Fix NeuStar regex to make State/Province optional since these are apparently not always present, and explicitly remove variable-width whitespace matching.

pythonwhois/parse.py

@@ -518,7 +518,7 @@ def parse_registrants(data):
 		"Holder of domain name:\n(?P<name>[\S\s]+)\n(?P<street>.+)\n(?P<postalcode>[A-Z0-9-]+)\s+(?P<city>.+)\n(?P<country>.+)\nContractual Language", # nic.ch
 		"\n\n(?:Owner)?\s+: (?P<name>.*)\n(?:\s+: (?P<organization>.*)\n)?\s+: (?P<street>.*)\n\s+: (?P<city>.*)\n\s+: (?P<state>.*)\n\s+: (?P<country>.*)\n", # nic.io
 		"Contact Information:\n\[Name\]\s*(?P<name>.*)\n\[Email\]\s*(?P<email>.*)\n\[Web Page\]\s*(?P<url>.*)\n\[Postal code\]\s*(?P<postalcode>.*)\n\[Postal Address\]\s*(?P<street1>.*)\n(?:\s+(?P<street2>.*)\n)?(?:\s+(?P<street3>.*)\n)?\[Phone\]\s*(?P<phone>.*)\n\[Fax\]\s*(?P<fax>.*)\n", # jprs.jp
-		"Registrant ID:[ ]*(?P<handle>.*)\nRegistrant Name:[ ]*(?P<name>.*)\n(?:Registrant Organization:[ ]*(?P<organization>.*)\n)?Registrant Address1:[ ]*(?P<street1>.*)\n(?:Registrant Address2:[ ]*(?P<street2>.*)\n)?(?:Registrant Address3:[ ]*(?P<street3>.*)\n)?Registrant City:[ ]*(?P<city>.*)\nRegistrant State/Province:[ ]*(?P<state>.*)\nRegistrant Postal Code:[ ]*(?P<postalcode>.*)\nRegistrant Country:[ ]*(?P<country>.*)\nRegistrant Country Code:[ ]*.*\nRegistrant Phone Number:[ ]*(?P<phone>.*)\n(?:Registrant Facsimile Number:[ ]*(?P<facsimile>.*)\n)?Registrant Email:[ ]*(?P<email>.*)", # .US (NeuStar)
+		"Registrant ID:(?P<handle>.*)\nRegistrant Name:(?P<name>.*)\n(?:Registrant Organization:(?P<organization>.*)\n)?Registrant Address1:(?P<street1>.*)\n(?:Registrant Address2:(?P<street2>.*)\n)?(?:Registrant Address3:(?P<street3>.*)\n)?Registrant City:(?P<city>.*)\n(?:Registrant State/Province:(?P<state>.*)\n)?Registrant Postal Code:(?P<postalcode>.*)\nRegistrant Country:(?P<country>.*)\nRegistrant Country Code:.*\nRegistrant Phone Number:(?P<phone>.*)\n(?:Registrant Facsimile Number:(?P<facsimile>.*)\n)?Registrant Email:(?P<email>.*)", # .US, .biz (NeuStar)
 		"  Organisation Name[.]* (?P<name>.*)\n  Organisation Address[.]* (?P<street1>.*)\n  Organisation Address[.]* (?P<street2>.*)\n(?:  Organisation Address[.]* (?P<street3>.*)\n)?  Organisation Address[.]* (?P<city>.*)\n  Organisation Address[.]* (?P<postalcode>.*)\n  Organisation Address[.]* (?P<state>.*)\n  Organisation Address[.]* (?P<country>.*)", # Melbourne IT (what a horrid format...)
