Changes in ModeRegistry.php

2d05a06d7893e8c100d5357c6301ffae08b00522 - fix(info): read parser modes from the registry, not the deprecated global

Wed, 08 Jul 2026 09:59:27 +0000

fix(info): read parser modes from the registry, not the deprecated globalThe info plugin's ~~INFO:syntaxtypes~~ read the $PARSER_MODES globaldirectly. Since the mode taxonomy moved into ModeRegistry (c8dd1b9d2),that global is only populated as a side effect of a parse, so on a warminstruction cache with no parse in the request it is unset and thesyntax-types table renders empty. ~~INFO:syntaxmodes~~ used thedeprecated p_get_parsermodes(), logging a deprecation on every render.Both now build a ModeRegistry and read the categories and modes from itdirectly, independent of parse or cache state. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

a7e1021671c06230d0a9890646b3531d28d5d29a - �� Rector and PHPCS fixes

Thu, 25 Jun 2026 10:48:18 +0000

�� Rector and PHPCS fixes List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

47a02a102092be9e1e6f1ddaf158bdfffdb13d4f - Parsing: make parse syntax a per-parse value, drop ModeInterface

Thu, 04 Jun 2026 12:27:59 +0000

Parsing: make parse syntax a per-parse value, drop ModeInterfaceThe active parse's syntax flavour is a per-parse question, not process-global state: within a single request a plugin can render bundledDokuWiki-syntax text inside an otherwise-Markdown page. Yet ModeRegistrywas a singleton that read $conf['syntax'] and the $PARSER_MODES global,and every mode reached it through ModeRegistry::getInstance() — so theflavour lived in shared mutable state that two parses in one requestwould fight over.Make the registry a short-lived value instead:- ModeRegistry is constructed once per parse with an explicit $syntax and injected into Parser, Handler and every mode. getSyntax() / isDwPreferred() / isMdPreferred() consult $this->syntax; the DOKU_UNITTEST-gated mode-list cache hack is gone (each registry is fresh, nothing to invalidate).- p_get_instructions() is now the single place in the pipeline where $conf['syntax'] is read; from there the flavour travels as a parameter. No code under inc/Parsing/ reads $conf['syntax'] directly anymore — the five syntax-reading modes (Preformatted, GfmHr, GfmEscape, Externallink, GfmQuote) route through $this->registry.Keep the two concepts apart, as documented in the ModeRegistry andAbstractMode docblocks: the user's configured *preference* stays in$conf['syntax'] for UI code (toolbar, settings), while the activeparse's syntax is a parameter carried by the registry.$PARSER_MODES is demoted to a deprecated, read-only mirror, publishedduring loadPluginModes() — third-party syntax plugins (columnlist,alphalist2, phpwikify, skipentity) and the bundled info plugin read theglobal directly, often from their constructors, so the taxonomy muststay visible there. No core code reads the mirror.Fold ModeInterface into AbstractMode while here: getSort()/handle() areabstract, the connect callbacks carry defaults, and the public $Lexer"FIXME should be done by setter" becomes setLexer()/getLexer() injectedby Parser::addMode() alongside the registry. Nested-content resolutionmoves to the allowedCategories()/filterAllowedModes() hooks, resolvedonce when the registry is attached.Tests build their own parser/registry through ParserTestBase::setSyntax()instead of mutating $conf and calling the removed ModeRegistry::reset(). List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

e7dae73bcd947f44c901faaac9dd45de67633a3b - fix: apply rector and code sniffer fixes

Tue, 12 May 2026 14:24:43 +0000

fix: apply rector and code sniffer fixes List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

