DIRECTORY
Convert, FileNames, GGFont, GGParseIn, Imager, ImagerFont, ImagerTransformation, IO, Real, Rope;


GGFontImpl: CEDAR --MONITOR-- PROGRAM IMPORTS Convert, FileNames, GGParseIn, Imager, ImagerFont, ImagerTransformation, IO, Real, Rope
EXPORTS GGFont = BEGIN

ROPE: TYPE = Rope.ROPE;
FontData: TYPE = GGFont.FontData;
FontDataRec: TYPE = GGFont.FontDataRec;

ParseError: PUBLIC ERROR[explanation: ROPE] = CODE;

boldList: LIST OF ROPE _ LIST["-B", "-BI"];
italicList: LIST OF ROPE _ LIST["-I", "-BI"];
endList: LIST OF ROPE _ LIST["mrr", "mir", "brr", "bir", "bold", "italic"];

DigitProc: IO.BreakProc = {
SELECT char FROM
IO.TAB, IO.CR, IO.SP => RETURN [break];
'0, '1, '2, '3, '4, '5, '6, '7, '8, '9 => RETURN [break];
ENDCASE => RETURN [other];
};

NonDigitProc: IO.BreakProc = {
SELECT char FROM
'0, '1, '2, '3, '4, '5, '6, '7, '8, '9 => RETURN [other];
ENDCASE => RETURN [break];
};

DefaultDefaultFontData: PUBLIC PROC RETURNS [data: FontData] = { 
data _ CopyFontData[defaultFontData];
};

CreateFontData: PUBLIC PROC RETURNS [data: FontData] = {
data _ NEW[FontDataRec _ []];
};

InitFontData: PUBLIC PROC [data: FontData] RETURNS [newData: FontData] = {
newData _ data;
IF newData#NIL THEN newData^ _ []; -- fill in from definition defaults
};

CopyFontData: PUBLIC PROC [data: FontData, oldCopy: FontData _ NIL] RETURNS [newCopy: FontData] = {
newCopy _ IF oldCopy#NIL THEN oldCopy ELSE CreateFontData[];
IF data#NIL THEN newCopy^ _ data^;
IF newCopy.transform#NIL THEN newCopy.transform _ ImagerTransformation.Copy[newCopy.transform];
};

scratchData: FontData _ NEW[FontDataRec _ []];
ParseFontData: PUBLIC --ENTRY-- PROC [data: FontData _ NIL, inStream: IO.STREAM, literalP, prefixP, familyP, faceP, transformP, scaleP, storedSizeP, designSizeP: BOOL _ FALSE] RETURNS [newData: FontData] = {
Inner: PROC = {
DisallowedEnding: PROC [s: Rope.ROPE] RETURNS [BOOL] = {
tail: Rope.ROPE _ FileNames.Tail[s, '-];
FOR endRope: LIST OF ROPE _ endList, endRope.rest UNTIL endRope=NIL DO
IF Rope.Find[s, endRope.first, 0, FALSE]#-1 THEN RETURN[TRUE];
ENDLOOP;
RETURN[FALSE];
};

