Update Dockerfile

Dockerfile

@@ -1,5 +1,6 @@
 # TransVar API for HuggingFace Spaces
 # 仓库: https://github.com/pzweuj/TransVar2API
+# 启动脚本: scripts/hf_startup.sh
 
 FROM python:3.9-slim
 
@@ -9,9 +10,6 @@ ENV DEBIAN_FRONTEND=noninteractive \
     PYTHONUNBUFFERED=1 \
     PORT=7860
 
-# HF Dataset 名称（首次构建为空，之后设置为你的数据集）
-ENV HF_DATASET=""
-
 # 安装系统依赖
 RUN apt-get update && apt-get install -y \
     wget \
@@ -25,9 +23,6 @@ RUN apt-get update && apt-get install -y \
     libcurl4-gnutls-dev \
     && rm -rf /var/lib/apt/lists/*
 
-# 安装 huggingface_hub
-RUN pip3 install --no-cache-dir huggingface_hub
-
 # 从 GitHub 克隆仓库
 WORKDIR /app
 RUN git clone https://github.com/pzweuj/TransVar2API.git .
@@ -41,45 +36,25 @@ RUN mkdir -p /data/transvar_db/refseq_hg38 /data/transvar_db/refseq_hg19
 
 # ========== 下载参考基因组数据 ==========
 WORKDIR /data/transvar_db/refseq_hg38
+# hg38 参考基因组
+RUN wget -q -O hg38.fa.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/hg38.fa.gz && \
+    gunzip -f hg38.fa.gz && \
+    samtools faidx hg38.fa
+# hg38 GTF 格式的 RefSeq 注释
+RUN wget -q -O hg38.ncbiRefSeq.gtf.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/genes/hg38.ncbiRefSeq.gtf.gz
 
-# 下载 hg38 参考基因组
-RUN if [ ! -f hg38.fa ]; then \
-    wget -q -O hg38.fa.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/hg38.fa.gz && gunzip -f hg38.fa.gz && samtools faidx hg38.fa; \
-    fi
-
-# 下载 hg38 RefSeq
-RUN if [ ! -f ncbiRefSeq.txt.gz ]; then \
-    wget -q -O ncbiRefSeq.txt.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg38/database/ncbiRefSeq.txt.gz; \
-    fi
-
-# 下载 hg19 参考基因组
 WORKDIR /data/transvar_db/refseq_hg19
-
-RUN if [ ! -f hg19.fa ]; then \
-    wget -q -O hg19.fa.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.fa.gz && gunzip -f hg19.fa.gz && samtools faidx hg19.fa; \
-    fi
-
-# 下载 hg19 RefSeq
-RUN if [ ! -f ncbiRefSeq.txt.gz ]; then \
-    wget -q -O ncbiRefSeq.txt.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/ncbiRefSeq.txt.gz; \
-    fi
-
-# ========== 创建启动脚本 ==========
+# hg19 参考基因组
+RUN wget -q -O hg19.fa.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.fa.gz && \
+    gunzip -f hg19.fa.gz && \
+    samtools faidx hg19.fa
+# hg19 GTF 格式的 RefSeq 注释
+RUN wget -q -O hg19.ncbiRefSeq.gtf.gz https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/genes/hg19.ncbiRefSeq.gtf.gz
+
+# 复制启动脚本并设置权限
 WORKDIR /app
-RUN printf '#!/bin/bash\n\
-echo "Setting up transvar database..."\n\
-if [ ! -f /data/transvar_db/refseq_hg38/ncbiRefSeq.txt ]; then\n\
-    gunzip -c /data/transvar_db/refseq_hg38/ncbiRefSeq.txt.gz > /data/transvar_db/refseq_hg38/ncbiRefSeq.txt\n\
-fi\n\
-transvar config -k reference -v /data/transvar_db/refseq_hg38/hg38.fa --refversion hg38_refseq\n\
-transvar config -k refseq -v /data/transvar_db/refseq_hg38/ncbiRefSeq.txt --refversion hg38_refseq\n\
-if [ ! -f /data/transvar_db/refseq_hg19/ncbiRefSeq.txt ]; then\n\
-    gunzip -c /data/transvar_db/refseq_hg19/ncbiRefSeq.txt.gz > /data/transvar_db/refseq_hg19/ncbiRefSeq.txt\n\
-fi\n\
-transvar config -k reference -v /data/transvar_db/refseq_hg19/hg19.fa --refversion hg19_refseq\n\
-transvar config -k refseq -v /data/transvar_db/refseq_hg19/ncbiRefSeq.txt --refversion hg19_refseq\n\
-echo "Starting server..."\n\
-exec python3 -c "import os; import uvicorn; from server import app; uvicorn.run(app, host=\0470.0.0.0\047, port=int(os.environ.get(\047PORT\047, 7860)))"\n' > /app/start.sh && chmod +x /app/start.sh
+COPY scripts/hf_startup.sh /app/start.sh
+RUN chmod +x /app/start.sh
 
 EXPOSE 7860