How do I use multiline regex match in Jenkins / Groovy

Question

In Jenkins I'm building a step that reads CHANGELOG.md and uses regex to extract the latest changelog from giving version.

The used .groovy code:

#!/usr/bin/env groovy

import java.nio.file.Files
import java.nio.file.Paths
import java.util.regex.Pattern

def String currentTag = '0.1.0';
def String changelogFileName = 'CHANGELOG.md';

def changelogText = new String(Files.readAllBytes(Paths.get(changelogFileName)))
def pattern = Pattern.compile("(?s)^## ${currentTag}.*?(?=\n\n## |\\z)", Pattern.MULTILINE)
def matcher = pattern.matcher(changelogText)

if (matcher.find()) {
    def latestChangelog = matcher.group()
    println latestChangelog.strip()
}

The used input file:

# Changelog - Project
All notable changes to this project will be documented in this file.
This project adheres to [Semantic Versioning](http://semver.org/).
Changelog based on guidelines from
[Keep a CHANGELOG](http://keepachangelog.com/).

## 0.3.0 - 2022-06-30
### Changed:
- MDN-XXXX: blabla
### Also Changed:
- MDN-XXXX: blabla

## 0.2.0 - 2022-06-09
### Added:
- MDN-XXXX: blabla

## 0.1.1 - 2022-06-08
### Fixed:
- MDN-XXXX: blabla

## 0.1.0 - 2022-06-08
### Added:
 - MDN-XXXX: blabla
 - MDN-XXXX: blabla

Locally this code works great. However, I want to use it in Jenkins, and there it's a security issues to directly import these:

import java.nio.file.Files
import java.nio.file.Paths
import java.util.regex.Pattern

Hence, I need to change it to Jenkins friendly code.

So I changed it to this:

    def changelogText = readFile(file: changelogFileName)

    def pattern = "(?s)^## ${currentTag}.*?(?=\n\n## |\\z)"
    def matcher = (changelogText =~ pattern)

    if (matcher.find()) {
        def latestChangelog = matcher.group();
        println(latestChangelog.strip());
    } else {
        println('Could not parse latest changelog.');
    }

However, I keep getting 'Could not parse latest changelog.'

It's quite hard to debug when I have to run it on Jenkins the whole time. Does anyone know how to make the Jenkins libraries available locally, or better, how I can match using Pattern.MULTILINE?

Or perhaps a suggestion for a whole different approach?

Answer 1

To match all symbols including newlines, you can use [\s\S]. This will work without any flags.

Consider the following demo:

String changelogText = """
# Changelog - Project
All notable changes to this project will be documented in this file.
This project adheres to [Semantic Versioning](http://semver.org/).
Changelog based on guidelines from
[Keep a CHANGELOG](http://keepachangelog.com/).

## 0.3.0 - 2022-06-30
### Changed:
- MDN-XXXX: blabla
### Also Changed:
- MDN-XXXX: blabla

## 0.2.0 - 2022-06-09
### Added:
- MDN-XXXX: blabla

## 0.1.1 - 2022-06-08
### Fixed:
- MDN-XXXX: blabla

## 0.1.0 - 2022-06-08
### Added:
 - MDN-XXXX: blabla
 - MDN-XXXX: blabla
"""

def currentTag = "0.2.0 - 2022-06-09"
def pattern = /## ${currentTag}[\s\S]*?(?=\n\n## |\z)/
def matcher = (changelogText =~ pattern)

if (matcher.find()) {
    def latestChangelog = matcher[0];
    println(latestChangelog);
} else {
    println('Could not parse latest changelog.');
}

Output:

## 0.2.0 - 2022-06-09
### Added:
- MDN-XXXX: blabla

Answer 2

In case someone wants to use awk instead, this in an awk solution, however, I still prefer the above one as it's better in case if the last entry 0.1.0 has some empty lines in it.

#!/usr/bin/env groovy

/**
 * Get latest changelog data from given filename and tag
 * @param currentTag string
 * @param changelogFileName string
 */
def call(
    String currentTag = '0.1.0',
    String changelogFileName = 'CHANGELOG.md'
) {
    def changelogText  = sh(returnStdout: true, script: "awk '{ RS=\"\" } /## ${currentTag}.*?(^## |\$)/' ${changelogFileName}").strip();
    if (changelogText) {
        return changelogText;
    } else {
        return "Could not parse the changelog of tag ${currentTag} in ${changelogFileName}. Is changelog pushed? Or perhaps the changelog contains a syntax error?";
    }
}