gitmole 0.6.4__tar.gz → 0.6.6__tar.gz

{gitmole-0.6.4 → gitmole-0.6.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gitmole
-Version: 0.6.4
+Version: 0.6.6
 Summary: Offline git repository analysis with a terminal report: hotspots, coupling, ownership, code age, secrets, repo health.
 License: MIT
 Project-URL: Homepage, https://github.com/antvinni/gitmole

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/__init__.py

@@ -1,3 +1,3 @@
 """gitmole: offline git repository analysis with a terminal report."""
 
-__version__ = "0.6.4"
+__version__ = "0.6.6"

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/cli.py

@@ -15,7 +15,7 @@ from rich.live import Live
 from rich.spinner import Spinner
 from rich.text import Text
 
-from . import __version__, banner, filetypes, findings, load, loss, run
+from . import __version__, banner, blame, filetypes, findings, load, loss, run
 
 
 def parse_args(argv):
@@ -294,6 +294,8 @@ def _meta_for_run(repo_dir: str, args, estimate, age_ok: bool, plots_ok: bool, p
     meta = run.collect_meta(repo_dir, since=args.since_date)
     meta["file_types"] = types_spec   # the loader filters scc's size data the way every other step was filtered
     meta["gone_months"] = args.gone
+    ignore = list(run.DATA_IGNORES if args.ignore_data else []) + list(args.ignore)
+    meta["generated"] = filetypes.generated_files(repo_dir, blame.text_files(repo_dir, ignore))   # hidden from the tables, out of the findings
     if args.since_date and meta["commits"] == 0:
         raise NoCommits(f"no commits since {args.since_date}; widen --since")
     if args.now:

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/coupling.py

@@ -32,12 +32,13 @@ def _components(pairs: list) -> list:
     return list(groups.values())
 
 
-def clusters(pairs: list, min_files: int = 4) -> tuple:
+def clusters(pairs: list, min_files: int = 4, min_density: float = 0.8) -> tuple:
     """Split `pairs` into (groups, rest). Pairs whose two files share a directory are gathered per
-    directory and then into connected groups; a group touching at least `min_files` distinct files
-    becomes one cluster with the file and pair counts, the weakest degree and the mean of the pairs'
-    average revisions. Every other pair comes back unchanged, in its original order. Two unrelated
-    pairs in one directory are not a cluster. Clusters are largest first."""
+    directory and then into connected groups; a group of at least `min_files` distinct files with at
+    least `min_density` of the possible pairs present becomes one cluster with the file and pair
+    counts, the weakest degree and the mean of the pairs' average revisions. Every other pair comes
+    back unchanged, in its original order. Two unrelated pairs in one directory, or a chain of pairs,
+    are not a cluster: "each other" has to be true. Clusters are largest first."""
     by_dir = defaultdict(list)
     for p in pairs:
         if _dir(p["entity"]) == _dir(p["coupled"]):
@@ -46,7 +47,8 @@ def clusters(pairs: list, min_files: int = 4) -> tuple:
     for directory, ps in by_dir.items():
         for component in _components(ps):
             files = {p["entity"] for p in component} | {p["coupled"] for p in component}
-            if len(files) < min_files:
+            possible = len(files) * (len(files) - 1) / 2
+            if len(files) < min_files or len(component) < min_density * possible:
                 continue
             groups.append({"dir": directory, "files": len(files), "pairs": len(component), "degree": min(p["degree"] for p in component),
                            "average-revs": round(sum(p["average-revs"] for p in component) / len(component))})

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/filetypes.py

@@ -2,6 +2,8 @@
 Standalone so blame.py and maat.py can import it as scripts."""
 from __future__ import annotations
 
+import fnmatch
+import os
 import re
 import subprocess
 from collections import Counter
@@ -83,6 +85,63 @@ def is_vendor_path(path: str) -> bool:
     return bool(_VENDOR_PATH.search(path))
 
 
+_RELEASE_NAMES = {"version", "version.rb", "version.py", "version.go", "version.rs", "version.txt", "__version__.py", "package.json",
+                  "package-lock.json", "yarn.lock", "pnpm-lock.yaml", "gemfile", "gemfile.lock", "cargo.toml", "cargo.lock",
+                  "pyproject.toml", "setup.py", "setup.cfg", "poetry.lock", "uv.lock", "go.mod", "go.sum", "composer.json", "composer.lock"}
+
+
+def is_release_path(path: str) -> bool:
+    """Release plumbing: version files, manifests, lock files and changelogs. Two of them changing
+    together is a release commit, not a dependency between them."""
+    name = path.rsplit("/", 1)[-1].lower()
+    return name in _RELEASE_NAMES or name.endswith(".gemspec") or name.startswith(("changelog", "changes.", "history.", "news."))
+
+
+# What a generated file says about itself in its first lines: protoc, ajv, code generators of every kind.
+_GENERATED = re.compile(r"auto[- ]?generated|generated (by|from|file|code|automatically|with)|do not (edit|modify)|@generated|code generated", re.I)
+GENERATED_HEAD_LINES = 5
+
+
+def _generated_patterns(repo: str) -> list:
+    """The .gitattributes patterns marked linguist-generated at the repository root."""
+    try:
+        with open(os.path.join(repo, ".gitattributes"), encoding="utf-8", errors="replace") as fh:
+            lines = fh.read().splitlines()
+    except OSError:
+        return []
+    out = []
+    for line in lines:
+        parts = line.split()
+        if len(parts) >= 2 and any(p in ("linguist-generated", "linguist-generated=true") for p in parts[1:]):
+            out.append(parts[0].lstrip("/"))
+    return out
+
+
+def _attribute_match(path: str, pattern: str) -> bool:
+    if "/" in pattern:
+        return fnmatch.fnmatchcase(path, pattern) or fnmatch.fnmatchcase(path, pattern.rstrip("/") + "/*")
+    return fnmatch.fnmatchcase(path.rsplit("/", 1)[-1], pattern)
+
+
+def generated_files(repo: str, paths: list) -> list:
+    """The tracked files that are generated: marked linguist-generated in .gitattributes, or saying so
+    in their first lines. Their complexity and churn are the generator's, not the repository's."""
+    patterns = _generated_patterns(repo)
+    out = []
+    for path in paths:
+        if any(_attribute_match(path, p) for p in patterns):
+            out.append(path)
+            continue
+        try:
+            with open(os.path.join(repo, path), "rb") as fh:
+                head = fh.read(2048)
+        except OSError:
+            continue
+        if any(_GENERATED.search(line) for line in head.decode("utf-8", "replace").splitlines()[:GENERATED_HEAD_LINES]):
+            out.append(path)
+    return sorted(out)
+
+
 def key(path: str) -> str:
     """The lowercased extension, or the whole lowercased name when there is none."""
     name = path.rsplit("/", 1)[-1].lower()

