introduction-to-lattice-qcd-latex

name	INTRODUCTION_TO_LATTICE_QCD_latex
description	将 R. Gupta《Introduction to Lattice QCD》转写为 LaTeX 的完整 skill — 从 pdftotext 提取逐节重建可编译 LaTeX（20 节 → 章），全局图号 Fig.1–35 / 表 1–8，公式按章自动编号

Conversion Guide — INTRODUCTION TO LATTICE QCD → LaTeX

Source: R. Gupta, "Introduction to Lattice QCD" (arXiv:hep-lat/9807028v1, 11 Jul 1998), a 150-page set of school lectures. This repo is a faithful LaTeX transcription.

Directory layout

main.tex                      master document (already written)
chapters/secNN_name.tex       one chapter file per original section (1–20)
extract/secNN.txt             raw pdftotext output per section (INPUT, read-only)
images/figN.png               extracted figures, Fig.1–Fig.35 (referenced from chapters)

Numbering model (important)

The source numbers sections 1–20, subsections 6.1, 6.2, …, sub-subsections 14.2.1, …, equations (5.1), …, but figures globally (Fig. 1–35) and tables globally (Table 1–8). main.tex is set up so:

• each original section → \chapter (auto-numbered 1–20, matching the source),
• each original subsection → \section, sub-subsection → \subsection (auto-numbering then reproduces the source's 6.1, 14.2.1, …),
• equations auto-number per chapter → (5.1), (6.2), … matching the source,
• figures and tables are forced to global numbering so they come out 1–35 / 1–8.

So: do not hand-number anything. Write \begin{equation}, \begin{table}, \begin{figure} with empty numbering and let the counter do the work.

Per-chapter transcription rules

1. Input: read extract/secNN.txt. It has page separators % -------- PDF page N --------. Strip the running heads ("INTRODUCTION TO LATTICE QCD" / "Rajan Gupta") and standalone page numbers (single integers that appear at the top or bottom of a page block).
2. Output: write chapters/secNN_name.tex (lower-case, underscore for the name; e.g. sec05_minkowski_euclidean.tex).
3. First line: \chapter{<exact source section title>} (chapter number is automatic).
4. Body paragraphs: plain text, blank line between paragraphs, \noindent not required.
5. Equations: the pdftotext output is a mangled linear rendering of the math. Reconstruct each equation into correct LaTeX using your knowledge of the physics. Examples:
   • x2E = 4 X x2i = x 2 − t2 = −x2M → x_E^2 = \sum_{i=1}^4 x_i^2 = \mathbf{x}^2 - t^2 = -x_M^2
   • p0 ≡ E → ip4 → p_0 \equiv E \to i p_4
   • γµ (1 − γ5 ) V → \gamma_\mu (1 - \gamma_5) V
   • W R T = 1/Z R dU WRT eβ/2N (W11 +W11 ) → a proper integral equation. Display equations go in \begin{equation} ... \end{equation}; inline math in \( ... \) or $ ... $.
6. Tables: transcribe as \begin{table}[htbp] \centering \begin{tabular}{...} ... \end{tabular} \caption{...} \label{tab:N} \end{table}.
7. Figures: wherever the source shows Fig. N. <caption>, insert \begin{figure}[htbp] \centering \includegraphics[width=0.8\textwidth]{images/figN.png} \caption{<caption text>} \label{fig:N} \end{figure}. Place it at the natural location (the paragraph discussing it). Omit the Fig. N. prefix from the caption text.
8. Faithfulness: transcribe the prose as-is — do not add commentary, editorial notes, or new content. Fix pdftotext artifacts (stray line breaks, split words, spacing). Preserve the meaning and all physics content.
9. Use LaTeX conventions: --- for em-dash, ` and ' for quotes, \textit for italic, \textbf for bold. Greek and math symbols belong in math mode, not as Unicode.
10. Sub-sections like "14.2.1. Wilson Loops" → \subsection{Wilson Loops}.

Do not

• Do not \tag equations, \setcounter, or hardcode any numbers.
• Do not invent figures — only the 35 extracted PNGs exist.
• Do not leave Unicode math symbols in the text; convert to LaTeX math.