This is the default ndJSON sanitizer function for log data being read into
the R session by read_logs(). This type of function is needed because since
loggit reimplements its own string-based JSON parser, and not a fancy one
built from an AST or something, it's very easy to have bad patterns break
your logs. You may also specify your own sanitizer function to pass to
loggit(), which takes a single string and returns an
(optionally-transformed) string, where each string is an individual element
of the log data.
default_ndjson_sanitizer(string, sanitize = TRUE)default_ndjson_unsanitizer(string)
Each element of the log data to operate on. Note that this is
each element, not each line in the logs. For example, each entry in the
log_msg field across all logs will be sanitized/unsanitized individually.
This is important because if writing your own sanitizer function, it must
take and return a single string as its argument.
Whether the operation will sanitize, or unsanitize the log
data. Defaults to TRUE, for sanitization on write.
A single string.
The default string patterns and their replacements are currently mapped as follows:
| Character | Replacement in log file |
{ |
__LEFTBRACE__ |
} |
__RIGHTBRACE__ |
" |
__DBLQUOTE__ |
, |
__COMMA__ |
\r |
__CR__ |
\n |
__LF__ |