feat: infer country from member location (CM-1355)

backend/src/services/member/memberAttributesService.ts

@@ -2,7 +2,7 @@
 import * as lodash from 'lodash'
 
 import { captureApiChange, memberEditProfileAction } from '@crowd/audit-logs'
-import { Error404 } from '@crowd/common'
+import { Error404, getAttributeValue, getCountry, hasAttributeValue } from '@crowd/common'
 import {
   deleteMemberBotSuggestion,
   deleteMemberNoBot,
@@ -78,6 +78,18 @@ export default class MemberAttributesService extends LoggerBase {
             }
           }
 
+          if (!hasAttributeValue(data.country)) {
+            const location = getAttributeValue(data.location)
+            const country = getCountry(location)
+            if (country) {
+              data.country = {
+                ...data.country,
+                system: country,
+                default: country,
+              }
+            }
+          }
+
           await updateMemberAttributes(qx, memberId, data)
 
           // Handle isBot status and maintain consistency with bot tracking tables

backend/src/services/memberService.ts

@@ -4,7 +4,15 @@ import moment from 'moment-timezone'
 import validator from 'validator'
 
 import { captureApiChange, memberUnmergeAction } from '@crowd/audit-logs'
-import { Error400, calculateReach, getProperDisplayName, isDomainExcluded } from '@crowd/common'
+import {
+  Error400,
+  calculateReach,
+  getAttributeValue,
+  getCountry,
+  getProperDisplayName,
+  hasAttributeValue,
+  isDomainExcluded,
+} from '@crowd/common'
 import {
   CommonMemberService,
   getGithubInstallationToken,
@@ -446,6 +454,17 @@ export default class MemberService extends LoggerBase {
         const toUpdate = CommonMemberService.membersMerge(existing, data)
 
         if (toUpdate.attributes) {
+          if (!hasAttributeValue(toUpdate.attributes.country)) {
+            const location = getAttributeValue(toUpdate.attributes.location)
+            const country = getCountry(location)
+            if (country) {
+              toUpdate.attributes.country = {
+                ...toUpdate.attributes.country,
+                system: country,
+              }
+            }
+          }
+
           toUpdate.attributes = await this.setAttributesDefaultValues(toUpdate.attributes)
         }
 
@@ -459,6 +478,17 @@ export default class MemberService extends LoggerBase {
         // It is important to call it with doPopulateRelations=false
         // because otherwise the performance is greatly decreased in integrations
         if (data.attributes) {
+          if (!hasAttributeValue(data.attributes.country)) {
+            const location = getAttributeValue(data.attributes.location)
+            const country = getCountry(location)
+            if (country) {
+              data.attributes.country = {
+                ...data.attributes.country,
+                system: country,
+              }
+            }
+          }
+
           data.attributes = await this.setAttributesDefaultValues(data.attributes)
         }
 
@@ -789,6 +819,20 @@ export default class MemberService extends LoggerBase {
         data.displayName = getProperDisplayName(data.displayName)
       }
 
+      if (data.attributes) {
+        if (!hasAttributeValue(data.attributes.country)) {
+          const location = getAttributeValue(data.attributes.location)
+          const country = getCountry(location)
+          if (country) {
+            data.attributes.country = {
+              ...data.attributes.country,
+              system: country,
+              default: data.attributes.country?.default ?? country,
+            }
+          }
+        }
+      }
+
       const record = await MemberRepository.update(id, data, repoOptions, {
         manualChange,
       })

services/apps/data_sink_worker/src/service/member.service.ts

@@ -5,8 +5,11 @@ import uniqby from 'lodash.uniqby'
 import {
   ApplicationError,
   DEFAULT_TENANT_ID,
+  getAttributeValue,
+  getCountry,
   getEarliestValidDate,
   getProperDisplayName,
+  hasAttributeValue,
   isDomainExcluded,
   isObjectEmpty,
   isSameMemberIdentity,
@@ -331,6 +334,14 @@ export default class MemberService extends LoggerBase {
               'memberService -> create -> validateAttributes',
             )
 
+            if (!hasAttributeValue(attributes.country)) {
+              const location = getAttributeValue(attributes.location)
+              const country = getCountry(location)
+              if (country) {
+                attributes.country = Object.assign({}, attributes.country, { system: country })
+              }
+            }
+
             attributes = await logExecutionTimeV2(
               () => memberAttributeService.setAttributesDefaultValues(attributes),
               this.log,
@@ -592,6 +603,17 @@ export default class MemberService extends LoggerBase {
 
           if (toUpdate.attributes) {
             this.log.trace({ memberId: id }, 'Setting attribute default values!')
+
+            if (!hasAttributeValue(toUpdate.attributes.country)) {
+              const location = getAttributeValue(toUpdate.attributes.location)
+              const country = getCountry(location)
+              if (country) {
+                toUpdate.attributes.country = Object.assign({}, toUpdate.attributes.country, {
+                  system: country,
+                })
+              }
+            }
+
             toUpdate.attributes = await logExecutionTimeV2(
               () => memberAttributeService.setAttributesDefaultValues(toUpdate.attributes),
               this.log,

services/apps/members_enrichment_worker/src/activities/enrichment.ts

@@ -4,6 +4,9 @@ import _ from 'lodash'
 
 import {
   generateUUIDv1,
+  getAttributeValue,
+  getCountry,
+  hasAttributeValue,
   hasIntersection,
   replaceDoubleQuotes,
   sanitizeMemberOrganizationDateRange,
@@ -45,6 +48,7 @@ import { refreshMaterializedView } from '@crowd/data-access-layer/src/utils'
 import { SearchSyncApiClient } from '@crowd/opensearch'
 import { RedisCache } from '@crowd/redis'
 import {
+  IAttributes,
   IEnrichableMember,
   IEnrichableMemberIdentityActivityAggregate,
   IMemberEnrichmentCache,
@@ -347,16 +351,28 @@ export async function updateMemberUsingSquashedPayload(
     }
 
