refactor

eric9n · Sep 12, 2024 · 8231d06 · 8231d06
1 parent 5c18ed0
commit 8231d06
Show file tree

Hide file tree

Showing 19 changed files with 299 additions and 1,622 deletions.
diff --git a/.github/workflows/release.yaml b/.github/workflows/release.yaml
@@ -0,0 +1,231 @@
+name: Release
+on:
+  push:
+    tags:
+    - 'v[0-9]+.[0-9]+.[0-9]+'
+
+env:
+  PROJECT_NAME: kun_peng
+  REPO_NAME: ${{ github.repository }}
+  BREW_TAP: eric9n/homebrew-tap
+  DESC: "An ultra-fast, low-memory footprint and accurate taxonomy classifier for all"
+
+jobs:
+  dist:
+    permissions:
+      contents: write
+    name: Dist
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        build: [x86_64-linux, x86_64-macos, x86_64-windows, aarch64-macos]
+        include:
+        - build: x86_64-linux
+          os: ubuntu-latest
+          rust: stable
+          target: x86_64-unknown-linux-gnu
+        - build: x86_64-macos
+          os: macos-latest
+          rust: stable
+          target: x86_64-apple-darwin
+        - build: x86_64-windows
+          os: windows-latest
+          rust: stable
+          target: x86_64-pc-windows-msvc
+        - build: aarch64-macos
+          os: macos-latest
+          rust: stable
+          target: aarch64-apple-darwin
+
+    steps:
+      - name: Checkout sources
+        uses: actions/checkout@v2
+        with:
+          submodules: true
+
+      - name: Install ${{ matrix.rust }} toolchain
+        uses: actions-rs/toolchain@v1
+        with:
+          profile: minimal
+          toolchain: ${{ matrix.rust }}
+          target: ${{ matrix.target }}
+          override: true
+
+      - name: Run cargo test
+        uses: actions-rs/cargo@v1
+        with:
+          use-cross: ${{ matrix.cross }}
+          command: test
+          args: --release --locked --target ${{ matrix.target }}
+
+      - name: Build release binary
+        uses: actions-rs/cargo@v1
+        with:
+          use-cross: ${{ matrix.cross }}
+          command: build
+          args: --release --locked --target ${{ matrix.target }}
+
+      - name: Build archive
+        shell: bash
+        run: |
+          mkdir dist
+          if [ "${{ matrix.os }}" = "windows-latest" ]; then
+            cp "target/${{ matrix.target }}/release/${{ env.PROJECT_NAME }}.exe" "dist/${{ env.PROJECT_NAME }}-${{ github.ref_name }}-${{ matrix.target }}.exe"
+          else
+            cp "target/${{ matrix.target }}/release/${{ env.PROJECT_NAME }}" "dist/${{ env.PROJECT_NAME }}-${{ github.ref_name }}-${{ matrix.target }}"
+          fi
+
+      # Set up the GitHub CLI
+      - name: Install GitHub CLI (macOS)
+        run: |
+          brew install gh
+        if: matrix.os == 'macos-latest'
+
+      - name: Install GitHub CLI (Ubuntu)
+        run: |
+          sudo apt install -y gh
+        if: matrix.os == 'ubuntu-latest'
+
+      - name: Install GitHub CLI (Windows)
+        run: |
+          choco install gh
+        if: matrix.os == 'windows-latest'
+
+      # Log in to the GitHub CLI
+      - name: Login to GitHub CLI
+        run: echo "${{ secrets.GITHUB_TOKEN }}" | gh auth login --with-token
+
+      - name: Upload Release Asset
+        run: |
+          if [ "${{ matrix.os }}" = "windows-latest" ]; then
+            ASSET_NAME="${{ env.PROJECT_NAME }}-${{ github.ref_name }}-${{ matrix.target }}.exe"
+          else
+            ASSET_NAME="${{ env.PROJECT_NAME }}-${{ github.ref_name }}-${{ matrix.target }}"
+          fi
+          gh release upload ${{ github.ref_name }} \
+            "./dist/$ASSET_NAME" \
+            --clobber
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+
+      - name: Set macOS artifact name
+        if: matrix.os == 'macos-latest' && matrix.target == 'x86_64-apple-darwin'
+        run: echo "macos_artifact=${{ env.PROJECT_NAME }}-${{ github.ref_name }}-${{ matrix.target }}" >> $GITHUB_OUTPUT
+        id: artifact_name
+
+      - name: Build on CentOS 7
+        if: matrix.os == 'ubuntu-latest'
+        run: |
+          docker run --name centos7-container -v $GITHUB_WORKSPACE:/github/workspace -w /github/workspace centos:7 \
+          /bin/bash -c "echo '[base]' > /etc/yum.repos.d/CentOS-Base.repo; \
+          echo 'name=CentOS-7 - Base' >> /etc/yum.repos.d/CentOS-Base.repo; \
+          echo 'baseurl=http://vault.centos.org/centos/7/os/x86_64/' >> /etc/yum.repos.d/CentOS-Base.repo; \
