@podlite/schema 0.0.38 → 0.0.40
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.podlite +14 -0
- package/esm/index.d.ts +2 -0
- package/esm/index.js +1 -0
- package/esm/index.js.map +1 -1
- package/esm/plugin-tables.js +430 -46
- package/esm/plugin-tables.js.map +1 -1
- package/esm/selectors.d.ts +18 -0
- package/esm/selectors.js +158 -0
- package/esm/selectors.js.map +1 -0
- package/lib/index.d.ts +2 -0
- package/lib/index.js +4 -1
- package/lib/plugin-tables.js +430 -46
- package/lib/selectors.d.ts +18 -0
- package/lib/selectors.js +163 -0
- package/package.json +1 -1
package/CHANGELOG.podlite
CHANGED
|
@@ -2,6 +2,20 @@
|
|
|
2
2
|
|
|
3
3
|
=head1 Upcoming
|
|
4
4
|
|
|
5
|
+
=head1 0.0.40
|
|
6
|
+
|
|
7
|
+
=item recognise the C<header> MIME parameter (RFC 4180 §3) on C<:mime-type> for CSV/TSV C<=data> blocks. When the C<=data> block declares C<:mime-type<'text/csv; header=present'>> (or the same with C<text/tab-separated-values>), C<=table data:X> marks the first row as C<:header>, which renders as C<<th>>/C<<thead>> in HTML and JSX. Default C<absent> preserves the previous behaviour. The MIME-type parser also accepts other parameters (e.g. C<charset>) without disturbing type matching. Spec: podlite-specs#25
|
|
8
|
+
|
|
9
|
+
=head1 0.0.39
|
|
10
|
+
|
|
11
|
+
=item resolve C<=table data:<key>> as CSV from a C<=data :mime-type<text/csv>> block (spec §1672) — table content is replaced with C<=row>/C<=cell> structure parsed per RFC 4180 (comma delimiter, C<"..."> quoted fields with C<""> escape). Spec §1672 does not define how to mark a header in CSV-sourced tables; authors who need a header row should use a structured table with C<=begin row :header> or a Markdown GFM table. C<file:<path>> scheme is left unresolved for host-driven readers
|
|
12
|
+
=item resolve C<=table data:<key>> as TSV when the C<=data> block carries C<:mime-type<text/tab-separated-values>> — split on tabs only, no quoting (C<"> is a literal character, unlike CSV)
|
|
13
|
+
=item add table error recovery — pad short rows with empty cells, truncate over-long rows; expected cell count is taken from a C<:header> row when present, otherwise from the row with the maximum cell count. Warn on each adjustment. Tables containing C<:colspan>/C<:rowspan> cells are skipped (naive count would mis-handle spanning)
|
|
14
|
+
=item warn on mixed column separator types within a single text-mode C<=table> (e.g. C<|> in one row, whitespace in another) — does not block parsing
|
|
15
|
+
=item improve CSV/TSV error recovery — missing C<=data> block leaves the table empty; a non-tabular mime-type is rendered as a C<=code> block so the underlying content remains visible
|
|
16
|
+
=item add per-line separator detection for text-mode C<=table> when lines disagree about separator type. A common pattern — header with C<|> for visual emphasis and data rows aligned by whitespace alone — used to collapse columns under the unified positional mask; each line is now parsed on its own and produces the expected cells. Tables with a single separator type, or with both visible kinds (C<|>+C<+>) handled by the shared mask, still go through the legacy positional path so continuation lines in multi-line rows keep their column alignment
|
|
17
|
+
=item move C<parseSelector> and C<runSelector> here from C<@podlite/publisher>. These are pure AST queries (no I/O, no filesystem) and belong in C<@podlite/schema>. C<runSelector> is now generic over a C<SelectorDoc> shape (C<{ file, node }>), so consumers without the publisher's richer C<publishRecord> type can use it directly. The publisher re-exports from C<@podlite/schema> for backward compatibility
|
|
18
|
+
|
|
5
19
|
=head1 0.0.38
|
|
6
20
|
|
|
7
21
|
=item added C<type: "map"> on config items parsed from C<:attr{k=>v, ...}> key-value syntax — previously the hash form returned the map value without a type annotation, making it indistinguishable from other value kinds
|
package/esm/index.d.ts
CHANGED
|
@@ -78,5 +78,7 @@ export { parse as parse };
|
|
|
78
78
|
export { default as toHtml } from './exportHtml';
|
|
79
79
|
export { default as toMarkdown } from './exportMarkdown';
|
|
80
80
|
export { default as Writer } from './writer';
|
|
81
|
+
export { parseSelector, runSelector } from './selectors';
|
|
82
|
+
export type { SelectorDoc, ParsedSelector } from './selectors';
|
|
81
83
|
declare const VERSION: any;
|
|
82
84
|
export { VERSION as version };
|
package/esm/index.js
CHANGED
|
@@ -140,6 +140,7 @@ export { parse as parse };
|
|
|
140
140
|
export { default as toHtml } from './exportHtml';
