sitescripts/extensions/pad/validation.py - Issue 29337957: Issue 3759 - Removed PAD generation code

Side by Side Diff: sitescripts/extensions/pad/validation.py

Issue 29337957: Issue 3759 - Removed PAD generation code (Closed)

Patch Set: Created March 8, 2016, 4:31 p.m.

Left:
Right:

Use n/p to move between diff chunks; N/P to move between comments.

Jump to:

View unified diff | Download patch

OLD	NEW
	(Empty)
1 # This file is part of the Adblock Plus web scripts,

2 # Copyright (C) 2006-2016 Eyeo GmbH

3 #

4 # Adblock Plus is free software: you can redistribute it and/or modify

5 # it under the terms of the GNU General Public License version 3 as

6 # published by the Free Software Foundation.

7 #

8 # Adblock Plus is distributed in the hope that it will be useful,

9 # but WITHOUT ANY WARRANTY; without even the implied warranty of

10 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

11 # GNU General Public License for more details.

12 #

13 # You should have received a copy of the GNU General Public License

14 # along with Adblock Plus. If not, see <http://www.gnu.org/licenses/>.

15

16 import itertools

17 import warnings

18 import re

19 import urllib2

20 from xml.dom import minidom

21 from sitescripts.extensions.pad.language import get_pad_languages

22

23 FIELDS = [

24 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'MASTER_PAD_VERSION'], r'^4\.0\Z' ),

25 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'MASTER_PAD_EDITOR'], r'^[^<\x09] {0,100}\Z'),

26 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'MASTER_PAD_INFO'], r'^[^<\x09]{0 ,1000}\Z'),

27 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'CERTIFIED'], r'^[YyNn]\Z'),

28 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'CERTIFICATE_ID'], r'^(crt\-[0-9A -Z]{12}\|)\Z'),

29 (['XML_DIZ_INFO', 'MASTER_PAD_VERSION_INFO', 'CERTIFICATE_LICENSE'], r'^(http\ :\/\/repository\.appvisor\.com\/crt\-[0-9a-z]{12}\|)\Z'),

30 (['XML_DIZ_INFO', 'Company_Info', 'PublisherID'], r'^(pid-[0-9a-z]{12}\|)\Z'),

31 (['XML_DIZ_INFO', 'Company_Info', 'Company_Name'], r'^[^<\x09]{2,40}\Z'),

32 (['XML_DIZ_INFO', 'Company_Info', 'Address_1'], r'^[a-zA-Z0-9\xbc-\xff .\-,#\/ \x27]{0,40}\Z'),

33 (['XML_DIZ_INFO', 'Company_Info', 'Address_2'], r'^[a-zA-Z\xbc-\xff0-9 .\-,#\/ \x27]{0,40}\Z'),

34 (['XML_DIZ_INFO', 'Company_Info', 'City_Town'], r'^[a-zA-Z\xbc-\xff0-9 .\-,#\/ \x27]{2,40}\Z'),

35 (['XML_DIZ_INFO', 'Company_Info', 'State_Province'], r'^[a-zA-Z\xbc-\xff0-9 .\ -,\/]{0,30}\Z'),

36 (['XML_DIZ_INFO', 'Company_Info', 'Zip_Postal_Code'], r'^[^<\x09]{0,20}\Z'),

37 (['XML_DIZ_INFO', 'Company_Info', 'Country'], r'^[a-z A-Z\xbc-\xff\x27-]{2,40} \Z'),

38 (['XML_DIZ_INFO', 'Company_Info', 'Company_WebSite_URL'], r'^(http\|https):\/\/ .{2,120}\Z'),

39 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Author_First_Name'], r'^[^< \x09]{2,30}\Z'),

40 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Author_Last_Name'], r'^[^<\ x09]{2,30}\Z'),

41 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Author_Email'], r'^.{2,30}\ @.{2,63}\..{2,20}\Z'),

42 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Contact_First_Name'], r'^[^ <\x09]{2,30}\Z'),

43 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Contact_Last_Name'], r'^[^< \x09]{2,30}\Z'),

44 (['XML_DIZ_INFO', 'Company_Info', 'Contact_Info', 'Contact_Email'], r'^.{2,30} \@.{2,63}\..{2,20}\Z'),

45 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'Sales_Email'], r'^.{2,30}\@ .{2,63}\..{2,20}\Z'),

46 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'Support_Email'], r'^.{2,30} \@.{2,63}\..{2,20}\Z'),

47 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'General_Email'], r'^.{2,30} \@.{2,63}\..{2,20}\Z'),

