@pseolint/core 0.1.0

package/dist/rules/tech/robots-sitemap-presence.js

@@ -0,0 +1,61 @@
+async function fetchTextIfOk(url) {
+    try {
+        const response = await fetch(url);
+        if (!response.ok) {
+            return null;
+        }
+        return await response.text();
+    }
+    catch {
+        return null;
+    }
+}
+export async function robotsSitemapPresenceRule(source) {
+    if (!/^https?:\/\//i.test(source)) {
+        return [];
+    }
+    const sourceUrl = new URL(source);
+    const origin = sourceUrl.origin;
+    const robotsUrl = `${origin}/robots.txt`;
+    const sitemapUrl = `${origin}/sitemap.xml`;
+    const findings = [];
+    const robotsText = await fetchTextIfOk(robotsUrl);
+    if (!robotsText) {
+        findings.push({
+            ruleId: "tech/robots-sitemap-presence",
+            severity: "warning",
+            message: `Could not fetch ${robotsUrl}.`,
+            fix: `Create a robots.txt file at ${robotsUrl} and include a Sitemap directive pointing to your sitemap.`
+        });
+        return findings;
+    }
+    if (!/^\s*sitemap\s*:/gim.test(robotsText)) {
+        findings.push({
+            ruleId: "tech/robots-sitemap-presence",
+            severity: "info",
+            message: `${robotsUrl} does not declare a Sitemap directive.`,
+            fix: `Add a Sitemap directive to ${robotsUrl}, e.g.: Sitemap: ${sitemapUrl}`
+        });
+    }
+    const sitemapText = await fetchTextIfOk(sitemapUrl);
+    if (!sitemapText) {
+        findings.push({
+            ruleId: "tech/robots-sitemap-presence",
+            severity: "warning",
+            message: `Could not fetch ${sitemapUrl}.`,
+            fix: `Create an XML sitemap at ${sitemapUrl} and reference it in ${robotsUrl}.`
+        });
+        return findings;
+    }
+    const lowered = sitemapText.toLowerCase();
+    if (!lowered.includes("<urlset") && !lowered.includes("<sitemapindex")) {
+        findings.push({
+            ruleId: "tech/robots-sitemap-presence",
+            severity: "warning",
+            message: `${sitemapUrl} was fetched but does not look like sitemap XML.`,
+            fix: `Ensure ${sitemapUrl} is valid sitemap XML containing a <urlset> or <sitemapindex> root element.`
+        });
+    }
+    return findings;
+}
+//# sourceMappingURL=robots-sitemap-presence.js.map

package/dist/rules/tech/robots-sitemap-presence.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"robots-sitemap-presence.js","sourceRoot":"","sources":["../../../src/rules/tech/robots-sitemap-presence.ts"],"names":[],"mappings":"AAEA,KAAK,UAAU,aAAa,CAAC,GAAW;IACtC,IAAI,CAAC;QACH,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,GAAG,CAAC,CAAC;QAClC,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;YACjB,OAAO,IAAI,CAAC;QACd,CAAC;QACD,OAAO,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;IAC/B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,yBAAyB,CAAC,MAAc;IAC5D,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC;QAClC,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,CAAC;IAClC,MAAM,MAAM,GAAG,SAAS,CAAC,MAAM,CAAC;IAChC,MAAM,SAAS,GAAG,GAAG,MAAM,aAAa,CAAC;IACzC,MAAM,UAAU,GAAG,GAAG,MAAM,cAAc,CAAC;IAC3C,MAAM,QAAQ,GAAiB,EAAE,CAAC;IAElC,MAAM,UAAU,GAAG,MAAM,aAAa,CAAC,SAAS,CAAC,CAAC;IAClD,IAAI,CAAC,UAAU,EAAE,CAAC;QAChB,QAAQ,CAAC,IAAI,CAAC;YACZ,MAAM,EAAE,8BAA8B;YACtC,QAAQ,EAAE,SAAS;YACnB,OAAO,EAAE,mBAAmB,SAAS,GAAG;YACxC,GAAG,EAAE,+BAA+B,SAAS,4DAA4D;SAC1G,CAAC,CAAC;QACH,OAAO,QAAQ,CAAC;IAClB,CAAC;IAED,IAAI,CAAC,oBAAoB,CAAC,IAAI,CAAC,UAAU,CAAC,EAAE,CAAC;QAC3C,QAAQ,CAAC,IAAI,CAAC;YACZ,MAAM,EAAE,8BAA8B;YACtC,QAAQ,EAAE,MAAM;YAChB,OAAO,EAAE,GAAG,SAAS,wCAAwC;YAC7D,GAAG,EAAE,8BAA8B,SAAS,oBAAoB,UAAU,EAAE;SAC7E,CAAC,CAAC;IACL,CAAC;IAED,MAAM,WAAW,GAAG,MAAM,aAAa,CAAC,UAAU,CAAC,CAAC;IACpD,IAAI,CAAC,WAAW,EAAE,CAAC;QACjB,QAAQ,CAAC,IAAI,CAAC;YACZ,MAAM,EAAE,8BAA8B;YACtC,QAAQ,EAAE,SAAS;YACnB,OAAO,EAAE,mBAAmB,UAAU,GAAG;YACzC,GAAG,EAAE,4BAA4B,UAAU,wBAAwB,SAAS,GAAG;SAChF,CAAC,CAAC;QACH,OAAO,QAAQ,CAAC;IAClB,CAAC;IAED,MAAM,OAAO,GAAG,WAAW,CAAC,WAAW,EAAE,CAAC;IAC1C,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,eAAe,CAAC,EAAE,CAAC;QACvE,QAAQ,CAAC,IAAI,CAAC;YACZ,MAAM,EAAE,8BAA8B;YACtC,QAAQ,EAAE,SAAS;YACnB,OAAO,EAAE,GAAG,UAAU,kDAAkD;YACxE,GAAG,EAAE,UAAU,UAAU,6EAA6E;SACvG,CAAC,CAAC;IACL,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/rules/tech/sitemap-completeness.d.ts

