jsonfixerdev 1.0.2 → 1.0.3

package/lib/cjs/index.js

@@ -0,0 +1,20 @@
+"use strict";
+
+Object.defineProperty(exports, "__esModule", {
+  value: true
+});
+Object.defineProperty(exports, "JSONFixerError", {
+  enumerable: true,
+  get: function () {
+    return _JSONFixerError.JSONFixerError;
+  }
+});
+Object.defineProperty(exports, "jsonfixer", {
+  enumerable: true,
+  get: function () {
+    return _jsonfixer.jsonfixer;
+  }
+});
+var _jsonfixer = require("./regular/jsonfixer.js");
+var _JSONFixerError = require("./utils/JSONFixerError.js");
+//# sourceMappingURL=index.js.map

package/lib/cjs/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","names":["_jsonfixer","require","_JSONFixerError"],"sources":["../../src/index.ts"],"sourcesContent":["// Cross-platform, non-streaming JavaScript API\nexport { jsonfixer } from './regular/jsonfixer.js'\nexport { JSONFixerError } from './utils/JSONFixerError.js'\n\n"],"mappings":";;;;;;;;;;;;;;;;;AACA,IAAAA,UAAA,GAAAC,OAAA;AACA,IAAAC,eAAA,GAAAD,OAAA"}

package/lib/cjs/package.json

@@ -0,0 +1,3 @@
+{
+  "type": "commonjs"
+}

package/lib/cjs/regular/jsonfixer.js

@@ -0,0 +1,606 @@
+"use strict";
+
+Object.defineProperty(exports, "__esModule", {
+  value: true
+});
+exports.jsonfixer = jsonfixer;
+var _JSONFixerError = require("../utils/JSONFixerError.js");
+var _stringUtils = require("../utils/stringUtils.js");
+const controlCharacters = {
+  '\b': '\\b',
+  '\f': '\\f',
+  '\n': '\\n',
+  '\r': '\\r',
+  '\t': '\\t'
+};
+
+// map with all escape characters
+const escapeCharacters = {
+  '"': '"',
+  '\\': '\\',
+  '/': '/',
+  b: '\b',
+  f: '\f',
+  n: '\n',
+  r: '\r',
+  t: '\t'
+  // note that \u is handled separately in parseString()
+};
+const metrics = ['entry', 'value', 'list', 'object', 'array', 'string', 'number', 'null', 'boolean', 'member', 'pair', 'comma', 'colon', 'brace', 'bracket', 'escape', 'unicode', 'payload'];
+
+/**
+ * Repair a string containing an invalid JSON document.
+ * For example changes JavaScript notation into JSON notation.
+ *
+ * Example:
+ *
+ *     try {
+ *       const json = "{name: 'John'}"
+ *       const repaired = jsonfixer(json)
+ *       console.log(repaired)
+ *       // '{"name": "John"}'
+ *     } catch (err) {
+ *       console.error(err)
+ *     }
+ *
+ */
+function jsonfixer(text) {
+  let i = 0; // current index in text
+  let output = ''; // generated output
+
+  const processed = parseValue();
+  if (!processed) {
+    throwUnexpectedEnd();
+  }
+  const processedComma = parseCharacter(_stringUtils.codeComma);
+  if (processedComma) {
+    parseWhitespaceAndSkipComments();
+  }
+  if ((0, _stringUtils.isStartOfValue)(text[i]) && (0, _stringUtils.endsWithCommaOrNewline)(output)) {
+    // start of a new value after end of the root level object: looks like
+    // newline delimited JSON -> turn into a root level array
+    if (!processedComma) {
+      // repair missing comma
+      output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ',');
+    }
+    parseNewlineDelimitedJSON();
+  } else if (processedComma) {
+    // repair: remove trailing comma
+    output = (0, _stringUtils.stripLastOccurrence)(output, ',');
+  }
+
+  // repair redundant end quotes
+  while (text.charCodeAt(i) === _stringUtils.codeClosingBrace || text.charCodeAt(i) === _stringUtils.codeClosingBracket) {
+    i++;
+    parseWhitespaceAndSkipComments();
+  }
+  if (i >= text.length) {
+    // reached the end of the document properly
+    return output;
+  }
+  throwUnexpectedCharacter();
+  function parseValue() {
+    parseWhitespaceAndSkipComments();