48 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'Sales_Phone'], r'^\+{0,2}(( [0-9#()-\/_] ){7,40})?\Z'),

49 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'Support_Phone'], r'^\+{0,2} (([0-9#()-\/_] ){7,40})?\Z'),

50 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'General_Phone'], r'^\+{0,2} (([0-9#()-\/_] ){7,40})?\Z'),

51 (['XML_DIZ_INFO', 'Company_Info', 'Support_Info', 'Fax_Phone'], r'^\+{0,2}(([0 -9#()-\/_] ){7,40})?\Z'),

52 (['XML_DIZ_INFO', 'Company_Info', 'GooglePlusPage'], r'^((http\|https):\/\/.{4, 120}\|)\Z'),

53 (['XML_DIZ_INFO', 'Company_Info', 'LinkedinPage'], r'^((http\|https):\/\/.{4,12 0}\|)\Z'),

54 (['XML_DIZ_INFO', 'Company_Info', 'TwitterCompanyPage'], r'^((http\|https):\/\/ .{4,120}\|)\Z'),

55 (['XML_DIZ_INFO', 'Company_Info', 'FacebookCompanyPage'], r'^((http\|https):\/\ /.{4,120}\|)\Z'),

56 (['XML_DIZ_INFO', 'Company_Info', 'CompanyStorePage'], r'^((http\|https):\/\/.{ 4,120}\|)\Z'),

57 (['XML_DIZ_INFO', 'Program_Info', 'AppID'], r'^(app-[0-9a-z]{12}\|)\Z'),

58 (['XML_DIZ_INFO', 'Program_Info', 'Program_Name'], r'^[^<\x09]{1,40}\Z'),

59 (['XML_DIZ_INFO', 'Program_Info', 'Program_Version'], r'^[a-zA-Z0-9_.\-]{1,15} \Z'),

60 (['XML_DIZ_INFO', 'Program_Info', 'Program_Release_Month'], r'^(0[1-9]\|1[0-2]) \Z'),

61 (['XML_DIZ_INFO', 'Program_Info', 'Program_Release_Day'], r'^(0[1-9]\|[12][0-9] \|3[01])\Z'),

62 (['XML_DIZ_INFO', 'Program_Info', 'Program_Release_Year'], r'^(19\|20\|21)[0-9]{ 2}\Z'),

63 (['XML_DIZ_INFO', 'Program_Info', 'Program_Cost_Dollars'], r'^([0-9]+(\.[0-9]{ 2})?)?\Z'),

64 (['XML_DIZ_INFO', 'Program_Info', 'Program_Cost_Other_Code'], r'^(AED\|AFN\|ALL\| AMD\|ANG\|AOA\|ARS\|AUD\|AWG\|AZM\|BAM\|BBD\|BDT\|BGN\|BHD\|BIF\|BMD\|BND\|BOB\|BRL\|BSD\|BTN\|BWP\| BYR\|BZD\|CAD\|CDF\|CHF\|CLP\|CNY\|COP\|COU\|CRC\|CSD\|CZK\|CUP\|CVE\|CYP\|DJF\|DKK\|DOP\|DZD\|EEK\| EGP\|ERN\|ETB\|EUR\|FJD\|FKP\|GBP\|GEL\|GHC\|GIP\|GMD\|GNF\|GTQ\|GYD\|HKD\|HNL\|HRK\|HTG\|HUF\|IDR\| ILS\|INR\|IQD\|IRR\|ISK\|JMD\|JOD\|JPY\|KES\|KGS\|KHR\|KMF\|KPW\|KRW\|KWD\|KYD\|KZT\|LAK\|LBP\|LKR\| LRD\|LSL\|LTL\|LVL\|LYD\|MAD\|MDL\|MGA\|MKD\|MMK\|MNT\|MOP\|MRO\|MTL\|MUR\|MVR\|MWK\|MXN\|MYR\|MZN\| NAD\|NGN\|NIO\|NOK\|NPR\|NZD\|OMR\|PAB\|PEN\|PGK\|PHP\|PKR\|PLN\|PYG\|QAR\|RON\|RUB\|RWF\|SAR\|SBD\| SCR\|SDD\|SEK\|SGD\|SHP\|SIT\|SKK\|SLL\|SOS\|SRD\|STD\|SYP\|SZL\|THB\|TJS\|TMM\|TND\|TOP\|TRY\|TTD\| TWD\|TZS\|UAH\|UGX\|USD\|UYU\|UZS\|VEB\|VND\|VUV\|WST\|XAF\|XCD\|XOF\|XPF\|YER\|ZAR\|ZMK\|ZWD)?\Z' ),

