Squeak-dev September 2017

squeak-dev@lists.squeakfoundation.org

30 participants
85 discussions

Retro conversion question
by Casimiro de Almeida Barreto 13 Sep '17

13 Sep '17

Hello, Perhaps someone can help me with this problem: I'm saving a key pair in a mariadb/mysql table as follows: str := 'INSERT INTO chaves_de_encriptacao (numero_funcional, chave_privada, chave_publica) '. str := str,'VALUES ('''. str := str, self personId leaId asString, ''', '''. str := str, self keyPair privateKey asAsn1DerBytes asByteArray hex, ''', '''. str := str, self keyPair publicKey asAsn1DerBytes asByteArray hex,''')'. [ dbConnection execute: str ] on: Exception do: [ ^ false ]. And I'd like to recover the keys (SELECT * FROM chaves_de_encriptacao WHERE numero_funcional = 1010) and then reconvert in order to recover the original keys. Obviously I can do that by writting a parser that will get the hex and bring it back to binary but I'd like to avoid it. Question is: is there a reverse for asByteArray hex?

3 2

Ubuntu script to ask image file?
by Alan Pinch 13 Sep '17

13 Sep '17

I am looking for a squeak startup she'll script on Ubuntu that would open an OS file prompt for the image file to load. I had seen one, at one time, but the ocean is deep. Thank you. - Alan

3 4

Here is a tiny proposal to make Smalltalk more functional
by Kjell Godo 13 Sep '17

13 Sep '17

Here is a proposal to make Smalltalk more functional. If a Block B is never assigned to any variable in a Method M and B is not returned by M and B is evaluated by M then B should be inlined by the Compiler as a way of allocating and assigning immutable local variables which can be part of a more functional local style of Smalltalk programming. Functional is good. Look at Haskell Scala Java F# Ocaml lisp etc which are all going functional. Look at category theory which is the theory of function composition and function combination. It is good to base programming on math. Look at Smalltalk Traits which are state free. Smalltalk can be more functional. No value Activation should be done by ( B value.. ) in M at runtime but only immutable local variable assignments followed by immutable variable accesses. Here is an example of a local tail recursive function using only immutable variables : [ :recurse | recurse value: input value: recurse ] value:[ :input1 :recurse | ( input1 message1 ) ifTrue:[ input1 message2 ] ifFalse:[ recurse value:( input1 message3 ) value: recurse ] ] Category Theory for programmers by Bartosz Milewski https://www.youtube.com/watch?v=I8LbkfSSR58&list=PLbgaMIhjbmEnaH_LTkxLI7FMa… Thank you Kjell E Godø

1 0

Seaside question won't post to Seaside news group
by Max Leske 11 Sep '17

11 Sep '17

Forwarded to seaside(a)lists.squeakfoundation.org. On 9 September 2017 at 14:00:04, squeak-dev-request(a)lists.squeakfoundation.org ( squeak-dev-request(a)lists.squeakfoundation.org) wrote: Date: Fri, 08 Sep 2017 11:30:25 -0400 <http://airmail.calendar/2017-09-08 17:30:25 GMT+2> From: Louis LaBrunda <Lou(a)Keystone-Software.com> To: squeak-dev(a)lists.squeakfoundation.org Subject: [squeak-dev] Seaside question won't post to Seaside news group Message-ID: <ipc5rchghu20meho39r22picl3e92abl4q(a)4ax.com> Content-Type: text/plain; charset=us-ascii Hi, I'm very sorry for posting this here but I have tried to post two messages to the Seaside news group that don't show up and I don't get any kind of rejection notice. I know Seaside is more a Pharo project than a Squeak project but I know I can post here (at least I have before). I want to register a sub-class of WARequestHandler as '..'. '...' works but '..' doesn't. I'm trying to use the same URL that one would use with Apache like "../folder/filename.txt". The request handler then reads and returns the file. Thanks for any ideas as to where to look. Lou -- Louis LaBrunda Keystone Software Corp. SkypeMe callto://PhotonDemon

