solpbc.org
personal memory agent
1# SPDX-License-Identifier: AGPL-3.0-only
2# Copyright (c) 2026 sol pbc
3
4"""CLI functionality for the indexer."""
5
6import argparse
7from typing import Any
8
9from think.utils import get_journal, journal_log, setup_cli
10
11from .journal import (
12 index_file,
13 reset_journal_index,
14 scan_journal,
15 search_counts,
16 search_journal,
17)
18
19
20def _format_count_column(
21 items: list[tuple[str, int]], total: int, top_n: int
22) -> list[str]:
23 """Format a column of count items with overflow indicator."""
24 lines = [f"{name} ({count})" for name, count in items[:top_n]]
25 if total > top_n:
26 lines.append(f"... +{total - top_n} more")
27 return lines
28
29
30def _display_counts(counts: dict[str, Any], top_n: int = 5) -> None:
31 """Display aggregated counts in a compact table format."""
32 total = counts["total"]
33 facets = counts["facets"] # Counter
34 agents = counts["agents"] # Counter
35 days = counts["days"] # Counter
36
37 print(f"Total: {total:,} chunks\n")
38
39 # Build columns
40 facet_col = _format_count_column(facets.most_common(top_n), len(facets), top_n)
41 agent_col = _format_count_column(agents.most_common(top_n), len(agents), top_n)
42 day_col = _format_count_column(
43 sorted(days.items(), reverse=True)[:top_n], len(days), top_n
44 )
45
46 # Header and rows
47 print(f"{'Facet':<20} {'Agent':<20} {'Day':<20}")
48 print("-" * 60)
49
50 from itertools import zip_longest
51
52 for f, a, d in zip_longest(facet_col, agent_col, day_col, fillvalue=""):
53 print(f"{f:<20} {a:<20} {d:<20}")
54
55 print()
56
57
58def _display_search_results(
59 results: list[dict[str, Any]], total: int, offset: int
60) -> None:
61 """Display search results in a consistent format."""
62 if total == 0 or not results:
63 print("No results found")
64 return
65
66 # Show pagination context
67 start = offset + 1
68 end = offset + len(results)
69 print(f"Showing {start}-{end} of {total} results\n")
70
71 for idx, r in enumerate(results, start):
72 meta = r.get("metadata", {})
73 text = r.get("text", "").replace("\n", " ")
74 snippet = text[:100] + "..." if len(text) > 100 else text
75 label = meta.get("agent") or meta.get("time") or ""
76 facet = meta.get("facet")
77 facet_str = f" ({facet})" if facet else ""
78 print(f"{idx}. {meta.get('day')} {label}{facet_str}: {snippet}")
79
80
81def main() -> None:
82 """Main CLI entry point for the indexer."""
83 parser = argparse.ArgumentParser(
84 description="Index journal content (insights, transcripts, events, entities, todos)"
85 )
86 parser.add_argument(
87 "--rescan",
88 action="store_true",
89 help="Scan and update index (light mode: today + facets/imports/apps, excludes historical days)",
90 )
91 parser.add_argument(
92 "--rescan-full",
93 action="store_true",
94 help="Full rescan including all historical day directories",
95 )
96 parser.add_argument(
97 "--rescan-file",
98 metavar="PATH",
99 help="Index a specific file (absolute or journal-relative path)",
100 )
101 parser.add_argument(
102 "--reset",
103 action="store_true",
104 help="Remove the index before rescan",
105 )
106 parser.add_argument(
107 "--day",
108 help="Filter search results by exact YYYYMMDD day",
109 )
110 parser.add_argument(
111 "--day-from",
112 help="Filter search results by date range start (YYYYMMDD, inclusive)",
113 )
114 parser.add_argument(
115 "--day-to",
116 help="Filter search results by date range end (YYYYMMDD, inclusive)",
117 )
118 parser.add_argument(
119 "--facet",
120 help="Filter search results by facet name",
121 )
122 parser.add_argument(
123 "--agent",
124 "-a",
125 help="Filter search results by agent (e.g., 'flow', 'event', 'news')",
126 )
127 parser.add_argument(
128 "--stream",
129 help="Filter search results by stream name (e.g., 'archon', 'import.apple')",
130 )
131 parser.add_argument(
132 "-q",
133 "--query",
134 nargs="?",
135 const="",
136 help="Run query (interactive mode if no query provided)",
137 )
138 parser.add_argument(
139 "--limit",
140 type=int,
141 default=10,
142 help="Maximum number of results to return (default: 10)",
143 )
144 parser.add_argument(
145 "--offset",
146 type=int,
147 default=0,
148 help="Number of results to skip for pagination (default: 0)",
149 )
150 parser.add_argument(
151 "--top",
152 type=int,
153 default=5,
154 help="Number of items to show per count column (default: 5)",
155 )
156
157 args = setup_cli(parser)
158 journal = get_journal()
159
160 if (
161 not args.rescan
162 and not args.rescan_full
163 and not args.rescan_file
164 and not args.reset
165 and args.query is None
166 ):
167 parser.print_help()
168 return
169
170 if args.reset:
171 reset_journal_index(journal)
172
173 if args.rescan_file:
174 # Single file indexing (incompatible with --rescan/--rescan-full)
175 if args.rescan or args.rescan_full:
176 parser.error("--rescan-file cannot be used with --rescan or --rescan-full")
177 try:
178 index_file(journal, args.rescan_file, verbose=args.verbose)
179 journal_log(f"indexer file indexed: {args.rescan_file}")
180 except (ValueError, FileNotFoundError) as e:
181 parser.error(str(e))
182 elif args.rescan or args.rescan_full:
183 changed = scan_journal(journal, verbose=args.verbose, full=args.rescan_full)
184 if changed:
185 journal_log("indexer journal rescan ok")
186
187 if args.query is not None:
188 query_kwargs: dict[str, Any] = {}
189 if args.day:
190 query_kwargs["day"] = args.day
191 if args.day_from:
192 query_kwargs["day_from"] = args.day_from
193 if args.day_to:
194 query_kwargs["day_to"] = args.day_to
195 if args.facet:
196 query_kwargs["facet"] = args.facet
197 if args.agent:
198 query_kwargs["agent"] = args.agent
199 if args.stream:
200 query_kwargs["stream"] = args.stream
201
202 if args.query:
203 # Single query mode - show counts then results
204 counts = search_counts(args.query, **query_kwargs)
205 _display_counts(counts, args.top)
206 total, results = search_journal(
207 args.query, args.limit, args.offset, **query_kwargs
208 )
209 _display_search_results(results, total, args.offset)
210 else:
211 # Interactive mode
212 while True:
213 try:
214 query = input("search> ").strip()
215 except EOFError:
216 break
217 if not query:
218 break
219 counts = search_counts(query, **query_kwargs)
220 _display_counts(counts, args.top)
221 total, results = search_journal(
222 query, args.limit, args.offset, **query_kwargs
223 )
224 _display_search_results(results, total, args.offset)