January 2012 - Oberon - mailman3.inf.ethz.ch

Why transform text via a file?
by Chris Glur 22 Jan '12

22 Jan '12

AFAICS the interesting Mail.Mod question raised was not answered and just evaporated. Although I doubt that it's related to the following: files [pipes] are nice for transforming text, because they automatically adjust their size, and eg. facilitate the economical [in human effort] concatenative style. <><><><><> Newsgroups: comp.lang.forth Subject: Re (2): existing code to build a 'sed-like'? References: <4f1ad6d4$0$293$14726298(a)news.sunsite.dk> In article <4f1ad6d4$0$293$14726298(a)news.sunsite.dk>, Doug Hoffman <glidedog(a)gmail.com> wrote: > On 1/21/12 9:34 AM, no.top.post(a)gmail.com wrote: > > > I get the impression that forth programmers don't code at a high level. > > Without getting sidetracked on the meaning of "high level" I believe > that is a fair assessment. There are of course exceptions such as > people like me that don't do embedded (resource constrained) work and > instead program for the desktop. > > > Is that because no libraries of convenient tools/functions exist, > > for such functions to just be concatenated? > > I believe so, at least historically. Recently there have been more > focused efforts toward centralizing/building/improving such libraries so > reinventing wheels in Forth is minimized. > > -Doug Recently: after 35 years? Here's an example of hi-level which can profitably use the concatinative style which I'm harping on about:- If I want to 'clean': > > > You have it the wrong way around. Interpretation is a special case of > > > execution. Very simply stated (because someone here is bound to take > > > issue ...), a VM is a loop, an interpreter, and an emulated processor, > > > implemented in machine code, via compiled C or assembled assembly, > > > which is executed. to get: >>>You have it the wrong way around. Interpretation is a special >>>case of execution. Very simply stated (because someone here >>>is bound to take issue ...), a VM is a loop, an interpreter, >>>and an emulated processor, implemented in machine code, >>>via compiled C or assembled assembly, which is executed. previously I'd be thinking in terms of indices and pointers, which is expensive. Now with knowledge of concatinative style and many good examples of *nix script I do: GetText | RemoveAll ">" | Replace <EndOfLine> with <space> | UseExistingFormatterFor1spaceBetweenWords&WrapLineLength 66 | DeleteLeading <spaces> Which the *nix 1-line-jockeys write as the 5 transformations: cat d2 | sed s/">"//g | tr "\012" "\040"| fmt -u -w 66 | sed s/^" "*// Which could be called eg. `cleanUSEnet` And 3 times: | sed s/^/">"/ adds the leading ">>>". I don't like the *nix syntax, but forth-users are 'almost there' in their method, and something like: "d2 cleanUSEnet 3 cite" would be a nice syntax. == Chris Glur.

1 0

Re: [Oberon] Mail args wrongly changed during processing?
by Jörg Straube 21 Jan '12

