[pull] main from firecrawl:main

.github/scripts/check_version_has_incremented.py

@@ -8,11 +8,16 @@
 Published version: 0.0.21
 true
 
-python .github/scripts/check_version_has_incremented.py python ./apps/python-sdk/firecrawl firecrawl-py 
+python .github/scripts/check_version_has_incremented.py python ./apps/python-sdk/firecrawl firecrawl-py
 Local version: 0.0.11
 Published version: 0.0.11
 false
 
+python .github/scripts/check_version_has_incremented.py java ./apps/java-sdk com.firecrawl:firecrawl-java
+Local version: 1.0.0
+Published version: 0.0.0  (0.0.0 means not yet published on Maven Central)
+true
+
 """
 import json
 import os
@@ -53,6 +58,30 @@ def get_npm_version(package_name: str) -> str:
     version = response.json()['version']
     return version.strip()
 
+def get_gradle_version(file_path: str) -> str:
+    """Extract version string from build.gradle.kts."""
+    build_file = Path(file_path).read_text()
+    version_match = re.search(r'^version\s*=\s*["\']([^"\']*)["\']', build_file, re.M)
+    if version_match:
+        return version_match.group(1).strip()
+    raise RuntimeError("Unable to find version string in build.gradle.kts.")
+
+def get_maven_central_version(package_name: str) -> str:
+    """Get latest version of Java package from Maven Central. package_name should be groupId:artifactId."""
+    group_id, artifact_id = package_name.split(":")
+    group_path = group_id.replace(".", "/")
+    url = f"https://repo1.maven.org/maven2/{group_path}/{artifact_id}/maven-metadata.xml"
+    response = requests.get(url)
+    if response.status_code == 404:
+        return "0.0.0"
+    response.raise_for_status()
+    version_match = re.search(r"<release>(.*?)</release>", response.text)
+    if not version_match:
+        version_match = re.search(r"<latest>(.*?)</latest>", response.text)
+    if version_match:
+        return version_match.group(1).strip()
+    return "0.0.0"
+
 # def get_rust_version(file_path: str) -> str:
 #     """Extract version string from Cargo.toml."""
 #     cargo_toml = toml.load(file_path)
@@ -87,14 +116,19 @@ def is_version_incremented(local_version: str, published_version: str) -> bool:
         current_version = get_js_version(os.path.join(package_path, 'package.json'))
         # Get published version from npm
         published_version = get_npm_version(package_name)
+    elif package_type == "java":
+        # Get current version from build.gradle.kts
+        current_version = get_gradle_version(os.path.join(package_path, 'build.gradle.kts'))
+        # Get published version from Maven Central
+        published_version = get_maven_central_version(package_name)
     # if package_type == "rust":
     #     # Get current version from Cargo.toml
     #     current_version = get_rust_version(os.path.join(package_path, 'Cargo.toml'))
     #     # Get published version from crates.io
     #     published_version = get_crates_version(package_name)
 
     else:
-        raise ValueError("Invalid package type. Use 'python' or 'js'.")
+        raise ValueError("Invalid package type. Use 'python', 'js', or 'java'.")
 
