| 864d6c6d | 21-Apr-2026 |
Andreas Gohr <gohr@cosmocode.de> |
fix Lexer so exit-pattern lookbehinds see chars consumed by prior tokens
Lexer::reduce used to hand PCRE a shrinking tail of the subject — each
matched token was chopped off the front of $raw and th
fix Lexer so exit-pattern lookbehinds see chars consumed by prior tokens
Lexer::reduce used to hand PCRE a shrinking tail of the subject — each
matched token was chopped off the front of $raw and the next preg_match
ran on what remained. Once a token was consumed, the bytes before the
cursor were gone, and any lookbehind assertion in a subsequent pattern
silently failed.
The bug was latent for DokuWiki's entire history because literal exit
patterns like `\*\*`, `</file>`, or `%%` don't care what's behind them.
It surfaced with c3755410a ("require non-whitespace adjacency for
inline formatting delimiters"), which added `(?<=[^\s])` to Strong,
Emphasis, Underline, Monospace, Subscript, Superscript and Deleted at
once. After that commit, `**[[link]]**` stopped closing — the `]` that
would satisfy the lookbehind had just been consumed by the link match,
so Strong stayed open until end-of-section and swallowed everything
after it (list items, headings, the lot).
Fix:
* Lexer::parse and Lexer::reduce track a byte offset into $raw instead
of mutating $raw. $initialLength and the shrinking-length arithmetic
for absolute match positions are replaced by straight offset
increments; the no-progress guard and the trailing-unmatched dispatch
both shift to the same cursor.
* ParallelRegex::split takes an optional $offset and passes it to
preg_match together with PREG_OFFSET_CAPTURE. PCRE scans from the
offset forward but still sees the whole subject, so lookbehinds work
across already-consumed tokens. The secondary preg_split call used
to carve out pre/post is no longer needed — PREG_OFFSET_CAPTURE
gives the match start for free, saving one regex operation per
reduce() step.
Regression tests at all three layers:
* ParallelRegexTest — offset plumbing and pre/match accounting.
* LexerTest::testIndexLookbehindAcrossConsumedToken — exit-pattern
lookbehind targeting the `/>` of a self-closing `<a/>` that was
consumed as a SPECIAL token on the previous step. Fails under the
old Lexer.
* FormattingTest — `**[[link]]**` and `**foo//bar//**` round-trip
with correct open/close instructions through the full pipeline.
Also updates ListsTest::testUnorderedListStrong, whose expectations
documented the pre-fix buggy behaviour ("formatting able to spread
across list items"). With the fix, bold correctly stays within a
single list item; the expected call sequence and the comment are
updated to match.
show more ...
|
| 8ab4ec30 | 16-Apr-2026 |
Andreas Gohr <gohr@cosmocode.de> |
remove dead ParallelRegex::apply() method
Remove apply() which was never called from production code. Rewrite
the inherited SimpleTest tests to use split() instead, and add a
test for pre/post-match
remove dead ParallelRegex::apply() method
Remove apply() which was never called from production code. Rewrite
the inherited SimpleTest tests to use split() instead, and add a
test for pre/post-match splitting.
show more ...
|