Merge branch 'loops/hs' into hs-f (E37 tokenizer, E40 fetch, DOM ref-eq, DOM tree fixes)

This commit is contained in:
2026-04-26 17:57:37 +00:00
10 changed files with 3556 additions and 3057 deletions

View File

@@ -125,19 +125,9 @@ SKIP_TEST_NAMES = {
"can ignore when target doesn't exist",
"can ignore when target doesn\\'t exist",
"can handle an or after a from clause",
# upstream 'fetch' category — depend on per-test sinon stubs for 404 / thrown errors,
# or on real DocumentFragment semantics (`its childElementCount` after `as html`).
# Our generic test-runner mock returns a fixed 200 response, so these cases
# (non-2xx handling, error path, before-fetch event, real DOM fragment) can't be
# exercised here.
# upstream 'fetch' category — real DocumentFragment semantics (`its childElementCount`
# after `as html`) not exercisable with our DOM mock.
"can do a simple fetch w/ html",
"triggers an event just before fetching",
"can catch an error that occurs when using fetch",
"throws on non-2xx response by default",
"do not throw passes through 404 response",
"don't throw passes through 404 response",
"as response does not throw on 404",
"Response can be converted to JSON via as JSON",
}
# Manually-written SX test bodies for tests whose upstream body cannot be
@@ -249,11 +239,18 @@ def parse_html(html):
# button HTML in `properly processes hyperscript X` tests). HTMLParser handles
# backslashes in attribute values as literal characters, so we leave them.
# HTML5 void elements — never have children, auto-pop from stack immediately.
VOID_TAGS = {'area','base','br','col','embed','hr','img','input','link',
'meta','param','source','track','wbr'}
elements = []
stack = []
class Parser(HTMLParser):
def handle_starttag(self, tag, attrs):
# Pop any void elements left on the stack (they have no close tag).
while stack and stack[-1]['tag'] in VOID_TAGS:
stack.pop()
el = {
'tag': tag, 'id': None, 'classes': [], 'hs': None,
'attrs': {}, 'inner': '', 'depth': len(stack),
@@ -283,6 +280,9 @@ def parse_html(html):
elements.append(el)
def handle_endtag(self, tag):
# Pop void elements first (they don't have close tags but may linger).
while stack and stack[-1]['tag'] in VOID_TAGS:
stack.pop()
if stack and stack[-1]['tag'] == tag:
stack.pop()
@@ -1002,6 +1002,24 @@ def parse_dev_body(body, elements, var_names):
else:
pre_setups.append(('__hs_config__', op_expr))
continue
# window.addEventListener(EVT, (param) => { param.target.PROP = 'VAL'; })
wa = re.search(
r"window\.addEventListener\(\s*(['\"])([^'\"]+)\1\s*,\s*"
r"\((\w+)\)\s*=>\s*\{\s*\3\.target\.(\w+)\s*=\s*['\"]([^'\"]+)['\"]\s*;?\s*\}",
m.group(1),
)
if wa:
ev_name = wa.group(2)
prop = wa.group(4)
val = wa.group(5)
attr = 'class' if prop == 'className' else prop
sx = (f'(host-call (host-global "window") "addEventListener" "{ev_name}" '
f'(fn (_event) (dom-set-attr (host-get _event "target") "{attr}" "{val}")))')
if seen_html:
ops.append(sx)
else:
pre_setups.append(('__hs_config__', sx))
continue
# fall through
# evaluate(() => _hyperscript.config.X = ...) single-line variant.
@@ -1293,7 +1311,9 @@ def process_hs_val(hs_val):
hs_val = hs_val.replace('\\n', '\n').replace('\\t', ' ')
# Preserve escaped quotes (\" → placeholder), strip remaining backslashes, restore
hs_val = hs_val.replace('\\"', '\x00QUOT\x00')
hs_val = hs_val.replace('\\$', '\x00DOLLAR\x00') # preserve \$ template escape
hs_val = hs_val.replace('\\', '')
hs_val = hs_val.replace('\x00DOLLAR\x00', '\\$') # restore \$
hs_val = hs_val.replace('\x00QUOT\x00', '\\"')
# Strip line comments BEFORE newline collapse — once newlines become `then`,
# an unterminated `//` / ` --` comment would consume the rest of the input.
@@ -1705,6 +1725,13 @@ def js_expr_to_sx(expr):
if s is None:
return None
arg_sx.append(s)
# Translate common array HO methods to SX primitives so SX lists work.
if method == 'reduce' and len(arg_sx) == 2:
return f'(reduce {arg_sx[0]} {arg_sx[1]} {obj})'
if method == 'map' and len(arg_sx) == 1:
return f'(map {arg_sx[0]} {obj})'
if method == 'filter' and len(arg_sx) == 1:
return f'(filter {arg_sx[0]} {obj})'
return f'(host-call {obj} "{method}" {" ".join(arg_sx)})'.strip()
# Property access: o.prop
@@ -1877,6 +1904,272 @@ def extract_hs_expr(raw):
return expr
def generate_tokenizer_test(test, safe_name):
"""Hardcoded SX translation for _hyperscript.internals.tokenizer tests (E37)."""