21 Jan '12

Chris Mail.Panel is very old (approx 10 years or so). So, better use newer Oberon, AOS, A2... distros with things like IMAP and SSL. To my knowledge, Mail.Panel does neither support SSL security feature nor IMAP. So, if the mail server wants you to use SSL, Mail.Panel is not the right tool for you. Just for fun I downloaded UnixAos on my Mac and run the old Oberon system in it. After I set all my parameters right, Mail.Panel worked like a charm. To retrieve email with Mail.Panel, you need THREE things: 1) The POP server address (in my case "pop.bluewin.ch") 2) your username (in my case "joerg.straube") 3) your password (in my case "ABCDEF") In Mail.Panel click on "Settings". Under "POP Server" enter 1). Under "POP User" enter 2). "EMail Address" and "SMTP Server" is only needed to send mails Close the Settings Panel Right to word "Server" click on "Get". As Oberon (or better NetSystem) does not know the password for your mail server yet, you will get in System.Log a command like NetSystem.SetUser pop:joerg.straube@pop.bluewin.ch ~ Execute that command. You will then be prompted for a password. Enter it and hit return. Your mails will be downloaded and the headers displayed in the Panel. Double clicking on the header will open a new window with the mails content. When reading your mail, it seems you have at least THREE different mail accounts: Address POP3 server Username crglur(a)gmail.com pop.googlemail.com crglur 0110246665(a)neomail.co.za mail.neomail.co.za 0110246665 ??? pop3.absamail.co.za ?? Google uses SSL, so Mail.Panel won't work. Don't know if SSL is needed to access neomail or absamail, but you should know. br Joerg On 20.01.2012, at 01:00, Chris Glur wrote: > Perhaps I can benefit from collaborator's newly found familiarity > with Mail.Mod [where I don't find a LEO version right now]. > > Because in Africa, you always need a planB, I've got gmail, which > is inconvenient, since I've not learned how to access it other than > via a monster-browser -- slowly. > > My planC is worse, and I access it rarely: via telnet. > > Since I've never suceeded with Mail.Panel before, I now > decided to try again [after a decade] for planC via Mail.Panel > > Without going down to source-code level, I'm already confused:- > > My <Trap Trace> is: ---------- > Trap 5.7 (Index out of range) > sp = BFF12DD4, bp = BFF12DEC, pc = 082844D8 > > Mail.Append @ 1168 > buffer = B774C478H : > heap = : Mail.Heap > i = 0 > idx = 0 > index = B76BBE58H : > j = 0 > len = 0 > str = "crglur(a)gmail.com" > Mail.Insert @ 1820 <-- stage N-1 > .. > str = "crglur(a)gmail.com" > Mail.AddMsgHead @ 8997 <-- stage N-2 > .. > str = "crglur(a)gmail.com" > Mail.Synchronize @ 11975 <-- stage N-3 > .. > user = "0110246665(a)neomail.co.za" <-- PlanC !! > .. > --------- > The Mail.Panel settings read > "Local Settings (override Oberon.Text)" > > The trace for PROC Mail.Synchronize > has correctly accepted my new settings in Mail.Panel: > for neomail. > > Yet, <neomail> changes to <gmail> further on ?!? > > And Oberon.Text has no knowledge about "crglur(a)gmail.com" >> POP = "pop.absamail.co.za" {* incoming mail server *} > > So how/why does "crglur(a)gmail.com" get to replace > "0110246665(a)neomail.co.za" ? > > I vaguely remember that I use a separate directory, for my > second <absamail> userID, which uses a second Oberon.Text > But that's not where the ghost comes from. > --------- > > Surely if PROC Mail.Synchronize has got the correct settings, > these should not change to other settings, further on. > > Wouldn't you expect the arg to be accessed from *ONE* > location and to be passed along without being changed? > > If I could get a simple solution [without analysing the > punishing Mail.Mod code] I could use my planC. > > == TIA, Chris Glur. > > > -- > Oberon(a)lists.inf.ethz.ch mailing list for ETH Oberon and related systems > https://lists.inf.ethz.ch/mailman/listinfo/oberon