IF literalP THEN {
scratchData.literal _ IO.GetTokenRope[inStream, IO.IDProc].token; -- "xerox/myFonts/FooBarFont"
scratchData.prefix _ FileNames.Directory[scratchData.literal]; -- "xerox/myFonts/"
scratchData.literalFSF _ FileNames.Tail[scratchData.literal, '/]; -- "FooBarFont"
}
ELSE {
lit: ROPE _ IO.GetTokenRope[inStream, IO.IDProc].token; -- "xerox/xc1-2-2/Modern-BI"
IF prefixP THEN {
scratchData.prefix _ FileNames.Directory[lit]; -- "xerox/xc1-2-2/"
lit _ FileNames.Tail[lit, '/]; -- "Modern-BI"
};
IF familyP THEN {
IF DisallowedEnding[lit] THEN ParseError["Literal font name not allowed here"];
scratchData.userFSF _ lit; -- "Modern-BI"
scratchData.family _ Before[lit, '-]; -- "Modern
}; 
IF faceP THEN { -- means figure out the faces from the userFSF
faceRope: ROPE;
faceS: IO.STREAM  _ IO.RIS[scratchData.userFSF]; -- "Modern-BI"
[] _ IO.GetTokenRope[faceS, IO.TokenProc]; -- read and discard family name "Modern"
faceRope  _ IO.GetTokenRope[faceS, IO.IDProc ! IO.EndOfStream => CONTINUE;].token; -- "-BI or SP in other cases
FOR boldRope: LIST OF ROPE _ boldList, boldRope.rest UNTIL boldRope=NIL DO
IF scratchData.bold THEN EXIT;
scratchData.bold _ Rope.Find[faceRope, boldRope.first, 0, FALSE]#-1;
ENDLOOP;
FOR italicRope: LIST OF ROPE _ italicList, italicRope.rest UNTIL italicRope=NIL DO
IF scratchData.italic THEN EXIT;
scratchData.italic _ Rope.Find[faceRope, italicRope.first, 0, FALSE]#-1;
ENDLOOP;
scratchData.faceKnown _ TRUE;
};
scratchData.comfortable _ TRUE;
};
IF transformP THEN scratchData.transform _ GGParseIn.ReadFactoredTransformationVEC[inStream];
IF scaleP THEN scratchData.scale _ Convert.RealFromRope[IO.GetTokenRope[inStream, IO.IDProc].token]; -- "12 or 20.123 or ..."
IF storedSizeP THEN scratchData.storedSize _ Convert.RealFromRope[IO.GetTokenRope[inStream, IO.IDProc].token]; -- "12 or 20.123 or ..."
IF designSizeP THEN scratchData.designSize _ Convert.RealFromRope[IO.GetTokenRope[inStream, IO.IDProc].token]; -- "12 or 20.123 or ..."
IF literalP THEN UserDataFromFontData[scratchData] ELSE LiteralDataFromFontData[scratchData];
}; -- Inner

errorRope: Rope.ROPE;
BEGIN
newData _ IF data=NIL THEN CreateFontData[] ELSE data; -- prepare newData
scratchData^ _ newData^;  -- operate on scratch copy in case failure happens while parsing
Inner[ ! IO.Error => {errorRope _ "FontName IO Error"; GOTO Error};
IO.EndOfStream => {errorRope _ "FontName EndOfStream"; GOTO Error};
Convert.Error  => {errorRope _ "FontName Convert Error"; GOTO Error};
GGParseIn.SyntaxError => {errorRope _ "FontName SyntaxError"; GOTO Error};
];
newData^ _ scratchData^;
EXITS
Error => ParseError[Rope.Concat[errorRope, ": See GGFontSampler.tioga for syntax"]];
END;
};

LiteralDataFromFontData: PUBLIC --ENTRY-- PROC [data: FontData] = {
Inner: PROC = {
HasEndDigits: PROC RETURNS [yep: BOOL _ FALSE] = {
ENABLE IO.Error, IO.EndOfStream, Convert.Error  => {
yep _ FALSE;
CONTINUE;
};
nameStream: IO.STREAM _ IO.RIS[scratchData.family]; -- Helvetica or cmbbi66
[] _ IO.GetTokenRope[nameStream, DigitProc]; -- toss the leading alpha characters
yep _ Convert.IntFromRope[IO.GetTokenRope[nameStream, NonDigitProc].token]>0; -- get any digit characters. Assume font names indicate positive values.
};
pressPrefix: Rope.ROPE _ "xerox/pressfonts/";
printPrefix: Rope.ROPE _ "xerox/xc1-2-2/";
screenPrefix: Rope.ROPE _ "xerox/tiogafonts/";
faceRope: Rope.ROPE;
SELECT TRUE FROM
Rope.Equal[scratchData.prefix, pressPrefix, FALSE] AND scratchData.faceKnown => {
faceRope _ SELECT TRUE FROM
scratchData.bold AND scratchData.italic => "-bir",
scratchData.bold  => "-brr",
scratchData.italic => "-mir",
ENDCASE => IF HasEndDigits[] THEN "" ELSE "-mrr";
};
Rope.Equal[scratchData.prefix, printPrefix, FALSE] AND scratchData.faceKnown => {
faceRope _ SELECT TRUE FROM
scratchData.bold AND scratchData.italic => "-bold-italic",
scratchData.bold  => "-bold",
scratchData.italic => "-italic",
ENDCASE => "";

