Navegador

feat: incremental ingestion with content hashing and watch mode Files are tracked by SHA-256 content hash on File nodes. With --incremental, unchanged files are skipped and changed files have their subgraph cleared before re-parsing. --watch enables polling-based continuous re-ingestion. Closes #21

lmata 2026-03-23 04:49 trunk

Commit df3d6fa8b5d1133a8671472c424cd4dac096f1802d4fe00ae597017db4cfe302

Parent ece88cab8dc2f26…

6 files changed +25 -3 +12 +1 -1 +86 -4 +38 +148

~ navegador/cli/commands.py ~ navegador/graph/queries.py ~ navegador/graph/schema.py ~ navegador/ingestion/parser.py ~ tests/test_cli.py ~ tests/test_ingestion_code.py

M navegador/cli/commands.py

+25 -3

		--- navegador/cli/commands.py
		+++ navegador/cli/commands.py
		@@ -104,24 +104,46 @@
104	104
105	105	@main.command()
106	106	@click.argument("repo_path", type=click.Path(exists=True))
107	107	@DB_OPTION
108	108	@click.option("--clear", is_flag=True, help="Clear existing graph before ingesting.")
	109	+@click.option("--incremental", is_flag=True, help="Only re-parse changed files.")
	110	+@click.option("--watch", is_flag=True, help="Watch for changes and re-ingest incrementally.")
	111	+@click.option("--interval", default=2.0, show_default=True, help="Watch poll interval (seconds).")
109	112	@click.option("--json", "as_json", is_flag=True, help="Output stats as JSON.")
110		-def ingest(repo_path: str, db: str, clear: bool, as_json: bool):
	113	+def ingest(repo_path: str, db: str, clear: bool, incremental: bool, watch: bool,
	114	+ interval: float, as_json: bool):
111	115	"""Ingest a repository's code into the graph (AST + call graph)."""
112	116	from navegador.ingestion import RepoIngester
113	117
114	118	store = _get_store(db)
115	119	ingester = RepoIngester(store)
116	120
	121	+ if watch:
	122	+ console.print(f"[bold]Watching[/bold] {repo_path} (interval={interval}s, Ctrl-C to stop)")
	123	+
	124	+ def _on_cycle(stats):
	125	+ changed = stats["files"]
	126	+ skipped = stats["skipped"]
	127	+ if changed:
	128	+ console.print(
	129	+ f" [green]{changed} changed[/green], {skipped} unchanged"
	130	+ )
	131	+ return True # keep watching
	132	+
	133	+ try:
	134	+ ingester.watch(repo_path, interval=interval, callback=_on_cycle)
	135	+ except KeyboardInterrupt:
	136	+ console.print("\n[yellow]Watch stopped.[/yellow]")
	137	+ return
	138	+
117	139	if as_json:
118		- stats = ingester.ingest(repo_path, clear=clear)
	140	+ stats = ingester.ingest(repo_path, clear=clear, incremental=incremental)
119	141	click.echo(json.dumps(stats, indent=2))
120	142	else:
121	143	with console.status(f"[bold]Ingesting[/bold] {repo_path}..."):
122		- stats = ingester.ingest(repo_path, clear=clear)
	144	+ stats = ingester.ingest(repo_path, clear=clear, incremental=incremental)
123	145	table = Table(title="Ingestion complete")
124	146	table.add_column("Metric", style="cyan")
125	147	table.add_column("Count", justify="right", style="green")
126	148	for k, v in stats.items():
127	149	table.add_row(k.capitalize(), str(v))
128	150

	--- navegador/cli/commands.py
	+++ navegador/cli/commands.py
	@@ -104,24 +104,46 @@
104
105	@main.command()
106	@click.argument("repo_path", type=click.Path(exists=True))
107	@DB_OPTION
108	@click.option("--clear", is_flag=True, help="Clear existing graph before ingesting.")



109	@click.option("--json", "as_json", is_flag=True, help="Output stats as JSON.")
110	def ingest(repo_path: str, db: str, clear: bool, as_json: bool):

111	"""Ingest a repository's code into the graph (AST + call graph)."""
112	from navegador.ingestion import RepoIngester
113
114	store = _get_store(db)
115	ingester = RepoIngester(store)
116


















117	if as_json:
118	stats = ingester.ingest(repo_path, clear=clear)
119	click.echo(json.dumps(stats, indent=2))
120	else:
121	with console.status(f"[bold]Ingesting[/bold] {repo_path}..."):
122	stats = ingester.ingest(repo_path, clear=clear)
123	table = Table(title="Ingestion complete")
124	table.add_column("Metric", style="cyan")
125	table.add_column("Count", justify="right", style="green")
126	for k, v in stats.items():
127	table.add_row(k.capitalize(), str(v))
128