@@ -0,0 +1,3 @@
+import type { ParsedPage, RuleResult } from "../../types.js";
+export declare function sitemapCompletenessRule(pages: ParsedPage[], sitemapUrls: Set<string>): RuleResult[];
+//# sourceMappingURL=sitemap-completeness.d.ts.map

package/dist/rules/tech/sitemap-completeness.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"sitemap-completeness.d.ts","sourceRoot":"","sources":["../../../src/rules/tech/sitemap-completeness.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AAE7D,wBAAgB,uBAAuB,CACrC,KAAK,EAAE,UAAU,EAAE,EACnB,WAAW,EAAE,GAAG,CAAC,MAAM,CAAC,GACvB,UAAU,EAAE,CA0Cd"}

package/dist/rules/tech/sitemap-completeness.js

@@ -0,0 +1,40 @@
+export function sitemapCompletenessRule(pages, sitemapUrls) {
+    if (sitemapUrls.size === 0)
+        return [];
+    const findings = [];
+    const missingFromSitemap = pages.filter((p) => !sitemapUrls.has(p.url));
+    if (missingFromSitemap.length > 0) {
+        findings.push({
+            ruleId: "tech/sitemap-completeness",
+            severity: "error",
+            message: `${missingFromSitemap.length} crawlable page(s) not in sitemap.`,
+            fix: "Add these pages to your sitemap.xml to ensure Google discovers them.",
+            relatedUrls: missingFromSitemap.map((p) => p.url).sort()
+        });
+    }
+    for (const page of pages) {
+        if (!page.httpMeta || !sitemapUrls.has(page.url))
+            continue;
+        if (page.httpMeta.statusCode >= 400) {
+            findings.push({
+                ruleId: "tech/sitemap-completeness",
+                severity: "error",
+                message: `Sitemap URL ${page.url} returns HTTP ${page.httpMeta.statusCode}.`,
+                pageUrl: page.url,
+                fix: "Remove this URL from sitemap.xml or fix the page to return HTTP 200."
+            });
+        }
+        if (page.httpMeta.redirectChain.length > 0 && page.httpMeta.finalUrl !== page.url) {
+            findings.push({
+                ruleId: "tech/sitemap-completeness",
+                severity: "warning",
+                message: `Sitemap URL ${page.url} redirects to ${page.httpMeta.finalUrl}.`,
+                pageUrl: page.url,
+                relatedUrls: [page.httpMeta.finalUrl],
+                fix: `Update sitemap.xml to use the final URL: ${page.httpMeta.finalUrl}`
+            });
+        }
+    }
+    return findings;
+}
+//# sourceMappingURL=sitemap-completeness.js.map

