inferJoinBlocks

private fn

CsvCandidate[] inferJoinBlocks(const  CsvTextBlock[]  blocks,
       size_t width,  const  CsvReadPolicy read,  bool  tryParseTemporal,
       bool  hasHeader,  ddn.data.csv.CsvDialect dialect)

Full-scan inference over the text blocks: each block joins its own per-column candidates (block 0 past the header), and the per-block results join in block order. The join is associative and commutative, so the block count and boundaries - which the split derives from the bytes - cannot change the schema. Any worker error (a ragged row) propagates; the caller's serial re-walk names the canonical first one.

Parameters

`blocks`	the text blocks (all rows; block 0 may start with the header row)
`width`	the schema width
`read`	the reading policy
`tryParseTemporal`	whether temporal candidates are tested
`hasHeader`	whether block 0's first row is the header
`dialect`	the dialect (header forced off)

Returns

the per-column joined candidates.