|
|
141
141
|
export { default as toMarkdown } from './exportMarkdown';
|
|
142
142
|
export { default as Writer } from './writer';
|
|
143
|
+
export { parseSelector, runSelector } from './selectors';
|
|
143
144
|
// Cannot be `import` as it's not under TS root dir
|
|
144
145
|
// https://stackoverflow.com/questions/51070138/how-to-import-package-json-into-typescript-file-without-including-it-in-the-comp
|
|
145
146
|
const { version: VERSION } = require('../package.json');
|
package/esm/index.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,GAAoC,MAAM,KAAK,CAAA;AACtD,OAAO,KAAK,OAAO,MAAM,cAAc,CAAA;AAGvC,OAAO,KAAK,UAAU,MAAM,WAAW,CAAA;AACvC,OAAO,EAAE,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAE9C,cAAc,SAAS,CAAA;AACvB,cAAc,kBAAkB,CAAA;AAChC,cAAc,iBAAiB,CAAA;AAC/B,cAAc,eAAe,CAAA;AAC7B,cAAc,oBAAoB,CAAA;AAElC,OAAO,EAAE,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAC9C,OAAO,EAAE,eAAe,EAAE,YAAY,EAAE,eAAe,EAAE,MAAM,2BAA2B,CAAA;AAC1F,OAAO,EAAE,KAAK,EAAE,MAAM,aAAa,CAAA;AACnC,OAAO,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAA;AAC5C,OAAO,EAAE,mBAAmB,EAAE,MAAM,yBAAyB,CAAA;AAC7D,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAA;AAE9C,OAAO,EAAE,gBAAgB,EAA0B,QAAQ,EAAE,SAAS,EAAE,MAAM,mBAAmB,CAAA;AAEjG,OAAO,KAAK,MAAM,MAAM,WAAW,CAAA;AACnC,OAAO,YAAY,MAAM,kBAAkB,CAAA;AAC3C,OAAO,oBAAoB,MAAM,2BAA2B,CAAA;AAC5D,OAAO,mBAAmB,MAAM,gBAAgB,CAAA;AAChD,OAAO,YAAY,MAAM,kBAAkB,CAAA;AAC3C,OAAO,cAAc,MAAM,qBAAqB,CAAA;AAChD,OAAO,eAAe,MAAM,sBAAsB,CAAA;AAClD,OAAO,cAAc,MAAM,yBAAyB,CAAA;AACpD,OAAO,UAAU,MAAM,iBAAiB,CAAA;AAExC,MAAM,UAAU,KAAK;IACnB,OAAO,EAAE,CAAA;AACX,CAAC;AAGD,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,eAAe,EAAE,IAAI,EAAE,CAAC,CAAA;AAE5D,MAAM,UAAU,sBAAsB,CAAC,IAAa;IAClD,IAAI,IAAI,GAAG,EAAE,CAAA;IACb,MAAM,KAAK,GAAG;QACZ,OAAO,EAAE,IAAI,CAAC,EAAE;YACd,IAAI,IAAI,IAAI,CAAC,KAAK,CAAA;QACpB,CAAC;QACD,WAAW,EAAE,IAAI,CAAC,EAAE;YAClB,IAAI,IAAI,IAAI,CAAC,KAAK,CAAA;QACpB,CAAC;KACF,CAAA;IACD,MAAM,WAAW,GAAG,aAAa,CAAC,KAAK,CAAC,CAAA;IACxC,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA;IACjC,OAAO,IAAI,CAAA;AACb,CAAC;AAED,MAAM,UAAU,qBAAqB,CAAC,IAAa;IACjD,IAAI,IAAI,GAAG,EAAE,CAAA;IACb,MAAM,KAAK,GAAG;QACZ,GAAG,EAAE,CAAC,IAAI,EAAE,GAAG,EAAE,OAAO,EAAE,EAAE;YAC1B,IAAI,IAAI,CAAC,IAAI,KAAK,WAAW,EAAE;gBAC7B,IAAI,IAAI,IAAI,CAAA;gBACZ,OAAM;aACP;YACD,IAAI,IAAI,CAAC,IAAI,EAAE;gBACb,IAAI,IAAI,IAAI,CAAC,IAAI,CAAA;gBACjB,OAAM;aACP;YAED,IAAI,IAAI,IAAI,IAAI,CAAC,IAAI,GAAG,CAAA;YACxB,IAAI,SAAS,IAAI,IAAI,EAAE;gBACrB,OAAO,CAAC,IAAI,CAAC,OAAO,EAAE,GAAG,CAAC,CAAA;aAC3B;YACD,OAAO,IAAI,CAAA;QACb,CAAC;KACF,CAAA;IACD,MAAM,WAAW,GAAG,aAAa,CAAC,KAAK,CAAC,CAAA;IACxC,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA;IACjC,OAAO,IAAI,CAAA;AACb,CAAC;AACD,MAAM,UAAU,kBAAkB,CAAC,IAAa;IAC9C,OAAO,WAAW,CAAC,IAAI,EAAE,iBAAiB,CAAC,CAAA;AAC7C,CAAC;AACD,MAAM,UAAU,eAAe,CAAC,IAAa;IAC3C,OAAO,WAAW,CAAC,IAAI,EAAE,SAAS,CAAC,CAAA;AACrC,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,IAAa,EAAE,OAAe,SAAS;IACjE,MAAM,aAAa,GAA4B,UAAU,CAAC,IAAI,CAAC,CAAA;IAC/D,IAAI,CAAC,aAAa,EAAE;QAClB,OAAO,CAAC,IAAI,CAAC,8BAA8B,IAAI,UAAU,CAAC,CAAA;KAC3D;IAED,MAAM,QAAQ,GAAG,GAAG,CAAC,OAAO,CAAU,aAAa,CAAC,CAAA;IACpD,IAAI,QAAQ,CAAC,IAAI,CAAC,EAAE;QAClB,OAAO,EAAE,CAAA;KACV;IACD,OAAO,QAAQ,CAAC,MAAM,IAAI,EAAE,CAAA;AAC9B,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,MAA+B,EAAE,GAAQ;IACvE,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE;QACrB,4BAA4B;QAC5B,IAAI,cAAc,GAAG,EAAE,CAAA;QACvB,MAAM,OAAO,GAAG,CAAC,GAAG,EAAE,KAAK,EAAE,EAAE;YAC7B,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAA;YACpD,OAAO,GAAG,CAAA;QACZ,CAAC,CAAA;QACD,MAAM,OAAO,GAAG,MAAM,CAAC,MAAM,CAAC,OAAO,EAAE,cAAc,CAAC,CAAA;QACtD,MAAM,WAAW,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAA;QAC3F,OAAO,OAAO,CAAC,GAAG,CAAC,GAAG,EAAE,WAAW,CAAC,CAAA;KACrC;IACD,OAAO,SAAS,CAAA;AAClB,CAAC;AAmDD,SAAS,QAAQ;IACf,IAAI,OAAO,GAAwB,EAAE,CAAA;IACrC,KAAK,CAAC,GAAG,GAAG,GAAG,CAAA;IACf,KAAK,CAAC,KAAK,GAAG,KAAK,CAAA;IACnB,KAAK,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;IACvB,KAAK,CAAC,GAAG,CAAC,mBAAmB,CAAC,CAAA;IAC9B,KAAK,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;IACvB,KAAK,CAAC,GAAG,CAAC,cAAc,CAAC,CAAA;IACzB,KAAK,CAAC,GAAG,CAAC,UAAU,CAAC,CAAA;IACrB,KAAK,CAAC,GAAG,CAAC,oBAAoB,CAAC,CAAA;IAE/B,sCAAsC;IACtC,KAAK,CAAC,GAAG,CAAC,eAAe,CAAC,CAAA;IAC1B,KAAK,CAAC,GAAG,CAAC,cAAc,CAAC,CAAA;IACzB,OAAO,KAAK,CAAA;IAEZ,SAAS,KAAK;QACZ,OAAM;IACR,CAAC;IAED,SAAS,GAAG,CAAC,MAAoB;QAC/B,IAAI,CAAC,CAAC,UAAU,CAAC,CAAC,QAAQ,CAAC,OAAO,MAAM,CAAC,EAAE;YACzC,MAAM,MAAM,CAAA;SACb;QAED,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;QACpB,OAAO,KAAK,CAAA;IACd,CAAC;IAED,SAAS,KAAK,CAAC,GAAW,EAAE,MAAgB,EAAE,SAAS,EAAE,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE;QACtE,MAAM,EAAE,SAAS,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,GAAG,GAAG,CAAA;QAC1C,IAAI,IAAI,GAAQ,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,EAAE,OAAO,EAAE,CAAC,CAAA;QAC9C,IAAI,CAAC,SAAS,EAAE;YACd,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;gBACvC,MAAM,MAAM,GAAG,OAAO,CAAC,CAAC,CAAC,CAAA;gBACzB,OAAO;gBACP,MAAM,OAAO,GAAG,MAAM,CAAC,GAAG,CAAC,CAAA;gBAC3B,eAAe;gBACf,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,CAAA;aACrB;SACF;QACD,OAAO,IAAI,CAAA;IACb,CAAC;AACH,CAAC;AAED,OAAO,EAAE,QAAQ,IAAI,MAAM,EAAE,CAAA;AAC7B,MAAM,KAAK,GAAa,QAAQ,EAAE,CAAC,KAAK,CAAA;AACxC,OAAO,EAAE,KAAK,IAAI,KAAK,EAAE,CAAA;AACzB,OAAO,EAAE,OAAO,IAAI,MAAM,EAAE,MAAM,cAAc,CAAA;AAChD,OAAO,EAAE,OAAO,IAAI,UAAU,EAAE,MAAM,kBAAkB,CAAA;AACxD,OAAO,EAAE,OAAO,IAAI,MAAM,EAAE,MAAM,UAAU,CAAA;AAC5C,mDAAmD;AACnD,gIAAgI;AAEhI,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,GAAG,OAAO,CAAC,iBAAiB,CAAC,CAAA;AACvD,gDAAgD;AAChD,OAAO,EAAE,OAAO,IAAI,OAAO,EAAE,CAAA"}
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,GAAoC,MAAM,KAAK,CAAA;AACtD,OAAO,KAAK,OAAO,MAAM,cAAc,CAAA;AAGvC,OAAO,KAAK,UAAU,MAAM,WAAW,CAAA;AACvC,OAAO,EAAE,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAE9C,cAAc,SAAS,CAAA;AACvB,cAAc,kBAAkB,CAAA;AAChC,cAAc,iBAAiB,CAAA;AAC/B,cAAc,eAAe,CAAA;AAC7B,cAAc,oBAAoB,CAAA;AAElC,OAAO,EAAE,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAC9C,OAAO,EAAE,eAAe,EAAE,YAAY,EAAE,eAAe,EAAE,MAAM,2BAA2B,CAAA;AAC1F,OAAO,EAAE,KAAK,EAAE,MAAM,aAAa,CAAA;AACnC,OAAO,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAA;AAC5C,OAAO,EAAE,mBAAmB,EAAE,MAAM,yBAAyB,CAAA;AAC7D,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAA;AAE9C,OAAO,EAAE,gBAAgB,EAA0B,QAAQ,EAAE,SAAS,EAAE,MAAM,mBAAmB,CAAA;AAEjG,OAAO,KAAK,MAAM,MAAM,WAAW,CAAA;AACnC,OAAO,YAAY,MAAM,kBAAkB,CAAA;AAC3C,OAAO,oBAAoB,MAAM,2BAA2B,CAAA;AAC5D,OAAO,mBAAmB,MAAM,gBAAgB,CAAA;AAChD,OAAO,YAAY,MAAM,kBAAkB,CAAA;AAC3C,OAAO,cAAc,MAAM,qBAAqB,CAAA;AAChD,OAAO,eAAe,MAAM,sBAAsB,CAAA;AAClD,OAAO,cAAc,MAAM,yBAAyB,CAAA;AACpD,OAAO,UAAU,MAAM,iBAAiB,CAAA;AAExC,MAAM,UAAU,KAAK;IACnB,OAAO,EAAE,CAAA;AACX,CAAC;AAGD,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,eAAe,EAAE,IAAI,EAAE,CAAC,CAAA;AAE5D,MAAM,UAAU,sBAAsB,CAAC,IAAa;IAClD,IAAI,IAAI,GAAG,EAAE,CAAA;IACb,MAAM,KAAK,GAAG;QACZ,OAAO,EAAE,IAAI,CAAC,EAAE;YACd,IAAI,IAAI,IAAI,CAAC,KAAK,CAAA;QACpB,CAAC;QACD,WAAW,EAAE,IAAI,CAAC,EAAE;YAClB,IAAI,IAAI,IAAI,CAAC,KAAK,CAAA;QACpB,CAAC;KACF,CAAA;IACD,MAAM,WAAW,GAAG,aAAa,CAAC,KAAK,CAAC,CAAA;IACxC,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA;IACjC,OAAO,IAAI,CAAA;AACb,CAAC;AAED,MAAM,UAAU,qBAAqB,CAAC,IAAa;IACjD,IAAI,IAAI,GAAG,EAAE,CAAA;IACb,MAAM,KAAK,GAAG;QACZ,GAAG,EAAE,CAAC,IAAI,EAAE,GAAG,EAAE,OAAO,EAAE,EAAE;YAC1B,IAAI,IAAI,CAAC,IAAI,KAAK,WAAW,EAAE;gBAC7B,IAAI,IAAI,IAAI,CAAA;gBACZ,OAAM;aACP;YACD,IAAI,IAAI,CAAC,IAAI,EAAE;gBACb,IAAI,IAAI,IAAI,CAAC,IAAI,CAAA;gBACjB,OAAM;aACP;YAED,IAAI,IAAI,IAAI,IAAI,CAAC,IAAI,GAAG,CAAA;YACxB,IAAI,SAAS,IAAI,IAAI,EAAE;gBACrB,OAAO,CAAC,IAAI,CAAC,OAAO,EAAE,GAAG,CAAC,CAAA;aAC3B;YACD,OAAO,IAAI,CAAA;QACb,CAAC;KACF,CAAA;IACD,MAAM,WAAW,GAAG,aAAa,CAAC,KAAK,CAAC,CAAA;IACxC,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA;IACjC,OAAO,IAAI,CAAA;AACb,CAAC;AACD,MAAM,UAAU,kBAAkB,CAAC,IAAa;IAC9C,OAAO,WAAW,CAAC,IAAI,EAAE,iBAAiB,CAAC,CAAA;AAC7C,CAAC;AACD,MAAM,UAAU,eAAe,CAAC,IAAa;IAC3C,OAAO,WAAW,CAAC,IAAI,EAAE,SAAS,CAAC,CAAA;AACrC,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,IAAa,EAAE,OAAe,SAAS;IACjE,MAAM,aAAa,GAA4B,UAAU,CAAC,IAAI,CAAC,CAAA;IAC/D,IAAI,CAAC,aAAa,EAAE;QAClB,OAAO,CAAC,IAAI,CAAC,8BAA8B,IAAI,UAAU,CAAC,CAAA;KAC3D;IAED,MAAM,QAAQ,GAAG,GAAG,CAAC,OAAO,CAAU,aAAa,CAAC,CAAA;IACpD,IAAI,QAAQ,CAAC,IAAI,CAAC,EAAE;QAClB,OAAO,EAAE,CAAA;KACV;IACD,OAAO,QAAQ,CAAC,MAAM,IAAI,EAAE,CAAA;AAC9B,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,MAA+B,EAAE,GAAQ;IACvE,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE;QACrB,4BAA4B;QAC5B,IAAI,cAAc,GAAG,EAAE,CAAA;QACvB,MAAM,OAAO,GAAG,CAAC,GAAG,EAAE,KAAK,EAAE,EAAE;YAC7B,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAA;YACpD,OAAO,GAAG,CAAA;QACZ,CAAC,CAAA;QACD,MAAM,OAAO,GAAG,MAAM,CAAC,MAAM,CAAC,OAAO,EAAE,cAAc,CAAC,CAAA;QACtD,MAAM,WAAW,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAA;QAC3F,OAAO,OAAO,CAAC,GAAG,CAAC,GAAG,EAAE,WAAW,CAAC,CAAA;KACrC;IACD,OAAO,SAAS,CAAA;AAClB,CAAC;AAmDD,SAAS,QAAQ;IACf,IAAI,OAAO,GAAwB,EAAE,CAAA;IACrC,KAAK,CAAC,GAAG,GAAG,GAAG,CAAA;IACf,KAAK,CAAC,KAAK,GAAG,KAAK,CAAA;IACnB,KAAK,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;IACvB,KAAK,CAAC,GAAG,CAAC,mBAAmB,CAAC,CAAA;IAC9B,KAAK,CAAC,GAAG,CAAC,YAAY,CAAC,CAAA;IACvB,KAAK,CAAC,GAAG,CAAC,cAAc,CAAC,CAAA;IACzB,KAAK,CAAC,GAAG,CAAC,UAAU,CAAC,CAAA;IACrB,KAAK,CAAC,GAAG,CAAC,oBAAoB,CAAC,CAAA;IAE/B,sCAAsC;IACtC,KAAK,CAAC,GAAG,CAAC,eAAe,CAAC,CAAA;IAC1B,KAAK,CAAC,GAAG,CAAC,cAAc,CAAC,CAAA;IACzB,OAAO,KAAK,CAAA;IAEZ,SAAS,KAAK;QACZ,OAAM;IACR,CAAC;IAED,SAAS,GAAG,CAAC,MAAoB;QAC/B,IAAI,CAAC,CAAC,UAAU,CAAC,CAAC,QAAQ,CAAC,OAAO,MAAM,CAAC,EAAE;YACzC,MAAM,MAAM,CAAA;SACb;QAED,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;QACpB,OAAO,KAAK,CAAA;IACd,CAAC;IAED,SAAS,KAAK,CAAC,GAAW,EAAE,MAAgB,EAAE,SAAS,EAAE,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE;QACtE,MAAM,EAAE,SAAS,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,GAAG,GAAG,CAAA;QAC1C,IAAI,IAAI,GAAQ,MAAM,CAAC,KAAK,CAAC,GAAG,EAAE,EAAE,OAAO,EAAE,CAAC,CAAA;QAC9C,IAAI,CAAC,SAAS,EAAE;YACd,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;gBACvC,MAAM,MAAM,GAAG,OAAO,CAAC,CAAC,CAAC,CAAA;gBACzB,OAAO;gBACP,MAAM,OAAO,GAAG,MAAM,CAAC,GAAG,CAAC,CAAA;gBAC3B,eAAe;gBACf,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,CAAA;aACrB;SACF;QACD,OAAO,IAAI,CAAA;IACb,CAAC;AACH,CAAC;AAED,OAAO,EAAE,QAAQ,IAAI,MAAM,EAAE,CAAA;AAC7B,MAAM,KAAK,GAAa,QAAQ,EAAE,CAAC,KAAK,CAAA;AACxC,OAAO,EAAE,KAAK,IAAI,KAAK,EAAE,CAAA;AACzB,OAAO,EAAE,OAAO,IAAI,MAAM,EAAE,MAAM,cAAc,CAAA;AAChD,OAAO,EAAE,OAAO,IAAI,UAAU,EAAE,MAAM,kBAAkB,CAAA;AACxD,OAAO,EAAE,OAAO,IAAI,MAAM,EAAE,MAAM,UAAU,CAAA;AAC5C,OAAO,EAAE,aAAa,EAAE,WAAW,EAAE,MAAM,aAAa,CAAA;AAExD,mDAAmD;AACnD,gIAAgI;AAEhI,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,GAAG,OAAO,CAAC,iBAAiB,CAAC,CAAA;AACvD,gDAAgD;AAChD,OAAO,EAAE,OAAO,IAAI,OAAO,EAAE,CAAA"}
|
package/esm/plugin-tables.js
CHANGED
|
@@ -4,13 +4,302 @@
|
|
|
4
4
|
* and the remaining content is treated as the definition for the term.
|
|
5
5
|
*/
|
|
6
6
|
import makeTransformer from './helpers/makeTransformer';
|
|
7
|
+
import makeAttrs from './helpers/config';
|
|
7
8
|
function flattenDeep(arr) {
|
|
8
9
|
return arr.reduce((acc, val) => (Array.isArray(val) ? acc.concat(flattenDeep(val)) : acc.concat(val)), []);
|
|
9
10
|
}
|
|
11
|
+
// ─── CSV source for =table (spec §1672) ────────────────────────────────────
|
|
12
|
+
// Parse CSV content per RFC 4180: comma delimiter, `"`-quoted fields with
|
|
13
|
+
// `""` as embedded-quote escape. Supports LF or CRLF line endings. Leading
|
|
14
|
+
// indentation typical of =data block bodies is stripped from each line.
|
|
15
|
+
function parseCsv(text) {
|
|
16
|
+
const rows = [];
|
|
17
|
+
let field = '';
|
|
18
|
+
let row = [];
|
|
19
|
+
let inQuote = false;
|
|
20
|
+
let i = 0;
|
|
21
|
+
while (i < text.length) {
|
|
22
|
+
const c = text[i];
|
|
23
|
+
if (inQuote) {
|
|
24
|
+
if (c === '"' && text[i + 1] === '"') {
|
|
25
|
+
field += '"';
|
|
26
|
+
i += 2;
|
|
27
|
+
continue;
|
|
28
|
+
}
|
|
29
|
+
if (c === '"') {
|
|
30
|
+
inQuote = false;
|
|
31
|
+
i++;
|
|
32
|
+
continue;
|
|
33
|
+
}
|
|
34
|
+
field += c;
|
|
35
|
+
i++;
|
|
36
|
+
continue;
|
|
37
|
+
}
|
|
38
|
+
if (c === '"' && field === '') {
|
|
39
|
+
inQuote = true;
|
|
40
|
+
i++;
|
|
41
|
+
continue;
|
|
42
|
+
}
|
|
43
|
+
if (c === ',') {
|
|
44
|
+
row.push(field);
|
|
45
|
+
field = '';
|
|
46
|
+
i++;
|
|
47
|
+
continue;
|
|
48
|
+
}
|
|
49
|
+
if (c === '\r') {
|
|
50
|
+
i++;
|
|
51
|
+
continue;
|
|
52
|
+
}
|
|
53
|
+
if (c === '\n') {
|
|
54
|
+
row.push(field);
|
|
55
|
+
rows.push(row);
|
|
56
|
+
row = [];
|
|
57
|
+
field = '';
|
|
58
|
+
i++;
|
|
59
|
+
continue;
|
|
60
|
+
}
|
|
61
|
+
field += c;
|
|
62
|
+
i++;
|
|
63
|
+
}
|
|
64
|
+
if (field !== '' || row.length > 0) {
|
|
65
|
+
row.push(field);
|
|
66
|
+
rows.push(row);
|
|
67
|
+
}
|
|
68
|
+
// Drop rows that are entirely blank (empty content after trim).
|
|
69
|
+
return rows.filter(r => !(r.length === 1 && r[0].trim() === ''));
|
|
70
|
+
}
|
|
71
|
+
// Parse a TSV (tab-separated values) blob. Unlike CSV, TSV has no quoting
|
|
72
|
+
// mechanism — fields are split strictly on tabs and `"` is a literal
|
|
73
|
+
// character. Tabs and newlines inside fields are not representable in TSV.
|
|
74
|
+
function parseTsv(text) {
|
|
75
|
+
const lines = text.split(/\r?\n/);
|
|
76
|
+
const rows = lines.map(line => line.split('\t'));
|
|
77
|
+
return rows.filter(r => !(r.length === 1 && r[0].trim() === ''));
|
|
78
|
+
}
|
|
79
|
+
// Parse a MIME type value into the bare type and a map of parameters
|
|
80
|
+
// (RFC 6838 / RFC 7231 §3.1.1.1). Used to recognise the `header` parameter
|
|
81
|
+
// (RFC 4180 §3) on `:mime-type` of CSV/TSV =data blocks.