06ab3bb2c83692e8fee0c14ad0cc085ac2b07162 - Merge branch 'master' into gfm

Tue, 12 May 2026 14:11:42 +0000

Merge branch 'master' into gfm* master: (176 commits) Delete inc/Search/concept.txt �� Rector and PHPCS fixes �� Update deleted files updated phpseclib dependency use new ModeRegistry constant in info plugin keep historic typo in value but not in constant SearchIndex: fix comment position �� Rector and PHPCS fixes Translation update (tr) Translation update (fr) refactor(changelog): persist external-edit detection on first read fix(infoutils): escape git log arguments for Windows compatibility fix(preview): remove deprecated X-XSS-Protection header fix: don't move the editor textarea into the toolbar fix EXIF-rotated images shown cropped in previews, closes #4482 fix: avoid picker close/reopen race on toggle button click subscriptions: include diff link in plain-text list mails too Translation update (ru) Translation update (pl) Translation update (pt-br) ...# Conflicts:# inc/Parsing/ParserMode/Quote.php List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

56c730b56ef2746acf3b1a27c69bada1239535bd - keep historic typo in value but not in constant

Wed, 06 May 2026 19:14:25 +0000

keep historic typo in value but not in constantWe need to keep the historic typo in the value ("substition"), but thereis no reason to keep it in the constant. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

f9d3b7bd008099dc4c61ce262a02a0ed8bc94254 - Externallink: add per-scheme angle-bracket autolinks for MD syntax

Tue, 05 May 2026 14:16:44 +0000

Externallink: add per-scheme angle-bracket autolinks for MD syntaxAdds CommonMark §6.5 autolinks to Externallink, gated tomd/md+dw/dw+md syntax via ModeRegistry::isMdPreferred(). Per-schemepatterns share the existing conf/scheme.conf allow-list so unknownschemes fall through to literal cdata instead of being silentlydropped by the renderer. Internal whitespace inside the bracketsdisqualifies the autolink and the whole envelope is emitted ascdata to keep the bare-URL detector off the URL.LinksTest gains 5 cases covering success, internal-whitespace andleading-whitespace disqualification, unregistered scheme fallthrough,and the dw-only no-op path. SpecCompatRenderer URL encoder is updatedto match cmark-gfm's HREF_SAFE table (square brackets and a few othercharacters move from safe to encoded). skip.php loses the obsolete#356 entry and gains #605/#606/#607/#609 explaining the unregistered-scheme cases that the per-scheme regex naturally rejects. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

d20858669cbb910f566e0b7d1ba9da293d1b794e - extend GfmNumericEntity to HTML5 named entities, rename to GfmHtmlEntity

Mon, 04 May 2026 12:22:32 +0000

extend GfmNumericEntity to HTML5 named entities, rename to GfmHtmlEntityNumeric refs are still decoded explicitly: PHP's html_entity_decodereturns the input unchanged for U+0000, surrogates, U+10FFFF, andBMP noncharacters where CommonMark requires U+FFFD or the literalcodepoint. Named refs delegate to html_entity_decode with ENT_HTML5,which carries the full HTML5 named-entity table (including multi-codepoint decodes like ≧̸ -> U+2267 + U+0338).Unknown names stay literal: the original &xxx; passes through ascdata and the renderer's &-escaping turns it into &xxx;. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

150dc5f2f6a0d3886854a0503a1bdac5f30f5bae - add GfmNumericEntity for CommonMark numeric character references

Mon, 04 May 2026 11:55:36 +0000

add GfmNumericEntity for CommonMark numeric character referencesDecodes &#nnn; (decimal, 1-7 digits) and &#xhhh; / &#Xhhh; (hex,1-6 digits) to the corresponding Unicode codepoint, emitted asplain cdata. Codepoint 0, codepoints above U+10FFFF, and thesurrogate range U+D800..U+DFFF map to U+FFFD per the spec.Distinct from the typography Entity mode, which is renderer-sideconfigurable via entities.conf. Numeric refs are not configurableso decoding happens at parse time and the renderer needs nochanges.Lexer leftmost-match consumes the run before any structuralpattern, so *foo* renders as literal *foo* and * foodoes not start a list - matching the spec rule that numeric refscannot stand in for structural markers. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

13a62f810fbd091d15ab734b467eaec0a6bf829a - rename syntax flavors 'dokuwiki' / 'markdown' to 'dw' / 'md'