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/findings.py

@@ -100,6 +100,14 @@ def _source_ownership(report: dict) -> list:
     return [r for r in report.get("ownership") or [] if not (filetypes.is_test_path(r["entity"]) or filetypes.is_vendor_path(r["entity"]))]
 
 
+def _present_areas(report: dict, rows: list, build=knowledge.areas) -> list:
+    """Areas built from the ownership rows of directories that still exist, then only those areas that
+    exist themselves: a directory the history knows but HEAD does not (the layout before a move to
+    src/ or crates/) is nowhere to pair anyone on. `build` is knowledge.areas or a wrapper of it."""
+    tree = _tree(report)
+    return [a for a in build(knowledge.present_rows(rows, tree)) if knowledge.in_tree(a["area"], tree)]
+
+
 def bus_factor(report: dict, threshold: float = 0.7, min_lines: int = 200) -> list:
     """One author owns most of the surviving code (whole history). The areas named in the advice
     come from lines added, which `--since` windows, so the advice says so when it applies."""
@@ -111,7 +119,7 @@ def bus_factor(report: dict, threshold: float = 0.7, min_lines: int = 200) -> li
     if lines / total <= threshold:
         return []
     theirs = []
-    for a in knowledge.areas(_source_ownership(report)):
+    for a in _present_areas(report, _source_ownership(report)):
         owned = dict(a["owners"]).get(name, 0)
         if a["lines"] >= min_lines and owned / a["lines"] >= 0.8:
             theirs.append((a["area"], round(100 * owned / a["lines"])))
@@ -177,10 +185,12 @@ def hotspot_dominance(report: dict, ratio: float = 2.0, minimum: int = 20) -> li
 
 def tight_coupling(report: dict, min_degree: int = 80, min_revs: int = 5) -> list:
     """A file and its test are expected to change together, so pairs with a test file on either side are
-    left out; so are pairs where either file is no longer in the tree, which are history, not a dependency."""
+    left out; so are pairs where either file is no longer in the tree, which are history, not a dependency,
+    and pairs of release plumbing (two version files, a manifest and its lock file), which are a release."""
     tree = _tree(report)
     pairs = [p for p in report.get("coupling") or [] if p["degree"] >= min_degree and p["average-revs"] >= min_revs
              and not (filetypes.is_test_path(p["entity"]) or filetypes.is_test_path(p["coupled"]))
+             and not (filetypes.is_release_path(p["entity"]) and filetypes.is_release_path(p["coupled"]))
              and not (tree and (p["entity"] not in tree or p["coupled"] not in tree))]
     if not pairs:
         return []
@@ -263,7 +273,9 @@ def reverts(report: dict, min_share: float = 0.05, min_count: int = 5, warn_shar
 
 
 def knowledge_islands(report: dict, min_lines: int = 200, min_share: float = 0.9) -> list:
-    areas = knowledge.areas(_source_ownership(report))
+    """Areas of the tree written almost entirely by one person. Areas that no longer exist are left
+    out, of the islands and of the total they are measured against."""
+    areas = _present_areas(report, _source_ownership(report))
     islands = knowledge.islands(areas, min_lines=min_lines, min_share=min_share)
     if not islands:
         return []
@@ -329,9 +341,10 @@ def _loss_people(by_person: dict, total: int) -> str:
 
 
 def _loss_areas(report: dict, names: set, source_rows: list) -> list:
-    """Areas at 200+ lines where 80%+ of the surviving code is theirs, tagged live or not and
-    sorted live-first: that is where the gap bites soonest."""
-    theirs = [a for a in loss.areas(source_rows, names) if a["lines"] >= 200 and a["lost_share"] >= 0.8]
+    """Areas at 200+ lines where 80%+ of the surviving code is theirs, still in the tree, tagged live
+    or not and sorted live-first: that is where the gap bites soonest."""
+    theirs = [a for a in _present_areas(report, source_rows, build=lambda rows: loss.areas(rows, names))
+              if a["lines"] >= 200 and a["lost_share"] >= 0.8]
     for a in theirs:
         a["live"] = _is_live(a["area"], report.get("age") or [])
     theirs.sort(key=lambda a: (not a["live"], -a["lines"], a["area"]))   # a live area first: that is where the gap bites
@@ -375,20 +388,36 @@ def _partial_functions(report: dict) -> str:
 
 
 def brain_methods(report: dict, min_ccn: int = 15, min_lines: int = 100) -> list:
-    """Functions that are both long and complex, in this repository's own source files: test files and
-    vendored code are left out. A warning when one sits in a hotspot."""
+    """Functions that are both long and complex, in this repository's own source files: test files,
+    vendored code and generated files are left out. A warning when one sits in a hotspot."""
+    generated = _generated(report)
     big = [f for f in report.get("functions") or [] if f["ccn"] >= min_ccn and f["nloc"] >= min_lines
-           and not (filetypes.is_test_path(f["file"]) or filetypes.is_vendor_path(f["file"]))]
+           and not (filetypes.is_test_path(f["file"]) or filetypes.is_vendor_path(f["file"]) or f["file"] in generated)]
     if not big:
         return []
     big.sort(key=lambda f: (-f["ccn"], -f["nloc"], f["file"], f["function"], f["start"]))
     hot = hotspots.top(report)
     sev = "warning" if any(f["file"] in hot for f in big) else "info"
-    listed = "; ".join(f"{f['function']} ({f['file']}) complexity {f['ccn']}, {f['nloc']} lines, {f['params']} params" for f in big[:5])
+    listed = "; ".join(f"{f['function']} ({_place(f)}) complexity {f['ccn']}, {f['nloc']} lines, {f['params']} params" for f in big[:5])
     more = f" and {len(big) - 5} more" if len(big) > 5 else ""
+    first = big[0]
+    which = f"the anonymous function at {_place(first)}" if first["function"] == ANONYMOUS else f"{first['function']} in {first['file']}"
     return [_f(sev, "Brain methods",
                f"{len(big)} function(s) are both long and complex: {listed}{more}.{_partial_functions(report)}",
-               f"Split {big[0]['function']} in {big[0]['file']} first, before the next change lands there.")]
+               f"Split {which} first, before the next change lands there.")]
+
+
+ANONYMOUS = "(anonymous)"
+
+
+def _place(f: dict) -> str:
+    """Where a function is: its file, or file:line when it has no name to find it by."""
+    return f"{f['file']}:{f['start']}" if f["function"] == ANONYMOUS else f["file"]
+
+
+def _generated(report: dict) -> set:
+    """Files the run found to be generated (a header marker or a linguist-generated attribute)."""
+    return set((report.get("meta") or {}).get("generated") or [])
 
 
 def complexity_growth(report: dict, min_growers: int = 3, min_pct: int = 25, top_n: int = 10) -> list:

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/identity.py

@@ -20,8 +20,26 @@ def _tokens(name: str) -> set:
     return {t for t in re.split(r"[^a-z0-9]+", name.lower()) if len(t) >= 3}
 
 
+# A bare first name under two emails may be two people; anything else spelled identically is one.
+_COMMON_FIRST_NAMES = {
+    "adam", "alex", "alexander", "andrew", "andy", "ann", "anna", "ben", "bob", "chris", "dan", "daniel", "dave", "david", "ed",
+    "eric", "frank", "george", "jack", "james", "jan", "jean", "jim", "joe", "john", "jon", "josh", "kevin", "lee", "li", "luke",
+    "mark", "martin", "matt", "max", "michael", "mike", "nick", "paul", "pete", "peter", "phil", "rob", "robert", "ryan", "sam",
+    "scott", "steve", "tim", "tom", "tony", "will",
+}
+
+
+def _plain(name: str) -> str:
+    return " ".join(name.lower().split())
+
+
 def same_person(a: dict, b: dict) -> bool:
-    return a["email"].lower() == b["email"].lower() or len(_tokens(a["name"]) & _tokens(b["name"])) >= 2
+    """Same email, two shared name tokens, or the same name spelled identically (a handle such as
+    KaKa under three emails), unless that name is a bare common first name."""
+    if a["email"].lower() == b["email"].lower() or len(_tokens(a["name"]) & _tokens(b["name"])) >= 2:
+        return True
+    name = _plain(a["name"])
+    return bool(name) and name == _plain(b["name"]) and name not in _COMMON_FIRST_NAMES
 
 
 def merge(identities: list) -> list:

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/knowledge.py

@@ -7,6 +7,16 @@ from collections import Counter, defaultdict
 ROOT = "(root files)"
 
 
+def in_tree(area: str, tree: dict) -> bool:
+    """Whether any tracked file sits under `area` (a directory prefix ending in "/", or ROOT). With
+    no tree listing every area counts: there is nothing to judge by."""
+    if not tree:
+        return True
+    if area == ROOT:
+        return any("/" not in path for path in tree)
+    return any(path.startswith(area) for path in tree)
+
+
 def _area(entity: str, depth: int) -> str:
     dirs = entity.split("/")[:-1]
     if not dirs:
@@ -14,6 +24,20 @@ def _area(entity: str, depth: int) -> str:
     return "/".join(dirs[:depth]) + "/"
 
 
+def top_area(entity: str) -> str:
+    """The top-level directory of a path, or ROOT."""
+    return _area(entity, 1)
+
+
+def present_rows(rows: list, tree: dict) -> list:
+    """Ownership rows for files whose top-level directory still exists in the tree. Filtering the rows
+    before areas are built keeps a vanished layout (the src/ before a move to crates/) from inflating
+    the total and hiding that one directory now holds almost everything."""
+    if not tree:
+        return rows
+    return [r for r in rows if in_tree(top_area(r["entity"]), tree)]
+
+
 def _aggregate(rows: list, depth: int) -> list:
     lines, per_author = Counter(), defaultdict(Counter)
     for r in rows:

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/leaks.py

@@ -33,10 +33,13 @@ RAW_FIELDS = ("Secret", "Match", "Line", "Message", "Attributes")
 
 # Shapes that cannot be a live secret: a version string (5.0.0-1667386184.dfbbb54), a token shortened
 # with an ellipsis, a whole-value template marker (your-project-id, <your-token>, XXXX-XXXX, changeme),
+# a whole value that is one of the words every example uses (`password: 'hello'` in a doc comment),
 # and a key block whose body holds no key material. Every rule is about the whole value; nothing is
 # skipped by prefix, since a public and a private key of the same service often share one.
 _VERSION = re.compile(r"^\d+\.\d+\.\d+(?:[-+][0-9A-Za-z.-]+)?$")
 _MARKER = re.compile(r"^(<[^<>]+>|x+|(?:x{2,}[-_ ]?)+|your[-_][\w-]+|change[-_]?me|replace[-_]?me)$", re.I)
+_EXAMPLE_WORDS = {"password", "passwd", "pass", "secret", "hello", "hey", "test", "example", "sample", "dummy", "foo", "bar",
+                  "baz", "admin", "root", "user", "123456", "12345678", "123456789", "abc123", "qwerty", "letmein", "welcome"}
 _KEY_BLOCK = re.compile(r"-----BEGIN [A-Z ]*KEY-----(.*?)-----END [A-Z ]*KEY-----", re.S)
 _KEY_MATERIAL = 64   # a real body is hundreds of base64 characters; a template has dots or a few x's
 
@@ -54,7 +57,7 @@ def digest(value: str, key: bytes) -> str:
 
 def is_placeholder(value: str) -> bool:
     value = (value or "").strip()
-    if _VERSION.match(value) or value.endswith("...") or value.endswith("…") or _MARKER.match(value):
+    if _VERSION.match(value) or value.endswith("...") or value.endswith("…") or _MARKER.match(value) or value.lower() in _EXAMPLE_WORDS:
         return True
     m = _KEY_BLOCK.search(value)
     if m:

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/load.py

@@ -151,7 +151,7 @@ def parse_functions(text: str) -> list:
     for r in csv.reader(io.StringIO(text)):
         if len(r) < 11:
             continue
-        rows.append({"file": _rel(r[6]), "function": r[7], "ccn": _num(r[1]), "nloc": _num(r[0]), "params": _num(r[3]),
+        rows.append({"file": _rel(r[6]), "function": r[7] or "(anonymous)", "ccn": _num(r[1]), "nloc": _num(r[0]), "params": _num(r[3]),
                      "start": _num(r[9]), "end": _num(r[10])})
     return rows
 