package/dist/rules/tech/sitemap-completeness.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"sitemap-completeness.js","sourceRoot":"","sources":["../../../src/rules/tech/sitemap-completeness.ts"],"names":[],"mappings":"AAEA,MAAM,UAAU,uBAAuB,CACrC,KAAmB,EACnB,WAAwB;IAExB,IAAI,WAAW,CAAC,IAAI,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEtC,MAAM,QAAQ,GAAiB,EAAE,CAAC;IAElC,MAAM,kBAAkB,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IACxE,IAAI,kBAAkB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAClC,QAAQ,CAAC,IAAI,CAAC;YACZ,MAAM,EAAE,2BAA2B;YACnC,QAAQ,EAAE,OAAO;YACjB,OAAO,EAAE,GAAG,kBAAkB,CAAC,MAAM,oCAAoC;YACzE,GAAG,EAAE,sEAAsE;YAC3E,WAAW,EAAE,kBAAkB,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE;SACzD,CAAC,CAAC;IACL,CAAC;IAED,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,IAAI,CAAC,IAAI,CAAC,QAAQ,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC;YAAE,SAAS;QAE3D,IAAI,IAAI,CAAC,QAAQ,CAAC,UAAU,IAAI,GAAG,EAAE,CAAC;YACpC,QAAQ,CAAC,IAAI,CAAC;gBACZ,MAAM,EAAE,2BAA2B;gBACnC,QAAQ,EAAE,OAAO;gBACjB,OAAO,EAAE,eAAe,IAAI,CAAC,GAAG,iBAAiB,IAAI,CAAC,QAAQ,CAAC,UAAU,GAAG;gBAC5E,OAAO,EAAE,IAAI,CAAC,GAAG;gBACjB,GAAG,EAAE,sEAAsE;aAC5E,CAAC,CAAC;QACL,CAAC;QAED,IAAI,IAAI,CAAC,QAAQ,CAAC,aAAa,CAAC,MAAM,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,QAAQ,KAAK,IAAI,CAAC,GAAG,EAAE,CAAC;YAClF,QAAQ,CAAC,IAAI,CAAC;gBACZ,MAAM,EAAE,2BAA2B;gBACnC,QAAQ,EAAE,SAAS;gBACnB,OAAO,EAAE,eAAe,IAAI,CAAC,GAAG,iBAAiB,IAAI,CAAC,QAAQ,CAAC,QAAQ,GAAG;gBAC1E,OAAO,EAAE,IAAI,CAAC,GAAG;gBACjB,WAAW,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC;gBACrC,GAAG,EAAE,4CAA4C,IAAI,CAAC,QAAQ,CAAC,QAAQ,EAAE;aAC1E,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/rules/tech/soft-404.d.ts

@@ -0,0 +1,3 @@
+import type { ParsedPage, RuleResult } from "../../types.js";
+export declare function soft404Rule(pages: ParsedPage[]): RuleResult[];
+//# sourceMappingURL=soft-404.d.ts.map

package/dist/rules/tech/soft-404.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"soft-404.d.ts","sourceRoot":"","sources":["../../../src/rules/tech/soft-404.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AAI7D,wBAAgB,WAAW,CAAC,KAAK,EAAE,UAAU,EAAE,GAAG,UAAU,EAAE,CAsB7D"}

package/dist/rules/tech/soft-404.js

@@ -0,0 +1,24 @@
+const SOFT_404_PATTERNS = /\b(not\s*found|404|page\s*missing|does\s*not\s*exist|no\s*longer\s*available)\b/i;
+export function soft404Rule(pages) {
+    const findings = [];
+    for (const page of pages) {
+        if (!page.httpMeta)
+            continue;
+        if (page.httpMeta.statusCode !== 200)
+            continue;
+        const wordCount = page.contentText.split(/\s+/).filter(Boolean).length;
+        if (wordCount >= 50)
+            continue;
+        if (SOFT_404_PATTERNS.test(page.title)) {
+            findings.push({
+                ruleId: "tech/soft-404",
+                severity: "error",
+                message: `${page.url} returns HTTP 200 but appears to be an error page (title: "${page.title}", ${wordCount} words).`,
+                pageUrl: page.url,
+                fix: "Return a proper HTTP 404 status code for error pages instead of 200."
+            });
+        }
+    }
+    return findings;
+}
+//# sourceMappingURL=soft-404.js.map

