add exhaustive tests for ANSI-C quoted strings

test/string-utils.cjs

@@ -1,6 +1,8 @@
-/* global describe, it */
+/* global describe, it, xit */
 
 const { strictEqual } = require('assert')
+const util = require('util')
+const exec = util.promisify(require('child_process').exec)
 const {
   camelCase,
   decamelize,
@@ -44,5 +46,101 @@ describe('string-utils', function () {
       const testCase = "$'\\c\\ '"
       await check(testCase)
     })
+
+    async function check (ansiString, encoding = 'utf8') {
+      const result = parseAnsiCQuotedString(ansiString)
+
+      const resultBuffer = Buffer.from(result, encoding)
+      const firstNull = resultBuffer.indexOf(0)
+      const resultUpToFirstNull = firstNull === -1 ? resultBuffer : resultBuffer.slice(0, firstNull)
+
+      const { stdout } = await exec(
+        'printf "%s" ' + ansiString,
+        { shell: '/bin/bash', encoding: 'buffer' }
+      )
+
+      if (!resultUpToFirstNull.equals(stdout)) {
+        console.log(JSON.stringify(ansiString), ansiString[4], ansiString.codePointAt(4))
+        console.log(resultUpToFirstNull, 'should be', stdout)
+        if (!resultUpToFirstNull.equals(resultBuffer)) {
+          console.log('(', resultUpToFirstNull, 'is actually', resultBuffer, ')')
+        }
+        console.log()
+      }
+
+      resultUpToFirstNull.should.deep.equal(stdout)
+    }
+
+    async function checkAll (prefix, value, maxLength, encoding = 'utf8') {
+      await check("$'\\" + prefix + value + "'", encoding)
+      while (value.length < maxLength) {
+        value = '0' + value
+        await check("$'\\" + prefix + value + "'", encoding)
+        if (value.toLowerCase() !== value) {
+          await check("$'\\" + prefix + value.toLowerCase() + "'", encoding)
+        }
+      }
+    }
+
+    // Skip exhaustive testing because it takes a long time
+    xit('parses all octal codes in ANSI-C quoted strings like bash', async () => {
+      for (let i = 0; i <= 0o777; i++) {
+        if (i % 256 === 0) { // null character
+          continue
+        }
+        await checkAll('', i.toString(8), 3, 'ascii')
+      }
+    }).timeout(5000)
+
+    xit('parses all hex codes in ANSI-C quoted strings like bash', async () => {
+      for (let i = 1; i <= 0xFF; i++) { // start at 1 to skip null
+        await checkAll('x', i.toString(16), 2, 'ascii')
+      }
+
+      for (let i = 1; i <= 0xFFFF; i++) {
+        // The high and low surrogates "\ud800" to "\udfff" fail this test
+        // because when the resulting string is encoded to a Buffer with UTF-8,
+        // it becomes a "?" replacement character, but the underlying data is correct.
+        // https://en.wikipedia.org/wiki/UTF-16#U+D800_to_U+DFFF
+        if (i >= 0xD800 && i <= 0xDFFF) {
+          continue
+        }
+        await checkAll('u', i.toString(16), 4)
+      }
+
+      for (let i = 1; i <= 0x10FFFF; i++) {
+        if (i >= 0xD800 && i <= 0xDFFF) {
+          continue
+        }
+        await checkAll('U', i.toString(16), 8)
+      }
+
+      // TODO: echo -n $'\U110000' produces f4 90 80 80 but String.fromCodePoint(0x110000)
+      // raises an error in node: Uncaught RangeError: Invalid code point 1114112
+    }).timeout(5 * 60 * 60 * 1000)
+
+    xit('parses control codes in ANSI-C quoted strings like bash', async () => {
+      // We skip \x01 and \x7F because those don't make sense as control codes
+      // and because bash parses them differently.
+      for (let i = 2; i < 0x7F; i++) {
+        const chr = String.fromCodePoint(i)
+        if (chr === "'" || chr === '\\') {
+          continue
+        }
+        await check("$'\\c" + chr + "'")
+      }
+    }).timeout(60 * 1000)
+
+    // We skip this because we do not handle non-ASCII characters after a control code,
+    // unlike bash.
+    xit('parses all non-ASCII codes in ANSI-C quoted strings like bash', async () => {
+      for (let i = 0x80; i <= 0x10FFFF; i++) {
+        const chr = String.fromCodePoint(i)
+        if (chr === "'" || chr === '\\' || (i >= 0xD800 && i <= 0xDFFF)) {
+          continue
+        }
+        await check("$'\\c" + chr + "'")
+      }
+    }).timeout(5 * 60 * 60 * 1000)
   })
 })