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole/render.py

@@ -111,6 +111,24 @@ def _hide_vendor(rows: list, path_of, full, noun="file in vendored code", plural
     return _hide_rows(rows, path_of, full, filetypes.is_vendor_path, noun, plural)
 
 
+def _hide_generated(rows: list, path_of, report: dict, full, noun="generated file", plural=None) -> tuple:
+    """Generated files (a header marker or a linguist-generated attribute, found at run time): the
+    generator's churn and complexity, not the repository's."""
+    generated = set((report.get("meta") or {}).get("generated") or [])
+    return _hide_rows(rows, path_of, full, lambda p: p in generated, noun, plural)
+
+
+def _hide_release(pairs: list, full) -> tuple:
+    """Coupled pairs where both files are release plumbing (version files, manifests, lock files,
+    changelogs): they change together because a release touches them all, not because one depends on
+    the other. A version file paired with real code stays."""
+    if full is True:
+        return pairs, None
+    kept = [p for p in pairs if not (filetypes.is_release_path(p["entity"]) and filetypes.is_release_path(p["coupled"]))]
+    hidden = len(pairs) - len(kept)
+    return kept, (f"{hidden} release pair{'s' if hidden != 1 else ''} hidden{HIDDEN_SUFFIX}" if hidden else None)
+
+
 def _join_hidden(*notes) -> str:
     """Several hidden-row notes as one caption phrase: 'A hidden; B hidden; --full shows them'."""
     parts = [n[:-len(HIDDEN_SUFFIX)] if n.endswith(HIDDEN_SUFFIX) else n for n in notes if n]
@@ -376,7 +394,8 @@ def hotspots_section(report: dict, full: bool = True, width=None) -> dict:
     scored = hotspots.ranked(report)
     scored, hidden_note = _hide_tests(scored, lambda h: h["entity"], full)
     scored, deleted_note = _hide_deleted(scored, report, full)
-    hidden_note = _join_hidden(hidden_note, deleted_note)
+    scored, generated_note = _hide_generated(scored, lambda h: h["entity"], report, full)
+    hidden_note = _join_hidden(hidden_note, deleted_note, generated_note)
     title = "Hotspots (score = revisions × lines of code)" if full is True else "Hotspots"
     limit = _limit("Hotspots", full)
     series = (report.get("trend") or {}).get("files") or {}
@@ -408,6 +427,8 @@ def coupling_section(report: dict, full: bool = True, width=None) -> dict:
     pairs = sorted((p for p in report.get("coupling") or [] if p["average-revs"] >= 5), key=lambda p: (-p["degree"], -p["average-revs"]))
     pairs, hidden_note = _hide_tests(pairs, lambda p: (p["entity"], p["coupled"]), full, noun="test pair")
     pairs, gone_note = _hide_gone(pairs, report, full)
+    pairs, release_note = _hide_release(pairs, full)
+    gone_note = _join_hidden(gone_note, release_note)
     groups, cluster_note = [], None
     if full is not True:
         # a directory whose files all change together is one row; --full lists every pair
@@ -474,7 +495,8 @@ def functions_section(report: dict, full: bool = True, width=None) -> dict:
     funcs = sorted((f for f in measured if f["ccn"] >= CCN_FLOOR), key=lambda f: (-f["ccn"], -f["nloc"], f["file"], f["function"], f["start"]))
     funcs, hidden_note = _hide_tests(funcs, lambda f: f["file"], full, noun="function in a test file", plural="functions in test files")
     funcs, vendor_note = _hide_vendor(funcs, lambda f: f["file"], full, noun="function in vendored code", plural="functions in vendored code")
-    hidden_note = _join_hidden(hidden_note, vendor_note)
+    funcs, generated_note = _hide_generated(funcs, lambda f: f["file"], report, full, noun="function in a generated file", plural="functions in generated files")
+    hidden_note = _join_hidden(hidden_note, vendor_note, generated_note)
     limit = _limit("Complex functions", full)
     rows = [(f["function"], f["file"], f["ccn"], f["nloc"], f["params"]) for f in funcs[:limit]]
     columns = [("function", {"overflow": "fold"}), ("file", PATH), ("ccn", RIGHT), ("lines", RIGHT), ("params", RIGHT)]
@@ -502,7 +524,16 @@ def knowledge_section(report: dict, full: bool = True, width=None) -> dict:
     """Ownership by area of the tree: who wrote most of each directory, gone owners marked."""
     months = report["meta"].get("gone_months", loss.DEFAULT_MONTHS)
     gone = {g["name"] for g in loss.gone(report, months)}
-    areas = loss.areas(report.get("ownership") or [], gone)   # every area the map showed before, tests included
+    rows_all = report.get("ownership") or []   # every area the map showed before, tests included
+    areas = loss.areas(rows_all, gone)
+    hidden_note = None
+    tree = (report.get("size") or {}).get("files") or {}
+    if full is not True and tree:
+        # a directory the history knows but HEAD does not is a layout that no longer exists; the rows are
+        # filtered before the areas are built so a vanished layout cannot hide that one directory now dominates
+        areas = [a for a in loss.areas(knowledge.present_rows(rows_all, tree), gone) if knowledge.in_tree(a["area"], tree)]
+        hidden = len({knowledge.top_area(r["entity"]) for r in rows_all if not knowledge.in_tree(knowledge.top_area(r["entity"]), tree)})
+        hidden_note = f"{hidden} historical area{'s' if hidden != 1 else ''} hidden{HIDDEN_SUFFIX}" if hidden else None
     limit = _limit("Knowledge map", full)
     rows = []
     for a in areas[:limit]:
@@ -512,7 +543,7 @@ def knowledge_section(report: dict, full: bool = True, width=None) -> dict:
     columns = [("area", PATH), ("lines added", RIGHT), ("authors", RIGHT), ("lost", RIGHT), ("main owner", {}), ("second", {})]
     if full is not True:
         columns, rows = _keep(columns, rows, ["area", "lines added", "main owner", "second"])
-    notes = [c for c in (_more(len(areas), limit),) if c]
+    notes = [c for c in (_more(len(areas), limit), hidden_note) if c]
     if gone:
         notes.append(f"gone = no commits in the {months} months before {report['meta'].get('last_date')}"
                      + ("; gone and lost are measured over the whole history" if report["meta"].get("since") else ""))

