[alloutflo]

With trimBlocks enabled, CRLF line endings after tags and comments are currently left in the output, producing the extra blank lines reported in #1087. Both token scanners only recognize LF in this path.

Consume exactly one immediately following LF or CRLF sequence in both scanners. Advance the line counter once and reset the column origin after the entire consumed sequence so subsequent token locations remain correct. Existing standalone-CR, ordinary text, and trailing-newline behavior is preserved. No API, configuration, or dependency changes.

Validation

• The original issue's interpreter-rendering example fails before the fix and passes afterward, producing exactly the three expected configuration lines with their CRLF endings.
• 10 new tests cover both scanners, tags/comments, LF/CRLF and mixed sequences, multiple newlines, EOF, standalone CR, intervening whitespace, expression tokens, disabled trimming, token locations, custom delimiters, and raw boundaries. The focused scanner/parser run passed all 104 tests.
• Full Maven verify passed with Temurin 17, 21, and 25, with Maven running on Java 25: 2,248 tests per run, zero failures/errors, 23 skipped. Java 21 and 25 used clean verify to rebuild with their respective toolchains.
• Spotless, Checkstyle, SpotBugs, PMD, independent diff review, and git diff --check passed.

The full runs use en_US for the previously reproduced, unrelated locale-sensitive number-format tests.

Implemented and independently reviewed with assistance from OpenAI Codex.

Fixes #1087