1 0

Mail args wrongly changed during processing?
by Chris Glur 20 Jan '12

20 Jan '12

Perhaps I can benefit from collaborator's newly found familiarity with Mail.Mod [where I don't find a LEO version right now]. Because in Africa, you always need a planB, I've got gmail, which is inconvenient, since I've not learned how to access it other than via a monster-browser -- slowly. My planC is worse, and I access it rarely: via telnet. Since I've never suceeded with Mail.Panel before, I now decided to try again [after a decade] for planC via Mail.Panel Without going down to source-code level, I'm already confused:- My <Trap Trace> is: ---------- Trap 5.7 (Index out of range) sp = BFF12DD4, bp = BFF12DEC, pc = 082844D8 Mail.Append @ 1168 buffer = B774C478H : heap = : Mail.Heap i = 0 idx = 0 index = B76BBE58H : j = 0 len = 0 str = "crglur(a)gmail.com" Mail.Insert @ 1820 <-- stage N-1 .. str = "crglur(a)gmail.com" Mail.AddMsgHead @ 8997 <-- stage N-2 .. str = "crglur(a)gmail.com" Mail.Synchronize @ 11975 <-- stage N-3 .. user = "0110246665(a)neomail.co.za" <-- PlanC !! .. --------- The Mail.Panel settings read "Local Settings (override Oberon.Text)" The trace for PROC Mail.Synchronize has correctly accepted my new settings in Mail.Panel: for neomail. Yet, <neomail> changes to <gmail> further on ?!? And Oberon.Text has no knowledge about "crglur(a)gmail.com" >POP = "pop.absamail.co.za" {* incoming mail server *} So how/why does "crglur(a)gmail.com" get to replace "0110246665(a)neomail.co.za" ? I vaguely remember that I use a separate directory, for my second <absamail> userID, which uses a second Oberon.Text But that's not where the ghost comes from. --------- Surely if PROC Mail.Synchronize has got the correct settings, these should not change to other settings, further on. Wouldn't you expect the arg to be accessed from *ONE* location and to be passed along without being changed? If I could get a simple solution [without analysing the punishing Mail.Mod code] I could use my planC. == TIA, Chris Glur.

1 0

Re: [Oberon] Re (2): Mail.GetUIDLs
by peasthope＠shaw.ca 20 Jan '12

20 Jan '12

Alexander, From: Alexander Iljin <ajsoft(a)yandex.ru> Date: Fri, 20 Jan 2012 08:21:03 +0400 > The Files.File object is created in memory with NEW, and it will be > garbage-collected later. ... keeping in mind the finalization > process, can be indirectly applied to other resources. Now I understand a little better. > PS: Excuse me for the Base64 encoding, if any. I currently use a > web interface for emailing, not my regular program. A few mouse clicks is no burden. Thanks, ... Peter E. -- Telephone 1 360 639 0202. bcc: peasthope ... shaw.ca "http://carnot.yi.org/" "http://members.shaw.ca/peasthope/index.html#Itinerary"

1 0

Mail.GetUIDLs
by peasthope＠shaw.ca 20 Jan '12

20 Jan '12

Hi, I'm working on a problem with POP mail retrieval which surfaced during the weekend. This question refers to ETH Oberon / PC Native 05.01.2003. I haven't looked for the analogue in Aos but a similar concept must be there. According to my understanding, the purpose of GetUIDLs is to retrieve a list of message UIDLs from a POP server; one UIDL per message. PROCEDURE GetUIDLs(S: NetTools.Session; VAR T: Texts.Text); VAR F: Files.File; R: Files.Rider; BEGIN SendCmd(S, "UIDL", ""); IF ReadState(S) THEN F := Files.New(""); Files.Set(R, F, 0); ReadText(S, R); NEW(T); Texts.LoadAscii(T, F) ELSE T := NIL END END GetUIDLs; Files.New("") creates an anonymous file. What becomes of this file when GetUIDLs exits? Why is the file involved rather than the data from Session put directly into T? Thanks, ... Peter E. -- Telephone 1 360 639 0202. Former telephone 1 360 450 2132. bcc: peasthope ... shaw.ca "http://carnot.yi.org/" "http://members.shaw.ca/peasthope/index.html#Itinerary"