	--- navegador/cli/commands.py
	+++ navegador/cli/commands.py
	@@ -104,24 +104,46 @@
104
105	@main.command()
106	@click.argument("repo_path", type=click.Path(exists=True))
107	@DB_OPTION
108	@click.option("--clear", is_flag=True, help="Clear existing graph before ingesting.")
109	@click.option("--incremental", is_flag=True, help="Only re-parse changed files.")
110	@click.option("--watch", is_flag=True, help="Watch for changes and re-ingest incrementally.")
111	@click.option("--interval", default=2.0, show_default=True, help="Watch poll interval (seconds).")
112	@click.option("--json", "as_json", is_flag=True, help="Output stats as JSON.")
113	def ingest(repo_path: str, db: str, clear: bool, incremental: bool, watch: bool,
114	interval: float, as_json: bool):
115	"""Ingest a repository's code into the graph (AST + call graph)."""
116	from navegador.ingestion import RepoIngester
117
118	store = _get_store(db)
119	ingester = RepoIngester(store)
120
121	if watch:
122	console.print(f"[bold]Watching[/bold] {repo_path} (interval={interval}s, Ctrl-C to stop)")
123
124	def _on_cycle(stats):
125	changed = stats["files"]
126	skipped = stats["skipped"]
127	if changed:
128	console.print(
129	f" [green]{changed} changed[/green], {skipped} unchanged"
130	)
131	return True # keep watching
132
133	try:
134	ingester.watch(repo_path, interval=interval, callback=_on_cycle)
135	except KeyboardInterrupt:
136	console.print("\n[yellow]Watch stopped.[/yellow]")
137	return
138
139	if as_json:
140	stats = ingester.ingest(repo_path, clear=clear, incremental=incremental)
141	click.echo(json.dumps(stats, indent=2))
142	else:
143	with console.status(f"[bold]Ingesting[/bold] {repo_path}..."):
144	stats = ingester.ingest(repo_path, clear=clear, incremental=incremental)
145	table = Table(title="Ingestion complete")
146	table.add_column("Metric", style="cyan")
147	table.add_column("Count", justify="right", style="green")
148	for k, v in stats.items():
149	table.add_row(k.capitalize(), str(v))
150

M navegador/graph/queries.py

+12

		--- navegador/graph/queries.py
		+++ navegador/graph/queries.py
		@@ -209,10 +209,22 @@
209	209	MATCH (n)-[:ASSIGNED_TO]->(p:Person)
210	210	WHERE n.name = $name AND ($file_path = '' OR n.file_path = $file_path)
211	211	RETURN labels(n)[0] AS node_type, n.name AS node_name,
212	212	p.name AS owner, p.email AS email, p.role AS role, p.team AS team
213	213	"""
	214	+
	215	+# ── Incremental ingestion ─────────────────────────────────────────────────────
	216	+
	217	+FILE_HASH = """
	218	+MATCH (f:File {path: $path})
	219	+RETURN f.content_hash AS hash
	220	+"""
	221	+
	222	+DELETE_FILE_SUBGRAPH = """
	223	+MATCH (f:File {path: $path})-[:CONTAINS]->(child)
	224	+DETACH DELETE child
	225	+"""
214	226
215	227	# ── Stats ─────────────────────────────────────────────────────────────────────
216	228
217	229	NODE_TYPE_COUNTS = """
218	230	MATCH (n)
219	231

	--- navegador/graph/queries.py
	+++ navegador/graph/queries.py
	@@ -209,10 +209,22 @@
209	MATCH (n)-[:ASSIGNED_TO]->(p:Person)
210	WHERE n.name = $name AND ($file_path = '' OR n.file_path = $file_path)
211	RETURN labels(n)[0] AS node_type, n.name AS node_name,
212	p.name AS owner, p.email AS email, p.role AS role, p.team AS team
213	"""












214
215	# ── Stats ─────────────────────────────────────────────────────────────────────
216
217	NODE_TYPE_COUNTS = """
218	MATCH (n)
219

	--- navegador/graph/queries.py
	+++ navegador/graph/queries.py
	@@ -209,10 +209,22 @@
209	MATCH (n)-[:ASSIGNED_TO]->(p:Person)
210	WHERE n.name = $name AND ($file_path = '' OR n.file_path = $file_path)
211	RETURN labels(n)[0] AS node_type, n.name AS node_name,
212	p.name AS owner, p.email AS email, p.role AS role, p.team AS team
213	"""
214
215	# ── Incremental ingestion ─────────────────────────────────────────────────────
216
217	FILE_HASH = """
218	MATCH (f:File {path: $path})
219	RETURN f.content_hash AS hash
220	"""
221
222	DELETE_FILE_SUBGRAPH = """
223	MATCH (f:File {path: $path})-[:CONTAINS]->(child)
224	DETACH DELETE child
225	"""
226
227	# ── Stats ─────────────────────────────────────────────────────────────────────
228
229	NODE_TYPE_COUNTS = """
230	MATCH (n)
231

M navegador/graph/schema.py

+1 -1

		--- navegador/graph/schema.py
		+++ navegador/graph/schema.py
		@@ -60,11 +60,11 @@
60	60	# ── Property keys per node label ──────────────────────────────────────────────
61	61
62	62	NODE_PROPS = {
63	63	# Code layer
64	64	NodeLabel.Repository: ["name", "path", "language", "description"],
65		- NodeLabel.File: ["name", "path", "language", "size", "line_count"],
	65	+ NodeLabel.File: ["name", "path", "language", "size", "line_count", "content_hash"],
66	66	NodeLabel.Module: ["name", "file_path", "docstring"],
67	67	NodeLabel.Class: ["name", "file_path", "line_start", "line_end", "docstring", "source"],
68	68	NodeLabel.Function: [
69	69	"name",
70	70	"file_path",
71	71

	--- navegador/graph/schema.py
	+++ navegador/graph/schema.py
	@@ -60,11 +60,11 @@
60	# ── Property keys per node label ──────────────────────────────────────────────
61
62	NODE_PROPS = {
63	# Code layer
64	NodeLabel.Repository: ["name", "path", "language", "description"],
65	NodeLabel.File: ["name", "path", "language", "size", "line_count"],
66	NodeLabel.Module: ["name", "file_path", "docstring"],
67	NodeLabel.Class: ["name", "file_path", "line_start", "line_end", "docstring", "source"],
68	NodeLabel.Function: [
69	"name",
70	"file_path",
71

