@datagrok/bio 2.4.5 → 2.4.7
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/detectors.js
CHANGED
|
@@ -165,7 +165,7 @@ class BioPackageDetectors extends DG.Package {
|
|
|
165
165
|
if (statsAsChars.sameLength) {
|
|
166
166
|
const stats = this.getStats(categoriesSample, seqMinLength, splitter);
|
|
167
167
|
const alphabet = this.detectAlphabet(stats.freq, candidateAlphabets, '-', colNameLikely);
|
|
168
|
-
if (alphabet === ALPHABET.UN
|
|
168
|
+
if (alphabet === ALPHABET.UN) return null;
|
|
169
169
|
|
|
170
170
|
col.setTag(DG.TAGS.UNITS, units);
|
|
171
171
|
if (separator) col.setTag(UnitsHandler.TAGS.separator, separator);
|
|
@@ -193,8 +193,7 @@ class BioPackageDetectors extends DG.Package {
|
|
|
193
193
|
|
|
194
194
|
// TODO: If separator detected, then extra efforts to detect alphabet are allowed.
|
|
195
195
|
const alphabet = this.detectAlphabet(stats.freq, candidateAlphabets, gapSymbol, colNameLikely);
|
|
196
|
-
|
|
197
|
-
if (units === NOTATION.FASTA && alphabet === ALPHABET.UN && !alphabetIsMultichar && !colNameLikely) return null;
|
|
196
|
+
if (units === NOTATION.FASTA && alphabet === ALPHABET.UN && !alphabetIsMultichar) return null;
|
|
198
197
|
|
|
199
198
|
// const forbidden = this.checkForbiddenWoSeparator(stats.freq);
|
|
200
199
|
col.setTag(DG.TAGS.UNITS, units);
|
|
@@ -248,8 +247,9 @@ class BioPackageDetectors extends DG.Package {
|
|
|
248
247
|
}
|
|
249
248
|
|
|
250
249
|
checkForbiddenSeparator(separator) {
|
|
251
|
-
//
|
|
252
|
-
|
|
250
|
+
// comma, ampersand, space, underscore, CRLF, CR, LF
|
|
251
|
+
// 2023-04-15: dot is allowed to allow Helm like separator in Helm MSA results (no Helm monomers contains dot)
|
|
252
|
+
const forbiddenSepRe = /,|&| |_|\r\n|\r|\n/i;
|
|
253
253
|
return forbiddenSepRe.test(separator);
|
|
254
254
|
}
|
|
255
255
|
|