};
Rope.Equal[scratchData.prefix, screenPrefix, FALSE] => { -- derive storedSize from the font name
storedSize: REAL _ 1.5;
nameStream: IO.STREAM _ IO.RIS[scratchData.userFSF]; -- Tioga10-BI or TERMINAL
[] _ IO.GetTokenRope[nameStream, DigitProc]; -- get the leading alpha characters
storedSize _ Convert.RealFromRope[IO.GetTokenRope[nameStream, NonDigitProc].token]; -- get any digit characters
IF Real.Float[Real.Fix[storedSize]] # storedSize THEN ParseError["StoredSize must be an integer for screen fonts"];
scratchData.storedSize _ storedSize;  -- so things come out the right size
faceRope _ SELECT TRUE FROM
scratchData.bold AND scratchData.italic => "BI",
scratchData.bold => "B",
scratchData.italic => "I",
ENDCASE => "";
};
ENDCASE => NULL;

scratchData.literalFSF _ Rope.Concat[scratchData.family, faceRope];
scratchData.literal _ Rope.Concat[scratchData.prefix, scratchData.literalFSF];
};

errorRope: Rope.ROPE;
BEGIN
IF data=NIL THEN ParseError["NIL Font Data cannot be parsed"];
scratchData^ _ data^;  -- operate on scratch copy in case failure happens while parsing
Inner[ ! IO.Error => {errorRope _ "FontName IO Error"; GOTO Error};
IO.EndOfStream => {errorRope _ "FontName EndOfStream"; GOTO Error};
Convert.Error  => {errorRope _ "FontName Convert Error"; GOTO Error};
GGParseIn.SyntaxError => {errorRope _ "FontName SyntaxError"; GOTO Error};
];
data^ _ scratchData^;
EXITS
Error => ParseError[Rope.Concat[errorRope, ": (example: Helvetica-BI 18 for SetPressFont)"]];
END;
};

UserDataFromFontData: PUBLIC --ENTRY-- PROC [data: FontData] = {
Inner: PROC = {
isBold, isItalic: BOOL _ FALSE;
userFace, literalFace: Rope.ROPE;
pressPrefix: Rope.ROPE _ "xerox/pressfonts/";
printPrefix: Rope.ROPE _ "xerox/xc1-2-2/";
screenPrefix: Rope.ROPE _ "xerox/tiogafonts/";
cmrFamily: Rope.ROPE _ "CMR";
dashBold: Rope.ROPE _ "-bold";
dashItalic: Rope.ROPE _ "-italic";

