parsimony-boc 0.0.1__tar.gz

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 +1,40 @@
1
+ __pycache__/
2
+ *.py[cod]
3
+ *$py.class
4
+ *.so
5
+
6
+ .Python
7
+ build/
8
+ dist/
9
+ *.egg-info/
10
+ *.egg
11
+
12
+ .venv/
13
+ .env
14
+ .env.*
15
+ !.env.example
16
+
17
+ .pytest_cache/
18
+ .mypy_cache/
19
+ .ruff_cache/
20
+ .coverage
21
+ htmlcov/
22
+ coverage.xml
23
+
24
+ uv.lock
25
+
26
+ .vscode/
27
+ .council/
28
+ PLAN-*.md
29
+ .idea/
30
+ *.swp
31
+ .DS_Store
32
+
33
+ outputs/
34
+ logs/
35
+ # Recorded HTTP cassettes must never be committed — respx mocks are hand-authored
36
+ # from upstream API documentation. A pre-commit / CI regex scan is the belt; this
37
+ # ignore is the braces. Override per-file via `!` if you need a hand-authored
38
+ # fixture checked in.
39
+ packages/*/tests/fixtures/**
40
+ !packages/*/tests/fixtures/README.md
@@ -0,0 +1,57 @@
1
+ # Changelog — parsimony-boc
2
+
3
+ All notable changes to `parsimony-boc` will be documented in this file. The
4
+ format is based on [Keep a Changelog](https://keepachangelog.com/) and
5
+ this project adheres to [Semantic Versioning](https://semver.org/).
6
+
7
+ ## [0.8.0] — 2026-06-09
8
+
9
+ Re-run through the full connector guidebook process and **live-verified** against
10
+ the production Valet API and bulk list endpoints.
11
+
12
+ ### Added
13
+
14
+ - **Observations URL-length guard.** Valet 302-redirects any `/observations`
15
+ request whose URL exceeds ~4096 bytes (a limit on URL length, not series
16
+ count). `boc_fetch` now rejects oversized requests pre-network with an
17
+ actionable `InvalidParameterError` (split the names, or fetch a whole panel
18
+ with `group:NAME`) instead of an opaque redirect surfacing as a `ParseError`.
19
+ - **Retired-group pruning.** The per-group membership fan-out doubles as a
20
+ liveness probe: ~29 dated one-off panels that 404 on both `/groups/{name}` and
21
+ `/observations/group/{name}` are pruned from the catalog so it never offers an
22
+ unfetchable panel. A *transient* (5xx/network) failure keeps the group
23
+ best-effort — only a definitive 404 prunes.
24
+ - A `failed/total` enumeration summary log (series + live groups + pruned +
25
+ transient failures), `catalog_tests/queries.yaml` (recall gate), and
26
+ `tests/test_public_surface.py`.
27
+
28
+ ### Changed
29
+
30
+ - **Package restructured** into `_http` / `outputs` / `connectors/{fetch,
31
+ enumerate,__init__}` / `search` / `catalog_build`; the monolithic
32
+ `__init__.py` is now a thin facade. No change to the connector surface
33
+ (`boc_fetch`, `enumerate_boc`, `boc_search`) or the catalog schema.
34
+ - Completeness re-verified live: `/lists/series/json` (15,638 series) is the
35
+ authoritative universe — a full fan-out over all 2,441 groups surfaces **0**
36
+ members absent from it; ~99.7% of listed series are fetchable (stale entries
37
+ return a clean `EmptyDataError`); every live group panel is fetchable.
38
+
39
+ ## [0.5.0] — 2026-05-06
40
+
41
+ ### Changed
42
+
43
+ - Adapted to `parsimony-core==0.5`. Connector code no longer constructs `Provenance` directly; the framework authors all provenance fields in `Connector._wrap_result`. Source-specific extras (where present) move to `Result.with_properties(**kwargs)`. Drops the `provenance=` and `params=` kwargs from `OutputConfig.build_table_result` / `Result.from_dataframe` call sites.
44
+ - Bump `parsimony-core` pin from `>=0.4.0,<0.5` to `>=0.5.0,<0.6` (and `[standard-onnx]` extra accordingly on catalog-publishing packages).
45
+ ## [0.4.0] — 2026-04-24
46
+
47
+ Part of the first coordinated release of the
48
+ [`parsimony-connectors`](https://github.com/ockham-sh/parsimony-connectors)
49
+ monorepo under `parsimony-core==0.4`.
50
+
51
+ ### Changed
52
+
53
+ - Connector rewritten against the kernel's `parsimony.discover` surface
54
+ (`iter_providers`, `load`, `load_all`) and the `@connector(env=...)`
55
+ decorator-level env-var declaration that replaced module-level
56
+ `ENV_VARS`.
57
+ - Pin bumped to `parsimony-core>=0.4,<0.5`.
@@ -0,0 +1,190 @@
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 the 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 the 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 any 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
+ Copyright 2026 Ockham.sh
179
+
180
+ Licensed under the Apache License, Version 2.0 (the "License");
181
+ you may not use this file except in compliance with the License.
182
+ You may obtain a copy of the License at
183
+
184
+ http://www.apache.org/licenses/LICENSE-2.0
185
+
186
+ Unless required by applicable law or agreed to in writing, software
187
+ distributed under the License is distributed on an "AS IS" BASIS,
188
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
189
+ See the License for the specific language governing permissions and
190
+ limitations under the License.
@@ -0,0 +1,137 @@
1
+ Metadata-Version: 2.4
2
+ Name: parsimony-boc
3
+ Version: 0.0.1
4
+ Summary: Bank of Canada connector for the parsimony framework
5
+ Project-URL: Homepage, https://www.bankofcanada.ca
6
+ Project-URL: Repository, https://github.com/ockham-sh/parsimony-connectors
7
+ Project-URL: Issues, https://github.com/ockham-sh/parsimony-connectors/issues
8
+ Author-email: "Ockham.sh" <team@ockham.sh>
9
+ License-Expression: Apache-2.0
10
+ License-File: LICENSE
11
+ Keywords: boc,connectors,data,finance,parsimony
12
+ Classifier: Development Status :: 4 - Beta
13
+ Classifier: Intended Audience :: Developers
14
+ Classifier: Intended Audience :: Financial and Insurance Industry
15
+ Classifier: License :: OSI Approved :: Apache Software License
16
+ Classifier: Programming Language :: Python :: 3
17
+ Classifier: Programming Language :: Python :: 3.11
18
+ Classifier: Programming Language :: Python :: 3.12
19
+ Classifier: Programming Language :: Python :: 3.13
20
+ Classifier: Topic :: Office/Business :: Financial
21
+ Classifier: Topic :: Software Development :: Libraries :: Python Modules
22
+ Classifier: Typing :: Typed
23
+ Requires-Python: >=3.11
24
+ Requires-Dist: pandas<3,>=2.3.0
25
+ Requires-Dist: parsimony-core[catalog]>=0.0.1
26
+ Requires-Dist: pydantic<3,>=2.11.1
27
+ Provides-Extra: dev
28
+ Requires-Dist: mypy>=1.10; extra == 'dev'
29
+ Requires-Dist: pytest-cov>=5.0; extra == 'dev'
30
+ Requires-Dist: pytest>=9.0.3; extra == 'dev'
31
+ Requires-Dist: respx>=0.22.0; extra == 'dev'
32
+ Requires-Dist: ruff>=0.15.10; extra == 'dev'
33
+ Description-Content-Type: text/markdown
34
+
35
+ # parsimony-boc
36
+
37
+ Bank of Canada connector — Canadian exchange rates, interest/bond yields, money
38
+ & credit aggregates, CPI, commodity price indices, and the data behind BoC
39
+ publications, as numeric time series via the **Valet** API.
40
+
41
+ Part of the [parsimony-connectors](https://github.com/ockham-sh/parsimony-connectors) monorepo. Distributed standalone on PyPI as `parsimony-boc`.
42
+
43
+ ## Connectors
44
+
45
+ | Name | Kind | Description |
46
+ |---|---|---|
47
+ | `boc_fetch` | connector | Fetch one or more series by name (e.g. `FXUSDCAD,FXEURCAD`) or a whole panel by group (e.g. `group:FX_RATES_DAILY`). Reaches the **entire** universe by name. |
48
+ | `enumerate_boc` | enumerator | Catalog feed: one row per series (~15.6k) and one per live group (~2.4k), from Valet's list endpoints plus a per-group membership fan-out. |
49
+ | `boc_search` | connector | Search the published BoC catalog (lexical title or `code:`/structured). Pass returned codes to `boc_fetch(series_name=...)`. |
50
+
51
+ ## Discovery model
52
+
53
+ The Valet API has **no native keyword search**, so discovery is a built catalog
54
+ (archetype A — one live full-index call lists the whole universe):
55
+
56
+ - `GET /lists/series/json` is the **authoritative enumeration path** (15,638
57
+ series, live 2026-06-09). It self-tracks BoC additions, so the catalog stays
58
+ current on each rebuild.
59
+ - **Groups are first-class catalog entities.** Each named panel (e.g.
60
+ `FX_RATES_DAILY`) gets its own row keyed `group:NAME` — the exact string
61
+ `boc_fetch` accepts — because group descriptions carry retrieval signal no
62
+ individual series has ("Month-end, Millions of dollars"). A fetch on a
63
+ `group:` code returns the whole panel in one request.
64
+ - The per-group membership fan-out doubles as a **liveness probe**: ~29 retired
65
+ one-off panels that 404 on every fetch path are pruned, so the catalog never
66
+ offers a panel you cannot fetch.
67
+
68
+ Every series stays fetchable by name regardless of catalog coverage — the
69
+ boundary is discovery, not access.
70
+
71
+ ## Install
72
+
73
+ ```bash
74
+ pip install parsimony-boc
75
+ ```
76
+
77
+ Pulls in a compatible `parsimony-core` automatically. Verify discovery:
78
+
79
+ ```bash
80
+ python -c "from parsimony import discover; print([p.name for p in discover.iter_providers()])"
81
+ ```
82
+
83
+ ## Configuration
84
+
85
+ No configuration required — the Valet API is open and unauthenticated.
86
+
87
+ `boc_search` reads published catalog snapshots (default root
88
+ `hf://parsimony-dev/boc`). Override with `PARSIMONY_BOC_CATALOG_URL` or
89
+ `catalog_url=` at call time; a missing snapshot is built on demand from the live
90
+ list endpoints and cached in an LRU.
91
+
92
+ > **Observations request limit.** Valet caps the `/observations` request URL near
93
+ > **4096 bytes** (roughly 100–160 comma-joined series, depending on name length).
94
+ > `boc_fetch` guards this pre-call and asks you to split the request or fetch a
95
+ > whole panel with `group:NAME`. The limit is on URL length, not series count.
96
+
97
+ ## Quick start
98
+
99
+ ```python
100
+ from parsimony_boc import CONNECTORS
101
+
102
+ # find a series (or a whole panel) in the catalog
103
+ hits = CONNECTORS["boc_search"](query="US dollar Canadian dollar exchange rate")
104
+ code = hits.data.iloc[0]["code"] # e.g. "FXUSDCAD" or "group:FX_RATES_DAILY"
105
+ # fetch observations
106
+ result = CONNECTORS["boc_fetch"](series_name=code, start_date="2024-01-01")
107
+ print(result.data.head())
108
+ ```
109
+
110
+ For multi-plugin composition (autoloads everything installed):
111
+
112
+ ```python
113
+ from parsimony import discover
114
+ connectors = discover.load_all()
115
+ ```
116
+
117
+ ## Catalog building
118
+
119
+ `scripts/build_catalog.py` builds the single `boc` catalog from the live list
120
+ endpoints and saves/pushes a snapshot:
121
+
122
+ ```bash
123
+ uv run python packages/boc/scripts/build_catalog.py \
124
+ --save file:///tmp/parsimony-catalogs/boc --push hf://parsimony-dev/boc
125
+ ```
126
+
127
+ ## Provider
128
+
129
+ - Homepage: https://www.bankofcanada.ca
130
+ - Valet API docs: https://www.bankofcanada.ca/valet/docs
131
+ - Terms: https://www.bankofcanada.ca/terms/ — free reuse **with attribution to
132
+ the Bank of Canada**. Data is © Bank of Canada; this connector and its catalog
133
+ are a derived index of series identifiers and titles.
134
+
135
+ ## License
136
+
137
+ See [LICENSE](./LICENSE).
@@ -0,0 +1,103 @@
1
+ # parsimony-boc
2
+
3
+ Bank of Canada connector — Canadian exchange rates, interest/bond yields, money
4
+ & credit aggregates, CPI, commodity price indices, and the data behind BoC
5
+ publications, as numeric time series via the **Valet** API.
6
+
7
+ Part of the [parsimony-connectors](https://github.com/ockham-sh/parsimony-connectors) monorepo. Distributed standalone on PyPI as `parsimony-boc`.
8
+
9
+ ## Connectors
10
+
11
+ | Name | Kind | Description |
12
+ |---|---|---|
13
+ | `boc_fetch` | connector | Fetch one or more series by name (e.g. `FXUSDCAD,FXEURCAD`) or a whole panel by group (e.g. `group:FX_RATES_DAILY`). Reaches the **entire** universe by name. |
14
+ | `enumerate_boc` | enumerator | Catalog feed: one row per series (~15.6k) and one per live group (~2.4k), from Valet's list endpoints plus a per-group membership fan-out. |
15
+ | `boc_search` | connector | Search the published BoC catalog (lexical title or `code:`/structured). Pass returned codes to `boc_fetch(series_name=...)`. |
16
+
17
+ ## Discovery model
18
+
19
+ The Valet API has **no native keyword search**, so discovery is a built catalog
20
+ (archetype A — one live full-index call lists the whole universe):
21
+
22
+ - `GET /lists/series/json` is the **authoritative enumeration path** (15,638
23
+ series, live 2026-06-09). It self-tracks BoC additions, so the catalog stays
24
+ current on each rebuild.
25
+ - **Groups are first-class catalog entities.** Each named panel (e.g.
26
+ `FX_RATES_DAILY`) gets its own row keyed `group:NAME` — the exact string
27
+ `boc_fetch` accepts — because group descriptions carry retrieval signal no
28
+ individual series has ("Month-end, Millions of dollars"). A fetch on a
29
+ `group:` code returns the whole panel in one request.
30
+ - The per-group membership fan-out doubles as a **liveness probe**: ~29 retired
31
+ one-off panels that 404 on every fetch path are pruned, so the catalog never
32
+ offers a panel you cannot fetch.
33
+
34
+ Every series stays fetchable by name regardless of catalog coverage — the
35
+ boundary is discovery, not access.
36
+
37
+ ## Install
38
+
39
+ ```bash
40
+ pip install parsimony-boc
41
+ ```
42
+
43
+ Pulls in a compatible `parsimony-core` automatically. Verify discovery:
44
+
45
+ ```bash
46
+ python -c "from parsimony import discover; print([p.name for p in discover.iter_providers()])"
47
+ ```
48
+
49
+ ## Configuration
50
+
51
+ No configuration required — the Valet API is open and unauthenticated.
52
+
53
+ `boc_search` reads published catalog snapshots (default root
54
+ `hf://parsimony-dev/boc`). Override with `PARSIMONY_BOC_CATALOG_URL` or
55
+ `catalog_url=` at call time; a missing snapshot is built on demand from the live
56
+ list endpoints and cached in an LRU.
57
+
58
+ > **Observations request limit.** Valet caps the `/observations` request URL near
59
+ > **4096 bytes** (roughly 100–160 comma-joined series, depending on name length).
60
+ > `boc_fetch` guards this pre-call and asks you to split the request or fetch a
61
+ > whole panel with `group:NAME`. The limit is on URL length, not series count.
62
+
63
+ ## Quick start
64
+
65
+ ```python
66
+ from parsimony_boc import CONNECTORS
67
+
68
+ # find a series (or a whole panel) in the catalog
69
+ hits = CONNECTORS["boc_search"](query="US dollar Canadian dollar exchange rate")
70
+ code = hits.data.iloc[0]["code"] # e.g. "FXUSDCAD" or "group:FX_RATES_DAILY"
71
+ # fetch observations
72
+ result = CONNECTORS["boc_fetch"](series_name=code, start_date="2024-01-01")
73
+ print(result.data.head())
74
+ ```
75
+
76
+ For multi-plugin composition (autoloads everything installed):
77
+
78
+ ```python
79
+ from parsimony import discover
80
+ connectors = discover.load_all()
81
+ ```
82
+
83
+ ## Catalog building
84
+
85
+ `scripts/build_catalog.py` builds the single `boc` catalog from the live list
86
+ endpoints and saves/pushes a snapshot:
87
+
88
+ ```bash
89
+ uv run python packages/boc/scripts/build_catalog.py \
90
+ --save file:///tmp/parsimony-catalogs/boc --push hf://parsimony-dev/boc
91
+ ```
92
+
93
+ ## Provider
94
+
95
+ - Homepage: https://www.bankofcanada.ca
96
+ - Valet API docs: https://www.bankofcanada.ca/valet/docs
97
+ - Terms: https://www.bankofcanada.ca/terms/ — free reuse **with attribution to
98
+ the Bank of Canada**. Data is © Bank of Canada; this connector and its catalog
99
+ are a derived index of series identifiers and titles.
100
+
101
+ ## License
102
+
103
+ See [LICENSE](./LICENSE).
@@ -0,0 +1,32 @@
1
+ """Bank of Canada (BoC) connector: fetch + catalog enumeration + search.
2
+
3
+ Valet API (``https://www.bankofcanada.ca/valet``) — keyless public JSON. Three
4
+ connectors, discovered as the top-level ``CONNECTORS`` bundle:
5
+
6
+ * ``boc_fetch`` — observations by series name(s) or ``group:NAME`` panel.
7
+ * ``enumerate_boc`` — the catalog feed (archetype A: live ``/lists/series``).
8
+ * ``boc_search`` — search over the published catalog snapshot.
9
+
10
+ This module is a thin facade; the implementation lives in ``_http`` / ``outputs``
11
+ / ``connectors/{fetch,enumerate}`` / ``search`` / ``catalog_build``.
12
+ """
13
+
14
+ from __future__ import annotations
15
+
16
+ from parsimony_boc.connectors import CONNECTORS, load
17
+ from parsimony_boc.connectors.enumerate import enumerate_boc
18
+ from parsimony_boc.connectors.fetch import boc_fetch
19
+ from parsimony_boc.outputs import BOC_ENUMERATE_OUTPUT, BOC_FETCH_OUTPUT
20
+ from parsimony_boc.search import BOC_SEARCH_OUTPUT, PARSIMONY_BOC_CATALOG_URL_ENV, boc_search
21
+
22
+ __all__ = [
23
+ "BOC_ENUMERATE_OUTPUT",
24
+ "BOC_FETCH_OUTPUT",
25
+ "BOC_SEARCH_OUTPUT",
26
+ "CONNECTORS",
27
+ "PARSIMONY_BOC_CATALOG_URL_ENV",
28
+ "boc_fetch",
29
+ "boc_search",
30
+ "enumerate_boc",
31
+ "load",
32
+ ]
@@ -0,0 +1,91 @@
1
+ """Bank of Canada (BoC) Valet transport: constants + the canonical client.
2
+
3
+ The Valet API is keyless JSON — there is no auth to resolve, so this module is
4
+ thin: a base URL, timeouts, the group-fan-out concurrency cap, and the one
5
+ non-obvious transport fact, the **observations request-URI limit**.
6
+
7
+ ``boc_fetch`` and the list endpoints all speak GET + JSON, so they go through
8
+ the kernel's canonical ``make_http_client`` + ``fetch_json`` pair (GET +
9
+ ``raise_for_status`` + ``map_http_error`` / ``map_timeout_error`` + ``json()`` +
10
+ ``None``-param dropping). There is no provider-specific status semantics to
11
+ intercept, so there is no per-package mapper chokepoint here.
12
+
13
+ Endpoints (base ``https://www.bankofcanada.ca/valet``):
14
+
15
+ * ``GET /lists/series/json`` (~15.6k) and ``GET /lists/groups/json`` (~2.4k) —
16
+ the catalog index.
17
+ * ``GET /groups/{name}/json`` — per-group series membership (the fan-out); a
18
+ retired group returns 404.
19
+ * ``GET /observations/{names}/json`` and
20
+ ``GET /observations/group/{name}/json`` — time-series observations.
21
+
22
+ **Observations request-URI cap (~4096 bytes).** Valet redirects (HTTP 302, to an
23
+ error page) any ``/observations`` request whose URL exceeds ~4 KB. The boundary
24
+ is on the *URL length*, not the series count: a full URL of 4087 bytes returns
25
+ 200 while 4127 bytes 302s (measured live 2026-06-09). So a count cap is wrong
26
+ (140 short names pass, 140 long names 302). We guard the assembled URL
27
+ pre-network and raise a clear :class:`InvalidParameterError` instead of letting
28
+ the agent hit an opaque redirect that ``fetch_json`` would surface as a
29
+ ``ParseError``.
30
+ """
31
+
32
+ from __future__ import annotations
33
+
34
+ from parsimony.errors import InvalidParameterError
35
+ from parsimony.transport import HttpClient
36
+ from parsimony.transport.helpers import make_http_client
37
+
38
+ PROVIDER = "boc"
39
+
40
+ BASE_URL = "https://www.bankofcanada.ca/valet"
41
+
42
+ #: Per-call timeout for both the fetch and the (large) list endpoints. The
43
+ #: series index is ~3.5 MB, so allow a long read.
44
+ FETCH_TIMEOUT = 60.0
45
+
46
+ #: Concurrency cap for the per-group fan-out that builds the series→group map.
47
+ #: Valet is unauthenticated and tolerates moderate concurrency; 16 keeps the
48
+ #: ~2,400-group sweep at ~70 s while staying well under any sensible limit.
49
+ GROUP_FETCH_CONCURRENCY = 16
50
+
51
+ #: Maximum assembled-URL length (host + path, query excluded) we will send to
52
+ #: ``/observations``. The server caps the request URI at ~4096 bytes; we sit
53
+ #: conservatively below that so the date query string still fits under the cap.
54
+ OBSERVATIONS_MAX_URL_BYTES = 4000
55
+
56
+
57
+ def make_valet_client(timeout: float = FETCH_TIMEOUT) -> HttpClient:
58
+ """Build the canonical keyless Valet client (GET + JSON via ``fetch_json``)."""
59
+ return make_http_client(BASE_URL, timeout=timeout)
60
+
61
+
62
+ def guard_observations_path(path: str, *, series_name: str) -> None:
63
+ """Reject an ``/observations`` path whose URL would exceed the server cap.
64
+
65
+ Valet 302-redirects any observations request URL over ~4 KB. Caught
66
+ pre-network, this becomes an actionable :class:`InvalidParameterError`
67
+ (split the request or fetch a whole panel with ``group:NAME``) instead of an
68
+ opaque redirect that downstream parsing would report as a ``ParseError``.
69
+ """
70
+ url_len = len(BASE_URL) + 1 + len(path)
71
+ if url_len > OBSERVATIONS_MAX_URL_BYTES:
72
+ raise InvalidParameterError(
73
+ PROVIDER,
74
+ (
75
+ f"too many/long series names: the request URL is {url_len} bytes but the "
76
+ f"Bank of Canada observations endpoint caps it near {OBSERVATIONS_MAX_URL_BYTES + 96}. "
77
+ "Split the names across multiple boc_fetch calls, or fetch a whole panel in one "
78
+ "call with series_name='group:GROUP_NAME'."
79
+ ),
80
+ )
81
+
82
+
83
+ __all__ = [
84
+ "BASE_URL",
85
+ "FETCH_TIMEOUT",
86
+ "GROUP_FETCH_CONCURRENCY",
87
+ "OBSERVATIONS_MAX_URL_BYTES",
88
+ "PROVIDER",
89
+ "guard_observations_path",
90
+ "make_valet_client",
91
+ ]
@@ -0,0 +1,24 @@
1
+ """Build the Bank of Canada catalog snapshot."""
2
+
3
+ from __future__ import annotations
4
+
5
+ from parsimony.catalog import Catalog
6
+ from parsimony.catalog.policy import discovery_indexes
7
+ from parsimony.catalog.source import entities_from_raw
8
+
9
+ from parsimony_boc.connectors.enumerate import enumerate_boc
10
+ from parsimony_boc.outputs import BOC_ENUMERATE_OUTPUT
11
+
12
+ CATALOG_NAMESPACE = "boc"
13
+
14
+
15
+ def build_boc_catalog() -> Catalog:
16
+ result = enumerate_boc()
17
+ entries = entities_from_raw(result, BOC_ENUMERATE_OUTPUT)
18
+ catalog = Catalog(CATALOG_NAMESPACE, indexes=discovery_indexes(entries), default_field="title")
19
+ catalog.set_entities(entries)
20
+ catalog.build()
21
+ return catalog
22
+
23
+
24
+ __all__ = ["CATALOG_NAMESPACE", "build_boc_catalog"]
@@ -0,0 +1,25 @@
1
+ """boc connector registry."""
2
+
3
+ from __future__ import annotations
4
+
5
+ from parsimony.connector import Connectors
6
+
7
+ from parsimony_boc.connectors.enumerate import enumerate_boc
8
+ from parsimony_boc.connectors.fetch import boc_fetch
9
+ from parsimony_boc.search import boc_search
10
+
11
+ CONNECTORS = Connectors([boc_fetch, enumerate_boc, boc_search])
12
+
13
+
14
+ def load(*, catalog_url: str | None = None) -> Connectors:
15
+ """Return :data:`CONNECTORS` with the optional search catalog URL bound.
16
+
17
+ BoC is keyless — there is no API key to bind. ``catalog_url`` lets an
18
+ operator point ``boc_search`` at a specific catalog snapshot.
19
+ """
20
+ if catalog_url is None:
21
+ return CONNECTORS
22
+ return CONNECTORS.bind(catalog_url=catalog_url)
23
+
24
+
25
+ __all__ = ["CONNECTORS", "load"]
@@ -0,0 +1,220 @@
1
+ """BoC catalog enumerator (archetype A: the live ``/lists/series`` full index).
2
+
3
+ Emits one row per series **and** one row per *live* group. The series index
4
+ (``/lists/series/json``, ~15.6k) is the authoritative universe — verified
5
+ complete: a full fan-out over every group surfaces zero members absent from it.
6
+
7
+ The per-group fan-out (``/groups/{name}/json`` × ~2.4k) does double duty:
8
+
9
+ 1. **Membership** — annotates each series with its group (97.7% of series carry
10
+ a group; the first encountered wins for the rare multi-group series).
11
+ 2. **Liveness** — a group whose detail endpoint 404s is *retired* (BoC leaves
12
+ ~29 dated one-off panels in ``/lists/groups`` that 404 on both
13
+ ``/groups/{name}`` and ``/observations/group/{name}``). We use the 404 signal
14
+ to **prune** those group rows so the catalog never offers a panel that cannot
15
+ be fetched. A *transient* failure (5xx / network) is kept best-effort — only
16
+ a definitive 404 prunes.
17
+
18
+ A ``failed/total`` summary is logged at the end so a quietly-shrunk crawl is
19
+ visible (guidebook §7.2).
20
+ """
21
+
22
+ from __future__ import annotations
23
+
24
+ import logging
25
+ from typing import Any
26
+
27
+ import httpx
28
+ import pandas as pd
29
+ from parsimony.connector import enumerator
30
+ from parsimony.errors import ParseError
31
+ from parsimony.transport import HttpClient, pooled_client
32
+ from parsimony.transport.helpers import fetch_json
33
+
34
+ from parsimony_boc._http import PROVIDER, make_valet_client
35
+ from parsimony_boc.outputs import BOC_ENUMERATE_OUTPUT, ENUMERATE_COLUMNS
36
+
37
+ logger = logging.getLogger(__name__)
38
+
39
+
40
+ def _list_groups(client: HttpClient) -> dict[str, dict[str, Any]]:
41
+ """Return BoC's group index (~2.4k entries) from ``/lists/groups/json``.
42
+
43
+ This is the **bounding seam** for live tests: monkeypatch this module global
44
+ to a 2–3 group slice so the per-group fan-out fires a handful of requests,
45
+ never the full ~2,400-request crawl. ``enumerate_boc`` reads it as a module
46
+ global at call time so the monkeypatch takes.
47
+ """
48
+ payload = fetch_json(client, path="lists/groups/json", provider=PROVIDER, op_name="groups/list")
49
+ if not isinstance(payload, dict):
50
+ raise ParseError(PROVIDER, "unexpected /lists/groups/json shape (expected object)")
51
+ groups = payload.get("groups") or {}
52
+ return groups if isinstance(groups, dict) else {}
53
+
54
+
55
+ def _fetch_group_membership(
56
+ client: HttpClient,
57
+ group_name: str,
58
+ ) -> tuple[str, list[str], bool]:
59
+ """Fetch one group's membership. Returns ``(name, members, alive)``.
60
+
61
+ ``alive`` is ``False`` **only** on a definitive 404 (a retired group): the
62
+ caller prunes those from the catalog. A transient transport error or a
63
+ 200-with-non-JSON body keeps ``alive=True`` (best-effort) — the group is
64
+ still catalogued, it just loses its membership annotation this run, so a
65
+ network blip never silently shrinks the catalog.
66
+ """
67
+ try:
68
+ resp = client.request("GET", f"/groups/{group_name}/json")
69
+ resp.raise_for_status()
70
+ body = resp.json()
71
+ except httpx.HTTPStatusError as exc:
72
+ if exc.response.status_code == 404:
73
+ # Retired-but-indexed group: prune it.
74
+ return group_name, [], False
75
+ logger.warning("BoC group fetch failed for %r: HTTP %s", group_name, exc.response.status_code)
76
+ return group_name, [], True
77
+ except (httpx.HTTPError, ValueError) as exc:
78
+ # Transport failure OR a 200-with-non-JSON body (BoC sometimes serves
79
+ # an HTML stub) — keep the group, drop only its membership this run.
80
+ logger.warning("BoC group fetch failed for %r: %s", group_name, exc)
81
+ return group_name, [], True
82
+
83
+ if not isinstance(body, dict):
84
+ return group_name, [], True
85
+ details = body.get("groupDetails") or {}
86
+ members = details.get("groupSeries") or {}
87
+ if not isinstance(members, dict):
88
+ return group_name, [], True
89
+ return group_name, [s for s in members if s], True
90
+
91
+
92
+ def _build_series_to_group_map(
93
+ client: HttpClient,
94
+ groups_index: dict[str, dict[str, Any]],
95
+ ) -> tuple[dict[str, tuple[str, str]], set[str], int]:
96
+ """Fan out over groups → ``(series_to_group, dead_groups, transient_failures)``.
97
+
98
+ ``series_to_group`` maps each series to its first ``(group_id, group_label)``.
99
+ ``dead_groups`` are the names that 404'd (retired) — the caller skips their
100
+ rows. ``transient_failures`` counts non-404 failures (kept, for the summary
101
+ log). Connections are pooled across the walk via :func:`pooled_client`.
102
+ """
103
+ with pooled_client(client) as shared:
104
+ results = [_fetch_group_membership(shared, name) for name in groups_index]
105
+
106
+ series_to_group: dict[str, tuple[str, str]] = {}
107
+ dead_groups: set[str] = set()
108
+ transient_failures = 0
109
+ for group_name, members, alive in results:
110
+ if not alive:
111
+ dead_groups.add(group_name)
112
+ continue
113
+ if not members:
114
+ # alive but no members → either an empty/transient result. We can't
115
+ # distinguish a live-but-empty group from a transient miss here, but
116
+ # live groups always carry members in practice, so a non-404 empty is
117
+ # a transient failure for summary purposes.
118
+ transient_failures += 1
119
+ info = groups_index.get(group_name) or {}
120
+ label = (info.get("label") if isinstance(info, dict) else "") or ""
121
+ for series_name in members:
122
+ if series_name not in series_to_group:
123
+ series_to_group[series_name] = (group_name, label)
124
+ return series_to_group, dead_groups, transient_failures
125
+
126
+
127
+ @enumerator(output=BOC_ENUMERATE_OUTPUT, tags=["macro", "ca"])
128
+ def enumerate_boc() -> pd.DataFrame:
129
+ """Enumerate every Bank of Canada series and live group via Valet.
130
+
131
+ Granularity is one row per series — Valet addresses observations per series,
132
+ so series-level keys are the right unit (~15.6k rows) — plus one row per live
133
+ group (keyed ``group:NAME``) so whole panels are discoverable.
134
+
135
+ Pipeline: ``/lists/series/json`` and ``/lists/groups/json``, then a
136
+ serial ``/groups/{name}/json`` fan-out for series→group membership and
137
+ group liveness (retired groups that 404 are pruned).
138
+ """
139
+ client = make_valet_client()
140
+
141
+ series_payload = fetch_json(client, path="lists/series/json", provider=PROVIDER, op_name="series/list")
142
+ if not isinstance(series_payload, dict):
143
+ raise ParseError(PROVIDER, "unexpected /lists/series/json shape (expected object)")
144
+
145
+ groups_index = _list_groups(client)
146
+ series_to_group, dead_groups, transient_failures = _build_series_to_group_map(client, groups_index)
147
+
148
+ series = series_payload.get("series") or {}
149
+ if not isinstance(series, dict):
150
+ series = {}
151
+
152
+ rows: list[dict[str, str]] = []
153
+ for series_name, info in series.items():
154
+ if not series_name:
155
+ continue
156
+ if isinstance(info, dict):
157
+ label = info.get("label") or series_name
158
+ desc = info.get("description") or ""
159
+ else:
160
+ label = str(info)
161
+ desc = ""
162
+
163
+ group_id, group_label = series_to_group.get(series_name, ("", ""))
164
+ rows.append(
165
+ {
166
+ "series_name": series_name,
167
+ "title": label,
168
+ "description": desc,
169
+ "source": "valet",
170
+ "entity_type": "series",
171
+ "group": group_id,
172
+ "group_label": group_label,
173
+ }
174
+ )
175
+
176
+ n_series = len(rows)
177
+
178
+ # One row per *live* group as a discoverable catalog entity. Groups are
179
+ # addressable via ``boc_fetch(series_name="group:NAME")``; cataloguing them
180
+ # lets agents search by the group-level description ("Month-end, Millions of
181
+ # dollars") and fetch a whole panel in one shot. Retired groups (404 on
182
+ # detail) are pruned so the catalog never offers an unfetchable panel.
183
+ n_groups = 0
184
+ for group_name, group_info in groups_index.items():
185
+ if not group_name or group_name in dead_groups:
186
+ continue
187
+ if isinstance(group_info, dict):
188
+ g_label = group_info.get("label") or group_name
189
+ g_desc = group_info.get("description") or ""
190
+ else:
191
+ g_label = str(group_info)
192
+ g_desc = ""
193
+ rows.append(
194
+ {
195
+ "series_name": f"group:{group_name}",
196
+ "title": g_label,
197
+ "description": g_desc,
198
+ "source": "valet",
199
+ "entity_type": "group",
200
+ "group": group_name,
201
+ "group_label": g_label,
202
+ }
203
+ )
204
+ n_groups += 1
205
+
206
+ logger.info(
207
+ "boc enumerate: %d series + %d live groups "
208
+ "(%d retired groups pruned, %d transient membership failures of %d groups)",
209
+ n_series,
210
+ n_groups,
211
+ len(dead_groups),
212
+ transient_failures,
213
+ len(groups_index),
214
+ )
215
+
216
+ columns = list(ENUMERATE_COLUMNS)
217
+ return pd.DataFrame(rows, columns=columns) if rows else pd.DataFrame(columns=columns)
218
+
219
+
220
+ __all__ = ["enumerate_boc"]
@@ -0,0 +1,119 @@
1
+ """BoC observations fetch connector (Valet ``/observations`` endpoint).
2
+
3
+ Fetches one or more series by name, or a whole named panel via ``group:NAME``.
4
+ The observations payload is a wide, date-keyed array — ``{"d": ..., "FXUSDCAD":
5
+ {"v": "1.38"}, ...}`` — which we melt into long ``(series_name, title, date,
6
+ value)`` rows. Suppressed/missing observations (``""`` / ``"NaN"`` / absent) come
7
+ back as a null ``value``.
8
+ """
9
+
10
+ from __future__ import annotations
11
+
12
+ from typing import Annotated, Any
13
+
14
+ import pandas as pd
15
+ from parsimony import Namespace
16
+ from parsimony.connector import connector
17
+ from parsimony.errors import EmptyDataError, InvalidParameterError, ParseError
18
+ from parsimony.transport.helpers import fetch_json
19
+
20
+ from parsimony_boc._http import PROVIDER, guard_observations_path, make_valet_client
21
+ from parsimony_boc.outputs import BOC_FETCH_OUTPUT
22
+
23
+ _FETCH_COLUMNS = ["series_name", "title", "date", "value"]
24
+
25
+
26
+ def _parse_observations(
27
+ json_data: dict[str, Any],
28
+ series_details: dict[str, Any] | None = None,
29
+ ) -> pd.DataFrame:
30
+ """Melt the Valet observations array into long-format rows.
31
+
32
+ Each observation entry looks like
33
+ ``{"d": "2024-01-15", "FXUSDCAD": {"v": "1.3456"}, ...}`` — one date key
34
+ plus one nested ``{"v": ...}`` per series. Titles are resolved from the
35
+ payload's ``seriesDetail`` block when present, else fall back to the name.
36
+ """
37
+ observations = json_data.get("observations", [])
38
+ if not observations:
39
+ return pd.DataFrame(columns=_FETCH_COLUMNS)
40
+
41
+ sample = observations[0]
42
+ series_cols = [k for k in sample if k != "d"]
43
+
44
+ rows: list[dict[str, Any]] = []
45
+ for obs in observations:
46
+ date = obs.get("d", "")
47
+ for col in series_cols:
48
+ raw = obs.get(col)
49
+ if raw is None:
50
+ continue
51
+ raw_value = raw.get("v") if isinstance(raw, dict) else raw
52
+ try:
53
+ value = float(raw_value) if raw_value is not None and raw_value not in ("", "NaN") else None
54
+ except (ValueError, TypeError):
55
+ value = None
56
+
57
+ title = col
58
+ if series_details and col in series_details:
59
+ detail = series_details[col]
60
+ title = detail.get("label", detail.get("description", col))
61
+
62
+ rows.append({"series_name": col, "title": title, "date": date, "value": value})
63
+
64
+ return pd.DataFrame(rows, columns=_FETCH_COLUMNS) if rows else pd.DataFrame(columns=_FETCH_COLUMNS)
65
+
66
+
67
+ @connector(output=BOC_FETCH_OUTPUT, tags=["macro", "ca"])
68
+ def boc_fetch(
69
+ series_name: Annotated[str, Namespace("boc")],
70
+ start_date: str | None = None,
71
+ end_date: str | None = None,
72
+ ) -> pd.DataFrame:
73
+ """Fetch Bank of Canada time series by series name(s) or group name.
74
+
75
+ Use 'group:GROUP_NAME' syntax for a whole panel (e.g. group:FX_RATES_DAILY).
76
+ Otherwise pass one or more comma-separated series names (e.g.
77
+ FXUSDCAD,FXEURCAD). Discover names with boc_search or enumerate_boc.
78
+ Observations come back as long-format (series_name, title, date, value)
79
+ rows; optional start_date/end_date (YYYY-MM-DD) bound the window.
80
+ """
81
+ series_name = series_name.strip()
82
+ if not series_name:
83
+ raise InvalidParameterError(PROVIDER, "series_name must be non-empty")
84
+
85
+ if series_name.startswith("group:"):
86
+ group_name = series_name[6:].strip()
87
+ if not group_name:
88
+ raise InvalidParameterError(PROVIDER, "group name must be non-empty after 'group:'")
89
+ path = f"observations/group/{group_name}/json"
90
+ else:
91
+ path = f"observations/{series_name}/json"
92
+
93
+ # Valet 302-redirects observations requests whose URL exceeds ~4 KB; reject
94
+ # those pre-network with actionable guidance rather than a downstream parse
95
+ # failure. Only the multi-series path realistically trips this.
96
+ guard_observations_path(path, series_name=series_name)
97
+
98
+ json_data = fetch_json(
99
+ make_valet_client(),
100
+ path=path,
101
+ params={"start_date": start_date, "end_date": end_date},
102
+ provider=PROVIDER,
103
+ op_name="observations",
104
+ )
105
+ if not isinstance(json_data, dict):
106
+ raise ParseError(PROVIDER, f"unexpected observations response shape for: {series_name}")
107
+
108
+ df = _parse_observations(json_data, json_data.get("seriesDetail"))
109
+ if df.empty:
110
+ raise EmptyDataError(
111
+ PROVIDER,
112
+ message=f"No observations returned for: {series_name}",
113
+ query_params={"series_name": series_name, "start_date": start_date, "end_date": end_date},
114
+ )
115
+
116
+ return df
117
+
118
+
119
+ __all__ = ["boc_fetch"]
@@ -0,0 +1,54 @@
1
+ """Output schemas for the Bank of Canada (BoC) connectors."""
2
+
3
+ from __future__ import annotations
4
+
5
+ from parsimony.result import Column, ColumnRole, OutputConfig
6
+
7
+ BOC_FETCH_OUTPUT = OutputConfig(
8
+ columns=[
9
+ Column(name="series_name", role=ColumnRole.KEY, namespace="boc"),
10
+ Column(name="title", role=ColumnRole.TITLE),
11
+ Column(name="date", dtype="datetime", role=ColumnRole.DATA),
12
+ Column(name="value", dtype="numeric", role=ColumnRole.DATA),
13
+ ]
14
+ )
15
+
16
+ BOC_ENUMERATE_OUTPUT = OutputConfig(
17
+ columns=[
18
+ # The KEY is either a series name (e.g. ``FXUSDCAD``) or a group entry
19
+ # prefixed with ``group:`` (e.g. ``group:FX_RATES_DAILY``). Groups are
20
+ # first-class addressable entities — ``boc_fetch`` accepts
21
+ # ``series_name="group:NAME"`` and Valet's
22
+ # ``/observations/group/{name}/json`` returns the full panel — so they
23
+ # get their own catalog rows for discovery alongside the per-series rows.
24
+ # The ``group:`` prefix is the exact string ``boc_fetch`` already
25
+ # expects, so a search hit routes straight to a fetch.
26
+ Column(name="series_name", role=ColumnRole.KEY, namespace="boc"),
27
+ Column(name="title", role=ColumnRole.TITLE),
28
+ # The upstream Valet ``description`` text. For group rows this carries
29
+ # the group's description (often the only place units/frequency live,
30
+ # e.g. "Month-end, Millions of dollars").
31
+ Column(name="description", role=ColumnRole.METADATA),
32
+ # ``source`` tells the agent which fetch connector to call. BoC has a
33
+ # single Valet source today; the column is future-proofing for a
34
+ # parallel source so dispatch is already wired (matches Treasury's
35
+ # ``fiscal_data``/``treasury_rates`` split).
36
+ Column(name="source", role=ColumnRole.METADATA),
37
+ # ``entity_type`` is ``"series"`` for individual series rows and
38
+ # ``"group"`` for group rows — lets agents filter or weight by
39
+ # entity granularity.
40
+ Column(name="entity_type", role=ColumnRole.METADATA),
41
+ # ``group`` carries the upstream group ID the series belongs to. A
42
+ # series can belong to several groups (rare); the first encountered
43
+ # group wins. Empty string when a series is in no catalogued group.
44
+ # For group rows this is the group's own ID.
45
+ Column(name="group", role=ColumnRole.METADATA),
46
+ Column(name="group_label", role=ColumnRole.METADATA),
47
+ ]
48
+ )
49
+
50
+ #: The exact column order the ``@enumerator`` must return (enumerators drop
51
+ #: unmapped columns then require an exact match against the declared schema).
52
+ ENUMERATE_COLUMNS: tuple[str, ...] = tuple(c.name for c in BOC_ENUMERATE_OUTPUT.columns)
53
+
54
+ __all__ = ["BOC_ENUMERATE_OUTPUT", "BOC_FETCH_OUTPUT", "ENUMERATE_COLUMNS"]
File without changes
@@ -0,0 +1,35 @@
1
+ """Semantic search over the published Bank of Canada (BoC) catalog."""
2
+
3
+ from __future__ import annotations
4
+
5
+ from parsimony.catalog.search import CatalogSearchParams, make_local_search_connector
6
+ from parsimony.result import Column, ColumnRole, OutputConfig
7
+
8
+ from parsimony_boc.catalog_build import build_boc_catalog
9
+
10
+ PARSIMONY_BOC_CATALOG_URL_ENV = "PARSIMONY_BOC_CATALOG_URL"
11
+
12
+ BocSearchParams = CatalogSearchParams
13
+
14
+ BOC_SEARCH_OUTPUT = OutputConfig(
15
+ columns=[
16
+ Column(name="code", role=ColumnRole.KEY, namespace="boc"),
17
+ Column(name="title", role=ColumnRole.TITLE),
18
+ Column(name="score", role=ColumnRole.DATA),
19
+ ]
20
+ )
21
+
22
+
23
+ boc_search = make_local_search_connector(
24
+ provider="boc",
25
+ default_url="hf://parsimony-dev/boc",
26
+ catalog_url_env_var=PARSIMONY_BOC_CATALOG_URL_ENV,
27
+ build_catalog=build_boc_catalog,
28
+ tags=["macro", "ca", "tool"],
29
+ description=(
30
+ "Search the Bank of Canada (BoC) Valet catalog. "
31
+ "Preferred: structured queries such as 'code: FXUSDCAD'. "
32
+ "Pass returned codes to boc_fetch(series_name=...)."
33
+ ),
34
+ output_columns=BOC_SEARCH_OUTPUT.columns,
35
+ )
@@ -0,0 +1,79 @@
1
+ [project]
2
+ name = "parsimony-boc"
3
+ version = "0.0.1"
4
+ description = "Bank of Canada connector for the parsimony framework"
5
+ authors = [{ name = "Ockham.sh", email = "team@ockham.sh" }]
6
+ license = "Apache-2.0"
7
+ readme = "README.md"
8
+ requires-python = ">=3.11"
9
+ keywords = ["finance", "data", "connectors", "parsimony", "boc"]
10
+ classifiers = [
11
+ "Development Status :: 4 - Beta",
12
+ "Intended Audience :: Developers",
13
+ "Intended Audience :: Financial and Insurance Industry",
14
+ "License :: OSI Approved :: Apache Software License",
15
+ "Programming Language :: Python :: 3",
16
+ "Programming Language :: Python :: 3.11",
17
+ "Programming Language :: Python :: 3.12",
18
+ "Programming Language :: Python :: 3.13",
19
+ "Topic :: Office/Business :: Financial",
20
+ "Topic :: Software Development :: Libraries :: Python Modules",
21
+ "Typing :: Typed",
22
+ ]
23
+ dependencies = [
24
+ "parsimony-core[catalog]>=0.0.1",
25
+ "pydantic>=2.11.1,<3",
26
+ "pandas>=2.3.0,<3",
27
+ ]
28
+
29
+ [project.optional-dependencies]
30
+ dev = [
31
+ "pytest>=9.0.3",
32
+ "pytest-cov>=5.0",
33
+ "respx>=0.22.0",
34
+ "ruff>=0.15.10",
35
+ "mypy>=1.10",
36
+ ]
37
+
38
+ [project.urls]
39
+ Homepage = "https://www.bankofcanada.ca"
40
+ Repository = "https://github.com/ockham-sh/parsimony-connectors"
41
+ Issues = "https://github.com/ockham-sh/parsimony-connectors/issues"
42
+
43
+ [project.entry-points."parsimony.providers"]
44
+ boc = "parsimony_boc"
45
+
46
+
47
+ [build-system]
48
+ requires = ["hatchling"]
49
+ build-backend = "hatchling.build"
50
+
51
+ [tool.hatch.build.targets.wheel]
52
+ packages = ["parsimony_boc"]
53
+
54
+ [tool.hatch.build.targets.sdist]
55
+ include = ["parsimony_boc", "README.md", "LICENSE", "CHANGELOG.md"]
56
+
57
+ [tool.ruff]
58
+ target-version = "py311"
59
+ line-length = 120
60
+
61
+ [tool.ruff.lint]
62
+ select = ["E", "F", "I", "UP", "B", "SIM"]
63
+
64
+ [tool.ruff.lint.per-file-ignores]
65
+ # Publish scripts call logging.basicConfig() before importing parsimony so
66
+ # the kernel's INFO logs surface during the long-running publish pipeline.
67
+ "scripts/*" = ["E402"]
68
+
69
+ [tool.mypy]
70
+ python_version = "3.11"
71
+ warn_return_any = true
72
+ warn_unused_ignores = true
73
+ ignore_missing_imports = true
74
+
75
+ [tool.pytest.ini_options]
76
+ addopts = "--import-mode=importlib -m 'not integration'"
77
+ markers = [
78
+ "integration: hits live APIs (may be slow, requires env vars)",
79
+ ]