add GfmMedia and extend GfmLink with image-as-label form

- New GfmMedia parses `` with the full DokuWiki media-parameter
vocabulary in the URL slot (?100x200, ?right, ?nolink, ?recache,

add GfmMedia and extend GfmLink with image-as-label form

- New GfmMedia parses `` with the full DokuWiki media-parameter
vocabulary in the URL slot (?100x200, ?right, ?nolink, ?recache, …).
Adds `?left`/`?right`/`?center` align keywords shared with DW `{{…}}`
— gives pure-Markdown users a way to align inline images.
- GfmLink now also matches `[](target)` — the GFM equivalent
of `[[target|{{img}}]]`. Detection is post-entry, mirroring
Internallink's `^{{…}}$` check; one mode covers the whole family.
- LinkDispatch trait replaced by Helpers::classifyLink and
Helpers::parseMediaParameters — two pure static methods, shared by
DW and GFM counterparts.
- Entry patterns for GfmLink / GfmMedia simplified (permissive URL slot,
handle-time parsing), following DW's Internallink style.
- GfmSpecTest drives a test-only SpecCompatRenderer that emits bare
<img> / <a> instead of DW's wiki-wrapped HTML, recovering 13 spec
tests that previously failed/skipped only because of renderer shape.

show more ...

add GfmLink for GFM inline links `[text](url)`

Extracts the URL-classification ladder from Internallink into a
LinkDispatch trait so both modes route identically across all six
DokuWiki link flavors

add GfmLink for GFM inline links `[text](url)`

Extracts the URL-classification ladder from Internallink into a
LinkDispatch trait so both modes route identically across all six
DokuWiki link flavors (internal, external, interwiki, email,
windowsshare, local anchor). GfmLink parses the `[text](url)` form
with optional `"title"` / `'title'` and hands the URL to the trait.
The GFM title attribute is discarded — DokuWiki link instructions
have no slot for it.

show more ...

add GfmHeader for ATX headings (`# text` through `###### text`)

Opener must sit at column 0. GFM tolerates 0-3 spaces before the `#`
but that collides with DokuWiki's 2-space-indent preformatted blo

add GfmHeader for ATX headings (`# text` through `###### text`)

Opener must sit at column 0. GFM tolerates 0-3 spaces before the `#`
but that collides with DokuWiki's 2-space-indent preformatted block,
so the tolerance is dropped rather than plumbed across modes.

Widen the XHTML renderer's section-node tracker from 5 slots to 6 so
h6 doesn't hit "Undefined array key 5". Extend GfmSpecTest's HTML
normalizer to strip DokuWiki's section-div wrappers, section-edit
comments, and header id/class attributes so heading spec examples
can validate semantic correctness.

show more ...

add GfmBacktickSingle / GfmBacktickDouble for GFM inline code spans

Two new inline formatting modes covering GFM code spans in their n=1
and n=2 forms:

GfmBacktickSingle `text` → <code>text<

add GfmBacktickSingle / GfmBacktickDouble for GFM inline code spans

Two new inline formatting modes covering GFM code spans in their n=1
and n=2 forms:

GfmBacktickSingle `text` → <code>text</code>
GfmBacktickDouble ``text`` → <code>text</code>

Both emit monospace_open and monospace_close around an unformatted()
call (the same instruction shape as DokuWiki's two-single-quote pair
wrapping a nowiki span), so renderers that distinguish verbatim text
from plain cdata — metadata, indexer, non-XHTML backends — treat the
body as literal.

GfmBacktickDouble extends GfmBacktickSingle to reuse handle() and the
body-normalization helper; only the delimiter length and the body
character class differ. Both share sort 165 and gate on Markdown
being loaded.

Design notes:

* The lexer has no backreferences, so each length is its own mode.
Length-boundary guards (?<!`)...(?!`) on every opener and closer
ensure a run of two-or-more backticks is never read as an n=1
delimiter and a run of three-or-more is never read as n=2. The two
modes never steal each other's input regardless of registration
order — sort can't reach this kind of cross-position constraint.