{gitmole-0.6.4 → gitmole-0.6.6}/gitmole.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gitmole
-Version: 0.6.4
+Version: 0.6.6
 Summary: Offline git repository analysis with a terminal report: hotspots, coupling, ownership, code age, secrets, repo health.
 License: MIT
 Project-URL: Homepage, https://github.com/antvinni/gitmole

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_cli.py

@@ -592,6 +592,28 @@ class Arguments(unittest.TestCase):
         self.assertNotIn("jar", text)
 
 
+class GeneratedFiles(unittest.TestCase):
+    def test_a_run_records_the_generated_files_in_meta(self):
+        with tempfile.TemporaryDirectory() as d:
+            _tiny_repo(d)
+            os.makedirs(os.path.join(d, "lib"))
+            with open(os.path.join(d, "lib", "validator.js"), "w") as fh:
+                fh.write("// This file is autogenerated by build/build.js, do not edit\nmodule.exports = 1\n")
+            with open(os.path.join(d, "lib", "app.js"), "w") as fh:
+                fh.write("module.exports = 2\n")
+            import subprocess
+            subprocess.run(["git", "-C", d, "add", "-A"], check=True)
+            subprocess.run(["git", "-C", d, "-c", "user.name=T", "-c", "user.email=t@x.com", "commit", "-q", "-m", "files"], check=True)
+            out = os.path.join(d, "out")
+            planner = lambda repo, o, branch="HEAD", **kw: [{"name": "q", "argv": ["true"], "stdout": None, "deps": []}]
+            rc = cli.main([d, "--out", out], console=console(), tool_check=lambda **kw: [], planner=planner,
+                          estimator=lambda repo, interval, **kw: {"files": 2, "samples": 1, "blames": 2})
+            with open(os.path.join(out, "meta.json")) as fh:
+                meta = json.load(fh)
+        self.assertEqual(rc, 0)
+        self.assertEqual(meta["generated"], ["lib/validator.js"])
+
+
 class Clean(unittest.TestCase):
     """--clean lists what gitmole left behind and deletes on a yes. TMPDIR is pointed at a scratch dir so the
     real temp folder is never listed or touched."""

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_coupling.py

@@ -45,8 +45,24 @@ class Clusters(unittest.TestCase):
         self.assertEqual([g["dir"] for g in groups], ["lib/", "(root files)"])
         self.assertEqual(rest, [])
 
-    def test_a_partial_clique_counts_the_files_it_touches(self):
-        pairs = [pair("d/a", "d/b"), pair("d/b", "d/c"), pair("d/c", "d/e")]   # 4 files, 3 pairs, not every pair present
-        [g], rest = coupling.clusters(pairs)
-        self.assertEqual((g["files"], g["pairs"]), (4, 3))
+    def test_a_chain_is_not_a_cluster(self):
+        pairs = [pair("d/a", "d/b"), pair("d/b", "d/c"), pair("d/c", "d/e")]   # 4 files, 3 of 6 possible pairs: connected, not a clique
+        groups, rest = coupling.clusters(pairs)
+        self.assertEqual(groups, [])
+        self.assertEqual(rest, pairs)
+
+    def test_a_near_complete_clique_is_a_cluster(self):
+        pairs = clique("d", ["a", "b", "c", "e", "f"])   # 5 files, 10 pairs
+        eight = pairs[:8]
+        [g], rest = coupling.clusters(eight)
+        self.assertEqual((g["files"], g["pairs"]), (5, 8))
         self.assertEqual(rest, [])
+        groups, rest = coupling.clusters(pairs[:7])   # 7 of 10 is under the 80% floor
+        self.assertEqual(groups, [])
+        self.assertEqual(len(rest), 7)
+
+    def test_two_unrelated_pairs_in_one_directory_are_not_a_cluster(self):
+        pairs = [pair("a", "b"), pair("c", "d")]
+        groups, rest = coupling.clusters(pairs)
+        self.assertEqual(groups, [])
+        self.assertEqual(rest, pairs)

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_filetypes.py

@@ -104,6 +104,33 @@ class TestPaths(unittest.TestCase):
         for path in ("app/settings.py", "examplesite/app.py", "src/rulesets/a.go", "sampler/x.py", "config/betterleaks.toml"):
             self.assertFalse(filetypes.is_sample_path(path), path)
 
+    def test_release_plumbing_files(self):
+        for path in ("lib/sinatra/version.rb", "VERSION", "src/pkg/__version__.py", "package.json", "package-lock.json", "Gemfile.lock",
+                     "Cargo.toml", "pyproject.toml", "go.sum", "CHANGELOG.md", "CHANGES.rst", "sinatra.gemspec", "uv.lock"):
+            self.assertTrue(filetypes.is_release_path(path), path)
+        for path in ("lib/version_check.py", "src/app.py", "docs/versions.md", "Makefile", "lib/sinatra/base.rb"):
+            self.assertFalse(filetypes.is_release_path(path), path)
+
+    def test_generated_files_by_header_marker_or_attribute(self):
+        with tempfile.TemporaryDirectory() as d:
+            files = {
+                "lib/config-validator.js": "// This file is autogenerated by build/build-validation.js, do not edit\n'use strict'\n",
+                "pb/api.pb.go": "// Code generated by protoc-gen-go. DO NOT EDIT.\npackage pb\n",
+                "gen/schema.py": "# @generated\nx = 1\n",
+                "src/app.py": "# the app; it generated reports once\ndef main():\n    pass\n",
+                "docs/notes.md": "generated notes are the best notes\n",
+                "dist/bundle.js": "var a = 1;\n",
+                "src/late.py": "\n" * 10 + "# generated by hand, do not edit\n",   # a marker past the first lines does not count
+            }
+            for path, text in files.items():
+                os.makedirs(os.path.join(d, os.path.dirname(path)), exist_ok=True)
+                with open(os.path.join(d, path), "w") as fh:
+                    fh.write(text)
+            with open(os.path.join(d, ".gitattributes"), "w") as fh:
+                fh.write("* text=auto\ndist/* linguist-generated=true\n*.min.js linguist-generated\n")
+            found = filetypes.generated_files(d, sorted(files))
+        self.assertEqual(found, ["dist/bundle.js", "gen/schema.py", "lib/config-validator.js", "pb/api.pb.go"])
+
     def test_vendored_trees(self):
         for path in ("vendor/github.com/x/y.go", "web/node_modules/a/index.js", "third_party/z/a.c", "thirdparty/a.c", "_vendor/a.py",
                      "external/lib/a.cpp"):

