lee/graphify
1
0
Fork 0
mirror of https://github.com/safishamsi/graphify.git synced 2026-08-27 16:56:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix hooks phantom dir on git < 2.31, save_manifest incremental data loss, cohesion rounding, C++ inheritance; add --resolution and --exclude-hubs

Browse Source 
- hooks.py: drop --path-format=absolute (added git 2.31), validate no newlines in path, anchor relative paths on repo root (#907)
- detect.py: seed save_manifest from existing manifest before loop so incremental callers don't erase untouched file entries (#917)
- cluster.py: drop round(..., 2) from cohesion_score so split threshold 0.05 fires correctly; add resolution param to _partition and cluster; add exclude_hubs_percentile to cluster with majority-vote reattachment (#919)
- report.py: format cohesion with :.2f for display
- __main__.py: wire --resolution and --exclude-hubs into extract and cluster-only commands (#919)
- C++ inheritance already written to disk by analysis agent (#915)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Safi
2026-05-17 23:05:48 +01:00
co-authored by Claude Sonnet 4.6
parent f7160c81c5
commit 2d783e569a
5 changed files with 113 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files
graphify/__main__.py
+23 -2
View File
@@ -1761,11 +1761,21 @@ def main() -> None:
args = sys.argv[2:]
watch_path: Path | None = None
graph_override: Path | None = None
co_resolution: float = 1.0
co_exclude_hubs: float | None = None
i_arg = 0
while i_arg < len(args):
a = args[i_arg]
if a == "--graph" and i_arg + 1 < len(args):
graph_override = Path(args[i_arg + 1]); i_arg += 2
elif a == "--resolution" and i_arg + 1 < len(args):
co_resolution = float(args[i_arg + 1]); i_arg += 2
elif a.startswith("--resolution="):
co_resolution = float(a.split("=", 1)[1]); i_arg += 1
elif a == "--exclude-hubs" and i_arg + 1 < len(args):
co_exclude_hubs = float(args[i_arg + 1]); i_arg += 2
elif a.startswith("--exclude-hubs="):
co_exclude_hubs = float(a.split("=", 1)[1]); i_arg += 1
elif a == "--no-viz" or a.startswith("--min-community-size="):
i_arg += 1
elif a.startswith("--"):
@@ -1792,7 +1802,7 @@ def main() -> None:
G = build_from_json(_raw, directed=_directed)
print(f"Graph: {G.number_of_nodes()} nodes, {G.number_of_edges()} edges")
print("Re-clustering...")
communities = cluster(G)
communities = cluster(G, resolution=co_resolution, exclude_hubs_percentile=co_exclude_hubs)
cohesion = score_all(G, communities)
gods = god_nodes(G)
surprises = surprising_connections(G, communities)
@@ -2415,6 +2425,9 @@ def main() -> None:
cli_token_budget: int | None = None
cli_max_concurrency: int | None = None
cli_api_timeout: float | None = None
# Clustering tuning knobs
cli_resolution: float = 1.0
cli_exclude_hubs: float | None = None
def _parse_int(name: str, raw: str) -> int:
try:
@@ -2480,6 +2493,14 @@ def main() -> None:
cli_api_timeout = _parse_float("--api-timeout", args[i + 1]); i += 2
elif a.startswith("--api-timeout="):
cli_api_timeout = _parse_float("--api-timeout", a.split("=", 1)[1]); i += 1
elif a == "--resolution" and i + 1 < len(args):
cli_resolution = _parse_float("--resolution", args[i + 1]); i += 2
elif a.startswith("--resolution="):
cli_resolution = _parse_float("--resolution", a.split("=", 1)[1]); i += 1
elif a == "--exclude-hubs" and i + 1 < len(args):
cli_exclude_hubs = float(args[i + 1]); i += 2
elif a.startswith("--exclude-hubs="):
cli_exclude_hubs = float(a.split("=", 1)[1]); i += 1
else:
i += 1
@@ -2796,7 +2817,7 @@ def main() -> None:
)
sys.exit(1)
communities = _cluster(G)
communities = _cluster(G, resolution=cli_resolution, exclude_hubs_percentile=cli_exclude_hubs)
cohesion = _score_all(G, communities)
try:
gods = _god_nodes(G)
graphify/cluster.py
+53 -7
View File
@@ -19,12 +19,15 @@ def _suppress_output():
return contextlib.redirect_stdout(io.StringIO())
def _partition(G: nx.Graph) -> dict[str, int]:
def _partition(G: nx.Graph, resolution: float = 1.0) -> dict[str, int]:
"""Run community detection. Returns {node_id: community_id}.
