Split on line feed when parsing transport files in the build (elastic#135978)

When parsing transport files that may contain comments we split on
newline. On windows systems these files _might_ have the line feed
replaced by CRLF. This commit adjusts the parsing to only split on line
feed so that regardless of what the system line separator is we will
split the lines as needed.

Author: rjernst

build-tools-internal/src/main/java/org/elasticsearch/gradle/internal/transport/TransportVersionDefinition.java

@@ -22,9 +22,11 @@ public static TransportVersionDefinition fromString(Path file, String contents,
 
         String idsLine = null;
         if (contents.isEmpty() == false) {
-            String[] lines = contents.split(System.lineSeparator());
+            // Regardless of whether windows newlines exist (they could be added by git), we split on line feed.
+            // All we care about skipping lines with the comment character, so the remaining \r won't matter
+            String[] lines = contents.split("\n");
             for (String line : lines) {
-                line = line.replaceAll("\\s+", "");
+                line = line.strip();
                 if (line.startsWith("#") == false) {
                     idsLine = line;
                     break;

build-tools-internal/src/main/java/org/elasticsearch/gradle/internal/transport/TransportVersionUpperBound.java

@@ -24,9 +24,11 @@ public static TransportVersionUpperBound fromString(Path file, String contents)
         String branch = filename.substring(slashIndex == -1 ? 0 : (slashIndex + 1), filename.length() - 4);
 
         String idsLine = null;
-        String[] lines = contents.split(System.lineSeparator());
+        // Regardless of whether windows newlines exist (they could be added by git), we split on line feed.
+        // All we care about skipping lines with the comment character, so the remaining \r won't matter
+        String[] lines = contents.split("\n");
         for (String line : lines) {
-            line = line.replaceAll("\\s+", "");
+            line = line.strip();
             if (line.startsWith("#") == false) {
                 idsLine = line;
                 break;