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_findings.py

@@ -145,6 +145,14 @@ class BusFactor(unittest.TestCase):
         self.assertEqual(findings.bus_factor(r)[0]["advice"], "Pair someone with Ann on core/ first; it is 95% theirs since 2025-01-01.",
                          "ownership is windowed while the headline share is not")
 
+    def test_areas_no_longer_in_the_tree_are_not_named_in_the_advice(self):
+        own = [{"entity": "flask/a.py", "author": "Ann", "added": 5000, "deleted": 0},   # the pre-src/ layout
+               {"entity": "src/a.py", "author": "Ann", "added": 900, "deleted": 0},
+               {"entity": "src/b.py", "author": "Bob", "added": 50, "deleted": 0}]
+        tree = {"files": {"src/a.py": {"code": 1, "complexity": 0}, "src/b.py": {"code": 1, "complexity": 0}}}
+        f = findings.bus_factor(report(theseus_authors={"Ann": 79, "Bob": 21}, ownership=own, size=tree))
+        self.assertEqual(f[0]["advice"], "Pair someone with Ann on src/ first; it is 95% theirs.")
+
     def test_vendored_trees_are_not_named_in_the_advice(self):
         own = [{"entity": "vendor/github.com/x/a.go", "author": "Ann", "added": 500000, "deleted": 0},
                {"entity": "core/a.py", "author": "Ann", "added": 900, "deleted": 0},
@@ -246,6 +254,15 @@ class TightCoupling(unittest.TestCase):
         f = findings.tight_coupling(report(coupling=pairs))
         self.assertIn("2 pairs", f[0]["detail"], "without a tree listing every pair counts")
 
+    def test_release_plumbing_pairs_are_not_a_dependency(self):
+        pairs = [{"entity": "lib/sinatra/version.rb", "coupled": "rack-protection/lib/rack/protection/version.rb", "degree": 100, "average-revs": 60},
+                 {"entity": "package.json", "coupled": "package-lock.json", "degree": 95, "average-revs": 40},
+                 {"entity": "lib/sinatra/version.rb", "coupled": "lib/sinatra/base.rb", "degree": 85, "average-revs": 10}]
+        f = findings.tight_coupling(report(coupling=pairs))
+        self.assertIn("1 pair changes together", f[0]["detail"], "a version file paired with real code still counts")
+        self.assertNotIn("package.json", f[0]["detail"])
+        self.assertEqual(findings.tight_coupling(report(coupling=pairs[:2])), [])
+
     def test_single_pair_reads_grammatically(self):
         pairs = [{"entity": "a", "coupled": "b", "degree": 100, "average-revs": 10}]
         f = findings.tight_coupling(report(coupling=pairs))
@@ -366,6 +383,21 @@ class BrainMethods(unittest.TestCase):
         self.assertNotIn("test_all", f[0]["detail"])
         self.assertEqual(findings.brain_methods(report(functions=fns[:1])), [])
 
+    def test_an_anonymous_function_is_named_by_its_place(self):
+        fns = [{"file": "completions.go", "function": "(anonymous)", "ccn": 47, "nloc": 136, "params": 1, "start": 316, "end": 585}]
+        f = findings.brain_methods(report(functions=fns))
+        self.assertIn("(anonymous) (completions.go:316) complexity 47, 136 lines, 1 params", f[0]["detail"])
+        self.assertEqual(f[0]["advice"], "Split the anonymous function at completions.go:316 first, before the next change lands there.")
+
+    def test_generated_files_are_not_brain_methods(self):
+        fns = [{"file": "lib/config-validator.js", "function": "validate10", "ccn": 373, "nloc": 1150, "params": 5, "start": 1, "end": 1150},
+               {"file": "lib/reply.js", "function": "onSendEnd", "ccn": 34, "nloc": 180, "params": 2, "start": 1, "end": 180}]
+        r = report(functions=fns)
+        r["meta"]["generated"] = ["lib/config-validator.js"]
+        f = findings.brain_methods(r)
+        self.assertEqual(f[0]["advice"], "Split onSendEnd in lib/reply.js first, before the next change lands there.")
+        self.assertNotIn("validate10", f[0]["detail"])
+
     def test_vendored_functions_are_not_brain_methods(self):
         fns = [{"file": "vendor/github.com/google/jsonschema-go/jsonschema/validate.go", "function": "validate", "ccn": 179, "nloc": 424, "params": 3, "start": 1, "end": 424},
                {"file": "processor/workers.go", "function": "countLoopGeneric", "ccn": 56, "nloc": 164, "params": 8, "start": 1, "end": 164}]
@@ -438,6 +470,19 @@ class KnowledgeIslands(unittest.TestCase):
         self.assertEqual(f[0]["advice"], "Pair someone with Bob on core/ first; it is the largest at 300 lines.")
         self.assertNotIn("tests/", f[0]["detail"])
 
+    def test_areas_no_longer_in_the_tree_are_not_islands(self):
+        own = [{"entity": "src/a.rs", "author": "Ann", "added": 30000, "deleted": 0},   # moved to crates/ years ago
+               {"entity": "grep-printer/a.rs", "author": "Ann", "added": 20000, "deleted": 0},
+               {"entity": "crates/core/a.rs", "author": "Bob", "added": 300, "deleted": 0}]
+        tree = {"files": {"crates/core/a.rs": {"code": 300, "complexity": 1}}}
+        f = findings.knowledge_islands(report(ownership=own, size=tree))
+        self.assertEqual(f[0]["advice"], "Pair someone with Bob on crates/core/ first; it is the largest at 300 lines.",
+                         "with the vanished directories gone, crates/ holds everything and the map descends into it")
+        self.assertNotIn("src/", f[0]["detail"])
+        self.assertIn("100% of all lines added", f[0]["detail"], "lines in vanished directories are not in the denominator")
+        f = findings.knowledge_islands(report(ownership=own))
+        self.assertIn("src/", f[0]["detail"], "without a tree listing every area counts")
+
     def test_vendored_trees_are_not_islands(self):
         own = [{"entity": "vendor/github.com/x/a.go", "author": "Ann", "added": 500000, "deleted": 0},
                {"entity": "web/node_modules/y/b.js", "author": "Ann", "added": 90000, "deleted": 0},
@@ -511,6 +556,18 @@ class KnowledgeLoss(unittest.TestCase):
         self.assertIn("Areas mostly theirs: old/ (100%), docs/ (100%)", f[0]["detail"])
         self.assertEqual(f[0]["advice"], "Pair someone on old/ first; nobody who wrote it is around to ask.")
 