|
|
82
|
+
//
|
|
83
|
+
// Examples:
|
|
84
|
+
// "text/csv" → { type: 'text/csv', params: {} }
|
|
85
|
+
// "text/csv; header=present" → { type: 'text/csv', params: { header: 'present' } }
|
|
86
|
+
// "text/csv;charset=utf-8;header=absent" → { ..., params: { charset: 'utf-8', header: 'absent' } }
|
|
87
|
+
function parseMimeType(raw) {
|
|
88
|
+
if (!raw || typeof raw !== 'string')
|
|
89
|
+
return { type: '', params: {} };
|
|
90
|
+
const parts = raw.split(';').map(s => s.trim());
|
|
91
|
+
const type = (parts.shift() || '').toLowerCase();
|
|
92
|
+
const params = {};
|
|
93
|
+
for (const p of parts) {
|
|
94
|
+
if (!p)
|
|
95
|
+
continue;
|
|
96
|
+
const eq = p.indexOf('=');
|
|
97
|
+
if (eq < 0)
|
|
98
|
+
continue;
|
|
99
|
+
const key = p.slice(0, eq).trim().toLowerCase();
|
|
100
|
+
let value = p.slice(eq + 1).trim();
|
|
101
|
+
if ((value.startsWith('"') && value.endsWith('"')) || (value.startsWith("'") && value.endsWith("'"))) {
|
|
102
|
+
value = value.slice(1, -1);
|
|
103
|
+
}
|
|
104
|
+
if (key)
|
|
105
|
+
params[key] = value;
|
|
106
|
+
}
|
|
107
|
+
return { type, params };
|
|
108
|
+
}
|
|
109
|
+
// Locate the first `=data` block with a matching `:key` attribute anywhere
|
|
110
|
+
// in the document tree.
|
|
111
|
+
function findDataBlockByKey(tree, key) {
|
|
112
|
+
let found = null;
|
|
113
|
+
const walk = node => {
|
|
114
|
+
if (found)
|
|
115
|
+
return;
|
|
116
|
+
if (Array.isArray(node)) {
|
|
117
|
+
node.forEach(walk);
|
|
118
|
+
return;
|
|
119
|
+
}
|
|
120
|
+
if (!node || typeof node !== 'object')
|
|
121
|
+
return;
|
|
122
|
+
if (node.type === 'block' && node.name === 'data') {
|
|
123
|
+
const attrs = makeAttrs(node, {});
|
|
124
|
+
if (attrs.getFirstValue('key') === key) {
|
|
125
|
+
found = node;
|
|
126
|
+
return;
|
|
127
|
+
}
|
|
128
|
+
}
|
|
129
|
+
if (Array.isArray(node.content))
|
|
130
|
+
node.content.forEach(walk);
|
|
131
|
+
};
|
|
132
|
+
walk(tree);
|
|
133
|
+
return found;
|
|
134
|
+
}
|
|
135
|
+
function extractDataText(dataNode) {
|
|
136
|
+
if (!dataNode || !Array.isArray(dataNode.content))
|
|
137
|
+
return '';
|
|
138
|
+
const verbatim = dataNode.content.find(c => c && c.type === 'verbatim');
|
|
139
|
+
return verbatim && typeof verbatim.value === 'string' ? verbatim.value : '';
|
|
140
|
+
}
|
|
141
|
+
// Detect whether the =table body is a single-line source reference like
|
|
142
|
+
// `data:key` or `file:path` (spec §1672). Returns the parsed reference or
|
|
143
|
+
// null if the body is ordinary table content.
|
|
144
|
+
function detectSourceReference(tableNode) {
|
|
145
|
+
const texts = [];
|
|
146
|
+
const walker = makeTransformer({
|
|
147
|
+
'row:text': r => {
|
|
148
|
+
texts.push(r.value);
|
|
149
|
+
return r;
|
|
150
|
+
},
|
|
151
|
+
'head:text': h => {
|
|
152
|
+
texts.push(h.value);
|
|
153
|
+
return h;
|
|
154
|
+
},
|
|
155
|
+
});
|
|
156
|
+
walker(tableNode, {});
|
|
157
|
+
const joined = texts.join('\n').trim();
|
|
158
|
+
const lines = joined
|
|
159
|
+
.split(/\r?\n/)
|
|
160
|
+
.map(l => l.trim())
|
|
161
|
+
.filter(l => l.length > 0);
|
|
162
|
+
if (lines.length !== 1)
|
|
163
|
+
return null;
|
|
164
|
+
const m = lines[0].match(/^(data|file):(\S+)$/);
|
|
165
|
+
if (!m)
|
|
166
|
+
return null;
|
|
167
|
+
return { scheme: m[1], target: m[2] };
|
|
168
|
+
}
|
|
169
|
+
function buildCellBlock(text) {
|
|
170
|
+
return {
|
|
171
|
+
name: 'cell',
|
|
172
|
+
type: 'block',
|
|
173
|
+
margin: '',
|
|
174
|
+
content: [{ type: 'text', value: text }],
|
|
175
|
+
};
|
|
176
|
+
}
|
|
177
|
+
function buildRowBlock(cells, isHeader) {
|
|
178
|
+
const block = {
|
|
179
|
+
name: 'row',
|
|
180
|
+
type: 'block',
|
|
181
|
+
margin: '',
|
|
182
|
+
content: cells,
|
|
183
|
+
};
|
|
184
|
+
if (isHeader) {
|
|
185
|
+
block.config = [{ name: 'header', value: true, type: 'boolean' }];
|
|
186
|
+
}
|
|
187
|
+
return block;
|
|
188
|
+
}
|
|
189
|
+
// Convert parsed CSV/TSV rows to plain `=row`/`=cell` blocks. When
|
|
190
|
+
// `hasHeader` is true (signalled by the `header=present` MIME parameter on
|
|
191
|
+
// the source `=data` block, per RFC 4180 §3), the first row is marked with
|
|
192
|
+
// `:header`. Default behaviour leaves all rows unmarked, matching authors
|
|
193
|
+
// who use a structured table with explicit `=begin row :header` or a
|
|
194
|
+
// Markdown GFM table.
|
|
195
|
+
function csvToTableContent(csvRows, hasHeader = false) {
|
|
196
|
+
return csvRows.map((row, i) => {
|
|
197
|
+
const cells = row.map(v => buildCellBlock(v.trim()));
|
|
198
|
+
return buildRowBlock(cells, hasHeader && i === 0);
|
|
199
|
+
});
|
|
200
|
+
}
|
|
201
|
+
// ─── Error recovery (design notes Rules 2-4) ───────────────────────────────
|
|
202
|
+
// Rule 2 — table-level cell count validation. Pad short rows with empty
|
|
203
|
+
// cells; truncate long rows. Emit a warning whenever a row is changed.
|
|
204
|
+
// Expected count is taken from the `:header` row if present, otherwise from
|
|
205
|
+
// the row with the maximum cell count.
|
|
206
|
+
//
|
|
207
|
+
// Skipped when any cell uses `:colspan` or `:rowspan`: a spanning cell
|
|
208
|
+
// occupies multiple columns, so naive cell counting would misreport row
|
|
209
|
+
// width and drop legitimate spanned cells.
