Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...


Code Block
do(
    $value = $args[1],
    $value = replaceAll( $value , "\b","\\\\b"),
    $value = replaceAll( $value , "\f","\\\\f"),
    $value = replaceAll( $value , "\t","\\\\t"),
    $value = replaceAll( $value , "\r","\\\\r"),
    $value = replaceAll( $value , "\n","\\\\n"),
    $value = replaceAll( $value , "\\\\","\\\\\\\\"),
    $value = replaceAll( $value , '"','\\\\"')
)

Un-escape Macro. This is useful when importing JSON into PhixFlow from an external source. Once JSON has been converted to XML (jsonToXML), and XML converted to Items (xmlToItems) it converts any escaped sequence (eg \\\\t) back into a usable character. The double \\ ahead of the \t in the output is necessary to escape \t when the jep expression is passed and will be converted to simply \t (the tab character). 


Code Block
do(
    $value= $args[1],
   
    $value = replaceAll($value, "\\\\t","\\\t"),
    $value = replaceAll($value, "\\\\b","\\\b"),
    $value = replaceAll($value, "\\\\r","\\\r"),
    $value = replaceAll($value, "\\\\n","\\\n"),
    $value = replaceAll($value, "\\\\f","\\\f"),
    $value
)

Tip

Check the format of the JSON you are importing to see what adjustments are required. Sometimes characters such as \\\\\\\\" may be helpfully un-escaped prior to import, making it unnecessary to unescape every special character. 

Validating Data Imported Into PhixFlow

JSON data (like any imported data format) can include characters that cannot be written to a database that is using the UTF-8 character set. For example, 4-bit Emojis are not supported by UTF-8.

The following regular expression lists the characters expected in a JSON file and removes invalid characters from it.

Code Block
replaceAll(JSON Code,"[^\\p{Space}0-9A-Za-z!:\\\"%&\\[*()\\],-/_\\\\{}\\.]","")

The backslash character escapes special characters such as [] , . and \ 

Tip

Check which database PhixFlow is running on to see how it handles 4-bit characters. An oracle database with the recommended configuration can store the full range of Uni-code characters whilst MySQL/MariaDB UFT-8mb3 databases cannot and invalid characters will need to be removed before writing to the database. See Database for recommended database configurations. 

See also: Regular Expressions


...