IF data.literal=NIL OR data.literalFSF=NIL THEN ERROR;
data.prefix  _ FileNames.Directory[data.literal]; -- xerox/...
SELECT TRUE FROM
Rope.Equal[data.prefix, pressPrefix, FALSE] => {
data.family _ Before[data.literalFSF, '-]; -- Helvetica or CMR
literalFace _ FileNames.Tail[data.literalFSF, '-]; -- bir or NIL
data.bold _ Rope.Equal[literalFace, "bir", FALSE] OR Rope.Equal[literalFace, "brr", FALSE];
data.italic _ Rope.Equal[literalFace, "bir", FALSE] OR Rope.Equal[literalFace, "mir", FALSE];
data.faceKnown _ TRUE;
data.comfortable _ TRUE;
userFace _ SELECT TRUE FROM
Rope.Equal[literalFace, "bir", FALSE] => "-BI",
Rope.Equal[literalFace, "brr", FALSE] => "-B",
Rope.Equal[literalFace, "mir", FALSE] => "-I",
Rope.Equal[literalFace, "mrr", FALSE] => "",
ENDCASE => "";
data.userFSF _ Rope.Concat[data.family, userFace];
};
Rope.Equal[data.prefix, printPrefix, FALSE] => {
data.family _ Before[data.literalFSF, '-]; -- Modern
data.bold _ Rope.Find[data.literalFSF, dashBold, 0, FALSE]#-1; -- has -bold
data.italic _ Rope.Find[data.literalFSF, dashItalic, 0, FALSE]#-1; -- has -italic
data.faceKnown _ TRUE;
data.comfortable _ TRUE;
userFace _ SELECT TRUE FROM
isBold AND isItalic => "-BI",
isBold => "-B",
isItalic => "-I",
ENDCASE => "";
data.userFSF _ Rope.Concat[data.family, userFace];
};
Rope.Equal[data.prefix, screenPrefix, FALSE] => { --Tioga10BI or TERMINAL
alphaRope, faceRope, storedSizeRope: Rope.ROPE;
storedSize: REAL _ 1.5; --, non integer default used below
nameStream: IO.STREAM _ IO.RIS[data.literalFSF]; -- Tioga10BI or TERMINAL
alphaRope _ IO.GetTokenRope[nameStream, DigitProc].token; -- get the leading alpha characters
storedSizeRope _ IO.GetTokenRope[nameStream, NonDigitProc ! IO.EndOfStream, IO.Error => CONTINUE;].token; -- get any digit characters
faceRope _ GGParseIn.ReadBlankAndWord[nameStream]; -- like BI or NIL
storedSize _ Convert.RealFromRope[storedSizeRope ! Convert.Error => CONTINUE;]; -- get any digit characters
IF Real.Float[Real.Fix[storedSize]] # storedSize THEN ParseError["StoredSize must be an integer for screen fonts"];
data.family _ Rope.Concat[alphaRope, storedSizeRope]; -- concat alpha and numeric parts
data.scale _ data.storedSize _ storedSize;  -- so things come out the right size. Note that this overides the user input during SetFontLiteral.  The rule we use is that if you use a TiogaFont with a size in the name, like Tioga10, we derive the storedSize and scale from the name.
IF faceRope#NIL THEN faceRope _ Rope.Concat["-", faceRope];
data.userFSF _ Rope.Concat[data.family, faceRope];
data.faceKnown _ TRUE;
data.comfortable _ TRUE;
FOR boldRope: LIST OF ROPE _ boldList, boldRope.rest UNTIL boldRope=NIL DO
IF data.bold THEN EXIT;
data.bold _ Rope.Find[faceRope, boldRope.first, 0, FALSE]#-1;
ENDLOOP;
FOR italicRope: LIST OF ROPE _ italicList, italicRope.rest UNTIL italicRope=NIL DO
IF data.italic THEN EXIT;
data.italic _ Rope.Find[faceRope, italicRope.first, 0, FALSE]#-1;
ENDLOOP;
};
ENDCASE => NULL;
};

errorRope: Rope.ROPE;
BEGIN
Inner[ ! IO.Error => {errorRope _ "IO.Error during font parse"; GOTO Error};
IO.EndOfStream => {errorRope _ "IO.EndOfStream during font parse"; GOTO Error};
Convert.Error  => {errorRope _ "Convert.Error during font parse"; GOTO Error};
GGParseIn.SyntaxError => {errorRope _ "GGParseIn.SyntaxError during font parse"; GOTO Error};
];
EXITS
Error => ParseError[errorRope];
END;
};

AlternateFont: PUBLIC PROC [data: FontData, font: ImagerFont.Font, op: ATOM] RETURNS [alternate: ImagerFont.Font] = {
IF font=NIL THEN RETURN[NIL];
IF op#$visible THEN RETURN[font];
IF Rope.Find[data.prefix, "pressfont", 0, FALSE]#-1 OR Rope.Find[data.prefix, "xc1", 0, FALSE]#-1 THEN {
dashIndex: INT _ Rope.Index[data.userFSF, 0, "-"]; -- is there a dash in the userFSF ??
altName: Rope.ROPE _ IF dashIndex=-1 THEN Rope.Cat["xerox/tiogafonts/", data.userFSF, "10"] ELSE Rope.Cat["xerox/tiogafonts/", Rope.Replace[data.userFSF, dashIndex, 1, "10"] ];
alternate _ ImagerFont.Scale[ImagerFont.Find[altName, substituteWithWarning ! Imager.Error, Imager.Warning => {alternate _ font; CONTINUE;}; ], 0.10];
}
ELSE alternate _ font;
};

