Chunk-Reading Long Sentences — Read Straight Through, No Backtracking

Take the long foreign sentence in the input and decode it by "chunk-reading" — breaking it into meaning units you understand front-to-back. The goal is not smooth translation but training to read without backtracking.

Work rules:
1. First show the original as a blockquote, inserting ' / ' at meaning boundaries (subject / verb / prepositional phrase / relative clause, 3-6 chunks).
2. Make a per-chunk table: (col1) the chunk, (col2) a front-to-back literal read. Keep the word order — render it "in the order it arrives," not paraphrased.
3. Below the table, "read together": one naturally connected sentence in order (smooth only here).
4. Give 1-3 "sticking points" — inserted clauses, inversion, long modifiers, tricky tenses — where beginners stall, with how to handle each (e.g., bracket the inserted clause and skip it).
5. End with a "takeaway": one line generalizing that language's word-order habit so it transfers to the next sentence.

Forbidden: Do not add meaning not in the original. Do not scare with excess grammar jargon — only what's needed. Do not call machine-literal awkwardness "natural."
Output format: markdown. Original (blockquote) -> chunk table -> read-together -> sticking points -> takeaway. Emoji only as head seasoning.
Output language: Explanations and literal reads in my language, original chunks in the source language. Works identically for any input language (English, Japanese, Chinese, etc.).
Fallback: If no sentence is given, invent one mid-level example learners often stall on, demo it, and state that in one line at the top.
Self-contained: Finish within this reply. No external sites or extra steps.

----
Input
- {{long sentence}} (optional): one foreign sentence to chunk-read. If blank, demo with an example.
- {{my level}} (optional): beginner / intermediate / advanced. Tunes explanation depth.