1 0

Seaside question won't post to Seaside news group
by Louis LaBrunda 09 Sep '17

09 Sep '17

Hi, I'm very sorry for posting this here but I have tried to post two messages to the Seaside news group that don't show up and I don't get any kind of rejection notice. I know Seaside is more a Pharo project than a Squeak project but I know I can post here (at least I have before). I want to register a sub-class of WARequestHandler as '..'. '...' works but '..' doesn't. I'm trying to use the same URL that one would use with Apache like "../folder/filename.txt". The request handler then reads and returns the file. Thanks for any ideas as to where to look. Lou -- Louis LaBrunda Keystone Software Corp. SkypeMe callto://PhotonDemon

3 3

The Trunk: Network-pre.204.mcz
by commits＠source.squeak.org 09 Sep '17

09 Sep '17

Patrick Rein uploaded a new version of Network to project The Trunk: http://source.squeak.org/trunk/Network-pre.204.mcz ==================== Summary ==================== Name: Network-pre.204 Author: pre Time: 9 September 2017, 11:49:05.258212 am UUID: 9c649cab-fd0e-f040-b203-268bcc758df6 Ancestors: Network-pre.203 Fixes minor issue with new MailAddressParser after using wrong message for trimming whitespace =============== Diff against Network-pre.203 =============== Item was changed: ----- Method: MailAddressParser>>grabName (in category 'parsing') ----- grabName | name | name := ''. [tokens isEmpty not and: [#(Atom QuotedString $. $@) includes: (tokens last type) ]] whileTrue: [ name := {tokens removeLast text copyWithoutAll: '"'. name} joinSeparatedBy: Character space ]. + ^ name withBlanksTrimmed! - ^ name trimBoth!