+          echo 'gpgcheck=1' >> /etc/yum.repos.d/CentOS-Base.repo; \
+          echo 'enabled=1' >> /etc/yum.repos.d/CentOS-Base.repo; \
+          echo 'gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-CentOS-7' >> /etc/yum.repos.d/CentOS-Base.repo; \
+          yum update -y && yum install -y gcc make openssl openssl-devel && \
+          curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && export PATH=\$HOME/.cargo/bin:\$PATH && cd /github/workspace && cargo build --release"
+
+          docker cp centos7-container:/github/workspace/target/release/${{ env.PROJECT_NAME }} ./dist/${{ env.PROJECT_NAME }}-${{ github.ref_name }}-centos7
+          docker rm centos7-container
+
+      - name: Upload CentOS 7 Release Asset
+        if: matrix.os == 'ubuntu-latest'
+        run: |
+          gh release upload ${{ github.ref_name }} \
+            "./dist/${{ env.PROJECT_NAME }}-${{ github.ref_name }}-centos7" \
+            --clobber
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+  update-formula:
+    needs: dist
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v2
+
+      - name: Extract version
+        id: extract-version
+        run: echo "tag-name=${GITHUB_REF#refs/tags/}" >> $GITHUB_OUTPUT
+
+      - name: Set environment variables
+        run: |
+          echo "PROJECT_NAME=${{ env.PROJECT_NAME }}" >> $GITHUB_ENV
+          echo "REPO_NAME=${{ env.REPO_NAME }}" >> $GITHUB_ENV
+
+      - name: Verify release assets
+        run: |
+          VERSION=${{ steps.extract-version.outputs.tag-name }}
+          REPO=${{ env.REPO_NAME }}
+          PROJECT=${{ env.PROJECT_NAME }}
+          X86_64_URL="https://github.com/${REPO}/releases/download/${VERSION}/${PROJECT}-${VERSION}-x86_64-apple-darwin"
+          AARCH64_URL="https://github.com/${REPO}/releases/download/${VERSION}/${PROJECT}-${VERSION}-aarch64-apple-darwin"
+
+          if curl --output /dev/null --silent --head --fail "$X86_64_URL"; then
+            echo "x86_64 binary exists"
+          else
+            echo "x86_64 binary does not exist"
+            exit 1
+          fi
+
+          if curl --output /dev/null --silent --head --fail "$AARCH64_URL"; then
+            echo "aarch64 binary exists"
+          else
+            echo "aarch64 binary does not exist"
+            exit 1
+          fi
+
+      - name: Update Homebrew formula
+        env:
+          COMMITTER_TOKEN: ${{ secrets.COMMITTER_TOKEN }}
+        run: |
+          VERSION=${{ steps.extract-version.outputs.tag-name }}
+          REPO=${{ env.REPO_NAME }}
+          PROJECT=${{ env.PROJECT_NAME }}
+          DESC="${{ env.DESC }}"
+          X86_64_URL="https://github.com/${REPO}/releases/download/${VERSION}/${PROJECT}-${VERSION}-x86_64-apple-darwin"
+          AARCH64_URL="https://github.com/${REPO}/releases/download/${VERSION}/${PROJECT}-${VERSION}-aarch64-apple-darwin"
+
+          # 下载并更新formula
+          git clone https://github.com/${{ env.BREW_TAP }}.git homebrew-tap
+          cd homebrew-tap
+
+          cat > Formula/${PROJECT}.rb <<EOL
+          class $(echo $PROJECT | perl -pe 's/(^|_|-)(\w)/\U$2/g') < Formula
+            desc "${DESC}"
+            homepage "https://github.com/${REPO}"
+            version "${VERSION}"
+
+            on_macos do
+              if Hardware::CPU.intel?
+                url "${X86_64_URL}"
+                sha256 "$(curl -sL ${X86_64_URL} | shasum -a 256 | cut -d ' ' -f 1)"
+              else
+                url "${AARCH64_URL}"
+                sha256 "$(curl -sL ${AARCH64_URL} | shasum -a 256 | cut -d ' ' -f 1)"
+              end
+            end
+
+            def install
+              if Hardware::CPU.intel?
+                bin.install "${PROJECT}-#{version}-x86_64-apple-darwin" => "${PROJECT}"
+              else
+                bin.install "${PROJECT}-#{version}-aarch64-apple-darwin" => "${PROJECT}"
+              end
+            end
+
+            test do
+              system "#{bin}/${PROJECT}", "--version"
+            end
+          end
+
+          EOL
+
+          git config user.name github-actions
+          git config user.email github-actions@github.com
+          git add Formula/${PROJECT}.rb
+          git commit -m "Updating formula for ${PROJECT} to ${VERSION}"
+          git push https://${{ secrets.COMMITTER_TOKEN }}@github.com/${{ env.BREW_TAP }}.git main
diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
@@ -1,15 +1,17 @@
 name: "publish"
 
