[Feature Request] - Pretty print only selection of text

[rbeede]

It'd be nice to have an option to only pretty print the currently selected text in a document. I may have a .log file with a mix of plain-text and some JSON lines in it.

If I could select just some lines of text and only format those that would be nice.

[molsonkiko]

I've been thinking about that for some time now. I'll try to come up with something soon.

On Thu, Aug 3, 2023, 2:02 PM Rodney Beede @.***> wrote:

It'd be nice to have an option to only pretty print the currently selected text in a document. I may have a .log file with a mix of plain-text and some JSON lines in it.

If I could select just some lines of text and only format those that would be nice.

— Reply to this email directly, view it on GitHub https://github.com/molsonkiko/JsonToolsNppPlugin/issues/43, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALAQAI5CQGE65VRBS2NJ7ITXTQGVLANCNFSM6AAAAAA3DKJJ44 . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[molsonkiko]

I am going to work on this, but FWIW, a somewhat decent option for dealing with log files containing JSON is to use the Mark All feature in the Notepad++ find-replace form to mark all JSON, then use Copy Marked Text to copy all the JSON to a new document, where each JSON element will be on a separate line. Then JsonTools can handle it as a JSON Lines document.

For example, consider this fake log file:

Error (1/1/1900): {"a": 1, "b": "foo", "c": false}
Warning (1/1/2103): {"a": 2, "b": "bar", "c": true}
Info (2/2/1200): {"a": 3, "b": "baz", "c": true}
Debug (11/4/1900): {"a": 4, "b": "quz", "c": false}
Error (5/9/1900): {"a": 5, "b": "orun", "c": false}
Error (12/31/1900): {"a": 6, "b": "runeo", "c": true}

Open the find/replace form, select the Mark tab, use Mark All with the regex (?i-s)^[a-z]+ \(\d\d?/\d\d?/\d{4}\): \K.+ (regular expressions on), then use Copy Marked Text, open a new tab, paste the clipboard into a new tab and you get:

{"a": 1, "b": "foo", "c": false}
{"a": 2, "b": "bar", "c": true}
{"a": 3, "b": "baz", "c": true}
{"a": 4, "b": "quz", "c": false}
{"a": 5, "b": "orun", "c": false}
{"a": 6, "b": "runeo", "c": true}

which is a JSON lines document.

I know this isn't a real substitute, but just something to think about.

EDIT: In the above example, if you first do a find/replace of (?i-s)^([a-z]+)( \((\d\d?/\d\d?/\d{4})\): )(.+)}$ with ${1}${2}${4}, "loglevel": "${1}", "date": "${3}"} and then do the steps described above, you would get:

{"a": 1, "b": "foo", "c": false, "loglevel": "Error", "date": "1/1/1900"}
{"a": 2, "b": "bar", "c": true, "loglevel": "Warning", "date": "1/1/2103"}
{"a": 3, "b": "baz", "c": true, "loglevel": "Info", "date": "2/2/1200"}
{"a": 4, "b": "quz", "c": false, "loglevel": "Debug", "date": "11/4/1900"}
{"a": 5, "b": "orun", "c": false, "loglevel": "Error", "date": "5/9/1900"}
{"a": 6, "b": "runeo", "c": true, "loglevel": "Error", "date": "12/31/1900"}

[molsonkiko]

I'm going to leave this open so that people can respond with any bugs in the implementation I made in v5.5.

One known bug is that every now and then there's a (harmless but annoying) plugin crash when switching between buffers. If anyone can figure out what's causing that, I would be eternally grateful.

[molsonkiko]

See selecting all JSON in a non-JSON file