 		"Registrant:[ ]*(?P<name>.+)\n[\s\S]*Eligibility Name:[ ]*(?P<organization>.+)\n[\s\S]*Registrant Contact ID:[ ]*(?P<handle>.+)\n", # .au business
 		"Eligibility Type:[ ]*Citizen\/Resident\n[\s\S]*Registrant Contact ID:[ ]*(?P<handle>.+)\n[\s\S]*Registrant Contact Name:[ ]*(?P<name>.+)\n", # .au individual
@@ -550,7 +550,7 @@ def parse_registrants(data):
 		"Technical Contact:\n tech_org: (?P<organization>.*)\n tech_name: (?P<name>.*)\n tech_email: (?P<email>.*)\n tech_address: (?P<address>.*)\n tech_city: (?P<city>.*)\n tech_state: (?P<state>.*)\n tech_zip: (?P<postalcode>.*)\n tech_country: (?P<country>.*)\n tech_phone: (?P<phone>.*)", # Bellnames
 		"Technical contact:\n(?P<name>[\S\s]+)\n(?P<street>.+)\n(?P<postalcode>[A-Z0-9-]+)\s+(?P<city>.+)\n(?P<country>.+)\n\n", # nic.ch
 		"Tech Contact ID:[ ]*(?P<handle>.+)\nTech Contact Name:[ ]*(?P<name>.+)", # .au
-		"Technical Contact ID:[ ]*(?P<handle>.*)\nTechnical Contact Name:[ ]*(?P<name>.*)\n(?:Technical Contact Organization:[ ]*(?P<organization>.*)\n)?Technical Contact Address1:[ ]*(?P<street1>.*)\n(?:Technical Contact Address2:[ ]*(?P<street2>.*)\n)?(?:Technical Contact Address3:[ ]*(?P<street3>.*)\n)?Technical Contact City:[ ]*(?P<city>.*)\nTechnical Contact State/Province:[ ]*(?P<state>.*)\nTechnical Contact Postal Code:[ ]*(?P<postalcode>.*)\nTechnical Contact Country:[ ]*(?P<country>.*)\nTechnical Contact Country Code:[ ]*.*\nTechnical Contact Phone Number:[ ]*(?P<phone>.*)\n(?:Technical Contact Facsimile Number:[ ]*(?P<facsimile>.*)\n)?Technical Contact Email:[ ]*(?P<email>.*)", # .US (NeuStar)
+		"Technical Contact ID:(?P<handle>.*)\nTechnical Contact Name:(?P<name>.*)\n(?:Technical Contact Organization:(?P<organization>.*)\n)?Technical Contact Address1:(?P<street1>.*)\n(?:Technical Contact Address2:(?P<street2>.*)\n)?(?:Technical Contact Address3:(?P<street3>.*)\n)?Technical Contact City:(?P<city>.*)\n(?:Technical Contact State/Province:(?P<state>.*)\n)?Technical Contact Postal Code:(?P<postalcode>.*)\nTechnical Contact Country:(?P<country>.*)\nTechnical Contact Country Code:.*\nTechnical Contact Phone Number:(?P<phone>.*)\n(?:Technical Contact Facsimile Number:(?P<facsimile>.*)\n)?Technical Contact Email:(?P<email>.*)", # .US, .biz (NeuStar)
 		"Tech Name[.]* (?P<name>.*)\n  Tech Address[.]* (?P<street1>.*)\n  Tech Address[.]* (?P<street2>.*)\n(?:  Tech Address[.]* (?P<street3>.*)\n)?  Tech Address[.]* (?P<city>.*)\n  Tech Address[.]* (?P<postalcode>.*)\n  Tech Address[.]* (?P<state>.*)\n  Tech Address[.]* (?P<country>.*)\n  Tech Email[.]* (?P<email>.*)\n  Tech Phone[.]* (?P<phone>.*)\n  Tech Fax[.]* (?P<fax>.*)", # Melbourne IT
 		"Technical contact:\n(?:  (?P<organization>.+)\n)?  (?P<name>.+)\n  (?P<email>.+)\n  (?P<street>.+)\n  (?P<city>.+), (?P<state>.+) (?P<postalcode>.+) (?P<country>.+)\n  Phone: (?P<phone>.*)\n  Fax: (?P<fax>.*)\n", # Fabulous.com
 		"tech-c-name:\s*(?P<name>.+)\ntech-c-type:\s*(?P<type>.+)\ntech-c-address:\s*(?P<street>.+)\ntech-c-postcode:\s*(?P<postalcode>.+)\ntech-c-city:\s*(?P<city>.+)\ntech-c-country:\s*(?P<country>.+)\n(?:tech-c-phone:\s*(?P<phone>.+)\n)?(?:tech-c-email:\s*(?P<email>.+)\n)?", # Hetzner
@@ -572,7 +572,7 @@ def parse_registrants(data):