Before: PROC [s: Rope.ROPE, char: CHAR] RETURNS [Rope.ROPE] = {
pos: INT _ 0;
len: INT _ s.Length[];
IF s=NIL THEN RETURN[NIL];
DO
IF s.Fetch[pos] = char THEN RETURN[s.Substr[0, pos]];
pos _ pos + 1;
IF pos = len THEN RETURN[s];
ENDLOOP;
};

OldParseFontData: PUBLIC PROC [inStream: IO.STREAM, prefixP, familyP, faceP, transformP, sizeP: BOOL _ FALSE] RETURNS [fail: BOOL, prefix, family, face: Rope.ROPE, transform: ImagerTransformation.Transformation, size: REAL _ 0.0] = {
ENABLE IO.Error, IO.EndOfStream, Convert.Error, GGParseIn.SyntaxError => {
fail _ TRUE;
CONTINUE; 
};
ReadWord: PROC [f: IO.STREAM] RETURNS [word: Rope.ROPE] = {
WordBreakProc: SAFE PROC [char: CHAR] RETURNS [IO.CharClass] = CHECKED {
SELECT char FROM
IO.TAB => RETURN [break];
IO.CR =>RETURN [break];
IO.SP => RETURN [break];
', => RETURN [break];
'] => RETURN [break];
') => RETURN [break];
ENDCASE => RETURN [other];
};
[word, ----] _ IO.GetTokenRope[f, WordBreakProc
!IO.EndOfStream => {word _ NIL; CONTINUE}];
};

nameStream: IO.STREAM;
fail _ FALSE;
IF prefixP THEN prefix _ IO.GetTokenRope[inStream, IO.IDProc].token; -- "xerox/myfonts/"
nameStream _ IO.RIS[IO.GetTokenRope[inStream, IO.IDProc].token]; -- "fontOne-BI"
IF familyP THEN family _ IO.GetTokenRope[nameStream, IO.TokenProc].token; -- "fontOne"
IF faceP THEN face _ ReadWord[nameStream]; -- "-BI" (or SP)
IF transformP THEN transform _ GGParseIn.ReadFactoredTransformation[inStream];
IF sizeP THEN size _ Convert.RealFromRope[IO.GetTokenRope[inStream, IO.IDProc].token]; -- "12"
};

defaultFontData: FontData _ NIL;

Init: PROC [] = {
defaultFontData _ CreateFontData[];
defaultFontData^ _ ["xerox/pressfonts/helvetica-mrr", "xerox/pressfonts/", "helvetica-mrr", "helvetica", "helvetica", ImagerTransformation.Scale[10.0], 1.0,1.0,1.0, TRUE, FALSE, FALSE, TRUE];
};

Init[];

END.

ΖGGFontImpl.mesa
Copyright (C) 1987 by Xerox Corporation.  All rights reserved.
Created by: Ken Pier, February 10, 1987 11:23:45 am PST
Last Edited by: Ken Pier, April 21, 1987 1:04:40 pm PDT

! ParseError[explanation] if something goes wrong.
Takes a FontData (if NIL, will create a new one), a stream, and BOOLEANS which denote what tokens to parse out of the stream and put in the newData.  If literalP is TRUE, then prefixP, familyP, faceP are ignored and the parser only attempts to find a literal string and parses it no further.  IF it is not literal, then if familyP then a userFSF string is parsed, and if faceP then the parser tries to extract face information from the userFSF. Fills in as many of the newData fields as possible.
IF transformP THEN scratchData.transform _ GGParseIn.ReadFactoredTransformation[inStream];
if you get here, you successfully parsed. Return good data.
! ParseError[explanation] if something goes wrong.
Constructs a literal font name from the (non NIL) values in the FontData and fills in the literal field in FontData.
This routine has the hairy specific knowledge of the formats of font names.
So far:	prefix		face 	fontName suffix
	xerox/xc1-2-2/	-B	-bold
	xerox/xc1-2-2/	-I	-italic
	xerox/xc1-2-2/	-BI	-bold-italic
	xerox/xc1-2-2/	-IB	-bold-italic
	xerox/xc1-2-2/	none	none