Mon, 04 May 2026 10:18:11 +0000

rename syntax flavors 'dokuwiki' / 'markdown' to 'dw' / 'md'Symmetry with the existing 'dw+md' / 'md+dw' setting values. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

c4bcbc2e0c397783de26c1e3c211d82d1ac21bb4 - add GfmLinebreak for GFM hard line breaks

Mon, 04 May 2026 09:21:04 +0000

add GfmLinebreak for GFM hard line breaksTwo-or-more trailing spaces, or a single backslash, immediately beforea non-final newline render as a `
`. Both delimiter forms share asingle SUBSTITION mode at sort 140, loaded under any MD-active syntax(markdown, dw+md, md+dw); pure dokuwiki is unaffected.Reuses the existing `linebreak` handler call and renderer; no newinstructions or renderer changes. SpecCompatRenderer overrideslinebreak() to emit the spec's `
` shape. Examples 662, 663(line break inside a raw HTML tag) are skipped — raw HTML is notpassed through by default. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

3e6baeff313fa406e4d4b5dd2e5ab85ec7d7816d - replace DW Hr with unified GfmHr

Thu, 30 Apr 2026 13:14:27 +0000

replace DW Hr with unified GfmHrSingle mode covers both DokuWiki (4+ dashes) and GFM (3+ of -/*/_)horizontal rules; pattern self-narrows on $conf['syntax']. Alwaysloaded across all four syntax settings, mirroring the GfmQuotereplacement pattern. Same `hr` handler call so renderers and thecall API are unchanged.Drops DW's old [ \t]* leading-whitespace tolerance — inert inpractice past 0-1 spaces (Preformatted at sort 20 interceptseverything ≥ 2 spaces or any tab).Spec examples 13, 20, 26-28, 224 turn green; 17, 21-24, 29, 30, 31go to skip.php as deliberate non-implementations (whitespacetolerance and list-precedence cases). List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

309a08521b24a6fff00f318e061096f69771bbad - replace DW Quote with unified GfmQuote

Thu, 30 Apr 2026 11:58:30 +0000

replace DW Quote with unified GfmQuoteGfmQuote covers blockquote parsing for both DokuWiki and GFM dialectsin a single mode. Same quote_open/quote_close handler instructions; aDW-preferred post-pass flattens sub-parsed paragraph wrapping intolinebreak calls so existing pages keep their
-between-linesrendering. MD-preferred keeps the

-wrapped spec shape.Block content (lists, fenced code, tables) inside `>` quotes nowrenders, since the body is sub-parsed. Headers stay excluded(BASEONLY) — TOC and section-edit anchors don't compose with

, same rationale as GfmListblock.Convert ModeRegistry's sub-parser cache into an acquire/release poolto support same-key re-entrancy: a list inside a quote re-entersgfm_quote during the list-item sub-parse, and the inner call needsits own parser instance even though the exclusion key matches.GfmListblock is updated to use the new acquire/release primitives. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

74031e463764923581b9204cebc0fc3f34ce881f - add GfmEscape for GFM backslash escapes

Tue, 28 Apr 2026 17:08:30 +0000

add GfmEscape for GFM backslash escapesImplements GFM §6.1 backslash-escape handling. GfmEscape is a sort-5inline mode in CATEGORY_SUBSTITION that claims `\X` for any escapableASCII punctuation char before competing delimiters can match. Theshared character class lives on Helpers\Escape so the lexer patternand the post-hoc unescape stay in lockstep.Whole-span captures (GfmCode info string, GfmLink label/URL) bypassthe lexer; those modes call Escape::unescapeBackslashes() on therelevant slot. GfmLink skips the unescape when the URL classifies asa windowssharelink so the leading \\host survives intact.GfmTable cells get a separate per-cell `\|` to `|` pass in therewriter to honour the tables-extension rule that pipes alwaysunescape, even inside code spans where standard §6.1 escapes don'tfire. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

