Platob
@Platob
Public Skills
yggdryl-fix
by Platob
Decodes, encodes and streams FIX messages with yggdryl against a FIX dictionary (FixRegistry), in Rust, Python and Node.js. Use when parsing tag=value/SOH frames, bridge rows or FIXML from logs (parse_line / parseLine, parse_text_arrow_reader / parseTextArrowReader), re-emitting wire bytes (into_text / intoText), landing captures in Arrow or Parquet (fix_schema, arrow_reader / arrowReader), chaining order lifecycles (lifecycle, crossuuid), turning FIX into market data and books (market_data, book_arrow_reader), building or storing a dictionary (FixRegistry.from_handle / fromHandle, from_cfb_file, merge_with, commit, code sets, FIX metadata) or running the yggdryl fix CLI.
yggdryl-market-data
by Platob
Models and streams market data with yggdryl's graph layer in Rust, Python and Node.js - orders, quotes, executions and trades as dated events (OrderEvent, QuoteEvent, ExecutionEvent, TradeEvent), identities and side-keyed chains (curruuid, crossuuid, with_previous / withPrevious, EventIterator), order books (BookIterator, BookEvent, limits, best_price / bestPrice, spread, depth, SnapshotEvent), the Side and MarketDataKind enums and the lifted marketdata Arrow row (MarketData.arrow_reader / arrowReader, from_arrow_reader / fromArrowReader, apply_view / applyView). Use when building market events, folding them into books, or persisting or querying marketdata batches.
yggdryl-records
by Platob
Reads and writes rows and Arrow batches on any yggdryl handle - Arrow IPC/Feather, Parquet, Avro, plain-text logs, Iceberg tables and hive-partitioned folders - with streamed readers, pushdown, append/merge (upsert) and commit cadence. Use when calling read_arrow_reader / readArrowReader, overwrite_arrow_* / append_* / merge_*, write_arrow_* with a mode, *_records / readRecords, read_arrow / write_arrow (SerieReader), read_arrow_field / row_size (a file's schema or row count without reading it), pandas or polars frames to and from a file (read_polars_frame, overwrite_pandas_frame, scan_polars), RecordOptions (field, safe, select, filter, merge_by, max_row_size, row_offset, commit_row_size, plan), TextOptions rowheader, iceberg Table create/append/merge/scan, partition pruning, or picking an encoding by suffix. Covers Rust, Python and Node.js.
yggdryl-types
by Platob
Declare yggdryl types and values in Rust, Python and Node.js - parse DataType expressions (Arrow, SQL, Hive, Spark, FIX spellings), build and edit Field schemas (non-null Struct root, metadata, PARQUET:field_id, comment, protocol views, FIELD:enum) and read values through DataType.scalar / Field.scalar into Scalar. Use when choosing a column type (decimal, timestamp/datetime64 zone, string or bytes leaf, uuid, geometry/geography WKB, ccy/isin/forex codes, the side, marketdatakind and state enums, an enumerated StringEnum column, serie/map/union), declaring a schema (Field::new, Field(...), new Field, yggdryl.int64 / fields.int64, @scalar dataclasses, into_field / intoField), adding, replacing or removing a column (set_field, remove_field, unnest_fields), an Arrow/pyarrow schema in or out (from_arrow_schema), converting values (Scalar.from_ / Scalar.from, as_py / asJs) or merging, diffing and walking schemas by path.
yggdryl
by Platob
Routes yggdryl work to the right layer and states the conventions every yggdryl API shares (install and features, naming, defaults, errors, streaming, zero copy) across the Rust crate, the Python wheel and the npm package - Arrow-native schemas (DataType, Field, Scalar), columns (Serie), storage handles (IOBase over local files, ZIP, S3/GCS/Azure, HTTP(S)), record media (Arrow IPC, Parquet, Avro, text, Iceberg), JSON/YAML/TOML/XML, URIs, expressions, xxHash/TxHash, FIX and market data. Use when installing or importing yggdryl, choosing which yggdryl API answers a task, translating yggdryl code between Rust, Python and Node.js, or before any other yggdryl-* skill.