|
|
210
|
+
function normalizeCellCounts(tableNode, source = 'table') {
|
|
211
|
+
if (!tableNode || !Array.isArray(tableNode.content))
|
|
212
|
+
return tableNode;
|
|
213
|
+
const rows = tableNode.content.filter(c => c && c.type === 'block' && c.name === 'row');
|
|
214
|
+
if (rows.length === 0)
|
|
215
|
+
return tableNode;
|
|
216
|
+
const cellsOf = row => (Array.isArray(row.content) ? row.content.filter(c => c && c.name === 'cell') : []);
|
|
217
|
+
const isHeaderRow = row => Array.isArray(row.config) && row.config.some(a => a.name === 'header' && a.value === true);
|
|
218
|
+
const cellHasSpan = cell => Array.isArray(cell.config) && cell.config.some(a => a.name === 'colspan' || a.name === 'rowspan');
|
|
219
|
+
const rowHasSpan = row => cellsOf(row).some(cellHasSpan);
|
|
220
|
+
if (rows.some(rowHasSpan))
|
|
221
|
+
return tableNode;
|
|
222
|
+
const headerRow = rows.find(isHeaderRow);
|
|
223
|
+
const expected = headerRow ? cellsOf(headerRow).length : Math.max(...rows.map(r => cellsOf(r).length));
|
|
224
|
+
if (expected === 0)
|
|
225
|
+
return tableNode;
|
|
226
|
+
let mutated = false;
|
|
227
|
+
const newContent = tableNode.content.map(child => {
|
|
228
|
+
if (!child || child.type !== 'block' || child.name !== 'row')
|
|
229
|
+
return child;
|
|
230
|
+
const cells = cellsOf(child);
|
|
231
|
+
if (cells.length === expected)
|
|
232
|
+
return child;
|
|
233
|
+
if (cells.length < expected) {
|
|
234
|
+
const padding = [];
|
|
235
|
+
for (let i = cells.length; i < expected; i++)
|
|
236
|
+
padding.push(buildCellBlock(''));
|
|
237
|
+
console.warn(`[${source}] row has ${cells.length} cells, expected ${expected} — padded with ${padding.length} empty`);
|
|
238
|
+
mutated = true;
|
|
239
|
+
return { ...child, content: [...child.content, ...padding] };
|
|
240
|
+
}
|
|
241
|
+
// cells.length > expected → truncate
|
|
242
|
+
const dropped = cells.length - expected;
|
|
243
|
+
console.warn(`[${source}] row has ${cells.length} cells, expected ${expected} — truncated ${dropped}`);
|
|
244
|
+
mutated = true;
|
|
245
|
+
// Keep non-cell entries (e.g. blanklines) and the first `expected` cells
|
|
246
|
+
let keepCells = expected;
|
|
247
|
+
const trimmed = [];
|
|
248
|
+
for (const c of child.content || []) {
|
|
249
|
+
if (c && c.type === 'block' && c.name === 'cell') {
|
|
250
|
+
if (keepCells > 0) {
|
|
251
|
+
trimmed.push(c);
|
|
252
|
+
keepCells--;
|
|
253
|
+
}
|
|
254
|
+
}
|
|
255
|
+
else {
|
|
256
|
+
trimmed.push(c);
|
|
257
|
+
}
|
|
258
|
+
}
|
|
259
|
+
return { ...child, content: trimmed };
|
|
260
|
+
});
|
|
261
|
+
return mutated ? { ...tableNode, content: newContent } : tableNode;
|
|
262
|
+
}
|
|
263
|
+
// Rule 3 — mixed separator detection (text-mode only). Inspects each line
|
|
264
|
+
// for visible separators (`|` / `+`) surrounded by whitespace; lines without
|
|
265
|
+
// any visible separator fall back to whitespace separation. Warns when more
|
|
266
|
+
// than one separator type is observed within a single table.
|
|
267
|
+
function detectMixedSeparators(lines) {
|
|
268
|
+
const seen = new Set();
|
|
269
|
+
for (const line of lines) {
|
|
270
|
+
if (!line || typeof line !== 'string')
|
|
271
|
+
continue;
|
|
272
|
+
if (/\s\|\s/.test(line))
|
|
273
|
+
seen.add('pipe');
|
|
274
|
+
else if (/\s\+\s/.test(line))
|
|
275
|
+
seen.add('plus');
|
|
276
|
+
else if (line.trim().length > 0)
|
|
277
|
+
seen.add('whitespace');
|
|
278
|
+
if (seen.size > 1)
|
|
279
|
+
break;
|
|
280
|
+
}
|
|
281
|
+
if (seen.size > 1) {
|
|
282
|
+
console.warn(`[table] mixed separator types detected: ${Array.from(seen).join(', ')} — recommend a single style`);
|
|
283
|
+
}
|
|
284
|
+
}
|
|
285
|
+
// Rule 4 — replace =table with =code block. Used when a referenced =data
|
|
286
|
+
// source has a non-CSV mime type: the spec mandates the source be rendered
|
|
287
|
+
// as a code block so the user can still see the underlying content.
|
|
288
|
+
function buildCodeFromDataBlock(tableNode, dataBlock) {
|
|
289
|
+
return {
|
|
290
|
+
type: 'block',
|
|
291
|
+
name: 'code',
|
|
292
|
+
margin: tableNode.margin || '',
|
|
293
|
+
content: Array.isArray(dataBlock.content) ? dataBlock.content : [],
|
|
294
|
+
config: Array.isArray(tableNode.config) ? tableNode.config : [],
|
|
295
|
+
};
|
|
296
|
+
}
|
|
10
297
|
/**
|
|
11
298
|
* Helpers section
|
|
12
299
|
*/
|
|
13
|
-
//
|
|
300
|
+
// Bit-mask helper used by the positional column extractor for multi-line
|
|
301
|
+
// rows. Compares two same-length digit strings character by character via
|
|
302
|
+
// `cb`, returning the joined result.
|
|
14
303
|
const strbin = (str1, str2, cb) => {
|
|
15
304
|
let res = [];
|
|
16
305
|
for (let i = 0; str1.length > i; i++) {
|
|
@@ -18,62 +307,109 @@ const strbin = (str1, str2, cb) => {
|
|
|
18
307
|
}
|
|
19
308
|
return res.join('');
|
|
20
309
|
};
|
|
21
|
-
//
|
|
310
|
+
// Build a unified column-position mask across a set of lines. Used as a
|
|
311
|
+
// fallback when a row spans multiple lines (continuation lines that align
|
|
312
|
+
// content by character position rather than by separator). Returns a binary
|
|
313
|
+
// string where `0` runs mark column ranges and `1` runs mark gaps.
