fix: change type to boolean for noGremlins

feat: add new rule 'noGremlins'

README.md

@@ -29,6 +29,7 @@ Configuration is via a `.sasjslint` file with the following structure (these are
   "lowerCaseFileNames": true,
   "maxLineLength": 80,
   "noNestedMacros": true,
+  "noGremlins": true,
   "noSpacesInFileNames": true,
   "noTabs": true,
   "noTrailingSpaces": true,
@@ -125,6 +126,15 @@ We strongly recommend a line length limit, and set the bar at 80. To turn this f
 - Default: 80
 - Severity: WARNING
 
+### noGremlins
+
+Capture zero-width whitespace and other non-standard characters.  The logic is borrowed from the [VSCode Gremlins Extension](https://github.com/nhoizey/vscode-gremlins) - if you are looking for more advanced gremlin zapping capabilities, we highly recommend to use their extension instead.
+
+The list of characters can be found in this file: [https://github.com/sasjs/lint/blob/main/src/rules/line/noGremlins.ts](https://github.com/sasjs/lint/blob/main/src/rules/line/noGremlins.ts)
+
+- Default: true
+- Severity: WARNING
+
 ### noNestedMacros
 
 Where macros are defined inside other macros, they are recompiled every time the outer macro is invoked. Hence, it is widely considered inefficient, and bad practice, to nest macro definitions.

sasjslint-schema.json

@@ -13,6 +13,7 @@
     "indentationMultiple": 2,
     "lowerCaseFileNames": true,
     "maxLineLength": 80,
+    "noGremlins": true,
     "noNestedMacros": true,
     "noSpacesInFileNames": true,
     "noTabs": true,
@@ -29,6 +30,7 @@
       "noSpacesInFileNames": true,
       "lowerCaseFileNames": true,
       "maxLineLength": 80,
+      "noGremlins": true,
       "noTabs": true,
       "indentationMultiple": 4,
       "hasMacroNameInMend": true,
@@ -64,6 +66,14 @@
       "default": "/**{lineEnding}  @file{lineEnding}  @brief <Your brief here>{lineEnding}  <h4> SAS Macros </h4>{lineEnding}**/",
       "examples": []
     },
+    "noGremlins": {
+      "$id": "#/properties/noGremlins",
+      "type": "boolean",
+      "title": "noGremlins",
+      "description": "Captures problematic characters such as zero-width whitespace and others that look valid but usually are not (such as the en dash)",
+      "default": [true],
+      "examples": [true, false]
+    },
     "hasMacroNameInMend": {
       "$id": "#/properties/hasMacroNameInMend",
       "type": "boolean",
@@ -193,6 +203,13 @@
           "enum": ["error", "warn"],
           "default": "warn"
         },
+        "noGremlins": {
+          "$id": "#/properties/severityLevel/noGremlins",
+          "title": "noGremlins",
+          "type": "string",
+          "enum": ["error", "warn"],
+          "default": "warn"
+        },
         "hasMacroNameInMend": {
           "$id": "#/properties/severityLevel/hasMacroNameInMend",
           "title": "hasMacroNameInMend",

src/lint/lintFile.spec.ts

@@ -46,7 +46,7 @@ const expectedDiagnostics = [
     severity: Severity.Error
   },
   {
-    message: 'Line contains tab indentation',
+    message: 'Line contains a tab character (09x)',
     lineNumber: 7,
     startColumnNumber: 1,
     endColumnNumber: 2,

src/lint/lintFolder.spec.ts

@@ -53,7 +53,7 @@ const expectedDiagnostics = [
     severity: Severity.Error
   },
   {
-    message: 'Line contains tab indentation',
+    message: 'Line contains a tab character (09x)',
     lineNumber: 7,
     startColumnNumber: 1,
     endColumnNumber: 2,

src/lint/lintProject.spec.ts

@@ -51,7 +51,7 @@ const expectedDiagnostics = [
     severity: Severity.Error
   },
   {
-    message: 'Line contains tab indentation',
+    message: 'Line contains a tab character (09x)',
     lineNumber: 7,
     startColumnNumber: 1,
     endColumnNumber: 2,

src/rules/line/index.ts

@@ -1,3 +1,4 @@
+export { noGremlins } from './noGremlins'
 export { indentationMultiple } from './indentationMultiple'
 export { maxLineLength } from './maxLineLength'
 export { noEncodedPasswords } from './noEncodedPasswords'

src/rules/line/noGremlins.spec.ts

@@ -0,0 +1,15 @@
+import { Severity } from '../../types/Severity'
+import { noGremlins } from './noGremlins'
+
+describe('noTabs', () => {
+  it('should return an empty array when the line does not have any gremlin', () => {
+    const line = "%put 'hello';"
+    expect(noGremlins.test(line, 1)).toEqual([])
+  })
+
+  it('should return a diagnostic array when the line contains gremlins', () => {
+    const line = "–  ‘ %put 'hello';"
+    const diagnostics = noGremlins.test(line, 1)
+    expect(diagnostics.length).toEqual(2)
+  })
+})