65 (['XML_DIZ_INFO', 'Program_Info', 'Program_Cost_Other'], r'^([0-9]+(\.[0-9]{2} )?)?\Z'),

66 (['XML_DIZ_INFO', 'Program_Info', 'Program_Type'], r'^(Shareware\|Freeware\|Adwa re\|Demo\|Commercial\|Data Only)\Z'),

67 (['XML_DIZ_INFO', 'Program_Info', 'Program_Release_Status'], r'^(Major Update\| Minor Update\|New Release\|Beta\|Alpha\|Media Only)\Z'),

68 (['XML_DIZ_INFO', 'Program_Info', 'Program_Install_Support'], r'^(Install and Uninstall\|Install Only\|No Install Support\|Uninstall Only)\Z'),

69 (['XML_DIZ_INFO', 'Program_Info', 'Program_OS_Support'], r'^((Android\|BlackBer ry\|Handheld\/Mobile Other\|iPhone\|iPad\|iPod\|iTouch\|Java\|Linux\|Linux Console\|Linux Gnome\|Linux GPL\|Linux Open Source\|Mac OS X\|Mac Other\|MS-DOS\|Netware\|OpenVMS\|Pal m\|Pocket PC\|Symbian\|Unix\|Win2000\|Win7 x32\|Win7 x64\|Win98\|WinMobile\|WinOther\|WinS erver\|WinVista\|WinVista x64\|WinXP\|Windows 8\|Windows Phone 7\|Windows Phone 8\|Wind ows RT\|Other\|Not Applicable)[, ]*)+\Z'),

70 (['XML_DIZ_INFO', 'Program_Info', 'Program_Language'], r'^(%s\|Other\|,)+\Z' % ' \|'.join(get_pad_languages())),

71 (['XML_DIZ_INFO', 'Program_Info', 'File_Info', 'File_Size_Bytes'], r'^[0-9]{3, 16}\Z'),

72 (['XML_DIZ_INFO', 'Program_Info', 'File_Info', 'File_Size_K'], r'^[0-9.]{1,12} \Z'),

73 (['XML_DIZ_INFO', 'Program_Info', 'File_Info', 'File_Size_MB'], r'^[0-9.]{1,8} \Z'),

74 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Has_Expire_Info'], r'^[YyNn] \Z'),

75 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Count'], r'^[0-9]{0,1 5}\Z'),

76 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Based_On'], r'^(Days\| Uses\|Either\/Or)?\Z'),

77 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Other_Info'], r'^[^<\ x09]{0,100}\Z'),

78 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Month'], r'^(0[1-9]\|1 [0-2])?\Z'),

79 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Day'], r'^(0[1-9]\|[12 ][0-9]\|3[01])?\Z'),

80 (['XML_DIZ_INFO', 'Program_Info', 'Expire_Info', 'Expire_Year'], r'^((19\|20\|21 )[0-9]{2})?\Z'),

81 (['XML_DIZ_INFO', 'Program_Info', 'Program_Change_Info'], r'^[^<\x09]{0,300}\Z '),

