rag-security-protocol 0.1.1rc1__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -0,0 +1,132 @@
1
+ Metadata-Version: 2.5
2
+ Name: rag-security-protocol
3
+ Version: 0.1.1rc1
4
+ Summary: RAG Security Protocol — where security tools hook into a RAG pipeline, and what they return
5
+ Project-URL: Homepage, https://github.com/PCBZ/rsp
6
+ Project-URL: Specification, https://github.com/PCBZ/rsp/blob/main/SPEC.md
7
+ Project-URL: Issues, https://github.com/PCBZ/rsp/issues
8
+ Project-URL: Changelog, https://github.com/PCBZ/rsp/releases
9
+ Author: PCBZ
10
+ License-Expression: Apache-2.0
11
+ License-File: LICENSE
12
+ Keywords: conformance,llm,protocol,rag,redaction,security
13
+ Classifier: Development Status :: 4 - Beta
14
+ Classifier: Intended Audience :: Developers
15
+ Classifier: Programming Language :: Python :: 3.11
16
+ Classifier: Programming Language :: Python :: 3.12
17
+ Classifier: Programming Language :: Python :: 3.13
18
+ Classifier: Programming Language :: Python :: 3.14
19
+ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
20
+ Classifier: Topic :: Security
21
+ Classifier: Typing :: Typed
22
+ Requires-Python: >=3.11
23
+ Provides-Extra: llamaindex
24
+ Requires-Dist: llama-index-core<0.15,>=0.14; extra == 'llamaindex'
25
+ Description-Content-Type: text/markdown
26
+
27
+ # RSP
28
+
29
+ A protocol for scanning content **before** it reaches a vector store.
30
+
31
+ Everything that ships today scans after storage, where an embedding already
32
+ exists and the text is recoverable from it. RSP puts the scanner at the
33
+ ingest hook instead, so that blocked means it was never there.
34
+
35
+ ```console
36
+ $ rsp ingest demo/sample-docs --config demo/rsp.toml
37
+ scanned 11 chunks
38
+ REDACT aws-access-token demo/sample-docs/deploy-notes.md:1-14
39
+ REDACT private-key demo/sample-docs/runbook-backups.md:1-18
40
+ 11 chunks indexed, 0 blocked, 2 redacted
41
+ ```
42
+
43
+ Nothing in that run is ours except the wiring: the detector is
44
+ [gitleaks](https://github.com/gitleaks/gitleaks), unmodified, and the pipeline
45
+ is LlamaIndex.
46
+
47
+ ## What it is
48
+
49
+ `SPEC.md` is the deliverable — twenty-eight clauses, each with the reasoning
50
+ that produced it, twenty-three of them settled by a conformance case rather
51
+ than by assertion. The code in `rsp/` is one host implementation, and exists
52
+ to prove the clauses are implementable.
53
+
54
+ A plugin is a command. It reads one JSON object on stdin, writes one on
55
+ stdout, and exits:
56
+
57
+ ```console
58
+ $ echo '{"rsp_version":"0.1","hook":"on_chunk","content":"key: AKIA47CQZHT2MVPF3JXB"}' | my-scanner
59
+ {"verdict":"REDACT","spans":[{"start":5,"end":25,"type":"aws"}],"replacement":"[REDACTED]"}
60
+ ```
61
+
62
+ Offsets are bytes into UTF-8, half-open. The host applies them; a plugin never
63
+ sees the content again. [Writing a plugin](https://github.com/PCBZ/rsp/blob/main/WRITING-A-PLUGIN.md) is the page to
64
+ start from — the four messages, and the three things that bite.
65
+
66
+ ## Try it
67
+
68
+ ```bash
69
+ uv sync --extra llamaindex
70
+ uv run rsp ingest demo/sample-docs --config demo/rsp.toml # needs gitleaks and node
71
+ ```
72
+
73
+ ## Use it
74
+
75
+ ```python
76
+ from rsp.config import load
77
+ from rsp.guards import RSPIngestGuard, RSPRetrieveGuard
78
+ from rsp.runtime import Runtime
79
+
80
+ runtime = Runtime(load("rsp.toml")) # every plugin handshakes here, or this raises
81
+
82
+ indexed = IngestionPipeline(
83
+ transformations=[SentenceSplitter(), RSPIngestGuard(runtime=runtime), embedding]
84
+ ).run(documents=documents)
85
+
86
+ kept = RSPRetrieveGuard(runtime=runtime).postprocess_nodes(retrieved, query_str=query)
87
+ ```
88
+
89
+ The ingest guard goes after the splitter and before the embedding, which is
90
+ the whole point: a blocked chunk is never embedded and never stored. The
91
+ retrieve guard catches what an index already held.
