Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours:

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006508.html

Name: Multilingual-nice.187
Ancestors: Multilingual-nice.186

Remove scanSelector/scanSelectorAt: now that they are replaced by new mechanism #scanMultibyteCharactersFrom:to:in:with:rightX:font:
Take property of specific *Multilingual methods from CharacterScanner

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006509.html

Name: Multilingual-nice.188
Ancestors: Multilingual-nice.187

Avoid using theFirstCharCrossedX, this has to be handled differently, and uniformly Multilingual WideString or not.

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006510.html

Name: Graphics-nice.258
Ancestors: Graphics-nice.257

Take back the scanCharactersFrom:to:with:rightX:font: methods from *Collections
For some reasons (renaming a protocol) this message is also required in Symbol.
Give back those exlcusive MultiByte and Japanese CharacterScanner methods to *Multilingual.
Remove the now obsolete CharacterScanner methods which tried to pass stopConditions:kern:
Remove theFirstCharCrossedX which must be handled differently and uniformely

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006511.html

Name: Collections-nice.540
Ancestors: Collections-tpr.539

Give back the scanCharactersFrom:to:with:rightX:font: methods to *Graphics

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006512.html

Name: ST80-nice.158
Ancestors: ST80-nice.157

DisplayScanner does not require a lineHeight inst. var., it can be a temporary.

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006513.html

Name: Graphics-nice.259
Ancestors: Graphics-nice.258

DisplayScanner does not require a lineHeight inst. var., it can be a temporary.

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006514.html

Name: Graphics-nice.260
Ancestors: Graphics-nice.259

Introduce uniform handling for the case when the first char does not fit in the composition rectangle: if we crossedX and have no breakable, then advanceIfFirstCharOfLine.
Move some pendingkernX reset (way too many of those...).
In DisplayScanner, this is more complex because we display BEFORE processing the stop conditions. Fix it with a temporary workaround that advanceIfFirstCharOfLine in the scan loop, but prepare a new instance variable lastDisplayableIndex to fix it properly (in a next stage, displaying is vital).

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006515.html

Name: GraphicsTests-nice.35
Ancestors: GraphicsTests-ul.34

An ersatz of CharacterScannerTest...
This illustrates several bugs that we recently corrected.
It would be interesting to try it in other images in order to measure our progress...
Of course, there are many more tests to write...

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006516.html

Name: Graphics-nice.261
Ancestors: Graphics-nice.260

If we perform the stopCondition before displaying the string in order to find the lastDisplayableIndex,
then the stopCondition will reset the font at each endOfRun,
and the display will occur with the wrong font.

It's impossible to keep a pointer on lastFont := font, because the font glyphs seem to be recycled for obscure optimization reasons.

Instead, add a new inst. var. stopConditionsMustBeReset to differ the setFont (setStopCondition) at next scan loop.

Set this inst. var. so that we can safely use it at next commit.

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006517.html

Name: Graphics-nice.262
Ancestors: Graphics-nice.261

Change the DisplayScanner loop:
- don't let endOfRun immediately change the font (setStopConditions) but just emit the stopConditionsMustBeReset request
- since the font is not reset, now perform: stopCondition before dispalying the string.
this enable using the lastDisplayableIndex rather than trying to double the job of stopCondition and grumbling against inhomogeneous lastIndex(ing).

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006518.html

Name: ST80-nice.159
Ancestors: ST80-nice.158

Update the MVC character display loop to use lastDisplayableIndex and stopConditionsMustBeReset like the Morphic one.
Reminder: this was for handling the edge case when the first char does not fit the composition rectangle.

=============================================

http://lists.squeakfoundation.org/pipermail/packages/2013-October/006519.html

Name: Multilingual-nice.189
Ancestors: Multilingual-nice.188

Put the charsetAt: send out of the scanJapaneseCharactersFrom:to:in:rightX: inner loop.

=============================================