bmad-module-skill-forge 5.0.0 → 6.0.0
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.
- package/.claude-plugin/marketplace.json +1 -1
- package/docs/_data/pinned.yaml +1 -1
- package/docs/agents.md +1 -1
- package/docs/architecture.md +3 -3
- package/docs/campaign.md +1 -1
- package/docs/troubleshooting.md +9 -3
- package/docs/verifying-a-skill.md +9 -9
- package/docs/workflows.md +4 -4
- package/package.json +1 -1
- package/src/knowledge/tool-resolution.md +6 -6
- package/src/shared/references/headless-gate-convention.md +4 -1
- package/src/shared/references/output-contract-schema.md +3 -3
- package/src/shared/references/pipeline-contracts.md +2 -1
- package/src/shared/scripts/schemas/skf-test-result-envelope.v1.json +1 -1
- package/src/shared/scripts/skf-build-change-manifest.py +332 -54
- package/src/shared/scripts/skf-classify-changed-files.py +15 -10
- package/src/shared/scripts/skf-emit-result-envelope.py +58 -15
- package/src/shared/scripts/skf-extract-public-api.py +33 -13
- package/src/shared/scripts/skf-parse-gaps.py +3 -2
- package/src/shared/scripts/skf-verify-provenance-completeness.py +108 -11
- package/src/skf-analyze-source/SKILL.md +1 -0
- package/src/skf-analyze-source/references/generate-briefs.md +1 -1
- package/src/skf-analyze-source/references/headless-contract.md +1 -1
- package/src/skf-audit-skill/SKILL.md +2 -1
- package/src/skf-audit-skill/references/report.md +1 -1
- package/src/skf-brief-skill/SKILL.md +1 -0
- package/src/skf-brief-skill/references/invocation-contract.md +2 -2
- package/src/skf-campaign/SKILL.md +1 -0
- package/src/skf-campaign/references/campaign-contracts.md +2 -0
- package/src/skf-campaign/references/step-05-skill-loop.md +2 -2
- package/src/skf-campaign/references/step-10-export.md +3 -2
- package/src/skf-create-skill/SKILL.md +2 -1
- package/src/skf-create-skill/references/ccc-index-check.md +1 -1
- package/src/skf-create-skill/references/component-extraction.md +1 -1
- package/src/skf-create-skill/references/extract.md +1 -1
- package/src/skf-create-skill/references/generate-artifacts.md +1 -1
- package/src/skf-create-skill/references/report.md +2 -2
- package/src/skf-create-skill/references/sub/ccc-discover.md +1 -1
- package/src/skf-create-skill/references/validate.md +1 -1
- package/src/skf-create-stack-skill/SKILL.md +1 -0
- package/src/skf-create-stack-skill/references/generate-output.md +1 -1
- package/src/skf-create-stack-skill/references/invocation-contract.md +1 -1
- package/src/skf-create-stack-skill/references/parallel-extract.md +2 -2
- package/src/skf-create-stack-skill/references/report.md +1 -1
- package/src/skf-drop-skill/SKILL.md +1 -0
- package/src/skf-drop-skill/references/invocation-contract.md +1 -1
- package/src/skf-drop-skill/references/report.md +1 -1
- package/src/skf-export-skill/SKILL.md +1 -0
- package/src/skf-export-skill/references/invocation-contract.md +1 -1
- package/src/skf-export-skill/references/summary.md +1 -1
- package/src/skf-forger/SKILL.md +1 -0
- package/src/skf-forger/references/pipeline-mode.md +2 -2
- package/src/skf-quick-skill/SKILL.md +2 -1
- package/src/skf-quick-skill/references/finalize.md +1 -1
- package/src/skf-quick-skill/references/halt-contract.md +2 -2
- package/src/skf-quick-skill/references/write-and-validate.md +1 -1
- package/src/skf-refine-architecture/SKILL.md +2 -1
- package/src/skf-refine-architecture/customize.toml +1 -1
- package/src/skf-refine-architecture/references/report.md +1 -1
- package/src/skf-rename-skill/SKILL.md +1 -0
- package/src/skf-rename-skill/references/invocation-contract.md +2 -2
- package/src/skf-setup/SKILL.md +1 -0
- package/src/skf-setup/references/ccc-index.md +1 -1
- package/src/skf-setup/references/invocation-contract.md +1 -1
- package/src/skf-test-skill/SKILL.md +1 -0
- package/src/skf-test-skill/assets/output-section-formats.md +4 -4
- package/src/skf-test-skill/references/coherence-check.md +8 -6
- package/src/skf-test-skill/references/coverage-check-branches.md +23 -3
- package/src/skf-test-skill/references/coverage-check-tiers.md +3 -3
- package/src/skf-test-skill/references/coverage-check.md +30 -43
- package/src/skf-test-skill/references/init.md +1 -1
- package/src/skf-test-skill/references/inventory-contract.md +27 -0
- package/src/skf-test-skill/references/invocation-contract.md +3 -3
- package/src/skf-test-skill/references/report.md +3 -3
- package/src/skf-test-skill/references/scoring-rules.md +6 -6
- package/src/skf-test-skill/references/signature-comparison-rules.md +72 -0
- package/src/skf-test-skill/references/source-access-protocol.md +12 -8
- package/src/skf-test-skill/references/step-hard-gate.md +1 -0
- package/src/skf-test-skill/scripts/build-result-context.py +20 -3
- package/src/skf-test-skill/scripts/gap-ledger.py +78 -13
- package/src/skf-test-skill/scripts/load-coverage-inputs.py +345 -9
- package/src/skf-test-skill/scripts/locate-export-segments.py +591 -63
- package/src/skf-test-skill/scripts/score-signatures.py +939 -74
- package/src/skf-test-skill/scripts/validate-inventory.py +245 -29
- package/src/skf-update-skill/SKILL.md +1 -0
- package/src/skf-update-skill/references/gap-driven.md +28 -22
- package/src/skf-update-skill/references/invocation-contract.md +1 -1
- package/src/skf-update-skill/references/merge.md +4 -4
- package/src/skf-update-skill/references/report.md +1 -1
- package/src/skf-update-skill/references/write.md +5 -4
- package/src/skf-verify-stack/SKILL.md +2 -1
|
@@ -16,17 +16,27 @@ ratios and the gap records of the wrong signatures.
|
|
|
16
16
|
plan the comparisons to make. For each documented export with a
|
|
17
17
|
signature (an inventory entry of any kind but `method` that carries
|
|
18
18
|
`params`, `return_type` or `usage_signature`; a function's `params`
|
|
19
|
-
counts when empty, its form for one that takes none
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
|
|
23
|
-
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
|
|
19
|
+
counts when empty, its form for one that takes none; a `constant`
|
|
20
|
+
holding a component carries the form the skill writes as
|
|
21
|
+
`usage_signature`, references/inventory-contract.md says, and a data
|
|
22
|
+
constant none) in the signature universe, grouped by source file,
|
|
23
|
+
with the line that defines it in the source, the documented
|
|
24
|
+
signature (with the entry's `signature_form` when it has one; of
|
|
25
|
+
several entries of one name, one that lists parameters or gives a
|
|
26
|
+
usage form before one that does neither, then a full one before a
|
|
27
|
+
partial one, then, of two full or two partial forms, one that lists
|
|
28
|
+
parameters before one that gives only a usage form, else the first)
|
|
29
|
+
and, as `alternatives`, the name's other distinct forms that list
|
|
30
|
+
parameters or give a usage form, in that order: the judge compares
|
|
31
|
+
the declaration with each, by its own reading rules, and reports a
|
|
32
|
+
mismatch for each form that does not match, naming it as `form`. A
|
|
33
|
+
form that lists no member is never an alternative, and two entries
|
|
34
|
+
that say the same thing are one form (a `full` label and none, a
|
|
35
|
+
list and its items joined with `, `, whitespace runs; of two that
|
|
36
|
+
differ only by the partial label, the full one). The signature universe is the
|
|
37
|
+
surface's `all` set plus every name on its export records, so the
|
|
38
|
+
names of a Python sub-package load-coverage-inputs.py keeps out of
|
|
39
|
+
`all` (excluded.nestedEntries) keep their checks. A check sits at the
|
|
30
40
|
name's first surface record (one with a file before one with none
|
|
31
41
|
or a blank one). For a name the surface records in more than one
|
|
32
42
|
file, --skill-dir (the skill package) picks the record whose file
|
|
@@ -34,8 +44,8 @@ ratios and the gap records of the wrong signatures.
|
|
|
34
44
|
(validate-inventory.py's _skill_cited_files) when exactly one of
|
|
35
45
|
those files is cited; otherwise the check stays at the first record
|
|
36
46
|
and lists every record with a file as `candidates` for the judge to
|
|
37
|
-
read. Also the whole documented-signature map,
|
|
38
|
-
that has no surface lines yet.
|
|
47
|
+
read. Also the whole documented-signature map, and each name's
|
|
48
|
+
alternatives, for the fallback scan that has no surface lines yet.
|
|
39
49
|
score the categories. Reads every subagent result, checks its schema,
|
|
40
50
|
and computes:
|
|
41
51
|
Signature Accuracy = matching / compared * 100, where compared is
|
|
@@ -51,8 +61,13 @@ ratios and the gap records of the wrong signatures.
|
|
|
51
61
|
interface) and documented types those the inventory documents
|
|
52
62
|
A category with nothing to compare scores 100 (nothing documented
|
|
53
63
|
is wrong) and says so in `warnings`. Each signature mismatch of a
|
|
54
|
-
compared name is graded (grade())
|
|
55
|
-
|
|
64
|
+
compared name is graded (grade()) against the documented form its
|
|
65
|
+
`form` names (0, or none, the check's `documented`; n its nth
|
|
66
|
+
`alternatives` entry; an index past the name's forms is listed in
|
|
67
|
+
`warnings` and graded against `documented`), by that form's own
|
|
68
|
+
`signature_form`. With --plan the forms are the plan's (a name whose
|
|
69
|
+
forms differ from the inventory's is listed in `warnings`). Each
|
|
70
|
+
becomes a record in the gap ledger's record format (gap-ledger.py append --input reads
|
|
56
71
|
--gaps-output). The judge's `difference` tag alone never lowers a
|
|
57
72
|
grade: the script checks the fact each one rests on.
|
|
58
73
|
Low `signature-mismatch` `difference` "parameter-name" reported
|
|
@@ -65,6 +80,58 @@ ratios and the gap records of the wrong signatures.
