As this is intended for use in CloudFormation templates (as an AllowedPattern), we are aiming for compatibility with Java's regular expression syntax.
However, due to the massive complexity and sheer scope of the overall expression we are maintiaining here, it will be broken up into various pieces for the cron() portion and versioned as native Perl regex.
The most recent working version of this pattern is also available (with unit tests!) on Regex101.
The complete pattern can be found in main.pl, in two different forms and assigned to two different global variables:
$expanded- here, the complete pattern is broken out across multiple lines and commented heavily (using the/xflag) for readability$collapsed- here, any non-leading and non-trailing comments and whitespace have been removed, making it fit for copy and paste elsewhere
Due to the complexity of the cron() matching, each individual field of the cron format has been broken out into its own file under the cron directory:
./cron
├── daysofmonth.pl
├── daysofweek.pl
├── hours.pl
├── minutes.pl
├── months.pl
└── years.pl
These individual field patterns are then merged back into the main pattern manually.
Currently, several valid rate and cron patterns are recorded as "positive" test cases as well as several invalid patterns as "negative" cases in tab delimited files under the tests directory.
You can test them against the latest complete patterns (again, housed in main.pl) by running the test script:
perl test.pl
# outputs TAP formatted results...The current multiline pattern can be condensed into a single line format (and compared against the currently versioned collapsed pattern) by running the build script:
Note: you may need to install the
Text::DiffCPAN module in order for this script to run successfully
perl build.pl
# prints a diff if anything in the main pattern has changed
# subsequently prints the condensed pattern to stdout