@schmock/faker 1.9.4 → 1.10.0

package/dist/field-name-matcher.d.ts

@@ -0,0 +1,31 @@
+import type { JSONSchema7 } from "json-schema";
+import type { FieldMapping } from "./field-mappings.js";
+/**
+ * Split a field name (camelCase, snake_case, kebab-case) into lowercase tokens.
+ * "userFirstName" → ["user", "first", "name"]
+ * "created_at"    → ["created", "at"]
+ * "HTMLParser"    → ["html", "parser"]
+ * "is_active"     → ["is", "active"]
+ */
+export declare function tokenizeFieldName(name: string): string[];
+/**
+ * Score how well a set of keyword tokens matches field name tokens.
+ * Returns 0-1:
+ *   1.0 — exact full match (joined tokens equal joined keywords)
+ *   0.9 — all keyword tokens found in field tokens with high coverage (>50%)
+ *   0.8 — field ends with keyword tokens
+ *   0.7 — substring match (keyword appears in joined field name)
+ *   0.65 — all keyword tokens found but low coverage (<=50%)
+ */
+export declare function scoreMatch(fieldTokens: string[], keywords: string[]): number;
+interface MatchResult {
+    mapping: FieldMapping;
+    score: number;
+}
+/**
+ * Find the best field mapping for a given field name and schema.
+ * Returns the highest-scoring match above its threshold, or undefined.
+ */
+export declare function findBestMapping(fieldName: string, schema: JSONSchema7, mappings?: FieldMapping[]): MatchResult | undefined;
+export {};
+//# sourceMappingURL=field-name-matcher.d.ts.map

package/dist/field-name-matcher.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"field-name-matcher.d.ts","sourceRoot":"","sources":["../src/field-name-matcher.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAC/C,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AAGxD;;;;;;GAMG;AACH,wBAAgB,iBAAiB,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,CAmBxD;AAED;;;;;;;;GAQG;AACH,wBAAgB,UAAU,CAAC,WAAW,EAAE,MAAM,EAAE,EAAE,QAAQ,EAAE,MAAM,EAAE,GAAG,MAAM,CAsC5E;AAED,UAAU,WAAW;IACnB,OAAO,EAAE,YAAY,CAAC;IACtB,KAAK,EAAE,MAAM,CAAC;CACf;AAED;;;GAGG;AACH,wBAAgB,eAAe,CAC7B,SAAS,EAAE,MAAM,EACjB,MAAM,EAAE,WAAW,EACnB,QAAQ,GAAE,YAAY,EAAuB,GAC5C,WAAW,GAAG,SAAS,CA6EzB"}

