
*/
static pm_splat_node_t *
parse_pattern_rest(pm_parser_t *parser, pm_constant_id_list_t *captures) {
parse_pattern_rest(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_pattern_capturing_t *capturing) {
assert(parser->previous.type == PM_TOKEN_USTAR);
pm_token_t operator = parser->previous;
pm_node_t *name = NULL;
Expand All @@ -17108,7 +17121,7 @@ parse_pattern_rest(pm_parser_t *parser, pm_constant_id_list_t *captures) {
pm_parser_local_add(parser, constant_id, identifier.start, identifier.end, 0);
}

parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&identifier));
parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&identifier), capturing);
name = (pm_node_t *) pm_local_variable_target_node_create(
parser,
&PM_LOCATION_TOKEN_VALUE(&identifier),
Expand All @@ -17125,7 +17138,7 @@ parse_pattern_rest(pm_parser_t *parser, pm_constant_id_list_t *captures) {
* Parse a keyword rest node.
*/
static pm_node_t *
parse_pattern_keyword_rest(pm_parser_t *parser, pm_constant_id_list_t *captures) {
parse_pattern_keyword_rest(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_pattern_capturing_t *capturing) {
assert(parser->current.type == PM_TOKEN_USTAR_STAR);
parser_lex(parser);

Expand All @@ -17144,7 +17157,7 @@ parse_pattern_keyword_rest(pm_parser_t *parser, pm_constant_id_list_t *captures)
pm_parser_local_add(parser, constant_id, parser->previous.start, parser->previous.end, 0);
}

parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous));
parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous), capturing);
value = (pm_node_t *) pm_local_variable_target_node_create(
parser,
&PM_LOCATION_TOKEN_VALUE(&parser->previous),
Expand Down Expand Up @@ -17188,7 +17201,7 @@ pm_slice_is_valid_local(const pm_parser_t *parser, const uint8_t *start, const u
* value. This will use an implicit local variable target.
*/
static pm_node_t *
parse_pattern_hash_implicit_value(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_symbol_node_t *key) {
parse_pattern_hash_implicit_value(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_symbol_node_t *key, pm_pattern_capturing_t *capturing) {
const pm_location_t *value_loc = &((pm_symbol_node_t *) key)->value_loc;

pm_constant_id_t constant_id = pm_parser_constant_id_location(parser, value_loc->start, value_loc->end);
Expand All @@ -17208,7 +17221,7 @@ parse_pattern_hash_implicit_value(pm_parser_t *parser, pm_constant_id_list_t *ca
pm_parser_local_add(parser, constant_id, value_loc->start, value_loc->end, 0);
}

parse_pattern_capture(parser, captures, constant_id, value_loc);
parse_pattern_capture(parser, captures, constant_id, value_loc, capturing);
pm_local_variable_target_node_t *target = pm_local_variable_target_node_create(
parser,
value_loc,
Expand All @@ -17234,7 +17247,7 @@ parse_pattern_hash_key(pm_parser_t *parser, pm_static_literals_t *keys, pm_node_
* Parse a hash pattern.
*/
static pm_hash_pattern_node_t *
parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node_t *first_node, uint16_t depth) {
parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node_t *first_node, pm_pattern_capturing_t *capturing, uint16_t depth) {
pm_node_list_t assocs = { 0 };
pm_static_literals_t keys = { 0 };
pm_node_t *rest = NULL;
Expand All @@ -17252,7 +17265,7 @@ parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node
if (match8(parser, PM_TOKEN_COMMA, PM_TOKEN_KEYWORD_THEN, PM_TOKEN_BRACE_RIGHT, PM_TOKEN_BRACKET_RIGHT, PM_TOKEN_PARENTHESIS_RIGHT, PM_TOKEN_NEWLINE, PM_TOKEN_SEMICOLON, PM_TOKEN_EOF)) {
// Otherwise, we will create an implicit local variable
// target for the value.
value = parse_pattern_hash_implicit_value(parser, captures, (pm_symbol_node_t *) first_node);
value = parse_pattern_hash_implicit_value(parser, captures, (pm_symbol_node_t *) first_node, capturing);
} else {
// Here we have a value for the first assoc in the list, so
// we will parse it now.
Expand Down Expand Up @@ -17296,7 +17309,7 @@ parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node
}