	--- navegador/graph/schema.py
	+++ navegador/graph/schema.py
	@@ -60,11 +60,11 @@
60	# ── Property keys per node label ──────────────────────────────────────────────
61
62	NODE_PROPS = {
63	# Code layer
64	NodeLabel.Repository: ["name", "path", "language", "description"],
65	NodeLabel.File: ["name", "path", "language", "size", "line_count", "content_hash"],
66	NodeLabel.Module: ["name", "file_path", "docstring"],
67	NodeLabel.Class: ["name", "file_path", "line_start", "line_end", "docstring", "source"],
68	NodeLabel.Function: [
69	"name",
70	"file_path",
71

M navegador/ingestion/parser.py

+86 -4

		--- navegador/ingestion/parser.py
		+++ navegador/ingestion/parser.py
		@@ -9,13 +9,16 @@
9	9	Go .go
10	10	Rust .rs
11	11	Java .java
12	12	"""
13	13
	14	+import hashlib
14	15	import logging
	16	+import time
15	17	from pathlib import Path
16	18
	19	+from navegador.graph import queries
17	20	from navegador.graph.schema import NodeLabel
18	21	from navegador.graph.store import GraphStore
19	22
20	23	logger = logging.getLogger(__name__)
21	24
		@@ -44,20 +47,26 @@
44	47
45	48	def __init__(self, store: GraphStore) -> None:
46	49	self.store = store
47	50	self._parsers: dict[str, "LanguageParser"] = {}
48	51
49		- def ingest(self, repo_path: str \| Path, clear: bool = False) -> dict[str, int]:
	52	+ def ingest(
	53	+ self,
	54	+ repo_path: str \| Path,
	55	+ clear: bool = False,
	56	+ incremental: bool = False,
	57	+ ) -> dict[str, int]:
50	58	"""
51	59	Ingest a repository into the graph.
52	60
53	61	Args:
54	62	repo_path: Path to the repository root.
55	63	clear: If True, wipe the graph before ingesting.
	64	+ incremental: If True, skip files whose content hash hasn't changed.
56	65
57	66	Returns:
58		- Dict with counts: files, functions, classes, edges.
	67	+ Dict with counts: files, functions, classes, edges, skipped.
59	68	"""
60	69	repo_path = Path(repo_path).resolve()
61	70	if not repo_path.exists():
62	71	raise FileNotFoundError(f"Repository not found: {repo_path}")
63	72
		@@ -71,34 +80,100 @@
71	80	"name": repo_path.name,
72	81	"path": str(repo_path),
73	82	},
74	83	)
75	84
76		- stats: dict[str, int] = {"files": 0, "functions": 0, "classes": 0, "edges": 0}
	85	+ stats: dict[str, int] = {
	86	+ "files": 0,
	87	+ "functions": 0,
	88	+ "classes": 0,
	89	+ "edges": 0,
	90	+ "skipped": 0,
	91	+ }
77	92
78	93	for source_file in self._iter_source_files(repo_path):
79	94	language = LANGUAGE_MAP.get(source_file.suffix)
80	95	if not language:
81	96	continue
	97	+
	98	+ rel_path = str(source_file.relative_to(repo_path))
	99	+ content_hash = _file_hash(source_file)
	100	+
	101	+ if incremental and self._file_unchanged(rel_path, content_hash):
	102	+ stats["skipped"] += 1
	103	+ continue
	104	+
	105	+ if incremental:
	106	+ self._clear_file_subgraph(rel_path)
	107	+
82	108	try:
83	109	parser = self._get_parser(language)
84	110	file_stats = parser.parse_file(source_file, repo_path, self.store)
85	111	stats["files"] += 1
86	112	stats["functions"] += file_stats.get("functions", 0)
87	113	stats["classes"] += file_stats.get("classes", 0)
88	114	stats["edges"] += file_stats.get("edges", 0)
	115	+
	116	+ self._store_file_hash(rel_path, content_hash)
89	117	except Exception:
90	118	logger.exception("Failed to parse %s", source_file)
91	119
92	120	logger.info(
93		- "Ingested %s: %d files, %d functions, %d classes",
	121	+ "Ingested %s: %d files, %d functions, %d classes, %d skipped",
94	122	repo_path.name,
95	123	stats["files"],
96	124	stats["functions"],
97	125	stats["classes"],
	126	+ stats["skipped"],
98	127	)
99	128	return stats
	129	+
	130	+ def watch(
	131	+ self,
	132	+ repo_path: str \| Path,
	133	+ interval: float = 2.0,
	134	+ callback=None,
	135	+ ) -> None:
	136	+ """
	137	+ Watch a repo for changes and re-ingest incrementally.
	138	+
	139	+ Args:
	140	+ repo_path: Path to the repository root.
	141	+ interval: Seconds between polls.
	142	+ callback: Optional callable receiving stats dict after each cycle.
	143	+ If callback returns False, the watch loop stops.
	144	+ """
	145	+ repo_path = Path(repo_path).resolve()
	146	+ if not repo_path.exists():
	147	+ raise FileNotFoundError(f"Repository not found: {repo_path}")
	148	+
	149	+ # Initial full ingest
	150	+ stats = self.ingest(repo_path, incremental=True)
	151	+ if callback and callback(stats) is False:
	152	+ return
	153	+
	154	+ while True:
	155	+ time.sleep(interval)
	156	+ stats = self.ingest(repo_path, incremental=True)
	157	+ if callback and callback(stats) is False:
	158	+ return
	159	+
	160	+ def _file_unchanged(self, rel_path: str, content_hash: str) -> bool:
	161	+ result = self.store.query(queries.FILE_HASH, {"path": rel_path})
	162	+ rows = result.result_set or []
	163	+ if not rows or rows[0][0] is None:
	164	+ return False
	165	+ return rows[0][0] == content_hash
	166	+
	167	+ def _clear_file_subgraph(self, rel_path: str) -> None:
	168	+ self.store.query(queries.DELETE_FILE_SUBGRAPH, {"path": rel_path})
	169	+
	170	+ def _store_file_hash(self, rel_path: str, content_hash: str) -> None:
	171	+ self.store.query(
	172	+ "MATCH (f:File {path: $path}) SET f.content_hash = $hash",
	173	+ {"path": rel_path, "hash": content_hash},
	174	+ )
100	175
101	176	def _iter_source_files(self, repo_path: Path):
102	177	skip_dirs = {
103	178	".git",
104	179	".venv",
		@@ -141,11 +216,18 @@
141	216	self._parsers[language] = JavaParser()
142	217	else:
143	218	raise ValueError(f"Unsupported language: {language}")
144	219	return self._parsers[language]
145	220
	221	+
	222	+def _file_hash(path: Path) -> str:
	223	+ """SHA-256 content hash for a file."""
	224	+ h = hashlib.sha256()
	225	+ h.update(path.read_bytes())
	226	+ return h.hexdigest()
	227	+
146	228
147	229	class LanguageParser:
148	230	"""Base class for language-specific AST parsers."""
149	231
150	232	def parse_file(self, path: Path, repo_root: Path, store: GraphStore) -> dict[str, int]:
151	233	raise NotImplementedError
152	234

