Misc

src/parseMetaString.ts

@@ -1,11 +1,41 @@
+// constructed with the assistance of claude AI
+function customSplit(str: string) {
+  const result = []
+  let current = ''
+  let inQuotes = false
+  let inBrackets = false
+
+  for (const char of str) {
+    if (char === '"') {
+      inQuotes = !inQuotes
+      current += char
+    } else if (char === '[') {
+      inBrackets = true
+      current += char
+    } else if (char === ']') {
+      inBrackets = false
+      current += char
+    } else if (char === ',' && !inQuotes && !inBrackets) {
+      result.push(current.trim())
+      current = ''
+    } else {
+      current += char
+    }
+  }
+
+  if (current) {
+    result.push(current.trim())
+  }
+
+  return result
+}
+
 export function parseMetaString(metaString: string) {
+  const inside = metaString.replace(/^<|>$/g, '')
   return Object.fromEntries(
-    [
-      // split string on comma except when in square bracket
-      // https://stackoverflow.com/questions/74238461/
-      ...metaString.replace(/^<|>$/g, '').matchAll(/(?:\[[^\][]*\]|[^,])+/g),
-    ].map(f => {
-      const [key, val] = f[0].split('=').map(f => f.trim())
+    customSplit(inside).map(f => {
+      const [key, val] = f.split('=').map(f => f.trim())
+      console.log({ key, val })
       if (val && val.startsWith('[') && val.endsWith(']')) {
         return [
           key,
@@ -17,7 +47,7 @@ export function parseMetaString(metaString: string) {
       } else if (val && val.startsWith('"') && val.endsWith('"')) {
         return [key, val.slice(1, -1)]
       } else {
-        return [key, val]
+        return [key, val?.replaceAll(/^"|"$/g, '')]
       }
     }),
   )

test/__snapshots__/parse.test.ts.snap

@@ -218,10 +218,9 @@ exports[`can get metadata from the header 1`] = `
       "Type": "Integer",
     },
     "DB": {
-      "Description": ""dbSNP membership",
+      "Description": "dbSNP membership, build 129",
       "Number": 0,
       "Type": "Flag",
-      "build 129"": undefined,
     },
     "DBRIPID": {
       "Description": "ID of this element in DBRIP",
@@ -460,10 +459,9 @@ exports[`can get metadata from the header 2`] = `
     "Type": "Integer",
   },
   "DB": {
-    "Description": ""dbSNP membership",
+    "Description": "dbSNP membership, build 129",
     "Number": 0,
     "Type": "Flag",
-    "build 129"": undefined,
   },
   "DBRIPID": {
     "Description": "ID of this element in DBRIP",
@@ -1346,16 +1344,14 @@ exports[`clinvar metadata 1`] = `
       "Type": "Integer",
     },
     "CLNDISDB": {
-      "Description": ""Tag-value pairs of disease database name and identifier submitted for germline classifications",
+      "Description": "Tag-value pairs of disease database name and identifier submitted for germline classifications, e.g. OMIM:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. OMIM:NNNNNN"": undefined,
     },
     "CLNDISDBINCL": {
-      "Description": ""For included Variant: Tag-value pairs of disease database name and identifier for germline classifications",
+      "Description": "For included Variant: Tag-value pairs of disease database name and identifier for germline classifications, e.g. OMIM:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. OMIM:NNNNNN"": undefined,
     },
     "CLNDN": {
       "Description": "ClinVar's preferred disease name for the concept specified by disease identifiers in CLNDISDB",
@@ -1368,11 +1364,9 @@ exports[`clinvar metadata 1`] = `
       "Type": "String",
     },
     "CLNHGVS": {
-      "Description": ""Top-level (primary assembly",
+      "Description": "Top-level (primary assembly, alt, or patch) HGVS expression.",
       "Number": ".",
       "Type": "String",
-      "alt": undefined,
-      "or patch) HGVS expression."": undefined,
     },
     "CLNREVSTAT": {
       "Description": "ClinVar review status of germline classification for the Variation ID",
@@ -1537,16 +1531,14 @@ exports[`clinvar metadata 1`] = `
       "Type": "String",
     },
     "ONCDISDB": {
-      "Description": ""Tag-value pairs of disease database name and identifier submitted for oncogenicity classifications",
+      "Description": "Tag-value pairs of disease database name and identifier submitted for oncogenicity classifications, e.g. MedGen:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. MedGen:NNNNNN"": undefined,
     },
     "ONCDISDBINCL": {
-      "Description": ""For included variant: Tag-value pairs of disease database name and identifier for oncogenicity classifications",
+      "Description": "For included variant: Tag-value pairs of disease database name and identifier for oncogenicity classifications, e.g. OMIM:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. OMIM:NNNNNN"": undefined,
     },
     "ONCDN": {
       "Description": "ClinVar's preferred disease name for the concept specified by disease identifiers in ONCDISDB",
@@ -1594,16 +1586,14 @@ exports[`clinvar metadata 1`] = `
       "Type": "String",
     },
     "SCIDISDB": {
-      "Description": ""Tag-value pairs of disease database name and identifier submitted for somatic clinial impact classifications",
+      "Description": "Tag-value pairs of disease database name and identifier submitted for somatic clinial impact classifications, e.g. MedGen:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. MedGen:NNNNNN"": undefined,
     },
     "SCIDISDBINCL": {
-      "Description": ""For included variant: Tag-value pairs of disease database name and identifier for somatic clinical impact classifications",
+      "Description": "For included variant: Tag-value pairs of disease database name and identifier for somatic clinical impact classifications, e.g. OMIM:NNNNNN",
       "Number": ".",
       "Type": "String",
-      "e.g. OMIM:NNNNNN"": undefined,
     },
     "SCIDN": {
       "Description": "ClinVar's preferred disease name for the concept specified by disease identifiers in SCIDISDB",

test/__snapshots__/parseMetaString.test.ts.snap

@@ -11,3 +11,12 @@ exports[`m1 1`] = `
   ],
 }
 `;
+
+exports[`m2 1`] = `
+{
+  "Description": "dbSNP membership, build 129",
+  "ID": "DB",
+  "Number": "0",
+  "Type": "Flag",
+}
+`;

test/parseMetaString.test.ts

@@ -1,7 +1,7 @@
 import { expect, test } from 'vitest'
 import { parseMetaString } from '../src/parseMetaString'
 
-test('m1', () => {
+test('array in values', () => {
   const result1 = parseMetaString(
     '<ID=Assay,Type=String,Number=.,Values=[WholeGenome, Exome]>',
   )
@@ -11,3 +11,11 @@ test('m1', () => {
   expect(result1).toEqual(result2)
   expect(result1).toMatchSnapshot()
 })
+
+test('quoted string with comma in description', () => {
+  expect(
+    parseMetaString(
+      '<ID=DB,Number=0,Type=Flag,Description="dbSNP membership, build 129">',
+    ),
+  ).toMatchSnapshot()
+})