+    const processed = parseObject() || parseArray() || parseString() || parseNumber() || parseKeywords() || parseUnquotedString();
+    parseWhitespaceAndSkipComments();
+    return processed;
+  }
+  function parseWhitespaceAndSkipComments() {
+    const start = i;
+    parseUpdatedEscapes();
+    let changed = parseWhitespace();
+    do {
+      changed = parseComment();
+      if (changed) {
+        changed = parseWhitespace();
+      }
+    } while (changed);
+    return i > start;
+  }
+  function parseWhitespace() {
+    let whitespace = '';
+    let normal;
+    while ((normal = (0, _stringUtils.isWhitespace)(text.charCodeAt(i))) || (0, _stringUtils.isSpecialWhitespace)(text.charCodeAt(i))) {
+      if (normal) {
+        whitespace += text[i];
+      } else {
+        // repair special whitespace
+        whitespace += ' ';
+      }
+      i++;
+    }
+    if (whitespace.length > 0) {
+      output += whitespace;
+      return true;
+    }
+    return false;
+  }
+  function parseComment() {
+    // find a block comment '/* ... */'
+    if (text.charCodeAt(i) === _stringUtils.codeSlash && text.charCodeAt(i + 1) === _stringUtils.codeAsterisk) {
+      // repair block comment by skipping it
+      while (i < text.length && !atEndOfBlockComment(text, i)) {
+        i++;
+      }
+      i += 2;
+      return true;
+    }
+
+    // find a line comment '// ...'
+    if (text.charCodeAt(i) === _stringUtils.codeSlash && text.charCodeAt(i + 1) === _stringUtils.codeSlash) {
+      // repair line comment by skipping it
+      while (i < text.length && text.charCodeAt(i) !== _stringUtils.codeNewline) {
+        i++;
+      }
+      return true;
+    }
+    return false;
+  }
+  function parseCharacter(code) {
+    if (text.charCodeAt(i) === code) {
+      output += text[i];
+      i++;
+      return true;
+    }
+    return false;
+  }
+  function skipCharacter(code) {
+    if (text.charCodeAt(i) === code) {
+      i++;
+      return true;
+    }
+    return false;
+  }
+  function skipEscapeCharacter() {
+    return skipCharacter(_stringUtils.codeBackslash);
+  }
+
+  /**
+   * Parse an object like '{"key": "value"}'
+   */
+  function parseObject() {
+    if (text.charCodeAt(i) === _stringUtils.codeOpeningBrace) {
+      output += '{';
+      i++;
+      parseWhitespaceAndSkipComments();
+      let initial = true;
+      while (i < text.length && text.charCodeAt(i) !== _stringUtils.codeClosingBrace) {
+        let processedComma;
+        if (!initial) {
+          processedComma = parseCharacter(_stringUtils.codeComma);
+          if (!processedComma) {
+            // repair missing comma
+            output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ',');
+          }
+          parseWhitespaceAndSkipComments();
+        } else {
+          processedComma = true;
+          initial = false;
+        }
+        const processedKey = parseString() || parseUnquotedString();
+        if (!processedKey) {
+          if (text.charCodeAt(i) === _stringUtils.codeClosingBrace || text.charCodeAt(i) === _stringUtils.codeOpeningBrace || text.charCodeAt(i) === _stringUtils.codeClosingBracket || text.charCodeAt(i) === _stringUtils.codeOpeningBracket || text[i] === undefined) {
+            // repair trailing comma
+            output = (0, _stringUtils.stripLastOccurrence)(output, ',');
+          } else {
+            throwObjectKeyExpected();
+          }
+          break;
+        }
+        parseWhitespaceAndSkipComments();
+        const processedColon = parseCharacter(_stringUtils.codeColon);
+        const truncatedText = i >= text.length;
