@databio/gtars 0.5.2 → 0.6.0

package/gtars_js.d.ts

@@ -1,146 +1,377 @@
 /* tslint:disable */
 /* eslint-disable */
-export function greet(name: string): void;
+
+export class BedClassificationOutput {
+    private constructor();
+    free(): void;
+    [Symbol.dispose](): void;
+    readonly bed_compliance: string;
+    readonly compliant_columns: number;
+    readonly data_format: string;
+    readonly non_compliant_columns: number;
+}
+
 export class ChromosomeStatistics {
-  private constructor();
-  free(): void;
-  [Symbol.dispose](): void;
-  readonly chromosome: string;
-  readonly start_nucleotide_position: number;
-  readonly end_nucleotide_position: number;
-  readonly number_of_regions: number;
-  readonly minimum_region_length: number;
-  readonly maximum_region_length: number;
-  readonly mean_region_length: number;
-  readonly median_region_length: number;
+    private constructor();
+    free(): void;
+    [Symbol.dispose](): void;
+    readonly chromosome: string;
+    readonly end_nucleotide_position: number;
+    readonly maximum_region_length: number;
+    readonly mean_region_length: number;
+    readonly median_region_length: number;
+    readonly minimum_region_length: number;
+    readonly number_of_regions: number;
+    readonly start_nucleotide_position: number;
 }
+
 export class Overlapper {
-  free(): void;
-  [Symbol.dispose](): void;
-  constructor(universe: any, backend: string);
-  get_backend(): string;
-  find(regions: any): any;
+    free(): void;
+    [Symbol.dispose](): void;
+    find(regions: any): any;
+    get_backend(): string;
+    constructor(universe: any, backend: string);
 }
+
 export class RegionDistribution {
-  private constructor();
-  free(): void;
-  [Symbol.dispose](): void;
+    private constructor();
+    free(): void;
+    [Symbol.dispose](): void;
 }
+
 export class RegionSet {
-  free(): void;
-  [Symbol.dispose](): void;
-  constructor(regions: any);
-  chromosomeStatistics(): any;
-  regionDistribution(n_bins: number): any;
-  readonly numberOfRegions: number;
-  readonly meanRegionWidth: number;
-  readonly nucleotidesLength: number;
-  readonly identifier: string;
+    free(): void;
+    [Symbol.dispose](): void;
+    chromosomeStatistics(): any;
+    constructor(regions: any);
+    regionDistribution(n_bins: number): any;
+    readonly classify: BedClassificationOutput;
+    readonly firstRegion: string;
+    readonly identifier: string;
+    readonly meanRegionWidth: number;
+    readonly nucleotidesLength: number;
+    readonly numberOfRegions: number;
 }
+
 export class Tokenizer {
-  free(): void;
-  [Symbol.dispose](): void;
-  constructor(universe: any, backend: string);
-  tokenize(regions: any): any;
-  encode(tokens: any): any;
-  decode(ids: any): any;
-  specialTokensMap(): any;
-  getVocab(): any;
-  readonly unkToken: string;
-  readonly padToken: string;
-  readonly maskToken: string;
-  readonly clsToken: string;
-  readonly bosToken: string;
-  readonly eosToken: string;
-  readonly sepToken: string;
-  readonly padTokenId: number;
-  readonly maskTokenId: number;
-  readonly clsTokenId: number;
-  readonly bosTokenId: number;
-  readonly eosTokenId: number;
-  readonly sepTokenId: number;
-  readonly unkTokenId: number;
-  readonly vocabSize: number;
-  readonly length: number;
+    free(): void;
+    [Symbol.dispose](): void;
+    decode(ids: any): any;
+    encode(tokens: any): any;
+    getVocab(): any;
+    constructor(universe: any, backend: string);
+    specialTokensMap(): any;
+    tokenize(regions: any): any;
+    readonly bosToken: string;
+    readonly bosTokenId: number;
+    readonly clsToken: string;
+    readonly clsTokenId: number;
+    readonly eosToken: string;
+    readonly eosTokenId: number;
+    readonly length: number;
+    readonly maskToken: string;
+    readonly maskTokenId: number;
+    readonly padToken: string;
+    readonly padTokenId: number;
+    readonly sepToken: string;
+    readonly sepTokenId: number;
+    readonly unkToken: string;
+    readonly unkTokenId: number;
+    readonly vocabSize: number;
 }
 
+/**
+ * Canonicalize a JSON string according to RFC-8785.