	--- navegador/ingestion/parser.py
	+++ navegador/ingestion/parser.py
	@@ -9,13 +9,16 @@
9	Go .go
10	Rust .rs
11	Java .java
12	"""
13

14	import logging

15	from pathlib import Path
16

17	from navegador.graph.schema import NodeLabel
18	from navegador.graph.store import GraphStore
19
20	logger = logging.getLogger(__name__)
21
	@@ -44,20 +47,26 @@
44
45	def __init__(self, store: GraphStore) -> None:
46	self.store = store
47	self._parsers: dict[str, "LanguageParser"] = {}
48
49	def ingest(self, repo_path: str \| Path, clear: bool = False) -> dict[str, int]:





50	"""
51	Ingest a repository into the graph.
52
53	Args:
54	repo_path: Path to the repository root.
55	clear: If True, wipe the graph before ingesting.

56
57	Returns:
58	Dict with counts: files, functions, classes, edges.
59	"""
60	repo_path = Path(repo_path).resolve()
61	if not repo_path.exists():
62	raise FileNotFoundError(f"Repository not found: {repo_path}")
63
	@@ -71,34 +80,100 @@
71	"name": repo_path.name,
72	"path": str(repo_path),
73	},
74	)
75
76	stats: dict[str, int] = {"files": 0, "functions": 0, "classes": 0, "edges": 0}






77
78	for source_file in self._iter_source_files(repo_path):
79	language = LANGUAGE_MAP.get(source_file.suffix)
80	if not language:
81	continue











82	try:
83	parser = self._get_parser(language)
84	file_stats = parser.parse_file(source_file, repo_path, self.store)
85	stats["files"] += 1
86	stats["functions"] += file_stats.get("functions", 0)
87	stats["classes"] += file_stats.get("classes", 0)
88	stats["edges"] += file_stats.get("edges", 0)


89	except Exception:
90	logger.exception("Failed to parse %s", source_file)
91
92	logger.info(
93	"Ingested %s: %d files, %d functions, %d classes",
94	repo_path.name,
95	stats["files"],
96	stats["functions"],
97	stats["classes"],

98	)
99	return stats














































100
101	def _iter_source_files(self, repo_path: Path):
102	skip_dirs = {
103	".git",
104	".venv",
	@@ -141,11 +216,18 @@
141	self._parsers[language] = JavaParser()
142	else:
143	raise ValueError(f"Unsupported language: {language}")
144	return self._parsers[language]
145







146
147	class LanguageParser:
148	"""Base class for language-specific AST parsers."""
149
150	def parse_file(self, path: Path, repo_root: Path, store: GraphStore) -> dict[str, int]:
151	raise NotImplementedError
152