Tries Leiden (graspologic) first — best quality.
Falls back to Louvain (built into networkx) if graspologic is not installed.
resolution > 1.0 → more, smaller communities.
resolution < 1.0 → fewer, larger communities.
Output from graspologic is suppressed to prevent ANSI escape codes
from corrupting terminal scroll buffers on Windows PowerShell 5.1.
"""
@@ -49,6 +52,8 @@ def _partition(G: nx.Graph) -> dict[str, int]:
kwargs["random_seed"] = 42
if "trials" in lsig:
kwargs["trials"] = 1
if "resolution" in lsig:
kwargs["resolution"] = resolution
# Suppress graspologic output to prevent ANSI escape codes from
# corrupting PowerShell 5.1 scroll buffer (issue #19)
old_stderr = sys.stderr
@@ -65,7 +70,7 @@ def _partition(G: nx.Graph) -> dict[str, int]:
# Fallback: networkx louvain (available since networkx 2.7).
# Inspect kwargs to stay compatible across NetworkX versions — max_level
# was added in a later release and prevents hangs on large sparse graphs.
kwargs: dict = {"seed": 42, "threshold": 1e-4}
kwargs: dict = {"seed": 42, "threshold": 1e-4, "resolution": resolution}
if "max_level" in inspect.signature(nx.community.louvain_communities).parameters:
kwargs["max_level"] = 10
communities = nx.community.louvain_communities(stable, **kwargs)
@@ -78,7 +83,11 @@ _COHESION_SPLIT_THRESHOLD = 0.05 # re-split communities with cohesion below this
_COHESION_SPLIT_MIN_SIZE = 50 # only cohesion-split if community has at least this many nodes
def cluster(G: nx.Graph) -> dict[int, list[str]]:
def cluster(
G: nx.Graph,
resolution: float = 1.0,
exclude_hubs_percentile: float | None = None,
) -> dict[int, list[str]]:
"""Run Leiden community detection. Returns {community_id: [node_ids]}.
Community IDs are stable across runs: 0 = largest community after splitting.
@@ -87,6 +96,13 @@ def cluster(G: nx.Graph) -> dict[int, list[str]]:
Accepts directed or undirected graphs. DiGraphs are converted to undirected
internally since Louvain/Leiden require undirected input.
resolution: passed to Leiden/Louvain. >1.0 = more smaller communities,
<1.0 = fewer larger communities. Default 1.0.
exclude_hubs_percentile: if set (0-100), nodes whose degree exceeds this
percentile are excluded from partitioning and reattached to their
majority-vote neighbour community afterwards. Useful for staging/utility
super-hubs that inflate god-node rankings (#919).
"""
if G.number_of_nodes() == 0:
return {}
@@ -95,14 +111,26 @@ def cluster(G: nx.Graph) -> dict[int, list[str]]:
if G.number_of_edges() == 0:
return {i: [n] for i, n in enumerate(sorted(G.nodes))}
# Compute hub exclusion set before removing anything so degree is based on full graph
hub_nodes: set[str] = set()
if exclude_hubs_percentile is not None:
degrees = sorted(d for _, d in G.degree())
if degrees:
idx = max(0, int(len(degrees) * exclude_hubs_percentile / 100) - 1)
threshold = degrees[idx]
hub_nodes = {n for n, d in G.degree() if d > threshold}
# Leiden warns and drops isolates - handle them separately
isolates = [n for n in G.nodes() if G.degree(n) == 0]
connected_nodes = [n for n in G.nodes() if G.degree(n) > 0]
# Also exclude hub nodes from partitioning so they don't pull unrelated
# subsystems into the same community
excluded = hub_nodes
isolates = [n for n in G.nodes() if G.degree(n) == 0 and n not in excluded]
connected_nodes = [n for n in G.nodes() if G.degree(n) > 0 and n not in excluded]
connected = G.subgraph(connected_nodes)
raw: dict[int, list[str]] = {}
if connected.number_of_nodes() > 0:
partition = _partition(connected)
partition = _partition(connected, resolution=resolution)
for node, cid in partition.items():
raw.setdefault(cid, []).append(node)
@@ -112,6 +140,24 @@ def cluster(G: nx.Graph) -> dict[int, list[str]]:
raw[next_cid] = [node]
next_cid += 1
# Reattach excluded hubs by majority-vote neighbour community
if hub_nodes:
node_community: dict[str, int] = {n: cid for cid, nodes in raw.items() for n in nodes}
for hub in sorted(hub_nodes):
votes: dict[int, int] = {}
for nb in G.neighbors(hub):
cid = node_community.get(nb)
if cid is not None:
votes[cid] = votes.get(cid, 0) + 1
if votes:
best = min(votes, key=lambda c: (-votes[c], c))
raw.setdefault(best, []).append(hub)
node_community[hub] = best
else:
raw[next_cid] = [hub]
node_community[hub] = next_cid
next_cid += 1
# Split oversized communities
max_size = max(_MIN_SPLIT_SIZE, int(G.number_of_nodes() * _MAX_COMMUNITY_FRACTION))
final_communities: list[list[str]] = []
@@ -163,7 +209,7 @@ def cohesion_score(G: nx.Graph, community_nodes: list[str]) -> float:
subgraph = G.subgraph(community_nodes)
actual = subgraph.number_of_edges()
possible = n * (n - 1) / 2
return round(actual / possible, 2) if possible > 0 else 0.0
return actual / possible if possible > 0 else 0.0
def score_all(G: nx.Graph, communities: dict[int, list[str]]) -> dict[int, float]:
graphify/detect.py
+25 -6
View File
@@ -856,7 +856,31 @@ def save_manifest(
kind="both" — full pipeline: stamps both hashes (default).
