Got an error: $msg
\";\n}\nsetmessage(\\&handleerrors);\n}\n\nIn order to correctly intercept compile-time errors, you should call setmessage() from within a\nBEGIN{} block.\n\nDOING MORE THAN PRINTING A MESSAGE IN THE EVENT OF PERL ERRORS\nIf fatalsToBrowser in conjunction with setmessage does not provide you with all of the\nfunctionality you need, you can go one step further by specifying a function to be executed any\ntime a script calls \"die\", has a syntax error, or dies unexpectedly at runtime with a line like\n\"undef->explode();\".\n\nuse CGI::Carp qw(setdiehandler);\nBEGIN {\nsub handleerrors {\nmy $msg = shift;\nprint \"content-type: text/html\\n\\n\";\nprint \"Got an error: $msg
\";\n\n#proceed to send an email to a system administrator,\n#write a detailed message to the browser and/or a log,\n#etc....\n}\nsetdiehandler(\\&handleerrors);\n}\n\nNotice that if you use setdiehandler(), you must handle sending HTML headers to the browser\nyourself if you are printing a message.\n\nIf you use setdiehandler(), you will most likely interfere with the behavior of\nfatalsToBrowser, so you must use this or that, not both.\n\nUsing setdiehandler() sets SIG{DIE} (as does fatalsToBrowser), and there is only one\nSIG{DIE}. This means that if you are attempting to set SIG{DIE} yourself, you may\ninterfere with this module's functionality, or this module may interfere with your module's\nfunctionality.\n\nSUPPRESSING PERL ERRORS APPEARING IN THE BROWSER WINDOW\nA problem sometimes encountered when using fatalsToBrowser is when a \"die()\" is done inside an\n\"eval\" body or expression. Even though the fatalsToBrower support takes precautions to avoid\nthis, you still may get the error message printed to STDOUT. This may have some undesirable\neffects when the purpose of doing the eval is to determine which of several algorithms is to be\nused.\n\nBy setting $CGI::Carp::TOBROWSER to 0 you can suppress printing the \"die\" messages but without\nall of the complexity of using \"setdiehandler\". You can localize this effect to inside \"eval\"\nbodies if this is desirable: For example:\n\neval {\nlocal $CGI::Carp::TOBROWSER = 0;\ndie \"Fatal error messages not sent browser\"\n}\n# $@ will contain error message\n" } ] }, "MAKING WARNINGS APPEAR AS HTML COMMENTS": { "content": "It is also possible to make non-fatal errors appear as HTML comments embedded in the output of\nyour program. To enable this feature, export the new \"warningsToBrowser\" subroutine. Since\nsending warnings to the browser before the HTTP headers have been sent would cause an error, any\nwarnings are stored in an internal buffer until you call the warningsToBrowser() subroutine with\na true argument:\n\nuse CGI::Carp qw(fatalsToBrowser warningsToBrowser);\nuse CGI qw(:standard);\nprint header();\nwarningsToBrowser(1);\n\nYou may also give a false argument to warningsToBrowser() to prevent warnings from being sent to\nthe browser while you are printing some content where HTML comments are not allowed:\n\nwarningsToBrowser(0); # disable warnings\nprint \"\\n\";\nwarningsToBrowser(1); # re-enable warnings\n\nNote: In this respect warningsToBrowser() differs fundamentally from fatalsToBrowser(), which\nyou should never call yourself!\n", "subsections": [] }, "OVERRIDING THE NAME OF THE PROGRAM": { "content": "CGI::Carp includes the name of the program that generated the error or warning in the messages\nwritten to the log and the browser window. Sometimes, Perl can get confused about what the\nactual name of the executed program was. In these cases, you can override the program name that\nCGI::Carp will use for all messages.\n\nThe quick way to do that is to tell CGI::Carp the name of the program in its use statement. You\ncan do that by adding \"name=cgicarplogname\" to your \"use\" statement. For example:\n\nuse CGI::Carp qw(name=cgicarplogname);\n\n. If you want to change the program name partway through the program, you can use the\n\"setprogname()\" function instead. It is not exported by default, you must import it explicitly\nby saying\n\nuse CGI::Carp qw(setprogname);\n\nOnce you've done that, you can change the logged name of the program at any time by calling\n\nsetprogname(newprogramname);\n\nYou can set the program back to the default by calling\n\nsetprogname(undef);\n\nNote that this override doesn't happen until after the program has compiled, so any compile-time\nerrors will still show up with the non-overridden program name\n", "subsections": [] }, "TURNING OFF TIMESTAMPS IN MESSAGES": { "content": "If your web server automatically adds a timestamp to each log line, you may not need CGI::Carp\nto add its own. You can disable timestamping by importing \"noTimestamp\":\n\nuse CGI::Carp qw(noTimestamp);\n\nAlternatively you can set $CGI::Carp::NOTIMESTAMP to 1.\n\nNote that the name of the program is still automatically included in the message.\n", "subsections": [] }, "GETTING THE FULL PATH OF THE SCRIPT IN MESSAGES": { "content": "Set $CGI::Carp::FULLPATH to 1.\n", "subsections": [] }, "AUTHOR INFORMATION": { "content": "The CGI.pm distribution is copyright 1995-2007, Lincoln D. Stein. It is distributed under the\nArtistic License 2.0. It is currently maintained by Lee Johnson with help from many\ncontributors.\n\nAddress bug reports and comments to: https://github.com/leejo/CGI.pm/issues\n\nThe original bug tracker can be found at:\nhttps://rt.cpan.org/Public/Dist/Display.html?Queue=CGI.pm\n\nWhen sending bug reports, please provide the version of CGI.pm, the version of Perl, the name\nand version of your Web server, and the name and version of the operating system you are using.\nIf the problem is even remotely browser dependent, please provide information about the affected\nbrowsers as well.\n", "subsections": [] }, "SEE ALSO": { "content": "Carp, CGI::Base, CGI::BasePlus, CGI::Request, CGI::MiniSvr, CGI::Form, CGI::Response.\n", "subsections": [] } }, "summary": "CGI::Carp - CGI routines for writing to the HTTPD (or other) error log", "flags": [], "examples": [], "see_also": [] }