Patch from Roy Wood to allow changing the characters that are in the

whitespace class.

6 files changed, 75 insertions, 9 deletions

diff --git a/doc/ScintillaDoc.html b/doc/ScintillaDoc.html
index bbdf57a90..221d141db 100644
--- a/doc/ScintillaDoc.html
+++ b/doc/ScintillaDoc.html
@@ -2217,6 +2217,8 @@ struct TextToFind {
      <a class="message" href="#SCI_GETCODEPAGE">SCI_GETCODEPAGE</a><br />
      <a class="message" href="#SCI_SETWORDCHARS">SCI_SETWORDCHARS(&lt;unused&gt;, const char
     *chars)</a><br />
+	 <a class="message" href="#SCI_SETWHITESPACECHARS">SCI_SETWHITESPACECHARS(&lt;unused&gt;, const char
+    *chars)</a><br />
      <a class="message" href="#SCI_GRABFOCUS">SCI_GRABFOCUS</a><br />
      <a class="message" href="#SCI_SETFOCUS">SCI_SETFOCUS(bool focus)</a><br />
      <a class="message" href="#SCI_GETFOCUS">SCI_GETFOCUS</a><br />
@@ -2308,6 +2310,13 @@ struct TextToFind {
     use:<br />
      <code>SCI_SETWORDCHARS(0, "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789")</code>;</p>
 
+    <p><b id="SCI_SETWHITESPACECHARS">SCI_SETWHITESPACECHARS(&lt;unused&gt;, const char *chars)</b><br />
+     Similar to <code>SCI_SETWORDCHARS</code>, this message allows the user to define which chars Scintilla considers
+	as whitespace.  If <code>chars</code> is null then the default set, space and all char codes less than
+	0x20, are used.  Setting the whitespace chars allows the user to fine-tune Scintilla's behaviour doing
+	such things as moving the cursor to the start or end of a word; for example, by defining punctuation chars
+	as whitespace, they will be skipped over when the user presses ctrl+left or ctrl+right.
+
     <p><b id="SCI_GRABFOCUS">SCI_GRABFOCUS</b><br />
      <b id="SCI_SETFOCUS">SCI_SETFOCUS(bool focus)</b><br />
      <b id="SCI_GETFOCUS">SCI_GETFOCUS</b><br />
diff --git a/include/Scintilla.h b/include/Scintilla.h
index d0b788e47..44a608d9f 100644
--- a/include/Scintilla.h
+++ b/include/Scintilla.h
@@ -567,6 +567,7 @@ typedef sptr_t (*SciFnDirect)(sptr_t ptr, unsigned int iMessage, uptr_t wParam,
 #define SCI_WORDLEFTENDEXTEND 2440
 #define SCI_WORDRIGHTEND 2441
 #define SCI_WORDRIGHTENDEXTEND 2442
+#define SCI_SETWHITESPACECHARS 2443
 #define SCI_STARTRECORD 3001
 #define SCI_STOPRECORD 3002
 #define SCI_SETLEXER 4001
diff --git a/include/Scintilla.iface b/include/Scintilla.iface
index 1e4d0db34..93173d07e 100644
--- a/include/Scintilla.iface
+++ b/include/Scintilla.iface
@@ -1538,6 +1538,8 @@ fun void WordRightEnd=2441(,)
 # Move caret right one word, position cursor at end of word, extending selection to new caret position.
 fun void WordRightEndExtend=2442(,)
 
+# Set the set of characters making up whitespace for when moving or selecting by word.
+set void SetWhitespaceChars=2443(, string characters)
 
 # Start notifying the container of all key presses and commands.
 fun void StartRecord=3001(,)
diff --git a/src/Document.cxx b/src/Document.cxx
index 130a920e1..53d01d2b8 100644
--- a/src/Document.cxx
+++ b/src/Document.cxx
@@ -50,7 +50,7 @@ Document::Document() {
 	stylingBits = 5;
 	stylingBitsMask = 0x1F;
 	stylingMask = 0;
-	SetWordChars(0);
+	SetDefaultCharClasses();
 	endStyled = 0;
 	styleClock = 0;
 	enteredCount = 0;
@@ -1119,29 +1119,74 @@ void Document::ChangeCase(Range r, bool makeUpperCase) {
 	}
 }
 