82 (['XML_DIZ_INFO', 'Program_Info', 'Program_Category_Class'], r'^(Audio & M ultimedia::Audio Encoders\/Decoders\|Audio & Multimedia::Audio File Players\|A udio & Multimedia::Audio File Recorders\|Audio & Multimedia::CD Burners\|A udio & Multimedia::CD Players\|Audio & Multimedia::Multimedia Creation To ols\|Audio & Multimedia::Music Composers\|Audio & Multimedia::Other\|Audio & Multimedia::Presentation Tools\|Audio & Multimedia::Rippers & Conve rters\|Audio & Multimedia::Speech\|Audio & Multimedia::Video Tools\|Busines s::Accounting & Finance\|Business::Calculators & Converters\|Business::Dat abases & Tools\|Business::Helpdesk & Remote PC\|Business::Inventory & Barcoding\|Business::Investment Tools\|Business::Math & Scientific Tools\|Busin ess::Office Suites & Tools\|Business::Other\|Business::PIMS & Calendars\|Bu siness::Project Management\|Business::Vertical Market Apps\|Communications::Chat & amp; Instant Messaging\|Communications::Dial Up & Connection Tools\|Communicat ions::E-Mail Clients\|Communications::E-Mail List Management\|Communications::Fax Tools\|Communications::Newsgroup Clients\|Communications::Other Comms Tools\|Commun ications::Other E-Mail Tools\|Communications::Pager Tools\|Communications::Telepho ny\|Communications::Web\/Video Cams\|Desktop::Clocks & Alarms\|Desktop::Cursors & Fonts\|Desktop::Icons\|Desktop::Other\|Desktop::Screen Savers: Art\|Desktop:: Screen Savers: Cartoons\|Desktop::Screen Savers: Nature\|Desktop::Screen Savers: O ther\|Desktop::Screen Savers: People\|Desktop::Screen Savers: Science\|Desktop::Scr een Savers: Seasonal\|Desktop::Screen Savers: Vehicles\|Desktop::Themes & Wall paper\|Development::Active X\|Development::Basic, VB, VB DotNet\|Development::C \/ C\+\+ \/ C\#\|Development::Compilers & Interpreters\|Development::Components & amp; Libraries\|Development::Debugging\|Development::Delphi\|Development::Help Tool s\|Development::Install & Setup\|Development::Management & Distribution\|De velopment::Other\|Development::Source Editors\|Education::Computer\|Education::Dict ionaries\|Education::Geography\|Education::Kids\|Education::Languages\|Education::Ma thematics\|Education::Other\|Education::Reference Tools\|Education::Science\|Educati on::Teaching & Training Tools\|Games & Entertainment::Action\|Games & Entertainment::Adventure & Roleplay\|Games & Entertainment::Arcade\|Games & Entertainment::Board\|Games & Entertainment::Card\|Games & Entertain ment::Casino & Gambling\|Games & Entertainment::Kids\|Games & Entertai nment::Online Gaming\|Games & Entertainment::Other\|Games & Entertainment: :Puzzle & Word Games\|Games & Entertainment::Simulation\|Games & Enter tainment::Sports\|Games & Entertainment::Strategy & War Games\|Games & Entertainment::Tools & Editors\|Graphic Apps::Animation Tools\|Graphic Apps:: CAD\|Graphic Apps::Converters & Optimizers\|Graphic Apps::Editors\|Graphic Apps ::Font Tools\|Graphic Apps::Gallery & Cataloging Tools\|Graphic Apps::Icon Too ls\|Graphic Apps::Other\|Graphic Apps::Screen Capture\|Graphic Apps::Viewers\|Home & amp; Hobby::Astrology\/Biorhythms\/Mystic\|Home & Hobby::Astronomy\|Home & Hobby::Cataloging\|Home & Hobby::Food & Drink\|Home & Hobby::Genealog y\|Home & Hobby::Health & Nutrition\|Home & Hobby::Other\|Home & Ho bby::Personal Finance\|Home & Hobby::Personal Interest\|Home & Hobby::Recr eation\|Home & Hobby::Religion\|Network & Internet::Ad Blockers\|Network &a mp; Internet::Browser Tools\|Network & Internet::Browsers\|Network & Inter net::Download Managers\|Network & Internet::File Sharing\/Peer to Peer\|Networ k & Internet::FTP Clients\|Network & Internet::Network Monitoring\|Network & Internet::Other\|Network & Internet::Remote Computing\|Network & In ternet::Search\/Lookup Tools\|Network & Internet::Terminal & Telnet Clien ts\|Network & Internet::Timers & Time Synch\|Network & Internet::Trace & Ping Tools\|Security & Privacy::Access Control\|Security & Privacy: :Anti-Spam & Anti-Spy Tools\|Security & Privacy::Anti-Virus Tools\|Securit y & Privacy::Covert Surveillance\|Security & Privacy::Encryption Tools\|Se curity & Privacy::Other\|Security & Privacy::Password Managers\|Servers::F irewall & Proxy Servers\|Servers::FTP Servers\|Servers::Mail Servers\|Servers:: News Servers\|Servers::Other Server Applications\|Servers::Telnet Servers\|Servers: :Web Servers\|System Utilities::Automation Tools\|System Utilities::Backup & R estore\|System Utilities::Benchmarking\|System Utilities::Clipboard Tools\|System U tilities::File & Disk Management\|System Utilities::File Compression\|System U tilities::Launchers & Task Managers\|System Utilities::Other\|System Utilities ::Printer\|System Utilities::Registry Tools\|System Utilities::Shell Tools\|System Utilities::System Maintenance\|System Utilities::Text\/Document Editors\|Web Devel opment::ASP & PHP\|Web Development::E-Commerce\|Web Development::Flash Tools\|W eb Development::HTML Tools\|Web Development::Java & JavaScript\|Web Developmen t::Log Analysers\|Web Development::Other\|Web Development::Site Administration\|Web Development::Wizards & Components\|Web Development::XML\/CSS Tools)\Z'),

