Extract Utf8Validator class

.github/workflows/lint.yml

@@ -36,7 +36,7 @@ jobs:
       - name: "Lint PHP files"
         run: |
           hasErrors=0
-          for f in $(find src/ tests/ -type f -name '*.php' ! -path '*/vendor/*' ! -path '*/Fixtures/*')
+          for f in bin/jsonlint $(find src/ tests/ -type f -name '*.php' ! -path '*/vendor/*' ! -path '*/Fixtures/*')
           do
             { error="$(php -derror_reporting=-1 -ddisplay_errors=1 -l -f $f 2>&1 1>&3 3>&-)"; } 3>&1;
             if [ "$error" != "" ]; then

README.md

@@ -36,6 +36,7 @@ You can also pass additional flags to `JsonParser::lint/parse` that tweak the fu
 - `JsonParser::PARSE_TO_ASSOC` parses to associative arrays instead of stdClass objects.
 - `JsonParser::ALLOW_COMMENTS` parses while allowing (and ignoring) inline `//` and multiline `/* */` comments in the JSON document.
 - `JsonParser::ALLOW_DUPLICATE_KEYS_TO_ARRAY` collects duplicate keys. e.g. if you have two `foo` keys the `foo` key will become an object (or array in assoc mode) with all `foo` values accessible as an array in `$result->foo->__duplicates__` (or `$result['foo']['__duplicates__']` in assoc mode).
+- `JsonParser::VALIDATE_UTF8_ENCODING` validates that the whole input is well-formed UTF-8 (as defined by [RFC 3629](https://www.rfc-editor.org/rfc/rfc3629)) before parsing, throwing an `InvalidEncodingException` that points at the first invalid byte otherwise.
 
 Example:
 
@@ -49,6 +50,28 @@ try {
 }
 ```
 
+Validating UTF-8 encoding
+-------------------------
+
+The UTF-8 validation behind the `VALIDATE_UTF8_ENCODING` flag is also available as a
+standalone, reusable utility through the `Seld\JsonLint\Utf8Validator` class, should you
+need to validate UTF-8 encoding outside of JSON parsing:
+
+```php
+use Seld\JsonLint\Utf8Validator;
+use Seld\JsonLint\InvalidEncodingException;
+
+try {
+    // returns void on success, throws on the first invalid byte
+    Utf8Validator::validate($string);
+} catch (InvalidEncodingException $e) {
+    // getMessage() describes the offending byte and its position
+    echo $e->getMessage();
+    // getDetails() returns that position information as an array
+    $details = $e->getDetails();
+}
+```
+
 > **Note:** This library is meant to parse JSON while providing good error messages on failure. There is no way it can be as fast as php native `json_decode()`.
 >
 > It is recommended to parse with `json_decode`, and when it fails parse again with seld/jsonlint to get a proper error message back to the user. See for example [how Composer uses this library](https://github.com/composer/composer/blob/56edd53046fd697d32b2fd2fbaf45af5d7951671/src/Composer/Json/JsonFile.php#L283-L318):

bin/jsonlint

@@ -54,11 +54,7 @@ if (isset($_SERVER['argc']) && $_SERVER['argc'] > 1) {
     }
 }
 
-$flags = getFlags(
-  $allowComments,
-  $detectKeyConflicts,
-  $validateUTF8Encoding,
-);
+$flags = getFlags($allowComments, $detectKeyConflicts, $validateUTF8Encoding);
 
 if (!empty($files)) {
     // file linting
@@ -80,15 +76,19 @@ if (!empty($files)) {
     }
 }
 
-function getFlags(
-    $allowComments = false,
-    $detectKeyConflicts = false,
-    $validateUTF8Encoding = false,
-){
+function getFlags($allowComments = false, $detectKeyConflicts = false, $validateUTF8Encoding = false)
+{
     $flags = 0;
-    if($allowComments){ $flags |= JsonParser::ALLOW_COMMENTS; }
-    if($detectKeyConflicts){ $flags |= JsonParser::DETECT_KEY_CONFLICTS; }
-    if($validateUTF8Encoding){ $flags |= JsonParser::VALIDATE_UTF8_ENCODING; }
+    if ($allowComments) {
+        $flags |= JsonParser::ALLOW_COMMENTS;
+    }
+    if ($detectKeyConflicts) {
+        $flags |= JsonParser::DETECT_KEY_CONFLICTS;
+    }
+    if ($validateUTF8Encoding) {
+        $flags |= JsonParser::VALIDATE_UTF8_ENCODING;
+    }
+
     return $flags;
 }
 
@@ -129,6 +129,7 @@ function lintFile($file, $quiet = false, $flags = 0)
     if (!$quiet) {
         echo 'Valid JSON (' . $file . ')' . PHP_EOL;
     }
+
     return true;
 }
 

phpstan.neon.dist

@@ -12,4 +12,4 @@ parameters:
         - tests/
 
     ignoreErrors:
-        - '/Method JsonParserTest::[a-zA-Z0-9]+\(\) has no return type/'
+        - '/Method [a-zA-Z0-9]+Test::[a-zA-Z0-9]+\(\) has no return type/'

src/Seld/JsonLint/InvalidEncodingException.php

@@ -17,14 +17,14 @@
 class InvalidEncodingException extends ParsingException
 {
     /**
-     * @var array{key: string, line: int}
+     * @var array{key: string, current_octet: int|null, continuation_octet_needed: int, offset_in_octets_from_string_start: int, offset_in_characters_from_string_start: int, character_start_position_from_string_start: int, line: int, offset_in_octets_from_line_start: int, offset_in_characters_from_line_start: int, character_start_position_from_line_start: int, current_continuation_octet_minimum: int, current_continuation_octet_maximum: int}
      */
     protected $details;
 
     /**
      * @param string $message
      * @param string $key
-     * @phpstan-param array{line: int} $details
+     * @phpstan-param array{current_octet: int|null, continuation_octet_needed: int, offset_in_octets_from_string_start: int, offset_in_characters_from_string_start: int, character_start_position_from_string_start: int, line: int, offset_in_octets_from_line_start: int, offset_in_characters_from_line_start: int, character_start_position_from_line_start: int, current_continuation_octet_minimum: int, current_continuation_octet_maximum: int} $details
      */
     public function __construct($message, $key, array $details)
     {
@@ -41,7 +41,7 @@ public function getKey()
     }
 
     /**
-     * @phpstan-return array{key: string, line: int}
+     * @phpstan-return array{key: string, current_octet: int|null, continuation_octet_needed: int, offset_in_octets_from_string_start: int, offset_in_characters_from_string_start: int, character_start_position_from_string_start: int, line: int, offset_in_octets_from_line_start: int, offset_in_characters_from_line_start: int, character_start_position_from_line_start: int, current_continuation_octet_minimum: int, current_continuation_octet_maximum: int}
      */
     public function getDetails()
     {