92
+
93
+ [`examples/host`](https://github.com/PCBZ/rsp/tree/main/examples/host) is that, whole and runnable: four documents
94
+ through one pipeline, judged by a Python plugin and a Go one at the same time.
95
+ The host code does not know which is which.
96
+
97
+ ## Check your own plugin
98
+
99
+ ```console
100
+ $ python conformance/run.py -- my-scanner
101
+ ```
102
+
103
+ Copy `conformance/` and `plugins/` next to your plugin — they are not in the
104
+ installed package, on purpose: you are checking against the protocol, not
105
+ against a version of it that shipped with a host. Standard library only, and
106
+ nothing in it imports this host. `conformance/README.md` explains roles
107
+ and the host cases.
108
+
109
+ ## Where things are
110
+
111
+ | | |
112
+ |---|---|
113
+ | `SPEC.md` | the protocol |
114
+ | `WRITING-A-PLUGIN.md` | the on-ramp, quoting a plugin under test |
115
+ | `conformance/` | cases, a runner, and a plugin that misbehaves on request |
116
+ | `rsp/` | the reference host — five layers, one-way imports, zero dependencies |
117
+ | `plugins/rsp-echo` | the reference plugin, whose verdicts are chosen by markers |
118
+ | `examples/host` | a whole host in one file, judged by a Python plugin and a Go one |
119
+ | `examples/gitleaks-{ts,go,rs,py,c,java,swift}` | the same real scanner wrapped seven times, in seven languages |
120
+ | `demo/` | eight documents, two of them with fabricated credentials |
121
+
122
+ Decisions (D1–D9) and open questions live in the
123
+ [wiki](https://github.com/PCBZ/rsp/wiki). Clause IDs — `S1`, `E4` — resolve
124
+ inside a clone; issue numbers do not, so the spec never cites one.
125
+
126
+ ## Status
127
+
128
+ v0.1 is the first freeze. Four clauses have no conformance case and
129
+ `SPEC.md` §10 says which and why. Two of them — blocked content never reaching
130
+ storage, and a dropped retrieval hit — are the ones a host case cannot observe
131
+ without the host offering somewhere to look, and they are the two the protocol
132
+ is most about. That is the honest state of it.
@@ -0,0 +1,16 @@
1
+ rsp/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
2
+ rsp/cli.py,sha256=GHv1oAEItdQUiHzPrvr02IN3lSMvcttjRdWmgd8h_R4,2683
3
+ rsp/codec.py,sha256=o-SGeLqUc5OOkc3Axr2t9f4BNWuDZ1B4g6Z2heJoD34,6256
4
+ rsp/config.py,sha256=qzs0W_40ooQH_mXrmVsiivSeymW10l5X10tLomrgJkg,5954
5
+ rsp/guards.py,sha256=33jdo6rzu4YawTeaFncaPQAnx7uA52Ayv4aMx3VMVZs,3304
6
+ rsp/handshake.py,sha256=2WJ-rvDA494ixmNOO2_yZQcfFsScLaWQutbImLS_NyU,2388
7
+ rsp/ingest.py,sha256=xddiatHxxnhzoouh_-N110nkPc-QimvD2yAoYMmmh5Q,5546
8
+ rsp/process.py,sha256=AChPxbTySX7nXqh8-UCg_klQyLNPFHER04d3_0caGBo,5488
9
+ rsp/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
10
+ rsp/runtime.py,sha256=LXv_oVGGHpoIuTUajGSHX7_Qe-WoaS6Z2ZmExZNW3A0,9907
11
+ rsp/spans.py,sha256=mIw9TmYahPX3K91SsSlLITlQBHo8T4ep0HX2Bp5tjMw,3709
12
+ rag_security_protocol-0.1.1rc1.dist-info/METADATA,sha256=RitIxzr_pLzws01On0pNNtgyuZQwxT4juWpxMOpbMoQ,5341
13
+ rag_security_protocol-0.1.1rc1.dist-info/WHEEL,sha256=W3fkpkm7-wf9vBI5Z-7s0eWkeM-spu78I8Neb98DeEg,87
14
+ rag_security_protocol-0.1.1rc1.dist-info/entry_points.txt,sha256=MpeFCB8yav-W9hF5Bd2yfQ-8G0_kfg9K9GPPgU5uY1w,37
15
+ rag_security_protocol-0.1.1rc1.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
16
+ rag_security_protocol-0.1.1rc1.dist-info/RECORD,,
@@ -0,0 +1,4 @@
1
+ Wheel-Version: 1.0
2
+ Generator: hatchling 1.32.4
3
+ Root-Is-Purelib: true
4
+ Tag: py3-none-any
@@ -0,0 +1,2 @@
1
+ [console_scripts]
2
+ rsp = rsp.cli:main
@@ -0,0 +1,201 @@
1
+ Apache License
2
+ Version 2.0, January 2004
3
+ http://www.apache.org/licenses/
4
+
5
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6
+
7
+ 1. Definitions.
