Increase String::num default decimal precision

Fixes #34541

Renamed MAX_DIGITS to MAX_DECIMALS, since it only changes the
amount of digits after the decimal point.

Increased MAX_DECIMALS to 32, and made String::num use
MAX_DECIMALS consistently. If -1 is passed as
decimal precision to String::num, it now gets changed to
the correct precision based on the number's magnitude,
instead of using printf default(which is 6)

String::num_real also calculates the correct precision now.

Also made the types used in floating-point math more
consistent in a few places.
This commit is contained in:
Heikki Simojoki
2019-12-23 17:38:33 +02:00
committed by Aaron Franke
parent b63f9b5961
commit 09a905ca80
7 changed files with 47 additions and 22 deletions

View File

@ -27,9 +27,12 @@
</argument>
<argument index="2" name="sort_keys" type="bool" default="false">
</argument>
<argument index="3" name="full_precision" type="bool" default="false">
</argument>
<description>
Converts a [Variant] var to JSON text and returns the result. Useful for serializing data to store or send over the network.
[b]Note:[/b] The JSON specification does not define integer or float types, but only a [i]number[/i] type. Therefore, converting a Variant to JSON text will convert all numerical values to [float] types.
[b]Note:[/b] If [code]full_precision[/code] is true, when printing floats, the unreliable digits are printed in addition to the reliable digits to guarantee exact decoding.
Use [code]indent[/code] parameter to pretty print the output.
[b]Example output:[/b]
[codeblock]