package/dist/field-name-matcher.js

@@ -0,0 +1,136 @@
+import { ALL_FIELD_MAPPINGS } from "./field-mappings.js";
+/**
+ * Split a field name (camelCase, snake_case, kebab-case) into lowercase tokens.
+ * "userFirstName" → ["user", "first", "name"]
+ * "created_at"    → ["created", "at"]
+ * "HTMLParser"    → ["html", "parser"]
+ * "is_active"     → ["is", "active"]
+ */
+export function tokenizeFieldName(name) {
+    // Split on _ and -
+    const parts = name.split(/[_-]/);
+    const tokens = [];
+    for (const part of parts) {
+        if (!part)
+            continue;
+        // Split camelCase and consecutive uppercase (e.g., HTMLParser → HTML, Parser)
+        const camelTokens = part
+            .replace(/([A-Z]+)([A-Z][a-z])/g, "$1_$2")
+            .replace(/([a-z0-9])([A-Z])/g, "$1_$2")
+            .split("_");
+        for (const t of camelTokens) {
+            if (t)
+                tokens.push(t.toLowerCase());
+        }
+    }
+    return tokens;
+}
+/**
+ * Score how well a set of keyword tokens matches field name tokens.
+ * Returns 0-1:
+ *   1.0 — exact full match (joined tokens equal joined keywords)
+ *   0.9 — all keyword tokens found in field tokens with high coverage (>50%)
+ *   0.8 — field ends with keyword tokens
+ *   0.7 — substring match (keyword appears in joined field name)
+ *   0.65 — all keyword tokens found but low coverage (<=50%)
+ */
+export function scoreMatch(fieldTokens, keywords) {
+    const fieldJoined = fieldTokens.join("");
+    let bestScore = 0;
+    // Try each keyword variant, keep the highest score
+    for (const keyword of keywords) {
+        const kwTokens = tokenizeFieldName(keyword);
+        const kwJoined = kwTokens.join("");
+        // Exact match: joined tokens are identical
+        if (fieldJoined === kwJoined)
+            return 1.0;
+        // All keyword tokens present in field tokens with high coverage
+        if (kwTokens.length > 0 &&
+            kwTokens.every((kt) => fieldTokens.includes(kt))) {
+            const coverage = kwTokens.length / fieldTokens.length;
+            const score = coverage > 0.5 ? 0.9 : 0.65;
+            bestScore = Math.max(bestScore, score);
+        }
+        // Field ends with keyword tokens
+        if (kwTokens.length > 0 && kwTokens.length <= fieldTokens.length) {
+            const tail = fieldTokens.slice(-kwTokens.length);
+            if (tail.every((t, i) => t === kwTokens[i])) {
+                bestScore = Math.max(bestScore, 0.8);
+            }
+        }
+        // Substring match: keyword joined appears in field joined
+        if (kwJoined.length >= 3 && fieldJoined.includes(kwJoined)) {
+            bestScore = Math.max(bestScore, 0.7);
+        }
+    }
+    return bestScore;
+}
+/**
+ * Find the best field mapping for a given field name and schema.
+ * Returns the highest-scoring match above its threshold, or undefined.
+ */
+export function findBestMapping(fieldName, schema, mappings = ALL_FIELD_MAPPINGS) {
+    const schemaAny = schema;
+    const schemaType = typeof schema.type === "string" ? schema.type : undefined;
+    // Priority: format:uuid always maps to string.uuid
+    if (schemaType === "string" && schema.format === "uuid") {
+        return {
+            mapping: {
+                keywords: ["uuid"],
+                fakerMethod: "string.uuid",
+                schemaType: "string",
+                minScore: 0.5,
+            },
+            score: 1.0,
+        };
+    }
+    // Skip if schema already has pattern, enum, or faker constraint
+    if (schemaAny.pattern || schemaAny.enum || schemaAny.faker) {
+        return undefined;
+    }
+    // Skip numeric faker mapping when schema already constrains the range
+    const hasNumericConstraints = schema.minimum !== undefined ||
+        schema.maximum !== undefined ||
+        schema.exclusiveMinimum !== undefined ||
+        schema.exclusiveMaximum !== undefined;
+    const tokens = tokenizeFieldName(fieldName);
+    let best;
+    for (const mapping of mappings) {
+        // Skip numeric faker mappings when schema has explicit constraints
+        if (hasNumericConstraints &&
+            (mapping.schemaType === "number" || mapping.schemaType === "integer")) {
+            continue;
+        }
+        // Check schema type constraint
+        if (mapping.schemaType && schemaType && mapping.schemaType !== schemaType) {
+            const isNumeric = (mapping.schemaType === "number" && schemaType === "integer") ||
+                (mapping.schemaType === "integer" && schemaType === "number");
+            if (!isNumeric)
+                continue;
+        }
+        const score = scoreMatch(tokens, mapping.keywords);
+        if (score >= mapping.minScore && (!best || score > best.score)) {
+            best = { mapping, score };
+        }
+    }
+    // ID suffix detection: fields ending in Id/_id with string type → UUID
+    // Only if no better match was found or the existing match is weak
+    if (schemaType === "string") {
+        const lastToken = tokens[tokens.length - 1];
+        if (lastToken === "id" && tokens.length > 1) {
+            const idScore = 0.8;
+            if (!best || best.score < idScore) {
+                best = {
+                    mapping: {
+                        keywords: ["id"],
+                        fakerMethod: "string.uuid",
+                        schemaType: "string",
+                        minScore: 0.5,
+                    },
+                    score: idScore,
+                };
+            }
+        }
+    }
+    return best;
+}

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAOA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAO/C,MAAM,WAAW,uBAAuB;IACtC,MAAM,EAAE,WAAW,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;IAChC,MAAM,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAChC,KAAK,CAAC,EAAE,GAAG,CAAC;IACZ,KAAK,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC/B,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,kBAAkB;IACjC,MAAM,EAAE,WAAW,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;IAChC,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,wBAAgB,WAAW,CAAC,OAAO,EAAE,kBAAkB,GAAG,OAAO,CAAC,MAAM,CA+CvE;AAED,wBAAgB,kBAAkB,CAAC,OAAO,EAAE,uBAAuB,GAAG,GAAG,CA6CxE"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAOA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAO/C,MAAM,WAAW,uBAAuB;IACtC,MAAM,EAAE,WAAW,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;IAChC,MAAM,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAChC,KAAK,CAAC,EAAE,GAAG,CAAC;IACZ,KAAK,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC/B,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,kBAAkB;IACjC,MAAM,EAAE,WAAW,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;IAChC,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,wBAAgB,WAAW,CAAC,OAAO,EAAE,kBAAkB,GAAG,OAAO,CAAC,MAAM,CA+CvE;AAED,wBAAgB,kBAAkB,CAAC,OAAO,EAAE,uBAAuB,GAAG,GAAG,CA8CxE"}