package/dist/rules/tech/soft-404.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"soft-404.js","sourceRoot":"","sources":["../../../src/rules/tech/soft-404.ts"],"names":[],"mappings":"AAEA,MAAM,iBAAiB,GAAG,kFAAkF,CAAC;AAE7G,MAAM,UAAU,WAAW,CAAC,KAAmB;IAC7C,MAAM,QAAQ,GAAiB,EAAE,CAAC;IAElC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,IAAI,CAAC,IAAI,CAAC,QAAQ;YAAE,SAAS;QAC7B,IAAI,IAAI,CAAC,QAAQ,CAAC,UAAU,KAAK,GAAG;YAAE,SAAS;QAE/C,MAAM,SAAS,GAAG,IAAI,CAAC,WAAW,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC;QACvE,IAAI,SAAS,IAAI,EAAE;YAAE,SAAS;QAE9B,IAAI,iBAAiB,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YACvC,QAAQ,CAAC,IAAI,CAAC;gBACZ,MAAM,EAAE,eAAe;gBACvB,QAAQ,EAAE,OAAO;gBACjB,OAAO,EAAE,GAAG,IAAI,CAAC,GAAG,8DAA8D,IAAI,CAAC,KAAK,MAAM,SAAS,UAAU;gBACrH,OAAO,EAAE,IAAI,CAAC,GAAG;gBACjB,GAAG,EAAE,sEAAsE;aAC5E,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/types.d.ts