if (match1(parser, PM_TOKEN_USTAR_STAR)) {
pm_node_t *assoc = parse_pattern_keyword_rest(parser, captures);
pm_node_t *assoc = parse_pattern_keyword_rest(parser, captures, capturing);

if (rest == NULL) {
rest = assoc;
Expand Down Expand Up @@ -17324,7 +17337,7 @@ parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node
pm_node_t *value = NULL;

if (match7(parser, PM_TOKEN_COMMA, PM_TOKEN_KEYWORD_THEN, PM_TOKEN_BRACE_RIGHT, PM_TOKEN_BRACKET_RIGHT, PM_TOKEN_PARENTHESIS_RIGHT, PM_TOKEN_NEWLINE, PM_TOKEN_SEMICOLON)) {
value = parse_pattern_hash_implicit_value(parser, captures, (pm_symbol_node_t *) key);
value = parse_pattern_hash_implicit_value(parser, captures, (pm_symbol_node_t *) key, capturing);
} else {
value = parse_pattern(parser, captures, PM_PARSE_PATTERN_SINGLE, PM_ERR_PATTERN_EXPRESSION_AFTER_KEY, (uint16_t) (depth + 1));
}
Expand All @@ -17351,7 +17364,7 @@ parse_pattern_hash(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node
* Parse a pattern expression primitive.
*/
static pm_node_t *
parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_diagnostic_id_t diag_id, uint16_t depth) {
parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_pattern_capturing_t *capturing, pm_diagnostic_id_t diag_id, uint16_t depth) {
switch (parser->current.type) {
case PM_TOKEN_IDENTIFIER:
case PM_TOKEN_METHOD_NAME: {
Expand All @@ -17363,7 +17376,7 @@ parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm
pm_parser_local_add(parser, constant_id, parser->previous.start, parser->previous.end, 0);
}

parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous));
parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous), capturing);
return (pm_node_t *) pm_local_variable_target_node_create(
parser,
&PM_LOCATION_TOKEN_VALUE(&parser->previous),
Expand Down Expand Up @@ -17447,7 +17460,7 @@ parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm
first_node = (pm_node_t *) pm_symbol_node_label_create(parser, &parser->previous);
break;
case PM_TOKEN_USTAR_STAR:
first_node = parse_pattern_keyword_rest(parser, captures);
first_node = parse_pattern_keyword_rest(parser, captures, capturing);
break;
case PM_TOKEN_STRING_BEGIN:
first_node = parse_expression(parser, PM_BINDING_POWER_MAX, false, true, PM_ERR_PATTERN_HASH_KEY_LABEL, (uint16_t) (depth + 1));
Expand All @@ -17461,7 +17474,7 @@ parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm
}
}

node = parse_pattern_hash(parser, captures, first_node, (uint16_t) (depth + 1));
node = parse_pattern_hash(parser, captures, first_node, capturing, (uint16_t) (depth + 1));

accept1(parser, PM_TOKEN_NEWLINE);
expect1(parser, PM_TOKEN_BRACE_RIGHT, PM_ERR_PATTERN_TERM_BRACE);
Expand Down Expand Up @@ -17629,10 +17642,18 @@ parse_pattern_primitive(pm_parser_t *parser, pm_constant_id_list_t *captures, pm
static pm_node_t *
parse_pattern_primitives(pm_parser_t *parser, pm_constant_id_list_t *captures, pm_node_t *first_node, pm_diagnostic_id_t diag_id, uint16_t depth) {
pm_node_t *node = first_node;
pm_pattern_capturing_t capturing = { false, false };