* Edge-whitespace handling and newline normalization live in handle(),
not in the regex. On DOKU_LEXER_UNMATCHED the body is normalized:
1. CR/LF and LF become single spaces (GFM line-ending rule).
2. If the body starts and ends with a space and is not entirely
whitespace, one space is stripped from each end.
That produces the right GFM output for the tricky cases without
special-casing the entry pattern:
` ` → <code> </code> (all-whitespace, no strip)
` a` → <code> a</code> (asymmetric, no strip)
` `` ` → <code>``</code> (interior run-of-2 + strip)
``foo`bar`` → <code>foo`bar</code>

* Body character classes admit exactly the runs that cannot be valid
closers for this mode's length: n=1 allows `[^`] | ``+`, n=2 allows
`[^`] | `(?!`)`. That is what lets a single-backtick span contain
a pair and a double-backtick span contain a lone backtick.

* allowedModes is empty — no other inline parsing runs inside a span.

Deliberately not implemented, with skip.php entries explaining why:

351 — code-span precedence over emphasis (*foo`*` expected to render
as *foo<code>*</code>). Cross-positional: the single-pass
lexer matches leftmost-first and cannot reject an earlier
emphasis opener because a later backtick span would consume
its closer. A proper fix would need a pre-scan pass; sort
values only break ties at the same position.
353 — the trailing " outside the code span gets converted to a
curly quote by DokuWiki typography, diverging from spec HTML.
354 — raw HTML tag pass-through; DokuWiki does not render raw HTML
by default.
356 — GFM angle-bracket autolink <http://…>: not implemented.

Per-mode unit tests cover basic matching, flanking via the length-
boundary guards, interior-run support in the body, edge-space
stripping, newline normalization, all-whitespace bodies, paragraph-
boundary rejection, content-is-literal, and sort values.
ModeRegistryTest's gating data provider picks up both modes.

Net effect on GfmSpecTest: eleven previously-red code-span examples
now pass (339, 340, 341, 342, 344, 345, 346, 347, 349, 350, 357, 359
— the simple pairs, edge-space, interior-run, newline-normalization,
and mismatched-run cases). Four skipped. Three remain pending outside
the code-span scope (emphasis interactions that need GfmLink once
that lands).

show more ...

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 ...

add GfmDeleted mode for GFM strikethrough (`~~text~~`)

Shares the deleted_open/deleted_close instructions with DW's <del> mode.
Entry/exit anchors `(?<!~)` / `(?!~)` reject runs of three or more til

add GfmDeleted mode for GFM strikethrough (`~~text~~`)

Shares the deleted_open/deleted_close instructions with DW's <del> mode.
Entry/exit anchors `(?<!~)` / `(?!~)` reject runs of three or more tildes
so fenced-code markers remain untouched. Also trim redundant class-level
docblocks on sibling Gfm test files.

show more ...

add GFM em-wrapping-strong modes for `***foo***` / `___foo___`

Two new inline formatting modes that render triple-delimiter runs as
em wrapping strong:

GfmEmphasisStrong `***text***`

add GFM em-wrapping-strong modes for `***foo***` / `___foo___`

Two new inline formatting modes that render triple-delimiter runs as
em wrapping strong:

GfmEmphasisStrong `***text***` → <em><strong>text</strong></em>
GfmEmphasisStrongUnderscore `___text___` → same (MD-preferred only)

Only the exact 3+3 symmetric case is handled. The other long-run and
asymmetric variants (4+4, 5+5, `***foo**`, etc.) require CommonMark's
stack-based delimiter-pairing algorithm with its flanking and
multiple-of-3 rules, which is explicitly out of scope; those examples
stay skipped in gfm-spec/skip.php.

Implementation notes:

* Patterns enforce exact 3+3 via `(?<!\*)` / `(?<!_)` lookbehinds
(preventing entry at the second `*` of a `****...` run) and
`(?!\*)` / `(?!_)` lookaheads after the closing triple (rejecting
`***foo****` etc.). Combined with the existing non-whitespace
adjacency lookaheads, all asymmetric cases cleanly fall through to
other modes or stay literal.

* GfmEmphasisStrong overrides handle() to emit two instructions on
entry (emphasis_open + strong_open) and two on exit (strong_close
+ emphasis_close). GfmEmphasisStrongUnderscore inherits that
handler — only delimiters and word-boundary rules differ.