+ *
+ * This is useful for computing digests of JSON objects in a deterministic way.
+ *
+ * # Arguments
+ * * `json_str` - A JSON string
+ *
+ * # Returns
+ * The canonicalized JSON string
+ */
+export function canonicalizeJsonString(json_str: string): string;
+
+/**
+ * Compute the MD5 digest of a string.
+ *
+ * Low-level function that does NOT uppercase the input.
+ *
+ * # Arguments
+ * * `input` - The input string
+ *
+ * # Returns
+ * The MD5 digest as a hex string
+ */
+export function computeMd5(input: string): string;
+
+/**
+ * Compute the sha512t24u digest of a string.
+ *
+ * This is a low-level function that computes the GA4GH standard digest
+ * of arbitrary input. Unlike `sequence_digest`, this does NOT uppercase
+ * the input.
+ *
+ * # Arguments
+ * * `input` - The input string
+ *
+ * # Returns
+ * The sha512t24u digest string (base64url encoded)
+ */
+export function computeSha512t24u(input: string): string;
+
+/**
+ * Compute a seqcol (sequence collection) digest from FASTA content.
+ *
+ * This function takes FASTA file content as bytes and returns a JSON object
+ * containing the sequence collection digest and metadata. Automatically handles
+ * gzip-compressed FASTA files.
+ *
+ * # Arguments
+ * * `fasta_content` - The FASTA file content as a Uint8Array (supports gzip)
+ *
+ * # Returns
+ * A JavaScript object containing:
+ * - `digest`: The top-level seqcol digest
+ * - `names_digest`: Level 1 digest of sequence names
+ * - `sequences_digest`: Level 1 digest of sequence digests
+ * - `lengths_digest`: Level 1 digest of sequence lengths
+ * - `n_sequences`: Number of sequences
+ * - `sequences`: Array of sequence metadata objects
+ *
+ * # Example (JavaScript)
+ * ```javascript
+ * const fastaContent = new TextEncoder().encode(">chr1\nACGT\n");
+ * const result = digestSeqcol(fastaContent);
+ * console.log(result.digest);  // Top-level seqcol digest
+ * ```
+ */
+export function digestSeqcol(fasta_content: Uint8Array): any;
+
+/**
+ * Finalize the hasher and return the results.
+ *
+ * This consumes the hasher and frees its resources.
+ * After calling this, the handle is no longer valid.
+ *
+ * # Arguments
+ * * `handle` - The hasher handle from `fastaHasherNew()`
+ *
+ * # Returns
+ * A JavaScript object with the sequence collection result,
+ * or an error if the handle is invalid.
+ *
+ * # Example (JavaScript)
+ * ```javascript
+ * const handle = fastaHasherNew();
+ * fastaHasherUpdate(handle, fastaData);
+ * const result = fastaHasherFinish(handle);
+ * console.log(result.digest);
+ * ```
+ */
+export function fastaHasherFinish(handle: number): any;
+
+/**
+ * Free a hasher without getting results.
+ *
+ * Use this if you need to cancel processing early.
+ * After calling this, the handle is no longer valid.
+ *
+ * # Arguments
+ * * `handle` - The hasher handle from `fastaHasherNew()`
+ *
+ * # Returns
+ * `true` if the handle was valid and freed, `false` otherwise.
+ */
+export function fastaHasherFree(handle: number): boolean;
+
+/**
+ * Create a new streaming FASTA hasher.
+ *
+ * Returns a handle (u32) that must be used in subsequent calls.
+ * The handle must be freed with `fastaHasherFinish()` or `fastaHasherFree()`.
+ *
+ * # Returns
+ * A handle ID (always > 0).
+ *
+ * # Example (JavaScript)
+ * ```javascript
+ * const handle = fastaHasherNew();
+ * // ... use handle with fastaHasherUpdate ...
+ * const result = fastaHasherFinish(handle);
+ * ```
+ */
+export function fastaHasherNew(): number;
+
+/**
+ * Get the current progress of a streaming hasher.
+ *
+ * # Arguments
+ * * `handle` - The hasher handle from `fastaHasherNew()`
+ *
+ * # Returns
+ * A JavaScript object with progress information, or null if handle is invalid.
