Implement a 'take' option.

doc/configuration.md

@@ -51,6 +51,14 @@ There are several configuration sections in the file:
   slash) from the sources root. Supports asterisk-style wildcards.
   Example: `['/src/tests/*','/data/Test_*.lut']`
 
+* **take**: a list (enclosed in square brackets) of items to include in
+  the export. Items are specified as a path (including leading slash)
+  from the sources root. Supports asterisk-style wildcards. Example:
+  `['/src/tests/*','/data/Test_*.lut']`. If this list is empty, all
+  items will be included. If an item is present in both the `skip` and
+  `take` lists (or the latter is empty), the `skip` list takes
+  precedence.
+
 ## Section CSP
 
 (This section and the CSP.parsers sections are only needed if CSP files

src/config.py

@@ -55,16 +55,17 @@ def get_config() -> ns.Namespace:
     # Make sure configuration is complete
     check(config)
 
-    # Converts specification(s) of files to skip to regexes
-    config.skip_regexes = []
-    for spec in config.Source.skip:
-        spec = spec.replace('\\', '\\\\')
-        spec = spec.replace('.', '\\.')
-        # Create valid regex for star
-        spec = spec.replace('*', '.*')
-        regex = re.compile(spec, re.I)
-        config.skip_regexes.append(regex)
-
+    # # Converts specification(s) of files to skip/take to regexes
+    for src, dst in (('skip', 'skip_regexes'), ('take', 'take_regexes')):
+        config[dst] = []
+        for spec in config.Source[src]:
+            spec = spec.replace('\\', '\\\\')
+            spec = spec.replace('.', '\\.')
+            # Create valid regex for star
+            spec = spec.replace('*', '.*')
+            regex = re.compile(spec, re.I)
+            config[dst].append(regex)
+
     # Load token contents from file, if specified
     if config.Source.type == 'github' and config.GitHub.token and config.GitHub.token[0] == '@':
         path = config.GitHub.token[1:]
@@ -131,6 +132,7 @@ def check(config:ns.Namespace):
     ns.check_default(src, 'datadir', '')
     ns.check_default(src, 'cspdir', '')
     ns.check_default(src, 'skip', [])
+    ns.check_default(src, 'take', [])
 
     # Strip leading slash if present, we don't need it
     if src.srcdir == '/':

src/fsrepo.py

@@ -44,9 +44,14 @@ def get_names(self):
             norm = '/' + (name.replace(os.sep, '/') if os.sep != '/' else name)
             skip = any(rx.match(norm) for rx in self.config.skip_regexes)
             if skip:
-                logging.debug('Skipping %s because config requested so', name)
+                logging.debug('Skipping %s because item in "skip" list', name)
                 continue
 
+            if self.config.take_regexes:
+                if not any(rx.match(norm) for rx in self.config.take_regexes):
+                    logging.debug('Skipping %s because not in "take" list', name)
+                    continue
+
             parts = name.split(os.sep)
 
             if path_matches(cspdir, parts):

src/ziprepo.py

@@ -51,13 +51,18 @@ def get_names(self):
             if parts[-1][0] == '.':
                 continue
 
-            # Check for items we should skip (remove base directory name first)
+            # Check for items we should skip/take (remove base directory name first)
             tmp = '/' + name.split('/', 1)[1]
             skip = any(rx.match(tmp) for rx in self.config.skip_regexes)
             if skip:
-                logging.debug('Skipping %s because config requested so', tmp)
+                logging.debug('Skipping %s because item in "skip" list', name)
                 continue
-
+
+            if self.config.take_regexes:
+                if not any(rx.match(tmp) for rx in self.config.take_regexes):
+                    logging.debug('Skipping %s because not in "take" list', name)
+                    continue
+
             # Configure separately for CSP and source?
             encoding = self.config.Source.encoding
 

template.toml

@@ -23,11 +23,18 @@ cspdir = 'csp'
 # Directory to search for data files to include, if any.
 datadir = 'data'
 
