@ansvar/eu-regulations-mcp 0.2.3 → 0.3.1
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/README.md +101 -7
- package/data/regulations.db +0 -0
- package/data/seed/gdpr.json +699 -1
- package/dist/index.js +32 -0
- package/dist/index.js.map +1 -1
- package/dist/tools/recital.d.ts +13 -0
- package/dist/tools/recital.d.ts.map +1 -0
- package/dist/tools/recital.js +23 -0
- package/dist/tools/recital.js.map +1 -0
- package/dist/tools/search.d.ts +1 -0
- package/dist/tools/search.d.ts.map +1 -1
- package/dist/tools/search.js +55 -18
- package/dist/tools/search.js.map +1 -1
- package/package.json +1 -1
- package/scripts/build-db.ts +62 -0
- package/scripts/ingest-eurlex.ts +87 -0
- package/scripts/reingest-all-with-recitals.sh +81 -0
- package/src/index.ts +33 -0
- package/src/tools/recital.ts +48 -0
- package/src/tools/search.ts +69 -21
package/src/tools/search.ts
CHANGED
|
@@ -12,6 +12,7 @@ export interface SearchResult {
|
|
|
12
12
|
title: string;
|
|
13
13
|
snippet: string;
|
|
14
14
|
relevance: number;
|
|
15
|
+
type?: 'article' | 'recital';
|
|
15
16
|
}
|
|
16
17
|
|
|
17
18
|
/**
|
|
@@ -45,46 +46,93 @@ export async function searchRegulations(
|
|
|
45
46
|
return [];
|
|
46
47
|
}
|
|
47
48
|
|
|
48
|
-
|
|
49
|
-
|
|
49
|
+
const params: (string | number)[] = [escapedQuery];
|
|
50
|
+
|
|
51
|
+
// Build optional regulation filter
|
|
52
|
+
let regulationFilter = '';
|
|
53
|
+
if (regulations && regulations.length > 0) {
|
|
54
|
+
const placeholders = regulations.map(() => '?').join(', ');
|
|
55
|
+
regulationFilter = ` AND regulation IN (${placeholders})`;
|
|
56
|
+
params.push(...regulations);
|
|
57
|
+
}
|
|
58
|
+
|
|
59
|
+
// Search in articles
|
|
60
|
+
const articlesQuery = `
|
|
50
61
|
SELECT
|
|
51
62
|
articles_fts.regulation,
|
|
52
63
|
articles_fts.article_number as article,
|
|
53
64
|
articles_fts.title,
|
|
54
65
|
snippet(articles_fts, 3, '>>>', '<<<', '...', 32) as snippet,
|
|
55
|
-
bm25(articles_fts) as relevance
|
|
66
|
+
bm25(articles_fts) as relevance,
|
|
67
|
+
'article' as type
|
|
56
68
|
FROM articles_fts
|
|
57
69
|
WHERE articles_fts MATCH ?
|
|
70
|
+
${regulationFilter}
|
|
71
|
+
ORDER BY bm25(articles_fts)
|
|
72
|
+
LIMIT ?
|
|
58
73
|
`;
|
|
59
74
|
|
|
60
|
-
|
|
75
|
+
// Search in recitals
|
|
76
|
+
const recitalsQuery = `
|
|
77
|
+
SELECT
|
|
78
|
+
recitals_fts.regulation,
|
|
79
|
+
CAST(recitals_fts.recital_number AS TEXT) as article,
|
|
80
|
+
'Recital ' || recitals_fts.recital_number as title,
|
|
81
|
+
snippet(recitals_fts, 2, '>>>', '<<<', '...', 32) as snippet,
|
|
82
|
+
bm25(recitals_fts) as relevance,
|
|
83
|
+
'recital' as type
|
|
84
|
+
FROM recitals_fts
|
|
85
|
+
WHERE recitals_fts MATCH ?
|
|
86
|
+
${regulationFilter}
|
|
87
|
+
ORDER BY bm25(recitals_fts)
|
|
88
|
+
LIMIT ?
|
|
89
|
+
`;
|
|
61
90
|
|
|
62
|
-
|
|
63
|
-
|
|
64
|
-
|
|
65
|
-
params
|
|
66
|
-
}
|
|
91
|
+
try {
|
|
92
|
+
// Execute both queries
|
|
93
|
+
const articlesParams = [...params, limit];
|
|
94
|
+
const recitalsParams = [...params, limit];
|
|
67
95
|
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
sql += ` LIMIT ?`;
|
|
71
|
-
params.push(limit);
|
|
96
|
+
const articleStmt = db.prepare(articlesQuery);
|
|
97
|
+
const recitalStmt = db.prepare(recitalsQuery);
|
|
72
98
|
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
99
|
+
const articleRows = articleStmt.all(...articlesParams) as Array<{
|
|
100
|
+
regulation: string;
|
|
101
|
+
article: string;
|
|
102
|
+
title: string;
|
|
103
|
+
snippet: string;
|
|
104
|
+
relevance: number;
|
|
105
|
+
type: 'article' | 'recital';
|
|
106
|
+
}>;
|
|
107
|
+
|
|
108
|
+
const recitalRows = recitalStmt.all(...recitalsParams) as Array<{
|
|
76
109
|
regulation: string;
|
|
77
110
|
article: string;
|
|
78
111
|
title: string;
|
|
79
112
|
snippet: string;
|
|
80
113
|
relevance: number;
|
|
114
|
+
type: 'article' | 'recital';
|
|
81
115
|
}>;
|
|
82
116
|
|
|
83
|
-
//
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
87
|
-
|
|
117
|
+
// Combine and sort by relevance, prioritizing articles
|
|
118
|
+
const combined = [...articleRows, ...recitalRows]
|
|
119
|
+
.map(row => ({
|
|
120
|
+
...row,
|
|
121
|
+
relevance: Math.abs(row.relevance),
|
|
122
|
+
}))
|
|
123
|
+
.sort((a, b) => {
|
|
124
|
+
// First sort by relevance
|
|
125
|
+
if (Math.abs(a.relevance - b.relevance) > 0.01) {
|
|
126
|
+
return b.relevance - a.relevance;
|
|
127
|
+
}
|
|
128
|
+
// If relevance is similar, prioritize articles over recitals
|
|
129
|
+
if (a.type === 'article' && b.type === 'recital') return -1;
|
|
130
|
+
if (a.type === 'recital' && b.type === 'article') return 1;
|
|
131
|
+
return 0;
|
|
132
|
+
})
|
|
133
|
+
.slice(0, limit);
|
|
134
|
+
|
|
135
|
+
return combined;
|
|
88
136
|
} catch (error) {
|
|
89
137
|
// If FTS5 query fails (e.g., syntax error), return empty results
|
|
90
138
|
if (error instanceof Error && error.message.includes('fts5')) {
|