Fix issue on linux where if a stage takes to long to start and previous stage exit the processing fails

.github/workflows/python-package.yml

@@ -12,14 +12,14 @@ on:
 jobs:
   build:
 
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-24.04
     strategy:
       matrix:
-        python-version: ["3.7", "3.9", "3.11"]
+        python-version: ["3.8", "3.9", "3.11"]
     steps:
     - uses: actions/checkout@v2
     - name: Set up Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.python-version }}
     - name: Install dependencies
@@ -34,8 +34,8 @@ jobs:
       run: |
         python setup.py bdist_wheel
     - name: Archive artifact
-      uses: actions/upload-artifact@v2
+      uses: actions/upload-artifact@v4
       with:
-        name: dist
+        name: dist-${{ matrix.python-version }}
         path: |
           dist

parallelpipe.py

@@ -5,10 +5,8 @@
          \ map /
 """
 from multiprocessing import Process, Queue
-from time import sleep
-import inspect
-import collections
 from collections.abc import Iterable
+import time
 
 import dill
 
@@ -87,6 +85,11 @@ def run(self):
             for i in range(self._num_followers):
                 put_item(EXIT)
             self._que_err.put(EXIT)
+            while not self._que_out.empty():
+                time.sleep(0.1)
+            while not self._que_err.empty():
+                time.sleep(0.1)
+
 
 class Stage(object):
     """Represent a pool of parallel tasks that perform the same type of action on the input."""

setup.py

@@ -12,7 +12,7 @@
 
 setup(
     name='parallelpipe',
-    version='0.3.0',
+    version='0.3.1',
     author='Giuseppe Tribulato',
     author_email='gtsystem@gmail.com',
     py_modules=['parallelpipe'],

tests/test_parallelpipe.py

@@ -1,5 +1,6 @@
 import unittest
 from parallelpipe import Stage, TaskException, stage, map_stage
+from time import sleep
 
 def t1(x, fail_at=None):
     """Produce values from the given input iterator.
@@ -124,15 +125,15 @@ def test_exception_propagation(self):
         reducer = Stage(t3, sum).setup(workers=2, qsize=3)
         pipe = producer | mapper | reducer
 
-        with self.assertRaisesRegexp(TaskException, "failed at 200"):
+        with self.assertRaisesRegex(TaskException, "failed at 200"):
             for res in pipe.results():
                 pass
 
 
         producer = Stage(t1, range(1000), 10).setup(workers=2, qsize=10)
         pipe = producer | mapper | reducer
 
-        with self.assertRaisesRegexp(TaskException, "failed at 10"):
+        with self.assertRaisesRegex(TaskException, "failed at 10"):
             for res in pipe.results():
                 pass
 
@@ -178,10 +179,37 @@ def consume(n):
         res = (range(1000) | fail2(5) | consume).execute()
         self.assertEqual(res, 10)
 
-        with self.assertRaisesRegexp(TaskException, "failure"):
+        with self.assertRaisesRegex(TaskException, "failure"):
             (range(1000) | fail1(5) | consume).execute()
+
+    def test_slow_second_stage(self):
+
+        @stage(workers=2)
+        def mapit(it):
+            for item in it:
+                yield item + 1
 
+        @stage(workers=1)
+        def reduce(it):
+            sleep(3)  # simulate a long startup time
+            tot = 0
+            for item in it:
+                tot += item
+                sleep(2)
+                yield 5
+            yield tot
+
+        @stage(workers=2)
+        def write(it):
+            for item in it:
+                yield item
+
+        res = list(([1] | mapit | reduce | write).results())
+        self.assertEqual(res, [5, 2])
+
 
 if __name__ == '__main__':
+    import multiprocessing
+    multiprocessing.set_start_method('fork', force=False)
     unittest.main()