Search with a reusable 'sassy' searcher — sassy_searcher

pattern and text must be lists of sequences. Each element must be a raw vector or a non-missing character scalar. Every pattern is searched against every text and the returned pattern_idx and text_idx columns identify the 0-based input indices. Use threads > 1 for larger batches.

Usage

sassy_searcher_search(
  searcher,
  pattern,
  text,
  k,
  all = FALSE,
  threads = 1L,
  strategy = "pairwise",
  pattern_id = NULL,
  text_id = NULL,
  match_region = FALSE,
  sam = FALSE
)

Arguments

searcher: A searcher created by sassy_searcher().
pattern: List of raw vectors or non-missing character scalars.
text: List of raw vectors or non-missing character scalars.
k: Maximum edit distance.
all: If FALSE, return the usual local-minimum matches. If TRUE, return every end position with score <= k; this can include overlapping and nested candidate alignments and requires strategy = "pairwise".
threads: Number of worker threads to request for bulk searches.
strategy: Search strategy. "pairwise" searches each pattern/text pair independently and is the general default. "batch_texts" uses one text per SIMD lane. "batch_patterns" and "encoded_patterns" (alias "v2") use Sassy's multi-pattern encoding, which in sassy 0.2.1 is implemented for alphabet = "iupac" and equal byte-length patterns.
pattern_id: Optional pattern identifiers. If supplied, must be a non-missing character vector with one entry per pattern and adds/replaces a pattern_id column. Names on pattern are not inspected.
text_id: Optional text identifiers. If supplied, must be a non-missing character vector with one entry per text and adds/replaces a text_id column. Names on text are not inspected.
match_region: If TRUE, include a match_region column. Reverse-strand regions are reverse-complemented so the region and CIGAR are in the input pattern direction.
sam: If TRUE, format reverse-strand match_region and cigar in the text direction used by SAM and by the upstream sassy --sam output.

Value

A data frame with 0-based indices and coordinates: pattern_idx, text_idx, text_start, text_end, pattern_start, pattern_end, cost, strand, and cigar. If pattern_id or text_id are supplied, mapped identifier columns are included. If requested, also includes match_region. Rows are ordered by input text, then text start/end coordinate, then pattern index.