3 4

Re: [Oberon] Mail.GetUIDLs
by Chris Glur 19 Jan '12

19 Jan '12

> > PROCEDURE GetUIDLs(S: NetTools.Session; VAR T: Texts.Text); > > VAR > > F: Files.File; > > R: Files.Rider; > > BEGIN > > SendCmd(S, "UIDL", ""); > > IF ReadState(S) THEN > > F := Files.New(""); Files.Set(R, F, 0); <-- initFile > > ReadText(S, R); <-- ?presume? S=>R > > NEW(T); Texts.LoadAscii(T, F) > > ELSE > > T := NIL <-- perhaps at error-flag for callingProc > > END > > END GetUIDLs; > > > > Files.New("") creates an anonymous file. What becomes of this file when > > GetUIDLs exits? Why is the file involved rather than the data from Session > > put directly into T? > > > > Thanks, ... Peter E. > > > > > > -- PROCEDURE GetUIDLs(S: NetTools.Session; VAR T: Texts.Text); S SendCmd(S, "UIDL", ""); IF ReadState(S) THEN F := Files.New(""); Files.Set(R, F, 0); <-- initFile S=>R ReadText(S, R); <-- ?presume? S=>R Aubrey McIntosh wrote:- > This code has the same interpretation under V4 or Component Pascal, so I > think my comment will be true for the ETH / PC Native. > > A Text is not a sequence of bytes. It is more abstract than a file is. It > is a sequence of "Runs" where each run has attributes such as font, color, > length, and a source file and the starting position. > > This structure makes it very fast to copy from one viewer to another. > > When a Text is saved to a file, the separate runs are all written anew to > that file. At that time it is very easy to confuse the concepts of Text > and File, so I always try to think of a text that is displayed in a viewer > after it has had items copied from other texts whenever I am writing code. > > So, there has to be a file that is referenced in the Run data structure. > In this code snippet, the new anonymous file is created to fill that role. > > In Component Pascal, Converters all appear to contain code that looks > almost identical this. I always understood that such low-level implementation details were hidden, and that the interface [to other modules] couldn't see that. ================= For no obvious reason Alexander Iljin wrapped his mail in inconvenient base64 encoding, that: ] The anonymous file will be garbage-collected at a later time. ] I would guess that Texts.LoadAscii contains some important data conversion, ] e.g. it may ensure that there are no garbage characters, etc. ] One has to look at the procedure to find out. ======> Lets see if switching from LEO to LNO can show N-O:Mail.Mod ? OK, LNO can mount my decades old N-O:<Source>.Mod and can write Mail.Mod to a linux file; but the <select BIG text via 2 viewers via shift-key> doesn't work. So I've just copied PROCEDURE ReadState and PROCEDURE ReadText I'm particularly interested in this, because I had to modify/extend the <mail> code to handle <Send Authenticate>. And now I remember that I discovered that Mail.Mod was terrible code, whereas TextMail.Mod was nicely structured -- which I used. When I started ETHO in the late 90's I couldn't get Mail.Panel to work; so I used TextMail.* and have been using it since then. It's call inertia and it's the reason why 98% of the population uses M$ even though it is such garbage. == Chris Glur. PS. for convenience I'm just going to mail this my: /home/Mail.Mod Don't bother to read the 2 attatched PROCs. Of course I could cut them. BTW, of course you can't see how I've coloured this text to help in the analysis [mental process aid]. IMO this <object view> of computing is powerful and flexible: you don't think ito applications like mailers and browsers. You think rather ito of files [& lines & bytes]. And being-emailed is just one of the actions that the object:file can do. So what is valuable and 'permanent' is YOUR data, and the applications which it uses are just transitory. So eg. *this* discussion about Mail.GetUIDLs doesn't belong to email, Usenet , http, *.pdf-apps ..etc. It is data that belongs at a suitable place in your data-repository [currently in a file directory] and what ever applications become suitable and available can use the data and add value. =========================================================== PROCEDURE ReadState(S: NetTools.Session): BOOLEAN; BEGIN NetSystem.ReadString(S.C, S.reply); IF trace THEN Texts.WriteString(W, "RCV: "); Texts.WriteString(W, S.reply); Texts.WriteLn(W); Texts.Append(Oberon.Log, W.buf) END; IF S.reply[0] = "+" THEN S.status := NetTools.Done; S.res := NetTools.Done ELSE S.status := NetTools.Failed; S.res := NetTools.Failed END; RETURN S.status = NetTools.Done END ReadState; ================ PROCEDURE ReadText(S: NetTools.Session; VAR R: Files.Rider); VAR buffer: ARRAY BufLen OF CHAR; len, rlen, i, offs: LONGINT; state: INTEGER; ch, old: CHAR; BEGIN old := 0X; offs := 1; len := NetSystem.Available(S.C); state := NetSystem.State(S.C); WHILE (len > 0) OR (state = NetSystem.inout) DO IF len > (BufLen-2) THEN rlen := BufLen-2 ELSE rlen := len END; NetSystem.ReadBytes(S.C, 0, rlen, buffer); i := 0; WHILE i < rlen DO ch := buffer[i]; IF ch = Strings.CR THEN Files.Write(R, ch); IF (offs = 2) & (old = ".") THEN Files.Write(R, Strings.LF); RETURN END; offs := 0 ELSE IF (offs > 0) OR (ch # ".") THEN Files.Write(R, ch) END; INC(offs) END; old := ch; INC(i) END; DEC(len, rlen); IF len <= 0 THEN len := NetSystem.Available(S.C); state := NetSystem.State(S.C) END END END ReadText; ================> LEO: *Mail*Mod*