     # Print versions for debugging
     # print(f"Local version: {current_version}")

.github/workflows/publish-java-sdk.yml

@@ -0,0 +1,69 @@
+name: Publish Java SDK
+
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - main
+    paths:
+      - 'apps/java-sdk/**'
+
+jobs:
+  publish:
+    name: Publish to Maven Central
+    runs-on: blacksmith-2vcpu-ubuntu-2404
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v5
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.x'
+
+      - name: Install version check dependencies
+        run: pip install requests packaging
+
+      - name: Run version check script
+        id: version_check
+        run: |
+          VERSION_INCREMENTED=$(python .github/scripts/check_version_has_incremented.py java ./apps/java-sdk com.firecrawl:firecrawl-java)
+          echo "VERSION_INCREMENTED=$VERSION_INCREMENTED" >> $GITHUB_ENV
+
+      - name: Set up JDK 11
+        if: ${{ env.VERSION_INCREMENTED == 'true' }}
+        uses: actions/setup-java@v4
+        with:
+          distribution: temurin
+          java-version: '11'
+
+      - name: Cache Gradle packages
+        if: ${{ env.VERSION_INCREMENTED == 'true' }}
+        uses: actions/cache@v4
+        with:
+          path: |
+            ~/.gradle/caches
+            ~/.gradle/wrapper
+          key: ${{ runner.os }}-gradle-${{ hashFiles('apps/java-sdk/**/*.gradle.kts', 'apps/java-sdk/gradle/wrapper/gradle-wrapper.properties') }}
+          restore-keys: |
+            ${{ runner.os }}-gradle-
+
+      - name: Grant execute permission for gradlew
+        if: ${{ env.VERSION_INCREMENTED == 'true' }}
+        working-directory: ./apps/java-sdk
+        run: chmod +x gradlew
+
+      - name: Build
+        if: ${{ env.VERSION_INCREMENTED == 'true' }}
+        working-directory: ./apps/java-sdk
+        run: ./gradlew build -x test
+
+      - name: Publish to Maven Central
+        if: ${{ env.VERSION_INCREMENTED == 'true' }}
+        working-directory: ./apps/java-sdk
+        env:
+          ORG_GRADLE_PROJECT_mavenCentralUsername: ${{ secrets.MAVEN_USERNAME }}
+          ORG_GRADLE_PROJECT_mavenCentralPassword: ${{ secrets.MAVEN_PASSWORD }}
+          ORG_GRADLE_PROJECT_signingInMemoryKey: ${{ secrets.GPG_SIGNING_KEY }}
+          ORG_GRADLE_PROJECT_signingInMemoryKeyPassword: ${{ secrets.GPG_SIGNING_PASSWORD }}
+        run: ./gradlew publishAndReleaseToMavenCentral --no-configuration-cache

apps/java-sdk/README.md

@@ -339,7 +339,7 @@ FirecrawlClient client = FirecrawlClient.builder()
 
 ```bash
 # Clone the repository (if you haven't already)
-git clone https://github.com/mendableai/firecrawl.git
+git clone https://github.com/firecrawl/firecrawl.git
 cd firecrawl/apps/java-sdk
 
 # Build the project

apps/java-sdk/build.gradle.kts

@@ -1,6 +1,6 @@
 plugins {
     `java-library`
-    `maven-publish`
+    id("com.vanniktech.maven.publish") version "0.30.0"
 }
 
 group = "com.firecrawl"
@@ -9,8 +9,6 @@ version = "1.0.0"
 java {
     sourceCompatibility = JavaVersion.VERSION_11
     targetCompatibility = JavaVersion.VERSION_11
-    withSourcesJar()
-    withJavadocJar()
 }
 
 repositories {
@@ -39,29 +37,35 @@ tasks.withType<Javadoc> {
     }
 }
 
-publishing {
-    publications {
-        create<MavenPublication>("mavenJava") {
-            from(components["java"])
+mavenPublishing {
+    publishToMavenCentral(com.vanniktech.maven.publish.SonatypeHost.CENTRAL_PORTAL)
+    signAllPublications()
 
-            pom {
-                name.set("Firecrawl Java SDK")
-                description.set("Java SDK for the Firecrawl web scraping API")
-                url.set("https://github.com/mendableai/firecrawl")
+    coordinates("com.firecrawl", "firecrawl-java", version.toString())
 
-                licenses {
-                    license {
-                        name.set("MIT License")
-                        url.set("https://opensource.org/licenses/MIT")
-                    }
-                }
+    pom {
+        name.set("Firecrawl Java SDK")
+        description.set("Java SDK for the Firecrawl web scraping API")
+        url.set("https://github.com/firecrawl/firecrawl")
 
-                scm {
-                    url.set("https://github.com/mendableai/firecrawl")
-                    connection.set("scm:git:git://github.com/mendableai/firecrawl.git")
-                    developerConnection.set("scm:git:ssh://github.com/mendableai/firecrawl.git")
-                }
+        licenses {
+            license {
+                name.set("MIT License")
+                url.set("https://opensource.org/licenses/MIT")
             }
         }
+
+        developers {
+            developer {
+                name.set("Firecrawl")
+                url.set("https://firecrawl.dev")
+            }
+        }
+
+        scm {
+            url.set("https://github.com/firecrawl/firecrawl")
+            connection.set("scm:git:git://github.com/firecrawl/firecrawl.git")
+            developerConnection.set("scm:git:ssh://github.com/firecrawl/firecrawl.git")
+        }
     }
 }