NAME MultiTail - Tail multiple files for Unix systems SYNOPSIS use File::MultiTail; DESCRIPTION This perl library uses perl5 objects to make it easy to tail a dynamic list of files and match/except lines using full regular expressions. File::MultiTail; will tail multiple files and return the records read to a Data Structure. The Data Structure can be processed by MultiTail functions The files specified are processed in accordance with the following rules: Note: File devices and inode number uniquely identify each entry in the UNIX filesystem. If the stat() command shows them to be the same, then only one will be used. (Check for links) (1) Files that exist at program start time will be positioned to Object attribute "NumLines" before input. (2) Files that become available subsequently will be read from the beginning. Attribute ScanForFiles must be set to True (>=1) for the option. (3) If a file that has been selected as per rules 1 or 2 is deleted or truncated input will continue until end-of-file is reached before the file is closed. (4) If a file is deleted and it is recreated it is treated as a new file, will be read from the beginning (5) To conserve file descriptors, files that are selected for input are not actually opened until data is present beyond the input point selected. For example, if a file exists when ptail starts, ptail will determine the file mtime at that time and only open the file when the mtime increases. Note: mtime = Time when data was last modified. Changed by the following functions: creat, mknod, pipe, utime, and write. (6) If an opened file has not been updated for MultiTail Object attribute "MaxAge" minutes it will be closed. It will be reopened if it is later updated. (7) Since MultiTail is OO you can alway change its attributes. If you change the list of file to be tailed (Files attribute) the attribute ScanForFiles will set to true and all dir and files ilists will be check for new files. METHODS new Creating a new PTAIL object $tail = File::MultiTail->new( OutputPrefix => 'tf', RemoveDuplicate => $True, Files => ['/var/log','/var/adm/*.log'] ); Or $tail = File::MultiTail->new(); $tail->Files(['/var/log','/var/adm/*.log']); $tail->RemoveDuplicate($True); class/object method takes arguments ( All have defaults ) and returning a reference to a newly created MultiTail object. File : File attribute accepts file names that includes both explicit file/dir names and file/dir expressions. Duplicate file paths are rejected, along with non-duplicate names that resolve to a device/inode combination that is currently being read. Pattern : Arguments can be a file name or an array of patterns Stores in object attribute "LineArray" all lines that contain the patterns (Default is *) ExceptPattern : Arguments can be a file name or an array of patterns Stores in object attribute "LineArray" all lines except those that contain the patterns (Default is *) MaxAge : Maximum time in minute that an open file will be held open without an update. (Default is 10 minute) NumLines : Files that exist at MultiTail start time will have up to NumLines lines from the end of file displayed. Similar to tail -NumLines. (Default is 10) Fuction : Reference to a function that will be run by the MultiTail object. MultiTail object will pass a ref array of all the lines read from the files and passed through any filters you set in the object to the function. 0 = (Default) No Fuction ScanForFiles : Maximum time in minute before Read will scan for new files. If you change the attribute "Files" with fuction update_attribute the next Read will scan for new files. 0 = (Default) Off RemoveDuplicate : Removes all duplicate lines from LineArray 0 = (Default) Off 1 = On : Turn Debuging messages on for MultiTail.pm. 0 = (Default) Off 1 = On OutputPrefix : Determines the prefix applied to each output record. Output records are store in MultiTail object attribute "LineArray" The prefix is separated from the record by ': '. Prefixes supported are: p : path name of the input file f : file name of the input file t : time in HHMMSS tg : time in HHMMSS GMT pt : path and time ptg: path and time GMT ft : file and time ftg: file and time GMT tp : time and path tpg: time and path GMT tf : time and file tfg: time and file GMT 0 = (Default) No prefix GMT = Greenwich time ZONE Exit Codes 1001 - MaxAge is less the zero 1002 - NumLines is less the zero 1003 - OutputPrefix must one ( ) 1004 - Pattern must is not a file or ARRAY 1005 - ExceptPattern is not a file or ARRAY 1006 - Debug is not 0 or 1 1007 - ScanForFiles is less the zero 1008 - RemoveDuplicate is not 0 or 1 1009 - Function is not ref to fuction 1010 - File attribute not set read Read all new data from file $tail->read Read all new date from tailed files and return new lines as part of the Data Structure (MultiTail Object attribute LineArray) print Print all line contained in MultiTail Object attribute LineArray $tail->print update_attribute Allow you to Update (Change) any MultiTail Object attribute $tail->update_attribute( Files => ["/var/log","/var/adm","/home/nnysgm/logwatcher/foo*"], ExceptPattern => /home/nnysgm/ExceptPattern.txt, RemoveDuplicate => $True ); This changes the Files, ExceptPattern and RemoveDuplicate attributes for the Object $tail. New files will be scanned for during next Read if "Files" attribute is changed. Also you can use supplied methods to set attribute values. $tail->RemoveDuplicate($True); $tail->NumLines(100); version Return version number on PTAIL package $tail->version debug Toggle the debug switch for MultiTail package $tail->debug There are a number of function in the MultiTail.pm module. o printstat : Print out stat output for each file. o printfilestates : Print out All file states. (See note in MultiTail.pm for function OpenUpdateFiles) o printpat : Print out lines from pattern file array. o printexceptpat : Print out line from pattern file except array. close_all_files Closes all file that are being tailed $tail->close_all_files EXAMPLE 1) use File::MultiTail; $tail1=File::MultiTail->new ( OutputPrefix => "f", Debug => "$True", Files => ["/var/adm/messages"] ); while(1) { $tail1->read; # $tail1->print; sleep 10; } $tail1=MultiTail->new : Create new ptail object - Files => Tail file /var/adm/messages - OutputPrefix => Prepend the name of the file beginning of each line in object attribute "LineArray" $tail1->read : Read all line from files $tail1->print : Print all line in object attribute "LineArray"; 2) use File::MultiTail; $tail1=File::MultiTail->new ( OutputPrefix => "tf", Pattern => "/home/nnysgm/logwatcher/pattern", ExceptPattern => "/home/nnysgm/logwatcher/epattern", Fuction = > \&want, Files => ["/var/adm","/var/log/*.log"] ); while(1) { $tail1->read; # $tail1->print; sleep 10; } sub want { (your code .... ); } $tail1=File::MultiTail->new : Create new ptail object - OutputPrefix => Prepend the name of the file and time to the beginning of each line in object attribute "LineArray" - ExceptPattern => Stores in object attribute "LineArray" all lines except those that contain the patterns from file "epattern" - Pattern => Stores in object attribute "LineArray" all lines that contain the patterns from file "pattern" - Fuction => ref to a function that will be run by MultiTail object. MultiTail object will pass a ref array to the function of all the lines read from the file and passed through any filters you set in the object. - Files => Tail all files in dir /var/adm and all .log files dir /var/log. $tail1->read : Read all line from files $tail1->print : Print all line in object attribute "LineArray"; 3) use File::MultiTail; $tail=File::MultiTail->new; $tail->OutputPrefix(tf); $tail->Fuction(\&want); $tail->Files(["/var/adm","/var/log/*.log"]); while(1) { $tail1->read; } sub want { (your code .... ); } AUTHOR Stephen Miano, stevem@esm.com SEE ALSO perl(1).