|
|
|
65
80
|
Info `observation`, titled `Partial signature: {name}`
|
|
66
81
|
`difference` "omitted-optional" on a documented signature whose
|
|
67
82
|
inventory entry is `signature_form` "partial"
|
|
83
|
+
Info `observation`, titled `Annotation contradicted: {name}`
|
|
84
|
+
`difference` "annotation-contradicted" (the documented return
|
|
85
|
+
type is the one the body returns, the declared annotation says
|
|
86
|
+
otherwise) whose facts check: the documented entry states a
|
|
87
|
+
return type (a non-blank `return_type`, or a `documented_sig`
|
|
88
|
+
that writes one); `documented_sig` and `source_sig` write the
|
|
89
|
+
same parameters as signature-comparison-rules.md reads them
|
|
90
|
+
(_parameter_difference, below); `evidence_line` comes after the
|
|
91
|
+
declaration line (of the name's surface records at the
|
|
92
|
+
mismatch's file, the closest at or before the mismatch's
|
|
93
|
+
`line`, else the first; with none, the mismatch's `line`) and
|
|
94
|
+
before the next surface record in that file; and, with
|
|
95
|
+
--source-root, that line of the file under it starts with
|
|
96
|
+
`return`, or is the declaration line itself with its
|
|
97
|
+
expression body after its parameters (an arrow function's
|
|
98
|
+
`=> expr`, a Python `def ...: return x`). The parameters are
|
|
99
|
+
read from each one's parameter group, past a decorator's, a Go
|
|
100
|
+
receiver's or a C++ `operator()`'s group (a qualifier such as
|
|
101
|
+
`noexcept(...)` after a group leaves it the parameter group),
|
|
102
|
+
and compared by name and in order, by name alone after a `*`
|
|
103
|
+
marker; by type wherever both write one, with the common
|
|
104
|
+
Equivalent spellings (_type_key), a documented type that names
|
|
105
|
+
some members of the source's union included, and, in a TS/JS
|
|
106
|
+
file, with a literal default's type for a source parameter
|
|
107
|
+
that writes none; by optionality when the documented list
|
|
108
|
+
states one for some member; and by the type parameters both
|
|
109
|
+
write before the group. A list of bare names need not write an
|
|
110
|
+
optional parameter (unless leaving it out shifts a caller's
|
|
111
|
+
positions in a POSITIONAL_EXTENSIONS file), and one holding an
|
|
112
|
+
ellipsis claims nothing about what it stands for. A keyword or
|
|
113
|
+
name before the group, a default's value, a `*` or `/` marker,
|
|
114
|
+
the names of `*args` and `**kwargs`, and a leading receiver the
|
|
115
|
+
source writes (`self`, `cls`, `&self`, `this: T`) are never
|
|
116
|
+
compared; the documented list's own receiver is dropped only
|
|
117
|
+
when the source's list leads with one too, so a module-level
|
|
118
|
+
function whose first parameter is named `self` or `cls` is read
|
|
119
|
+
as a method. A signature with no group, on either side, is
|
|
120
|
+
compared by the text before its return. Still refused, on the
|
|
121
|
+
strict side, where the rules would read the two as one: type
|
|
122
|
+
variables renamed consistently, a type alias against the type
|
|
123
|
+
it names, an overload set, a union of callables against one
|
|
124
|
+
callable returning the union, a call example's `name=value`
|
|
125
|
+
(read as a default, so its list claims to be complete) or
|
|
126
|
+
positional literal (`RecursiveSplitter(512)`, read as a name),
|
|
127
|
+
a destructured parameter (compared as one name, so the order of
|
|
128
|
+
its properties and the ones it leaves out count), a parameter
|
|
129
|
+
in another language's order (Go's `id string`, C's `int n`,
|
|
130
|
+
read as one name), and a `;`-joined text, read as written. Its
|
|
131
|
+
remediation changes nothing in the skill: the annotation is
|
|
132
|
+
reported upstream. A tagged mismatch whose facts do not check is
|
|
133
|
+
graded as an untagged one, and `warnings` names the fact that
|
|
134
|
+
failed
|
|
68
135
|
Critical `signature-mismatch` every other mismatch
|
|
69
136
|
A mismatch's `file` (normalized: `/` separated, no `./`) is the file
|
|
70
137
|
it is graded and recorded at (the gap's Source) when the plan
|
|
@@ -74,10 +141,19 @@ ratios and the gap records of the wrong signatures.
|
|
|
74
141
|
the result's `file` stands, as it does when there is none. Results
|
|
75
142
|
that report one declaration (name, file and line) more than once
|
|
76
143
|
keep its most severe grade: Critical, then Low, then Info. Of the
|
|
77
|
-
mismatches of a check that offered several files, the
|
|
78
|
-
stands (the closest candidate), the first by file and
|
|
79
|
-
tie: one documented
|
|
80
|
-
|
|
144
|
+
mismatches of one form of a check that offered several files, the
|
|
145
|
+
least severe stands (the closest candidate), the first by file and
|
|
146
|
+
line on a tie: one documented form, one gap. Each form's mismatch
|
|
147
|
+
stands on its own (each usage of a `usage_signature` list too: the
|
|
148
|
+
mismatch's `documented_sig` is part of the declaration's key), and
|
|
149
|
+
a gap of a name with several forms names its form, counted from 1,
|
|
150
|
+
in its title (`Signature mismatch: {name} (form n of m)`) and its
|
|
151
|
+
Issue. For such a name the form's own label lowers a grade only when
|
|
152
|
+
the mismatch's `documented_sig` is that form's text (whitespace runs
|
|
153
|
+
as one, outer parentheses off, a list joined): otherwise `warnings`
|
|
154
|
+
says so and the mismatch is graded against the form it quotes, or
|
|
155
|
+
`documented` when it quotes none. A Low or an Info mismatch still
|
|
156
|
+
counts as matching. A mismatch a result reports for any
|
|
81
157
|
other name (one the skill does not document, or the surface lacks)
|
|
82
158
|
is listed in `warnings` instead, and no gap.
|
|
83
159
|
|
|
@@ -96,44 +172,62 @@ Inputs, all files in the run folder:
|
|
|
96
172
|
"types_found": ["..."], (fallback scan only)
|
|
97
173
|
"signature_mismatches": [
|
|
98
174
|
{"name", "line", "source_sig", "documented_sig", "issue",
|
|
99
|
-
"difference", "file"}]}
|
|
175
|
+
"difference", "file", "evidence_line", "form"}]}
|
|
176
|
+
(difference optional: one of
|
|
100
177
|
DIFFERENCES, absent or null for none;
|
|
101
178
|
file optional: the candidate the
|
|
102
179
|
mismatch was found at, a non-empty
|
|
103
180
|
string, absent or null for the
|
|
104
|
-
result's `file
|
|
181
|
+
result's `file`; evidence_line, the
|
|
182
|
+
body line that returns the documented
|
|
183
|
+
type, an integer from 1, required
|
|
184
|
+
with "annotation-contradicted" and
|
|
185
|
+
refused with any other difference;
|
|
186
|
+
form optional: the index of the
|
|
187
|
+
documented form the mismatch is
|
|
188
|
+
reported for, an integer from 0
|
|
189
|
+
(`documented`; n the check's nth
|
|
190
|
+
`alternatives` entry), absent or null
|
|
191
|
+
for 0)
|
|
105
192
|
--skill-dir plan only: the skill package under test (the folder that
|
|
106
193
|
holds SKILL.md), whose citations pick the declaration of a
|
|
107
194
|
name the surface records in several files
|
|
108
195
|
--plan score only: the plan's output, whose checks name the files a
|
|
109
196
|
mismatch of each name may be recorded at
|
|
197
|
+
--source-root score only: the local source tree the judges read (a
|
|
198
|
+
drifted workspace included), where an annotation-contradicted
|
|
199
|
+
mismatch's `evidence_line` is checked in its file
|
|
110
200
|
|
|
111
201
|
Rounding is compute-score.py's (JS-compatible half-up, two decimals).
|
|
112
202
|
|
|
113
203
|
Usage:
|
|
114
204
|
uv run score-signatures.py plan --inventory <file> [--surface <file>] [--skill-dir <dir>] [--output <file>]
|
|
115
205
|
uv run score-signatures.py score --inventory <file> --surface <file> [--surface-set <set>] [--plan <file>] \\
|
|
116
|
-
--results <file> [--results <file> ...] [--output <file>] [--gaps-output <file>]
|
|
206
|
+
[--source-root <dir>] --results <file> [--results <file> ...] [--output <file>] [--gaps-output <file>]
|
|
117
207
|
|
|
118
208
|
Output (stdout; --output also writes it, --gaps-output writes the records):
|
|
119
209
|
plan: {"compared": N, "files": [{"file", "checks": [{"name", "line",
|
|
120
|
-
"signatureLine", "documented": {...}, "
|
|
121
|
-
|
|
122
|
-
|
|
210
|
+
"signatureLine", "documented": {...}, "alternatives": [{...}]
|
|
211
|
+
(only when the name has another form that lists members),
|
|
212
|
+
"candidates": [{"file", "line", "signatureLine"}] (only when the
|
|
213
|
+
citation does not settle the declaration)}]}],
|
|
214
|
+
"documentedSignatures": {...}, "documentedAlternatives": {name:
|
|
215
|
+
[{...}], for each name with alternatives}}
|
|
123
216
|
score: {"valid": true, "signatureAccuracy", "matchingSignatures",
|
|
124
217
|
"totalDocumented", "comparedNames": [the names compared],
|
|
125
218
|
"typeCoverage", "documentedTypes", "totalTypes",
|
|
126
219
|
"missingTypes": [...], "mismatches": [{"name", "file", "line",
|
|
127
220
|
"source_sig", "documented_sig", "issue", "difference" (null for
|
|
128
|
-
none), "
|
|
221
|
+
none), "evidence_line" (with "annotation-contradicted" alone),
|
|
222
|
+
"severity" (Critical, Low or Info)}], "gapRecords": [...],
|
|
129
223
|
"warnings": [...]}
|
|
130
224
|
or {"valid": false, "violations": [...]} when a result breaks the schema
|
|
131
225
|
|
|
132
226
|
Exit codes:
|
|
133
227
|
0 the result was printed
|
|
134
228
|
1 an input file cannot be read or is not the JSON it should be, the
|
|
135
|
-
surface has no --surface-set set,
|
|
136
|
-
readable SKILL.md
|
|
229
|
+
surface has no --surface-set set, the --skill-dir folder holds no
|
|
230
|
+
readable SKILL.md, or the --source-root folder does not exist
|
|
137
231
|
2 a subagent result breaks the schema (violations[] names each breach;
|
|
138
232
|
nothing is written to --output or --gaps-output)
|
|
139
233
|
"""
|
|
@@ -144,6 +238,7 @@ import argparse
|
|
|
144
238
|
import importlib.util
|
|
145
239
|
import json
|
|
146
240
|
import math
|
|
241
|
+
import re
|
|
147
242
|
import sys
|
|
148
243
|
from pathlib import Path, PurePosixPath
|
|
149
244
|
|
|
@@ -152,10 +247,63 @@ from pathlib import Path, PurePosixPath
|
|
|
152
247
|
TYPE_KINDS = frozenset({"interface", "type", "enum", "class", "struct", "trait"})
|
|
153
248
|
SIGNATURE_FIELDS = ("params", "return_type", "usage_signature")
|
|
154
249
|
MISMATCH_FIELDS = ("name", "line", "source_sig", "documented_sig", "issue")
|
|
155
|
-
# The optional `difference` a judge tags a mismatch with
|
|
156
|
-
#
|
|
157
|
-
#
|
|
158
|
-
|
|
250
|
+
# The optional `difference` a judge tags a mismatch with, defined in
|
|
251
|
+
# references/signature-comparison-rules.md (Difference tags): only the
|
|
252
|
+
# parameter names differ, only members the documented form leaves out that
|
|
253
|
+
# the source marks optional, or only a return annotation the body contradicts
|
|
254
|
+
# (the body returns the documented type at the mismatch's `evidence_line`).
|
|
255
|
+
DIFFERENCES = ("parameter-name", "omitted-optional", "annotation-contradicted")
|
|
256
|
+
CONTRADICTED = "annotation-contradicted"
|
|
257
|
+
# grade()'s default: an annotation-contradicted mismatch whose facts no one checked stays Critical.
|
|
258
|
+
UNCHECKED = "its facts were not checked"
|
|
259
|
+
# A body line that returns starts with the `return` keyword. The declaration
|
|
260
|
+
# line itself returns when its expression body follows its parameters: an
|
|
261
|
+
# arrow function assigned with `=` or `export default` (`=> expr`, not a
|
|
262
|
+
# `{` block), or a Python `def ...: return x`. A `type` alias's arrow is a type.
|
|
263
|
+
RETURN_START_RE = re.compile(r"return(?![\w$])")
|
|
264
|
+
ARROW_OPENER_RE = re.compile(r"(?:=|\bdefault)\s*(?:async\s*)?(?:<[^()]*>\s*)?\(")
|
|
265
|
+
ARROW_BODY_RE = re.compile(r"\s*(?::[^=]*)?=>\s*[^\s{]")
|
|
266
|
+
PYTHON_DEF_RE = re.compile(r"(?:async\s+)?def\s+\w+\s*\(")
|
|
267
|
+
PYTHON_BODY_RE = re.compile(r"\s*(?:->[^:]*)?:\s*return(?![\w$])")
|
|
268
|
+
TYPE_DECLARATION_RE = re.compile(r"(?:export\s+)?(?:declare\s+)?type\s")
|
|
269
|
+
# Where a signature's return type starts, with no parenthesized parameters.
|
|
270
|
+
RETURN_ARROW_RE = re.compile(r"->|=>|→")
|
|
271
|
+
# The prefixes a type is compared without, as signature-comparison-rules.md's Equivalent spellings reads a name
|
|
272
|
+
# with or without one: `typing.List[int]` is `List[int]`, `React.ReactNode` is `ReactNode`.
|
|
273
|
+
TYPE_PREFIX_RE = re.compile(r"(?<![\w.$])(?:typing|collections\.abc|React)\.")
|
|
274
|
+
# The `typing` names Equivalent spellings reads as the builtins they alias.
|
|
275
|
+
TYPING_ALIASES = {"List": "list", "Dict": "dict", "Tuple": "tuple", "Set": "set", "FrozenSet": "frozenset",
|
|
276
|
+
"Type": "type"}
|
|
277
|
+
# A quoted forward reference (`"Node"`), read as the type it quotes.
|
|
278
|
+
FORWARD_REFERENCE_RE = re.compile(r"(['\"])([A-Za-z_][^'\"]*)\1")
|
|
279
|
+
# The name a type starts with, before its type arguments (`dict[str, Any]`, `Promise<T>`).
|
|
280
|
+
TYPE_NAME_RE = re.compile(r"[A-Za-z_$][\w.$]*")
|
|
281
|
+
# A Rust lifetime's `'a`, followed by whitespace, `,`, `>`, `+`, `)` or the end (_lifetime says where it stands).
|
|
282
|
+
LIFETIME_RE = re.compile(r"'[A-Za-z_]\w*(?=[\s,>+)]|$)")
|
|
283
|
+
# Each closing bracket and the opening one it closes (_levels: a `>` closes only a `<`).
|
|
284
|
+
CLOSERS = {")": "(", "]": "[", "}": "{"}
|
|
285
|
+
OPENERS = {"(": ")", "[": "]", "{": "}", "<": ">"}
|
|
286
|
+
# A group a decorator or a Java annotation writes: the text before it ends with `@name`.
|
|
287
|
+
DECORATED_RE = re.compile(r"@[\w.$]+\s*$")
|
|
288
|
+
# What follows a prefix group: another `(` with no space between (C++ `operator()(int x)`), or a name and a `(`
|
|
289
|
+
# after whitespace (a Go receiver, `func (s *Server) Start(port int)`), the name no C or C++ qualifier that takes
|
|
290
|
+
# arguments after a parameter group (`f(int x) noexcept(true)`).
|
|
291
|
+
PREFIX_GROUP_RE = re.compile(r"\(|\s+(?!(?:noexcept|throw|__attribute__|requires|decltype)\b)[A-Za-z_$][\w$]*\s*\(")
|
|
292
|
+
# A receiver a method's parameter list may lead with: Python's `self` and `cls`, Rust's `self`, `&self`,
|
|
293
|
+
# `&mut self` and `mut self` (with a lifetime too, `&'a self`), and TS's `this: T` (_parameter_difference).
|
|
294
|
+
RECEIVER_RE = re.compile(r"self|cls|this|mut self|&(?:'\w+ )?(?:mut )?self")
|
|
295
|
+
# An ellipsis item, standing for parameters a form does not show (a TS/JS rest `...name` is none).
|
|
296
|
+
ELLIPSES = ("...", "…")
|
|
297
|
+
# The TS/JS source files, where a parameter with no type and a literal default has that literal's type.
|
|
298
|
+
TS_JS_EXTENSIONS = frozenset({".ts", ".tsx", ".mts", ".cts", ".js", ".jsx", ".mjs", ".cjs"})
|
|
299
|
+
# A literal default's type (signature-comparison-rules.md, Annotations): a bigint, a number, a string or template
|
|
300
|
+
# literal, a boolean.
|
|
301
|
+
LITERAL_TYPES = (
|
|
302
|
+
(re.compile(r"[-+]?(?:0[xob][\da-f_]+|\d[\d_]*)n", re.I), "bigint"),
|
|
303
|
+
(re.compile(r"[-+]?(?:0[xob][\da-f_]+|(?:\d[\d_]*(?:\.[\d_]*)?|\.\d[\d_]*)(?:e[-+]?\d+)?)", re.I), "number"),
|
|
304
|
+
(re.compile(r"(['\"`]).*\1", re.S), "string"),
|
|
305
|
+
(re.compile(r"true|false"), "boolean"),
|
|
306
|
+
)
|
|
159
307
|
# Source files of the languages with no keyword arguments, where a caller
|
|
160
308
|
# never writes a parameter's name: TS/JS, Go, Rust, Java, C and C++. Not
|
|
161
309
|
# Python, which has keyword arguments (the script cannot see which
|
|
@@ -260,6 +408,26 @@ def load_plan(path: str) -> dict[str, list[str]]:
|
|
|
260
408
|
return offered
|
|
261
409
|
|
|
262
410
|
|
|
411
|
+
def load_plan_forms(path: str) -> dict[str, list[dict]]:
|
|
412
|
+
"""{name: the documented forms plan offered it, `documented` first, then
|
|
413
|
+
its `alternatives`} from plan's output: its documentedSignatures and
|
|
414
|
+
documentedAlternatives maps (the fallback scan's), then each check's."""
|
|
415
|
+
data = _read_json(path, "--plan")
|
|
416
|
+
data = data if isinstance(data, dict) else {}
|
|
417
|
+
signatures, alternatives = data.get("documentedSignatures"), data.get("documentedAlternatives")
|
|
418
|
+
alternatives = alternatives if isinstance(alternatives, dict) else {}
|
|
419
|
+
pairs = [(n, d, alternatives.get(n)) for n, d in (signatures.items() if isinstance(signatures, dict) else [])]
|
|
420
|
+
for group in data.get("files") if isinstance(data.get("files"), list) else []:
|
|
421
|
+
checks = group.get("checks") if isinstance(group, dict) else None
|
|
422
|
+
pairs += [(c.get("name"), c.get("documented"), c.get("alternatives"))
|
|
423
|
+
for c in (checks if isinstance(checks, list) else []) if isinstance(c, dict)]
|
|
424
|
+
forms: dict[str, list[dict]] = {}
|
|
425
|
+
for name, documented, others in pairs:
|
|
426
|
+
if isinstance(name, str) and isinstance(documented, dict):
|
|
427
|
+
forms[name] = [documented, *(f for f in (others if isinstance(others, list) else []) if isinstance(f, dict))]
|
|
428
|
+
return forms
|
|
429
|
+
|
|
430
|
+
|
|
263
431
|
def check_result(item: object, where: str) -> list[str]:
|
|
264
432
|
"""The schema breaches of one per-file result object."""
|
|
265
433
|
if not isinstance(item, dict):
|
|
@@ -293,9 +461,22 @@ def check_result(item: object, where: str) -> list[str]:
|
|
|
293
461
|
if entry.get("difference") is not None and entry["difference"] not in DIFFERENCES:
|
|
294
462
|
errors.append(f"{at}: `difference` must be one of {', '.join(DIFFERENCES)} when present, "
|
|
295
463
|
f"not {json.dumps(entry['difference'])}")
|
|
464
|
+
# The body line that returns the documented type: the fact an annotation-contradicted mismatch rests on.
|
|
465
|
+
evidence = entry.get("evidence_line")
|
|
466
|
+
if entry.get("difference") == CONTRADICTED:
|
|
467
|
+
if isinstance(evidence, bool) or not isinstance(evidence, int) or evidence < 1:
|
|
468
|
+
errors.append(f"{at}: `evidence_line` must be a line number (an integer from 1) with `difference` "
|
|
469
|
+
f"{CONTRADICTED}, not {json.dumps(evidence)}")
|
|
470
|
+
elif evidence is not None:
|
|
471
|
+
errors.append(f"{at}: `evidence_line` is given only with `difference` {CONTRADICTED}, "
|
|
472
|
+
f"not {json.dumps(entry.get('difference'))}")
|
|
296
473
|
# The candidate a mismatch was found at, when the plan handed the judge several.
|
|
297
474
|
if entry.get("file") is not None and not (isinstance(entry["file"], str) and entry["file"].strip()):
|
|
298
475
|
errors.append(f"{at}: `file` must be a non-empty string when present, not {json.dumps(entry['file'])}")
|
|
476
|
+
# The documented form a mismatch is reported for, when the check offered `alternatives`.
|
|
477
|
+
form = entry.get("form")
|
|
478
|
+
if form is not None and (isinstance(form, bool) or not isinstance(form, int) or form < 0):
|
|
479
|
+
errors.append(f"{at}: `form` must be an index (an integer from 0) when present, not {json.dumps(form)}")
|
|
299
480
|
return errors
|
|
300
481
|
|
|
301
482
|
|
|
@@ -369,29 +550,72 @@ def _rank(fields: dict) -> tuple[bool, bool, bool]:
|
|
|
369
550
|
return not _lists_members(fields), _partial(fields), not _lists_params(fields)
|
|
370
551
|
|
|
371
552
|
|
|
372
|
-
def
|
|
373
|
-
"""
|
|
374
|
-
|
|
375
|
-
|
|
376
|
-
|
|
377
|
-
|
|
378
|
-
|
|
379
|
-
|
|
380
|
-
|
|
381
|
-
|
|
382
|
-
|
|
553
|
+
def _form(entry: dict) -> dict:
|
|
554
|
+
"""The documented signature fields of one inventory entry, with its `signature_form` when it has one."""
|
|
555
|
+
fields = {f: entry[f] for f in SIGNATURE_FIELDS if f in entry}
|
|
556
|
+
if entry.get("signature_form") is not None:
|
|
557
|
+
fields["signature_form"] = entry["signature_form"]
|
|
558
|
+
return fields
|
|
559
|
+
|
|
560
|
+
|
|
561
|
+
def _field_text(value: object) -> str:
|
|
562
|
+
"""A signature field as one text: a list's items joined with `, `, whitespace runs as one space, "" for null."""
|
|
563
|
+
if value is None:
|
|
564
|
+
return ""
|
|
565
|
+
if isinstance(value, list):
|
|
566
|
+
value = ", ".join(v if isinstance(v, str) else json.dumps(v, sort_keys=True) for v in value)
|
|
567
|
+
return " ".join(value.split()) if isinstance(value, str) else json.dumps(value, sort_keys=True)
|
|
568
|
+
|
|
569
|
+
|
|
570
|
+
def _form_key(fields: dict) -> tuple:
|
|
571
|
+
"""What a form says, for telling two entries apart (_field_text), a
|
|
572
|
+
null, blank or empty key read as absent, and the `signature_form` label
|
|
573
|
+
left out (documented_forms keeps the full one)."""
|
|
574
|
+
return tuple((f, _field_text(fields[f])) for f in SIGNATURE_FIELDS if f in fields and _field_text(fields[f]))
|
|
575
|
+
|
|
576
|
+
|
|
577
|
+
def documented_forms(inventory: list[dict]) -> dict[str, list[dict]]:
|
|
578
|
+
"""{name: [its documented signature, *its alternatives]} for each
|
|
579
|
+
documented export with a signature: its distinct forms (_form), ranked.
|
|
580
|
+
Of several entries of one name, one that lists parameters or gives a
|
|
581
|
+
usage form (_lists_members) is compared before one that does neither
|
|
582
|
+
(empty or no `params` and no usage form: it never hides one that says
|
|
583
|
+
what the name takes), then a full one (no `signature_form`, or `full`)
|
|
584
|
+
before a partial one (a form the skill also gives in full is held to
|
|
585
|
+
it), then, of two full or two partial forms, one that lists parameters
|
|
586
|
+
is compared before one that gives only a usage form, else the first.
|
|
587
|
+
The first is the documented signature; the others that list members
|
|
588
|
+
follow it, in that order, as the alternatives the judge compares too (a
|
|
589
|
+
skill that documents a name briefly in one place and in full in
|
|
590
|
+
another). A form that lists no member is never an alternative. Two
|
|
591
|
+
entries that say the same thing (_form_key) are one form; of two that
|
|
592
|
+
differ only by the partial label, the full one (the stricter) stands."""
