The buffer that files are read into is not zero-terminated, but deserializeJson needs a zero-terminated string. It seems that deserializeJson normally stops when it's read a full JSON document hence it usually works fine, but it's possible (especially if a file is corrupt for some reason) that it could keep processing past the end of the file. deserialiseJson accepts an optional buffer size which saves having to extend the buffer for the extra \0 char.
There are some other cases of deserializeJson, but having looked at them it seems to me that they are operating on zero-terminated buffers, so are good as-is.
matjack1
commented
3 months ago
The buffer that files are read into is not zero-terminated, but deserializeJson needs a zero-terminated string. It seems that deserializeJson normally stops when it's read a full JSON document hence it usually works fine, but it's possible (especially if a file is corrupt for some reason) that it could keep processing past the end of the file. deserialiseJson accepts an optional buffer size which saves having to extend the buffer for the extra \0 char.
There are some other cases of deserializeJson, but having looked at them it seems to me that they are operating on zero-terminated buffers, so are good as-is.