Validate part(s) of a file that should be valid JSON text.
For JSON text that is allowed to have comments (//) there is a different validate command.
The JSON Lines format decribes a single JSON object ({}) on each line.
When you have in your source code a string that is a JSON object after parsing. With this extension you can validate if the text is able to parse to a JSON object.
- Select part of the file (can be multiple parts with Multi Cursor).
- Select from the Command Palette or the editor context menu:
- Validate JSON
- Validate JSON with Comments
- Validate JSON Lines
- Each selection is parsed and if an error is found you get a warning message with the possibility to go to the line with the error.
Only the first error in each selection is shown. Repeat the check until you don't get any error.
If nothing is selected the whole file is validated.
The errors shown in the PROBLEMS panel will not be removed when you fix the problem, the validator does work at the background (it does not know the syntax of your file). You have to validate the region again.
Configuration
jsonvalidate.placeCursorAfterPreviousChar : When the next character expected is a , or : use the location after the previous non whitespace character. Comments are considered whitespace. (default: true)jsonvalidate.errorsInProblemsPanel : Report an error in the PROBLEMS panel (red squiggles). (default: true)jsonvalidate.errorsByMessages : Report an error with a Warning message. (default: true)jsonvalidate.blocks : Define blocks of the file to validate automatically.jsonvalidate.showValidateJsonInContextMenu : Show "Validate JSON" in editor context menu. (default: true)jsonvalidate.showValidateJsonWithCommentsInContextMenu : Show "Validate JSON with Comments" in editor context menu. (default: true)jsonvalidate.showValidateJsonLinesInContextMenu : Show "Validate JSON Lines" in editor context menu. (default: true)
Blocks
If the part of the file that contains the JSON text has a known begin and end you can have that block validated automatically. The PROBLEMS panel will be updated. If you have the messages enabled you might get a lot of them.
The definition of a block is done in the configuration variable jsonvalidate.blocks. It is an object with:
- the key for the block can have any name, it can be used to override a definition.
- the properties for each block are
flags: a string with the regex flags "i" and/or "m" (default: "")beforeStart: regular expression to search for the block startafterEnd: regular expression to search for after the block startfilterLanguageID: (Optional) search for blocks in this file if LanguageID matches this regular expression (case insensitive) (default: undefined)filterFilePath: (Optional) search for blocks in this file if the file path matches this regular expression (case insensitive) (default: undefined)format: (Optional) the format of the block (default: "json")
Possible values are:
"json" : standard JSON"jsonc" : JSON with Comments"jsonl" : JSON Lines
allowComments: (Optional) are comments allowed in this block (default: false)
If format property is defined this is ignored.
If a filter property is not defined it will pass all files.
If a block is found, a different block is searched for in the rest of the file.
Example
If you end the name of a variable with JSON you can define a block to validate.
"jsonvalidate.blocks": {
"Python String": {
"beforeStart": "\\w+JSON\\s*=\\s*(\"\"\"|''')",
"afterEnd": "(\"\"\"|''')",
"filterFilePath": "\\.py$"
},
"JavaScript String": {
"beforeStart": "\\w+JSON\\s*=\\s*`",
"afterEnd": "`",
"filterLanguageID": "javascript"
}
}
In JavaScript we use the backtick (template strings).
Build in definition
The extension has a build in block definition (that can be overridden).
The Liquid template definition can contain schema information between specific tags.
"Liquid schema": {
"beforeStart": "\\{%\\s*schema\\s*%\\}",
"afterEnd": "\\{%\\s*endschema\\s*%\\}",
"filterLanguageID": "liquid"
}
You change this definition by using a name with the same key: "Liquid schema"