"""
existing = load_manifest(manifest_path)
def _normalise_entry(entry):
if isinstance(entry, (int, float)):
return {"mtime": entry, "ast_hash": "", "semantic_hash": ""}
if isinstance(entry, dict) and "hash" in entry and "ast_hash" not in entry:
return {"mtime": entry.get("mtime", 0), "ast_hash": entry["hash"], "semantic_hash": ""}
if isinstance(entry, dict):
return entry
return None
# Seed from the existing manifest so incremental callers passing a subset
# of files don't silently erase entries for untouched files (#917).
# Prune entries whose file no longer exists on disk — those are genuine
# deletions that detect_incremental() should treat as gone.
manifest: dict[str, dict] = {}
for f, entry in existing.items():
normalised = _normalise_entry(entry)
if normalised is None:
continue
try:
if Path(f).exists():
manifest[f] = normalised
except OSError:
continue
for file_list in files.values():
for f in file_list:
try:
@@ -865,12 +889,7 @@ def save_manifest(
h = _md5_file(p)
except OSError:
continue # file deleted between detect() and manifest write
prev = existing.get(f, {})
# Normalise legacy {mtime, hash} entries to new schema
if isinstance(prev, (int, float)):
prev = {"mtime": prev, "ast_hash": "", "semantic_hash": ""}
elif isinstance(prev, dict) and "hash" in prev and "ast_hash" not in prev:
prev = {"mtime": prev.get("mtime", 0), "ast_hash": prev["hash"], "semantic_hash": ""}
prev = _normalise_entry(existing.get(f, {})) or {}
entry: dict = {"mtime": mtime}
if kind in ("ast", "both"):
entry["ast_hash"] = h
graphify/hooks.py
+11 -3
View File
@@ -199,14 +199,22 @@ def _hooks_dir(root: Path) -> Path:
)
# In a linked worktree .git is a file not a directory, so constructing
# root/.git/hooks directly fails. Ask git for the real hooks path instead.
# NOTE: do NOT pass --path-format=absolute — added in git 2.31; older git
# echoes it back as a literal argument, contaminating stdout and causing a
# phantom directory to be created (#907). git -C <root> already returns an
# absolute path for worktree/external-gitdir cases, and a path relative to
# <root> for normal repos — anchoring on root covers both.
import subprocess as _sp
try:
res = _sp.run(
["git", "-C", str(root), "rev-parse", "--path-format=absolute", "--git-path", "hooks"],
["git", "-C", str(root), "rev-parse", "--git-path", "hooks"],
capture_output=True, text=True,
)
if res.returncode == 0:
d = Path(res.stdout.strip())
raw = res.stdout.strip()
# A valid hooks path can never contain newlines or NUL. Their presence
# means git echoed an unrecognised flag back (old git behaviour).
if res.returncode == 0 and raw and not any(c in raw for c in ("\n", "\r", "\x00")):
d = (root / raw).resolve()
d.mkdir(parents=True, exist_ok=True)
return d
except (OSError, FileNotFoundError):
graphify/report.py
+1 -1
View File
@@ -144,7 +144,7 @@ def generate(
lines += [
"",
f"### Community {cid} - \"{label}\"",
f"Cohesion: {score}",
f"Cohesion: {score:.2f}",
f"Nodes ({len(real_nodes)}): {', '.join(display)}{suffix}",
]