while ((node == NULL) || accept1(parser, PM_TOKEN_PIPE)) {
pm_token_t operator = parser->previous;

if (node) {
if (capturing.capture_in_pattern) {
pm_parser_err(parser, operator.start, operator.end, PM_ERR_PATTERN_ALTERNATIVE_AFTER_CAPTURE);
}
capturing.in_alternative_pattern = true;
}

switch (parser->current.type) {
case PM_TOKEN_IDENTIFIER:
case PM_TOKEN_BRACKET_LEFT_ARRAY:
Expand All @@ -17644,9 +17665,9 @@ parse_pattern_primitives(pm_parser_t *parser, pm_constant_id_list_t *captures, p
case PM_TOKEN_UDOT_DOT_DOT:
case PM_CASE_PRIMITIVE: {
if (node == NULL) {
node = parse_pattern_primitive(parser, captures, diag_id, (uint16_t) (depth + 1));
node = parse_pattern_primitive(parser, captures, &capturing, diag_id, (uint16_t) (depth + 1));
} else {
pm_node_t *right = parse_pattern_primitive(parser, captures, PM_ERR_PATTERN_EXPRESSION_AFTER_PIPE, (uint16_t) (depth + 1));
pm_node_t *right = parse_pattern_primitive(parser, captures, &capturing, PM_ERR_PATTERN_EXPRESSION_AFTER_PIPE, (uint16_t) (depth + 1));
node = (pm_node_t *) pm_alternation_pattern_node_create(parser, node, right, &operator);
}

Expand Down Expand Up @@ -17698,7 +17719,7 @@ parse_pattern_primitives(pm_parser_t *parser, pm_constant_id_list_t *captures, p
pm_parser_local_add(parser, constant_id, parser->previous.start, parser->previous.end, 0);
}

parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous));
parse_pattern_capture(parser, captures, constant_id, &PM_LOCATION_TOKEN_VALUE(&parser->previous), &capturing);
pm_local_variable_target_node_t *target = pm_local_variable_target_node_create(
parser,
&PM_LOCATION_TOKEN_VALUE(&parser->previous),
Expand All @@ -17721,12 +17742,13 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag

bool leading_rest = false;
bool trailing_rest = false;
pm_pattern_capturing_t capturing = { false, false };

switch (parser->current.type) {
case PM_TOKEN_LABEL: {
parser_lex(parser);
pm_node_t *key = (pm_node_t *) pm_symbol_node_label_create(parser, &parser->previous);
node = (pm_node_t *) parse_pattern_hash(parser, captures, key, (uint16_t) (depth + 1));
node = (pm_node_t *) parse_pattern_hash(parser, captures, key, &capturing, (uint16_t) (depth + 1));

if (!(flags & PM_PARSE_PATTERN_TOP)) {
pm_parser_err_node(parser, node, PM_ERR_PATTERN_HASH_IMPLICIT);
Expand All @@ -17735,8 +17757,8 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
return node;
}
case PM_TOKEN_USTAR_STAR: {
node = parse_pattern_keyword_rest(parser, captures);
node = (pm_node_t *) parse_pattern_hash(parser, captures, node, (uint16_t) (depth + 1));
node = parse_pattern_keyword_rest(parser, captures, &capturing);
node = (pm_node_t *) parse_pattern_hash(parser, captures, node, &capturing, (uint16_t) (depth + 1));

if (!(flags & PM_PARSE_PATTERN_TOP)) {
pm_parser_err_node(parser, node, PM_ERR_PATTERN_HASH_IMPLICIT);
Expand All @@ -17747,10 +17769,10 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
case PM_TOKEN_STRING_BEGIN: {
// We need special handling for string beginnings because they could
// be dynamic symbols leading to hash patterns.
node = parse_pattern_primitive(parser, captures, diag_id, (uint16_t) (depth + 1));
node = parse_pattern_primitive(parser, captures, &capturing, diag_id, (uint16_t) (depth + 1));