* Sort 65 — below Strong (70) and GfmEmphasis (80) so the em+strong
modes win the lexer race for `***`/`___` runs. Underscore variant
is MD-preferred-only, matching the existing gating of
GfmEmphasisUnderscore and GfmStrongUnderscore.

Per-mode unit tests cover basic matching, single-char bodies,
whitespace flanking rejection, paragraph-boundary rejection,
longer-run rejection, asymmetric rejection, multibyte intraword
protection, and sort values. ModeRegistryTest's gating data provider
picks up the two new rules.

Net effect on GfmSpecTest: example #476 (`***foo***`) now passes;
473/474/475/477 remain skipped as documented in skip.php.

show more ...

add GFM emphasis and underscore-delimited strong modes

Three new inline formatting modes for GitHub Flavored Markdown:

GfmEmphasis `*text*` → <em>
GfmEmphasisUnderscore `_text_`

add GFM emphasis and underscore-delimited strong modes

Three new inline formatting modes for GitHub Flavored Markdown:

GfmEmphasis `*text*` → <em>
GfmEmphasisUnderscore `_text_` → <em> (MD-preferred only)
GfmStrongUnderscore `__text__` → <strong> (MD-preferred only)

All three emit the same handler instructions as DokuWiki's Emphasis /
Strong, so existing renderers need no changes.

Design notes:

* Lexer mode names use snake_case (gfm_emphasis, gfm_emphasis_underscore,
gfm_strong_underscore) to keep PascalCase readable at the class level.
The asterisk variant emits `emphasis_open`/`emphasis_close` via the
getInstructionName() hook, so DW's Emphasis (`//...//`) and
GfmEmphasis (`*...*`) can coexist in mixed modes without a lexer
state collision while still producing the same <em> output.

* Underscore variants gate on Markdown-preferred syntax (`markdown`,
`md+dw`) because `__` otherwise means DW underline. GfmStrongUnderscore
sorts at 70 (matching Strong) — below Underline at 90 — so when loaded
it wins the lexer race for `__` runs. Underline is already gated out
of MD-preferred modes in the previous commit.

* Entry patterns enforce the simplified CommonMark flanking rules
already shared across DW inline modes (non-whitespace adjacency,
no paragraph-boundary crossing) plus the word-boundary check for
underscore variants using NO_WORD_BEFORE / NO_WORD_AFTER. The
positive non-word-char enumeration makes them multibyte-safe without
requiring the `u` flag: `für_etwas` and `пристаням_стремятся_`
correctly stay literal.

Per-mode unit tests cover basic matching, single-char bodies,
leading/trailing-whitespace rejection, empty-delimiter rejection,
paragraph-boundary rejection, multibyte intraword protection, and
sort values. ModeRegistryTest's gating data provider picks up the
three new rules.

show more ...

gate Underline on DokuWiki-preferred syntax; tidy registry plumbing

Three related changes to ModeRegistry, prep work for the Markdown modes
that follow.

1. Underline (`__text__`) is moved out of lo

gate Underline on DokuWiki-preferred syntax; tidy registry plumbing

Three related changes to ModeRegistry, prep work for the Markdown modes
that follow.

1. Underline (`__text__`) is moved out of loadAlwaysModes() and into
loadDokuWikiModes(), gated on a new `\$dwPreferred` check that
evaluates true for 'dokuwiki' and 'dw+md'. In MD-preferred settings
('markdown' and 'md+dw') `__` will mean GFM strong, so loading
Underline there would conflict at the lexer level. Underline is
unchanged in the default 'dokuwiki' setting.

2. resolveModeClass() now PascalCases every `_`-separated segment of
the mode name, so `gfm_emphasis_underscore` resolves to
`GfmEmphasisUnderscore`. Existing lowercase-compound names like
`internallink` still resolve to `Internallink` (one segment,
ucfirst-ed) — no behaviour change for current modes. This prepares
the registry to load Gfm mode classes whose PascalCase filenames
preserve word boundaries for readability.

