relaxed python dependency, added headers for C/C++ execution of onion

.github/workflows/ci.yml

@@ -55,8 +55,10 @@ jobs:
         include:
           - os: ubuntu-latest
             artifact_name: onion-linux-x86_64
-          - os: macos-latest
+          - os: macos-13
             artifact_name: onion-darwin-x86_64
+          - os: macos-14
+            artifact_name: onion-darwin-arm64
 
     steps:
     - uses: actions/checkout@v4

.github/workflows/release.yml

@@ -17,8 +17,10 @@ jobs:
         include:
           - os: ubuntu-latest
             artifact_name: onion-linux-x86_64
-          - os: macos-latest
+          - os: macos-13
             artifact_name: onion-darwin-x86_64
+          - os: macos-14
+            artifact_name: onion-darwin-arm64
 
     steps:
     - uses: actions/checkout@v4
@@ -55,8 +57,10 @@ jobs:
         include:
           - os: ubuntu-latest
             onion_artifact: onion-linux-x86_64
-          - os: macos-latest
+          - os: macos-13
             onion_artifact: onion-darwin-x86_64
+          - os: macos-14
+            onion_artifact: onion-darwin-arm64
           - os: windows-latest
             onion_artifact: none
 
@@ -141,10 +145,11 @@ jobs:
     environment: release
 
     steps:
-    # Only use Linux wheel to avoid conflicts
+    # Download all platform-specific wheels
     - uses: actions/download-artifact@v4
       with:
-        name: wheels-ubuntu-latest
+        pattern: wheels-*
+        merge-multiple: true
         path: dist/
 
     - uses: actions/download-artifact@v4

dalla_data_processing/deduplication/onion/src/Makefile

@@ -1,7 +1,7 @@
 include ../Makefile.config
 
 CC=g++
-CFLAGS=-Wall -O3
+CFLAGS=-Wall -O3 -std=c++11 -I/opt/homebrew/opt/google-sparsehash/include
 
 OBJS=version.o buzhash.o
 TARGETS=hashgen hashdup onion

dalla_data_processing/deduplication/onion/src/Makefile.g

@@ -2,7 +2,7 @@ include ../Makefile.config
 
 CC=g++
 #CFLAGS=-Wall -O3
-CFLAGS=-Wall -g
+CFLAGS=-Wall -g -std=c++11 -I/opt/homebrew/opt/google-sparsehash/include
 
 OBJS=version.o buzhash.o
 TARGETS=hashgen hashdup onion

dalla_data_processing/deduplication/onion/src_sc/Makefile

@@ -1,7 +1,7 @@
 include ../Makefile.config
 
 CC=g++
-CFLAGS=-Wall -O3
+CFLAGS=-Wall -O3 -std=c++11 -I/opt/homebrew/opt/google-sparsehash/include
 
 OBJS=version.o buzhash.o
 TARGETS=hashgen hashdup onion

dalla_data_processing/deduplication/onion/src_sc/Makefile.g

@@ -2,7 +2,7 @@ include ../Makefile.config
 
 CC=g++
 #CFLAGS=-Wall -O3
-CFLAGS=-Wall -g
+CFLAGS=-Wall -g -std=c++11 -I/opt/homebrew/opt/google-sparsehash/include
 
 OBJS=version.o buzhash.o
 TARGETS=hashgen hashdup onion

pyproject.toml

@@ -11,13 +11,16 @@ authors = [
     {name = "Digital Research Unit - Arab Center", email = "dru@dohainstitute.edu.qa"}
 ]
 readme = "README.md"
-requires-python = ">=3.12,<3.13"
+requires-python = ">=3.10"
 keywords = ["arabic", "nlp", "data-processing", "deduplication", "stemming", "readability", "quality"]
 classifiers = [
     "Intended Audience :: Developers",
     "Intended Audience :: Science/Research",
     "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "Topic :: Scientific/Engineering :: Artificial Intelligence",
     "Topic :: Text Processing :: Linguistic",
 ]
@@ -83,7 +86,7 @@ dalla_data_processing = ["py.typed"]
 
 [tool.ruff]
 line-length = 100
-target-version = "py312"
+target-version = "py310"
 src = ["dalla_data_processing"]
 
 [tool.ruff.lint]

scripts/build_onion.sh

@@ -63,14 +63,16 @@ if [[ "$OSTYPE" == "darwin"* ]]; then
     # On macOS, add Homebrew paths for sparsehash
     if command -v brew &> /dev/null; then
         BREW_PREFIX=$(brew --prefix)
-        EXTRA_CFLAGS="-I${BREW_PREFIX}/include"
+        SPARSEHASH_PREFIX=$(brew --prefix google-sparsehash 2>/dev/null || echo "${BREW_PREFIX}")
+        EXTRA_CFLAGS="-I${SPARSEHASH_PREFIX}/include"
         echo -e "${YELLOW}Using Homebrew prefix: ${BREW_PREFIX}${NC}"
+        echo -e "${YELLOW}Using sparsehash include: ${SPARSEHASH_PREFIX}/include${NC}"
     fi
 fi
 
 # Build onion
 echo -e "${YELLOW}Compiling Onion...${NC}"
-if make CFLAGS="-Wall -O3 ${EXTRA_CFLAGS}"; then
+if make CFLAGS="-Wall -O3 -std=c++11 ${EXTRA_CFLAGS}"; then
     echo -e "${GREEN}✓ Compilation successful${NC}"
 else
     echo -e "${RED}✗ Compilation failed${NC}"