Remove magic for logdir; adapt merge_vocab for split SKOS

CHANGELOG.md

@@ -1,5 +1,19 @@
 # Change log
 
+## Release 0.6.x (2023-08-dd)
+
+New features:
+
+- The `merge_vocab` script gained support for directories containing vocabularies created with `voc4cat transform --split`. This feature is used in gh-actions of [voc4cat-template](https://github.com/nfdi4cat/voc4cat-template)-based vocabularies.
+
+Changes:
+
+- The log file will now always be written to the given directory. Previously the log file directory depended on the presence of the `--outdir` option.
+
+Bug fixes:
+
+- Sub-command `voc4cat docs` failed if `--outdir` was not given.
+
 ## Release 0.6.0 (2023-08-09)
 
 New features:

src/voc4cat/cli.py

@@ -34,9 +34,7 @@ def process_common_options(args):
     if logfile is None:
         setup_logging(loglevel)
     else:
-        if outdir is not None:
-            logfile = Path(outdir) / logfile
-        elif not logfile.parents[0].exists():
+        if not logfile.parents[0].exists():
             logfile.parents[0].mkdir(exist_ok=True, parents=True)
         setup_logging(loglevel, logfile)
 

src/voc4cat/merge_vocab.py

@@ -1,7 +1,6 @@
 # This script is mainly useful for CI.
 import argparse
 import logging
-import os
 import shutil
 import subprocess
 import sys
@@ -19,12 +18,18 @@ def main(ttl_inbox: Path, vocab: Path) -> int:
     New files are copied, existing files are synced by git merge-file.
     """
     retcode = 0
-    for p in os.listdir(ttl_inbox):
-        new = ttl_inbox / Path(p)
-        if new.suffix != ".ttl" or new.is_dir():
-            logger.info('Skipping "%s"', new)
+    for new in ttl_inbox.iterdir():
+        if new.is_dir():
+            logger.debug('Entering directory "%s"', new)
+            (vocab / new.name).mkdir(exist_ok=True)
+            retcode = main(new, vocab / new.name)
+            if retcode != 0:
+                break
+            continue
+        if new.suffix != ".ttl":
+            logger.debug('Skipping "%s"', new)
             continue
-        if os.path.exists(vocab / Path(new).name):
+        if (vocab / Path(new).name).exists():
             exists = vocab / Path(new).name
             cmd = ["git", "merge-file", "--theirs", str(exists), str(exists), str(new)]
             logger.info("Running cmd: %s", " ".join(cmd))
@@ -64,7 +69,6 @@ def main_cli(args=None) -> int:
         setup_logging()
     else:
         outbox.mkdir(exist_ok=True, parents=True)
-        logfile = Path(outbox) / logfile
         setup_logging(logfile=logfile)
 
     if not has_outbox or not vocab.exists():

tests/test_convert.py

@@ -45,7 +45,7 @@ def test_run_vocexcel_outputdir(monkeypatch, datadir, tmp_path, outputdir, testf
     shutil.copy(datadir / testfile, tmp_path)
     monkeypatch.chdir(tmp_path)
     # Check if log is placed in out folder.
-    log = "test-run.log"
+    log = Path(outputdir) / "test-run.log"
     main_cli(
         ["convert", "--logfile", str(log)]
         + (["--outdir", str(outputdir)] if outputdir else [])
@@ -56,7 +56,7 @@ def test_run_vocexcel_outputdir(monkeypatch, datadir, tmp_path, outputdir, testf
         assert (outdir / testfile).with_suffix(".ttl").exists()
     else:
         assert (outdir / testfile).with_suffix(".xlsx").exists()
-    assert (outdir / log).exists()
+    assert (outdir / log.name).exists()
 
 
 def test_duplicates(datadir, tmp_path, caplog):

tests/test_merge_vocab.py

@@ -3,7 +3,7 @@
 from unittest import mock
 
 import pytest
-from test_cli import CS_CYCLES_TURTLE
+from test_cli import CS_CYCLES_TURTLE, CS_SIMPLE_TURTLE
 from voc4cat.merge_vocab import main_cli
 
 
@@ -36,17 +36,40 @@ def test_main_merge_dirs(datadir, tmp_path, caplog):
     vocab.mkdir()
     ttl_inbox = tmp_path / "ttl_inbox"
     ttl_inbox.mkdir()
-    extra = ttl_inbox / "extra"
-    extra.mkdir()
+    extra = ttl_inbox / "extra.txt"
+    extra.touch()
+    splitvoc = ttl_inbox / "splitvoc"
+    splitvoc.mkdir()
+
     shutil.copy(datadir / CS_CYCLES_TURTLE, ttl_inbox / CS_CYCLES_TURTLE)
+    shutil.copy(datadir / CS_SIMPLE_TURTLE, splitvoc)
 
-    with caplog.at_level(logging.INFO):
+    with caplog.at_level(logging.DEBUG):
         exit_code = main_cli([str(ttl_inbox), str(vocab)])
     assert f'Skipping "{extra}"' in caplog.text
     assert (vocab / CS_CYCLES_TURTLE).exists()
+    assert (vocab / "splitvoc" / CS_SIMPLE_TURTLE).exists()
     assert exit_code == 0
 
 
+def test_main_merge_split_vocab_dir(datadir, tmp_path, caplog):
+    """Check merge of dir with split vocab."""
+    vocab = tmp_path / "vocab"
+    (vocab / "splitvoc").mkdir(parents=True)
+    ttl_inbox = tmp_path / "ttl_inbox"
+    (ttl_inbox / "splitvoc").mkdir(parents=True)
+    shutil.copy(datadir / CS_SIMPLE_TURTLE, ttl_inbox / "splitvoc")
+    shutil.copy(datadir / CS_SIMPLE_TURTLE, vocab / "splitvoc")
+
+    with caplog.at_level(logging.DEBUG), mock.patch(
+        "voc4cat.merge_vocab.subprocess"
+    ) as subprocess:
+        subprocess.Popen.return_value.returncode = 1
+        exit_code = main_cli([str(ttl_inbox), str(vocab)])
+    assert "Entering directory" in caplog.text
+    assert exit_code == 1
+
+
 def test_main_merge_files(datadir, tmp_path, caplog):
     """Check merge that merges the content of files."""
     vocab = tmp_path / "vocab"