	xerox/pressfonts/	-B	-brr
	xerox/pressfonts/	-I	-mir
	xerox/pressfonts/	-BI	-bir
	xerox/pressfonts/	-IB	-bir
	xerox/pressfonts/	none	-mrr (unless CMR font)

	xerox/tiogafonts/	-B	Fix[size]B
	xerox/tiogafonts/	-I	Fix[size]I
	xerox/tiogafonts/	-BI	Fix[size]BI
	xerox/tiogafonts/	-IB	Fix[size]BI
	xerox/tiogafonts/	none	Fix[size]

For PressFonts Only; digits on the end indicate a "stored size" but ImagerFont produces a unit font with no face characters like -mrr; e.g. cmr60, cmbbi66, ... 
For TiogaFonts, scratchData.family includes storedSize, like Tioga10
if you get here, you successfully parsed. Return good data.
! ParseError[explanation] if something goes wrong.
Parses the literal font string in FontData, filling in the other values in FontData as possible.
xerox/pressfonts/Helvetica-bir or xerox/pressfonts/CMR63
xerox/xc1-2-2/Modern-bold-italic
Takes an existing font and fontData and returns an alternate. Only implemented op is $visible.
Returns font if no alternate is possible.  Remember that the alternate font has to work with the transformation in data.transform.
So, what do we do here, Bunkie?? Well, if we have a PressFont or an XC (Print) font, with a suitable data.userFSF, we concoct a TiogaFont name and see if we can Find it.  Otherwise, we punt.
Before returns the part of a rope before the first instance of char.
Before returns the entire rope if char is not found
Used to read in a rope which is data.
Κͺ˜™Jš
œ>™>Jšœ4Οk™7Jš
œ7™7J™—š
	˜	JšœQœ
˜`—J˜Icode˜Kš
Οn
œœ
œ