     // process attributes
-    let attributes = existingMemberData.attributes as Record<string, unknown>
+    let attributes = existingMemberData.attributes as IAttributes
 
     if (squashedPayload.attributes) {
       svc.log.debug({ memberId }, 'Updating member attributes!')
 
       attributes = _.merge({}, attributes, squashedPayload.attributes)
 
       if (Object.keys(attributes).length > 0) {
+        // Infer country from location when no country source is set.
+        if (!hasAttributeValue(attributes.country)) {
+          const location = getAttributeValue(attributes.location)
+          const country = getCountry(location)
+          if (country) {
+            attributes.country = {
+              ...attributes.country,
+              system: country,
+            }
+          }
+        }
+
         const priorities = await getPriorityArray()
-        attributes = await setAttributesDefaultValues(attributes, priorities)
+        attributes = (await setAttributesDefaultValues(attributes, priorities)) as IAttributes
       }
       didUpdate = true
       await updateMemberAttributes(qx, memberId, attributes)

services/libs/common/package.json

@@ -16,6 +16,7 @@
   "dependencies": {
     "@crowd/types": "workspace:*",
     "config": "^3.3.9",
+    "i18n-iso-countries": "^7.14.0",
     "lodash.clonedeep": "^4.5.0",
     "lodash.get": "~4.4.2",
     "lodash.isarray": "^4.0.0",

services/libs/common/src/constants/index.ts

@@ -1,3 +1,4 @@
 export * from './email-providers'
 export * from './bots'
 export * from './disposable-email-domains'
+export * from './location'

services/libs/common/src/constants/location.ts

@@ -0,0 +1,108 @@
+/**
+ * US state / district names (not postal codes).
+ * Excludes georgia and washington — too easy to confuse with other places.
+ */
+export const US_STATE_NAMES = new Set([
+  'alabama',
+  'alaska',
+  'arizona',
+  'arkansas',
+  'california',
+  'colorado',
+  'connecticut',
+  'delaware',
+  'florida',
+  'hawaii',
+  'idaho',
+  'illinois',
+  'indiana',
+  'iowa',
+  'kansas',
+  'kentucky',
+  'louisiana',
+  'maine',
+  'maryland',
+  'massachusetts',
+  'michigan',
+  'minnesota',
+  'mississippi',
+  'missouri',
+  'montana',
+  'nebraska',
+  'nevada',
+  'new hampshire',
+  'new jersey',
+  'new mexico',
+  'new york',
+  'north carolina',
+  'north dakota',
+  'ohio',
+  'oklahoma',
+  'oregon',
+  'pennsylvania',
+  'rhode island',
+  'south carolina',
+  'south dakota',
+  'tennessee',
+  'texas',
+  'utah',
+  'vermont',
+  'virginia',
+  'west virginia',
+  'wisconsin',
+  'wyoming',
+  'district of columbia',
+])
+
+/** Common alternate spellings → a name the country package recognizes. */
+export const COUNTRY_NAME_INPUT_ALIASES = new Map<string, string>([
+  ['brasil', 'Brazil'],
+  ['turkiye', 'Türkiye'],
+  ['viet nam', 'Vietnam'],
+  ['korea', 'South Korea'],
+  ['england', 'United Kingdom'],
+  ['scotland', 'United Kingdom'],
+  ['wales', 'United Kingdom'],
+])
+
+/** Tokens that should never resolve to a country. */
+export const JUNK_LOCATION_TOKENS = new Set([
+  'earth',
+  'mars',
+  'moon',
+  'remote',
+  'worldwide',
+  'world',
+  'global',
+  'internet',
+  'somewhere',
+  'nowhere',
+  'unknown',
+  'n/a',
+  'na',
+  'null',
+  'undefined',
+  'home',
+  'here',
+  'africa',
+  'europe',
+  'asia',
+  'antarctica',
+])
+
+/** Names that could mean more than one place — skip instead of guessing. */
+export const AMBIGUOUS_LOCATION_TOKENS = new Set(['georgia', 'washington'])
+
+/** Preferred display names for a few ISO codes with awkward official forms. */
+export const COUNTRY_DISPLAY_NAME_BY_ALPHA2 = new Map<string, string>([
+  ['US', 'United States'],
+  ['GB', 'United Kingdom'],
+  ['CN', 'China'],
+  ['TW', 'Taiwan'],
+  ['RU', 'Russia'],
+  ['TR', 'Turkey'],
+  ['KR', 'South Korea'],
+  ['NL', 'Netherlands'],
+  ['CZ', 'Czech Republic'],
+  ['AE', 'United Arab Emirates'],
+])