obitools/obitools4
2
0
Fork 0
mirror of https://github.com/metabarcoding/obitools4.git synced 2026-03-25 13:30:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

Refactor rope scanner and add FASTQ rope parser

Browse Source 
This commit refactors the rope scanner implementation by renaming gbRopeScanner to ropeScanner and extracting the common functionality into a new file. It also introduces a new FastqChunkParserRope function that parses FASTQ chunks directly from a rope without Pack(), enabling more efficient memory usage. The existing parsers are updated to use the new rope-based parser when available. The BioSequence type is enhanced with a TakeQualities method for more efficient quality data handling.
This commit is contained in:
Eric Coissac
2026-03-10 16:46:53 +01:00
parent 623116ab13
commit 3d2e205722
5 changed files with 176 additions and 83 deletions
Download Patch File Download Diff File Expand all files Collapse all files

81
pkg/obiformats/genbank_read.go
View File

@@ -29,70 +29,11 @@ const (
var _seqlenght_rx = regexp.MustCompile(" +([0-9]+) bp")
// gbRopeScanner reads lines from a PieceOfChunk rope without heap allocation.
// The carry buffer (stack) handles lines that span two rope nodes.
type gbRopeScanner struct {
current *PieceOfChunk
pos int
carry [256]byte // max GenBank line = 80 chars; 256 gives ample margin
carryN int
}
func newGbRopeScanner(rope *PieceOfChunk) *gbRopeScanner {
return &gbRopeScanner{current: rope}
}
// ReadLine returns the next line without the trailing \n (or \r\n).
// Returns nil at end of rope. The returned slice aliases carry[] or the node
// data and is valid only until the next ReadLine call.
func (s *gbRopeScanner) ReadLine() []byte {
for {
if s.current == nil {
if s.carryN > 0 {
n := s.carryN
s.carryN = 0
return s.carry[:n]
}
return nil
}
data := s.current.data[s.pos:]
idx := bytes.IndexByte(data, '\n')
if idx >= 0 {
var line []byte
if s.carryN == 0 {
line = data[:idx]
} else {
n := copy(s.carry[s.carryN:], data[:idx])
s.carryN += n
line = s.carry[:s.carryN]
s.carryN = 0
}
s.pos += idx + 1
if s.pos >= len(s.current.data) {
s.current = s.current.Next()
s.pos = 0
}
if len(line) > 0 && line[len(line)-1] == '\r' {
line = line[:len(line)-1]
}
return line
}
// No \n in this node: accumulate into carry and advance
n := copy(s.carry[s.carryN:], data)
s.carryN += n
s.current = s.current.Next()
s.pos = 0
}
}
// extractSequence scans the ORIGIN section byte-by-byte directly on the rope,
// appending compacted bases to dest. Returns the extended slice.
// Stops and returns when "//" is found at the start of a line.
// The scanner is left positioned after the "//" line.
func (s *gbRopeScanner) extractSequence(dest []byte, UtoT bool) []byte {
func (s *ropeScanner) extractSequence(dest []byte, UtoT bool) []byte {
lineStart := true
skipDigits := true
@@ -139,24 +80,6 @@ func (s *gbRopeScanner) extractSequence(dest []byte, UtoT bool) []byte {
return dest
}
// skipToNewline advances the scanner past the next '\n'.
func (s *gbRopeScanner) skipToNewline() {
for s.current != nil {
data := s.current.data[s.pos:]
idx := bytes.IndexByte(data, '\n')
if idx >= 0 {
s.pos += idx + 1
if s.pos >= len(s.current.data) {
s.current = s.current.Next()
s.pos = 0
}
return
}
s.current = s.current.Next()
s.pos = 0
}
}
// parseLseqFromLocus extracts the declared sequence length from a LOCUS line.
// Format: "LOCUS <id> <length> bp ..."
// Returns -1 if not found or parse error.
@@ -205,7 +128,7 @@ func GenbankChunkParserRope(source string, rope *PieceOfChunk,
withFeatureTable, UtoT bool) (obiseq.BioSequenceSlice, error) {
state := inHeader
scanner := newGbRopeScanner(rope)
scanner := newRopeScanner(rope)
sequences := obiseq.MakeBioSequenceSlice(100)[:0]
id := ""