python-materialsdb 0.3.0__tar.gz → 0.3.1__tar.gz

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-materialsdb
-Version: 0.3.0
+Version: 0.3.1
 Summary: A library to work with materialsdb.org open standard for building materials.
 Author-email: Cyril Waechter <cyrwae@hotmail.com>
 License-Expression: GPL-3.0-or-later

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "python-materialsdb"
-version = "0.3.0"
+version = "0.3.1"
 description = "A library to work with materialsdb.org open standard for building materials."
 readme = "README.md"
 license = "GPL-3.0-or-later"

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/src/materialsdb/cache.py

@@ -2,6 +2,7 @@ import os
 import pathlib
 import urllib.request
 from collections import namedtuple
+from concurrent.futures import FIRST_COMPLETED, ThreadPoolExecutor, wait
 
 from lxml import etree
 
@@ -85,25 +86,77 @@ def _plan_index(index):
     return _IndexPlan(index, new_index, unchanged, todos)
 
 
-def update_producers_data(url_list=MATERIALSDBINDEXURLLIST, on_progress=None):
-    """on_progress(done, total, name) is called after each producer download
-    with a global running total; returning a truthy value stops the update
-    between two downloads (the partial Report is returned as-is)."""
+def update_producers_data(url_list=MATERIALSDBINDEXURLLIST, on_progress=None, max_workers: int = 8):
+    """Update the cached producer files from the index URLs.
+
+    Downloads run on a bounded thread pool (max_workers). on_progress(done,
+    total, name) is called once per completed download from the calling
+    thread with a monotonic running total; returning a truthy value stops
+    NEW submissions (queued jobs are abandoned, in-flight ones finish but
+    no longer count toward the partial Report). A download error stops the
+    run and the first exception is re-raised after the pool drains.
+    """
     plans = [_plan_index(index) for index in url_list]
-    total = sum(len(plan.todos) for plan in plans)
-    existing, updated, deleted = [], [], []
+    jobs = [
+        (plan, company, cached_producer, producer_path)
+        for plan in plans
+        for (company, cached_producer, producer_path) in plan.todos
+    ]
+    total = len(jobs)
+    workers = max(1, min(int(max_workers), total or 1))
+
+    existing: list = []
+    updated: list = []
+    deleted: list = []
     done = 0
+    cancel_requested = False
+    failure = None
+
+    def _download(job):
+        _, company, cached_producer, producer_path = job
+        deleted_path = None
+        if cached_producer is not None:
+            deleted_path = get_producers_dir() / pathlib.Path(cached_producer.get("href")).name
+            deleted_path.unlink(True)
+        urllib.request.urlretrieve(company.get("href"), producer_path)
+        return deleted_path
+
+    pool = ThreadPoolExecutor(max_workers=workers)
+    pending = list(jobs)  # pops from the FRONT, submission order preserved
+    futures: dict = {}
+    try:
+        while futures or pending:
+            while pending and len(futures) < workers:
+                job = pending.pop(0)
+                futures[pool.submit(_download, job)] = job
+            if not futures:
+                break
+            ready, _ = wait(set(futures), return_when=FIRST_COMPLETED)
+            for future in ready:
+                job = futures.pop(future)
+                try:
+                    deleted_path = future.result()
+                except Exception as err:  # noqa: BLE001 - re-raised below, caller decides
+                    if failure is None:
+                        failure = err
+                    break
+                if deleted_path is not None:
+                    deleted.append(deleted_path)
+                updated.append(job[3])
+                done += 1
+                if on_progress is not None and on_progress(done, total, pathlib.Path(job[3]).name):
+                    cancel_requested = True
+                    break
+            if cancel_requested or failure is not None:
+                break
+    finally:
+        pool.shutdown(wait=True, cancel_futures=True)
+
+    if failure is not None:
+        raise failure
+    if cancel_requested:
+        return Report(existing, updated, deleted)
     for plan in plans:
-        for company, cached_producer, producer_path in plan.todos:
-            if cached_producer is not None:
-                cached_path = get_producers_dir() / pathlib.Path(cached_producer.get("href")).name
-                deleted.append(cached_path)
-                cached_path.unlink(True)
-            urllib.request.urlretrieve(company.get("href"), producer_path)
-            updated.append(producer_path)
-            done += 1
-            if on_progress is not None and on_progress(done, total, pathlib.Path(company.get("href")).name):
-                return Report(existing, updated, deleted)
         if plan.todos:
             plan.new_index.write(str(get_cached_index_path(plan.index)))
         existing.extend(plan.unchanged)