yggdryl-uri
by Platob
Parse, canonicalize, inspect and edit resource identifiers with yggdryl - Uri, Url, Urn, Arn and Scheme - including platform file paths (Windows drives, UNC), percent escapes, credentials, S3/GCS/Azure bucket-key-region, filename suffixes and media type, query parameters, navigation, globs and Hive partitions. Use when parsing a URL or path, from_path / into_path / fromPath / intoPath, into_url / into_urn / intoUrl, joinpath, parameters, is_glob / glob_parts / matches_glob, hive_partitions / partitions / childrenWhere, a URN or ARN locator, bucket vs host rules, or url/urn columns. Covers Rust, Python and Node.js.
yggdryl-storage
by Platob
Read and write bytes through yggdryl's one positional handle (IOBase / Holder) on memory, local files, Arrow filesystems, S3 / Google Cloud Storage / Azure Blob, HTTP(S) URLs, page caches and ZIP archives, with gzip/zlib/zstd codings and charsets; talk HTTP like requests (http.Session, get/post, send_all fan-outs, resumable streams, HTTP/2 and HTTP/3, an in-process test Server). Use when opening a path or URL, reading a range or streaming chunks (read_range_bytes / readRangeBytes, pstream_bytes / pstreamBytes), cursors, listing or globbing a folder (ls, glob, rglob), clear/remove, open/close scopes, compress_into / decompressInto, gzip.dumps / zstd.loads, Charset / charset.decode, digests or read_scalar on a handle, S3Options / S3File, AWS credentials (aws::Session, profiles, SSO, assume role) or a MinIO / S3-compatible endpoint, Buffered caches, Counted call budgets. Covers Rust, Python and Node.js.
yggdryl-hashing
by Platob
Digest bytes, handles, values and Arrow rows with yggdryl's xxHash (XXH32, XXH64, XXH3-64, XXH3-128) and couple digests with instants as TxHash. Use when hashing bytes or a file (xxh3, read_digest / readDigest), streaming a resumable state (Xxh3, write_bytes / writeBytes), seeding XXH3 or giving it a secret, computing a cross-language stable_hash / stableHash of a value, declaring DIGEST:role=holder row-digest columns and filling them (apply_arrow_batch / applyArrowBatch, row_digests), building TxHash / TxHasher values, sortable UUIDv7 keys (into_uuid / intoUuid) or DIGEST:time coupled holders. Covers Rust, Python and Node.js.
yggdryl-arrow
by Platob
Moves columns, tables and streams across the Apache Arrow boundary with yggdryl's Serie (one column), ChunkedSerie (chunked arrays and tables kept apart) and SerieReader (a stream under one compiled plan), and casts them with ArrowCastPlan and ArrowCastOptions (safe, representation). Use when landing arrow-rs arrays / RecordBatches / readers, pyarrow or Arrow JS (apache-arrow) tables and vectors under a Field (from_arrow_array / fromArrowArray, from_arrow_batch, SerieReader.from_arrow_reader), taking pyarrow, pandas, polars or NumPy in via Serie.from_, reading typed buffers (as_int64().values()), casting a column or a stream, or handing data back (into_arrow_array / intoArrowTable, into_pandas / into_polars). Covers Rust, Python and Node.js.
yggdryl-documents
by Platob
Parses and writes JSON, JSON Lines, YAML (multi-document), TOML and XML documents as yggdryl Scalar values or native objects, typed and validated by an optional Field, bounded by limits, with opt-in {{ }} and environment-variable placeholders. Use when loading or validating a config file, calling json/yaml/toml/xml loads, dumps, loads_all / loadsAll, dump_all / dumpAll, load_all / loadAll, from_json_scalar / into_json_scalar / from_json_scalar_with_field (and the yaml/toml/xml twins), json::from_utf8 / from_reader / into_writer, decoding into a dataclass with cls=, typing decimals/dates with field=, mapping XML attributes (@name) and #text, inferring a document's format, or read_scalar / write_scalar on a handle. Covers Rust, Python and Node.js.
yggdryl-expressions
by Platob
Parse, bind and evaluate yggdryl expressions - Term, Filter, Selector, Plan, Expression, FieldPath - over rows, Arrow batches and streams, and push where/select into record reads. Use when writing a where or select clause or an SQL-like plan (insert into, upsert ... by, delete from, execute), filtering or projecting Arrow data (apply_arrow_reader / applyArrowReader, apply_arrow_batch / applyArrowBatch, apply_records / applyRecords), binding parameters, pruning by statistics (Bounds, statistics_prune), partition_split, user-defined functions, derived (TRANSFORM:expression) columns or nested field paths. Covers Rust, Python and Node.js.