mirror of
https://github.com/hanzoai/datastore.git
synced 2026-09-26 13:55:07 +00:00
Alexey MilovidovandClaude Opus 4.8
4c956bc220
Extend the generator to formats (generate-formats.sql + format_generators in the port, same mechanism as engines). Convert the 65 format pages whose structured docs reproduce the page (content-neutral / whitespace-only). Deferred (kept hand-written): pages that include shared Mintlify snippets (<PrettyFormatSettings/>, <RowBinaryFormatSettings/>, <DataTypeMapping/> from /snippets/) which the Docusaurus->Mintlify transform can't synthesize, and a few (Values, HiveText, Arrow, ArrowStream, Parquet) whose structured description is less complete than the page and needs enriching. Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
21 lines
1.0 KiB
SQL
21 lines
1.0 KiB
SQL
-- Format documentation generator.
|
|
-- Usage: one page per format, driven by the doc page's basename, e.g.:
|
|
-- clickhouse --param_format='JSONEachRow' --queries-file generate-formats.sql
|
|
--
|
|
-- system.formats.description holds the full Markdown body of the format's
|
|
-- reference page (populated from the `Documentation` attached via
|
|
-- FormatFactory::setDocumentation). We emit that body verbatim; the Python
|
|
-- port then applies the same Docusaurus->Mintlify body transforms it runs on
|
|
-- migrated pages. The page basename is matched to the format name case- and
|
|
-- separator-insensitively; format aliases carry only a short "Alias of ..."
|
|
-- description, so notEmpty()+LIMIT 1 selects the canonical format. Pages that
|
|
-- are not a single format (index) produce no output and are skipped.
|
|
SELECT description
|
|
FROM system.formats
|
|
WHERE lower(replaceRegexpAll(name, '[^0-9a-zA-Z]', ''))
|
|
= lower(replaceRegexpAll({format:String}, '[^0-9a-zA-Z]', ''))
|
|
AND notEmpty(description)
|
|
ORDER BY name ASC
|
|
LIMIT 1
|
|
INTO OUTFILE 'temp-format.md' TRUNCATE FORMAT LineAsString
|