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/src/materialsdb/serialiser.py

@@ -25,9 +25,31 @@ def get_xml_schema() -> str:
     return str(Path(__file__).parent / "schema/materialsdb103.xsd")
 
 
+@cache
+def cached_type_hints(cls) -> dict:
+    return typing.get_type_hints(cls)
+
+
+@cache
+def _tag_local_name(tag: str) -> str:
+    m = re.search("{.*}(.*)", tag)
+    return m.group(1) if m else tag
+
+
 def get_element_name(element: objectify.ObjectifiedElement) -> str:
-    m = re.search("{.*}(.*)", element.tag)
-    return m.group(1) if m else element.tag
+    return _tag_local_name(element.tag)
+
+
+@cache
+def _is_optional(type_hint):
+    return typing.get_origin(type_hint) is typing.Union and typing.get_args(type_hint)[1] is type(None)
+
+
+@cache
+def _strip_optional(type_hint):
+    if _is_optional(type_hint):
+        return typing.get_args(type_hint)[0]
+    return type_hint
 
 
 def create_element_maker():
@@ -49,11 +71,6 @@ def get_valid_root(tree: objectify.ObjectifiedElement) -> objectify.ObjectifiedE
     return root
 
 
-@cache
-def cached_type_hints(cls) -> dict:
-    return typing.get_type_hints(cls)
-
-
 class XmlDeserialiser:
     def __init__(self):
         self.schema = etree.XMLSchema(file=get_xml_schema())
@@ -93,7 +110,7 @@ class XmlDeserialiser:
             value = element.get(attrib)
             if value is None:
                 continue
-            base_class = self.strip_optional(type_hints[attrib])
+            base_class = _strip_optional(type_hints[attrib])
 
             kwargs[attrib] = base_class(value)
         for child_name in getattr(element_class, "xml_elements", ()):
@@ -127,12 +144,10 @@ class XmlDeserialiser:
         return instance
 
     def strip_optional(self, type_hint):
-        if self.is_optional(type_hint):
-            return typing.get_args(type_hint)[0]
-        return type_hint
+        return _strip_optional(type_hint)
 
     def is_optional(self, type_hint):
-        return typing.get_origin(type_hint) is typing.Union and typing.get_args(type_hint)[1] is type(None)
+        return _is_optional(type_hint)
 
     def cls_name(self, name: str) -> str:
         return name if name[0].isupper() else name.title()

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/src/python_materialsdb.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-materialsdb
-Version: 0.3.0
+Version: 0.3.1
 Summary: A library to work with materialsdb.org open standard for building materials.
 Author-email: Cyril Waechter <cyrwae@hotmail.com>
 License-Expression: GPL-3.0-or-later

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/tests/test_build_bonsai_addon.py

@@ -1,3 +1,4 @@
+import re
 import sys
 import zipfile
 from pathlib import Path
@@ -8,13 +9,22 @@ sys.path.insert(0, str(ROOT / "dev_utils"))
 import build_bonsai_addon  # ty: ignore[unresolved-import] - dev_util on a side path
 
 
+def _manifest_version():
+    text = (ROOT / "bonsai_addon" / "blender_manifest.toml").read_text(encoding="utf-8")
+    match = re.search(r'(?m)^version = "([^"]+)"$', text)
+    assert match is not None
+    return match.group(1)
+
+
 def test_default_build_unchanged(tmp_path):
+    manifest_version = _manifest_version()
+
     out = build_bonsai_addon.build(out_dir=tmp_path)
 
     assert out.name == "materialsdb_listener.zip"
     with zipfile.ZipFile(out) as zf:
         assert "blender_manifest.toml" in zf.namelist()
-        assert 'version = "0.2.0"' in zf.read("blender_manifest.toml").decode("utf-8")
+        assert f'version = "{manifest_version}"' in zf.read("blender_manifest.toml").decode("utf-8")
         assert "blender_manifest.toml".replace(".", "/") not in zf.namelist()  # manifest at zip ROOT
 