83 (['XML_DIZ_INFO', 'Program_Info', 'Program_Specific_Category'], r'^(Audio\|Busi ness\|Development Tools\|Education\|Games\|Graphics\|Home\/Hobby\|Internet\|Miscellaneo us\|Screen Savers\|Utilities)?\Z'),

84 (['XML_DIZ_INFO', 'Program_Info', 'Program_System_Requirements'], r'^[^<\x09]{ 0,100}\Z'),

85 (['XML_DIZ_INFO', 'Program_Info', 'FacebookProductPage'], r'^((http\|https):\/\ /.{4,120}\|)\Z'),

86 (['XML_DIZ_INFO', 'Program_Info', 'GooglePlusProductPage'], r'^((http\|https):\ /\/.{4,120}\|)\Z'),

87 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Application_Info_URL'], r'^ (http\|https):\/\/.{2,120}\Z'),

88 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Application_Order_URL'], r' ^((http\|https):\/\/.{2,120})?\Z'),

89 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Application_Screenshot_URL' ], r'^(http\|https):\/\/.{2,120}\.(gif\|jpg\|png)\Z'),

90 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Application_Icon_URL'], r'^ (http\|https):\/\/.{2,120}\.(gif\|jpg\|png)\Z'),

91 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Application_XML_File_URL'], r'^(http\|https):\/\/.{2,120}\.(xml\|cgi\|php\|asp)\Z'),

92 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Video_Link_1_URL'], r'^((ht tp\|https):\/\/.{4,120}\|)\Z'),

93 (['XML_DIZ_INFO', 'Web_Info', 'Application_URLs', 'Video_Link_2_URL'], r'^((ht tp\|https):\/\/.{4,120}\|)\Z'),

94 (['XML_DIZ_INFO', 'Web_Info', 'Download_URLs', 'Primary_Download_URL'], r'^(ht tp\|https\|ftp):\/\/.{2,120}\Z'),

95 (['XML_DIZ_INFO', 'Web_Info', 'Download_URLs', 'Secondary_Download_URL'], r'^( (http\|https\|ftp):\/\/.{2,120})?\Z'),

96 (['XML_DIZ_INFO', 'Web_Info', 'Download_URLs', 'Additional_Download_URL_1'], r '^((http\|https\|ftp):\/\/.{2,120})?\Z'),

97 (['XML_DIZ_INFO', 'Web_Info', 'Download_URLs', 'Additional_Download_URL_2'], r '^((http\|https\|ftp):\/\/.{2,120})?\Z'),

98 (['XML_DIZ_INFO', 'Permissions', 'Distribution_Permissions'], r'^[^<]{0,2000}\ Z'),

99 (['XML_DIZ_INFO', 'Permissions', 'EULA'], r'^[^<]{0,20000}\Z'),

100 (['XML_DIZ_INFO', 'Press_Release', 'Keywords'], r'^[^<\x09\x0a\x0d]{0,250}\Z') ,

101 (['XML_DIZ_INFO', 'Press_Release', 'Headline'], r'^([^<\x09\x0a\x0d]{20,100}\|) \Z'),

102 (['XML_DIZ_INFO', 'Press_Release', 'Summary'], r'^([^<\x09\x0a\x0d]{20,250}\|)\ Z'),

103 (['XML_DIZ_INFO', 'Press_Release', 'Press_Release'], r'^[^<\x09\x0a\x0d]{0,300 0}\Z'),

104 (['XML_DIZ_INFO', 'Press_Release', 'Press_Release_Plain'], r'^[^<\x09\x0a\x0d] {0,3000}\Z'),

105 (['XML_DIZ_INFO', 'Press_Release', 'Related_URL'], r'^((http\|https):\/\/.{0,10 0}\|)\Z'),

106 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Feed_URL'], r'^((http\|https):\/\/.{0,1 00}\|)\Z'),

107 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Type'], r'^(RSS 0\.90\|RSS 0\.91\|RSS 0\ .92\|RSS 0\.93\|RSS 0\.94\|RSS 1\.0\|RSS 2\.0\|Atom 0\.3\|Atom 1\.0\|)\Z'),

108 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Title'], r'^[^<\x09\x0a\x0d]{0,60}\Z') ,

109 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Keywords'], r'^[^<\x09]{0,250}\Z'),