1 0

Re: [Oberon] LNO revive.
by Chris Glur 11 Jan '12

11 Jan '12

On 1/9/2012 4:00 PM, Chris Glur wrote: > BlackBox/ComponentPascal was a real dissapointment for me. > Chris, I am probably taking this out of context but what, exactly, disappointed you in BlackBox/Component Pascal? Was it the interface which did not look like V4 or something else? How would it: = show me the list of mounted partitions, for me to select the suitable one, = to search for files: *rame*uffr* = Repeat those 2 steps for files: *ill*haos* and *umry* = From the 3 sets of files search for 4 <keywords> and copy the enclosing sentences., to the destination file, whose name I can't remember. But that doesn't matter, because it's writen on my <job sheet> which is HERE, and I just have to klux the name, which I RECOGNISE [if need be because I've given it a meaningful colour/font]. Or a task that I do every day: =extract the set of links from a `lynx -dump` to a file of URLs to lynx-get, next time on-line; where the extracted URLs are at the bottom of the page, whereas you're reading the top of the page, to decide if the link may be of interest -- like foot notes work. 2 attributes of the HCI are essential, which I don't remember BB having [an infuriating aspect of M$ directory is it's hiding of files - for use by kiddies]: 1. the ability to read/write cutNpast between multiple text on the same screen. 2. the ability to work heads-up for 'those' operations. ie. whith no need to look down to the keyboard, to find special keys. How does Micro$loth do this ? 3 files, have 3 sections [of multiple consecutive lines] each: f1: A1y f2: xB3 f3: 2zC The contents of the [3 per file] section-terminator-lines is known; say: "<><><><><><><>" You must construct: f1: ABC f2: xyz f3: 123 and make TableOfContents for each file like: *<A's description> *<B's description> *<C's description> so that you can quickly go to f2 , section 'y' and search for <stringN> Such tasks are commonly done by *nix scripts, but I mean cases where you need a human to read/recognise; so that it can't be automated. == Chris Glur.