|
|
22
314
|
const makeMask = (lines, separators) => {
|
|
23
|
-
// calculate template length
|
|
24
315
|
const tmplLength = Math.max(...[...lines, ...separators].map(s => s.length));
|
|
25
|
-
// make bin mask for each string
|
|
26
316
|
const masks = lines.map(str => {
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
'0000000011111111111110001111111111111000001111111111111111111111111111' ]
|
|
30
|
-
then not(mask) ... then & masks
|
|
31
|
-
*/
|
|
32
|
-
// enlarge string to tmplLength
|
|
33
|
-
let tstr = str + ' '.repeat(tmplLength - str.length);
|
|
34
|
-
let mask = [];
|
|
317
|
+
const tstr = str + ' '.repeat(tmplLength - str.length);
|
|
318
|
+
const mask = [];
|
|
35
319
|
const re = /\s+[+|\s]\s/g;
|
|
36
320
|
let match;
|
|
37
321
|
while ((match = re.exec(tstr)) != null) {
|
|
38
322
|
const tmpMask = '1'.repeat(match.index) + '0'.repeat(match[0].length);
|
|
39
323
|
mask.push(tmpMask + '1'.repeat(tmplLength - tmpMask.length));
|
|
40
324
|
}
|
|
41
|
-
return mask.reduce((a, b) =>
|
|
42
|
-
return strbin(a, b, (i1, i2) => i1 & i2);
|
|
43
|
-
}, '1'.repeat(tmplLength));
|
|
325
|
+
return mask.reduce((a, b) => strbin(a, b, (i1, i2) => i1 & i2), '1'.repeat(tmplLength));
|
|
44
326
|
});
|
|
45
|
-
// make result mask
|
|
46
327
|
const inverted = masks.map(m => strbin(m, '', i1 => (i1 == 0 ? 1 : 0)));
|
|
47
|
-
|
|
48
|
-
return strbin(a, b, (i1, i2) => i1 & i2);
|
|
49
|
-
}, '1'.repeat(tmplLength));
|
|
50
|
-
return columnTemplate;
|
|
328
|
+
return inverted.reduce((a, b) => strbin(a, b, (i1, i2) => i1 & i2), '1'.repeat(tmplLength));
|
|
51
329
|
};
|
|
330
|
+
// Apply a column-position template to a multi-line text block, extracting
|
|
331
|
+
// per-column substrings and aggregating across lines (continuation lines
|
|
332
|
+
// append to the cell at the same column position).
|
|
52
333
|
const extractColumnsByTemplate = (text, template) => {
|
|
53
|
-
const lines = flattenDeep(text
|
|
54
|
-
.split(/\n/) // split each row by eol
|
|
55
|
-
.filter(str => str.length > 0));
|
|
334
|
+
const lines = flattenDeep(text.split(/\n/).filter(s => s.length > 0));
|
|
56
335
|
const cols = lines.map(line => {
|
|
57
336
|
const re = /((1+|0+))/g;
|
|
58
|
-
|
|
337
|
+
const columns = [];
|
|
59
338
|
let match;
|
|
60
339
|
while ((match = re.exec(template)) != null) {
|
|
61
|
-
if (match[0][0] == 1)
|
|
340
|
+
if (match[0][0] == '1')
|
|
62
341
|
continue;
|
|
63
|
-
|
|
64
|
-
columns.push(s);
|
|
342
|
+
columns.push(line.substring(match.index, match.index + match[0].length));
|
|
65
343
|
}
|
|
66
344
|
return columns;
|
|
67
345
|
});
|
|
68
|
-
|
|
69
|
-
|
|
346
|
+
const result = [];
|
|
347
|
+
cols.reduce((a, b) => {
|
|
70
348
|
for (let i = 0; i < b.length; i++) {
|
|
71
349
|
a[i] = (a[i] === undefined ? '' : a[i]) + ' ' + b[i];
|
|
72
350
|
}
|
|
73
351
|
return a;
|
|
74
|
-
},
|
|
352
|
+
}, result);
|
|
75
353
|
return result;
|
|
76
354
|
};
|
|
355
|
+
const detectLineSeparator = (line) => {
|
|
356
|
+
if (/(?:^|\s)\|(?:\s|$)/.test(line))
|
|
357
|
+
return 'pipe';
|
|
358
|
+
if (/(?:^|\s)\+(?:\s|$)/.test(line))
|
|
359
|
+
return 'plus';
|
|
360
|
+
return 'whitespace';
|
|
361
|
+
};
|
|
362
|
+
const trimEdgeEmpty = (cells) => {
|
|
363
|
+
// A leading `|` produces an empty cell at index 0; a trailing `|` does the
|
|
364
|
+
// same at the end. Drop those edge artifacts. Empty cells in the middle of
|
|
365
|
+
// the row are preserved.
|
|
366
|
+
let start = 0;
|
|
367
|
+
let end = cells.length;
|
|
368
|
+
if (cells[start] === '')
|
|
369
|
+
start++;
|
|
370
|
+
if (end > start && cells[end - 1] === '')
|
|
371
|
+
end--;
|
|
372
|
+
return cells.slice(start, end);
|
|
373
|
+
};
|
|
374
|
+
const splitLineByPipe = (line) => trimEdgeEmpty(line.split(/\s*\|\s*/).map(c => c.trim()));
|
|
375
|
+
const splitLineByPlus = (line) => trimEdgeEmpty(line.split(/\s*\+\s*/).map(c => c.trim()));
|
|
376
|
+
const splitLineByWhitespace = (line) => line
|
|
377
|
+
.trim()
|
|
378
|
+
.split(/\s{2,}/)
|
|
379
|
+
.filter(c => c !== '');
|
|
380
|
+
const splitLineCells = (line) => {
|
|
381
|
+
const trimmed = line.trim();
|
|
382
|
+
if (trimmed === '')
|
|
383
|
+
return [];
|
|
384
|
+
const kind = detectLineSeparator(line);
|
|
385
|
+
if (kind === 'pipe')
|
|
386
|
+
return splitLineByPipe(line);
|
|
387
|
+
if (kind === 'plus')
|
|
388
|
+
return splitLineByPlus(line);
|
|
389
|
+
return splitLineByWhitespace(line);
|
|
390
|
+
};
|
|
391
|
+
// Convert a row's raw text (which may span multiple lines) to cell values.
|
|
392
|
+
// Multi-line rows: take the separator kind from the first non-blank line and
|
|
393
|
+
// apply it to every line of the row, then aggregate column-wise so that a
|
|
394
|
+
// continuation line (e.g. wrapped cell content) appends to the cell from the
|
|
395
|
+
// previous line in the same column.