name = test['name']
def to_(src, tmpl=False):
"""Return (hs-tokens-of <sx-str> [:template]) for HS source string src."""
escaped = (src
.replace('\\', '\\\\')
.replace('"', '\\"')
.replace('\n', '\\n')
.replace('\r', '\\r')
.replace('\t', '\\t'))
q = '"' + escaped + '"'
suffix = ' :template' if tmpl else ''
return f'(hs-tokens-of {q}{suffix})'
def consume(s):
return f'(hs-stream-consume {s})'
def tok_i(s, i):
return f'(hs-stream-token {s} {i})'
def has_more(s):
return f'(hs-stream-has-more {s})'
def t_type(t):
return f'(hs-token-type {t})'
def t_val(t):
return f'(hs-token-value {t})'
def t_op(t):
return f'(hs-token-op? {t})'
def nth_list(s, i):
return f'(nth (get {s} "list") {i})'
def list_len(s):
return f'(len (get {s} "list"))'
def ae(actual, expected):
return f' (assert= {actual} {expected})'
def throws(expr):
return (
f' (let ((threw false))\n'
f' (guard (e (true (set! threw true))) {expr})\n'
f' (assert threw))'
)
lines = [f' (deftest "{safe_name}"']
if name == 'handles $ in template properly':
s = to_('"', tmpl=True)
lines.append(ae(t_val(tok_i(s, 0)), sx_str('"')))
elif name == 'handles all special escapes properly':
for src, exp in [
('"\\b"', '(char-from-code 8)'),
('"\\f"', '(char-from-code 12)'),
('"\\n"', '"\\n"'),
('"\\r"', '"\\r"'),
('"\\t"', '"\\t"'),
('"\\v"', '(char-from-code 11)'),
]:
lines.append(ae(t_val(consume(to_(src))), exp))
elif name == 'handles basic token types':
lines.append(ae(t_type(consume(to_('foo'))), '"IDENTIFIER"'))
lines.append(ae(t_type(consume(to_('1'))), '"NUMBER"'))
for src in ['1.1', '1e6', '1e-6', '1.1e6', '1.1e-6']:
sq = to_(src)
lines.append(f' (let ((s {sq}))')
lines.append(f' (let ((tok (hs-stream-consume s)))')
lines.append(f' (assert= (hs-token-type tok) "NUMBER")')
lines.append(f' (assert= (hs-stream-has-more s) false)))')
lines.append(ae(t_type(consume(to_('.a'))), '"CLASS_REF"'))
lines.append(ae(t_type(consume(to_('#a'))), '"ID_REF"'))
lines.append(ae(t_type(consume(to_('"asdf"'))), '"STRING"'))
elif name == 'handles class identifiers properly':
for src, idx, exp_type, exp_val in [
('.a', None, 'CLASS_REF', '.a'),
(' .a', None, 'CLASS_REF', '.a'),
('a.a', None, 'IDENTIFIER', 'a'),
('(a).a', 4, 'IDENTIFIER', 'a'),
('{a}.a', 4, 'IDENTIFIER', 'a'),
('[a].a', 4, 'IDENTIFIER', 'a'),
('(a(.a', 3, 'CLASS_REF', '.a'),
('{a{.a', 3, 'CLASS_REF', '.a'),
('[a[.a', 3, 'CLASS_REF', '.a'),
]:
if idx is None:
tok_expr = consume(to_(src))
else:
tok_expr = nth_list(to_(src), idx)
lines.append(ae(t_type(tok_expr), f'"{exp_type}"'))
lines.append(ae(t_val(tok_expr), sx_str(exp_val)))
elif name == 'handles comments properly':
for src, expected in [
('--', 0),
('asdf--', 1),
('-- asdf', 0),
('--\nasdf', 1),
('--\nasdf--', 1),
('---asdf', 0),
('----\n---asdf', 0),
('----asdf----', 0),
('---\nasdf---', 1),
('// asdf', 0),
('///asdf', 0),
('asdf//', 1),
('asdf\n//', 2),
]:
lines.append(ae(list_len(to_(src)), str(expected)))
elif name == 'handles hex escapes properly':
lines.append(ae(t_val(consume(to_('"\\x1f"'))), '(char-from-code 31)'))
lines.append(ae(t_val(consume(to_('"\\x41"'))), '"A"'))
lines.append(ae(t_val(consume(to_('"\\x41\\x61"'))), '"Aa"'))
for bad in ['"\\x"', '"\\xGG"', '"\\x4"']:
lines.append(throws(consume(to_(bad))))
elif name == 'handles id references properly':
for src, idx, exp_type, exp_val in [
('#a', None, 'ID_REF', '#a'),
(' #a', None, 'ID_REF', '#a'),
('a#a', None, 'IDENTIFIER', 'a'),
('(a)#a', 4, 'IDENTIFIER', 'a'),
('{a}#a', 4, 'IDENTIFIER', 'a'),
('[a]#a', 4, 'IDENTIFIER', 'a'),
('(a(#a', 3, 'ID_REF', '#a'),
('{a{#a', 3, 'ID_REF', '#a'),
('[a[#a', 3, 'ID_REF', '#a'),
]:
if idx is None:
tok_expr = consume(to_(src))
else:
tok_expr = nth_list(to_(src), idx)
lines.append(ae(t_type(tok_expr), f'"{exp_type}"'))
lines.append(ae(t_val(tok_expr), sx_str(exp_val)))
elif name == 'handles identifiers properly':
lines.append(ae(t_type(consume(to_('foo'))), '"IDENTIFIER"'))
lines.append(ae(t_val(consume(to_('foo'))), '"foo"'))
lines.append(ae(t_type(consume(to_(' foo '))), '"IDENTIFIER"'))
lines.append(ae(t_val(consume(to_(' foo '))), '"foo"'))
for src, v1, v2 in [
(' foo bar', 'foo', 'bar'),
(' foo\n-- a comment\n bar', 'foo', 'bar'),
]:
sq = to_(src)
lines.append(f' (let ((s {sq}))')
lines.append(f' (let ((tok1 (hs-stream-consume s)))')
lines.append(f' (assert= (hs-token-type tok1) "IDENTIFIER")')
lines.append(f' (assert= (hs-token-value tok1) {sx_str(v1)})')
lines.append(f' (let ((tok2 (hs-stream-consume s)))')
lines.append(f' (assert= (hs-token-type tok2) "IDENTIFIER")')
lines.append(f' (assert= (hs-token-value tok2) {sx_str(v2)}))))')
elif name == 'handles identifiers with numbers properly':
for src in ['f1oo', 'fo1o', 'foo1']:
lines.append(ae(t_type(consume(to_(src))), '"IDENTIFIER"'))
lines.append(ae(t_val(consume(to_(src))), sx_str(src)))
elif name == 'handles look ahead property':
s = to_('a 1 + 1')
for i, v in [(0, 'a'), (1, '1'), (2, '+'), (3, '1'), (4, '<<<EOF>>>')]:
lines.append(ae(t_val(tok_i(s, i)), sx_str(v)))
elif name == 'handles numbers properly':
for src, v in [
('1', '1'),
('1.1', '1.1'),
('1234567890.1234567890', '1234567890.1234567890'),
('1e6', '1e6'),
('1e-6', '1e-6'),
('1.1e6', '1.1e6'),
('1.1e-6', '1.1e-6'),
]:
lines.append(ae(t_type(consume(to_(src))), '"NUMBER"'))
lines.append(ae(t_val(consume(to_(src))), sx_str(v)))
s = to_('1.1.1')
toks = f'(get {s} "list")'
lines.append(ae(f'(hs-token-type (nth {toks} 0))', '"NUMBER"'))
lines.append(ae(f'(hs-token-type (nth {toks} 1))', '"PERIOD"'))
lines.append(ae(f'(hs-token-type (nth {toks} 2))', '"NUMBER"'))
lines.append(ae(f'(len {toks})', '3'))
elif name == 'handles operators properly':
optable = [
('+', 'PLUS'), ('-', 'MINUS'), ('*', 'MULTIPLY'),
('.', 'PERIOD'), ('\\', 'BACKSLASH'), (':', 'COLON'),
('%', 'PERCENT'), ('|', 'PIPE'), ('!', 'EXCLAMATION'),
('?', 'QUESTION'), ('#', 'POUND'), ('&', 'AMPERSAND'),
(';', 'SEMI'), (',', 'COMMA'), ('(', 'L_PAREN'),
(')', 'R_PAREN'), ('<', 'L_ANG'), ('>', 'R_ANG'),
('{', 'L_BRACE'), ('}', 'R_BRACE'), ('[', 'L_BRACKET'),
(']', 'R_BRACKET'), ('=', 'EQUALS'),
('<=', 'LTE_ANG'), ('>=', 'GTE_ANG'),
('==', 'EQ'), ('===', 'EQQ'),
]
for op_char, _op_name in optable:
tok_expr = consume(to_(op_char))
lines.append(ae(t_op(tok_expr), 'true'))
lines.append(ae(t_val(tok_expr), sx_str(op_char)))
elif name == 'handles strings properly':
for src, v in [
('"foo"', 'foo'),
('"fo\'o"', "fo'o"),
('"fo\\"o"', 'fo"o'),
("'foo'", 'foo'),