|
|
593
|
+
entries: dict[str, list[dict]] = {}
|
|
383
594
|
for entry in inventory:
|
|
384
595
|
name = entry.get("name")
|
|
385
596
|
if entry.get("kind") == "method" or not (isinstance(name, str) and name) or not _has_signature(entry):
|
|
386
597
|
continue
|
|
387
|
-
fields =
|
|
388
|
-
if
|
|
389
|
-
|
|
390
|
-
|
|
391
|
-
|
|
598
|
+
forms, fields = entries.setdefault(name, []), _form(entry)
|
|
599
|
+
same = next((i for i, f in enumerate(forms) if _form_key(f) == _form_key(fields)), None)
|
|
600
|
+
if same is None:
|
|
601
|
+
forms.append(fields)
|
|
602
|
+
elif _partial(forms[same]) and not _partial(fields):
|
|
603
|
+
# in the inventory order the full entry has, as the rank would have ordered it
|
|
604
|
+
del forms[same]
|
|
605
|
+
forms.append(fields)
|
|
606
|
+
out: dict[str, list[dict]] = {}
|
|
607
|
+
for name, forms in entries.items():
|
|
608
|
+
first, *others = sorted(forms, key=_rank) # stable: of forms that rank alike, inventory order
|
|
609
|
+
out[name] = [first, *(f for f in others if _lists_members(f))]
|
|
392
610
|
return out
|
|
393
611
|
|
|
394
612
|
|
|
613
|
+
def documented_signatures(inventory: list[dict]) -> dict[str, dict]:
|
|
614
|
+
"""{name: its documented signature fields}: the first of its
|
|
615
|
+
documented_forms, the one a check holds as `documented`."""
|
|
616
|
+
return {name: forms[0] for name, forms in documented_forms(inventory).items()}
|
|
617
|
+
|
|
618
|
+
|
|
395
619
|
def documented_names(inventory: list[dict]) -> set[str]:
|
|
396
620
|
return {e["name"] for e in inventory
|
|
397
621
|
if isinstance(e.get("name"), str) and e["name"] and e.get("kind") != "method"}
|
|
@@ -451,29 +675,35 @@ def plan(inventory: list[dict], names: list[str] | None, exports: list[dict], sk
|
|
|
451
675
|
"""The comparisons over the signature universe of `names` (the `all`
|
|
452
676
|
set) and `exports`; none without a surface (`names` None). `skill_dir`,
|
|
453
677
|
the skill package, settles which declaration a name the surface records
|
|
454
|
-
in several files is compared with (_check_at).
|
|
455
|
-
|
|
678
|
+
in several files is compared with (_check_at). A name documented in
|
|
679
|
+
several forms carries the others that list members as `alternatives`
|
|
680
|
+
(documented_forms)."""
|
|
681
|
+
forms = documented_forms(inventory)
|
|
456
682
|
files: dict[str, list[dict]] = {}
|
|
457
683
|
compared = 0
|
|
458
684
|
if names is not None:
|
|
459
685
|
by_name = _records_by_name(exports)
|
|
460
686
|
texts = _inventory().skill_markdown_lines(Path(skill_dir)) if skill_dir is not None else None
|
|
461
|
-
for name in sorted(signature_universe(names, exports) & set(
|
|
687
|
+
for name in sorted(signature_universe(names, exports) & set(forms)):
|
|
462
688
|
record, candidates = _check_at(name, by_name.get(name) or [], skill_dir, texts)
|
|
463
689
|
compared += 1
|
|
464
690
|
check = {
|
|
465
691
|
"name": name,
|
|
466
692
|
"line": record.get("line"),
|
|
467
693
|
"signatureLine": record.get("signatureLine"),
|
|
468
|
-
"documented":
|
|
694
|
+
"documented": forms[name][0],
|
|
469
695
|
}
|
|
696
|
+
if len(forms[name]) > 1:
|
|
697
|
+
check["alternatives"] = forms[name][1:]
|
|
470
698
|
if candidates:
|
|
471
699
|
check["candidates"] = candidates
|
|
472
700
|
files.setdefault(_record_file(record), []).append(check)
|
|
473
701
|
return {
|
|
474
702
|
"compared": compared,
|
|
475
703
|
"files": [{"file": f, "checks": checks} for f, checks in sorted(files.items())],
|
|
476
|
-
"documentedSignatures":
|
|
704
|
+
"documentedSignatures": documented_signatures(inventory),
|
|
705
|
+
# The fallback scan's alternatives: a name's form n is its nth entry here, as in a check's `alternatives`.
|
|
706
|
+
"documentedAlternatives": {name: fs[1:] for name, fs in forms.items() if len(fs) > 1},
|
|
477
707
|
}
|
|
478
708
|
|
|
479
709
|
|
|
@@ -502,20 +732,545 @@ def _is_offered_file(files: list[str], file: str) -> bool:
|
|
|
502
732
|
return _path(file) in {_path(f) for f in files}
|
|
503
733
|
|
|
504
734
|
|
|
505
|
-
def
|
|
735
|
+
def _is_line(value: object) -> bool:
|
|
736
|
+
return isinstance(value, int) and not isinstance(value, bool) and value >= 1
|
|
737
|
+
|
|
738
|
+
|
|
739
|
+
def _lines_in(exports: list[dict], file: str) -> list[int]:
|
|
740
|
+
"""The lines of the surface records in `file`, however either writes its path."""
|
|
741
|
+
return [r["line"] for r in exports
|
|
742
|
+
if _record_file(r) and _path(_record_file(r)) == _path(file) and _is_line(r.get("line"))]
|
|
743
|
+
|
|
744
|
+
|
|
745
|
+
def _declaration_line(by_name: dict[str, list[dict]], name: str, file: str, line: int) -> int:
|
|
746
|
+
"""The line `name` is declared at in `file`: of its surface records there
|
|
747
|
+
(the plan's check, or the candidate it offered), the closest one at or
|
|
748
|
+
before the mismatch's `line` (a name declared twice in one file), else
|
|
749
|
+
the first; with none there, the mismatch's own `line`."""
|
|
750
|
+
lines = _lines_in(by_name.get(name) or [], file)
|
|
751
|
+
before = [at for at in lines if at <= line]
|
|
752
|
+
return max(before) if before else min(lines, default=line)
|
|
753
|
+
|
|
754
|
+
|
|
755
|
+
def _next_declaration(exports: list[dict], file: str, line: int) -> int | None:
|
|
756
|
+
"""The line of the first surface record in `file` after `line`, None when there is none."""
|
|
757
|
+
return min((at for at in _lines_in(exports, file) if at > line), default=None)
|
|
758
|
+
|
|
759
|
+
|
|
760
|
+
def _source_lines(source_root: str, file: str) -> list[str] | None:
|
|
761
|
+
"""The lines of `file` under `source_root`, None when it is no readable
|
|
762
|
+
file there (a path that leaves the root included)."""
|
|
763
|
+
root = Path(source_root).resolve()
|
|
764
|
+
path = (root / file).resolve()
|
|
765
|
+
try:
|
|
766
|
+
path.relative_to(root)
|
|
767
|
+
return path.read_bytes().decode("utf-8-sig", errors="replace").splitlines()
|
|
768
|
+
except (ValueError, OSError):
|
|
769
|
+
return None
|
|
770
|
+
|
|
771
|
+
|
|
772
|
+
def _after_parens(text: str, start: int) -> int:
|
|
773
|
+
"""The index past the `)` that closes the `(` at `start`, -1 when none does."""
|
|
774
|
+
depth = 0
|
|
775
|
+
for i in range(start, len(text)):
|
|
776
|
+
depth += {"(": 1, ")": -1}.get(text[i], 0)
|
|
777
|
+
if depth == 0:
|
|
778
|
+
return i + 1
|
|
779
|
+
return -1
|
|
780
|
+
|
|
781
|
+
|
|
782
|
+
def _string_end(text: str, start: int) -> int:
|
|
783
|
+
"""The index past the string literal whose quote is at `start`: up to the same quote not escaped, the end of
|
|
784
|
+
`text` when none closes it."""
|
|
785
|
+
quote, i = text[start], start + 1
|
|
786
|
+
while i < len(text):
|
|
787
|
+
if text[i] == "\\":
|
|
788
|
+
i += 2
|
|
789
|
+
elif text[i] == quote:
|
|
790
|
+
return i + 1
|
|
791
|
+
else:
|
|
792
|
+
i += 1
|
|
793
|
+
return len(text)
|
|
794
|
+
|
|
795
|
+
|
|
796
|
+
def _lifetime(text: str, i: int, in_angle: bool, extension: str) -> bool:
|
|
797
|
+
"""Whether the `'` at `i` starts a Rust lifetime: only in a `.rs` file, `'ident` followed by whitespace, `,`,
|
|
798
|
+
`>`, `+`, `)` or the end, coming after a `&` or a `+` (whitespace aside) or inside a `<...>` (`in_angle`), as
|
|
799
|
+
in `&'a str`, `impl Tr + 'a` and `<'a>`. Any other `'` opens a string, so a TS `Omit<P, 'a,b'>` holds one."""
|
|
800
|
+
return (extension == ".rs" and bool(LIFETIME_RE.match(text, i))
|
|
801
|
+
and (in_angle or text[:i].rstrip()[-1:] in ("&", "+")))
|
|
802
|
+
|
|
803
|
+
|
|
804
|
+
def _opens(text: str, i: int, extension: str) -> bool:
|
|
805
|
+
"""Whether the `<` at `i` opens a pair: it follows a letter, a digit, `_`, `$`, `.`, `]` or `>` with no
|
|
806
|
+
whitespace between (`Map<`, `f<T>`, `obj.get<T>`), and a `>` that is not part of `->` or `=>` closes it
|
|
807
|
+
before the bracket pair around it closes or the text ends. Any other `<` is an operator (`x < y`, `1 << 20`,
|
|
808
|
+
`x<3)`)."""
|
|
809
|
+
if i == 0 or not (text[i - 1].isalnum() or text[i - 1] in "_$.]>"):
|
|
810
|
+
return False
|
|
811
|
+
depth, j = 0, i + 1
|
|
812
|
+
while j < len(text):
|
|
813
|
+
ch = text[j]
|
|
814
|
+
if ch in "\"`" or (ch == "'" and not _lifetime(text, j, depth == 0, extension)):
|
|
815
|
+
j = _string_end(text, j)
|
|
816
|
+
continue
|
|
817
|
+
if ch in "([{":
|
|
818
|
+
depth += 1
|
|
819
|
+
elif ch in ")]}":
|
|
820
|
+
if depth == 0:
|
|
821
|
+
return False
|
|
822
|
+
depth -= 1
|
|
823
|
+
elif ch == ">" and depth == 0 and text[j - 1] not in "-=":
|
|
824
|
+
return True
|
|
825
|
+
j += 1
|
|
826
|
+
return False
|
|
827
|
+
|
|
828
|
+
|
|
829
|
+
def _levels(text: str, extension: str = "") -> list[int | None]:
|
|
830
|
+
"""For each character of `text`, in a file of `extension`, the number of bracket pairs open around it, None
|
|
831
|
+
inside a string literal (its quotes included). `(`, `[` and `{` open a pair, and a `<` opens one only when it
|
|
832
|
+
follows a name, a `.`, a `]` or a `>` with no whitespace between and a `>` that is not part of `->` or `=>`
|
|
833
|
+
closes it before the bracket pair around it closes or the text ends (_opens); any other `<` is an operator
|
|
834
|
+
and nests nothing, so `(path, size=1 << 20, mode)` holds three parameters,
|
|
835
|
+
`(a: boolean = x < y, b: Map<string, number>)` two, and `Map<string, number>` keeps its comma. An opening
|
|
836
|
+
bracket and the one that closes it stand at the level outside them. A closing bracket closes the pair it
|
|
837
|
+
matches and any `<` left open inside it, and a `>` closes only a `<`, never as part of `->` or `=>`. A `"`, a
|
|
838
|
+
backquote or a `'` opens a string (_string_end), except the `'` of a Rust lifetime in a `.rs` file
|
|
839
|
+
(_lifetime)."""
|
|
840
|
+
levels: list[int | None] = []
|
|
841
|
+
stack: list[str] = []
|
|
842
|
+
i = 0
|
|
843
|
+
while i < len(text):
|
|
844
|
+
ch = text[i]
|
|
845
|
+
if ch in "\"`" or (ch == "'" and not _lifetime(text, i, bool(stack) and stack[-1] == "<", extension)):
|
|
846
|
+
end = _string_end(text, i)
|
|
847
|
+
levels.extend([None] * (end - i))
|
|
848
|
+
i = end
|
|
849
|
+
continue
|
|
850
|
+
if ch in "([{" or (ch == "<" and _opens(text, i, extension)):
|
|
851
|
+
levels.append(len(stack))
|
|
852
|
+
stack.append(ch)
|
|
853
|
+
elif ch in CLOSERS and CLOSERS[ch] in stack:
|
|
854
|
+
while stack.pop() != CLOSERS[ch]:
|
|
855
|
+
pass
|
|
856
|
+
levels.append(len(stack))
|
|
857
|
+
elif ch == ">" and stack and stack[-1] == "<" and text[i - 1:i] not in ("-", "="):
|
|
858
|
+
stack.pop()
|
|
859
|
+
levels.append(len(stack))
|
|
860
|
+
else:
|
|
861
|
+
levels.append(len(stack))
|
|
862
|
+
i += 1
|
|
863
|
+
return levels
|
|
864
|
+
|
|
865
|
+
|
|
866
|
+
def _split(text: str, separator: str, extension: str = "") -> list[str]:
|
|
867
|
+
"""`text` cut at each `separator` outside every bracket pair and string (_levels)."""
|
|
868
|
+
levels = _levels(text, extension)
|
|
869
|
+
cuts = [-1, *(i for i, ch in enumerate(text) if ch == separator and levels[i] == 0), len(text)]
|
|
870
|
+
return [text[a + 1:b] for a, b in zip(cuts, cuts[1:])]
|
|
871
|
+
|
|
872
|
+
|
|
873
|
+
def _encloses(text: str, start: int, extension: str = "") -> bool:
|
|
874
|
+
"""Whether the bracket at `start` of `text` opens a pair that the last character of `text` closes."""
|
|
875
|
+
levels = _levels(text, extension)
|
|
876
|
+
return (OPENERS.get(text[start]) == text[-1] and levels[start] == 0 == levels[-1] and len(text) - start > 1
|
|
877
|
+
and all(level is None or level > 0 for level in levels[start + 1:-1]))
|
|
878
|
+
|
|
879
|
+
|
|
880
|
+
def _group(text: str, extension: str = "") -> tuple[int, int] | None:
|
|
881
|
+
"""The indices of the `(` and the `)` of the parameter group of `text`: the first pair of parentheses outside
|
|
882
|
+
every bracket pair and string (_levels), before any return arrow there, that is no prefix group. A prefix group
|
|
883
|
+
is one a decorator or a Java annotation writes (the text before it ends with `@name`:
|
|
884
|
+
`@lru_cache(maxsize=None) def f(x)`), one directly followed by another `(` (C++ `operator()(int x)`), or one
|
|
885
|
+
followed, after whitespace, by a name and a `(` (a Go receiver, `func (s *Server) Start(port int)`), unless
|
|
886
|
+
that name is a C or C++ qualifier (`noexcept`, `throw`, `__attribute__`, `requires`, `decltype`: `f(int x)
|
|
887
|
+
noexcept(true)` keeps its group). A `(` after a space is no prefix either (Go's result list,
|
|
888
|
+
`Start(port int) (int, error)`). So a keyword, a name or type parameters before the group are never part of
|
|
889
|
+
it. None when there is no such group: no `(` comes before the return arrow (`make -> X`), or a group does not
|
|
890
|
+
close."""
|
|
891
|
+
levels = _levels(text, extension)
|
|
892
|
+
i = 0
|
|
893
|
+
while i < len(text):
|
|
894
|
+
if levels[i] == 0 and RETURN_ARROW_RE.match(text, i):
|
|
895
|
+
return None
|
|
896
|
+
if text[i] == "(" and levels[i] == 0:
|
|
897
|
+
end = next((j for j in range(i + 1, len(text)) if text[j] == ")" and levels[j] == 0), None)
|
|
898
|
+
if end is None:
|
|
899
|
+
return None
|
|
900
|
+
if not (DECORATED_RE.search(text, 0, i) or PREFIX_GROUP_RE.match(text, end + 1)):
|
|
901
|
+
return i, end
|
|
902
|
+
i = end
|
|
903
|
+
i += 1
|
|
904
|
+
return None
|
|
905
|
+
|
|
906
|
+
|
|
907
|
+
def _signature_parts(sig: str, extension: str = "") -> tuple[str, str]:
|
|
908
|
+
"""(the text before the return, its whitespace runs read as one space; the return type it writes, "" for
|
|
909
|
+
none) of a signature, in a file of `extension`. The return follows the parameter group (_group, the one
|
|
910
|
+
_parameters reads), after a `->`, `=>`, `→` or `:`, or, with no group, the first such arrow. A `;` is read as
|
|
911
|
+
any other character."""
|
|
912
|
+
text = sig.strip()
|
|
913
|
+
group = _group(text, extension)
|
|
914
|
+
if group is None:
|
|
915
|
+
arrow = RETURN_ARROW_RE.search(text)
|
|
916
|
+
head, tail = (text[:arrow.start()], text[arrow.end():]) if arrow else (text, "")
|
|
917
|
+
else:
|
|
918
|
+
head, tail = text[:group[1] + 1], re.sub(r"^\s*(?:->|=>|→|:)", "", text[group[1] + 1:])
|
|
919
|
+
return " ".join(head.split()), tail.strip().rstrip("{:").strip()
|
|
920
|
+
|
|
921
|
+
|
|
922
|
+
def _compact(text: str) -> str:
|
|
923
|
+
"""`text` with its whitespace runs read as one space, and none beside a character that is no letter, digit
|
|
924
|
+
or `_`, so `{ title }` is `{title}` and `dict[str, Any]` is `dict[str,Any]`."""
|
|
925
|
+
return re.sub(r"\s*([^\w\s])\s*", r"\1", " ".join(text.split()))
|
|
926
|
+
|
|
927
|
+
|
|
928
|
+
def _type_parameters(sig: str, extension: str = "") -> str:
|
|
929
|
+
"""The type parameters a signature writes between its name and its parameter group (`<T extends X>`, or
|
|
930
|
+
Python's `[T: int]`), compacted and without a trailing comma; "" for none, or no group."""
|
|
931
|
+
text = sig.strip()
|
|
932
|
+
group = _group(text, extension)
|
|
933
|
+
if group is None:
|
|
934
|
+
return ""
|
|
935
|
+
levels = _levels(text, extension)
|
|
936
|
+
k = len(text[:group[0]].rstrip()) - 1
|
|
937
|
+
if k < 1 or text[k] not in ">]" or levels[k] is None:
|
|
938
|
+
return ""
|
|
939
|
+
j = next((j for j in range(k - 1, -1, -1) if levels[j] is not None and levels[j] <= levels[k]), None)
|
|
940
|
+
if j is None or OPENERS.get(text[j]) != text[k] or levels[j] != levels[k]:
|
|
941
|
+
return ""
|
|
942
|
+
return _compact(text[j + 1:k]).rstrip(",")
|
|
943
|
+
|
|
944
|
+
|
|
945
|
+
def _parameter(text: str, extension: str = "") -> dict:
|
|
946
|
+
"""One parameter as _parameters reads it, but for its "by_name" and "receiver": its name is the text before
|
|
947
|
+
the `:` of its type and the `=` of its default (each outside every bracket pair and string; the `=` of `==`,
|
|
948
|
+
`=>`, `<=`, `>=` or `!=` is no default), the whole text when it writes neither, so a parameter in another
|
|
949
|
+
language's order (Go's `id string`) or a destructured one (`{ a, b }`) is read as one name."""
|
|
950
|
+
levels = _levels(text, extension)
|
|
951
|
+
at = next((i for i, ch in enumerate(text) if ch == "=" and levels[i] == 0
|
|
952
|
+
and text[i + 1:i + 2] not in ("=", ">") and text[i - 1:i] not in ("=", "!", "<", ">")), None)
|
|
953
|
+
head = text if at is None else text[:at]
|
|
954
|
+
colon = next((i for i, ch in enumerate(head) if ch == ":" and levels[i] == 0), None)
|
|
955
|
+
name = _compact(head if colon is None else head[:colon])
|
|
956
|
+
optional = at is not None or name.endswith("?")
|
|
957
|
+
name = name.rstrip("?")
|
|
958
|
+
key = "**" if name.startswith("**") else "*" if name.startswith("*") else name
|
|
959
|
+
return {"key": key, "name": name, "type": "" if colon is None else _compact(head[colon + 1:]),
|
|
960
|
+
"default": None if at is None else text[at + 1:].strip(), "optional": optional,
|
|
961
|
+
"variadic": key in ("*", "**") or (name.startswith("...") and name not in ELLIPSES),
|
|
962
|
+
"ellipsis": name in ELLIPSES and colon is None and at is None}
|
|
963
|
+
|
|
964
|
+
|
|
965
|
+
def _parameters(sig: str, extension: str = "") -> list[dict] | None:
|
|
966
|
+
"""The parameters the group of `sig` writes (_group), in a file of `extension`, None when it has no group:
|
|
967
|
+
the group split at its commas outside every bracket pair and string (_levels), so a comma inside a nested
|
|
968
|
+
type, a callback's parameters or a default's string splits nothing, without a bare `*` or `/` marker. Each is
|
|
969
|
+
{"key": its name (_parameter), `*` for any `*args` and `**` for any `**kwargs`, whose names are never
|
|
970
|
+
compared; "name": its name as written; "type": the type it writes, compacted, "" for none; "default": the text
|
|
971
|
+
of its default, None for none; "optional": whether a default or a `?` on its name marks it optional;
|
|
972
|
+
"variadic": whether it is a `*args`, a `**kwargs` or a TS/JS rest `...name`; "ellipsis": whether it is an
|
|
973
|
+
ellipsis item (`...`, `…`) standing for members it does not show; "by_name": whether it comes after a `*`
|
|
974
|
+
marker, a bare `*` or a `*args`; "receiver": whether it is the group's first item and a receiver
|
|
975
|
+
(RECEIVER_RE: `self`, `cls`, `&self`, `&mut self`, `mut self`, `this: T`), which _parameter_difference
|
|
976
|
+
drops}. A `sig` that holds a `;` outside every bracket pair and string (shared text before a name's own
|
|
977
|
+
members, an overload set or a usage list joined by `;`) has none either, so it is compared as text, read as
|
|
978
|
+
written: its first group may be a call inside the shared text."""
|
|
979
|
+
text = sig.strip()
|
|
980
|
+
group = _group(text, extension)
|
|
981
|
+
if group is None or len(_split(text, ";", extension)) > 1:
|
|
982
|
+
return None
|
|
983
|
+
params: list[dict] = []
|
|
984
|
+
by_name = False
|
|
985
|
+
for index, item in enumerate(_split(text[group[0] + 1:group[1]], ",", extension)):
|
|
986
|
+
item = item.strip()
|
|
987
|
+
if item in ("", "/"):
|
|
988
|
+
continue
|
|
989
|
+
if item == "*":
|
|
990
|
+
by_name = True
|
|
991
|
+
continue
|
|
992
|
+
param = _parameter(item, extension)
|
|
993
|
+
param.update(by_name=by_name, receiver=index == 0 and bool(RECEIVER_RE.fullmatch(param["name"])))
|
|
994
|
+
by_name = by_name or param["key"] == "*"
|
|
995
|
+
params.append(param)
|
|
996
|
+
return params
|
|
997
|
+
|
|
998
|
+
|
|
999
|
+
def _type_key(text: str, extension: str = "") -> frozenset:
|
|
1000
|
+
"""A written type, in a file of `extension`, as signature-comparison-rules.md's Equivalent spellings commonly
|
|
1001
|
+
reads it: the set of its union's members (cut at each `|` outside every bracket pair and string), each (its
|
|
1002
|
+
base name, the keys of its type arguments, None when it writes none), so `Optional[X]`, `Union[X, None]` and
|
|
1003
|
+
`X | None` are one key, and so are `Union[A, B]`, `A | B` and `B | A`. A `typing.`, `collections.abc.` or
|
|
1004
|
+
`React.` prefix is dropped, `typing`'s `List`, `Dict`, `Tuple`, `Set`, `FrozenSet` and `Type` are the
|
|
1005
|
+
builtins they alias, a quoted forward reference is the type it quotes outside a TS/JS file (where a quoted
|
|
1006
|
+
type is a string literal type), `Array<T>` is `T[]`, `ReadonlyArray<T>` is `readonly T[]`, and a
|
|
1007
|
+
parenthesized type is the type inside. A type it reads no other way (a callable's arrow, a literal) is one
|
|
1008
|
+
member, its compacted text. Type variables renamed consistently, a type alias against the type it names and a
|
|
1009
|
+
union of callables against one callable returning the union are different keys: the script cannot see them
|
|
1010
|
+
as one."""
|
|
1011
|
+
members: set = set()
|
|
1012
|
+
for part in _split(TYPE_PREFIX_RE.sub("", _compact(text)), "|", extension):
|
|
1013
|
+
members |= _member_keys(part.strip(), extension)
|
|
1014
|
+
return frozenset(members)
|
|
1015
|
+
|
|
1016
|
+
|
|
1017
|
+
def _member_keys(text: str, extension: str = "") -> frozenset:
|
|
1018
|
+
"""The union members (_type_key) one part of a type with no `|` outside brackets stands for."""
|
|
1019
|
+
quoted = FORWARD_REFERENCE_RE.fullmatch(text) if extension not in TS_JS_EXTENSIONS else None
|
|
1020
|
+
if quoted:
|
|
1021
|
+
return _type_key(quoted.group(2), extension)
|
|
1022
|
+
if text.startswith("(") and _encloses(text, 0, extension):
|
|
1023
|
+
return _type_key(text[1:-1], extension)
|
|
1024
|
+
if text.endswith("[]") and len(text) > 2:
|
|
1025
|
+
inner = text[:-2].strip()
|
|
1026
|
+
if inner.startswith("readonly "):
|
|
1027
|
+
return frozenset({("ReadonlyArray", (_type_key(inner[len("readonly "):], extension),))})
|
|
1028
|
+
return frozenset({("Array", (_type_key(inner, extension),))})
|
|
1029
|
+
name = TYPE_NAME_RE.match(text)
|
|
1030
|
+
if name and name.end() == len(text):
|
|
1031
|
+
return frozenset({(TYPING_ALIASES.get(text, text), None)})
|
|
1032
|
+
if name and text[name.end()] in "[<" and _encloses(text, name.end(), extension):
|
|
1033
|
+
args = tuple(_type_key(arg, extension) for arg in _split(text[name.end() + 1:-1], ",", extension)
|
|
1034
|
+
if arg.strip())
|
|
1035
|
+
base = TYPING_ALIASES.get(name.group(), name.group())
|
|
1036
|
+
if base == "Optional" and len(args) == 1:
|
|
1037
|
+
return args[0] | {("None", None)}
|
|
1038
|
+
if base == "Union" and args:
|
|
1039
|
+
return frozenset().union(*args)
|
|
1040
|
+
return frozenset({(base, args)})
|
|
1041
|
+
return frozenset({(text, None)})
|
|
1042
|
+
|
|
1043
|
+
|
|
1044
|
+
def _members_match(documented: tuple, source: tuple) -> bool:
|
|
1045
|
+
"""Whether two union members (_type_key) are one type: one base name and, unless either side leaves its type
|
|
1046
|
+
arguments out (`dict` against `dict[str, Any]`, which compares the base alone), as many arguments, each pair
|
|
1047
|
+
one type (_types_match)."""
|
|
1048
|
+
(base, args), (other_base, other_args) = documented, source
|
|
1049
|
+
if base != other_base:
|
|
1050
|
+
return False
|
|
1051
|
+
if args is None or other_args is None:
|
|
1052
|
+
return True
|
|
1053
|
+
return len(args) == len(other_args) and all(_types_match(a, b) for a, b in zip(args, other_args))
|
|
1054
|
+
|
|
1055
|
+
|
|
1056
|
+
def _types_match(documented: frozenset, source: frozenset, subset: bool = False) -> bool:
|
|
1057
|
+
"""Whether every documented member is a source member (_members_match) and, unless `subset`, every source
|
|
1058
|
+
member a documented one. A parameter's own type is compared with `subset`: a documented type that names some
|
|
1059
|
+
of the members of the source's union is no difference, one that names a member the source lacks is."""
|
|
1060
|
+
if not all(any(_members_match(d, s) for s in source) for d in documented):
|
|
1061
|
+
return False
|
|
1062
|
+
return subset or all(any(_members_match(d, s) for d in documented) for s in source)
|
|
1063
|
+
|
|
1064
|
+
|
|
1065
|
+
def _literal_type(default: str | None) -> str:
|
|
1066
|
+
"""The type of a TS/JS literal default (LITERAL_TYPES: `number`, `string`, `boolean` or `bigint`), "" for any
|
|
1067
|
+
other default."""
|
|
1068
|
+
if default is None:
|
|
1069
|
+
return ""
|
|
1070
|
+
return next((name for pattern, name in LITERAL_TYPES if pattern.fullmatch(default)), "")
|
|
1071
|
+
|
|
1072
|
+
|
|
1073
|
+
def _count(n: int) -> str:
|
|
1074
|
+
return f"{n} parameter" if n == 1 else f"{n} parameters"
|
|
1075
|
+
|
|
1076
|
+
|
|
1077
|
+
def _pair_difference(documented: dict, source: dict, stated: bool, extension: str) -> str | None:
|
|
1078
|
+
"""Why a documented parameter differs from the source parameter it pairs with, None when it does not: its
|
|
1079
|
+
type, where both write one (_types_match with `subset`), or, when the documented list states optionality for
|
|
1080
|
+
any member (`stated`), its optionality, a rest or variadic member's aside."""
|
|
1081
|
+
if documented["type"] and source["type"] and not _types_match(
|
|
1082
|
+
_type_key(documented["type"], extension), _type_key(source["type"], extension), subset=True):
|
|
1083
|
+
return (f"`{documented['name']}` is typed `{documented['type']}` in the documented list, "
|
|
1084
|
+
f"`{source['type']}` in the source")
|
|
1085
|
+
if stated and not documented["variadic"] and documented["optional"] != source["optional"]:
|
|
1086
|
+
def word(optional: bool) -> str:
|
|
1087
|
+
return "optional" if optional else "required"
|
|
1088
|
+
return (f"`{documented['name']}` is {word(documented['optional'])} in the documented list, "
|
|
1089
|
+
f"{word(source['optional'])} in the source")
|
|
1090
|
+
return None
|
|
1091
|
+
|
|
1092
|
+
|
|
1093
|
+
def _complete_difference(documented: list[dict], source: list[dict], stated: bool, extension: str) -> str | None:
|
|
1094
|
+
"""Why a documented list that claims to be complete (it writes a type, a default or a `?` for some member)
|
|
1095
|
+
differs from the source's, None when it does not: it must write as many parameters, paired in order up to the
|
|
1096
|
+
first that either list writes after a `*` marker and by name from there, each pair with one name and no
|
|
1097
|
+
_pair_difference."""
|
|
1098
|
+
if len(documented) != len(source):
|
|
1099
|
+
return f"the documented list writes {_count(len(documented))}, the source {_count(len(source))}"
|
|
1100
|
+
first = min(next((i for i, p in enumerate(params) if p["by_name"]), len(params)) for params in (documented, source))
|
|
1101
|
+
for n, (d, s) in enumerate(zip(documented[:first], source[:first]), 1):
|
|
1102
|
+
if d["key"] != s["key"]:
|
|
1103
|
+
return f"parameter {n} is `{d['name']}` in the documented list, `{s['name']}` in the source"
|
|
1104
|
+
reason = _pair_difference(d, s, stated, extension)
|
|
1105
|
+
if reason:
|
|
1106
|
+
return f"parameter {n}: {reason}"
|
|
1107
|
+
named = [{p["key"]: p for p in params[first:]} for params in (documented, source)]
|
|
1108
|
+
if any(len(keys) != len(documented) - first for keys in named):
|
|
1109
|
+
return "a parameter is written twice"
|
|
1110
|
+
for key, d in named[0].items():
|
|
1111
|
+
if key not in named[1]:
|
|
1112
|
+
return f"`{d['name']}`, written after a `*` marker, is not a source parameter"
|
|
1113
|
+
reason = _pair_difference(d, named[1][key], stated, extension)
|
|
1114
|
+
if reason:
|
|
1115
|
+
return reason
|
|
1116
|
+
return None
|
|
1117
|
+
|
|
1118
|
+
|
|
1119
|
+
def _partial_difference(documented: list[dict], source: list[dict], stated: bool, bare: bool,
|
|
1120
|
+
extension: str) -> str | None:
|
|
1121
|
+
"""Why a documented list that does not claim to be complete differs from the source's, None when it does
|
|
1122
|
+
not: each member it writes must be a source parameter (`*args` and `**kwargs` by kind), in the source's order
|
|
1123
|
+
(a member after a `*` marker on either side by name alone), with no _pair_difference. A list of bare names
|
|
1124
|
+
(`bare`) must also leave out only optional source parameters (a default, a `?`, `*args`, `**kwargs`, a rest),
|
|
1125
|
+
and, in a file of a language without keyword arguments (POSITIONAL_EXTENSIONS), none before one it writes,
|
|
1126
|
+
which shifts a caller's positions. A list holding an ellipsis claims nothing about what it stands for."""
|
|
1127
|
+
used: list[int] = []
|
|
1128
|
+
last = -1
|
|
1129
|
+
for d in documented:
|
|
1130
|
+
at = next((i for i, s in enumerate(source) if i not in used and s["key"] == d["key"]), None)
|
|
1131
|
+
if at is None:
|
|
1132
|
+
return f"`{d['name']}` is not a source parameter"
|
|
1133
|
+
if not (d["by_name"] or source[at]["by_name"]):
|
|
1134
|
+
if at < last:
|
|
1135
|
+
return (f"`{d['name']}` comes before `{source[last]['name']}` in the source, after it in the "
|
|
1136
|
+
"documented list")
|
|
1137
|
+
last = at
|
|
1138
|
+
used.append(at)
|
|
1139
|
+
reason = _pair_difference(d, source[at], stated, extension)
|
|
1140
|
+
if reason:
|
|
1141
|
+
return reason
|
|
1142
|
+
if not bare:
|
|
1143
|
+
return None
|
|
1144
|
+
for i, s in enumerate(source):
|
|
1145
|
+
if i in used:
|
|
1146
|
+
continue
|
|
1147
|
+
if not (s["optional"] or s["variadic"]):
|
|
1148
|
+
return f"the required source parameter `{s['name']}` is left out"
|
|
1149
|
+
later = min((j for j in used if j > i), default=None)
|
|
1150
|
+
if extension in POSITIONAL_EXTENSIONS and later is not None:
|
|
1151
|
+
return (f"the optional `{s['name']}`, left out before `{source[later]['name']}`, shifts a caller's "
|
|
1152
|
+
f"positions in a `{extension}` file")
|
|
1153
|
+
return None
|
|
1154
|
+
|
|
1155
|
+
|
|
1156
|
+
def _parameter_difference(documented_sig: str, source_sig: str, extension: str = "") -> str | None:
|
|
1157
|
+
"""Why `documented_sig` and `source_sig` do not write the same parameters as signature-comparison-rules.md
|
|
1158
|
+
reads them in a file of `extension`, None when they do. When either has no parameter group (_parameters; a
|
|
1159
|
+
`;`-joined text has none), the texts before their returns (_signature_parts) are compared, and the reason says
|
|
1160
|
+
which side has none.
|
|
1161
|
+
Otherwise: the source's leading receiver is dropped, and the documented list's only when the source leads
|
|
1162
|
+
with one too, so a documented `(self, a)` against `def f(a)` differs; type parameters both write between the
|
|
1163
|
+
name and the group (_type_parameters) must be one text; in a TS/JS file (TS_JS_EXTENSIONS) a source parameter
|
|
1164
|
+
with no type and a literal default has that literal's type (_literal_type); a list of bare names (no type,
|
|
1165
|
+
default or `?` for any member) or one holding an ellipsis need not write every parameter
|
|
1166
|
+
(_partial_difference), and any other list claims to be complete (_complete_difference). A keyword or name
|
|
1167
|
+
before the group, a type only one side writes, a default's value, a `*` or `/` marker and the names of `*args`
|
|
1168
|
+
and `**kwargs` are never compared. The reason names the count, or the parameter (by position or by name) and
|
|
1169
|
+
the fact that differs: its name, its order, its type, its optionality, or the type parameters."""