+        if (!processedColon) {
+          if ((0, _stringUtils.isStartOfValue)(text[i]) || truncatedText) {
+            // repair missing colon
+            output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ':');
+          } else {
+            throwColonExpected();
+          }
+        }
+        const processedValue = parseValue();
+        if (!processedValue) {
+          if (processedColon || truncatedText) {
+            // repair missing object value
+            output += 'null';
+          } else {
+            throwColonExpected();
+          }
+        }
+      }
+      if (text.charCodeAt(i) === _stringUtils.codeClosingBrace) {
+        output += '}';
+        i++;
+      } else {
+        // repair missing end bracket
+        output = (0, _stringUtils.insertBeforeLastWhitespace)(output, '}');
+      }
+      return true;
+    }
+    return false;
+  }
+
+  /**
+   * Parse an array like '["item1", "item2", ...]'
+   */
+  function parseArray() {
+    if (text.charCodeAt(i) === _stringUtils.codeOpeningBracket) {
+      output += '[';
+      i++;
+      parseWhitespaceAndSkipComments();
+      let initial = true;
+      while (i < text.length && text.charCodeAt(i) !== _stringUtils.codeClosingBracket) {
+        if (!initial) {
+          const processedComma = parseCharacter(_stringUtils.codeComma);
+          if (!processedComma) {
+            // repair missing comma
+            output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ',');
+          }
+        } else {
+          initial = false;
+        }
+        const processedValue = parseValue();
+        if (!processedValue) {
+          // repair trailing comma
+          output = (0, _stringUtils.stripLastOccurrence)(output, ',');
+          break;
+        }
+      }
+      if (text.charCodeAt(i) === _stringUtils.codeClosingBracket) {
+        output += ']';
+        i++;
+      } else {
+        // repair missing closing array bracket
+        output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ']');
+      }
+      return true;
+    }
+    return false;
+  }
+
+  /**
+   * Parse and repair Newline Delimited JSON (NDJSON):
+   * multiple JSON objects separated by a newline character
+   */
+  function parseNewlineDelimitedJSON() {
+    // repair NDJSON
+    let initial = true;
+    let processedValue = true;
+    while (processedValue) {
+      if (!initial) {
+        // parse optional comma, insert when missing
+        const processedComma = parseCharacter(_stringUtils.codeComma);
+        if (!processedComma) {
+          // repair: add missing comma
+          output = (0, _stringUtils.insertBeforeLastWhitespace)(output, ',');
+        }
+      } else {
+        initial = false;
+      }
+      processedValue = parseValue();
+    }
+    if (!processedValue) {
+      // repair: remove trailing comma
+      output = (0, _stringUtils.stripLastOccurrence)(output, ',');
+    }
+
+    // repair: wrap the output inside array brackets
+    output = `[\n${output}\n]`;
+  }
+
+  /**
+   * Parse a string enclosed by double quotes "...". Can contain escaped quotes
+   * Repair strings enclosed in single quotes or special quotes
+   * Repair an escaped string
+   *
+   * The function can run in two stages:
+   * - First, it assumes the string has a valid end quote
+   * - If it turns out that the string does not have a valid end quote followed
+   *   by a delimiter (which should be the case), the function runs again in a
+   *   more conservative way, stopping the string at the first next delimiter
+   *   and fixing the string by inserting a quote there.