+ * The object contains:
+ * - `completed_sequences`: Number of fully processed sequences
+ * - `current_sequence_name`: Name of sequence being processed (if any)
+ * - `current_sequence_length`: Length of sequence being processed
+ */
+export function fastaHasherProgress(handle: number): any;
+
+/**
+ * Process a chunk of FASTA data.
+ *
+ * This can be called multiple times with successive chunks.
+ * Handles both plain text and gzip-compressed FASTA.
+ *
+ * # Arguments
+ * * `handle` - The hasher handle from `fastaHasherNew()`
+ * * `chunk` - A chunk of FASTA data as Uint8Array
+ *
+ * # Returns
+ * Ok on success, or an error with details about what went wrong.
+ *
+ * # Example (JavaScript)
+ * ```javascript
+ * const handle = fastaHasherNew();
+ * fastaHasherUpdate(handle, new TextEncoder().encode(">chr1\nACGT"));
+ * fastaHasherUpdate(handle, new TextEncoder().encode("\n>chr2\nTGCA\n"));
+ * ```
+ */
+export function fastaHasherUpdate(handle: number, chunk: Uint8Array): void;
+
+export function greet(name: string): void;
+
+/**
+ * Compute a sequence digest (sha512t24u) from raw sequence data.
+ *
+ * This is the GA4GH standard digest for sequences. The sequence is uppercased
+ * before computing the digest.
+ *
+ * # Arguments
+ * * `sequence` - The sequence string (A, C, G, T, etc.)
+ *
+ * # Returns
+ * The sha512t24u digest string (base64url encoded)
+ *
+ * # Example (JavaScript)
+ * ```javascript
+ * const digest = sequenceDigest("ACGT");
+ * console.log(digest);  // "aKF498dAxcJAqme6QYQ7EZ07-fiw8Kw2"
+ * ```
+ */
+export function sequenceDigest(sequence: string): string;
+
+/**
+ * Compute an MD5 digest from raw sequence data.
+ *
+ * The sequence is uppercased before computing the digest.
+ *
+ * # Arguments
+ * * `sequence` - The sequence string
+ *
+ * # Returns
+ * The MD5 digest as a hex string
+ */
+export function sequenceMd5(sequence: string): string;
+
 export type InitInput = RequestInfo | URL | Response | BufferSource | WebAssembly.Module;
 
 export interface InitOutput {
-  readonly memory: WebAssembly.Memory;
-  readonly __wbg_tokenizer_free: (a: number, b: number) => void;
-  readonly tokenizer_new: (a: any, b: number, c: number) => [number, number, number];
-  readonly tokenizer_tokenize: (a: number, b: any) => [number, number, number];
-  readonly tokenizer_encode: (a: number, b: any) => [number, number, number];
-  readonly tokenizer_decode: (a: number, b: any) => [number, number, number];
-  readonly tokenizer_unkToken: (a: number) => [number, number];
-  readonly tokenizer_padToken: (a: number) => [number, number];
-  readonly tokenizer_maskToken: (a: number) => [number, number];
-  readonly tokenizer_clsToken: (a: number) => [number, number];
-  readonly tokenizer_bosToken: (a: number) => [number, number];
-  readonly tokenizer_eosToken: (a: number) => [number, number];
-  readonly tokenizer_sepToken: (a: number) => [number, number];
-  readonly tokenizer_padTokenId: (a: number) => number;
-  readonly tokenizer_maskTokenId: (a: number) => number;
-  readonly tokenizer_clsTokenId: (a: number) => number;
-  readonly tokenizer_bosTokenId: (a: number) => number;
-  readonly tokenizer_eosTokenId: (a: number) => number;
-  readonly tokenizer_sepTokenId: (a: number) => number;
-  readonly tokenizer_unkTokenId: (a: number) => number;
-  readonly tokenizer_specialTokensMap: (a: number) => [number, number, number];
-  readonly tokenizer_getVocab: (a: number) => [number, number, number];
-  readonly tokenizer_length: (a: number) => number;
-  readonly tokenizer_vocabSize: (a: number) => number;
-  readonly __wbg_chromosomestatistics_free: (a: number, b: number) => void;
-  readonly chromosomestatistics_chromosome: (a: number) => [number, number];
-  readonly chromosomestatistics_start_nucleotide_position: (a: number) => number;
-  readonly chromosomestatistics_end_nucleotide_position: (a: number) => number;
-  readonly chromosomestatistics_number_of_regions: (a: number) => number;
-  readonly chromosomestatistics_minimum_region_length: (a: number) => number;
-  readonly chromosomestatistics_maximum_region_length: (a: number) => number;
-  readonly chromosomestatistics_mean_region_length: (a: number) => number;
-  readonly chromosomestatistics_median_region_length: (a: number) => number;
-  readonly __wbg_regiondistribution_free: (a: number, b: number) => void;
-  readonly __wbg_regionset_free: (a: number, b: number) => void;
-  readonly regionset_new: (a: any) => [number, number, number];
-  readonly regionset_numberOfRegions: (a: number) => number;
-  readonly regionset_meanRegionWidth: (a: number) => number;
-  readonly regionset_nucleotidesLength: (a: number) => number;
-  readonly regionset_identifier: (a: number) => [number, number];
-  readonly regionset_chromosomeStatistics: (a: number) => [number, number, number];
-  readonly regionset_regionDistribution: (a: number, b: number) => [number, number, number];
-  readonly __wbg_overlapper_free: (a: number, b: number) => void;
-  readonly overlapper_new: (a: any, b: number, c: number) => [number, number, number];
-  readonly overlapper_get_backend: (a: number) => [number, number];
-  readonly overlapper_find: (a: number, b: any) => [number, number, number];
-  readonly greet: (a: number, b: number) => void;
-  readonly __wbindgen_malloc: (a: number, b: number) => number;
-  readonly __wbindgen_realloc: (a: number, b: number, c: number, d: number) => number;
-  readonly __wbindgen_exn_store: (a: number) => void;
-  readonly __externref_table_alloc: () => number;
-  readonly __wbindgen_externrefs: WebAssembly.Table;
-  readonly __externref_table_dealloc: (a: number) => void;
-  readonly __wbindgen_free: (a: number, b: number, c: number) => void;
-  readonly __wbindgen_start: () => void;
+    readonly memory: WebAssembly.Memory;
+    readonly greet: (a: number, b: number) => void;
+    readonly canonicalizeJsonString: (a: number, b: number) => [number, number, number, number];
+    readonly computeMd5: (a: number, b: number) => [number, number];
+    readonly computeSha512t24u: (a: number, b: number) => [number, number];
+    readonly digestSeqcol: (a: number, b: number) => [number, number, number];
+    readonly fastaHasherFinish: (a: number) => [number, number, number];
+    readonly fastaHasherFree: (a: number) => number;
+    readonly fastaHasherNew: () => number;
+    readonly fastaHasherProgress: (a: number) => [number, number, number];
+    readonly fastaHasherUpdate: (a: number, b: number, c: number) => [number, number];
+    readonly sequenceDigest: (a: number, b: number) => [number, number];
+    readonly sequenceMd5: (a: number, b: number) => [number, number];
+    readonly __wbg_bedclassificationoutput_free: (a: number, b: number) => void;
+    readonly __wbg_chromosomestatistics_free: (a: number, b: number) => void;
+    readonly __wbg_regiondistribution_free: (a: number, b: number) => void;
+    readonly __wbg_regionset_free: (a: number, b: number) => void;
+    readonly bedclassificationoutput_bed_compliance: (a: number) => [number, number];
+    readonly bedclassificationoutput_compliant_columns: (a: number) => number;
+    readonly bedclassificationoutput_data_format: (a: number) => [number, number];
+    readonly bedclassificationoutput_non_compliant_columns: (a: number) => number;
+    readonly chromosomestatistics_chromosome: (a: number) => [number, number];
+    readonly chromosomestatistics_end_nucleotide_position: (a: number) => number;
+    readonly chromosomestatistics_maximum_region_length: (a: number) => number;
+    readonly chromosomestatistics_mean_region_length: (a: number) => number;
+    readonly chromosomestatistics_median_region_length: (a: number) => number;
+    readonly chromosomestatistics_number_of_regions: (a: number) => number;
+    readonly chromosomestatistics_start_nucleotide_position: (a: number) => number;
+    readonly regionset_chromosomeStatistics: (a: number) => [number, number, number];
+    readonly regionset_classify: (a: number) => number;
+    readonly regionset_firstRegion: (a: number) => [number, number];
+    readonly regionset_identifier: (a: number) => [number, number];
+    readonly regionset_meanRegionWidth: (a: number) => number;
+    readonly regionset_new: (a: any) => [number, number, number];
+    readonly regionset_nucleotidesLength: (a: number) => number;
+    readonly regionset_numberOfRegions: (a: number) => number;
+    readonly regionset_regionDistribution: (a: number, b: number) => [number, number, number];
+    readonly chromosomestatistics_minimum_region_length: (a: number) => number;
+    readonly __wbg_tokenizer_free: (a: number, b: number) => void;
+    readonly tokenizer_bosToken: (a: number) => [number, number];
+    readonly tokenizer_bosTokenId: (a: number) => number;
+    readonly tokenizer_clsToken: (a: number) => [number, number];
+    readonly tokenizer_clsTokenId: (a: number) => number;
+    readonly tokenizer_decode: (a: number, b: any) => [number, number, number];
+    readonly tokenizer_encode: (a: number, b: any) => [number, number, number];
+    readonly tokenizer_eosToken: (a: number) => [number, number];
+    readonly tokenizer_eosTokenId: (a: number) => number;
+    readonly tokenizer_getVocab: (a: number) => [number, number, number];
+    readonly tokenizer_length: (a: number) => number;
+    readonly tokenizer_maskToken: (a: number) => [number, number];
+    readonly tokenizer_maskTokenId: (a: number) => number;
+    readonly tokenizer_new: (a: any, b: number, c: number) => [number, number, number];
+    readonly tokenizer_padToken: (a: number) => [number, number];
+    readonly tokenizer_padTokenId: (a: number) => number;
+    readonly tokenizer_sepToken: (a: number) => [number, number];
+    readonly tokenizer_sepTokenId: (a: number) => number;
+    readonly tokenizer_specialTokensMap: (a: number) => [number, number, number];
+    readonly tokenizer_tokenize: (a: number, b: any) => [number, number, number];
+    readonly tokenizer_unkToken: (a: number) => [number, number];
+    readonly tokenizer_unkTokenId: (a: number) => number;
+    readonly tokenizer_vocabSize: (a: number) => number;
+    readonly __wbg_overlapper_free: (a: number, b: number) => void;
+    readonly overlapper_find: (a: number, b: any) => [number, number, number];
+    readonly overlapper_get_backend: (a: number) => [number, number];
+    readonly overlapper_new: (a: any, b: number, c: number) => [number, number, number];
+    readonly __wbindgen_malloc: (a: number, b: number) => number;
+    readonly __wbindgen_realloc: (a: number, b: number, c: number, d: number) => number;
+    readonly __wbindgen_exn_store: (a: number) => void;
+    readonly __externref_table_alloc: () => number;
+    readonly __wbindgen_externrefs: WebAssembly.Table;
+    readonly __wbindgen_free: (a: number, b: number, c: number) => void;
+    readonly __externref_table_dealloc: (a: number) => void;
+    readonly __wbindgen_start: () => void;
 }
 
 export type SyncInitInput = BufferSource | WebAssembly.Module;
+
 /**
-* Instantiates the given `module`, which can either be bytes or
-* a precompiled `WebAssembly.Module`.
-*
-* @param {{ module: SyncInitInput }} module - Passing `SyncInitInput` directly is deprecated.
-*
-* @returns {InitOutput}
-*/
+ * Instantiates the given `module`, which can either be bytes or
+ * a precompiled `WebAssembly.Module`.
+ *
+ * @param {{ module: SyncInitInput }} module - Passing `SyncInitInput` directly is deprecated.
+ *
+ * @returns {InitOutput}
+ */
 export function initSync(module: { module: SyncInitInput } | SyncInitInput): InitOutput;
 
 /**
-* If `module_or_path` is {RequestInfo} or {URL}, makes a request and
-* for everything else, calls `WebAssembly.instantiate` directly.
-*
-* @param {{ module_or_path: InitInput | Promise<InitInput> }} module_or_path - Passing `InitInput` directly is deprecated.
-*
-* @returns {Promise<InitOutput>}
-*/
+ * If `module_or_path` is {RequestInfo} or {URL}, makes a request and
+ * for everything else, calls `WebAssembly.instantiate` directly.
+ *
+ * @param {{ module_or_path: InitInput | Promise<InitInput> }} module_or_path - Passing `InitInput` directly is deprecated.
+ *
+ * @returns {Promise<InitOutput>}
+ */
 export default function __wbg_init (module_or_path?: { module_or_path: InitInput | Promise<InitInput> } | InitInput | Promise<InitInput>): Promise<InitOutput>;