Hi All,
I discovered that the "unicode spew" was not what I thought it was. When the "spew" happens it is because the html output is dumped within the body tag like this: <html> <head></head> <body><�!DOCTYPE html><�html><�head><�title>Seaside<�/title><�meta charset="utf-8"/><�link rel="stylesheet" type="text/css" href="/files/WADevelopmentFiles/development.css"/><�/head><�body onload="onLoad()"><�a href="/waDebug?_s=ccLzvr_77iZO7bVN&amp;_k=KdqQxjU4LkFiBg30&amp;1">standard dash<�/a>&nbsp;<�a h </body> </html> Its wierd and its a heisenbug. I created a copy of WACounter , named it WADebug and the callbacks "dash" and "mdash" totally isolate the Character. http://menmachinesmaterials.com/waDebug *renderContentOn: html html anchor callback: [ self dash ]; with: 'standard dash'. html space. html anchor callback: [ self ndash ]; with: 'n-dash'. html break. html html: count. dash count := '-' ndash count := (Character value:8211) asString * What is really weird is that it works on the server when it did not work on my workstation. For some reason, it just started working on my Workstation. I made a small edit to my Workstation, unrelated to that issue, and it stopped working. I am running Seaside on Squeak5.2 latest update: #18233 Anybody have a clue why it would mangle the output like that? I will be trying a Pharo install possibly tomorrow to see if that negates the bug. cheers. t -- Sent from: http://forum.world.st/Seaside-General-f86180.html _______________________________________________ seaside mailing list [hidden email] http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside |
Hi All,
First, my apologies for the multiple posts. When I did not see my message appear, I assumed I had messed up and reposted. I believe that the "Unicode Spew" is a result of the parse taking longer than the Seaside Render loop and/or errors in the parse not being handled. If you go here: http://menmachinesmaterials.com/WikitextParser click the Hamburger Icon->Database->Blade Runner, it creates the unicode spew every time. If I run the same parse within the image as a workspace Doit, it works just fine. Here is the bulky thing, it is a sub-message of renderContentOn:html *renderParsedOn: html | wikiGrammar wikiParser input actor s| actor := PEGWikiMediaGeneratorTables new. actor transcripton ifTrue:[ Transcript clear]. wikicode isNil ifTrue:[input := '== Welcome To WikitextParserBrowser =='] ifFalse:[input := wikicode]. wikiGrammar := PEGParser grammarWikiMediaTables reading positioning. wikiParser := PEGParser parserPEG parse: 'Grammar' stream: wikiGrammar actor: PEGParserParser new. [[output := wikiParser parse: 'Page' stream: input actor: actor. ] on: Error do:[:ex | output := ' Error parsing. see Wikicode tab for source <code>', ex description, '</code>' ]] ensure:[ s := ((output asString copyReplaceTokens: '<body>' with:'' ) copyReplaceTokens:'</body>' with:'') . s := s copyReplaceAll:'>' with:'>' asTokens:false. s := s copyReplaceAll:'<' with:'<' asTokens:false. html break;break. html html: s. ].* I think the "Render Loop" just cannot handle that call when things go south. Thank you all for your patience and advise. -- Sent from: http://forum.world.st/Seaside-General-f86180.html _______________________________________________ seaside mailing list [hidden email] http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside |
Free forum by Nabble | Edit this page |