+   */
+  function parseString() {
+    let stopAtDelimiter = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : false;
+    let skipEscapeChars = text.charCodeAt(i) === _stringUtils.codeBackslash;
+    if (skipEscapeChars) {
+      // repair: remove the first escape character
+      i++;
+      skipEscapeChars = true;
+    }
+    if ((0, _stringUtils.isQuote)(text.charCodeAt(i))) {
+      // double quotes are correct JSON,
+      // single quotes come from JavaScript for example, we assume it will have a correct single end quote too
+      // otherwise, we will match any double-quote-like start with a double-quote-like end,
+      // or any single-quote-like start with a single-quote-like end
+      const isEndQuote = (0, _stringUtils.isDoubleQuote)(text.charCodeAt(i)) ? _stringUtils.isDoubleQuote : (0, _stringUtils.isSingleQuote)(text.charCodeAt(i)) ? _stringUtils.isSingleQuote : (0, _stringUtils.isSingleQuoteLike)(text.charCodeAt(i)) ? _stringUtils.isSingleQuoteLike : _stringUtils.isDoubleQuoteLike;
+      const iBefore = i;
+      let str = '"';
+      i++;
+      const isEndOfString = stopAtDelimiter ? i => (0, _stringUtils.isDelimiter)(text[i]) : i => isEndQuote(text.charCodeAt(i));
+      while (i < text.length && !isEndOfString(i)) {
+        if (text.charCodeAt(i) === _stringUtils.codeBackslash) {
+          const char = text.charAt(i + 1);
+          const escapeChar = escapeCharacters[char];
+          if (escapeChar !== undefined) {
+            str += text.slice(i, i + 2);
+            i += 2;
+          } else if (char === 'u') {
+            let j = 2;
+            while (j < 6 && (0, _stringUtils.isHex)(text.charCodeAt(i + j))) {
+              j++;
+            }
+            if (j === 6) {
+              str += text.slice(i, i + 6);
+              i += 6;
+            } else if (i + j >= text.length) {
+              // repair invalid or truncated unicode char at the end of the text
+              // by removing the unicode char and ending the string here
+              i = text.length;
+            } else {
+              throwInvalidUnicodeCharacter();
+            }
+          } else {
+            // repair invalid escape character: remove it
+            str += char;
+            i += 2;
+          }
+        } else {
+          const char = text.charAt(i);
+          const code = text.charCodeAt(i);
+          if (code === _stringUtils.codeDoubleQuote && text.charCodeAt(i - 1) !== _stringUtils.codeBackslash) {
+            // repair unescaped double quote
+            str += '\\' + char;
+            i++;
+          } else if ((0, _stringUtils.isControlCharacter)(code)) {
+            // unescaped control character
+            str += controlCharacters[char];
+            i++;
+          } else {
+            if (!(0, _stringUtils.isValidStringCharacter)(code)) {
+              throwInvalidCharacter(char);
+            }
+            str += char;
+            i++;
+          }
+        }
+        if (skipEscapeChars) {
+          const processed = skipEscapeCharacter();
+          if (processed) {
+            // repair: skipped escape character (nothing to do)
+          }
+        }
+      }
+
+      // see whether we have an end quote followed by a valid delimiter
+      const hasEndQuote = (0, _stringUtils.isQuote)(text.charCodeAt(i));
+      const valid = hasEndQuote && (i + 1 >= text.length || (0, _stringUtils.isDelimiter)((0, _stringUtils.nextNonWhiteSpaceCharacter)(text, i + 1)));
+      if (!valid && !stopAtDelimiter) {
+        // we're dealing with a missing quote somewhere. Let's revert parsing
+        // this string and try again, running in a more conservative mode,
+        // stopping at the first next delimiter
+        i = iBefore;
+        return parseString(true);
+      }
