datannurpy v0.22.3

• add: non-quiet exports now print a per-table size report for .json, .json.js, and gzipped .json output
• improve: log prefix spacing is now consistent after leading icons across scan, metadata, export, and post-export output
• fix: log_file now writes in UTF-8 in datannurpy.utils.log
• improve: folder and database scan logs now use clearer resource labels and summaries, with root-relative paths for folder scans and file/table counts shown before dataset totals
• fix: time-series scans now canonicalize header aliases before variable ID build, avoiding duplicate variable IDs from encoding-damaged column names
• fix: time-series grouping now handles more realistic file/table layouts while rejecting incomplete no-year periods
• doc: add shared time-series grouping guide and a make check-py39 validation target
• fix: .xls files that are actually HTML reports are now detected early and skipped cleanly as untreatable, including in schema-only and remote scans