+    def test_areas_no_longer_in_the_tree_are_not_named(self):
+        r = self._report(theseus_authors={"Ann": 60, "Bob": 40},
+                         age=[{"entity": "flask/a.py", "age-months": 2}, {"entity": "src/x.py", "age-months": 2}],
+                         ownership=[{"entity": "flask/a.py", "author": "Bob", "added": 8000, "deleted": 0},   # the old layout, all Bob's
+                                    {"entity": "src/x.py", "author": "Bob", "added": 300, "deleted": 0},
+                                    {"entity": "app/b.py", "author": "Ann", "added": 900, "deleted": 0}],
+                         size={"files": {"src/x.py": {"code": 1, "complexity": 0}, "app/b.py": {"code": 1, "complexity": 0}}})
+        f = findings.knowledge_loss(r)
+        self.assertIn("Areas mostly theirs: src/ (100%).", f[0]["detail"])
+        self.assertNotIn("flask/", f[0]["detail"])
+        self.assertEqual(f[0]["advice"], "Pair someone on src/ first; nobody who wrote it is around to ask.")
+
     def test_a_live_area_is_preferred_over_a_bigger_idle_one(self):
         r = self._report(theseus_authors={"Ann": 60, "Bob": 40},
                          age=[{"entity": "old/a.py", "age-months": 30}, {"entity": "live/b.py", "age-months": 3}],

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_identity.py

@@ -18,6 +18,19 @@ class Merge(unittest.TestCase):
         names = [m["name"] for m in merged]
         self.assertEqual(names, ["Bob", "Grzegorz Bankosz", "Ann"])
 
+    def test_an_identical_handle_under_several_emails_is_one_person(self):
+        ids = [{"name": "KaKa", "email": "kaka@a.com", "commits": 57}, {"name": "KaKa", "email": "23028015+climba@users.noreply.github.com", "commits": 56},
+               {"name": "kaka", "email": "climba@b.com", "commits": 10}, {"name": "namusyaka", "email": "n@a.com", "commits": 180},
+               {"name": "namusyaka", "email": "n@b.com", "commits": 8}, {"name": "Li Yu", "email": "li@a.com", "commits": 5},
+               {"name": "Li Yu", "email": "li@b.com", "commits": 3}]
+        merged = {m["name"]: m["commits"] for m in identity.merge(ids)}
+        self.assertEqual(merged, {"KaKa": 123, "namusyaka": 188, "Li Yu": 8})
+
+    def test_a_bare_common_first_name_is_not_enough(self):
+        ids = [{"name": "Jean", "email": "jean@a.com", "commits": 24}, {"name": "Jean", "email": "jean@b.com", "commits": 18},
+               {"name": "Alex", "email": "alex@a.com", "commits": 3}, {"name": "alex", "email": "alex@b.com", "commits": 2}]
+        self.assertEqual(len(identity.merge(ids)), 4, "two Jeans and two Alexes may be four people")
+
     def test_merged_row_sums_commits_and_lists_aliases(self):
         merged = {m["name"]: m for m in identity.merge(IDS)}
         self.assertEqual(merged["Grzegorz Bankosz"]["commits"], 41)

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_knowledge.py

@@ -48,6 +48,36 @@ class OddNames(unittest.TestCase):
         self.assertEqual(areas[0]["lines"], 1000)
 
 
+class InTree(unittest.TestCase):
+    TREE = {"crates/core/a.rs": {}, "crates/ignore/src/b.rs": {}, "build.rs": {}}
+
+    def test_an_area_is_in_the_tree_when_any_tracked_file_sits_under_it(self):
+        self.assertTrue(knowledge.in_tree("crates/", self.TREE))
+        self.assertTrue(knowledge.in_tree("crates/ignore/", self.TREE))
+        self.assertFalse(knowledge.in_tree("src/", self.TREE))
+        self.assertFalse(knowledge.in_tree("crate/", self.TREE), "a prefix of a directory name is not that directory")
+
+    def test_root_files_are_in_the_tree_when_any_file_has_no_directory(self):
+        self.assertTrue(knowledge.in_tree(knowledge.ROOT, self.TREE))
+        self.assertFalse(knowledge.in_tree(knowledge.ROOT, {"crates/core/a.rs": {}}))
+
+    def test_no_tree_listing_means_every_area_counts(self):
+        self.assertTrue(knowledge.in_tree("src/", {}))
+
+    def test_present_rows_drop_vanished_top_level_directories_so_the_survivor_can_dominate(self):
+        rows = [{"entity": "src/a.rs", "author": "Ann", "added": 30000, "deleted": 0},          # the layout before crates/
+                {"entity": "crates/core/a.rs", "author": "Ann", "added": 9000, "deleted": 0},
+                {"entity": "crates/ignore/b.rs", "author": "Bob", "added": 900, "deleted": 0},
+                {"entity": "ci/x.sh", "author": "Ann", "added": 500, "deleted": 0}]
+        tree = {"crates/core/a.rs": {}, "crates/ignore/b.rs": {}, "ci/x.sh": {}}
+        kept = knowledge.present_rows(rows, tree)
+        self.assertEqual([r["entity"] for r in kept], ["crates/core/a.rs", "crates/ignore/b.rs", "ci/x.sh"])
+        self.assertEqual([a["area"] for a in knowledge.areas(kept)], ["crates/core/", "crates/ignore/", "ci/"],
+                         "with src/ gone, crates/ holds over 80% and the map descends into it")
+        self.assertEqual([a["area"] for a in knowledge.areas(rows)], ["src/", "crates/", "ci/"], "the vanished src/ used to hide that")
+        self.assertEqual(knowledge.present_rows(rows, {}), rows)
+
+
 class Islands(unittest.TestCase):
     def test_areas_dominated_by_one_author(self):
         areas = knowledge.areas(OWNERSHIP)

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_leaks.py

@@ -59,6 +59,13 @@ class Placeholder(unittest.TestCase):
         for value in ["AKIA" + "X" * 16, "6L" + "x" * 38, "ghp_" + "a1" * 18, "yourkey" + "9" * 20]:
             self.assertFalse(leaks.is_placeholder(value), "a marker inside real-looking material is not enough: " + value)
 