3dabe4e0a0d70b79a7aced8ac8a36d4b37a61024 - add GfmTable for GFM tables

Tue, 28 Apr 2026 12:34:24 +0000

add GfmTable for GFM tablesImplements the GFM pipe-table extension as a CONTAINER mode at sort 55,one below DW Table at 60. A lookahead-validated entry pattern asserts aheader line plus a `:?-+:?` delimiter row before consuming any input, sonon-table paragraphs containing pipes flow through unchanged. Cells areinline-only per spec.Handler\GfmTable rewrites the flat token stream into the canonicaltable_open / tablethead_* / tabletbody_* / table_close sequence, derivingper-column alignment from the delimiter row, padding short body rows(spec 202), truncating long ones (spec 204), and falling back to a singlecdata when the column count mismatches (spec 203).`tabletbody_open` / `tabletbody_close` are emitted for the first time;they are part of the base renderer API but DW Table never used them.Added to Block's blockOpen / blockClose lists alongside `tabletfoot_*`for symmetry. SpecCompatRenderer gains minimal table-element overridesso spec roundtrip output matches GFM's `

` shapewithout DW's wrapper div, row/col counter classes, or align-as-class. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

685560eb3044321b3bdd0be40985871ced5f1d05 - add GfmListblock for GFM lists

Tue, 28 Apr 2026 06:23:11 +0000

add GfmListblock for GFM listsGfmListblock captures an entire list block atomically with oneaddSpecialPattern match, then walks the captured text in handle()grouping lines into items. Each item's body is dedented to itscontent column and parsed by ModeRegistry::getSubParser() soblock content (paragraphs, fenced code, blockquotes, pluginblocks) works inside items uniformly. Sub-parsed calls are wrappedin a Nest call before they reach the outer handler, matching theFootnote pattern: the main handler's Block rewriter treats nestas opaque and the renderer base class unwraps it transparently,so multi-paragraph items don't get double-wrapped in

.Marker syntax: -, *, + (unordered) or 1-9 digits followed by. or ) (ordered). Indentation is a 2-space-multiple step startingat 0; depth = (indent / 2) + 1, odd indents round down, tabs becometwo spaces. The first ordered item's number drives the startattribute on

via the listo_open $start parameter.GfmLists subclasses AbstractListsRewriter with the GFM markerparser; the state machine on the base class is shared with DW Lists.GfmListblock loads only when $conf['syntax'] is markdown or md+dw.Under those settings the DW Listblock is suppressed because the twolist models conflict — DW's mandatory 2-space indent rule vs GFM'szero-indent top-level rule, and -/*/+ markers shared. Plugins thatrelied on Listblock loading under md+dw will see it absent there.Sub-parser exclusion set: CATEGORY_BASEONLY (no Header inside listitems) and gfm_listblock itself (defensive guard against re-entryon pathological inputs; nested lists are handled by the outerpattern, not by re-entry).Tests cover marker variants, ordered start numbers, nested lists attwo and three levels, inline formatting inside items, marker-character switches keeping one list, type switches splitting thelist, fenced code inside items, multi-paragraph (loose) items, andtwo regressions on blank-line tolerance inside the captured block.SpecCompatRenderer learns to render the list call sequence, andspec.txt tests for digit/marker-width/lazy-continuation behaviorthat GfmListblock deliberately doesn't implement are documented ingfm-spec/skip.php with the per-bucket reasons (A-F).Drops two now-obsolete entries from skip.php (image escapes thatland via earlier GfmLink/GfmMedia work) and inlines the Setextexplanation that previously pointed at SPEC.md. Replaces theSPEC.md reference in GfmEmphasisTest with the inline reason. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

9172eccff11a70df61d7aad697f84bbd815147ce - add sub-parser support to Handler / Parser / ModeRegistry