@@ -0,0 +1,170 @@
+export type Severity = "info" | "warning" | "error" | "critical";
+/** Options for `normalizeAuditUrl` (HTTP identity). */
+export interface NormalizeUrlOptions {
+    /** When true (default), drop `?query` for URL identity. */
+    stripQuery?: boolean;
+    /** When true, strip a leading `www.` from the hostname (opt-in; can be wrong for some TLDs). */
+    stripWwwHost?: boolean;
+}
+export interface ParseHtmlOptions {
+    normalizeUrl?: NormalizeUrlOptions;
+}
+export type FixEffort = "quick" | "moderate" | "structural";
+export type FindingContext = {
+    type: "cluster";
+    clusterSize: number;
+    members: string[];
+    worstPairs: Array<{
+        left: string;
+        right: string;
+        similarity: number;
+    }>;
+    similarityRange: [number, number];
+} | {
+    type: "contentBreakdown";
+    sharedBlocks: Array<{
+        text: string;
+        wordCount: number;
+    }>;
+    sharedWordCount: number;
+    uniqueWordCount: number;
+    totalWordCount: number;
+};
+export interface RuleResult {
+    ruleId: string;
+    severity: Severity;
+    message: string;
+    /** What to do about this finding. */
+    fix?: string;
+    /** Google documentation URL backing this finding. */
+    ref?: string;
+    /** Primary page this finding refers to, when applicable. */
+    pageUrl?: string;
+    /** Other URLs involved (e.g. cluster members, related pairs). */
+    relatedUrls?: string[];
+    /** Page group this finding belongs to, if page classification is configured. */
+    group?: string;
+    /** Numeric similarity score (0-1) for pairwise rules. Used by enrichment clustering. */
+    similarity?: number;
+    /** Structured context attached by the enrichment pipeline. */
+    context?: FindingContext;
+    /** Fix effort level assigned by the enrichment pipeline. */
+    effort?: FixEffort;
+}
+export interface CategoryScores {
+    spam: number;
+    content: number;
+    links: number;
+    tech: number;
+    schema: number;
+    cannibal: number;
+}
+export interface AuditSummary {
+    score: number;
+    categoryScores: CategoryScores;
+    groupScores?: Record<string, number>;
+    groupPageCounts?: Record<string, number>;
+    pageCount: number;
+    findings: RuleResult[];
+    /** True when the enrichment pipeline detects template-generated content. */
+    templateDetected?: boolean;
+    /** Pre-enrichment finding count, for backward compatibility with CI scripts. */
+    rawFindingCount?: number;
+}
+export interface PageGroupConfig {
+    /** Glob pattern(s) to match page URLs/paths. */
+    match: string | string[];
+    /** Rule globs to enable for this group. If omitted, all rules apply. */
+    rules?: string[];
+    /** Per-rule threshold or severity overrides. Keys are rule IDs. */
+    overrides?: Record<string, Record<string, unknown>>;
+}
+export interface AuditOptions {
+    rules?: {
+        /** Passed to `normalizeAuditUrl` for crawl URLs and resolved links (default: strip query, keep www). */
+        stripUrlQuery?: boolean;
+        stripWwwHost?: boolean;
+        nearDuplicateThreshold?: number;
+        entitySwapThreshold?: number;
+        thinContentMinWords?: number;
+        publicationVelocityMaxPerDay?: number;
+        boilerplateMaxRatio?: number;
+        templateDiversityMinUniqueRatio?: number;
+        uniqueValueMinWords?: number;
+        metaUniquenessMinJaccard?: number;
+        linkDepthMaxClicks?: number;
+        /** Minimum pages in one directory before hub/index coverage is required. */
+        hubPagesMinSiblings?: number;
+        /** Skip hub/index checks when a directory has more than this many pages (e.g. large blogs). */
+        hubPagesMaxSiblings?: number;
+        titleOverlapThreshold?: number;
+        keywordCollisionMinShared?: number;
+        templateCoverageMinPages?: number;
+    };
+    /** Max parallel HTTP fetches when auditing a remote sitemap (default: 5). */
+    concurrency?: number;
+    /** Per-request timeout in milliseconds (default: 30000). */
+    timeout?: number;
+    /** Audit a random subset of N pages. 0 means all pages (default: 0). */
+    sampleSize?: number;
+    /** URL/path glob patterns to exclude from the audit. */
+    ignore?: string[];
+    crawlDiscovery?: boolean;
+    /** Page groups with per-group rule sets and threshold overrides. */
+    pageGroups?: Record<string, PageGroupConfig>;
+    /** Browser rendering options for client-rendered pages. */
+    render?: {
+        browserWsEndpoint?: string;
+    };
+    /** Override template auto-detection. When set, skips heuristic detection. */
+    templateGenerated?: boolean;
+}
+export interface EntityMaskPattern {
+    placeholder: string;
+    pattern: RegExp;
+}
+export interface HttpMeta {
+    statusCode: number;
+    finalUrl: string;
+    redirectChain: string[];
+    xRobotsTag: string;
+    linkHeader: string;
+}
+export interface ParsedPage {
+    url: string;
+    title: string;
+    metaDescription: string;
+    canonical: string;
+    robotsMeta: string;
+    og: {
+        title: string;
+        description: string;
+        image: string;
+    };
+    hreflangs: Array<{
+        lang: string;
+        href: string;
+    }>;
+    headings: {
+        h1: string[];
+        h2: string[];
+    };
+    /**
+     * Resolved `a[href]` targets. For HTTP(S) page URLs, only `http:` / `https:` targets are kept
+     * (`javascript:`, `data:`, etc. are dropped). For filesystem page URLs, paths are normalized.
+     */
+    resolvedHrefs: string[];
+    publishedDate?: string;
+    structureSignature: string;
+    jsonLd: unknown[];
+    authorSignals: {
+        metaAuthor: string;
+        schemaAuthor: boolean;
+        bylineElement: boolean;
+        relAuthorLink: boolean;
+    };
+    contentText: string;
+    html: string;
+    httpMeta?: HttpMeta;
+}
+//# sourceMappingURL=types.d.ts.map