-on:
-  create:
-    tags:
-      - '*'
+# Commenting out the current trigger
+# on:
+#   create:
+#     tags:
+#       - '*'
+
 
 # This is the example from the readme.
 # On each push to the `release` branch it will create or update a GitHub release, build your app, and upload the artifacts to the release.
 env:
   CARGO_TERM_COLOR: always
-  BINARIES_LIST: 'ncbi_dl kun_peng'
+  BINARIES_LIST: 'kun_peng'
   PROJECT_PREFIX: 'Kun-peng-'
 
 jobs:

diff --git a/Cargo.toml b/Cargo.toml
@@ -1,5 +1,5 @@
 [workspace]
-members = ["ncbi", "kr2r", "seqkmer"]
+members = ["kr2r", "seqkmer"]
 resolver = "2"
 
 [profile.release]

diff --git a/README.md b/README.md
@@ -41,14 +41,56 @@ Follow these steps to install Kun-peng and run the examples.
 
 If you prefer not to build from source, you can download the pre-built binaries for your platform from the GitHub [releases page](https://github.com/eric9n/Kun-peng/releases).
 
+For Linux users (CentOS 7 compatible):
+
 ```bash
-mkdir kun_peng_v0.6.15
-tar -xvf Kun-peng-v0.6.15-centos7.tar.gz -C kun_peng_v0.6.15
-# Add environment variable
-echo 'export PATH=$PATH:~/biosoft/kun_peng_v0.6.15' >> ~/.bashrc
+# Replace X.Y.Z with the latest version number
+VERSION=vX.Y.Z
+mkdir kun_peng_$VERSION
+wget https://github.com/eric9n/Kun-peng/releases/download/$VERSION/kun_peng-$VERSION-centos7
+mv kun_peng-$VERSION-centos7 kun_peng_$VERSION/kun_peng
+chmod +x kun_peng_$VERSION/kun_peng
+# Add to PATH
+echo "export PATH=\$PATH:$PWD/kun_peng_$VERSION" >> ~/.bashrc
 source ~/.bashrc
 ```
 
+For macOS users:
+
+```bash
+# Replace X.Y.Z with the latest version number
+VERSION=vX.Y.Z
+mkdir kun_peng_$VERSION
+# For Intel Macs
+wget https://github.com/eric9n/Kun-peng/releases/download/$VERSION/kun_peng-$VERSION-x86_64-apple-darwin
+mv kun_peng-$VERSION-x86_64-apple-darwin kun_peng_$VERSION/kun_peng
+# For Apple Silicon Macs
+# wget https://github.com/eric9n/Kun-peng/releases/download/$VERSION/kun_peng-$VERSION-aarch64-apple-darwin
+# mv kun_peng-$VERSION-aarch64-apple-darwin kun_peng_$VERSION/kun_peng
+chmod +x kun_peng_$VERSION/kun_peng
+# Add to PATH
+echo "export PATH=\$PATH:$PWD/kun_peng_$VERSION" >> ~/.zshrc  # or ~/.bash_profile for Bash
+source ~/.zshrc  # or source ~/.bash_profile for Bash
+```
+
+For Windows users:
+
+```powershell
+# Replace X.Y.Z with the latest version number
+$VERSION = "vX.Y.Z"
+New-Item -ItemType Directory -Force -Path kun_peng_$VERSION
+Invoke-WebRequest -Uri "https://github.com/eric9n/Kun-peng/releases/download/$VERSION/kun_peng-$VERSION-x86_64-pc-windows-msvc.exe" -OutFile "kun_peng_$VERSION\kun_peng.exe"
+# Add to PATH
+$env:Path += ";$PWD\kun_peng_$VERSION"
+[Environment]::SetEnvironmentVariable("Path", $env:Path, [EnvironmentVariableTarget]::User)
+```
+
+After installation, you can verify the installation by running:
+
+```bash
+kun_peng --version
+```
+
 #### Run the `kun_peng` example
 
