@slicedearth/whoisleuth-cli 2.0.0 → 2.1.0

package/lib/whois-common-formats.mjs

@@ -0,0 +1,276 @@
+// Common contact, bracketed-record, nameserver and status formats applied
+// after generic fields and marker-gated registry dialects.
+import { parseIndentedContactBlock, resolveFredContact, } from "./whois-contacts.mjs";
+import { MAX_WHOIS_NAMESERVERS, MAX_WHOIS_STATUSES, addBoundedWhoisSetValue, } from "./whois-parser-support.mjs";
+import { boundedWhoisValue } from "./whois-values.mjs";
+function applyWhoisCommonFormats(text, isRootHop, dialects, context) {
+    const { expandedStreetFields, fields, nameservers, statuses, truncatedFields, } = context;
+    const { isAtiTn, isDnsBelgium, isEif, isHkirc, isIsocIl, isNicKz, isNicMd, isPunktum, isRnids, isTwnic, } = dialects;
+    if (!isRootHop) {
+        for (const [prefix, label] of [
+            ['registrant', 'Registrant'], ['admin', 'Admin(?:istrative)?'],
+            ['tech', 'Tech(?:nical)?'], ['billing', 'Billing'],
+        ]) {
+            const key = `${prefix}Street`;
+            if (expandedStreetFields.has(key))
+                continue;
+            const streetRe = new RegExp(`^[ \\t*]*${label} (?:Contact )?Street[ \\t.]*:[ \\t]*(.+)$`, 'gim');
+            const lines = [];
+            for (const match of text.matchAll(streetRe)) {
+                const bounded = boundedWhoisValue(match[1], 300);
+                if (bounded.truncated)
+                    truncatedFields.add(key);
+                if (!bounded.value || lines.includes(bounded.value))
+                    continue;
+                if (lines.length >= 4) {
+                    truncatedFields.add(key);
+                    break;
+                }
+                lines.push(bounded.value);
+            }
+            if (lines.length)
+                fields[key] = lines.join(', ');
+            expandedStreetFields.add(key);
+        }
+        // The documented .nz protocol numbers up to two address lines using
+        // underscore field names. Aggregate only those exact fields, keeping
+        // the same four-line/300-character bounds as repeated ICANN-style
+        // street fields and leaving city/province/postcode separately typed.
+        for (const [prefix, nzPrefix] of [
+            ['registrant', 'registrant'], ['admin', 'admin'], ['tech', 'technical'],
+        ]) {
+            const key = `${prefix}Street`;
+            if (fields[key])
+                continue;
+            const addressRe = new RegExp(`^[ \\t]*${nzPrefix}_contact_address(?:1|2)[ \\t]*:[ \\t]*(.+)$`, 'gim');
+            const lines = [];
+            for (const match of text.matchAll(addressRe)) {
+                const bounded = boundedWhoisValue(match[1], 300);
+                if (bounded.truncated)
+                    truncatedFields.add(key);
+                if (!bounded.value || lines.includes(bounded.value))
+                    continue;
+                if (lines.length >= 4) {
+                    truncatedFields.add(key);
+                    break;
+                }
+                lines.push(bounded.value);
+            }
+            if (lines.length)
+                fields[key] = lines.join(', ');
+        }
+    }
+    // If registrantName looks like it's actually a handle (a matching
+    // "contact: <handle>" block exists in this same hop), resolve it for
+    // the real name/org/email/phone/address. Harmless no-op otherwise -
+    // registries where "Registrant:" is already the real name (e.g. .kr)
+    // simply won't have a matching contact block to find.
+    if (!isRootHop && fields.registrantName) {
+        const resolved = resolveFredContact(text, fields.registrantName);
+        if (resolved) {
+            if (resolved.name)
+                fields.registrantName = resolved.name;
+            if (resolved.org && !fields.registrantOrg)
+                fields.registrantOrg = resolved.org;
+            if (resolved.email && !fields.registrantEmail)
+                fields.registrantEmail = resolved.email;
+            if (resolved.phone && !fields.registrantPhone)
+                fields.registrantPhone = resolved.phone;
+            if (resolved.address && !fields.registrantAddress)
+                fields.registrantAddress = resolved.address;
+        }
+    }
+    // EDUCAUSE (.edu) and similar legacy registries list registrant/admin/
+    // technical contacts as indented blocks rather than "Field: value"
+    // pairs - see parseIndentedContactBlock. The plain "Registrant:" header
+    // has no separate person name on .edu (registrants are institutions),
+    // so its block's first line maps to the org, not a name.