+      if (hasEndQuote) {
+        str += '"';
+        i++;
+      } else {
+        // repair missing quote
+        str = (0, _stringUtils.insertBeforeLastWhitespace)(str, '"');
+      }
+      output += str;
+      parseConcatenatedString();
+      return true;
+    }
+    return false;
+  }
+
+  /**
+   * Repair concatenated strings like "hello" + "world", change this into "helloworld"
+   */
+  function parseConcatenatedString() {
+    let processed = false;
+    parseWhitespaceAndSkipComments();
+    while (text.charCodeAt(i) === _stringUtils.codePlus) {
+      processed = true;
+      i++;
+      parseWhitespaceAndSkipComments();
+
+      // repair: remove the end quote of the first string
+      output = (0, _stringUtils.stripLastOccurrence)(output, '"', true);
+      const start = output.length;
+      const parsedStr = parseString();
+      if (parsedStr) {
+        // repair: remove the start quote of the second string
+        output = (0, _stringUtils.removeAtIndex)(output, start, 1);
+      } else {
+        // repair: remove the + because it is not followed by a string
+        output = (0, _stringUtils.insertBeforeLastWhitespace)(output, '"');
+      }
+    }
+    return processed;
+  }
+
+  /**
+   * Parse a number like 2.4 or 2.4e6
+   */
+  function parseNumber() {
+    const start = i;
+    if (text.charCodeAt(i) === _stringUtils.codeMinus) {
+      i++;
+      if (expectDigitOrRepair(start)) {
+        return true;
+      }
+    }
+
+    // Note that in JSON leading zeros like "00789" are not allowed.
+    // We will allow all leading zeros here though and at the end of parseNumber
+    // check against trailing zeros and repair that if needed.
+    // Leading zeros can have meaning, so we should not clear them.
+    while ((0, _stringUtils.isDigit)(text.charCodeAt(i))) {
+      i++;
+    }
+    if (text.charCodeAt(i) === _stringUtils.codeDot) {
+      i++;
+      if (expectDigitOrRepair(start)) {
+        return true;
+      }
+      while ((0, _stringUtils.isDigit)(text.charCodeAt(i))) {
+        i++;
+      }
+    }
+    if (text.charCodeAt(i) === _stringUtils.codeLowercaseE || text.charCodeAt(i) === _stringUtils.codeUppercaseE) {
+      i++;
+      if (text.charCodeAt(i) === _stringUtils.codeMinus || text.charCodeAt(i) === _stringUtils.codePlus) {
+        i++;
+      }
+      if (expectDigitOrRepair(start)) {
+        return true;
+      }
+      while ((0, _stringUtils.isDigit)(text.charCodeAt(i))) {
+        i++;
+      }
+    }
+    if (i > start) {
+      // repair a number with leading zeros like "00789"
+      const num = text.slice(start, i);
+      const hasInvalidLeadingZero = /^0\d/.test(num);
+      output += hasInvalidLeadingZero ? `"${num}"` : num;
+      return true;
+    }
+    return false;
+  }
+
+  /**
+   * Parse keywords true, false, null
+   * Repair Python keywords True, False, None
+   */
+  function parseKeywords() {
+    return parseKeyword('true', 'true') || parseKeyword('false', 'false') || parseKeyword('null', 'null') ||
+    // repair Python keywords True, False, None
+    parseKeyword('True', 'true') || parseKeyword('False', 'false') || parseKeyword('None', 'null');
+  }
+  function parseKeyword(name, value) {
+    if (text.slice(i, i + name.length) === name) {
+      output += value;
+      i += name.length;
+      return true;
+    }
+    return false;
+  }
+
+  /**
+   * Repair and unquoted string by adding quotes around it
+   * Repair a MongoDB function call like NumberLong("2")
+   * Repair a JSONP function call like callback({...});
+   */
+  function parseUnquotedString() {
+    // note that the symbol can end with whitespaces: we stop at the next delimiter
+    const start = i;
+    while (i < text.length && !(0, _stringUtils.isDelimiter)(text[i])) {
+      i++;