-# List of items to skip when building the export. Supports *-style wildcards.
-# The items are matched against their full path from the root of the repository.
+# List of items to skip when building the export. Supports *-style
+# wildcards. The items are matched against their full path in the
+# repository, e.g. '/src/Strix/Lib/String.cls'.
 skip = [
 ]
 
+# List of items to include in the export. Supports *-style wildcards. If
+# this list is empty, all items will be included. Items both in the skip
+# and take lists will be excluded. Items are matched against their full
+# path in the repository, e.g. '/src/Strix/Lib/String.cls'.
+take = [
+]
 
 # Details on what to do with CSP items
 [CSP]

tests/test_take.py

@@ -0,0 +1,91 @@
+from importlib import import_module
+from io import BytesIO
+
+from lxml import etree
+
+import pytest
+
+builder = import_module("build-export")
+
+
+CFG = """
+[Source]
+type = "directory"
+srctype = "xml"
+srcdir = 'src'
+cspdir = 'csp'
+datadir = 'data'
+skip = [{skip}]
+take = [{take}]
+[CSP]
+export = 'embed'
+[Data]
+export = 'embed'
+[[CSP.parsers]]
+regex = '((/csp)?/[^/]+)/(.+)'
+app = '\\1'
+item = '\\3'
+nomatch = 'error'
+[Directory]
+path = '{path}'
+[Local]
+outfile = 'out.xml'
+loglevel = 'debug'
+"""
+
+
+@pytest.mark.usefixtures("reload_modules")
+def test_take_basic(src_tree, tmp_path, get_build, validate_schema):
+    """Basic test for take"""
+
+    # Get parsed configuration
+    paths = '/csp/app/hello.csp', '/data/Test*', '/src/a.cls.xml'
+    takes = ','.join((f"'{p}'" for p in paths))
+
+    cfg = CFG.format(path=src_tree, skip='', take=takes)
+    xml = get_build(cfg, tmp_path)
+
+    # Parse with ElementTree
+    tree = etree.parse(BytesIO(xml))
+    assert tree.docinfo.root_name == 'Export'
+
+    export = tree.getroot()
+    assert len(export) ==  3, "Expect 3 items in export"
+
+    # Make sure the things we skipped are not in the export
+    assert tree.find('./CSP[@name="hello.csp"]') is not None, "hello.csp in export"
+    assert tree.find('./Document[@name="Test.LUT"]') is not None, "Test.LUT in export"
+    assert tree.find('./Class[@name="tmp.a"]') is not None, "tmp.a in export"
+
+    validate_schema(xml)
+
+
+@pytest.mark.usefixtures("reload_modules")
+def test_skip_overrides_take(src_tree, tmp_path, get_build, validate_schema):
+    """Tests that skip has priority over take"""
+
+    # Get parsed configuration
+    paths = '/csp/app/hello.csp', '/data/Test*', '/src/a.cls.xml'
+    takes = ','.join((f"'{p}'" for p in paths))
+    skips = "'/csp/app/hello.csp'"
+
+    cfg = CFG.format(path=src_tree, skip=skips, take=takes)
+    xml = get_build(cfg, tmp_path)
+
+    # Parse with ElementTree
+    tree = etree.parse(BytesIO(xml))
+    assert tree.docinfo.root_name == 'Export'
+
+    export = tree.getroot()
+    assert len(export) ==  2, "Expect 2 items in export"
+
+    # Make sure the things we skipped are not in the export
+    assert tree.find('./CSP[@name="hello.csp"]') is None, "hello.csp must not in export"
+    assert tree.find('./Document[@name="Test.LUT"]') is not None, "Test.LUT in export"
+    assert tree.find('./Class[@name="tmp.a"]') is not None, "tmp.a in export"
+
+    validate_schema(xml)
+
+
+
+