dsq

• Correctly handle queries on nested JSON data courtesy of @xwjdsh (see https://github.com/multiprocessio/dsq/issues/91 and https://github.com/multiprocessio/dsq/issues/96)

• Add logfmt support courtesy of @0michalsokolowski0 (see https://github.com/multiprocessio/dsq/pull/85)

• Adds support for reading YAML files (see https://github.com/multiprocessio/datastation/pull/275) courtesy of @Sajuno
• Fixes Parquet reader (see #81 ) courtesy of @Sajuno

• Properly quote fields when using new CSV writer (see #74 )
• Handle --convert-numbers flag in new SQLiteWriter by using NUMERIC field type, no conversions in Go
• Properly handle no query case by disabling SQLiteWriter in this case so intermediate JSON file does get written

• Limit SQLite writer to only CSV, TSV and Regexp newline files for now since nested objects aren't translated the same way yet in the SQLite writer.

• Speed up most runs of dsq on large files 2x by skipping intermediate writes to JSON, writing directly from file to SQLite (see https://github.com/multiprocessio/datastation/pull/267 for details). If this breaks things for you you can add the --no-sqlite-writer flag to go back to the old code path. This path isn't taken for JSON, Excel, and ODS files or when schema or --convert-numbers is set.

• -n, --convert-numbers flag to automatically convert values to integers or floats when reading CSV/TSV; courtesy of @fritzgrabo

• New hashing/encoding functions: base64, from_base64, md5, sha1, sha256, sha512, etc. See here for the full list.
• New string functions: len, split_part, regexp_count, regexp_split_part. See here for details.

• Bring in go-sqlite3-stdlib with functions for parsing URLs, best-effort date parsing, string and math helpers, and aggregate functions like percentiles, stddev, median, mode, etc. See that repo README for full details.

• Support for a basic interactive REPL with the -i or --interactive flag courtesy of @Ghibranalj

• Fixes dsq -v, dsq -h panic after printing version and help text. Introduced in 0.15.0

• Support for -f|--file flag to pass query as file instead of string in terminal, courtesy of @Ghibranalj
• Support for --cache|-C flag to cache imported data, courtesy of @Posrabi (see README for details)

• Fix for empty cells in XLSX (see https://github.com/multiprocessio/datastation/pull/218 for details) courtesy of @steirico

• Support for Apache Avro (courtesy of @Posrabi)
• Support for dumping inferred file schema (rather than dumping data or running a query) with --schema flag

• Bump JSONL line length limit to 1MB (from 64k) (see https://github.com/multiprocessio/datastation/pull/209 for details) courtesy of @fritzgrabo

• Prevent over-eager merging of Scalars of same type into Varied kind (see https://github.com/multiprocessio/datastation/pull/205 for details)

• Consistent sort --pretty table columns

• Support for loading and querying nested array columns

Fix for https://github.com/multiprocessio/dsq/issues/34

Adds -v/--version flag to show version.