package/dist/types.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,QAAQ,GAAG,MAAM,GAAG,SAAS,GAAG,OAAO,GAAG,UAAU,CAAC;AAEjE,uDAAuD;AACvD,MAAM,WAAW,mBAAmB;IAClC,2DAA2D;IAC3D,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB,gGAAgG;IAChG,YAAY,CAAC,EAAE,OAAO,CAAC;CACxB;AAED,MAAM,WAAW,gBAAgB;IAC/B,YAAY,CAAC,EAAE,mBAAmB,CAAC;CACpC;AAED,MAAM,MAAM,SAAS,GAAG,OAAO,GAAG,UAAU,GAAG,YAAY,CAAC;AAE5D,MAAM,MAAM,cAAc,GACtB;IACE,IAAI,EAAE,SAAS,CAAC;IAChB,WAAW,EAAE,MAAM,CAAC;IACpB,OAAO,EAAE,MAAM,EAAE,CAAC;IAClB,UAAU,EAAE,KAAK,CAAC;QAChB,IAAI,EAAE,MAAM,CAAC;QACb,KAAK,EAAE,MAAM,CAAC;QACd,UAAU,EAAE,MAAM,CAAC;KACpB,CAAC,CAAC;IACH,eAAe,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;CACnC,GACD;IACE,IAAI,EAAE,kBAAkB,CAAC;IACzB,YAAY,EAAE,KAAK,CAAC;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IACzD,eAAe,EAAE,MAAM,CAAC;IACxB,eAAe,EAAE,MAAM,CAAC;IACxB,cAAc,EAAE,MAAM,CAAC;CACxB,CAAC;AAEN,MAAM,WAAW,UAAU;IACzB,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,QAAQ,CAAC;IACnB,OAAO,EAAE,MAAM,CAAC;IAChB,qCAAqC;IACrC,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,qDAAqD;IACrD,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,4DAA4D;IAC5D,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,iEAAiE;IACjE,WAAW,CAAC,EAAE,MAAM,EAAE,CAAC;IACvB,gFAAgF;IAChF,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,wFAAwF;IACxF,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,8DAA8D;IAC9D,OAAO,CAAC,EAAE,cAAc,CAAC;IACzB,4DAA4D;IAC5D,MAAM,CAAC,EAAE,SAAS,CAAC;CACpB;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE,MAAM,CAAC;IACd,cAAc,EAAE,cAAc,CAAC;IAC/B,WAAW,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACrC,eAAe,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACzC,SAAS,EAAE,MAAM,CAAC;IAClB,QAAQ,EAAE,UAAU,EAAE,CAAC;IACvB,4EAA4E;IAC5E,gBAAgB,CAAC,EAAE,OAAO,CAAC;IAC3B,gFAAgF;IAChF,eAAe,CAAC,EAAE,MAAM,CAAC;CAC1B;AAED,MAAM,WAAW,eAAe;IAC9B,gDAAgD;IAChD,KAAK,EAAE,MAAM,GAAG,MAAM,EAAE,CAAC;IACzB,wEAAwE;IACxE,KAAK,CAAC,EAAE,MAAM,EAAE,CAAC;IACjB,mEAAmE;IACnE,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC,CAAC;CACrD;AAED,MAAM,WAAW,YAAY;IAC3B,KAAK,CAAC,EAAE;QACN,wGAAwG;QACxG,aAAa,CAAC,EAAE,OAAO,CAAC;QACxB,YAAY,CAAC,EAAE,OAAO,CAAC;QACvB,sBAAsB,CAAC,EAAE,MAAM,CAAC;QAChC,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,4BAA4B,CAAC,EAAE,MAAM,CAAC;QACtC,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,+BAA+B,CAAC,EAAE,MAAM,CAAC;QACzC,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,wBAAwB,CAAC,EAAE,MAAM,CAAC;QAClC,kBAAkB,CAAC,EAAE,MAAM,CAAC;QAC5B,4EAA4E;QAC5E,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,+FAA+F;QAC/F,mBAAmB,CAAC,EAAE,MAAM,CAAC;QAC7B,qBAAqB,CAAC,EAAE,MAAM,CAAC;QAC/B,yBAAyB,CAAC,EAAE,MAAM,CAAC;QACnC,wBAAwB,CAAC,EAAE,MAAM,CAAC;KACnC,CAAC;IACF,6EAA6E;IAC7E,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,4DAA4D;IAC5D,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,wEAAwE;IACxE,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,wDAAwD;IACxD,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,cAAc,CAAC,EAAE,OAAO,CAAC;IACzB,oEAAoE;IACpE,UAAU,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,eAAe,CAAC,CAAC;IAC7C,2DAA2D;IAC3D,MAAM,CAAC,EAAE;QACP,iBAAiB,CAAC,EAAE,MAAM,CAAC;KAC5B,CAAC;IACF,6EAA6E;IAC7E,iBAAiB,CAAC,EAAE,OAAO,CAAC;CAC7B;AAED,MAAM,WAAW,iBAAiB;IAChC,WAAW,EAAE,MAAM,CAAC;IACpB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,QAAQ;IACvB,UAAU,EAAE,MAAM,CAAC;IACnB,QAAQ,EAAE,MAAM,CAAC;IACjB,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,UAAU;IACzB,GAAG,EAAE,MAAM,CAAC;IACZ,KAAK,EAAE,MAAM,CAAC;IACd,eAAe,EAAE,MAAM,CAAC;IACxB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;IACnB,EAAE,EAAE;QACF,KAAK,EAAE,MAAM,CAAC;QACd,WAAW,EAAE,MAAM,CAAC;QACpB,KAAK,EAAE,MAAM,CAAC;KACf,CAAC;IACF,SAAS,EAAE,KAAK,CAAC;QACf,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;IACH,QAAQ,EAAE;QACR,EAAE,EAAE,MAAM,EAAE,CAAC;QACb,EAAE,EAAE,MAAM,EAAE,CAAC;KACd,CAAC;IACF;;;OAGG;IACH,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,kBAAkB,EAAE,MAAM,CAAC;IAC3B,MAAM,EAAE,OAAO,EAAE,CAAC;IAClB,aAAa,EAAE;QACb,UAAU,EAAE,MAAM,CAAC;QACnB,YAAY,EAAE,OAAO,CAAC;QACtB,aAAa,EAAE,OAAO,CAAC;QACvB,aAAa,EAAE,OAAO,CAAC;KACxB,CAAC;IACF,WAAW,EAAE,MAAM,CAAC;IACpB,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,QAAQ,CAAC;CACrB"}