8
+
9
+ "License" shall mean the terms and conditions for use, reproduction,
10
+ and distribution as defined by Sections 1 through 9 of this document.
11
+
12
+ "Licensor" shall mean the copyright owner or entity authorized by
13
+ the copyright owner that is granting the License.
14
+
15
+ "Legal Entity" shall mean the union of the acting entity and all
16
+ other entities that control, are controlled by, or are under common
17
+ control with that entity. For the purposes of this definition,
18
+ "control" means (i) the power, direct or indirect, to cause the
19
+ direction or management of such entity, whether by contract or
20
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
21
+ outstanding shares, or (iii) beneficial ownership of such entity.
22
+
23
+ "You" (or "Your") shall mean an individual or Legal Entity
24
+ exercising permissions granted by this License.
25
+
26
+ "Source" form shall mean the preferred form for making modifications,
27
+ including but not limited to software source code, documentation
28
+ source, and configuration files.
29
+
30
+ "Object" form shall mean any form resulting from mechanical
31
+ transformation or translation of a Source form, including but
32
+ not limited to compiled object code, generated documentation,
33
+ and conversions to other media types.
34
+
35
+ "Work" shall mean the work of authorship, whether in Source or
36
+ Object form, made available under the License, as indicated by a
37
+ copyright notice that is included in or attached to the work
38
+ (an example is provided in the Appendix below).
39
+
40
+ "Derivative Works" shall mean any work, whether in Source or Object
41
+ form, that is based on (or derived from) the Work and for which the
42
+ editorial revisions, annotations, elaborations, or other modifications
43
+ represent, as a whole, an original work of authorship. For the purposes
44
+ of this License, Derivative Works shall not include works that remain
45
+ separable from, or merely link (or bind by name) to the interfaces of,
46
+ the Work and Derivative Works thereof.
47
+
48
+ "Contribution" shall mean any work of authorship, including
49
+ the original version of the Work and any modifications or additions
50
+ to that Work or Derivative Works thereof, that is intentionally
51
+ submitted to Licensor for inclusion in the Work by the copyright owner
52
+ or by an individual or Legal Entity authorized to submit on behalf of
53
+ the copyright owner. For the purposes of this definition, "submitted"
54
+ means any form of electronic, verbal, or written communication sent
55
+ to the Licensor or its representatives, including but not limited to
56
+ communication on electronic mailing lists, source code control systems,
57
+ and issue tracking systems that are managed by, or on behalf of, the
58
+ Licensor for the purpose of discussing and improving the Work, but
59
+ excluding communication that is conspicuously marked or otherwise
60
+ designated in writing by the copyright owner as "Not a Contribution."
61
+
62
+ "Contributor" shall mean Licensor and any individual or Legal Entity
63
+ on behalf of whom a Contribution has been received by Licensor and
64
+ subsequently incorporated within the Work.
65
+
66
+ 2. Grant of Copyright License. Subject to the terms and conditions of
67
+ this License, each Contributor hereby grants to You a perpetual,
68
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69
+ copyright license to reproduce, prepare Derivative Works of,
70
+ publicly display, publicly perform, sublicense, and distribute the
71
+ Work and such Derivative Works in Source or Object form.