|
|
1170
|
+
documented, source = _parameters(documented_sig, extension), _parameters(source_sig, extension)
|
|
1171
|
+
if documented is None or source is None:
|
|
1172
|
+
if _signature_parts(documented_sig, extension)[0] == _signature_parts(source_sig, extension)[0]:
|
|
1173
|
+
return None
|
|
1174
|
+
side = ("neither signature has a" if documented is None and source is None
|
|
1175
|
+
else "the documented signature has no" if documented is None else "the source signature has no")
|
|
1176
|
+
return f"{side} parameter group, so the texts before the return are compared"
|
|
1177
|
+
if source and source[0]["receiver"]:
|
|
1178
|
+
source = source[1:]
|
|
1179
|
+
if documented and documented[0]["receiver"]:
|
|
1180
|
+
documented = documented[1:]
|
|
1181
|
+
written, other = _type_parameters(documented_sig, extension), _type_parameters(source_sig, extension)
|
|
1182
|
+
if written and other and written != other:
|
|
1183
|
+
return f"the type parameters differ: documented `{written}`, source `{other}`"
|
|
1184
|
+
if extension in TS_JS_EXTENSIONS:
|
|
1185
|
+
source = [{**p, "type": _literal_type(p["default"])} if not p["type"] else p for p in source]
|
|
1186
|
+
stated = any(p["optional"] for p in documented if not p["variadic"])
|
|
1187
|
+
members = [p for p in documented if not p["ellipsis"]]
|
|
1188
|
+
bare = not any(p["type"] or p["optional"] for p in documented)
|
|
1189
|
+
if bare or len(members) < len(documented):
|
|
1190
|
+
return _partial_difference(members, source, stated, bare and len(members) == len(documented), extension)
|
|
1191
|
+
return _complete_difference(documented, source, stated, extension)
|
|
1192
|
+
|
|
1193
|
+
|
|
1194
|
+
def _returns(lines: list[str], number: int, declaration_line: int) -> bool:
|
|
1195
|
+
"""Whether line `number` (from 1) of `lines` returns: a line after the
|
|
1196
|
+
declaration whose text starts with the `return` keyword, or the
|
|
1197
|
+
declaration line itself when its own expression body follows its
|
|
1198
|
+
parameters there (an arrow function's `=> expr`, a Python
|
|
1199
|
+
`def ...: return x`). A type arrow, a callback parameter, a match arm, a
|
|
1200
|
+
string or a comment never does."""
|
|
1201
|
+
code = lines[number - 1].strip()
|
|
1202
|
+
if number != declaration_line:
|
|
1203
|
+
return bool(RETURN_START_RE.match(code))
|
|
1204
|
+
if TYPE_DECLARATION_RE.match(code):
|
|
1205
|
+
return False
|
|
1206
|
+
for opener, body in ((PYTHON_DEF_RE.match(code), PYTHON_BODY_RE), (ARROW_OPENER_RE.search(code), ARROW_BODY_RE)):
|
|
1207
|
+
end = _after_parens(code, opener.end() - 1) if opener else -1
|
|
1208
|
+
if end != -1 and body.match(code, end):
|
|
1209
|
+
return True
|
|
1210
|
+
return False
|
|
1211
|
+
|
|
1212
|
+
|
|
1213
|
+
def contradiction_fact(entry: dict, documented: dict, file: str, declaration_line: int,
|
|
1214
|
+
next_line: int | None = None, source_root: str | None = None) -> str | None:
|
|
1215
|
+
"""The first fact an annotation-contradicted mismatch rests on that does
|
|
1216
|
+
not check, None when each does: the documented entry states a return
|
|
1217
|
+
type (a non-blank `return_type`, or a `documented_sig` that writes one);
|
|
1218
|
+
`documented_sig` and `source_sig` write the same parameters, as the
|
|
1219
|
+
comparison rules read them in a file of `file`'s language
|
|
1220
|
+
(_parameter_difference, which compares the texts before the return when
|
|
1221
|
+
either has no parameter group); `evidence_line` comes
|
|
1222
|
+
after the declaration line and before `next_line`, the next surface
|
|
1223
|
+
record in the file; and, with `source_root`, that line of `file` under it
|
|
1224
|
+
returns (_returns), which alone lets the declaration line itself stand.
|
|
1225
|
+
A failed parameter fact quotes the text before each one's return
|
|
1226
|
+
(_signature_parts), then says why the two differ."""
|
|
1227
|
+
evidence = entry.get("evidence_line")
|
|
1228
|
+
extension = _extension(file)
|
|
1229
|
+
documented_params, documented_return = _signature_parts(entry["documented_sig"], extension)
|
|
1230
|
+
source_params, _ = _signature_parts(entry["source_sig"], extension)
|
|
1231
|
+
return_type = documented.get("return_type")
|
|
1232
|
+
if not (isinstance(return_type, str) and return_type.strip()) and not documented_return:
|
|
1233
|
+
return "the documented signature states no return type"
|
|
1234
|
+
difference = _parameter_difference(entry["documented_sig"], entry["source_sig"], extension)
|
|
1235
|
+
if difference is not None:
|
|
1236
|
+
return f"its parameters differ: documented `{documented_params}`, source `{source_params}`: {difference}"
|
|
1237
|
+
if not _is_line(evidence) or evidence < declaration_line:
|
|
1238
|
+
return f"evidence_line {json.dumps(evidence)} is before the declaration"
|
|
1239
|
+
if next_line is not None and evidence >= next_line:
|
|
1240
|
+
return f"evidence_line {evidence} is not before the next declaration in the file, at line {next_line}"
|
|
1241
|
+
if source_root is None:
|
|
1242
|
+
return None if evidence > declaration_line else (
|
|
1243
|
+
f"evidence_line {evidence} is the declaration line, which only --source-root can check")
|
|
1244
|
+
lines = _source_lines(source_root, file)
|
|
1245
|
+
if lines is None:
|
|
1246
|
+
return f"evidence_line {evidence} is in no readable file under --source-root"
|
|
1247
|
+
if evidence > len(lines):
|
|
1248
|
+
return f"evidence_line {evidence} is past the end of the file ({len(lines)} lines)"
|
|
1249
|
+
if not _returns(lines, evidence, declaration_line):
|
|
1250
|
+
if evidence == declaration_line:
|
|
1251
|
+
return f"evidence_line {evidence}, the declaration line, has no expression body after its parameters"
|
|
1252
|
+
return f"evidence_line {evidence} does not start with `return`"
|
|
1253
|
+
return None
|
|
1254
|
+
|
|
1255
|
+
|
|
1256
|
+
def grade(difference: str | None, file: str, documented: dict, declared: str | None = None,
|
|
1257
|
+
contradiction: str | None = UNCHECKED) -> str:
|
|
506
1258
|
"""The severity of one mismatch the result reports at `file`. The judge's
|
|
507
1259
|
`difference` lowers it only where the script confirms the fact it rests
|
|
508
1260
|
on: a parameter-name difference is Low when `declared`, the file the
|
|
509
1261
|
plan offered the check at (else `file`), is the mismatch's file and of a
|
|
510
1262
|
language with no keyword arguments; an omitted-optional difference on a
|
|
511
|
-
form the inventory marks partial is Info;
|
|
512
|
-
|
|
1263
|
+
form the inventory marks partial is Info; an annotation-contradicted
|
|
1264
|
+
difference is Info when `contradiction`, the fact contradiction_fact
|
|
1265
|
+
found failing, is None; every other mismatch is Critical."""
|
|
513
1266
|
if difference == "parameter-name":
|
|
514
1267
|
at = file if declared is None else declared
|
|
515
1268
|
if at.replace("\\", "/") == file.replace("\\", "/") and _extension(at) in POSITIONAL_EXTENSIONS:
|
|
516
1269
|
return "Low"
|
|
517
1270
|
if difference == "omitted-optional" and _partial(documented):
|
|
518
1271
|
return "Info"
|
|
1272
|
+
if difference == CONTRADICTED and contradiction is None:
|
|
1273
|
+
return "Info"
|
|
519
1274
|
return "Critical"
|
|
520
1275
|
|
|
521
1276
|
|
|
@@ -524,6 +1279,19 @@ def _gap_record(m: dict) -> dict:
|
|
|
524
1279
|
issue = f"documented `{m['documented_sig']}`, source `{m['source_sig']}`: {m['issue']}"
|
|
525
1280
|
remediation = (f"Update the documented signature of `{name}` from `{m['documented_sig']}` "
|
|
526
1281
|
f"to `{m['source_sig']}`, the signature the source defines at `{where}`.")
|
|
1282
|
+
if m["severity"] == "Info" and m["difference"] == CONTRADICTED:
|
|
1283
|
+
body = f"{m['file']}:{m['evidence_line']}"
|
|
1284
|
+
return {
|
|
1285
|
+
"severity": "Info",
|
|
1286
|
+
"category": "observation",
|
|
1287
|
+
"title": f"Annotation contradicted: {name}",
|
|
1288
|
+
"source": where,
|
|
1289
|
+
"export": name,
|
|
1290
|
+
"issue": (f"the source annotates the return of `{name}` otherwise than the skill documents, and its "
|
|
1291
|
+
f"body returns the documented type at `{body}`: {issue}"),
|
|
1292
|
+
"remediation": (f"No change to the skill: it documents what `{name}` returns, as `{body}` shows. "
|
|
1293
|
+
f"Report the return annotation at `{where}` upstream, since the body contradicts it."),
|
|
1294
|
+
}
|
|
527
1295
|
if m["severity"] == "Info":
|
|
528
1296
|
return {
|
|
529
1297
|
"severity": "Info",
|
|
@@ -550,21 +1318,72 @@ def _gap_record(m: dict) -> dict:
|
|
|
550
1318
|
}
|
|
551
1319
|
|
|
552
1320
|
|
|
1321
|
+
def _form_text(form: dict) -> str:
|
|
1322
|
+
"""A documented form on one line: its usage form, else its parameters and return type."""
|
|
1323
|
+
def text(value: object, sep: str = ", ") -> str:
|
|
1324
|
+
return sep.join(str(v) for v in value) if isinstance(value, list) else str(value)
|
|
1325
|
+
if _gives_usage(form):
|
|
1326
|
+
return text(form["usage_signature"], "; ")
|
|
1327
|
+
return_type = form.get("return_type")
|
|
1328
|
+
tail = f" -> {text(return_type)}" if isinstance(return_type, (str, list)) and text(return_type).strip() else ""
|
|
1329
|
+
return f"({text(form.get('params', ''))}){tail}"
|
|
1330
|
+
|
|
1331
|
+
|
|
1332
|
+
def _sig_text(text: str) -> str:
|
|
1333
|
+
"""A signature as a mismatch's `documented_sig` and a form compare: whitespace runs as one space, outer parentheses off."""
|
|
1334
|
+
text = " ".join(text.split())
|
|
1335
|
+
while text.startswith("(") and _after_parens(text, 0) == len(text):
|
|
1336
|
+
text = text[1:-1].strip()
|
|
1337
|
+
return text
|
|
1338
|
+
|
|
1339
|
+
|
|
1340
|
+
def _quotes(documented_sig: str, form: dict) -> bool:
|
|
1341
|
+
"""Whether `documented_sig` is the text of `form`: its parameters (a list joined), or one of its usages, whole
|
|
1342
|
+
or before the return type it writes."""
|
|
1343
|
+
usage = form.get("usage_signature")
|
|
1344
|
+
texts = {_sig_text(_field_text(form.get("params")))} | {
|
|
1345
|
+
_sig_text(u) for u in (usage if isinstance(usage, list) else [usage]) if isinstance(u, str)}
|
|
1346
|
+
texts.discard("")
|
|
1347
|
+
return bool({_sig_text(documented_sig), _sig_text(_signature_parts(documented_sig)[0])} & texts)
|
|
1348
|
+
|
|
1349
|
+
|
|
1350
|
+
def _form_gap_record(m: dict, index: int, forms: list[dict]) -> dict:
|
|
1351
|
+
"""_gap_record's record. For a name with several forms its title and Issue name the documented form the mismatch
|
|
1352
|
+
concerns, counted from 1, so the ledger keeps one gap per form (its dedupe reads the title, never the Issue)."""
|
|
1353
|
+
record = _gap_record(m)
|
|
1354
|
+
if len(forms) > 1:
|
|
1355
|
+
record["title"] += f" (form {index + 1} of {len(forms)})"
|
|
1356
|
+
record["issue"] += f" (documented form {index + 1} of {len(forms)}: `{_form_text(forms[index])}`)"
|
|
1357
|
+
return record
|
|
1358
|
+
|
|
1359
|
+
|
|
553
1360
|
def score(inventory: list[dict], names: list[str], exports: list[dict], results: list[dict],
|
|
554
|
-
type_names: list[str] | None = None, offered: dict[str, list[str]] | None = None
|
|
1361
|
+
type_names: list[str] | None = None, offered: dict[str, list[str]] | None = None,
|
|
1362
|
+
source_root: str | None = None, planned: dict[str, list[dict]] | None = None) -> dict:
|
|
555
1363
|
"""The two categories: Signature Accuracy over the signature universe of
|
|
556
1364
|
`names` (the `all` set) and `exports`, Type Coverage over `type_names`
|
|
557
1365
|
when given, else over `names`. A fallback scan's exports_found join the
|
|
558
1366
|
universe, and `names` when no `type_names` is given. `offered` (read
|
|
559
|
-
from --plan) names the files the plan offered each check at
|
|
1367
|
+
from --plan) names the files the plan offered each check at;
|
|
1368
|
+
`source_root` (--source-root) is the local tree an
|
|
1369
|
+
annotation-contradicted mismatch's `evidence_line` is checked in. A
|
|
1370
|
+
mismatch is graded against the documented form its `form` names: of
|
|
1371
|
+
documented_forms, or of `planned` (read from --plan) for a name it
|
|
1372
|
+
holds, with a warning where the two differ."""
|
|
560
1373
|
by_name = _records_by_name(exports)
|
|
561
1374
|
# A fallback scan reports its own names and types.
|
|
562
1375
|
fallback_types = {n for r in results for n in r.get("types_found") or []}
|
|
563
1376
|
surface = set(names) | {n for r in results for n in r.get("exports_found") or []}
|
|
564
1377
|
documented = documented_names(inventory)
|
|
565
|
-
|
|
566
|
-
compared = (surface | signature_universe(names, exports)) & set(
|
|
1378
|
+
forms = documented_forms(inventory)
|
|
1379
|
+
compared = (surface | signature_universe(names, exports)) & set(forms)
|
|
567
1380
|
warnings = []
|
|
1381
|
+
for name in sorted(set(forms) & set(planned or {})):
|
|
1382
|
+
if planned[name] != forms[name]:
|
|
1383
|
+
differ = (f"{len(planned[name])} documented form(s), not the {len(forms[name])} the inventory gives"
|
|
1384
|
+
if len(planned[name]) != len(forms[name]) else "different forms than the inventory gives")
|
|
1385
|
+
warnings.append(f"`{name}`: the plan offered {differ}: graded against the plan's")
|
|
1386
|
+
forms[name] = planned[name]
|
|
568
1387
|
|
|
569
1388
|
graded: dict[tuple, dict] = {}
|
|
570
1389
|
seen = set()
|
|
@@ -582,32 +1401,71 @@ def score(inventory: list[dict], names: list[str], exports: list[dict], results:
|
|
|
582
1401
|
f"`{entry['file']}`, no file the plan offered its check at: recorded at "
|
|
583
1402
|
"the result's file")
|
|
584
1403
|
file = _path(file)
|
|
585
|
-
key = (name, file, entry["line"])
|
|
586
1404
|
if name not in compared:
|
|
587
|
-
if
|
|
1405
|
+
if (name, file, entry["line"]) not in seen:
|
|
588
1406
|
warnings.append(f"{file}:{entry['line']}: a signature mismatch for "
|
|
589
1407
|
f"`{name}`, which is no documented signature on the source surface: "
|
|
590
1408
|
"not scored, no gap")
|
|
591
|
-
seen.add(
|
|
1409
|
+
seen.add((name, file, entry["line"]))
|
|
592
1410
|
continue
|
|
1411
|
+
# The documented form the mismatch is reported for: 0 is `documented`, n the check's nth alternative.
|
|
1412
|
+
index = entry.get("form") or 0
|
|
1413
|
+
if index >= len(forms[name]):
|
|
1414
|
+
warning = (f"{file}:{entry['line']}: a signature mismatch for `{name}` names `form` {index}, no "
|
|
1415
|
+
f"form its check offered (0 to {len(forms[name]) - 1}): graded against `documented`")
|
|
1416
|
+
if warning not in warnings:
|
|
1417
|
+
warnings.append(warning)
|
|
1418
|
+
index = 0
|
|
1419
|
+
# A form's own label lowers a grade only for the form the mismatch quotes as `documented_sig`.
|
|
1420
|
+
if len(forms[name]) > 1 and not _quotes(entry["documented_sig"], forms[name][index]):
|
|
1421
|
+
quoted = next((j for j, f in enumerate(forms[name]) if _quotes(entry["documented_sig"], f)), None)
|
|
1422
|
+
if quoted is not None or index != 0:
|
|
1423
|
+
target = "documented" if quoted is None else f"form {quoted}, the one it quotes"
|
|
1424
|
+
warning = (f"{file}:{entry['line']}: a signature mismatch for `{name}` names `form` {index}, whose "
|
|
1425
|
+
f"text its `documented_sig` is not: graded against {target}")
|
|
1426
|
+
if warning not in warnings:
|
|
1427
|
+
warnings.append(warning)
|
|
1428
|
+
index = quoted or 0
|
|
1429
|
+
form = forms[name][index]
|
|
1430
|
+
# Each usage of a list is a form of its own: one key per documented text.
|
|
1431
|
+
key = (name, file, entry["line"], index, _sig_text(entry["documented_sig"]))
|
|
593
1432
|
difference = entry.get("difference")
|
|
594
1433
|
# Graded at the file only when the plan offered it: never Low at a namesake it did not offer.
|
|
595
1434
|
declared = (file if _is_offered_file(files, file) else files[0]) if files else None
|
|
596
|
-
|
|
597
|
-
|
|
1435
|
+
# The facts of an annotation-contradicted mismatch, checked once and handed to grade().
|
|
1436
|
+
fact = None
|
|
1437
|
+
if difference == CONTRADICTED:
|
|
1438
|
+
at_line = _declaration_line(by_name, name, file, entry["line"])
|
|
1439
|
+
fact = contradiction_fact(entry, form, file, at_line,
|
|
1440
|
+
_next_declaration(exports, file, at_line), source_root)
|
|
1441
|
+
warning = (f"{file}:{at_line}: the mismatch for `{name}` is tagged {CONTRADICTED}, but {fact}: "
|
|
1442
|
+
"graded as an untagged mismatch (Critical)")
|
|
1443
|
+
if fact is not None and warning not in warnings:
|
|
1444
|
+
warnings.append(warning)
|
|
1445
|
+
severity = grade(difference, file, form, declared, fact)
|
|
1446
|
+
# Every entry is graded: one declaration reported twice against one form keeps its most severe grade.
|
|
598
1447
|
if key in graded and GRADES.index(graded[key]["severity"]) <= GRADES.index(severity):
|
|
599
1448
|
continue
|
|
600
1449
|
graded[key] = {"name": name, "file": file, "line": entry["line"],
|
|
601
1450
|
"source_sig": entry["source_sig"], "documented_sig": entry["documented_sig"],
|
|
602
1451
|
"issue": entry["issue"], "difference": difference, "severity": severity}
|
|
603
|
-
|
|
604
|
-
|
|
605
|
-
|
|
1452
|
+
if difference == CONTRADICTED:
|
|
1453
|
+
graded[key]["evidence_line"] = entry.get("evidence_line")
|
|
1454
|
+
# A check that offered several files keeps one mismatch per form, its least severe (the closest candidate).
|
|
1455
|
+
# Each form's mismatch stands on its own.
|
|
1456
|
+
def order(item: tuple) -> tuple:
|
|
1457
|
+
(_, _, _, index, text), m = item
|
|
1458
|
+
return m["file"], m["line"], m["name"], index, text
|
|
1459
|
+
|
|
1460
|
+
kept, closest = [], {}
|
|
1461
|
+
for key, m in sorted(graded.items(), key=order):
|
|
1462
|
+
bucket = (m["name"], key[3], key[4])
|
|
606
1463
|
if len({_path(f) for f in _offered_files(offered, by_name, m["name"])}) < 2:
|
|
607
|
-
|
|
608
|
-
elif
|
|
609
|
-
closest[
|
|
610
|
-
|
|
1464
|
+
kept.append((key, m))
|
|
1465
|
+
elif bucket not in closest or GRADES.index(m["severity"]) > GRADES.index(closest[bucket][1]["severity"]):
|
|
1466
|
+
closest[bucket] = (key, m)
|
|
1467
|
+
kept = sorted(kept + list(closest.values()), key=order)
|
|
1468
|
+
mismatches = [m for _, m in kept]
|
|
611
1469
|
|
|
612
1470
|
# A Low or an Info mismatch is a caller-invisible difference: it still matches.
|
|
613
1471
|
wrong = {m["name"] for m in mismatches if m["severity"] == "Critical"}
|
|
@@ -643,7 +1501,7 @@ def score(inventory: list[dict], names: list[str], exports: list[dict], results:
|
|
|
643
1501
|
"totalTypes": len(types),
|
|
644
1502
|
"missingTypes": sorted(types - documented),
|
|
645
1503
|
"mismatches": mismatches,
|
|
646
|
-
"gapRecords": [
|
|
1504
|
+
"gapRecords": [_form_gap_record(m, key[3], forms[m["name"]]) for key, m in kept],
|
|
647
1505
|
"warnings": warnings,
|
|
648
1506
|
}
|
|
649
1507
|
|
|
@@ -675,6 +1533,8 @@ def _cmd_plan(args: argparse.Namespace) -> tuple[dict, int]:
|
|
|
675
1533
|
|
|
676
1534
|
|
|
677
1535
|
def _cmd_score(args: argparse.Namespace) -> tuple[dict, int]:
|
|
1536
|
+
if args.source_root is not None and not (args.source_root.strip() and Path(args.source_root).is_dir()):
|
|
1537
|
+
raise InputError(f"--source-root {args.source_root!r} is no folder: pass the local source tree the judges read")
|
|
678
1538
|
inventory = load_inventory(args.inventory)
|
|
679
1539
|
chosen, exports, names = load_surface(args.surface, args.surface_set)
|
|
680
1540
|
results, violations = load_results(args.results or [])
|
|
@@ -684,7 +1544,9 @@ def _cmd_score(args: argparse.Namespace) -> tuple[dict, int]:
|
|
|
684
1544
|
Path(path).unlink(missing_ok=True)
|
|
685
1545
|
return {"valid": False, "violations": violations}, 2
|
|
686
1546
|
offered = load_plan(args.plan) if args.plan is not None else None
|
|
687
|
-
|
|
1547
|
+
planned = load_plan_forms(args.plan) if args.plan is not None else None
|
|
1548
|
+
out = score(inventory, names, exports, results, None if args.surface_set == "all" else chosen, offered,
|
|
1549
|
+
args.source_root, planned)
|
|
688
1550
|
if args.output:
|
|
689
1551
|
_write_json(args.output, out)
|
|
690
1552
|
if args.gaps_output:
|
|
@@ -716,6 +1578,9 @@ def _build_parser() -> argparse.ArgumentParser:
|
|
|
716
1578
|
p.add_argument("--surface-set", default="all",
|
|
717
1579
|
help="the surface name set Type Coverage counts (default: all)")
|
|
718
1580
|
p.add_argument("--plan", help="the plan's output: the files each check offered a mismatch at")
|
|
1581
|
+
p.add_argument("--source-root", help="the local source tree the judges read: an annotation-contradicted "
|
|
1582
|
+
"mismatch's evidence_line must start with `return` there, or be the "
|
|
1583
|
+
"declaration line with its expression body")
|
|
719
1584
|
p.add_argument("--results", action="append", help="a subagent result file (repeatable)")
|
|
720
1585
|
p.add_argument("--output", help="also write the scores to this file")
|
|
721
1586
|
p.add_argument("--gaps-output", help="write the gap ledger records to this file")
|