|
|
396
|
+
const rowToCells = (rowValue) => {
|
|
397
|
+
const lines = rowValue.split(/\r?\n/).filter(l => l.trim() !== '');
|
|
398
|
+
if (lines.length === 0)
|
|
399
|
+
return [];
|
|
400
|
+
if (lines.length === 1)
|
|
401
|
+
return splitLineCells(lines[0]);
|
|
402
|
+
const kind = detectLineSeparator(lines[0]);
|
|
403
|
+
const splitFn = kind === 'pipe' ? splitLineByPipe : kind === 'plus' ? splitLineByPlus : splitLineByWhitespace;
|
|
404
|
+
const lineCells = lines.map(splitFn);
|
|
405
|
+
const maxCols = Math.max(...lineCells.map(c => c.length));
|
|
406
|
+
const merged = [];
|
|
407
|
+
for (let i = 0; i < maxCols; i++) {
|
|
408
|
+
const parts = lineCells.map(line => line[i] ?? '').filter(p => p !== '');
|
|
409
|
+
merged.push(parts.join(' '));
|
|
410
|
+
}
|
|
411
|
+
return merged;
|
|
412
|
+
};
|
|
77
413
|
/**
|
|
78
414
|
* Main transforms
|
|
79
415
|
*/
|
|
@@ -97,14 +433,45 @@ const isStructured = tableNode => Array.isArray(tableNode.content) &&
|
|
|
97
433
|
export default () => tree => {
|
|
98
434
|
const transformer = makeTransformer({
|
|
99
435
|
table: node => {
|
|
100
|
-
//
|
|
436
|
+
// CSV/data source reference (spec §1672):
|
|
437
|
+
// =table data:<key> → resolve =data block with :key<key>
|
|
438
|
+
// =table file:<path> → defer to host reader (not implemented here)
|
|
439
|
+
const ref = detectSourceReference(node);
|
|
440
|
+
if (ref && ref.scheme === 'data') {
|
|
441
|
+
const dataBlock = findDataBlockByKey(tree, ref.target);
|
|
442
|
+
if (!dataBlock) {
|
|
443
|
+
// Rule 4: source not found → empty table (still a =table block)
|
|
444
|
+
console.warn(`[table] no =data block found for data:${ref.target} — rendered as empty`);
|
|
445
|
+
return { ...node, content: [] };
|
|
446
|
+
}
|
|
447
|
+
const rawMime = makeAttrs(dataBlock, {}).getFirstValue('mime-type');
|
|
448
|
+
const { type: mimeType, params: mimeParams } = parseMimeType(rawMime);
|
|
449
|
+
const isCsv = mimeType === 'text/csv';
|
|
450
|
+
const isTsv = mimeType === 'text/tab-separated-values';
|
|
451
|
+
if (isCsv || isTsv) {
|
|
452
|
+
const text = extractDataText(dataBlock);
|
|
453
|
+
const rows = isCsv ? parseCsv(text) : parseTsv(text);
|
|
454
|
+
if (rows.length === 0) {
|
|
455
|
+
console.warn(`[table] ${isCsv ? 'CSV' : 'TSV'} parse produced no rows for data:${ref.target} — rendered as empty`);
|
|
456
|
+
return { ...node, content: [] };
|
|
457
|
+
}
|
|
458
|
+
const hasHeader = mimeParams.header === 'present';
|
|
459
|
+
const filledNode = { ...node, content: csvToTableContent(rows, hasHeader) };
|
|
460
|
+
return normalizeCellCounts(filledNode, `table data:${ref.target}`);
|
|
461
|
+
}
|
|
462
|
+
// Rule 4: source not tabular → render as code block so content remains visible
|
|
463
|
+
console.warn(`[table] =data :key<${ref.target}> has non-tabular mime-type ${rawMime || '(none)'} — rendered as =code`);
|
|
464
|
+
return buildCodeFromDataBlock(node, dataBlock);
|
|
465
|
+
}
|
|
466
|
+
// structured mode: transform row children (wrap implicit cells), then
|
|
467
|
+
// apply Rule 2 cell count normalization.
|
|
101
468
|
if (isStructured(node)) {
|
|
102
469
|
const transformedContent = (node.content || []).map(c => {
|
|
103
470
|
if (c && c.name === 'row')
|
|
104
471
|
return wrapImplicitCells(c);
|
|
105
472
|
return c;
|
|
106
473
|
});
|
|
107
|
-
return { ...node, content: transformedContent };
|
|
474
|
+
return normalizeCellCounts({ ...node, content: transformedContent }, 'table');
|
|
108
475
|
}
|
|
109
476
|
let rows = [];
|
|
110
477
|
const collectValues = row => {
|
|
@@ -129,6 +496,8 @@ export default () => tree => {
|
|
|
129
496
|
// split each row into lines
|
|
130
497
|
const lines = flattenDeep(rows.map(splitToLines));
|
|
131
498
|
const separators = flattenDeep(seps.map(splitToLines));
|
|
499
|
+
// Rule 3: warn on mixed separator types within a single table
|
|
500
|
+
detectMixedSeparators(lines);
|
|
132
501
|
// collect text rows
|
|
133
502
|
let textRows = [];
|
|
134
503
|
makeTransformer({
|
|
@@ -136,33 +505,48 @@ export default () => tree => {
|
|
|
136
505
|
textRows.push(row.value);
|
|
137
506
|
},
|
|
138
507
|
})(node);
|
|
139
|
-
const columnTemplate = makeMask(lines, separators);
|
|
140
508
|
const makeBlock = (name, content, extra = {}) => {
|
|
141
509
|
return { ...extra, name, type: 'block', content: Array.isArray(content) ? content : [content] };
|
|
142
510
|
};
|
|
143
511
|
const makeRow = cells => makeBlock('row', cells);
|
|
144
512
|
const makeHeaderRow = cells => makeBlock('row', cells, { config: [{ name: 'header', value: true, type: 'boolean' }] });
|
|
145
513
|
const makeCell = text => makeBlock('cell', { type: 'text', value: text });
|
|
146
|
-
//
|
|
514
|
+
// Routing: per-line separator detection (Rule 1) is used only when a
|
|
515
|
+
// line with a visible separator (`|` or `+`) coexists with one that
|
|
516
|
+
// has only whitespace separation — the Scenario 2 case where the
|
|
517
|
+
// legacy positional mask collapses columns. Tables with a uniform
|
|
518
|
+
// separator, or with both visible kinds (pipe + plus) handled by the
|
|
519
|
+
// shared mask, fall back to the legacy positional template, which
|
|
520
|
+
// preserves continuation-line alignment in multi-line rows and keeps
|
|
521
|
+
// byte-for-byte AST/HTML output stable.
|
|
522
|
+
const columnTemplate = makeMask(lines, separators);
|
|
523
|
+
const seenSeparatorKinds = new Set(lines.map(detectLineSeparator));
|
|
524
|
+
const hasVisible = seenSeparatorKinds.has('pipe') || seenSeparatorKinds.has('plus');
|
|
525
|
+
const hasWhitespace = seenSeparatorKinds.has('whitespace');
|
|
526
|
+
const useMixedSplitting = hasVisible && hasWhitespace;
|
|
527
|
+
const splitToCells = (rowValue) => {
|
|
528
|
+
if (useMixedSplitting) {
|
|
529
|
+
const rowLines = rowValue.split(/\r?\n/).filter(l => l.trim() !== '');
|
|
530
|
+
if (rowLines.length <= 1)
|
|
531
|
+
return rowToCells(rowValue);
|
|
532
|
+
}
|
|
533
|
+
return extractColumnsByTemplate(rowValue, columnTemplate);
|
|
534
|
+
};
|
|
147
535
|
const res = makeTransformer({
|
|
148
536
|
'row:text': row => {
|
|
149
537
|
if (textRows.length == 1) {
|
|
150
|
-
//
|
|
538
|
+
// No separator blocks: each line of the only text row becomes its own row
|
|
151
539
|
const textRowsLines = flattenDeep([row.value].map(splitToLines));
|
|
152
|
-
|
|
153
|
-
|
|
154
|
-
|
|
155
|
-
|
|
540
|
+
if (useMixedSplitting) {
|
|
541
|
+
return textRowsLines.map(line => makeRow(splitLineCells(line).map(makeCell)));
|
|
542
|
+
}
|
|
543
|
+
return textRowsLines.map(line => makeRow(extractColumnsByTemplate(line, columnTemplate).map(makeCell)));
|
|
156
544
|
}
|
|
157
|
-
|
|
158
|
-
return makeRow(cols.map(makeCell));
|
|
159
|
-
},
|
|
160
|
-
'head:text': head => {
|
|
161
|
-
const cols = extractColumnsByTemplate(head.value, columnTemplate);
|
|
162
|
-
return makeHeaderRow(cols.map(makeCell));
|
|
545
|
+
return makeRow(splitToCells(row.value).map(makeCell));
|
|
163
546
|
},
|
|
547
|
+
'head:text': head => makeHeaderRow(splitToCells(head.value).map(makeCell)),
|
|
164
548
|
})(node);
|
|
165
|
-
return res;
|
|
549
|
+
return normalizeCellCounts(res, 'table');
|
|
166
550
|
},
|
|
167
551
|
});
|
|
168
552
|
return transformer(tree, {});
|