("'fo\"o'", 'fo"o'),
("'fo\\'o'", "fo'o"),
]:
lines.append(ae(t_type(consume(to_(src))), '"STRING"'))
lines.append(ae(t_val(consume(to_(src))), sx_str(v)))
lines.append(throws(consume(to_("'"))))
lines.append(throws(consume(to_('"'))))
elif name == 'handles strings properly 2':
tok_expr = consume(to_("'foo'"))
lines.append(ae(t_type(tok_expr), '"STRING"'))
lines.append(ae(t_val(tok_expr), '"foo"'))
elif name == 'handles template bootstrap properly':
s1 = to_('"', tmpl=True)
lines.append(ae(t_val(tok_i(s1, 0)), sx_str('"')))
s2 = to_('"$', tmpl=True)
lines.append(ae(t_val(tok_i(s2, 0)), sx_str('"')))
lines.append(ae(t_val(tok_i(s2, 1)), '"$"'))
s3 = to_('"${', tmpl=True)
lines.append(ae(t_val(tok_i(s3, 0)), sx_str('"')))
lines.append(ae(t_val(tok_i(s3, 1)), '"$"'))
lines.append(ae(t_val(tok_i(s3, 2)), '"{"'))
s4 = to_('"${"asdf"', tmpl=True)
lines.append(ae(t_val(tok_i(s4, 0)), sx_str('"')))
lines.append(ae(t_val(tok_i(s4, 1)), '"$"'))
lines.append(ae(t_val(tok_i(s4, 2)), '"{"'))
lines.append(ae(t_val(tok_i(s4, 3)), '"asdf"'))
s5 = to_('"${"asdf"}"', tmpl=True)
lines.append(ae(t_val(tok_i(s5, 0)), sx_str('"')))
lines.append(ae(t_val(tok_i(s5, 1)), '"$"'))
lines.append(ae(t_val(tok_i(s5, 2)), '"{"'))
lines.append(ae(t_val(tok_i(s5, 3)), '"asdf"'))
lines.append(ae(t_val(tok_i(s5, 4)), '"}"'))
lines.append(ae(t_val(tok_i(s5, 5)), sx_str('"')))
elif name == 'handles whitespace properly':
for src, expected in [
(' ', 0), (' asdf', 1), (' asdf ', 2), ('asdf ', 2),
('\n', 0), ('\nasdf', 1), ('\nasdf\n', 2), ('asdf\n', 2),
('\r', 0), ('\rasdf', 1), ('\rasdf\r', 2), ('asdf\r', 2),
('\t', 0), ('\tasdf', 1), ('\tasdf\t', 2), ('asdf\t', 2),
]:
lines.append(ae(list_len(to_(src)), str(expected)))
else:
return None # not a tokenizer test we handle
lines.append(' )')
return '\n'.join(lines)
def generate_eval_only_test(test, idx):
"""Generate SX deftest for no-HTML tests using eval-hs.
Handles patterns:
@@ -2095,6 +2388,9 @@ def generate_eval_only_test(test, idx):
f" (assert= (hs-line-at \"{src}\" (list :true-branch :next)) \" log 'it was true'\"))"
)
if '_hyperscript.internals.tokenizer' in body:
return generate_tokenizer_test(test, safe_name)
lines.append(f' (deftest "{safe_name}"')
assertions = []
@@ -2106,13 +2402,20 @@ def generate_eval_only_test(test, idx):
def emit_eval(hs_expr, expected_sx, extra_locals=None):
"""Emit an assertion using eval-hs / eval-hs-locals / eval-hs-with-me
as appropriate, given the window setups and any per-call locals.
Uses assert-equal (deep equal?) when expected contains dicts; assert= otherwise.
"""
pairs = list(window_setups) + list(extra_locals or [])
# assert= uses = (reference equality for dicts); assert-equal uses equal? (deep)
use_deep = '{' in expected_sx
if pairs:
locals_sx = '(list ' + ' '.join(
f'(list (quote {n}) {v})' for n, v in pairs
) + ')'
if use_deep:
return f' (assert-equal {expected_sx} (eval-hs-locals "{hs_expr}" {locals_sx}))'
return f' (assert= (eval-hs-locals "{hs_expr}" {locals_sx}) {expected_sx})'
if use_deep:
return f' (assert-equal {expected_sx} (eval-hs "{hs_expr}"))'
return f' (assert= (eval-hs "{hs_expr}") {expected_sx})'
# Shared sub-pattern for run() call with optional String.raw and extra args: