use conda.api instead of parallel calls to the conda binary (#4775)

* use conda.api instead of parallel calls to the conda binary

* don't select releases without release dates

* update the format to be wide enough to fit matplotlib-base

* don't verify the version using the filename

* filter invalid / missing dates before retrieving the metadata

Author: keewis

ci/min_deps_check.py

@@ -2,14 +2,15 @@
 publication date. Compare it against requirements/py36-min-all-deps.yml to verify the
 policy on obsolete dependencies is being followed. Print a pretty report :)
 """
-import subprocess
+import itertools
 import sys
-from concurrent.futures import ThreadPoolExecutor
 from datetime import datetime, timedelta
 from typing import Dict, Iterator, Optional, Tuple
 
+import conda.api
 import yaml
 
+CHANNELS = ["conda-forge", "defaults"]
 IGNORE_DEPS = {
     "black",
     "coveralls",
@@ -91,30 +92,23 @@ def query_conda(pkg: str) -> Dict[Tuple[int, int], datetime]:
 
     Return map of {(major version, minor version): publication date}
     """
-    stdout = subprocess.check_output(
-        ["conda", "search", pkg, "--info", "-c", "defaults", "-c", "conda-forge"]
-    )
-    out = {}  # type: Dict[Tuple[int, int], datetime]
-    major = None
-    minor = None
-
-    for row in stdout.decode("utf-8").splitlines():
-        label, _, value = row.partition(":")
-        label = label.strip()
-        if label == "file name":
-            value = value.strip()[len(pkg) :]
-            smajor, sminor = value.split("-")[1].split(".")[:2]
-            major = int(smajor)
-            minor = int(sminor)
-        if label == "timestamp":
-            assert major is not None
-            assert minor is not None
-            ts = datetime.strptime(value.split()[0].strip(), "%Y-%m-%d")
-
-            if (major, minor) in out:
-                out[major, minor] = min(out[major, minor], ts)
-            else:
-                out[major, minor] = ts
+
+    def metadata(entry):
+        version = entry.version
+
+        time = datetime.fromtimestamp(entry.timestamp)
+        major, minor = map(int, version.split(".")[:2])
+
+        return (major, minor), time
+
+    raw_data = conda.api.SubdirData.query_all(pkg, channels=CHANNELS)
+    data = sorted(metadata(entry) for entry in raw_data if entry.timestamp != 0)
+
+    release_dates = {
+        version: [time for _, time in group if time is not None]
+        for version, group in itertools.groupby(data, key=lambda x: x[0])
+    }
+    out = {version: min(dates) for version, dates in release_dates.items() if dates}
 
     # Hardcoded fix to work around incorrect dates in conda
     if pkg == "python":
@@ -202,16 +196,14 @@ def fmt_version(major: int, minor: int, patch: int = None) -> str:
 
 def main() -> None:
     fname = sys.argv[1]
-    with ThreadPoolExecutor(8) as ex:
-        futures = [
-            ex.submit(process_pkg, pkg, major, minor, patch)
-            for pkg, major, minor, patch in parse_requirements(fname)
-        ]
-        rows = [f.result() for f in futures]
-
-    print("Package       Required             Policy               Status")
-    print("------------- -------------------- -------------------- ------")
-    fmt = "{:13} {:7} ({:10}) {:7} ({:10}) {}"
+    rows = [
+        process_pkg(pkg, major, minor, patch)
+        for pkg, major, minor, patch in parse_requirements(fname)
+    ]
+
+    print("Package           Required             Policy               Status")
+    print("----------------- -------------------- -------------------- ------")
+    fmt = "{:17} {:7} ({:10}) {:7} ({:10}) {}"
     for row in rows:
         print(fmt.format(*row))