72
+
73
+ 3. Grant of Patent License. Subject to the terms and conditions of
74
+ this License, each Contributor hereby grants to You a perpetual,
75
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76
+ (except as stated in this section) patent license to make, have made,
77
+ use, offer to sell, sell, import, and otherwise transfer the Work,
78
+ where such license applies only to those patent claims licensable
79
+ by such Contributor that are necessarily infringed by their
80
+ Contribution(s) alone or by combination of their Contribution(s)
81
+ with the Work to which such Contribution(s) was submitted. If You
82
+ institute patent litigation against any entity (including a
83
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
84
+ or a Contribution incorporated within the Work constitutes direct
85
+ or contributory patent infringement, then any patent licenses
86
+ granted to You under this License for that Work shall terminate
87
+ as of the date such litigation is filed.
88
+
89
+ 4. Redistribution. You may reproduce and distribute copies of the
90
+ Work or Derivative Works thereof in any medium, with or without
91
+ modifications, and in Source or Object form, provided that You
92
+ meet the following conditions:
93
+
94
+ (a) You must give any other recipients of the Work or
95
+ Derivative Works a copy of this License; and
96
+
97
+ (b) You must cause any modified files to carry prominent notices
98
+ stating that You changed the files; and
99
+
100
+ (c) You must retain, in the Source form of any Derivative Works
101
+ that You distribute, all copyright, patent, trademark, and
102
+ attribution notices from the Source form of the Work,
103
+ excluding those notices that do not pertain to any part of
104
+ the Derivative Works; and
105
+
106
+ (d) If the Work includes a "NOTICE" text file as part of its
107
+ distribution, then any Derivative Works that You distribute must
108
+ include a readable copy of the attribution notices contained
109
+ within such NOTICE file, excluding those notices that do not
110
+ pertain to any part of the Derivative Works, in at least one
111
+ of the following places: within a NOTICE text file distributed
112
+ as part of the Derivative Works; within the Source form or
113
+ documentation, if provided along with the Derivative Works; or,
114
+ within a display generated by the Derivative Works, if and
115
+ wherever such third-party notices normally appear. The contents
116
+ of the NOTICE file are for informational purposes only and
117
+ do not modify the License. You may add Your own attribution
118
+ notices within Derivative Works that You distribute, alongside
119
+ or as an addendum to the NOTICE text from the Work, provided
120
+ that such additional attribution notices cannot be construed
121
+ as modifying the License.
122
+
123
+ You may add Your own copyright statement to Your modifications and
124
+ may provide additional or different license terms and conditions
125
+ for use, reproduction, or distribution of Your modifications, or
126
+ for any such Derivative Works as a whole, provided Your use,
127
+ reproduction, and distribution of the Work otherwise complies with
128
+ the conditions stated in this License.
129
+
130
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
131
+ any Contribution intentionally submitted for inclusion in the Work
132
+ by You to the Licensor shall be under the terms and conditions of
133
+ this License, without any additional terms or conditions.
134
+ Notwithstanding the above, nothing herein shall supersede or modify
135
+ the terms of any separate license agreement you may have executed
136
+ with Licensor regarding such Contributions.
137
+
138
+ 6. Trademarks. This License does not grant permission to use the trade
139
+ names, trademarks, service marks, or product names of the Licensor,
140
+ except as required for reasonable and customary use in describing the
141
+ origin of the Work and reproducing the content of the NOTICE file.
142
+
143
+ 7. Disclaimer of Warranty. Unless required by applicable law or
144
+ agreed to in writing, Licensor provides the Work (and each
145
+ Contributor provides its Contributions) on an "AS IS" BASIS,
146
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147
+ implied, including, without limitation, any warranties or conditions
148
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149
+ PARTICULAR PURPOSE. You are solely responsible for determining the
150
+ appropriateness of using or redistributing the Work and assume any
151
+ risks associated with Your exercise of permissions under this License.
152
+
153
+ 8. Limitation of Liability. In no event and under no legal theory,
154
+ whether in tort (including negligence), contract, or otherwise,
155
+ unless required by applicable law (such as deliberate and grossly
156
+ negligent acts) or agreed to in writing, shall any Contributor be
157
+ liable to You for damages, including any direct, indirect, special,
158
+ incidental, or consequential damages of any character arising as a
159
+ result of this License or out of the use or inability to use the
160
+ Work (including but not limited to damages for loss of goodwill,
161
+ work stoppage, computer failure or malfunction, or any and all
162
+ other commercial damages or losses), even if such Contributor
163
+ has been advised of the possibility of such damages.
164
+
165
+ 9. Accepting Warranty or Additional Liability. While redistributing
166
+ the Work or Derivative Works thereof, You may choose to offer,
167
+ and charge a fee for, acceptance of support, warranty, indemnity,
168
+ or other liability obligations and/or rights consistent with this
169
+ License. However, in accepting such obligations, You may act only
170
+ on Your own behalf and on Your sole responsibility, not on behalf
171
+ of any other Contributor, and only if You agree to indemnify,
172
+ defend, and hold each Contributor harmless for any liability
173
+ incurred by, or claims asserted against, such Contributor by reason
174
+ of your accepting any such warranty or additional liability.
175
+
176
+ END OF TERMS AND CONDITIONS
177
+
178
+ APPENDIX: How to apply the Apache License to your work.
179
+
180
+ To apply the Apache License to your work, attach the following
181
+ boilerplate notice, with the fields enclosed by brackets "[]"
182
+ replaced with your own identifying information. (Don't include
183
+ the brackets!) The text should be enclosed in the appropriate
184
+ comment syntax for the file format. We also recommend that a
185
+ file or class name and description of purpose be included on the
186
+ same "printed page" as the copyright notice for easier
187
+ identification within third-party archives.
188
+
189
+ Copyright [yyyy] [name of copyright owner]
190
+
191
+ Licensed under the Apache License, Version 2.0 (the "License");
192
+ you may not use this file except in compliance with the License.
193
+ You may obtain a copy of the License at
194
+
195
+ http://www.apache.org/licenses/LICENSE-2.0
196
+
197
+ Unless required by applicable law or agreed to in writing, software
198
+ distributed under the License is distributed on an "AS IS" BASIS,
199
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200
+ See the License for the specific language governing permissions and
201
+ limitations under the License.
rsp/__init__.py ADDED
File without changes
rsp/cli.py ADDED
@@ -0,0 +1,80 @@
1
+ """`rsp validate` checks a config; `rsp ingest` runs a directory through it.
2
+
3
+ `load` stops at the first problem, because a host with half its guards is worse
4
+ than one that does not start; `validate` turns that into a message and an exit
5
+ code for someone editing the file.
6
+ """
7
+
8
+ from __future__ import annotations
9
+
10
+ import argparse
11
+ import pathlib
12
+ import sys
13
+ from collections.abc import Sequence
14
+
15
+ from rsp.config import load
16
+ from rsp.runtime import ConfigError, Plugin
17
+
18
+
19
+ def _validate(plugins: list[Plugin]) -> int:
20
+ for plugin in plugins:
21
+ print(
22
+ f"{plugin.name}: {' '.join(plugin.command)} "
23
+ f"[on_error={plugin.on_error.value} timeout={plugin.timeout:g}s]"
24
+ )
25
+ # No handshake: a binary installed elsewhere does not make the config invalid.
26
+ print(f"{len(plugins)} plugin(s), not started", file=sys.stderr)
27
+ return 0
28
+
29
+
30
+ def _ingest(directory: pathlib.Path, plugins: list[Plugin]) -> int:
31
+ try:
32
+ # rsp.ingest imports llama_index lazily, so a missing extra surfaces at the call.
33
+ from rsp.ingest import ingest
34
+
35
+ report = ingest(directory, plugins)
36
+ except ImportError:
37
+ print(
38
+ "rsp ingest needs the llamaindex extra: pip install 'rsp[llamaindex]'", file=sys.stderr
39
+ )
40
+ return 1
41
+ except (ConfigError, FileNotFoundError) as exc:
42
+ print(exc, file=sys.stderr)
43
+ return 1
44
+
45
+ print(f"scanned {report.scanned} chunks")
46
+ for item in report.findings:
47
+ # A chunk's lines, not a position: the host never sees a span (S4).
48
+ print(f" {item.verdict:7}{item.types:18} {item.source}:{item.lines}")
49
+ print(
50
+ f"{report.indexed} chunks indexed, "
51
+ f"{len(report.blocked)} blocked, {len(report.redacted)} redacted"
52
+ )
53
+ return 0
54
+
55
+
56
+ def main(argv: Sequence[str] | None = None) -> int:
57
+ parser = argparse.ArgumentParser(prog="rsp")
58
+ commands = parser.add_subparsers(dest="command", required=True)
59
+
60
+ check = commands.add_parser("validate", help="check a config file")
61
+ check.add_argument("config", help="path to a TOML config")
62
+
63
+ run = commands.add_parser("ingest", help="run a directory through the plugins")
64
+ run.add_argument("directory", help="directory of documents")
65
+ run.add_argument("--config", default="rsp.toml", help="path to a TOML config")
66
+
67
+ arguments = parser.parse_args(argv)
68
+ try:
69
+ plugins = load(arguments.config)
70
+ except ConfigError as exc:
71
+ print(exc, file=sys.stderr)
72
+ return 1
73
+
74
+ if arguments.command == "validate":
75
+ return _validate(plugins)
76
+ return _ingest(pathlib.Path(arguments.directory), plugins)
77
+
78
+
79
+ if __name__ == "__main__":
80
+ raise SystemExit(main())
rsp/codec.py ADDED
@@ -0,0 +1,167 @@
1
+ """Wire codec: a request mapping in, a parsed response out, or why not.
2
+
3
+ Sits between rsp.process and plugin identity, so both directions can be tested on
4
+ raw bytes without spawning anything.
5
+ """
6
+
7
+ from __future__ import annotations
8
+
9
+ import json
10
+ from collections.abc import Mapping, Sequence
11
+ from dataclasses import dataclass
12
+ from typing import Any
13
+
14
+ from rsp.process import DEFAULT_MAX_OUTPUT, DEFAULT_TIMEOUT, Invocation, Outcome, invoke
15
+
16
+ RSP_VERSION = "0.1"
17
+ # Past this, parsers lose precision and two hosts disagree about a span (RFC 8259 §6).
18
+ _SAFE_INTEGER = 2**53 - 1
19
+ # RFC 8259 §2 whitespace. str.strip() also removes characters a strict parser refuses.
20
+ _WHITESPACE = " \t\n\r"
21
+
22
+
23
+ def _sendable(value: Any, seen: frozenset[int] = frozenset()) -> None:
24
+ """Raise ValueError if any part of a request would break T4 on the wire.
25
+
26
+ `seen` holds the containers above this one, so a request nested inside
27
+ itself is named rather than reaching the stack limit. Depth that is merely
28
+ great still raises RecursionError, which `call` catches alongside the rest
29
+ (E2).
30
+ """
31
+ if isinstance(value, (Mapping, list, tuple)):
32
+ if id(value) in seen:
33
+ raise ValueError("a request cannot contain itself")
34
+ seen = seen | {id(value)}
35
+ if isinstance(value, Mapping):
36
+ # json.dumps writes the key 1 as "1", so {1: "a", "1": "b"} repeats a key.
37
+ written = [str(key) if isinstance(key, (int, float, bool)) else key for key in value]
38
+ if len(set(written)) != len(written):
39
+ raise ValueError("a key would be repeated once written")
40
+ for nested in value.values():
41
+ _sendable(nested, seen)
42
+ elif isinstance(value, (list, tuple)):
43
+ for nested in value:
44
+ _sendable(nested, seen)
45
+ elif isinstance(value, int) and not isinstance(value, bool) and abs(value) > _SAFE_INTEGER:
46
+ raise ValueError(f"{value} is outside the interoperable range")
47
+
48
+
49
+ def encode(request: Mapping[str, Any]) -> bytes:
50
+ """One request as compact UTF-8 JSON. Raises ValueError if it cannot be.
51
+
52
+ `ensure_ascii=False` keeps the bytes a plugin counts the bytes we sent (S1);
53
+ `allow_nan=False` stops Python writing NaN, which no other parser accepts.
54
+ """
55
+ _sendable(request) # a host does not send what it would refuse (T4)
56
+ return json.dumps(request, ensure_ascii=False, allow_nan=False, separators=(",", ":")).encode(
57
+ "utf-8"
58
+ )
59
+
60
+
61
+ def _reject_constant(token: str) -> Any:
62
+ """NaN and Infinity are not JSON (RFC 8259); Python's parser takes them."""
63
+ raise ValueError(f"{token} is not JSON")
64
+
65
+
66
+ def _reject_duplicate_keys(pairs: list[tuple[str, Any]]) -> dict[str, Any]:
67
+ """Refuse a repeated key: parsers disagree about which one wins (RFC 8259 §4)."""
68
+ seen: set[str] = set()
69
+ for key, _ in pairs:
70
+ if key in seen:
71
+ raise ValueError(f"duplicate key {key!r}")
72
+ seen.add(key)
73
+ return dict(pairs)
74
+
75
+
76
+ def _checked_number(token: str) -> int:
77
+ """Integers a JavaScript host could not read back unchanged."""
78
+ value = int(token)
79
+ if abs(value) > _SAFE_INTEGER:
80
+ raise ValueError(f"{token} is outside the interoperable range")
81
+ return value
82
+
83
+
84
+ _DECODER = json.JSONDecoder(
85
+ parse_constant=_reject_constant,
86
+ object_pairs_hook=_reject_duplicate_keys,
87
+ parse_int=_checked_number,
88
+ )
89
+
90
+
91
+ def decode(raw: bytes) -> tuple[Outcome, dict[str, Any] | None]:
92
+ """Exactly one JSON object and nothing else, or why not (T2).
93
+
94
+ Not the first object that parses: anything else on stdout makes the protocol
95
+ channel a log, where a stray line looks like a response.
96
+ """
97
+ try:
98
+ text = raw.decode("utf-8")
99
+ except UnicodeDecodeError:
100
+ return Outcome.MALFORMED, None
101
+
102
+ # Stripped with JSON's four, not Python's: a plugin whose whole output is
103
+ # some other space character said something, and EMPTY means it did not.
104
+ body = text.lstrip(_WHITESPACE)
105
+ if not body:
106
+ return Outcome.EMPTY, None
107
+
108
+ try:
109
+ payload, end = _DECODER.raw_decode(body)
110
+ except (ValueError, RecursionError):
111
+ # ValueError: JSONDecodeError, or a value the hooks above refuse.
112
+ # RecursionError: nesting deep enough to exhaust the stack, which the
113
+ # output cap does not bound — 200KB of brackets is well under it. A
114
+ # plugin does not get to choose when the host falls over (D2).
115
+ return Outcome.MALFORMED, None
116
+
117
+ if body[end:].strip(_WHITESPACE):
118
+ return Outcome.MALFORMED, None # a second object, or trailing noise
119
+ if not isinstance(payload, dict):
120
+ return Outcome.MALFORMED, None # a list or a bare string is not a response
121
+ try:
122
+ # `\ud800` parses and cannot be written back as UTF-8 (M1). Through
123
+ # `encode`, whose walk also bounds the nesting: the C scanner reads two
124
+ # thousand brackets without complaint, and a plugin does not get to
125
+ # choose when the host falls over (D2).
126
+ encode(payload)
127
+ except (ValueError, RecursionError):
128
+ return Outcome.MALFORMED, None
129
+
130
+ return Outcome.OK, payload
131
+
132
+
133
+ @dataclass(frozen=True)
134
+ class Reply:
135
+ outcome: Outcome
136
+ payload: dict[str, Any] | None
137
+ invocation: Invocation | None # None when no process was ever started
138
+
139
+ @property
140
+ def ok(self) -> bool:
141
+ return self.outcome is Outcome.OK
142
+
143
+
144
+ def call(
145
+ command: Sequence[str],
146
+ request: Mapping[str, Any],
147
+ *,
148
+ timeout: float = DEFAULT_TIMEOUT,
149
+ max_output: int = DEFAULT_MAX_OUTPUT,
150
+ ) -> Reply:
151
+ """One plugin call, from a request mapping to a parsed response. Never raises (E2).
152
+
153
+ Only a clean exit has its output parsed: from a killed or truncated call it is
154
+ a fragment, whether or not it happens to parse.
155
+ """
156
+ try:
157
+ payload_bytes = encode(request)
158
+ except (ValueError, TypeError, RecursionError):
159
+ # No plugin is at fault and none ran, but there is no verdict either (E3).
160
+ return Reply(Outcome.UNENCODABLE, None, None)
161
+
162
+ invocation = invoke(command, payload_bytes, timeout=timeout, max_output=max_output)
163
+ if not invocation.ok:
164
+ return Reply(invocation.outcome, None, invocation)
165
+
166
+ outcome, payload = decode(invocation.stdout)
167
+ return Reply(outcome, payload, invocation)