package/dist/types.js

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.js.map

package/dist/types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.js","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":""}

package/dist/url-normalize.d.ts

@@ -0,0 +1,10 @@
+import type { NormalizeUrlOptions } from "./types.js";
+export declare function mergeNormalizeUrlOptions(options?: NormalizeUrlOptions): Required<NormalizeUrlOptions>;
+/**
+ * Stable URL/path identity for audits: deduping crawl entries and matching edges.
+ * HTTP(S): lowercase host, strip fragment, optional query strip, optional `www.` strip,
+ * drop default ports, trim trailing slash on non-root paths.
+ * Filesystem: Node path.normalize.
+ */
+export declare function normalizeAuditUrl(url: string, options?: NormalizeUrlOptions): string;
+//# sourceMappingURL=url-normalize.d.ts.map

package/dist/url-normalize.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"url-normalize.d.ts","sourceRoot":"","sources":["../src/url-normalize.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,mBAAmB,EAAE,MAAM,YAAY,CAAC;AAOtD,wBAAgB,wBAAwB,CAAC,OAAO,CAAC,EAAE,mBAAmB,GAAG,QAAQ,CAAC,mBAAmB,CAAC,CAKrG;AAED;;;;;GAKG;AACH,wBAAgB,iBAAiB,CAAC,GAAG,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,mBAAmB,GAAG,MAAM,CASpF"}

package/dist/url-normalize.js

@@ -0,0 +1,52 @@
+import { normalize } from "node:path";
+const DEFAULTS = {
+    stripQuery: true,
+    stripWwwHost: false
+};
+export function mergeNormalizeUrlOptions(options) {
+    return {
+        stripQuery: options?.stripQuery ?? DEFAULTS.stripQuery,
+        stripWwwHost: options?.stripWwwHost ?? DEFAULTS.stripWwwHost
+    };
+}
+/**
+ * Stable URL/path identity for audits: deduping crawl entries and matching edges.
+ * HTTP(S): lowercase host, strip fragment, optional query strip, optional `www.` strip,
+ * drop default ports, trim trailing slash on non-root paths.
+ * Filesystem: Node path.normalize.
+ */
+export function normalizeAuditUrl(url, options) {
+    const trimmed = url.trim();
+    if (!trimmed) {
+        return trimmed;
+    }
+    if (/^https?:\/\//i.test(trimmed)) {
+        return normalizeHttpUrl(trimmed, mergeNormalizeUrlOptions(options));
+    }
+    return normalize(trimmed);
+}
+function normalizeHttpUrl(url, opts) {
+    const u = new URL(url);
+    u.hash = "";
+    if (opts.stripQuery) {
+        u.search = "";
+    }
+    let hostname = u.hostname.toLowerCase();
+    if (opts.stripWwwHost && hostname.startsWith("www.")) {
+        hostname = hostname.slice(4);
+    }
+    u.hostname = hostname;
+    if (u.protocol === "http:" && u.port === "80") {
+        u.port = "";
+    }
+    if (u.protocol === "https:" && u.port === "443") {
+        u.port = "";
+    }
+    let pathname = u.pathname;
+    if (pathname.length > 1 && pathname.endsWith("/")) {
+        pathname = pathname.slice(0, -1);
+    }
+    u.pathname = pathname || "/";
+    return u.href;
+}
+//# sourceMappingURL=url-normalize.js.map