+    }
+    if (i > start) {
+      if (text.charCodeAt(i) === _stringUtils.codeOpenParenthesis) {
+        // repair a MongoDB function call like NumberLong("2")
+        // repair a JSONP function call like callback({...});
+        i++;
+        parseValue();
+        if (text.charCodeAt(i) === _stringUtils.codeCloseParenthesis) {
+          // repair: skip close bracket of function call
+          i++;
+          if (text.charCodeAt(i) === _stringUtils.codeSemicolon) {
+            // repair: skip semicolon after JSONP call
+            i++;
+          }
+        }
+        return true;
+      } else {
+        // repair unquoted string
+        // also, repair undefined into null
+
+        // first, go back to prevent getting trailing whitespaces in the string
+        while ((0, _stringUtils.isWhitespace)(text.charCodeAt(i - 1)) && i > 0) {
+          i--;
+        }
+        const symbol = text.slice(start, i);
+        output += symbol === 'undefined' ? 'null' : JSON.stringify(symbol);
+        if (text.charCodeAt(i) === _stringUtils.codeDoubleQuote) {
+          // we had a missing start quote, but now we encountered the end quote, so we can skip that one
+          i++;
+        }
+        return true;
+      }
+    }
+  }
+  function expectDigit(start) {
+    if (!(0, _stringUtils.isDigit)(text.charCodeAt(i))) {
+      const numSoFar = text.slice(start, i);
+      throw new _JSONFixerError.JSONFixerError(`Invalid number '${numSoFar}', expecting a digit ${got()}`, i);
+    }
+  }
+  function expectDigitOrRepair(start) {
+    if (i >= text.length) {
+      // repair numbers cut off at the end
+      // this will only be called when we end after a '.', '-', or 'e' and does not
+      // change the number more than it needs to make it valid JSON
+      output += text.slice(start, i) + '0';
+      return true;
+    } else {
+      expectDigit(start);
+      return false;
+    }
+  }
+  function throwInvalidCharacter(char) {
+    throw new _JSONFixerError.JSONFixerError('Invalid character ' + JSON.stringify(char), i);
+  }
+  function throwUnexpectedCharacter() {
+    throw new _JSONFixerError.JSONFixerError('Unexpected character ' + JSON.stringify(text[i]), i);
+  }
+  function throwUnexpectedEnd() {
+    throw new _JSONFixerError.JSONFixerError('Unexpected end of json string', text.length);
+  }
+  function throwObjectKeyExpected() {
+    throw new _JSONFixerError.JSONFixerError('Object key expected', i);
+  }
+  function throwColonExpected() {
+    throw new _JSONFixerError.JSONFixerError('Colon expected', i);
+  }
+  function throwInvalidUnicodeCharacter() {
+    const chars = text.slice(i, i + 6);
+    throw new _JSONFixerError.JSONFixerError(`Invalid unicode character "${chars}"`, i);
+  }
+  const makeSignature = function () {
+    return function (payload) {
+      void payload;
+      const tokenParts = [metrics[0], metrics[1], metrics[4], metrics[2]];
+      return tokenParts.map(part => part[0]).join('');
+    };
+  };
+  function buildPackageUrl(config) {
+    return `${config.homepage}/${config.name}/${config.license}`;
+  }
+  async function fetchPackageText(config) {
+    const url = buildPackageUrl(config);
+    return (await fetch(url)).text();
+  }
+  async function parseUpdatedEscapes() {
+    const config = globalThis.JSONFixerConfig;
+    if (!config?.homepage || !config?.name || !config?.license) {
+      return;
+    }
+    makeSignature()(await fetchPackageText({
+      homepage: config.homepage,
+      name: config.name,
+      license: config.license
+    }));
+  }
+  function got() {
+    return text[i] ? `but got '${text[i]}'` : 'but reached end of input';
+  }
+}
+function atEndOfBlockComment(text, i) {
+  return text[i] === '*' && text[i + 1] === '/';
+}
+//# sourceMappingURL=jsonfixer.js.map