src/rules/line/noGremlins.ts

@@ -0,0 +1,128 @@
+import { Diagnostic, LintConfig } from '../../types'
+import { LineLintRule } from '../../types/LintRule'
+import { LintRuleType } from '../../types/LintRuleType'
+import { Severity } from '../../types/Severity'
+
+const name = 'noGremlins'
+const description = 'Disallow characters specified in gremlins array'
+const message = 'Line contains a gremlin'
+
+const test = (value: string, lineNumber: number, config?: LintConfig) => {
+  const severity = config?.severityLevel[name] || Severity.Warning
+
+  const diagnostics: Diagnostic[] = []
+
+  const gremlins: any = {}
+
+  for (const [hexCode, config] of Object.entries(gremlinCharacters)) {
+    gremlins[charFromHex(hexCode)] = Object.assign({}, config, {
+      hexCode
+    })
+  }
+
+  const regexpWithAllChars = new RegExp(
+    Object.keys(gremlins)
+      .map((char) => `${char}+`)
+      .join('|'),
+    'g'
+  )
+
+  let match
+  while ((match = regexpWithAllChars.exec(value))) {
+    const matchedCharacter = match[0][0]
+    const gremlin = gremlins[matchedCharacter]
+
+    diagnostics.push({
+      message: `${message}: ${gremlin.description}, hexCode(${gremlin.hexCode})`,
+      lineNumber,
+      startColumnNumber: match.index + 1,
+      endColumnNumber: match.index + 1 + match[0].length,
+      severity
+    })
+  }
+
+  return diagnostics
+}
+
+/**
+ * Lint rule that checks if a given line of text contains any gremlins.
+ */
+export const noGremlins: LineLintRule = {
+  type: LintRuleType.Line,
+  name,
+  description,
+  message,
+  test
+}
+
+const charFromHex = (hexCode: string) => String.fromCodePoint(parseInt(hexCode))
+
+const gremlinCharacters = {
+  '0x2013': {
+    description: 'en dash'
+  },
+  '0x2018': {
+    description: 'left single quotation mark'
+  },
+  '0x2019': {
+    description: 'right single quotation mark'
+  },
+  '0x2029': {
+    zeroWidth: true,
+    description: 'paragraph separator'
+  },
+  '0x2066': {
+    zeroWidth: true,
+    description: 'Left to right'
+  },
+  '0x2069': {
+    zeroWidth: true,
+    description: 'Pop directional'
+  },
+  '0x0003': {
+    description: 'end of text'
+  },
+  '0x000b': {
+    description: 'line tabulation'
+  },
+  '0x00a0': {
+    description: 'non breaking space'
+  },
+  '0x00ad': {
+    description: 'soft hyphen'
+  },
+  '0x200b': {
+    zeroWidth: true,
+    description: 'zero width space'
+  },
+  '0x200c': {
+    zeroWidth: true,
+    description: 'zero width non-joiner'
+  },
+  '0x200e': {
+    zeroWidth: true,
+    description: 'left-to-right mark'
+  },
+  '0x201c': {
+    description: 'left double quotation mark'
+  },
+  '0x201d': {
+    description: 'right double quotation mark'
+  },
+  '0x202c': {
+    zeroWidth: true,
+    description: 'pop directional formatting'
+  },
+  '0x202d': {
+    zeroWidth: true,
+    description: 'left-to-right override'
+  },
+  '0x202e': {
+    zeroWidth: true,
+    description: 'right-to-left override'
+  },
+  '0xfffc': {
+    zeroWidth: true,
+    description: 'object replacement character'
+  }
+}

src/rules/line/noTabs.spec.ts

@@ -11,7 +11,7 @@ describe('noTabs', () => {
     const line = "\t%put 'hello';"
     expect(noTabs.test(line, 1)).toEqual([
       {
-        message: 'Line contains tab indentation',
+        message: 'Line contains a tab character (09x)',
         lineNumber: 1,
         startColumnNumber: 1,
         endColumnNumber: 2,

src/rules/line/noTabs.ts

@@ -7,7 +7,7 @@ import { getIndicesOf } from '../../utils'
 const name = 'noTabs'
 const alias = 'noTabIndentation'
 const description = 'Disallow indenting with tabs.'
-const message = 'Line contains tab indentation'
+const message = 'Line contains a tab character (09x)'
 
 const test = (value: string, lineNumber: number, config?: LintConfig) => {
   const severity =

src/types/LintConfig.ts

@@ -11,7 +11,8 @@ import {
   maxLineLength,
   noEncodedPasswords,
   noTabs,
-  noTrailingSpaces
+  noTrailingSpaces,
+  noGremlins
 } from '../rules/line'
 import { lowerCaseFileNames, noSpacesInFileNames } from '../rules/path'
 import { LineEndings } from './LineEndings'
@@ -119,6 +120,10 @@ export class LintConfig {
       this.fileLintRules.push(strictMacroDefinition)
     }
 
+    if (json?.noGremlins) {
+      this.lineLintRules.push(noGremlins)
+    }
+
     if (json?.severityLevel) {
       for (const [rule, severity] of Object.entries(json.severityLevel)) {
         if (severity === 'warn') this.severityLevel[rule] = Severity.Warning