Update

Author: RainPPR

images/exporter-build/Dockerfile

@@ -36,8 +36,8 @@ RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
     --mount=type=cache,target=/var/lib/apt,sharing=locked \
     set -eux; \
     apt-get update; \
-    # 基础工具
-    apt-get install -y --no-install-recommends \
+    apt-get install -y eatmydata; \
+    eatmydata apt-get install -y \
         python3 python3-pip python3-dev \
         bash tini tar zstd unzip git ca-certificates curl wget \
         liblmdb0 build-essential libffi-dev ghostscript \
@@ -70,10 +70,10 @@ RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
         libxcomposite1 libxcursor1 libxdamage1 libxext6 libxfixes3 libxi6 libxrandr2 libxrender1 libxss1 \
         libxtst6 ca-certificates fonts-liberation libappindicator1 libnss3 lsb-release xdg-utils wget; \
     # 清理
+    apt-get purge -y --auto-remove eatmydata; \
     apt-get clean; \
-    rm -rf /var/lib/apt/lists/*; \
-    fc-cache -fv; \
-    git config --system --add safe.directory '/__w/*'
+    git config --system --add safe.directory '/__w/*'; \
+    rm -rf /var/lib/apt/lists/*
 
 WORKDIR /app
 
@@ -89,9 +89,10 @@ RUN --mount=type=bind,source=.python-version,target=.python-version \
     uv sync --locked --no-cache
 
 # 6. 拷贝应用代码
-COPY src/ ./src/
-COPY scripts/ ./scripts/
-COPY templates/ ./templates/
+COPY src/ /app/src/
+COPY scripts/ /app/scripts/
+COPY templates/ /app/templates/
+COPY --chmod=755 bin/ /usr/local/bin/
 
 ENTRYPOINT ["/usr/bin/tini", "--"]
-CMD ["npm", "start"]
+CMD ["export-pdf"]

images/exporter-build/bin/export-pdf

@@ -0,0 +1,29 @@
+#!/bin/bash
+set -e
+
+BOOK_NAME="$1"
+
+if [ -z "$BOOK_NAME" ]; then
+    echo "Usage: export-pdf <book_name>"
+    exit 1
+fi
+
+echo "[Export-PDF] Starting Puppeteer printing..."
+node /app/src/index.js "$BOOK_NAME"
+
+echo "[Export-PDF] Generating TeX templates..."
+uv run /app/scripts/processor.py "${BOOK_NAME}.json" --plan-only
+
+echo "[Export-PDF] Compiling TeX with LaTeX..."
+# Note: We assume latexmk is available in the environment OR this part 
+# is handled by a separate step in GHA using the texlive-full container.
+# According to the workflow, we use the texlive-full container separately.
+# So this script might just handle the parts that run in exporter-build.
+
+# If we want it to be a complete wrapper, we'd need latexmk here too.
+# But keeping it aligned with the user's 3-step GHA workflow is safer.
+
+echo "[Export-PDF] Finalizing synthesis and bookmarks..."
+uv run /app/scripts/processor.py "${BOOK_NAME}.json" --merge
+
+echo "[Export-PDF] Done."

images/exporter-build/scripts/processor.py

@@ -1,9 +1,6 @@
 import fitz
 import json
 import os
-import sys
-import yaml
-import subprocess
 from pathlib import Path
 from jinja2 import Environment, FileSystemLoader
 
@@ -16,7 +13,9 @@ def __init__(self, book_json_path, output_dir="build"):
         with open(self.book_json_path, "r", encoding="utf-8") as f:
             self.book_data = json.load(f)
 
-        self.jinja_env = Environment(loader=FileSystemLoader("templates"))
+        # 锚定模板目录到镜像内的绝对路径
+        template_dir = os.environ.get("TEMPLATES_DIR", "/app/templates")
+        self.jinja_env = Environment(loader=FileSystemLoader(template_dir))
         self.final_doc = fitz.open()
         self.page_offset = 0
         self.toc_data = []
@@ -33,8 +32,8 @@ def extract_precise_toc(self, doc, offset):
             blocks = page.get_text("dict")["blocks"]
             for b in blocks:
                 if "lines" in b:
-                    for l in b["lines"]:
-                        for s in l["spans"]:
+                    for line in b["lines"]:
+                        for s in line["spans"]:
                             # 粗放式匹配：字体大且粗的可能是标题
                             if s["size"] > 12:
                                 text = s["text"].strip()

images/exporter-build/src/index.js

@@ -134,8 +134,8 @@ async function main() {
         process.exit(1);
     }
 
-    const buildDir = path.join(process.cwd(), 'site', 'build');
     const siteDir = path.join(process.cwd(), 'site');
+    const buildDir = path.join(siteDir, 'build');
     const tasksFile = path.join(buildDir, `download_${bookName}.json`);
 
     if (!fs.existsSync(tasksFile)) {