+    if (!isRootHop && !fields.registrantOrg && !fields.registrantName) {
+        const block = parseIndentedContactBlock(text, /^[ \t]*Registrant:[ \t]*$/im);
+        if (block) {
+            if (block.name)
+                fields.registrantOrg = block.name;
+            if (block.address)
+                fields.registrantAddress = block.address;
+            if (block.phone)
+                fields.registrantPhone = block.phone;
+            if (block.email)
+                fields.registrantEmail = block.email;
+            if (block.truncated)
+                truncatedFields.add('registrantAddress');
+        }
+    }
+    if (!isRootHop && !fields.adminName) {
+        const block = parseIndentedContactBlock(text, /^[ \t]*Administrative Contact:[ \t]*$/im);
+        if (block) {
+            if (block.name)
+                fields.adminName = block.name;
+            if (block.address)
+                fields.adminAddress = block.address;
+            if (block.phone)
+                fields.adminPhone = block.phone;
+            if (block.email)
+                fields.adminEmail = block.email;
+            if (block.truncated)
+                truncatedFields.add('adminAddress');
+        }
+    }
+    if (!isRootHop && !fields.techName) {
+        const block = parseIndentedContactBlock(text, /^[ \t]*Technical Contact:[ \t]*$/im);
+        if (block) {
+            if (block.name)
+                fields.techName = block.name;
+            if (block.address)
+                fields.techAddress = block.address;
+            if (block.phone)
+                fields.techPhone = block.phone;
+            if (block.email)
+                fields.techEmail = block.email;
+            if (block.truncated)
+                truncatedFields.add('techAddress');
+        }
+    }
+    if (!isRootHop && !fields.billingName) {
+        const block = parseIndentedContactBlock(text, /^[ \t]*Billing Contact:[ \t]*$/im);
+        if (block) {
+            if (block.name)
+                fields.billingName = block.name;
+            if (block.address)
+                fields.billingAddress = block.address;
+            if (block.phone)
+                fields.billingPhone = block.phone;
+            if (block.email)
+                fields.billingEmail = block.email;
+            if (block.truncated)
+                truncatedFields.add('billingAddress');
+        }
+    }
+    // .jp (JPRS) uses a bracketed dual-language format instead of
+    // "Label: value" - e.g. "[Domain Name]   GOO.JP", with Japanese-only
+    // labels alongside English ones for the same field. The bracket syntax
+    // is distinctive enough to not need root-hop gating.
+    if (!fields.domainName) {
+        const m = text.match(/\[Domain Name\][ \t]*(.+)/i);
+        if (m?.[1])
+            fields.domainName = m[1].trim();
+    }
+    if (!fields.registrantName) {
+        const m = text.match(/\[Registrant\][ \t]*(.+)/i);
+        if (m?.[1])
+            fields.registrantName = m[1].trim();
+    }
+    if (!fields.createdDate) {
+        const m = text.match(/\[登録年月日\][ \t]*(.+)/);
+        if (m?.[1])
+            fields.createdDate = m[1].trim();
+    }
+    if (!fields.expiryDate) {
+        const m = text.match(/\[有効期限\][ \t]*(.+)/);
+        if (m?.[1])
+            fields.expiryDate = m[1].trim();
+    }
+    for (const m of text.matchAll(/\[状態\][ \t]*(.+)/g)) {
+        if (addBoundedWhoisSetValue(statuses, m[1], {
+            maxEntries: MAX_WHOIS_STATUSES, maxLength: 160,
+            field: 'statuses', truncatedFields,
+        }) === 'capped')
+            break;
+    }
+    for (const m of text.matchAll(/\[Name Server\][ \t]*([a-zA-Z0-9.\-]+)/gi)) {
+        if (addBoundedWhoisSetValue(nameservers, m[1], {
+            maxEntries: MAX_WHOIS_NAMESERVERS, maxLength: 253,
+            field: 'nameservers', truncatedFields,
+        }) === 'capped')
+            break;
+    }
+    // "Name Server:" never collides with the IANA root hop (which uses
+    // lowercase "nserver:" for the TLD's own root nameservers) so it's safe
+    // on every hop; "nserver:"/"Host Name:"/"DNS:" are real per-domain
+    // labels on some registries (e.g. .ru, .kr, .mx) but only once we're
+    // past the root hop, for the same reason as the field patterns above.