3. ModeRegistryTest's multiple near-identical per-mode gating tests
are consolidated into a single data-provider-driven
testModeLoadingBySyntax, fed by a `\$rules` table that lists each
mode against its four-setting expected load state. Adding a new
gated mode now means one line in the provider. Currently only
Underline is listed; upcoming Gfm-mode commits will add theirs.

show more ...

add spec.txt-driven roundtrip test infrastructure for Markdown

Imports GitHub Flavored Markdown's test/spec.txt (671 examples combining
the CommonMark baseline with GFM extensions: tables, strikethr

add spec.txt-driven roundtrip test infrastructure for Markdown

Imports GitHub Flavored Markdown's test/spec.txt (671 examples combining
the CommonMark baseline with GFM extensions: tables, strikethrough, task
lists, autolink, disallowed raw HTML) and runs each example as a data-
provider-driven roundtrip test against DokuWiki's full parser + XHTML
renderer.

Files:

_test/tests/Parsing/Markdown/SpecReader.php
Parses the fenced-example format. Each example is ``` example
[optional-label] ... . ... ``` delimited by 10+ backticks with a
`.` separator; tracks most recent `## Heading` as section context
and numbers examples sequentially from 1 to match the spec's
rendered "Example N" labels.

_test/tests/Parsing/Markdown/SpecReaderTest.php
Hand-crafted fixtures covering ordinary examples, section tracking,
extension labels, multiline bodies, nested backticks, unclosed
fences (throws).

_test/tests/Parsing/Markdown/GfmSpecTest.php
Data-provider test. Renders each example's markdown through
p_get_instructions + p_render('xhtml') and compares to the expected
HTML with block-level-aware whitespace normalisation (DokuWiki
emits \n around block tags; inline-tag whitespace is preserved
because `<em>x</em> y` != `<em>x</em>y`).

_test/tests/Parsing/Markdown/gfm-spec/
spec.txt — verbatim from github/cmark-gfm, commit 587a12bb
LICENSE — CC-BY-SA 4.0 full legal text
README.md — upstream URL, pinned commit, resync notes
skip.php — map example-number => reason for SPEC-excluded
CommonMark behaviour (flanking-delimiter analysis,
multiple-of-3 rule, excess-drop). Unimplemented
features are NOT listed here — they show as real
failures so they remain the visible TODO list.

show more ...

require non-whitespace adjacency for inline formatting delimiters

An opening delimiter must now be followed by a non-whitespace character,
and a closing delimiter must be preceded by one. Empty deli

require non-whitespace adjacency for inline formatting delimiters

An opening delimiter must now be followed by a non-whitespace character,
and a closing delimiter must be preceded by one. Empty delimiter pairs
(****, ____, '''', , , <del></del>) no longer
match and stay literal.

Rationale: this matches Markdown's flanking-delimiter
rules and eliminates accidental bolding of sequences like `** note**`
at the start of a sentence. Well-formed uses (**bold**, //italic//,
__underline__) are unchanged.

Affected modes: Strong, Emphasis, Underline, Monospace, Subscript,
Superscript, Deleted.

BREAKING: content that was already malformed but
previously rendered as formatted (e.g. `**foo bar **`) now stays
literal.

show more ...

prevent inline formatting from matching across paragraph boundaries

The Lexer compiles all patterns with the `s` (DOTALL) flag via
ParallelRegex::getPerlMatchingFlags(), which makes `.` match newlin

prevent inline formatting from matching across paragraph boundaries

The Lexer compiles all patterns with the `s` (DOTALL) flag via
ParallelRegex::getPerlMatchingFlags(), which makes `.` match newlines.
Inline formatting modes use lookaheads like `\*\*(?=.*\*\*)` to verify
a closing delimiter exists, so with DOTALL a lone `**` happily matched
its "closer" many paragraphs later, swallowing blank lines into a
single <strong> run.

Add CONTENT_UNTIL_PARA on AbstractMode — a regex snippet matching any
character unless it would start a paragraph break (blank line, possibly
with horizontal whitespace). Update all inline formatting entry patterns
(Strong, Emphasis, Underline, Monospace, Subscript, Superscript,
Deleted) to use it in their closing-delimiter lookaheads.

Emphasis also gets a real closing-`//` check; its previous lookahead
just verified "content exists with a non-colon char" without requiring
the closing delimiter at all.