1 0

The Trunk: Network-pre.203.mcz
by commits＠source.squeak.org 06 Sep '17

06 Sep '17

Patrick Rein uploaded a new version of Network to project The Trunk: http://source.squeak.org/trunk/Network-pre.203.mcz ==================== Summary ==================== Name: Network-pre.203 Author: pre Time: 6 September 2017, 7:55:04.916204 pm UUID: a4494ca0-d9ef-1347-94b6-7719eb9a3e26 Ancestors: Network-tpr.202 Changes the mail address parser to only accept addresses in a valid string of email addresses. Adds logic to the mail address parser to parse out the names in front of email addresses. Changes the tokenizer to accept non-ascii characters in email addresses which is allowed since 2012. =============== Diff against Network-tpr.202 =============== Item was changed: Object subclass: #MailAddressParser + instanceVariableNames: 'tokens addresses storeNames curAddrTokens' - instanceVariableNames: 'tokens addresses curAddrTokens' classVariableNames: '' poolDictionaries: '' category: 'Network-RFC822'! !MailAddressParser commentStamp: '<historical>' prior: 0! Parse mail addresses. The basic syntax is: addressList := MailAddressParser addressesIn: aString This currently only returns the bare addresses, but it could also return a list of the address "source codes". For example, if you give it "Joe <joe@foo>, <jane>", it will currently return a list ('joe@foo' 'jane'). It would be nice to also get a list ('Joe <joe@foo>' '<jane>').! Item was added: + ----- Method: MailAddressParser class>>addressesAndNamePairsIn: (in category 'parsing') ----- + addressesAndNamePairsIn: aString + "return a collection of the addresses and the corresponding names listed in aString" + | tokens | + tokens := MailAddressTokenizer tokensIn: aString. + ^(self new initialize: tokens) grabAddressesAndNames! Item was changed: ----- Method: MailAddressParser>>grabAddressWithRoute (in category 'parsing') ----- grabAddressWithRoute "grad an address of the form 'Descriptive Text <real.address(a)c.d.e>" + | name | self startNewAddress. + - tokens removeLast. "remove the >" "grab until we see a $<" + [ tokens isEmpty ifTrue: [self error: '<> are not matched' ]. + tokens last type = $<] + whileFalse: [ self addToAddress ]. - [ - tokens isEmpty ifTrue: [ - self error: '<> are not matched' ]. - tokens last type = $< - ] whileFalse: [ self addToAddress ]. tokens removeLast. "remove the <" + self finishAddress. + + name := self grabName. + + storeNames ifTrue: [addresses addFirst: {name . addresses removeFirst}].! - - - self removePhrase. - - self finishAddress! Item was changed: ----- Method: MailAddressParser>>grabAddresses (in category 'parsing') ----- grabAddresses + - "grab all the addresses in the string" | token | "remove comments" tokens removeAllSuchThat: [:t | t type == #Comment]. + "grab one address or address group each time through this loop" + [ [tokens isEmpty not and: [ tokens last type = $, ]] + whileTrue: [ tokens removeLast ]. + tokens isEmpty ] whileFalse: [ + - [ - "remove commas" - [ - tokens isEmpty not and: [ tokens last type = $, ] - ] whileTrue: [ tokens removeLast ]. - - "check whether any tokens are left" - tokens isEmpty - ] whileFalse: [ token := tokens last. "delegate, depending on what form the address is in" "the from can be determined from the last token" + token type = $> + ifTrue: [self grabAddressWithRoute ] + ifFalse: [ (#(Atom DomainLiteral QuotedString) includes: token type) + ifTrue: [self grabBasicAddress ] + ifFalse: [token type = $; + ifTrue: [self grabGroupAddress ] + ifFalse: [self error: 'un-recognized address format' ]]]]. + ^ addresses! - token type = $> ifTrue: [ - self grabAddressWithRoute ] - ifFalse: [ - (#(Atom DomainLiteral QuotedString) includes: token type) ifTrue: [ - self grabBasicAddress ] - ifFalse: [ - token type = $; ifTrue: [ - self grabGroupAddress ] - ifFalse: [ - ^self error: 'un-recognized address format' ] ] ] - ]. - - ^addresses! Item was added: + ----- Method: MailAddressParser>>grabAddressesAndNames (in category 'parsing') ----- + grabAddressesAndNames + + storeNames := true. + + self grabAddresses. + + addresses := addresses collect: [:a | + a isString + ifTrue: [{'' . a}] + ifFalse: [a] ]. + + ^ addresses! Item was changed: ----- Method: MailAddressParser>>grabBasicAddress (in category 'parsing') ----- grabBasicAddress "grad an address of the form a.b(a)c.d.e" + | name | self startNewAddress. "grab either the domain if specified, or the domain if not" self addToAddress. [tokens isEmpty not and: [ tokens last type = $.] ] whileTrue: ["add name-dot pairs of tokens" self addToAddress. (#(Atom QuotedString ) includes: tokens last type) ifFalse: [self error: 'bad token in address: ' , tokens last text]. self addToAddress]. (tokens isEmpty or: [tokens last type ~= $@]) ifTrue: ["no domain specified" self finishAddress] ifFalse: ["that was the domain. check that no QuotedString's slipped in" curAddrTokens do: [:tok | tok type = #QuotedString ifTrue: [self error: 'quote marks are not allowed within a domain name (' , tok text , ')']]. "add the @ sign" self addToAddress. "add the local part" (#(Atom QuotedString ) includes: tokens last type) ifFalse: [self error: 'invalid local part for address: ' , tokens last text]. self addToAddress. "add word-dot pairs if there are any" [tokens isEmpty not and: [tokens last type = $.]] whileTrue: [self addToAddress. (tokens isEmpty not and: [#(Atom QuotedString ) includes: tokens last type]) ifTrue: [self addToAddress]]. + self finishAddress]. + + name := self grabName. + storeNames ifTrue: [addresses addFirst: {name . addresses removeFirst}].! - self finishAddress]! Item was changed: ----- Method: MailAddressParser>>grabGroupAddress (in category 'parsing') ----- grabGroupAddress "grab an address of the form 'phrase : address, address, ..., address;'" "I'm not 100% sure what this format means, so I'm just returningthe list of addresses between the : and ; -ls (if this sounds right to someone, feel free to remove this comment :)" "remove the $; " tokens removeLast. "grab one address each time through this loop" + [ "remove commas" + [tokens isEmpty not and: [ tokens last type = $, ]] + whileTrue: [ tokens removeLast ]. - [ - "remove commas" - [ - tokens isEmpty not and: [ tokens last type = $, ] - ] whileTrue: [ tokens removeLast ]. + tokens isEmpty ifTrue: ["no matching :" ^ self error: 'stray ; in address list']. + tokens last type = $:] + whileFalse: [ + + "delegate to either grabAddressWithRoute, or grabBasicAddress. nested groups are not allowed" + tokens last type = $> + ifTrue: [self grabAddressWithRoute ] + ifFalse: [ (#(Atom DomainLiteral QuotedString) includes: tokens last type) + ifTrue: [self grabBasicAddress ] + ifFalse: [^ self error: 'un-recognized address format' ]]]. - tokens isEmpty ifTrue: [ - "no matching :" - ^self error: 'stray ; in address list'. ]. - tokens last type = $: - ] whileFalse: [ - "delegate to either grabAddressWithRoute, or grabBasicAddress. nested groups are not allowed" - - tokens last type = $> ifTrue: [ - self grabAddressWithRoute ] - ifFalse: [ - (#(Atom DomainLiteral QuotedString) includes: tokens last type) ifTrue: [ - self grabBasicAddress ] - ifFalse: [ - ^self error: 'un-recognized address format' ] ] - ]. - tokens removeLast. "remove the :" self removePhrase.! Item was added: + ----- Method: MailAddressParser>>grabName (in category 'parsing') ----- + grabName + + | name | + name := ''. + [tokens isEmpty not and: [#(Atom QuotedString $. $@) includes: (tokens last type) ]] + whileTrue: [ name := {tokens removeLast text copyWithoutAll: '"'. name} joinSeparatedBy: Character space ]. + ^ name trimBoth! Item was added: + ----- Method: MailAddressParser>>initialize (in category 'private-initialization') ----- + initialize + + storeNames := false.! Item was changed: ----- Method: MailAddressParser>>removePhrase (in category 'parsing') ----- removePhrase "skip most characters to the left of this" + [tokens isEmpty not and: [#(Atom QuotedString $. $@) includes: (tokens last type)]] + whileTrue: [ tokens removeLast ]. - [ - tokens isEmpty not and: [ - #(Atom QuotedString $. $@) includes: (tokens last type) ] - ] whileTrue: [ tokens removeLast ]. ! Item was changed: ----- Method: MailAddressTokenizer class>>nonAtomSet (in category 'class initialization') ----- nonAtomSet + "(from RFC 6531)" - "(from RFC 2822)" + ^CSNonAtom ifNil: [ | notAtoms | + notAtoms := CharacterSet new - ^CSNonAtom ifNil: [ - CSNonAtom := CharacterSet new addAll: ($A to: $Z); addAll: ($a to: $z); addAll: ($0 to: $9); addAll: '!!#$%&''*+-/=?^_`{|}~'; + complement. + CSNonAtom := Set new. + 0 to: 127 do: [:v | | c | + c := Character value: v. + (notAtoms includes: c) ifTrue: [CSNonAtom add: c]]]! - complement ]!