	--- navegador/ingestion/parser.py
	+++ navegador/ingestion/parser.py
	@@ -9,13 +9,16 @@
9	Go .go
10	Rust .rs
11	Java .java
12	"""
13
14	import hashlib
15	import logging
16	import time
17	from pathlib import Path
18
19	from navegador.graph import queries
20	from navegador.graph.schema import NodeLabel
21	from navegador.graph.store import GraphStore
22
23	logger = logging.getLogger(__name__)
24
	@@ -44,20 +47,26 @@
47
48	def __init__(self, store: GraphStore) -> None:
49	self.store = store
50	self._parsers: dict[str, "LanguageParser"] = {}
51
52	def ingest(
53	self,
54	repo_path: str \| Path,
55	clear: bool = False,
56	incremental: bool = False,
57	) -> dict[str, int]:
58	"""
59	Ingest a repository into the graph.
60
61	Args:
62	repo_path: Path to the repository root.
63	clear: If True, wipe the graph before ingesting.
64	incremental: If True, skip files whose content hash hasn't changed.
65
66	Returns:
67	Dict with counts: files, functions, classes, edges, skipped.
68	"""
69	repo_path = Path(repo_path).resolve()
70	if not repo_path.exists():
71	raise FileNotFoundError(f"Repository not found: {repo_path}")
72
	@@ -71,34 +80,100 @@
80	"name": repo_path.name,
81	"path": str(repo_path),
82	},
83	)
84
85	stats: dict[str, int] = {
86	"files": 0,
87	"functions": 0,
88	"classes": 0,
89	"edges": 0,
90	"skipped": 0,
91	}
92
93	for source_file in self._iter_source_files(repo_path):
94	language = LANGUAGE_MAP.get(source_file.suffix)
95	if not language:
96	continue
97
98	rel_path = str(source_file.relative_to(repo_path))
99	content_hash = _file_hash(source_file)
100
101	if incremental and self._file_unchanged(rel_path, content_hash):
102	stats["skipped"] += 1
103	continue
104
105	if incremental:
106	self._clear_file_subgraph(rel_path)
107
108	try:
109	parser = self._get_parser(language)
110	file_stats = parser.parse_file(source_file, repo_path, self.store)
111	stats["files"] += 1
112	stats["functions"] += file_stats.get("functions", 0)
113	stats["classes"] += file_stats.get("classes", 0)
114	stats["edges"] += file_stats.get("edges", 0)
115
116	self._store_file_hash(rel_path, content_hash)
117	except Exception:
118	logger.exception("Failed to parse %s", source_file)
119
120	logger.info(
121	"Ingested %s: %d files, %d functions, %d classes, %d skipped",
122	repo_path.name,
123	stats["files"],
124	stats["functions"],
125	stats["classes"],
126	stats["skipped"],
127	)
128	return stats
129
130	def watch(
131	self,
132	repo_path: str \| Path,
133	interval: float = 2.0,
134	callback=None,
135	) -> None:
136	"""
137	Watch a repo for changes and re-ingest incrementally.
138
139	Args:
140	repo_path: Path to the repository root.
141	interval: Seconds between polls.
142	callback: Optional callable receiving stats dict after each cycle.
143	If callback returns False, the watch loop stops.
144	"""
145	repo_path = Path(repo_path).resolve()
146	if not repo_path.exists():
147	raise FileNotFoundError(f"Repository not found: {repo_path}")
148
149	# Initial full ingest
150	stats = self.ingest(repo_path, incremental=True)
151	if callback and callback(stats) is False:
152	return
153
154	while True:
155	time.sleep(interval)
156	stats = self.ingest(repo_path, incremental=True)
157	if callback and callback(stats) is False:
158	return
159
160	def _file_unchanged(self, rel_path: str, content_hash: str) -> bool:
161	result = self.store.query(queries.FILE_HASH, {"path": rel_path})
162	rows = result.result_set or []
163	if not rows or rows[0][0] is None:
164	return False
165	return rows[0][0] == content_hash
166
167	def _clear_file_subgraph(self, rel_path: str) -> None:
168	self.store.query(queries.DELETE_FILE_SUBGRAPH, {"path": rel_path})
169
170	def _store_file_hash(self, rel_path: str, content_hash: str) -> None:
171	self.store.query(
172	"MATCH (f:File {path: $path}) SET f.content_hash = $hash",
173	{"path": rel_path, "hash": content_hash},
174	)
175
176	def _iter_source_files(self, repo_path: Path):
177	skip_dirs = {
178	".git",
179	".venv",
	@@ -141,11 +216,18 @@
216	self._parsers[language] = JavaParser()
217	else:
218	raise ValueError(f"Unsupported language: {language}")
219	return self._parsers[language]
220
221
222	def _file_hash(path: Path) -> str:
223	"""SHA-256 content hash for a file."""
224	h = hashlib.sha256()
225	h.update(path.read_bytes())
226	return h.hexdigest()
227
228
229	class LanguageParser:
230	"""Base class for language-specific AST parsers."""
231
232	def parse_file(self, path: Path, repo_root: Path, store: GraphStore) -> dict[str, int]:
233	raise NotImplementedError
234

M tests/test_cli.py

+38

		--- tests/test_cli.py
		+++ tests/test_cli.py
		@@ -73,10 +73,48 @@
73	73	MockRI.return_value.ingest.return_value = {"files": 5}
74	74	result = runner.invoke(main, ["ingest", "src", "--json"])
75	75	assert result.exit_code == 0
76	76	data = json.loads(result.output)
77	77	assert data["files"] == 5
	78	+
	79	+ def test_incremental_flag_passes_through(self):
	80	+ runner = CliRunner()
	81	+ with runner.isolated_filesystem():
	82	+ Path("src").mkdir()
	83	+ with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
	84	+ patch("navegador.ingestion.RepoIngester") as MockRI:
	85	+ MockRI.return_value.ingest.return_value = {
	86	+ "files": 2, "functions": 5, "classes": 1, "edges": 3, "skipped": 8
	87	+ }
	88	+ result = runner.invoke(main, ["ingest", "src", "--incremental"])
	89	+ assert result.exit_code == 0
	90	+ MockRI.return_value.ingest.assert_called_once()
	91	+ _, kwargs = MockRI.return_value.ingest.call_args
	92	+ assert kwargs["incremental"] is True
	93	+
	94	+ def test_watch_flag_calls_watch(self):
	95	+ runner = CliRunner()
	96	+ with runner.isolated_filesystem():
	97	+ Path("src").mkdir()
	98	+ with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
	99	+ patch("navegador.ingestion.RepoIngester") as MockRI:
	100	+ # watch should be called, simulate immediate stop
	101	+ MockRI.return_value.watch.side_effect = KeyboardInterrupt()
	102	+ result = runner.invoke(main, ["ingest", "src", "--watch", "--interval", "0.1"])
	103	+ assert result.exit_code == 0
	104	+ MockRI.return_value.watch.assert_called_once()
	105	+
	106	+ def test_watch_with_interval(self):
	107	+ runner = CliRunner()
	108	+ with runner.isolated_filesystem():
	109	+ Path("src").mkdir()
	110	+ with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
	111	+ patch("navegador.ingestion.RepoIngester") as MockRI:
	112	+ MockRI.return_value.watch.side_effect = KeyboardInterrupt()
	113	+ runner.invoke(main, ["ingest", "src", "--watch", "--interval", "5.0"])
	114	+ _, kwargs = MockRI.return_value.watch.call_args
	115	+ assert kwargs["interval"] == 5.0
78	116
79	117
80	118	# ── context ───────────────────────────────────────────────────────────────────
81	119
82	120	class TestContextCommand:
83	121