if (pm_symbol_node_label_p(node)) {
node = (pm_node_t *) parse_pattern_hash(parser, captures, node, (uint16_t) (depth + 1));
node = (pm_node_t *) parse_pattern_hash(parser, captures, node, &capturing, (uint16_t) (depth + 1));

if (!(flags & PM_PARSE_PATTERN_TOP)) {
pm_parser_err_node(parser, node, PM_ERR_PATTERN_HASH_IMPLICIT);
Expand All @@ -17765,7 +17787,7 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
case PM_TOKEN_USTAR: {
if (flags & (PM_PARSE_PATTERN_TOP | PM_PARSE_PATTERN_MULTI)) {
parser_lex(parser);
node = (pm_node_t *) parse_pattern_rest(parser, captures);
node = (pm_node_t *) parse_pattern_rest(parser, captures, &capturing);
leading_rest = true;
break;
}
Expand All @@ -17779,7 +17801,7 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
// If we got a dynamic label symbol, then we need to treat it like the
// beginning of a hash pattern.
if (pm_symbol_node_label_p(node)) {
return (pm_node_t *) parse_pattern_hash(parser, captures, node, (uint16_t) (depth + 1));
return (pm_node_t *) parse_pattern_hash(parser, captures, node, &capturing, (uint16_t) (depth + 1));
}

if ((flags & PM_PARSE_PATTERN_MULTI) && match1(parser, PM_TOKEN_COMMA)) {
Expand All @@ -17800,7 +17822,7 @@ parse_pattern(pm_parser_t *parser, pm_constant_id_list_t *captures, uint8_t flag
}

if (accept1(parser, PM_TOKEN_USTAR)) {
node = (pm_node_t *) parse_pattern_rest(parser, captures);
node = (pm_node_t *) parse_pattern_rest(parser, captures, &capturing);

// If we have already parsed a splat pattern, then this is an
// error. We will continue to parse the rest of the patterns,
Expand Down
2 changes: 2 additions & 0 deletions templates/src/diagnostic.c.erb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -299,8 +299,10 @@ static const pm_diagnostic_data_t diagnostic_messages[PM_DIAGNOSTIC_ID_MAX] = {
[PM_ERR_PARAMETER_UNEXPECTED_FWD] = { "unexpected `...` in parameters", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PARAMETER_WILD_LOOSE_COMMA] = { "unexpected `,` in parameters", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PARAMETER_UNEXPECTED_NO_KW] = { "unexpected **nil; no keywords marker disallowed after keywords", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_ALTERNATIVE_AFTER_CAPTURE] = { "alternative pattern after variable capture", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_ARRAY_MULTIPLE_RESTS] = { "unexpected multiple '*' rest patterns in an array pattern", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_CAPTURE_DUPLICATE] = { "duplicated variable name", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_CAPTURE_IN_ALTERNATIVE] = { "variable capture in alternative pattern", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_EXPRESSION_AFTER_BRACKET] = { "expected a pattern expression after the `[` operator", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_EXPRESSION_AFTER_COMMA] = { "expected a pattern expression after `,`", PM_ERROR_LEVEL_SYNTAX },
[PM_ERR_PATTERN_EXPRESSION_AFTER_HROCKET] = { "expected a pattern expression after `=>`", PM_ERROR_LEVEL_SYNTAX },
Expand Down
2 changes: 1 addition & 1 deletion test/prism/result/source_location_test.rb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ def test_AliasMethodNode
end

def test_AlternationPatternNode
assert_location(AlternationPatternNode, "foo => bar | baz", 7...16, &:pattern)
assert_location(AlternationPatternNode, "foo => 0 | 1", 7...12, &:pattern)
end

def test_AndNode
Expand Down