1 0

The Trunk: NetworkTests-pre.50.mcz
by commits＠source.squeak.org 06 Sep '17

06 Sep '17

Patrick Rein uploaded a new version of NetworkTests to project The Trunk: http://source.squeak.org/trunk/NetworkTests-pre.50.mcz ==================== Summary ==================== Name: NetworkTests-pre.50 Author: pre Time: 6 September 2017, 7:53:00.334204 pm UUID: b7cc6ccd-9d08-0141-a4d7-b3ad3eaca877 Ancestors: NetworkTests-pre.49 Changes the test case for the logic of the mail address parser to only accept actual lists of email addresses, also adds a test case for parsing out the names in front of email addresses. =============== Diff against NetworkTests-pre.49 =============== Item was added: + ----- Method: MailAddressParserTest>>stringOfMailAddresses (in category 'fixtures') ----- + stringOfMailAddresses + ^ 'joe(a)lama.com, joe2(a)lama.com, joe3(a)lama.com, joe4 , Not an Address <joe5@address>, joe.(annoying (nested) comment)literal(a)[1.2.3.4], "an annoying" group : joe1@groupie, joe2@groupie, "Joey3" joe3@groupy, "joe6"."joe8"@group.com;, Lex''s email account <lex>, foo+bar(a)baz.com, romeo&juliet(a)shakespeare.uk, "Name" mail(a)email.com'.! Item was added: + ----- Method: MailAddressParserTest>>testAddressesAndNamePairsIn (in category 'tests') ----- + testAddressesAndNamePairsIn + + | correctAnswer parsedParis | + + correctAnswer := #( + ('' 'joe(a)lama.com') + ('' 'joe2(a)lama.com') + ('' 'joe3(a)lama.com') + ('' 'joe4') + ('Not an Address' 'joe5@address') + ('' 'joe.literal(a)[1.2.3.4]') + ('' 'joe1@groupie') + ('' 'joe2@groupie') + ('Joey3' 'joe3@groupy') + ('' '"joe6"."joe8"@group.com') + ('Lex''s email account' 'lex') + ('' 'foo+bar(a)baz.com') + ('' 'romeo&juliet(a)shakespeare.uk') + ('Name' 'mail(a)email.com')) asOrderedCollection. + + parsedParis := MailAddressParser addressesAndNamePairsIn: self stringOfMailAddresses. + self assert: correctAnswer equals: parsedParis.! Item was changed: ----- Method: MailAddressParserTest>>testAddressesIn (in category 'tests') ----- testAddressesIn + | correctAnswer | - | testString correctAnswer | + correctAnswer := #( + 'joe(a)lama.com' + 'joe2(a)lama.com' + 'joe3(a)lama.com' + 'joe4' + 'joe5@address' + 'joe.literal(a)[1.2.3.4]' + 'joe1@groupie' + 'joe2@groupie' + 'joe3@groupy' + '"joe6"."joe8"@group.com' + 'lex' + 'foo+bar(a)baz.com' + 'romeo&juliet(a)shakespeare.uk' + 'mail(a)email.com') asOrderedCollection. - testString := 'joe(a)lama.com, joe2(a)lama.com joe3(a)lama.com joe4 , Not an Address <joe5@address>, joe.(annoying (nested) comment)literal(a)[1.2.3.4], "an annoying" group : joe1@groupie, joe2@groupie, "Joey" joe3@groupy, "joe6"."joe8"@group.com;, Lex''s email account <lex>, foo+bar(a)baz.com romeo&juliet(a)shakespeare.uk'. + self assert: correctAnswer equals: (MailAddressParser addressesIn: self stringOfMailAddresses).! - correctAnswer := #('joe(a)lama.com' 'joe2(a)lama.com' 'joe3(a)lama.com' 'joe4' 'joe5@address' 'joe.literal(a)[1.2.3.4]' 'joe1@groupie' 'joe2@groupie' '"Joey"' 'joe3@groupy' '"joe6"."joe8"@group.com' 'lex' 'foo+bar(a)baz.com' 'romeo&juliet(a)shakespeare.uk') asOrderedCollection. - - self assert: ((MailAddressParser addressesIn: testString) = correctAnswer).!

1 0

Inbox/Files-monty.172
by tim Rowledge 06 Sep '17

06 Sep '17

I happened upon Monty’s suggested changes to file opening. They look pretty reasonable to me but since I don’t use Windows and it looks mostly like changes that relate to the old Windows issue of running out of file descriptors and needing to GC, I really can’t have a meaningful opinion. Are we going to incorporate the changes? tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: RIW: Re-Invent Wheel

4 6

Magic mirror
by tim Rowledge 05 Sep '17

05 Sep '17

I’ve been musing about making a Pi powered magic mirror for a while. The problem is that the typical software is nasty js/python/php stuff and I really don’t want that. Anyone interested in collaborating on a better way? See http://michaelteeuw.nl/tagged/magicmirror for a bunch of posts about an early and nicely done exemplar, maybe https://magicmirror.builders and generally google for ‘raspberry pi magic mirror’. We have JSON reading, MQTT handling, web fetching, morph animating, decent fonts (via Pango & the UnicodePlugin) and even some neat sound abilities. We could very reasonably run on a Pi Zero W, which is actually something that most web browser based code can’t do well. With a usb microphone we might even do Siri/Alexa integration? tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: PWB: Put to Waste Basket

1 0

← Newer
1
2
3
4
5
6
7
8
9
Older →