œJœ˜…
Kšœ
˜˜Kšœœœ
˜Kšœ
œ˜!Kšœ
œ˜'K˜—š
ž
œœ
œœœ
˜3K˜—Kš	œ
œ
œ
œœ˜+Kš	œœ
œ
œœ˜-Kš	œ	œ
œ
œœ/˜KK˜šž	œœ˜
procšœ˜Lšœ
œœ
œœ
œœ	˜'Lšœ*œ	˜9Lšœœ	˜—K˜K˜—šžœœ˜šœ˜Lšœ*œ	˜9Lšœœ	˜—K˜—K˜šžœœ
œ
œ˜AKšœžœ˜%K˜K˜—šžœœ
œ
œ˜8Kšœœ˜K˜K˜—šžœœ
œœ˜JK˜Kšœ	œ
œΟc#˜FK˜K˜—š
žœœ
œ&œœ˜cKš	œ
œ	œ
œ	œ˜<Kšœœ
œ˜"K•StartOfExpansion*[m: 
ImagerTransformation.Transformation]šœœ
œB˜_K˜—K˜Kšœœ˜.šž
œœ
œœœ
œSœœœ˜Ο
K™2Kš	œœ(œ]œ|œΙ
™πšžœœ˜š
žœœ
œœœ˜8Kšœœ˜(–>[s1: 
ROPE, s2: 
ROPE, pos1: 
INT _ 0, case: 
BOOL _ TRUE]š
œ
œ
œ
œœ	œ
˜FKš
œ œœ
œ
œ˜>Kšœ
˜—Kšœ
œ˜K˜K˜—šœ
œ˜KšœœœŸ˜_K–[s: 
ROPE, char: 
CHAR]šœ?Ÿ˜RK–[s: 
ROPE, char: 
CHAR]šœBŸ˜QKš
œ
˜
—šœ˜KšœœœœŸ˜Tšœ	œ˜Kšœ/Ÿ˜BKšœŸ˜-K˜—šœ	œ˜Kšœœ2˜OKšœŸ˜)Kšœ&Ÿ
˜0Kš
œ˜—šœœŸ.˜>Kšœ
œ
˜Kš
œœ
œœ
œŸ˜?Kšœœœ
Ÿ(˜SKš
œœœ
œœ
Ÿ˜o–>[s1: 
ROPE, s2: 
ROPE, pos1: 
INT _ 0, case: 
BOOL _ TRUE]š
œœ
œ
œœ
œ
˜JKšœœ
œ
˜Kšœ:œ˜DKšœ
˜—–>[s1: 
ROPE, s2: 
ROPE, pos1: 
INT _ 0, case: 
BOOL _ TRUE]š
œ
œ
œ
œœœ
˜RKšœœ
œ
˜ Kšœ>œ˜HKšœ
˜—Kšœœ
˜K˜—Kšœœ
˜K˜—K–
[f: 
STREAM]šœœH™ZK–
[f: 
STREAM]šœœK˜]K–[r: 
ROPE]š	œœ*œœŸ˜}K–[r: 
ROPE]š	œ
œ/œœŸ˜‡
K–[r: 
ROPE]š	œ
œ/œœŸ˜‡
Kšœ
œ#œ&˜]KšœŸ˜—K˜Kšœœ
˜š
˜Kš
œ
œœ
œœŸ˜IKšœŸ@˜Zšœ	œΟtœœ˜CKšœ 	œœ˜CKšœ  œœ˜EKšœ' œœ˜JKš
œ˜Kš
Οb;™;—K˜š
˜Kš
œT˜T—Kšœ
˜—K˜K˜—šžœœ
œ˜CK™2Kšœ-œD™tš
œK™KKš
œ%™%Kš
œ™Kš
œ™Kšœœ
™ Kšœœ
™ Kš
œ™K™Kš
œ™Kš
œ™Kšœœ™Kšœœ™Kšœ%œ™.K™Kš
œ ™ Kš
œ ™ Kšœœ
™"Kšœœ
™"Kš
œ!™!K™—šžœœ˜šžœœ
œœ˜2Kšœ“
Ÿœ™ 
šΠbkœ
œœ!˜4Kšœœ
˜Kšœ
˜	Kš
œ˜—Kš
œœ
œœ
œŸ˜KKšœœ&Ÿ$˜QKšœœ2ŸH˜–
K˜—Kšœœ˜-Kšœœ˜*Kšœœ˜.Kšœœ
˜šœ
œ
˜šœ,œœ˜Qšœœ
œ
˜Kšœœ˜2Kš
œ˜Kš
œ˜K–9[base: 
ROPE, start: 
INT _ 0, len: 
INT _ 2147483647]šœœœœ˜1—Kš