	--- tests/test_cli.py
	+++ tests/test_cli.py
	@@ -73,10 +73,48 @@
73	MockRI.return_value.ingest.return_value = {"files": 5}
74	result = runner.invoke(main, ["ingest", "src", "--json"])
75	assert result.exit_code == 0
76	data = json.loads(result.output)
77	assert data["files"] == 5






































78
79
80	# ── context ───────────────────────────────────────────────────────────────────
81
82	class TestContextCommand:
83

	--- tests/test_cli.py
	+++ tests/test_cli.py
	@@ -73,10 +73,48 @@
73	MockRI.return_value.ingest.return_value = {"files": 5}
74	result = runner.invoke(main, ["ingest", "src", "--json"])
75	assert result.exit_code == 0
76	data = json.loads(result.output)
77	assert data["files"] == 5
78
79	def test_incremental_flag_passes_through(self):
80	runner = CliRunner()
81	with runner.isolated_filesystem():
82	Path("src").mkdir()
83	with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
84	patch("navegador.ingestion.RepoIngester") as MockRI:
85	MockRI.return_value.ingest.return_value = {
86	"files": 2, "functions": 5, "classes": 1, "edges": 3, "skipped": 8
87	}
88	result = runner.invoke(main, ["ingest", "src", "--incremental"])
89	assert result.exit_code == 0
90	MockRI.return_value.ingest.assert_called_once()
91	_, kwargs = MockRI.return_value.ingest.call_args
92	assert kwargs["incremental"] is True
93
94	def test_watch_flag_calls_watch(self):
95	runner = CliRunner()
96	with runner.isolated_filesystem():
97	Path("src").mkdir()
98	with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
99	patch("navegador.ingestion.RepoIngester") as MockRI:
100	# watch should be called, simulate immediate stop
101	MockRI.return_value.watch.side_effect = KeyboardInterrupt()
102	result = runner.invoke(main, ["ingest", "src", "--watch", "--interval", "0.1"])
103	assert result.exit_code == 0
104	MockRI.return_value.watch.assert_called_once()
105
106	def test_watch_with_interval(self):
107	runner = CliRunner()
108	with runner.isolated_filesystem():
109	Path("src").mkdir()
110	with patch("navegador.cli.commands._get_store", return_value=_mock_store()), \
111	patch("navegador.ingestion.RepoIngester") as MockRI:
112	MockRI.return_value.watch.side_effect = KeyboardInterrupt()
113	runner.invoke(main, ["ingest", "src", "--watch", "--interval", "5.0"])
114	_, kwargs = MockRI.return_value.watch.call_args
115	assert kwargs["interval"] == 5.0
116
117
118	# ── context ───────────────────────────────────────────────────────────────────
119
120	class TestContextCommand:
121