 		" Administrative Contact Details:[ ]*\n    (?P<organization>.*)\n    (?P<name>.*)[ ]{2,}\((?P<email>.*)\)\n    (?P<street1>.*)\n(?:    (?P<street2>.*)\n)?(?:    (?P<street3>.*)\n)?    (?P<city>.*)\n    (?P<state>.*),(?P<postalcode>.*)\n    (?P<country>.*)\n    Tel. (?P<phone>.*)", # Whois.com
 		"admin-id:[ ]*(?P<handle>.*)\n(?:admin-organization:[ ]*(?P<organization>.*)\n)?admin-name:[ ]*(?P<name>.*)\nadmin-street:[ ]*(?P<street>.*)\nadmin-city:[ ]*(?P<city>.*)\nadmin-zip:[ ]*(?P<postalcode>.*)\nadmin-country:[ ]*(?P<country>.*)\n(?:admin-phone:[ ]*(?P<phone>.*)\n)?(?:admin-fax:[ ]*(?P<fax>.*)\n)?admin-email:[ ]*(?P<email>.*)", # InterNetworX
 		"Administrative Contact:\n admin_org: (?P<organization>.*)\n admin_name: (?P<name>.*)\n admin_email: (?P<email>.*)\n admin_address: (?P<address>.*)\n admin_city: (?P<city>.*)\n admin_state: (?P<state>.*)\n admin_zip: (?P<postalcode>.*)\n admin_country: (?P<country>.*)\n admin_phone: (?P<phone>.*)", # Bellnames
-		"Administrative Contact ID:[ ]*(?P<handle>.*)\nAdministrative Contact Name:[ ]*(?P<name>.*)\n(?:Administrative Contact Organization:[ ]*(?P<organization>.*)\n)?Administrative Contact Address1:[ ]*(?P<street1>.*)\n(?:Administrative Contact Address2:[ ]*(?P<street2>.*)\n)?(?:Administrative Contact Address3:[ ]*(?P<street3>.*)\n)?Administrative Contact City:[ ]*(?P<city>.*)\nAdministrative Contact State/Province:[ ]*(?P<state>.*)\nAdministrative Contact Postal Code:[ ]*(?P<postalcode>.*)\nAdministrative Contact Country:[ ]*(?P<country>.*)\nAdministrative Contact Country Code:[ ]*.*\nAdministrative Contact Phone Number:[ ]*(?P<phone>.*)\n(?:Administrative Contact Facsimile Number:[ ]*(?P<facsimile>.*)\n)?Administrative Contact Email:[ ]*(?P<email>.*)", # .US (NeuStar)
+		"Administrative Contact ID:(?P<handle>.*)\nAdministrative Contact Name:(?P<name>.*)\n(?:Administrative Contact Organization:(?P<organization>.*)\n)?Administrative Contact Address1:(?P<street1>.*)\n(?:Administrative Contact Address2:(?P<street2>.*)\n)?(?:Administrative Contact Address3:(?P<street3>.*)\n)?Administrative Contact City:(?P<city>.*)\n(?:Administrative Contact State/Province:(?P<state>.*)\n)?Administrative Contact Postal Code:(?P<postalcode>.*)\nAdministrative Contact Country:(?P<country>.*)\nAdministrative Contact Country Code:.*\nAdministrative Contact Phone Number:(?P<phone>.*)\n(?:Administrative Contact Facsimile Number:(?P<facsimile>.*)\n)?Administrative Contact Email:(?P<email>.*)", # .US, .biz (NeuStar)
 		"Admin Name[.]* (?P<name>.*)\n  Admin Address[.]* (?P<street1>.*)\n  Admin Address[.]* (?P<street2>.*)\n(?:  Admin Address[.]* (?P<street3>.*)\n)?  Admin Address[.]* (?P<city>.*)\n  Admin Address[.]* (?P<postalcode>.*)\n  Admin Address[.]* (?P<state>.*)\n  Admin Address[.]* (?P<country>.*)\n  Admin Email[.]* (?P<email>.*)\n  Admin Phone[.]* (?P<phone>.*)\n  Admin Fax[.]* (?P<fax>.*)", # Melbourne IT
 		"Administrative contact:\n(?:  (?P<organization>.+)\n)?  (?P<name>.+)\n  (?P<email>.+)\n  (?P<street>.+)\n  (?P<city>.+), (?P<state>.+) (?P<postalcode>.+) (?P<country>.+)\n  Phone: (?P<phone>.*)\n  Fax: (?P<fax>.*)\n", # Fabulous.com