 We will use a very small virus database on the GitHub homepage as an example:
@@ -148,54 +190,31 @@ This output confirms that the `kun_peng` commands were executed successfully and
 
 ## ncbi_dl tool
 
+For detailed information and usage instructions for the ncbi_dl tool, please refer to the [ncbi_dl repository](https://github.com/eric9n/ncbi_dl.git).
 
-#### Run the `ncbi` Example
-
-Run the example script in the ncbi project to download the necessary files. Execute the following command from the root of the workspace:
+The ncbi_dl tool is used to download resources from the NCBI website, including taxonomy files and genome data. It provides a convenient way to obtain the necessary data for building Kun-peng databases.
 
-``` sh
-cargo run --release --example run_download --package ncbi_dl
-```
-
-This will run the run_download.rs example located in the ncbi project's examples directory. The script will:
-
-1.  Ensure the necessary directories exist.
-2.  Download the required files using the ncbi binary with the following commands:
+### Downloading Genome Databases
 
--   ./target/release/ncbi_dl -d downloads gen -g archaea
--   ./target/release/ncbi_dl -d downloads tax
+To download genome databases using ncbi_dl, you can use the `genomes` (or `gen`) command. Here's a basic example:
 
-Example Output You should see output similar to the following:
-
-``` txt
-Executing command: /path/to/workspace/target/release/ncbi_dl -d /path/to/workspace/downloads gen -g archaea
-NCBI binary output: [download output here]
-
-Executing command: /path/to/workspace/target/release/ncbi_dl -d /path/to/workspace/downloads tax
-NCBI binary output: [download output here]
+```sh
+ncbi_dl -d /path/to/download/directory gen -g bacteria
 ```
 
-The ncbi_dl binary is used to download resources from the NCBI website. Here is the help manual for the ncbi_dl binary:
+This command will download bacterial genomes to the specified directory. You can replace `bacteria` with other genome groups like `archaea`, `fungi`, `protozoa`, or `viral` depending on your needs.
 
-``` sh
-./target/release/ncbi_dl -h
-ncbi_dl download resource
+Some key options for the `genomes` command include:
 
-Usage: ncbi_dl [OPTIONS] <COMMAND>
+- `-g, --groups <GROUPS>`: Specify which genome groups to download (e.g., bacteria, archaea, viral)
+- `-f, --file-types <FILE_TYPES>`: Choose which file types to download (default is genomic.fna.gz)
+- `-l, --assembly-level <ASSEMBLY_LEVEL>`: Set the assembly level (e.g., complete, chromosome, scaffold, contig)
 
-Commands:
-  taxonomy  Download taxonomy files from NCBI (alias: tax)
-  genomes   Download genomes data from NCBI (alias: gen)
-  help      Print this message or the help of the given subcommand(s)
+For a full list of options and more detailed usage instructions, please refer to the ncbi_dl repository documentation.
 
-Options:
-  -d, --download-dir <DOWNLOAD_DIR>  Directory to store downloaded files [default: lib]
-  -n, --num-threads <NUM_THREADS>    Number of threads to use for downloading [default: 20]
-  -h, --help                         Print help (see more with '--help')
-  -V, --version                      Print version
-```
+For installation, additional usage examples, and more detailed documentation, please visit the ncbi_dl repository linked above.
 
-## kun_peng tool
+## kun_peng
 
 ``` sh
 Usage: kun_peng <COMMAND>