M tests/test_ingestion_code.py

+148

		--- tests/test_ingestion_code.py
		+++ tests/test_ingestion_code.py
		@@ -345,10 +345,158 @@
345	345	stats = ingester.ingest(tmpdir)
346	346	assert stats["files"] == 0
347	347
348	348
349	349	# ── LanguageParser base class ─────────────────────────────────────────────────
	350	+
	351	+# ── Incremental ingestion ─────────────────────────────────────────────────────
	352	+
	353	+class TestIncrementalIngestion:
	354	+ def test_incremental_returns_skipped_count(self):
	355	+ store = _make_store()
	356	+ ingester = RepoIngester(store)
	357	+ with tempfile.TemporaryDirectory() as tmpdir:
	358	+ stats = ingester.ingest(tmpdir, incremental=True)
	359	+ assert "skipped" in stats
	360	+
	361	+ def test_incremental_skips_unchanged_file(self):
	362	+ store = _make_store()
	363	+ ingester = RepoIngester(store)
	364	+ mock_parser = MagicMock()
	365	+ mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
	366	+ ingester._parsers["python"] = mock_parser
	367	+
	368	+ with tempfile.TemporaryDirectory() as tmpdir:
	369	+ py_file = Path(tmpdir) / "app.py"
	370	+ py_file.write_text("def foo(): pass")
	371	+
	372	+ # First ingest: file is new, should be parsed
	373	+ stats1 = ingester.ingest(tmpdir, incremental=True)
	374	+ assert stats1["files"] == 1
	375	+ assert stats1["skipped"] == 0
	376	+
	377	+ # Simulate stored hash matching
	378	+ from navegador.ingestion.parser import _file_hash
	379	+ current_hash = _file_hash(py_file)
	380	+ rel_path = "app.py"
	381	+
	382	+ # Mock _file_unchanged to return True
	383	+ ingester._file_unchanged = MagicMock(return_value=True)
	384	+ stats2 = ingester.ingest(tmpdir, incremental=True)
	385	+ assert stats2["files"] == 0
	386	+ assert stats2["skipped"] == 1
	387	+
	388	+ def test_incremental_reparses_changed_file(self):
	389	+ store = _make_store()
	390	+ ingester = RepoIngester(store)
	391	+ mock_parser = MagicMock()
	392	+ mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
	393	+ ingester._parsers["python"] = mock_parser
	394	+
	395	+ with tempfile.TemporaryDirectory() as tmpdir:
	396	+ py_file = Path(tmpdir) / "app.py"
	397	+ py_file.write_text("def foo(): pass")
	398	+
	399	+ ingester._file_unchanged = MagicMock(return_value=False)
	400	+ ingester._clear_file_subgraph = MagicMock()
	401	+ stats = ingester.ingest(tmpdir, incremental=True)
	402	+ assert stats["files"] == 1
	403	+ ingester._clear_file_subgraph.assert_called_once()
	404	+
	405	+ def test_non_incremental_does_not_check_hash(self):
	406	+ store = _make_store()
	407	+ ingester = RepoIngester(store)
	408	+ mock_parser = MagicMock()
	409	+ mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
	410	+ ingester._parsers["python"] = mock_parser
	411	+
	412	+ with tempfile.TemporaryDirectory() as tmpdir:
	413	+ (Path(tmpdir) / "app.py").write_text("def foo(): pass")
	414	+ ingester._file_unchanged = MagicMock()
	415	+ ingester.ingest(tmpdir, incremental=False)
	416	+ ingester._file_unchanged.assert_not_called()
	417	+
	418	+ def test_file_hash_is_deterministic(self):
	419	+ from navegador.ingestion.parser import _file_hash
	420	+ with tempfile.TemporaryDirectory() as tmpdir:
	421	+ f = Path(tmpdir) / "test.py"
	422	+ f.write_text("x = 1")
	423	+ h1 = _file_hash(f)
	424	+ h2 = _file_hash(f)
	425	+ assert h1 == h2
	426	+ assert len(h1) == 64 # SHA-256 hex
	427	+
	428	+ def test_file_hash_changes_on_content_change(self):
	429	+ from navegador.ingestion.parser import _file_hash
	430	+ with tempfile.TemporaryDirectory() as tmpdir:
	431	+ f = Path(tmpdir) / "test.py"
	432	+ f.write_text("x = 1")
	433	+ h1 = _file_hash(f)
	434	+ f.write_text("x = 2")
	435	+ h2 = _file_hash(f)
	436	+ assert h1 != h2
	437	+
	438	+
	439	+class TestFileUnchanged:
	440	+ def test_returns_false_for_new_file(self):
	441	+ store = _make_store()
	442	+ store.query.return_value = MagicMock(result_set=[])
	443	+ ingester = RepoIngester(store)
	444	+ assert ingester._file_unchanged("app.py", "abc123") is False
	445	+
	446	+ def test_returns_false_for_null_hash(self):
	447	+ store = _make_store()
	448	+ store.query.return_value = MagicMock(result_set=[[None]])
	449	+ ingester = RepoIngester(store)
	450	+ assert ingester._file_unchanged("app.py", "abc123") is False
	451	+
	452	+ def test_returns_true_when_hash_matches(self):
	453	+ store = _make_store()
	454	+ store.query.return_value = MagicMock(result_set=[["abc123"]])
	455	+ ingester = RepoIngester(store)
	456	+ assert ingester._file_unchanged("app.py", "abc123") is True
	457	+
	458	+ def test_returns_false_when_hash_differs(self):
	459	+ store = _make_store()
	460	+ store.query.return_value = MagicMock(result_set=[["old_hash"]])
	461	+ ingester = RepoIngester(store)
	462	+ assert ingester._file_unchanged("app.py", "new_hash") is False
	463	+
	464	+
	465	+class TestWatch:
	466	+ def test_watch_raises_on_missing_dir(self):
	467	+ store = _make_store()
	468	+ ingester = RepoIngester(store)
	469	+ with pytest.raises(FileNotFoundError):
	470	+ ingester.watch("/nonexistent/repo")
	471	+
	472	+ def test_watch_calls_callback_and_stops_on_false(self):
	473	+ store = _make_store()
	474	+ ingester = RepoIngester(store)
	475	+ call_count = [0]
	476	+
	477	+ def callback(stats):
	478	+ call_count[0] += 1
	479	+ return False # stop immediately
	480	+
	481	+ with tempfile.TemporaryDirectory() as tmpdir:
	482	+ ingester.watch(tmpdir, interval=0.01, callback=callback)
	483	+ assert call_count[0] == 1
	484	+
	485	+ def test_watch_runs_multiple_cycles(self):
	486	+ store = _make_store()
	487	+ ingester = RepoIngester(store)
	488	+ call_count = [0]
	489	+
	490	+ def callback(stats):
	491	+ call_count[0] += 1
	492	+ return call_count[0] < 3 # run 3 times then stop
	493	+
	494	+ with tempfile.TemporaryDirectory() as tmpdir:
	495	+ ingester.watch(tmpdir, interval=0.01, callback=callback)
	496	+ assert call_count[0] == 3
	497	+
350	498
351	499	class TestLanguageParserBase:
352	500	def test_parse_file_raises_not_implemented(self):
353	501	from pathlib import Path
354	502
355	503

	--- tests/test_ingestion_code.py
	+++ tests/test_ingestion_code.py
	@@ -345,10 +345,158 @@
345	stats = ingester.ingest(tmpdir)
346	assert stats["files"] == 0
347
348
349	# ── LanguageParser base class ─────────────────────────────────────────────────




















































































































































350
351	class TestLanguageParserBase:
352	def test_parse_file_raises_not_implemented(self):
353	from pathlib import Path
354
355