-void Document::SetWordChars(unsigned char *chars) {
+
+void Document::SetDefaultCharClasses() {
 	int ch;
+	// Initialize all char classes to default values
 	for (ch = 0; ch < 256; ch++) {
 		if (ch == '\r' || ch == '\n')
 			charClass[ch] = ccNewLine;
 		else if (ch < 0x20 || ch == ' ')
 			charClass[ch] = ccSpace;
+		else if (ch >= 0x80 || isalnum(ch) || ch == '_')
+			charClass[ch] = ccWord;
 		else
 			charClass[ch] = ccPunctuation;
 	}
+}
+
+void Document::SetCharClasses(unsigned char *chars, charClassification newCharClass) {
+	int ch;
+	// The old code always reset all chars to their default charClass and then applied the new charClass 
+	// to a specific set of chars, so the most reasonable way to honour the old promise is to reset all chars
+	// that currently are of class newCharClass, and then apply the newCharClass to any specified chars.  Remember,
+	// the point is to allow the caller to explicitly define all chars which are to be of class newCharClass.
+	// The only other tricky thing is that the old promise was that if the user passes in NULL for the chars
+	// parameter, we are supposed to reset all chars that default to newCharClass, which is not the
+	// same as resetting the char class of any chars that are currently of newCharClass (user might have altered
+	// some char classes from their default class).
+	
+	// If all of that seems a little complex, well, it is.  If we could break the old promise and make some
+	// simplifying assumptions, this would all be waaaaay simpler.  This way though, Scintilla will do the right
+	// thing if lots of changes of charClass are applied during the lifetime of the Document object (not very likely,
+	// but as long as I'm doing this, might as well get it right).
+	
+	// So, first reset the char class of any chars that are currently of the class newCharClass
+	for (ch = 0; ch < 256; ch++) {
+		if (charClass[ch] == newCharClass) {
+			if (ch == '\r' || ch == '\n')
+				charClass[ch] = ccNewLine;
+			else if (ch < 0x20 || ch == ' ')
+				charClass[ch] = ccSpace;
+			else if (ch >= 0x80 || isalnum(ch) || ch == '_')
+				charClass[ch] = ccWord;
+			else
+				charClass[ch] = ccPunctuation;
+		}
+	}
+	
+	// Next, apply the newCharClass to the specifed chars
 	if (chars) {
 		while (*chars) {
-			charClass[*chars] = ccWord;
+			charClass[*chars] = newCharClass;
 			chars++;
 		}
 	} else {
+		// If user passed NULL for chars, reinitialize only the specified class of chars
 		for (ch = 0; ch < 256; ch++) {
-			if (ch >= 0x80 || isalnum(ch) || ch == '_')
+			if ((ch == '\r' || ch == '\n') && newCharClass == ccNewLine)
+				charClass[ch] = ccNewLine;
+			else if ((ch < 0x20 || ch == ' ') && newCharClass == ccSpace)
+				charClass[ch] = ccSpace;
+			else if ((ch >= 0x80 || isalnum(ch) || ch == '_') && newCharClass == ccWord)
 				charClass[ch] = ccWord;
+			else if (newCharClass == ccPunctuation)
+				charClass[ch] = ccPunctuation;
 		}
 	}
 }
 
+
 void Document::SetStylingBits(int bits) {
 	stylingBits = bits;
 	stylingBitsMask = 0;
diff --git a/src/Document.h b/src/Document.h
index f811bf077..dc9e38e21 100644
--- a/src/Document.h
+++ b/src/Document.h
@@ -87,11 +87,12 @@ public:
 			userData = 0;
 		}
 	};
+	
+	enum charClassification { ccSpace, ccNewLine, ccWord, ccPunctuation };
 
 private:
 	int refCount;
 	CellBuffer cb;
-	enum charClassification { ccSpace, ccNewLine, ccWord, ccPunctuation };
 	charClassification charClass[256];
 	char stylingMask;
 	int endStyled;
@@ -199,8 +200,9 @@ public:
 	int LinesTotal();
 
 	void ChangeCase(Range r, bool makeUpperCase);
-
-	void SetWordChars(unsigned char *chars);
+	
+	void SetDefaultCharClasses();
+	void SetCharClasses(unsigned char *chars, charClassification newCharClass);
 	void SetStylingBits(int bits);
 	void StartStyling(int position, char mask);
 	bool SetStyleFor(int length, char style);
diff --git a/src/Editor.cxx b/src/Editor.cxx
index b1811e075..c4e6860e9 100644
--- a/src/Editor.cxx
+++ b/src/Editor.cxx
@@ -1122,7 +1122,7 @@ This way, we favour the displaying of useful information: the begining of lines,
 where most code reside, and the lines after the caret, eg. the body of a function.
 
      |        |       |      |                                            |
-slop | strict | jumps | even | Caret can go to the margin                 | When reaching limit (caret going out of
+slop | strict | jumps | even | Caret can go to the margin                 | When reaching limitƯ(caret going out of
      |        |       |      |                                            | visibility or going into the UZ) display is...
 -----+--------+-------+------+--------------------------------------------+--------------------------------------------------------------
   0  |   0    |   0   |   0  | Yes                                        | moved to put caret on top/on right
@@ -5733,7 +5733,14 @@ sptr_t Editor::WndProc(unsigned int iMessage, uptr_t wParam, sptr_t lParam) {
 	case SCI_SETWORDCHARS: {
 			if (lParam == 0)
 				return 0;
-			pdoc->SetWordChars(reinterpret_cast<unsigned char *>(lParam));
+			pdoc->SetCharClasses(reinterpret_cast<unsigned char *>(lParam), Document::ccWord);
+		}
+		break;
+
+	case SCI_SETWHITESPACECHARS: {
+			if (lParam == 0)
+				return 0;
+			pdoc->SetCharClasses(reinterpret_cast<unsigned char *>(lParam), Document::ccSpace);
 		}
 		break;