110 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Description_70'], r'^[^<\x09\x0a\x0d]{ 0,70}\Z'),

111 (['XML_DIZ_INFO', 'NewsFeed', 'NewsFeed_Description_250'], r'^[^<\x09\x0a\x0d] {0,250}\Z'),

112 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_ShareIt_Order_Page'], None),

113 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_ShareIt_Vendor_ID'], None),

114 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_ShareIt_Product_ID'], None),

115 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_ShareIt_Maximum_Commission_Rate'], None),

116 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_PayPro_Order_Page'], None),

117 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_PayPro_Vendor_ID'], None),

118 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_PayPro_Product_ID'], None),

119 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_PayPro_Maximum_Commission_Rate'], None),

120 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_Avangate_Order_Page'], None),

121 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_Avangate_Vendor_ID'], None),

122 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_Avangate_Product_ID'], None),

123 (['XML_DIZ_INFO', 'Affiliates', 'Affiliates_Avangate_Maximum_Commission_Rate'] , None),

124 (['XML_DIZ_INFO', 'ASP', 'ASP_Member'], r'^[YyNn]\Z'),

125 (['XML_DIZ_INFO', 'ASP', 'ASP_Member_Number'], None),

126 ]

127

128 for language in get_pad_languages():

129 FIELDS.extend([

130 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Keywords'], r'^[^<\x09] {0,250}\Z'),

131 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Char_Desc_45'], r'^[^<\ x09\x0a\x0d]{0,45}\Z'),

132 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Char_Desc_80'], r'^[^<\ x09\x0a\x0d]{0,80}\Z'),

133 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Char_Desc_250'], r'^[^< \x09\x0a\x0d]{0,250}\Z'),

134 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Char_Desc_450'], r'^[^< \x09\x0a\x0d]{0,450}\Z'),

135 (['XML_DIZ_INFO', 'Program_Descriptions', language, 'Char_Desc_2000'], r'^[^ <]{0,2000}\Z'),

136 ])

137

138 def validate_fields(fields, nodes, filename):

139 expected_nodes = set()

140

141 for node_name, fields in itertools.groupby(fields, lambda (path, regex): path[ 0]):

142 expected_nodes.add(node_name)

143

144 regex = None

145 nested_fields = []

146 for path, regex_ in fields:

147 if path == [node_name]:

148 regex = regex_

149 else:

150 nested_fields.append((path[1:], regex_))

151

152 found = False

153 for node in nodes:

154 if node.nodeName == node_name:

155 if found:

156 warnings.warn('invalid PAD file (duplicate node)\n'

157 'filename: %s\n'

158 'node: %s' % (filename, node_name))

159

160 if regex:

161 value = ''.join(child.toxml() for child in node.childNodes)

162 if not re.match(regex, value):

163 warnings.warn('invalid PAD file (invalid value)\n'

164 'filename: %s\n'

165 'node: %s\n'

166 'value: %s\n'

167 'regex: %s' % (filename, node_name, value, regex))

168

169 if nested_fields:

170 validate_fields(nested_fields, node.childNodes, filename)

171

172 found = True

173

174 if not found:

175 if regex and not re.match(regex, ''):

176 warnings.warn('invalid PAD file (missing node)\n'

177 'filename: %s\n'

178 'node: %s' % (filename, node_name))

179

180 validate_fields(nested_fields, [], filename)

181

182 for node in nodes:

183 if node.nodeType == node.COMMENT_NODE:

184 continue

185 if node.nodeType == node.TEXT_NODE and node.nodeValue.strip() == '':

186 continue

187

188 if node.nodeName not in expected_nodes:

189 warnings.warn('invalid PAD file (unexpected node)\n'

190 'filename: %s\n'

191 'node: %s' % (filename, node.nodeName))

192

193 def validate_pad(pad, filename):

194 validate_fields(FIELDS, minidom.parseString(pad).childNodes, filename)

195

196 def print_fields():

197 doc = minidom.parse(urllib2.urlopen('http://repository.appvisor.com/padspec/fi les/padspec40.xml'))

198

199 print '['

200 for field in doc.getElementsByTagName('Field'):

201 path, regex = [

202 ''.join(node.nodeValue for node in field.getElementsByTagName(name)[0].chi ldNodes)

203 for name in ('Path', 'RegEx')

204 ]

205 print ' (%r, %s),' % (str(path).split('/'), "r'%s'" % regex if regex else ' None')

206 print ']'

207

208 if __name__ == '__main__':

209 print_fields()

OLD	NEW