+    const nsLinePatterns = [/^[ \t*]*Name Server[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim];
+    if (!isRootHop) {
+        nsLinePatterns.push(/^[ \t*]*nserver[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim, /^[ \t*]*Nameserver[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim, /^[ \t*]*Host Name[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim, /^[ \t]*ns_name_\d{2}[ \t]*:[ \t]*([a-zA-Z0-9.\-]+)/gim);
+        // Punktum dk uses `DNS: example.dk` for the queried domain, then
+        // `Hostname:` inside its nameserver section. Other supported
+        // registries use `DNS:` for an actual nameserver, so switch aliases
+        // only when the full .dk marker set is present.
+        nsLinePatterns.push(isPunktum
+            ? /^[ \t*]*Hostname[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim
+            : /^[ \t*]*DNS[ \t.]*:[ \t]*([a-zA-Z0-9.\-]+)/gim);
+    }
+    for (const re of nsLinePatterns) {
+        for (const m of text.matchAll(re)) {
+            if (addBoundedWhoisSetValue(nameservers, m[1], {
+                maxEntries: MAX_WHOIS_NAMESERVERS, maxLength: 253,
+                field: 'nameservers', truncatedFields,
+            }) === 'capped')
+                break;
+        }
+    }
+    // Same root-hop exclusion as above: bare "status:" is also how IANA
+    // reports the TLD's own delegation status (e.g. "status: ACTIVE" for
+    // .gt itself), not the queried domain's status. Deliberately not adding
+    // "state:" as an alternate here (some registries, e.g. .ru/.se, use it
+    // for domain status) - it's also the standard label for a postal
+    // address's state/province in Name/City/State/Country contact blocks
+    // (seen on .mx), and there's no reliable way to tell those apart from
+    // the label alone - a missing status is safer than a wrong one.
+    const statusRe = isRootHop
+        ? /^[ \t*]*Domain Status[ \t.]*:[ \t]*([a-zA-Z][a-zA-Z0-9_-]*)/gim
+        : (isDnsBelgium || isPunktum || isIsocIl || isTwnic || isEif
+            || isHkirc || isNicKz || isNicMd || isAtiTn)
+            ? /^[ \t*]*(?:Domain Status|Status)[ \t.]*:[ \t]*([^\r\n]+)/gim
+            : /^[ \t*]*(?:Domain Status|Status)[ \t.]*:[ \t]*([a-zA-Z][a-zA-Z0-9_-]*)/gim;
+    if (!isRnids) {
+        for (const m of text.matchAll(statusRe)) {
+            if (addBoundedWhoisSetValue(statuses, m[1], {
+                maxEntries: MAX_WHOIS_STATUSES, maxLength: 160,
+                field: 'statuses', truncatedFields,
+            }) === 'capped')
+                break;
+        }
+    }
+    if (!isRootHop) {
+        for (const m of text.matchAll(/^[ \t]*query_status[ \t]*:[ \t]*(\d{3}(?:[ \t]+[^\r\n]+)?)/gim)) {
+            if (addBoundedWhoisSetValue(statuses, m[1], {
+                maxEntries: MAX_WHOIS_STATUSES, maxLength: 160,
+                field: 'statuses', truncatedFields,
+            }) === 'capped')
+                break;
+        }
+    }
+    // Some registries (e.g. .it, .tr) list nameservers as a bare header
+    // ("Nameservers", "Domain Servers") followed by unlabeled lines -
+    // sometimes just a hostname, sometimes "hostname  ip.addr" - rather
+    // than a per-line "Name Server:" label. Only meaningful on non-root hops.
+    if (!isRootHop && nameservers.size === 0) {
+        const headerMatch = text.match(/^[ \t*]*(?:Name ?[Ss]ervers|Domain Servers|DNS servers)[ \t.]*:?[ \t]*$/mi);
+        if (headerMatch) {
+            let found = 0;
+            for (const line of text.slice((headerMatch.index ?? 0) + headerMatch[0].length).split('\n')) {
+                const trimmed = line.trim();
+                if (!trimmed) {
+                    if (found > 0)
+                        break; // blank line after >=1 hostname ends the section
+                    continue; // the header line's own line break - not a real gap yet
+                }
+                const hostMatch = trimmed.match(/^([a-zA-Z0-9.\-]+\.[a-zA-Z]{2,})(?:\s|$)/);
+                if (hostMatch) {
+                    const result = addBoundedWhoisSetValue(nameservers, hostMatch[1], {
+                        maxEntries: MAX_WHOIS_NAMESERVERS, maxLength: 253,
+                        field: 'nameservers', truncatedFields,
+                    });
+                    if (result === 'capped')
+                        break;
+                    if (result === 'added')
+                        found += 1;
+                }
+                else {
+                    break;
+                }
+            }
+        }
+    }
+}
+export { applyWhoisCommonFormats };