	--- tests/test_ingestion_code.py
	+++ tests/test_ingestion_code.py
	@@ -345,10 +345,158 @@
345	stats = ingester.ingest(tmpdir)
346	assert stats["files"] == 0
347
348
349	# ── LanguageParser base class ─────────────────────────────────────────────────
350
351	# ── Incremental ingestion ─────────────────────────────────────────────────────
352
353	class TestIncrementalIngestion:
354	def test_incremental_returns_skipped_count(self):
355	store = _make_store()
356	ingester = RepoIngester(store)
357	with tempfile.TemporaryDirectory() as tmpdir:
358	stats = ingester.ingest(tmpdir, incremental=True)
359	assert "skipped" in stats
360
361	def test_incremental_skips_unchanged_file(self):
362	store = _make_store()
363	ingester = RepoIngester(store)
364	mock_parser = MagicMock()
365	mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
366	ingester._parsers["python"] = mock_parser
367
368	with tempfile.TemporaryDirectory() as tmpdir:
369	py_file = Path(tmpdir) / "app.py"
370	py_file.write_text("def foo(): pass")
371
372	# First ingest: file is new, should be parsed
373	stats1 = ingester.ingest(tmpdir, incremental=True)
374	assert stats1["files"] == 1
375	assert stats1["skipped"] == 0
376
377	# Simulate stored hash matching
378	from navegador.ingestion.parser import _file_hash
379	current_hash = _file_hash(py_file)
380	rel_path = "app.py"
381
382	# Mock _file_unchanged to return True
383	ingester._file_unchanged = MagicMock(return_value=True)
384	stats2 = ingester.ingest(tmpdir, incremental=True)
385	assert stats2["files"] == 0
386	assert stats2["skipped"] == 1
387
388	def test_incremental_reparses_changed_file(self):
389	store = _make_store()
390	ingester = RepoIngester(store)
391	mock_parser = MagicMock()
392	mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
393	ingester._parsers["python"] = mock_parser
394
395	with tempfile.TemporaryDirectory() as tmpdir:
396	py_file = Path(tmpdir) / "app.py"
397	py_file.write_text("def foo(): pass")
398
399	ingester._file_unchanged = MagicMock(return_value=False)
400	ingester._clear_file_subgraph = MagicMock()
401	stats = ingester.ingest(tmpdir, incremental=True)
402	assert stats["files"] == 1
403	ingester._clear_file_subgraph.assert_called_once()
404
405	def test_non_incremental_does_not_check_hash(self):
406	store = _make_store()
407	ingester = RepoIngester(store)
408	mock_parser = MagicMock()
409	mock_parser.parse_file.return_value = {"functions": 1, "classes": 0, "edges": 0}
410	ingester._parsers["python"] = mock_parser
411
412	with tempfile.TemporaryDirectory() as tmpdir:
413	(Path(tmpdir) / "app.py").write_text("def foo(): pass")
414	ingester._file_unchanged = MagicMock()
415	ingester.ingest(tmpdir, incremental=False)
416	ingester._file_unchanged.assert_not_called()
417
418	def test_file_hash_is_deterministic(self):
419	from navegador.ingestion.parser import _file_hash
420	with tempfile.TemporaryDirectory() as tmpdir:
421	f = Path(tmpdir) / "test.py"
422	f.write_text("x = 1")
423	h1 = _file_hash(f)
424	h2 = _file_hash(f)
425	assert h1 == h2
426	assert len(h1) == 64 # SHA-256 hex
427
428	def test_file_hash_changes_on_content_change(self):
429	from navegador.ingestion.parser import _file_hash
430	with tempfile.TemporaryDirectory() as tmpdir:
431	f = Path(tmpdir) / "test.py"
432	f.write_text("x = 1")
433	h1 = _file_hash(f)
434	f.write_text("x = 2")
435	h2 = _file_hash(f)
436	assert h1 != h2
437
438
439	class TestFileUnchanged:
440	def test_returns_false_for_new_file(self):
441	store = _make_store()
442	store.query.return_value = MagicMock(result_set=[])
443	ingester = RepoIngester(store)
444	assert ingester._file_unchanged("app.py", "abc123") is False
445
446	def test_returns_false_for_null_hash(self):
447	store = _make_store()
448	store.query.return_value = MagicMock(result_set=[[None]])
449	ingester = RepoIngester(store)
450	assert ingester._file_unchanged("app.py", "abc123") is False
451
452	def test_returns_true_when_hash_matches(self):
453	store = _make_store()
454	store.query.return_value = MagicMock(result_set=[["abc123"]])
455	ingester = RepoIngester(store)
456	assert ingester._file_unchanged("app.py", "abc123") is True
457
458	def test_returns_false_when_hash_differs(self):
459	store = _make_store()
460	store.query.return_value = MagicMock(result_set=[["old_hash"]])
461	ingester = RepoIngester(store)
462	assert ingester._file_unchanged("app.py", "new_hash") is False
463
464
465	class TestWatch:
466	def test_watch_raises_on_missing_dir(self):
467	store = _make_store()
468	ingester = RepoIngester(store)
469	with pytest.raises(FileNotFoundError):
470	ingester.watch("/nonexistent/repo")
471
472	def test_watch_calls_callback_and_stops_on_false(self):
473	store = _make_store()
474	ingester = RepoIngester(store)
475	call_count = [0]
476
477	def callback(stats):
478	call_count[0] += 1
479	return False # stop immediately
480
481	with tempfile.TemporaryDirectory() as tmpdir:
482	ingester.watch(tmpdir, interval=0.01, callback=callback)
483	assert call_count[0] == 1
484
485	def test_watch_runs_multiple_cycles(self):
486	store = _make_store()
487	ingester = RepoIngester(store)
488	call_count = [0]
489
490	def callback(stats):
491	call_count[0] += 1
492	return call_count[0] < 3 # run 3 times then stop
493
494	with tempfile.TemporaryDirectory() as tmpdir:
495	ingester.watch(tmpdir, interval=0.01, callback=callback)
496	assert call_count[0] == 3
497
498
499	class TestLanguageParserBase:
500	def test_parse_file_raises_not_implemented(self):
501	from pathlib import Path
502
503

Navegador

Keyboard Shortcuts