 		"admin-c-name:\s*(?P<name>.+)\nadmin-c-type:\s*(?P<type>.+)\nadmin-c-address:\s*(?P<street>.+)\nadmin-c-postcode:\s*(?P<postalcode>.+)\nadmin-c-city:\s*(?P<city>.+)\nadmin-c-country:\s*(?P<country>.+)\n(?:admin-c-phone:\s*(?P<phone>.+)\n)?(?:admin-c-email:\s*(?P<email>.+)\n)?", # Hetzner
@@ -591,7 +591,7 @@ def parse_registrants(data):
 		" Billing Contact Details:[ ]*\n    (?P<organization>.*)\n    (?P<name>.*)[ ]{2,}\((?P<email>.*)\)\n    (?P<street1>.*)\n(?:    (?P<street2>.*)\n)?(?:    (?P<street3>.*)\n)?    (?P<city>.*)\n    (?P<state>.*),(?P<postalcode>.*)\n    (?P<country>.*)\n    Tel. (?P<phone>.*)", # Whois.com
 		"billing-id:[ ]*(?P<handle>.*)\n(?:billing-organization:[ ]*(?P<organization>.*)\n)?billing-name:[ ]*(?P<name>.*)\nbilling-street:[ ]*(?P<street>.*)\nbilling-city:[ ]*(?P<city>.*)\nbilling-zip:[ ]*(?P<postalcode>.*)\nbilling-country:[ ]*(?P<country>.*)\n(?:billing-phone:[ ]*(?P<phone>.*)\n)?(?:billing-fax:[ ]*(?P<fax>.*)\n)?billing-email:[ ]*(?P<email>.*)", # InterNetworX
 		"Billing Contact:\n bill_org: (?P<organization>.*)\n bill_name: (?P<name>.*)\n bill_email: (?P<email>.*)\n bill_address: (?P<address>.*)\n bill_city: (?P<city>.*)\n bill_state: (?P<state>.*)\n bill_zip: (?P<postalcode>.*)\n bill_country: (?P<country>.*)\n bill_phone: (?P<phone>.*)", # Bellnames
-		"Billing Contact ID:[ ]*(?P<handle>.*)\nBilling Contact Name:[ ]*(?P<name>.*)\n(?:Billing Contact Organization:[ ]*(?P<organization>.*)\n)?Billing Contact Address1:[ ]*(?P<street1>.*)\n(?:Billing Contact Address2:[ ]*(?P<street2>.*)\n)?(?:Billing Contact Address3:[ ]*(?P<street3>.*)\n)?Billing Contact City:[ ]*(?P<city>.*)\nBilling Contact State/Province:[ ]*(?P<state>.*)\nBilling Contact Postal Code:[ ]*(?P<postalcode>.*)\nBilling Contact Country:[ ]*(?P<country>.*)\nBilling Contact Country Code:[ ]*.*\nBilling Contact Phone Number:[ ]*(?P<phone>.*)\n(?:Billing Contact Facsimile Number:[ ]*(?P<facsimile>.*)\n)?Billing Contact Email:[ ]*(?P<email>.*)", # .US (NeuStar)
+		"Billing Contact ID:(?P<handle>.*)\nBilling Contact Name:(?P<name>.*)\n(?:Billing Contact Organization:(?P<organization>.*)\n)?Billing Contact Address1:(?P<street1>.*)\n(?:Billing Contact Address2:(?P<street2>.*)\n)?(?:Billing Contact Address3:(?P<street3>.*)\n)?Billing Contact City:(?P<city>.*)\n(?:Billing Contact State/Province:(?P<state>.*)\n)?Billing Contact Postal Code:(?P<postalcode>.*)\nBilling Contact Country:(?P<country>.*)\nBilling Contact Country Code:.*\nBilling Contact Phone Number:(?P<phone>.*)\n(?:Billing Contact Facsimile Number:(?P<facsimile>.*)\n)?Billing Contact Email:(?P<email>.*)", # .US, .biz (NeuStar)
 		"Billing contact:\n(?:  (?P<organization>.+)\n)?  (?P<name>.+)\n  (?P<email>.+)\n  (?P<street>.+)\n  (?P<city>.+), (?P<state>.+) (?P<postalcode>.+) (?P<country>.+)\n  Phone: (?P<phone>.*)\n  Fax: (?P<fax>.*)\n", # Fabulous.com
 		"Billing Contact Information :[ ]*\n[ ]+(?P<firstname>.*)\n[ ]+(?P<lastname>.*)\n[ ]+(?P<organization>.*)\n[ ]+(?P<email>.*)\n[ ]+(?P<street>.*)\n[ ]+(?P<city>.*)\n[ ]+(?P<postalcode>.*)\n[ ]+(?P<phone>.*)\n[ ]+(?P<fax>.*)\n\n", # GAL Communication
 	]

test/data/drpciv.biz

@@ -0,0 +1,82 @@
+Domain Name:                                 DRPCIV.BIZ
+Domain ID:                                   D48726051-BIZ
+Sponsoring Registrar:                        INTERNET.BS CORP.