 
@@ -30,8 +40,12 @@ def test_version_injected_into_manifest_and_name(tmp_path):
 def test_version_injection_is_literal_independent(tmp_path):
     """The injection must not depend on the manifest's current version
     literal (it changes on every release bump)."""
+
+    current = _manifest_version()
+    stale = "0.0.1" if current != "0.0.1" else "0.0.2"
+
     out = build_bonsai_addon.build(version="3.2.1", out_dir=tmp_path)
 
     with zipfile.ZipFile(out) as zf:
         assert 'version = "3.2.1"' in zf.read("blender_manifest.toml").decode("utf-8")
-        assert 'version = "0.2.0"' not in zf.read("blender_manifest.toml").decode("utf-8")
+        assert f'version = "{stale}"' not in zf.read("blender_manifest.toml").decode("utf-8")

{python_materialsdb-0.3.0 → python_materialsdb-0.3.1}/tests/test_cache.py

@@ -33,7 +33,7 @@ def cache_env(tmp_path, monkeypatch):
     # defaults are bound at def time, so patch __defaults__ too
     single = [cache.MATERIALSDBINDEXURLLIST[0]]
     monkeypatch.setattr(cache, "MATERIALSDBINDEXURLLIST", single)
-    monkeypatch.setattr(cache.update_producers_data, "__defaults__", (single, None))
+    monkeypatch.setattr(cache.update_producers_data, "__defaults__", (single, None, 8))
     monkeypatch.setattr(cache.updates_available, "__defaults__", (single,))
     recorded: _Recorded = {"urls": [], "dests": [], "index": INDEX_XML}
 
@@ -57,9 +57,11 @@ def test_first_refresh_downloads_all_with_global_total(cache_env):
         on_progress=lambda done, total, name: progress.append((done, total, name)) or False
     )
 
-    assert recorded["urls"] == ["https://x.org/a.xml", "https://x.org/b.xml"]
-    assert progress == [(1, 2, "a.xml"), (2, 2, "b.xml")]
-    assert [p.name for p in report.updated] == ["a.xml", "b.xml"]
+    assert set(recorded["urls"]) == {"https://x.org/a.xml", "https://x.org/b.xml"}
+    assert [p[0] for p in progress] == [1, 2]  # monotonic done
+    assert {p[1] for p in progress} == {2}  # global total reached before any download finished
+    assert {p[2] for p in progress} == {"a.xml", "b.xml"}
+    assert {p.name for p in report.updated} == {"a.xml", "b.xml"}
     assert report.existing == []
     assert (cache_dir / "Producers" / "a.xml").exists()  # downloads land in Producers dir
     assert (cache_dir / "ProducerIndex.xml").exists()  # new index persisted
@@ -68,7 +70,7 @@ def test_first_refresh_downloads_all_with_global_total(cache_env):
 def test_second_refresh_uses_existing_files(cache_env):
     cache, _, recorded = cache_env
     report = cache.update_producers_data()
-    assert [p.name for p in report.updated] == ["a.xml", "b.xml"]
+    assert {p.name for p in report.updated} == {"a.xml", "b.xml"}
     recorded["urls"].clear()
 
     progress = []
@@ -100,9 +102,9 @@ def test_outdated_single_producer_updates_with_total_one(cache_env):
         on_progress=lambda done, total, name: progress.append((done, total, name)) or False
     )
 
-    assert recorded["urls"] == ["https://x.org/a.xml"]
+    assert recorded["urls"] == ["https://x.org/a.xml"]  # single job — deterministic
     assert progress == [(1, 1, "a.xml")]
-    assert [p.name for p in record.updated] == ["a.xml"]
+    assert {p.name for p in record.updated} == {"a.xml"}
     assert [p.name for p in record.deleted] == ["a.xml"]  # stale file removed before re-download
     assert [p.name for p in record.existing] == ["b.xml"]
 
@@ -128,7 +130,8 @@ def test_cancel_stops_between_downloads(cache_env):
     def cancel_after_first(done, total, name):
         return done >= 1  # truthy → abort
 
-    report = cache.update_producers_data(on_progress=cancel_after_first)
+    # single worker: b.html cannot start before the callback runs
+    report = cache.update_producers_data(on_progress=cancel_after_first, max_workers=1)
 
     assert recorded["urls"] == ["https://x.org/a.xml"]  # second download never starts
     assert [p.name for p in report.updated] == ["a.xml"]