package/dist/url-normalize.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"url-normalize.js","sourceRoot":"","sources":["../src/url-normalize.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AAGtC,MAAM,QAAQ,GAAkC;IAC9C,UAAU,EAAE,IAAI;IAChB,YAAY,EAAE,KAAK;CACpB,CAAC;AAEF,MAAM,UAAU,wBAAwB,CAAC,OAA6B;IACpE,OAAO;QACL,UAAU,EAAE,OAAO,EAAE,UAAU,IAAI,QAAQ,CAAC,UAAU;QACtD,YAAY,EAAE,OAAO,EAAE,YAAY,IAAI,QAAQ,CAAC,YAAY;KAC7D,CAAC;AACJ,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,iBAAiB,CAAC,GAAW,EAAE,OAA6B;IAC1E,MAAM,OAAO,GAAG,GAAG,CAAC,IAAI,EAAE,CAAC;IAC3B,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,OAAO,OAAO,CAAC;IACjB,CAAC;IACD,IAAI,eAAe,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;QAClC,OAAO,gBAAgB,CAAC,OAAO,EAAE,wBAAwB,CAAC,OAAO,CAAC,CAAC,CAAC;IACtE,CAAC;IACD,OAAO,SAAS,CAAC,OAAO,CAAC,CAAC;AAC5B,CAAC;AAED,SAAS,gBAAgB,CAAC,GAAW,EAAE,IAAmC;IACxE,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,CAAC,CAAC,IAAI,GAAG,EAAE,CAAC;IACZ,IAAI,IAAI,CAAC,UAAU,EAAE,CAAC;QACpB,CAAC,CAAC,MAAM,GAAG,EAAE,CAAC;IAChB,CAAC;IACD,IAAI,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC;IACxC,IAAI,IAAI,CAAC,YAAY,IAAI,QAAQ,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;QACrD,QAAQ,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IAC/B,CAAC;IACD,CAAC,CAAC,QAAQ,GAAG,QAAQ,CAAC;IACtB,IAAI,CAAC,CAAC,QAAQ,KAAK,OAAO,IAAI,CAAC,CAAC,IAAI,KAAK,IAAI,EAAE,CAAC;QAC9C,CAAC,CAAC,IAAI,GAAG,EAAE,CAAC;IACd,CAAC;IACD,IAAI,CAAC,CAAC,QAAQ,KAAK,QAAQ,IAAI,CAAC,CAAC,IAAI,KAAK,KAAK,EAAE,CAAC;QAChD,CAAC,CAAC,IAAI,GAAG,EAAE,CAAC;IACd,CAAC;IACD,IAAI,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC;IAC1B,IAAI,QAAQ,CAAC,MAAM,GAAG,CAAC,IAAI,QAAQ,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;QAClD,QAAQ,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IACnC,CAAC;IACD,CAAC,CAAC,QAAQ,GAAG,QAAQ,IAAI,GAAG,CAAC;IAC7B,OAAO,CAAC,CAAC,IAAI,CAAC;AAChB,CAAC"}

package/package.json

@@ -0,0 +1,46 @@
+{
+  "name": "@pseolint/core",
+  "version": "0.1.0",
+  "description": "Programmatic SEO audit engine — 34 rules across 6 categories for SpamBrain risk detection",
+  "license": "MIT",
+  "author": "Ouranos Labs <contact@ouranos-labs.dev>",
+  "homepage": "https://pseolint.dev",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/ouranos-labs/pseolint.git",
+    "directory": "packages/core"
+  },
+  "keywords": ["seo", "pseo", "programmatic-seo", "spambrain", "lint", "audit"],
+  "type": "module",
+  "exports": {
+    ".": {
+      "import": "./dist/index.js",
+      "types": "./dist/index.d.ts"
+    }
+  },
+  "main": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "engines": {
+    "node": ">=18"
+  },
+  "files": [
+    "dist"
+  ],
+  "scripts": {
+    "build": "tsc -p tsconfig.json",
+    "lint": "tsc --noEmit -p tsconfig.json",
+    "test": "vitest run --passWithNoTests tests/**/*.test.ts",
+    "typecheck": "tsc --noEmit -p tsconfig.json"
+  },
+  "dependencies": {
+    "cheerio": "^1.2.0"
+  },
+  "peerDependencies": {
+    "playwright-core": ">=1.40.0"
+  },
+  "peerDependenciesMeta": {
+    "playwright-core": {
+      "optional": true
+    }
+  }
+}