+Sponsoring Registrar IANA ID:                814
+Registrar URL (registration services):       www.internet.bs
+Domain Status:                               clientTransferProhibited
+Registrant ID:                               INTEJ39MZQIM27GY
+Registrant Name:                             Domain Administrator
+Registrant Organization:                     Fundacion Private Whois
+Registrant Address1:                         Attn: drpciv.biz
+Registrant Address2:                         Aptds. 0850-00056
+Registrant City:                             Panama
+Registrant Postal Code:                      Zona 15
+Registrant Country:                          Panama
+Registrant Country Code:                     PA
+Registrant Phone Number:                     +507.65967959
+Registrant Email:                            522e20e76bmlbpe0@5225b4d0pi3627q9.privatewhois.net
+Administrative Contact ID:                   INTEDIIYRYASW5MQ
+Administrative Contact Name:                 Domain Administrator
+Administrative Contact Organization:         Fundacion Private Whois
+Administrative Contact Address1:             Attn: drpciv.biz
+Administrative Contact Address2:             Aptds. 0850-00056
+Administrative Contact City:                 Panama
+Administrative Contact Postal Code:          Zona 15
+Administrative Contact Country:              Panama
+Administrative Contact Country Code:         PA
+Administrative Contact Phone Number:         +507.65967959
+Administrative Contact Email:                522e20e76q1lbtbk@5225b4d0pi3627q9.privatewhois.net
+Billing Contact ID:                          INTE9CPTNSAGCZVW
+Billing Contact Name:                        Domain Administrator
+Billing Contact Organization:                Fundacion Private Whois
+Billing Contact Address1:                    Attn: drpciv.biz
+Billing Contact Address2:                    Aptds. 0850-00056
+Billing Contact City:                        Panama
+Billing Contact Postal Code:                 Zona 15
+Billing Contact Country:                     Panama
+Billing Contact Country Code:                PA
+Billing Contact Phone Number:                +507.65995877
+Billing Contact Email:                       522e20e7hvddrqc3@5225b4d0pi3627q9.privatewhois.net
+Technical Contact ID:                        INTEHQ98UIY8AG7R
+Technical Contact Name:                      Domain Administrator
+Technical Contact Organization:              Fundacion Private Whois
+Technical Contact Address1:                  Attn: drpciv.biz
+Technical Contact Address2:                  Aptds. 0850-00056
+Technical Contact City:                      Panama
+Technical Contact Postal Code:               Zona 15
+Technical Contact Country:                   Panama
+Technical Contact Country Code:              PA
+Technical Contact Phone Number:              +507.65967959
+Technical Contact Email:                     522e20e7o2mhz6ui@5225b4d0pi3627q9.privatewhois.net
+Name Server:                                 HANK.NS.CLOUDFLARE.COM
+Name Server:                                 AMY.NS.CLOUDFLARE.COM
+Created by Registrar:                        INTERNET.BS CORP.
+Last Updated by Registrar:                   INTERNET.BS CORP.
+Domain Registration Date:                    Fri Feb 03 15:34:57 GMT 2012
+Domain Expiration Date:                      Mon Feb 02 23:59:59 GMT 2015
+Domain Last Updated Date:                    Tue Jan 28 14:02:20 GMT 2014
+
+>>>> Whois database was last updated on: Wed Jun 04 22:33:43 GMT 2014 <<<<
+
+NeuStar, Inc., the Registry Operator for .BIZ, has collected this information
+for the WHOIS database through an ICANN-Accredited Registrar. This information
+is provided to you for informational purposes only and is designed to assist
+persons in determining contents of a domain name registration record in the
+NeuStar registry database. NeuStar makes this information available to you
+"as is" and does not guarantee its accuracy. By submitting a WHOIS query, you
+agree that you will use this data only for lawful purposes and that, under no
+circumstances will you use this data: (1) to allow, enable, or otherwise
+support the transmission of mass unsolicited, commercial advertising or
+solicitations via direct mail, electronic mail, or by telephone; (2) in
+contravention of any applicable data and privacy protection acts; or (3) to
+enable high volume, automated, electronic processes that apply to the registry
+(or its systems). Compilation, repackaging, dissemination, or other use of the
+WHOIS database in its entirety, or of a substantial portion thereof, is not
+allowed without NeuStar's prior written permission. NeuStar reserves the
+right to modify or change these conditions at any time without prior or
+subsequent notification of any kind. By executing this query, in any manner
+whatsoever, you agree to abide by these terms.
+
+NOTE: FAILURE TO LOCATE A RECORD IN THE WHOIS DATABASE IS NOT INDICATIVE
+OF THE AVAILABILITY OF A DOMAIN NAME.
+