alex/yomichan-import
1
Fork 0
Code Releases Activity

Rename variables for consistency

Browse Source
This commit is contained in:
stephenmk 2023-01-23 14:09:50 -06:00
parent d8a3b420ee
commit 6726c5245b
No known key found for this signature in database
GPG Key ID: B6DA730DB06235F1
1 changed files with 19 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
jmdictReferences.go
View File

@ -87,10 +87,10 @@ func (meta *jmdictMetadata) MakeReferenceToSeqMap() {
func (meta *jmdictMetadata) MakeHashToSearchValuesMap() { func (meta *jmdictMetadata) MakeHashToSearchValuesMap() {
meta.hashToSearchValues = make(map[hash][]searchValue) meta.hashToSearchValues = make(map[hash][]searchValue)
for seq, searchHashes := range meta.seqToSearchHashes { for seq, searchHashes := range meta.seqToSearchHashes {
for score, searchHash := range searchHashes { for idx, searchHash := range searchHashes {
searchValue := searchValue{ searchValue := searchValue{
sequence: seq, sequence: seq,
index: score, index: idx,
isPriority: searchHash.isPriority, isPriority: searchHash.isPriority,
} }
meta.hashToSearchValues[searchHash.hash] = meta.hashToSearchValues[searchHash.hash] =
@ -100,6 +100,10 @@ func (meta *jmdictMetadata) MakeHashToSearchValuesMap() {
} }
/* /*
* This function attemps to convert a JMdict reference string into a
* single definite sequence number. These reference strings are often
* ambiguous, so we have to resort to using heuristics.
*
* Generally, correspondence is determined by the order in which term * Generally, correspondence is determined by the order in which term
* pairs are extracted from each JMdict entry. Take for example the * pairs are extracted from each JMdict entry. Take for example the
* JMdict entry for ご本, which contains a reference to 本 (without a * JMdict entry for ご本, which contains a reference to 本 (without a
@ -115,7 +119,7 @@ func (meta *jmdictMetadata) MakeHashToSearchValuesMap() {
* returned. * returned.
* *
* In situations in which multiple sequences are found with the same * In situations in which multiple sequences are found with the same
* score, the entry with a priority tag ("news1", "ichi1", "spec1", * index, the entry with a priority tag ("news1", "ichi1", "spec1",
* "spec2", "gai1") is given preference. This mostly affects * "spec2", "gai1") is given preference. This mostly affects
* katakana-only loanwords like ラグ. * katakana-only loanwords like ラグ.
* *
@ -129,8 +133,8 @@ func (meta *jmdictMetadata) MakeHashToSearchValuesMap() {
* *
* All else being equal, the entry with the smallest sequence number * All else being equal, the entry with the smallest sequence number
* is chosen. References in the JMdict file are currently ambiguous, * is chosen. References in the JMdict file are currently ambiguous,
* and getting this perfect won't be possible until sequence numbers * and getting this perfect won't be possible until reference sequence
* are explictly identified in these references. See: * numbers are included in the file. See:
* https://github.com/JMdictProject/JMdictIssues/issues/61 * https://github.com/JMdictProject/JMdictIssues/issues/61
*/ */
func (meta *jmdictMetadata) FindBestSequence(reference string) sequence { func (meta *jmdictMetadata) FindBestSequence(reference string) sequence {
@ -142,24 +146,24 @@ func (meta *jmdictMetadata) FindBestSequence(reference string) sequence {
return bestSeq return bestSeq
} }
hash := headword.Hash() hash := headword.Hash()
for _, seqScore := range meta.hashToSearchValues[hash] { for _, v := range meta.hashToSearchValues[hash] {
if meta.seqToSenseCount[seqScore.sequence] < senseNumber { if meta.seqToSenseCount[v.sequence] < senseNumber {
// entry must contain the specified sense // entry must contain the specified sense
continue continue
} else if lowestIndex < seqScore.index { } else if lowestIndex < v.index {
// lower indices are better // lower indices are better
continue continue
} else if (lowestIndex == seqScore.index) && (bestIsPriority && !seqScore.isPriority) { } else if (lowestIndex == v.index) && (bestIsPriority && !v.isPriority) {
// if scores match, check priority // if indices match, check priority
continue continue
} else if (lowestIndex == seqScore.index) && (bestIsPriority == seqScore.isPriority) && (bestSeq < seqScore.sequence) { } else if (lowestIndex == v.index) && (bestIsPriority == v.isPriority) && (bestSeq < v.sequence) {
// if scores and priority match, check sequence number. // if indices and priority match, check sequence number.
// lower sequence numbers are better // lower sequence numbers are better
continue continue
} else { } else {
lowestIndex = seqScore.index lowestIndex = v.index
bestSeq = seqScore.sequence bestSeq = v.sequence
bestIsPriority = seqScore.isPriority bestIsPriority = v.isPriority
} }
} }
return bestSeq return bestSeq