Single newlines inside a delimiter pair still match (multi-line
formatting); only blank lines end it.

BREAKING: This means you no longer can mark multiple paragraphs as bold
or strike them out. On the other hand it prevents accidentally breaking
the page layout by missing a closing delimiter (as reported many many
times over the years) eg. #1025 #3588 #1056

show more ...

add $conf['syntax'] setting and conditional mode loading in ModeRegistry

Introduce a new 'syntax' configuration setting (dokuwiki, markdown, dw+md, md+dw)
that controls which parser modes are loaded

add $conf['syntax'] setting and conditional mode loading in ModeRegistry

Introduce a new 'syntax' configuration setting (dokuwiki, markdown, dw+md, md+dw)
that controls which parser modes are loaded. Built-in modes are split into
always-loaded (no Markdown equivalent), DW-only, and MD-only groups.
Refactor getModes() into focused sub-methods for each group.

No Gfm mode classes exist yet, so only 'dokuwiki' is functional.
The change is a strict no-op for existing behavior.

show more ...

migrate parser tests to modern namespaced classes

Move old-style parser tests from _test/tests/inc/parser/ to namespaced
test classes under _test/tests/Parsing/ParserMode/ and
_test/tests/Parsing/Le

migrate parser tests to modern namespaced classes

Move old-style parser tests from _test/tests/inc/parser/ to namespaced
test classes under _test/tests/Parsing/ParserMode/ and
_test/tests/Parsing/Lexer/.

- Add ParserTestBase with assertCalls() helper for comparing handler
call sequences with byte index stripping
- Split lexer.test.php into ParallelRegexTest, StateStackTest, and
LexerTest with a RecordingHandler that extends Handler
- Merge handler_parse_highlight_options tests into CodeTest
- Add new tests for previously untested modes: Nocache, Notoc, Rss,
and all individual formatting modes (Strong, Emphasis, etc.)
- Modernize test code: [] syntax, lowercase null, correct assertEquals
argument order, replace deprecated withConsecutive and string callables
- Renderer tests remain in old location (renderers not yet migrated)

show more ...

move handler methods into ParserMode classes and rename Handler

Each ParserMode class now implements handle() from ModeInterface,
containing the token handling logic that previously lived as individ

move handler methods into ParserMode classes and rename Handler

Each ParserMode class now implements handle() from ModeInterface,
containing the token handling logic that previously lived as individual
methods on Doku_Handler.

The Handler class (formerly Doku_Handler) is the single dispatch point:
Lexer passes tokens to Handler::handleToken() which routes to mode
objects, plugins, or returns false. The Lexer only tokenizes and
resolves mapHandler aliases.

Key changes:
- Add handle() to ModeInterface, implemented by all mode classes
- Move Doku_Handler to dokuwiki\Parsing\Handler namespace
- File extends Code (shared parsing via $type property)
- Quotes uses mapHandler() + Handler::getModeName() for sub-modes
- Media::parseMedia() replaces Doku_Handler_Parse_Media()
- Code::parseHighlightOptions() replaces parse_highlight_options()
- Per-parse state (footnote, doublequote) stays on Handler
- Deprecated wrappers kept for base/header/internallink/media
- Class alias and rector rules added for backward compatibility

show more ...

decouple hardcoded mode names in Eol and Preformatted

Eol.php hardcoded ['listblock', 'table'] as modes to skip, and
Preformatted.php hardcoded [\*\-] as a negative lookahead for list
markers. Both

decouple hardcoded mode names in Eol and Preformatted

Eol.php hardcoded ['listblock', 'table'] as modes to skip, and
Preformatted.php hardcoded [\*\-] as a negative lookahead for list
markers. Both embed knowledge that belongs to the respective block
modes, not to Eol/Preformatted. Adding a new block mode that handles
its own EOL or uses different line start markers would require editing
these unrelated files — a hidden coupling.

Listblock and Table now register themselves on ModeRegistry during
preConnect(). Eol queries getBlockEolModes() and Preformatted queries
getLineStartMarkers() to build its lookahead dynamically. Each mode
owns its own data, and new block modes can participate without touching
unrelated files.

show more ...

add unit tests for ModeRegistry