+    def test_common_example_words_are_placeholders(self):
+        # `password: 'hello'` in a doc comment, `secret` in a sample config: the words every example uses
+        for value in ["hello", "Hello", "secret", "password", "PASSWORD", "example", "123456", "qwerty", "letmein", "foo", "dummy"]:
+            self.assertTrue(leaks.is_placeholder(value), value)
+        for value in ["hello123", "secret-9f8a7b6c5d4e", "s3cr3t!Passw0rd", "foobarbaz2024"]:
+            self.assertFalse(leaks.is_placeholder(value), "a word inside other material is not a placeholder: " + value)
+
     def test_anything_else_is_taken_seriously(self):
         # built at runtime: a literal in these shapes would trip secret scanners on this very file
         key_id, long_key = "AKIA" + "X" * 16, "6L" + "x" * 38

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_load.py

@@ -138,6 +138,11 @@ class ParseFunctions(unittest.TestCase):
     def test_empty(self):
         self.assertEqual(load.parse_functions(""), [])
 
+    def test_a_nameless_function_is_called_anonymous(self):
+        # lizard names Go function literals with an empty string where it names JavaScript's "(anonymous)"
+        rows = load.parse_functions('136,47,926,1,270,"@316-585@completions.go","completions.go",""," c * Command",316,585\n')
+        self.assertEqual((rows[0]["function"], rows[0]["start"]), ("(anonymous)", 316))
+
     def test_a_row_cut_short_by_a_killed_step_does_not_abort_the_report(self):
         rows = load.parse_functions(self.CSV + '5,3,40,1,5,"g@1-5@a.py","a.py","g","g( )",1,\n')
         self.assertEqual(len(rows), 3)

{gitmole-0.6.4 → gitmole-0.6.6}/tests/test_render.py

@@ -212,6 +212,7 @@ class Report(unittest.TestCase):
     def test_knowledge_map_caption_says_gone_is_measured_over_the_whole_history(self):
         r = sample_report()
         r["meta"].update({"last_date": "2026-09-10", "bots": []})
+        r["size"]["files"]["tests/t.py"] = {"code": 1, "complexity": 0}   # every area in the tree, so nothing is hidden
         r["activity"]["authors"] = {"Ann": {"commits": 1, "added": 0, "deleted": 0, "first": "2025-01-01", "last": "2026-09-01"},
                                     "Bob": {"commits": 1, "added": 0, "deleted": 0, "first": "2025-01-01", "last": "2025-01-01"}}
         def caption(rep):
@@ -317,6 +318,36 @@ class Report(unittest.TestCase):
         full = _section_text(rendered(r, [], width=200, full=True), "Complex functions")
         self.assertIn("vendor/github.com/x/y.go", full)
 
+    def test_default_tables_hide_generated_files_and_say_so(self):
+        r = sample_report()
+        r["meta"]["generated"] = ["lib/config-validator.js"]
+        r["size"]["files"]["lib/config-validator.js"] = {"code": 1153, "complexity": 373}
+        r["revisions"].append({"entity": "lib/config-validator.js", "n-revs": 8})
+        r["functions"].append({"file": "lib/config-validator.js", "function": "validate10", "ccn": 373, "nloc": 1150, "params": 5, "start": 1, "end": 1150})
+        text = rendered(r, [], width=200)
+        hot = text[text.index("◆ Hotspots"):text.index("Change coupling")]
+        self.assertNotIn("config-validator", hot)
+        self.assertIn("1 generated file hidden; --full shows them", hot)
+        fn = _section_text(text, "Complex functions")
+        self.assertNotIn("validate10", fn)
+        self.assertIn("1 function in a generated file hidden; --full shows them", fn)
+        full = rendered(r, [], width=200, full=True)
+        self.assertIn("validate10", full)
+
+    def test_default_coupling_hides_release_plumbing_pairs_and_says_so(self):
+        r = sample_report()
+        for f in ("lib/version.rb", "contrib/version.rb", "Gemfile", "Gemfile.lock"):
+            r["size"]["files"][f] = {"code": 3, "complexity": 0}
+        r["coupling"] = [{"entity": "lib/version.rb", "coupled": "contrib/version.rb", "degree": 64, "average-revs": 60},
+                         {"entity": "Gemfile", "coupled": "Gemfile.lock", "degree": 90, "average-revs": 20},
+                         {"entity": "static/index.html", "coupled": "static/apps-metadata.json", "degree": 90, "average-revs": 11}]
+        coupling = _section_text(rendered(r, [], width=200), "Change coupling")
+        self.assertIn("static/index.html", coupling)
+        self.assertNotIn("version.rb", coupling)
+        self.assertIn("2 release pairs hidden; --full shows them", coupling)
+        full = _section_text(rendered(r, [], width=200, full=True), "Change coupling")
+        self.assertIn("version.rb", full)
+
     def test_hotspots_with_only_test_files_say_what_was_hidden(self):
         r = sample_report()
         r["revisions"] = [{"entity": "tests/test_a.py", "n-revs": 200}]
@@ -636,6 +667,25 @@ class KnowledgeMap(unittest.TestCase):
         r["ownership"] = []
         self.assertIn("no ownership data", rendered(r, []))
 
+    def test_default_map_hides_areas_no_longer_in_the_tree_and_says_so(self):
+        r = sample_report()   # the tree holds static/ files only; tests/t.py in the ownership rows is history
+        r["ownership"].append({"entity": "flask/app.py", "author": "Ann", "added": 4000, "deleted": 0})
+        km = _section_text(rendered(r, [], width=200), "Knowledge map")
+        self.assertIn("static/", km)
+        self.assertNotIn("flask/", km)
+        self.assertNotIn("tests/", km)
+        self.assertIn("2 historical areas hidden; --full shows them", km)
+        full = _section_text(rendered(r, [], width=200, full=True), "Knowledge map")
+        self.assertIn("flask/", full)
+        self.assertNotIn("hidden", full)
+
+    def test_map_without_a_tree_listing_hides_nothing(self):
+        r = sample_report()
+        r["size"]["files"] = {}
+        km = _section_text(rendered(r, [], width=200), "Knowledge map")
+        self.assertIn("tests/", km)
+        self.assertNotIn("historical", km)
+
 
 class Timeline(unittest.TestCase):
     def test_last_twelve_months_per_author_with_dots_for_zero(self):