Tue, 28 Apr 2026 06:22:27 +0000

add sub-parser support to Handler / Parser / ModeRegistryA block mode that wants to parse the body of one of its capturedmatches needs a second Parser instance configured with the activemodes minus whatever would re-enter the outer mode. Doing this byhand is verbose and easy to get wrong — modes hold a $Lexer slotthat addMode() overwrites, so the same mode object can't be sharedbetween the main parser and a sub-parser.Three small additions: Handler::reset() — clears calls, status, currentModeName, and installs a fresh CallWriter. Lets one Handler instance be parsed against repeatedly without state bleed. Parser::getHandler() — accessor; sub-parser callers need it to reach the handler for reset() and for harvesting the produced call list. ModeRegistry::getSubParser($excludeCategories, $excludeModes) — returns a cached Parser preconfigured with every active mode except those excluded. Mode objects are cloned before being attached so connectTo()'s assignment to $Lexer does not clobber the main parser's references. Cache key is the exclusion-set; default exclusion is CATEGORY_BASEONLY (no Header inside the sub-parsed content).Tests cover Handler::reset's full clear, sub-parser caching,default and custom exclusions, registry-reset propagation, andthe clone-not-share invariant for $Lexer. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

96d096f148d7def1456456590bd2910ff9115c0f - remove getLineStartMarkers registry — sort order already wins

Mon, 27 Apr 2026 14:36:55 +0000

remove getLineStartMarkers registry — sort order already winsPreformatted's entry pattern carried a `(?![\*\-])` negativelookahead to defer to list modes on indented bullet lines.0cecf9d50 (2005, "new parser added") introduced it hardcoded;7958e6980 (2026, "decouple hardcoded mode names in Eol andPreformatted") refactored that hardcoded knowledge intoregister/getLineStartMarkers on ModeRegistry so each list modeowned its marker chars. Both preserved the behavior verbatim;neither documented why it was needed.Tracing the lexer, it isn't. ParallelRegex merges all entrypatterns into one PCRE expression; PCRE returns the leftmostmatch and breaks ties on expression order. Modes are added insort order via ModeRegistry::getModes(), so Listblock (sort 10)always precedes Preformatted (sort 20) and wins the tie on" - foo" without any lookahead. The only test that caught adifference was testPreformattedList, which happened to registermodes in non-canonical order - that was a test bug.This patch drops the lookahead in Preformatted::connectTo, theregisterLineStartMarkers call in Listblock::preConnect, theregister/getLineStartMarkers methods on ModeRegistry, and thethree registry-API unit tests. testPreformattedList nowregisters Listblock before Preformatted. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

781f5c717ce516c65e49ebc3c46ab298520231d3 - gate monospace, unformatted, file on DokuWiki syntax

Thu, 23 Apr 2026 14:19:44 +0000

gate monospace, unformatted, file on DokuWiki syntaxThese DokuWiki specific modes should only be loaded when DokuWiki syntaxis still wanted, not in Markdown-only mode.Expands the ModeRegistryTest data provider to cover the full always-loadedand DW-always sets. List of files: /dokuwiki/inc/Parsing/ModeRegistry.php

b1c59bed2e3645a1f5f11438cdbe7d1596f4a3a4 - add GfmCode / GfmFile for fenced code blocks

Thu, 23 Apr 2026 14:01:03 +0000

add GfmCode / GfmFile for fenced code blocksGfmCode (backticks) emits the `code` handler instruction; GfmFile(tildes) emits `file`. Column-0 fences only, no length pairingbetween opener and closer, and unclosed fences stay literal —matching DokuWiki's `

` tag convention. The info string acceptsDW's full attribute vocabulary (language, filename, [options])through a new shared `Helpers::parseCodeAttributes` that `Code`also uses, with `html` aliased to `html4strict` and `-` meaning "nolanguage".Preformatted's indent threshold is now preference-gated: 2 spacesin DW-preferred settings, 4 spaces in MD-preferred, matching GFM'sindented code block rule. A single tab is a trigger in both.

            List of files:
            /dokuwiki/inc/Parsing/ModeRegistry.php