2 1

LNO revive.
by Chris Glur 10 Jan '12

10 Jan '12

Someone mentioned using V4 on some mimimal hardware. Can V4 run on these new Win7 based netbooks? Or is there something that's now replaced Win7? It's increasingly difficult to by-pass the Win? OS. I couldn't get DOSbased Oberon to run because of <display incompatibility>. Even old DOS:TurboPascal failed because of a known bug which causes problems in newer/faster hardware. BlackBox/ComponentPascal was a real dissapointment for me. Firstly, instead of using meaningless marketing terms, like 'good experience', we need to state WHY we want the ETHO/S3/V4 interface: * to be able to easily have 6 or more text frames available together, * to be able to easily cutNpaste between text frames, * to be able to easily color any text-stretches on any text frames, * to have <live commands> embedded in any text, * to be able to work: heads-up without needing to look down for special keys.... These minature CF are amazing, but I don't know how reliable they are for long term running of code, with the need to continually read/write. Keeping photos, where a false bit wouldn't be noticed is easy; or do the photos have CRC. Well the CF seems fine so far. What about various *nix file systems which journal/update the disk every few seconds: effect on the limited write-cyle life of solid-state menory? Since I failed to get the good ETHO/S3/V4 interface on the netbook via Win?DOS, I tried linux via a CF or USBstik. LinuxETHOberon [LEO] is very good, and I use it all the time - like now. But it seems a bit of an overkill for a portable/netbook. And X is known as a monster. Since LNO runs with a framebuffer also, if X is missing, you can get the good ETHO-interface, reasonably with a CF or USBstik. LEO was able to read/write up to Win98 files, but linux seems unable to write Win7 files, so you need to copy them [the FAT files] to linux and LNO can manipulate them there. This complexity can be handled convenietly by LNO, like N-O handled the multiple partitions by simply:-- [I'm copying this manually from the netbook to PC:LEO, so could make a mistake] ET.Do Cat ET.Do Dog .. ET.Do Cow where 'Cow' is eg.:------ OFSTools.Mount Cow LinuxFS /mnt/CF3/cow System.Directory Cow:*\d ! OFSTools.UnMount Cow ----------- So a single klux on <ET.Do Cow> mounts and shows all the files in that work-area and makes the later unMounting available via a single klux too. Of course colouring the commands appropriately, helps too. It's a pity that ETHO/LNO doesn't, like V4 does, automatically open FAT [or *nux] files 'as such' instead of as AosFiles. Has anybody got a code patch to check the <file type> before deciding to use *.Open, *.OpenUnix or *.OpenAscii ? Is LNO open sourced? It's more than twice the effort to have to select manually. LNO is your link between your N-O partitions and the power of *nix and LEO. So you could build your own live-books [evolved over time] from existing *.pdf files, by capturing the images via <Linux screen capture to eg. *pgn> and editing/ improving the text, to get a superior product like our ETHO: Chapter5.Text ,instead of the bloated garbage available from the industry leaders. I've been informed that a framebuffer can't be zoomed to occupy more than the display size. Why would it not be able to zoom as X can? == Chris Glur.

2 1

Re. Degeneration via proprietry software
by Chris Glur 04 Jan '12

04 Jan '12

Chris Burrows wrote; > Just guessing, but it reminds me of the Boyer-Moore fast string search > algorithm. The time parameter is probably just a remnant from some > performance checking runs. Yes, even http://en.wikipedia.org/wiki/Boyer%E2%80%93Moore_string_search_algorithm if difficult to understand. The motive that led me to to this is directly related to how socio-economic forces are killing technological advances: Remember when you had to buy-the-next-6-issues to get that valuable article? http is worse, because they are trained in the same school. Typically a 'book' may consist of 20 nodes/http-pages, where the TOC and other redundant contents is repeated 20 times. Several text-only-browsers can fetch and append the text-only, which conveniently form the chapters of your book. But you want to delete the redundant garbage-wrapping, which swamps each fetched http-page. It's bad enough having to pay to fetch the redundant text, but it's infuriating [noise], to have to see it every time you read & refine [eg. via colouring which ETHO does so well] your evolving book. Under ETHO you [can] evolve/morph your 'books of knowledge' by eg. deleting and hi-liting ..etc. the text. This mirrors the change in interconnections in your brain. So, the ETHO intelligently coloured text evolves like a live brain. ============== For this you need a utility to <delete the redundant text-fetches>. But ETHO's Search & Replace seem to handle too small a string. Ideally the deleted garbage should be per-line. I see good reason why *nix is line-based. It's a pity that ETHO hasn't got more line-based utilities. RX* is one that is line-based. The concatinative programming style, where text is serially filtered through <transformers> is powerfull and economical in human effort; compared to ETHO style where procedures are tangled/couple in complex/confusing ways. Consider how conceptually simple the following is, compared to an Oberon implimentation:- GetAllFilesIn(dirX) | LessThanDaysOld(N) | WhichContainString(S1) | .. WhichContainString(SN) | ListThe1stLinesThatFollowsThe1stString("<><><><>") == Chris Glur. PS. LNO seems one way of getting ETHO on a netbook's CF or stik, which has problem getting the full X system, since it runs on a mere frame-buffer. So far I don't know how/what determines if linux gets /dev/fb*