œ˜—šœ,œœ˜Qšœœ
œ
˜Kšœœ&˜:Kš
œ˜Kš
œ ˜ Kšœ˜—K˜Kš
œ˜—šœ-œŸ'˜`K™DKšœœ˜Kš
œœ
œœ
œŸ˜NKšœœ&Ÿ#˜PKšœ"œ0Ÿ˜oKšœ/œ>˜sKšœ&Ÿ$˜Jšœœ
œ
˜Kšœœ˜0Kš
œ˜Kš
œ˜Kšœ˜—Kš
œ˜—Kšœœ
˜K˜—Kš
œC˜CKš
œN˜NK˜K˜—Kšœœ
˜š
˜Kšœœ
œ.˜>KšœŸ@˜Wšœ	œ œœ˜CKšœ 	œœ˜CKšœ  œœ˜EKšœ' œœ˜JKš
œ˜Kš
‘;™;—K˜š
˜Kš
œ]˜]—Kšœ
˜—K˜K˜—šžœœ
œ˜@K™2Kš
œ`™`šžœœ˜Kšœœœ
˜Kšœœ
˜!Kšœœ˜-Kšœœ˜*Kšœœ˜.Kšœœ	˜Kšœœ˜Kšœœ
˜"K˜K–[path: 
ROPE]šœœ
œœ
œ
œ
˜6Kšœ2Ÿ˜>šœ
œ
˜šœ%œ˜0KšŸœŸ™8Kšœ+Ÿ˜>Kšœ3Ÿ
˜@Kšœ+œœ œ˜[Kšœ-œœ œ˜]Kšœœ
˜Kšœœ
˜šœœ
œ
˜Kšœœ˜/Kšœœ
˜.Kšœœ
˜.Kšœœ˜,Kšœ˜—Kš
œ2˜2K˜—šœ%œ˜0KšœŸ™ Kšœ+Ÿ	˜4Kšœ4œŸ˜KKšœ8œŸ˜QKšœœ
˜Kšœœ
˜šœœ
œ
˜Kšœœ˜Kš
œ˜Kš
œ˜Kšœ˜—Kš
œ2˜2K˜—šœ&œŸ˜IKšœ*œ
˜/KšœœŸ"˜:Kš
œœ
œœ
œŸ˜IKšœœ,Ÿ#˜]Kš
œœ)œœ
œ
Ÿ˜…
Kšœ3Ÿ˜DKšœDœŸ˜kKšœ/œ>˜sKšœ6Ÿ!˜WKšœ,Ÿμ
˜˜Kšœ
œ
œ'˜;Kš
œ2˜2Kšœœ
˜Kšœœ
˜–>[s1: 
ROPE, s2: 
ROPE, pos1: 
INT _ 0, case: 
BOOL _ TRUE]š
œœ
œ
œœ
œ
˜JKšœœ
œ
˜Kšœ3œ˜=Kšœ
˜—–>[s1: 
ROPE, s2: 
ROPE, pos1: 
INT _ 0, case: 
BOOL _ TRUE]š
œ
œ
œ
œœœ
˜RKšœ
œ
œ
˜Kšœ7œ˜AKšœ
˜—K˜—Kšœœ
˜—K˜K˜—Kšœœ
˜š
˜šœ	œ œœ˜LKšœ œ œœ˜OKšœ  œœ˜NKšœ< œœ˜]Kš
œ˜—š
˜Kš
œ˜—Kšœ
˜—K˜K˜—šž
œœ-œœ!˜uJ™^J™‚
Kš
œœ
œ
œ
œ˜Kšœ
œ
œ˜!KšœDœx™Ύ
š
œ(œœ"œœ˜hKšœœ%Ÿ$˜WKš	œœœœ3œP˜°
Kšœ‘œ‘œ)žœ	žœœ
˜–
K˜
—Kšœ˜J˜J˜—šžœœ
œœœœ˜?K™DK™3Kšœœ˜
Kšœœ˜Kš
œœ
œ
œ
œ˜š
˜Kšœœ
œ˜5K˜Kšœœ
œ˜Kšœ
˜—K˜K˜—šžœœ
œœ
œ.œœœœœ8œ˜ι
šœ
œœ7˜JKšœœ
˜Kšœ˜
Kš
œ˜—šžœœœ
œœ
œ˜;Lš
œ%™%šž
œœ
œœœœœ˜Hšœ˜Lšœ
œœ	˜Lšœ
œœ	˜Lšœ
œœ	˜Lšœœ	˜Lšœœ	˜Lšœœ	˜Lšœœ	˜—Lš
œ˜—šœŸœœ˜/Lšœ
œœœ˜+—Lš
œ˜L˜—Kšœœ
œ
˜Kšœœ
˜
Kš	œ	œ
œœŸ˜XKš
œ
œ
œ
œœŸ˜PKš	œ	œ
œœŸ˜VKšœœŸ˜;K–
[f: 
STREAM]šœœ<˜NK–[r: 
ROPE]š	œœœœŸ˜^Jš
œ˜—J˜Kšœœ
˜ K˜šžœœ˜K˜#Kš	œ₯
œœœœ˜Ώ
K˜—K˜K˜K˜Kšœ
˜—…—6Tr