Title: | Debug R Packages |
---|---|
Description: | Specify debug messages as special string constants, and control debugging of packages via environment variables. |
Authors: | Gábor Csárdi [aut, cre], Posit Software, PBC [cph, fnd] |
Maintainer: | Gábor Csárdi <[email protected]> |
License: | MIT + file LICENSE |
Version: | 1.2.0.9000 |
Built: | 2024-11-26 05:07:07 UTC |
Source: | https://github.com/r-lib/debugme |
Normally this function is not called directly, but debug strings
are used. See debugme()
.
debug(msg, pkg = environmentName(topenv(parent.frame())), level = 2)
debug(msg, pkg = environmentName(topenv(parent.frame())), level = 2)
msg |
Message to print, character constant. |
pkg |
Package name to which the message belongs. Detected automatically. |
level |
The maximum debug level to show this message at. |
Invisibly, the message if it is shown, otherwise NULL
.
Specify debug messages as special string constants, and control debugging of packages via environment variables.
debugme(env = topenv(parent.frame()), pkg = environmentName(env))
debugme(env = topenv(parent.frame()), pkg = environmentName(env))
env |
Environment to instument debugging in. Defaults to the package environment of the calling package. |
pkg |
Name of the calling package. The default should be fine for almost all cases. |
To add debugging to your package, you need to
Import the debugme
package.
Define an .onLoad
function in your package, that calls debugme
.
An example:
.onLoad <- function(libname, pkgname) { debugme::debugme() }
By default debugging is off. To turn on debugging, set the DEBUGME
environment variable to the names of the packages you want to debug.
Package names can be separated by commas.
This environment variable is read when the package is loaded,
and with every call to debugme()
.
Example debugme
entries:
"!DEBUG Start Shiny app"
It is often desired that the debug messages contain values of R expressions evaluated at runtime. For example, when starting a Shiny app, it is useful to also print out the path to the app. Similarly, when debugging an HTTP response, it is desired to log the HTTP status code.
debugme
allows embedding R code into the debug messages, within
backticks. The code will be evaluated at runtime. Here are some
examples:
"!DEBUG Start Shiny app at `path`" "!DEBUG Got HTTP response `httr::status_code(reponse)`"
Note that parsing the debug strings for code is not very sophisticated currently, and you cannot embed backticks into the code itself.
debugme
has two ways to organize log messages into log levels:
a quick informal way, and a more formal one.
For the informal way, you can start the !DEBUG
token with multiple
!
characters. You can then select the desired level of logging via
!
characters before the package name in the DEBUGME
environment
variable. E.g. DEBUGME=!!mypackage
means that only debug messages
with two or less !
marks will be printed.
A more formal way is to use log level names: "FATAL"
, "ERROR"
,
"WARNING"
, "INFO"
, "DEBUG"
, and "VERBOSE"
. To specify the log
level of the message, append the log level to "!DEBUG"
, with a dash.
E.g.:
"!DEBUG-INFO Just letting you know that..."
To select the log level of a package, you can specify the level either
with the number of !
characters, as above, or adding the log level
as a suffix to the package name, separated by a dash. E.g.:
Sys.setenv(DEBUGME = "mypackage-INFO")
(Use either methods to set the log level, but do not mix them.)
By default debugme
prints the debug stack at the beginning of
the debug messages. The debug stack contains the functions in the call
stack that have (and emit) debug messages. To suppress printing the call
stack set the DEBUGME_SHOW_STACK
environment variable to no
.
If the DEBUGME_OUTPUT_FILE
environment variable is set to
a filename, then the output is written there instead of the standard
output stream of the R process.
If DEBUGME_OUTPUT_FILE
is not set, but DEBUGME_OUTPUT_DIR
is, then
a log file is created there, and the name of the file will contain
the process id. This is is useful for logging from several parallel R
processes.