1 0

Degeneration via proprietry software
by Chris Glur 03 Jan '12

03 Jan '12

> I have the A2 up and running under the VMWare (4.0.1 build-528992, Win7 > 64bit home premium). ] After many years out of the Oberon world, I'd like to return (ideally ] through VMWare). ] Can someone on this list give me a hand getting A2 (or Native Oberon) up ] and running under VMWare? How we have degenerated, from the 1M4 floppy based ETH:N-O to '64bit home premium:VMWare' based poor imitation. My Win7 netbook is remarkable hardware, with facilities to block me from doing anything more usefull than a twittering kiddie. The reason for using the original 1M4 based version is grounded in our evolution as humans. If we were horses we'd be comfortable standing for hours. So, mobile devices are only valuable toys. Other creatures may be evolved to not need to be seated, with arms resting on an optimally positioned table, head-up, viewing an optimally positioned view-frame; with the ability to open multiple frame/files, so that eg. any reasonably complex task needing reference [i.e. copy-paste & text-stretch hi-lite/colouring] to: the progressLog/Notes & USEnetArticle/S & email/S & googSearches & ... can be coordinated ALL TOGETHER. For analysing complex text, the heads-up: no need to look away from the viewer facility seems never to have been repeated. == Chris Glur. PS. Newsgroups: comp.software-eng,comp.theory Subject: weird string search algorithm? Wanting to just modify some existing code, to find and delete text-strings, I came across this weird 'algorithm' in my system. And google:'search+string'+'displacement+vector' doesn't lead to Knuth, which my gut told it it would. Pasting the code verbatim:-- VAR tF: TextFrames.Frame; .... Texts.OpenReader(R, text, beg); <-read the selected pattern into sPat[i] i := 0; pos := beg; REPEAT Texts.Read(R, sPat[i]); INC(i); INC(pos) UNTIL (i = MaxPatLen) OR (pos = end); sPatLen := i; CalcDispVec(time) (* calculate displacement vector *) <-what's this?! END; IF tF.car > 0 THEN pos := tF.carloc.pos ELSE pos := 0 END; IF SPatFound(tF, tF.text, pos) THEN MarkPatPos(tF, pos) ----- PROCEDURE CalcDispVec(time: LONGINT); VAR i, j, d: INTEGER; BEGIN lTime := time; i := 1; d := 1; (* calculate displacement vector *) WHILE i <= sPatLen DO j := 0; WHILE (j + d < sPatLen) & (sPat[j] = sPat[j + d]) DO INC(j) END; WHILE i <= j + d DO sDv[i] := d; INC(i) END; INC(d) END END CalcDispVec; ----- Can someone point me to an explanation of this algorithm? == TIA. PSS. this is in ET.Mod & our other <editors>. If CalcDispVec want's the time, it's perhaps random-generating, or checking <event sequences> ?? PSS. how can we handle complex ideas without being able to cutNpaste [without losing sight by switching screens] of: mail, Usenet, general text ...etc.

2 1