adding syntax coloring (issue #46 )

3 files changed, 420 insertions, 0 deletions

diff --git a/Source/CMTextView.m b/Source/CMTextView.m
index 7ca2f782..12b253da 100644
--- a/Source/CMTextView.m
+++ b/Source/CMTextView.m
@@ -24,6 +24,17 @@
 #import "CMTextView.h"
 #import "SPStringAdditions.h"
 
+/*
+all the extern variables and prototypes required for flex (syntax highlighting)
+*/
+#import "tokens.h"
+extern int yylex();
+extern int yyuoffset, yyuleng;
+typedef struct yy_buffer_state *YY_BUFFER_STATE;
+void yy_switch_to_buffer(YY_BUFFER_STATE);
+YY_BUFFER_STATE yy_scan_string (const char *);
+
+
 @implementation CMTextView
 
 - (NSArray *)completionsForPartialWordRange:(NSRange)charRange indexOfSelectedItem:(int *)index
@@ -350,4 +361,75 @@
 	nil];
 }
 
+/*******************
+SYNTAX HIGHLIGHTING!
+*******************/
+- (void)awakeFromNib
+/*
+sets self as delegate for the textView's textStorage to enable syntax highlighting
+*/
+{
+    [[self textStorage] setDelegate:self];
+}
+
+- (void)textStorageDidProcessEditing:(NSNotification *)notification
+/*
+performs syntax highlighting
+this method recolors the entire text on every keypress. for a single sql query,
+that should be no problem, but it could be a nuisance when editing large sql queries
+*/
+{
+    NSTextStorage *textStore = [notification object];
+    
+    //make sure that the notification is from the correct textStorage object
+    if (textStore==[self textStorage])
+    {
+    
+        NSColor *reservedColor = [NSColor blueColor];
+        NSColor *quoteColor    = [NSColor grayColor];
+        NSColor *commentColor  = [NSColor redColor ];
+        
+        NSString *string = [textStore string];
+        unsigned int length = [string length];
+        int token;
+        NSRange textRange, tokenRange;
+        
+        textRange = NSMakeRange(0, length);
+        
+        //first remove the old colors
+        [textStore removeAttribute:NSForegroundColorAttributeName range:textRange];
+
+        //initialise flex
+        yyuoffset = 0; yyuleng = 0;
+        yy_switch_to_buffer(yy_scan_string([[textStore string] UTF8String]));
+
+        //now loop through all the tokens
+        while (token=yylex()){
+            tokenRange = NSMakeRange(yyuoffset, yyuleng);  //convert the result from flex to an NSRange
+            tokenRange = NSIntersectionRange(tokenRange, textRange); // make sure that tokenRange is valid (and therefore within textRange)
+                                                                     // otherwise a bug in the lex code could cause the the TextView to crash
+            
+            switch (token) {
+            case SPT_SINGLE_QUOTED_TEXT:
+            case SPT_DOUBLE_QUOTED_TEXT:
+                [textStore addAttribute: NSForegroundColorAttributeName
+                                  value: quoteColor
+                                  range: tokenRange ];
+                break;
+            case SPT_RESERVED_WORD:
+                [textStore addAttribute: NSForegroundColorAttributeName
+                                  value: reservedColor
+                                  range: tokenRange ];
+                break;
+            case SPT_COMMENT:
+                [textStore addAttribute: NSForegroundColorAttributeName
+                                  value: commentColor
+                                  range: tokenRange ];
+                break;
+        }}
+        
+    }
+}
+
+
 @end
diff --git a/Source/tokens.h b/Source/tokens.h
new file mode 100644
index 00000000..81ea830f
--- /dev/null
+++ b/Source/tokens.h
@@ -0,0 +1,15 @@
+/*
+ *  tokens.h
+ *  sequel-pro
+ *
+ *  Created by Jakob on 3/15/09.
+ *  
+ *  This file defines all the tokens used for parsing the source code
+ */
+
+#define SPT_DOUBLE_QUOTED_TEXT   1
+#define SPT_SINGLE_QUOTED_TEXT   2
+#define SPT_BACKTICK_QUOTED_TEXT 3
+#define SPT_RESERVED_WORD        4
+#define SPT_COMMENT              5
+#define SPT_OTHER                6
\ No newline at end of file
diff --git a/Source/tokens.l b/Source/tokens.l
new file mode 100644
index 00000000..941378c3
--- /dev/null
+++ b/Source/tokens.l
@@ -0,0 +1,323 @@
+%{
+#import "tokens.h"
+int utf8strlen(const char *s);
+
+#define YY_NO_UNPUT
+
+int yyuoffset, yyuleng;
+  
+#define YY_USER_ACTION { yyuoffset += yyuleng; yyuleng = utf8strlen(yytext); }
+%}
+%option noyywrap
+%option case-insensitive
+s		[ \t\n]+
+word	[a-z_0-9]
+nonword	[^a-z_0-9#]
+%x comment
+%%
+\"([^"\\]|\\(.|\n))*\"?				{ return SPT_DOUBLE_QUOTED_TEXT;   } /* double quoted strings */
+'([^'\\]|\\(.|\n))*'?				{ return SPT_SINGLE_QUOTED_TEXT;   } /* single quoted strings */
+`[^`]*`?						{ return SPT_BACKTICK_QUOTED_TEXT; } /* identifier quoting    */
+"/*"							{ BEGIN(comment); return SPT_COMMENT; }
+<comment>[^*]* 				    { return SPT_COMMENT; }
+<comment>"*"+                   { return SPT_COMMENT; }
+<comment>"*"+"/" 				{ BEGIN(INITIAL); return SPT_COMMENT; }
+							 
+#[^\n]*\n?				        |									 
+--[ \t][^\n]*\n?						{ return SPT_COMMENT;  }		     /* -- Comments           */
+{s}								{ return SPT_OTHER;    } 								 /* ignore spaces         */
+ADD                         	|
+ALL                         	|
+ALTER{s}TABLE               	|
+ALTER{s}VIEW                	|
+ALTER{s}SCHEMA              	|
+ALTER{s}SCHEMA              	|
+ALTER{s}FUNCTION            	|
+ALTER{s}COLUMN              	|
+ALTER{s}DATABASE            	|
+ALTER{s}PROCEDURE           	|
+ANALYZE                     	|
+AND                         	|
+ASC                         	|
+ASENSITIVE                  	|
+BEFORE                      	|
+BETWEEN                     	|
+BIGINT                      	|
+BINARY                      	|
+BLOB                        	|
+BOTH                        	|
+CALL                        	|
+CASCADE                     	|
+CASE                        	|
+CHANGE                      	|
+CHAR                        	|
+CHARACTER                   	|
+CHECK                       	|
+COLLATE                     	|
+COLUMN                      	|
+COLUMNS                     	|
+CONDITION                   	|
+CONNECTION                  	|
+CONSTRAINT                  	|
+CONTINUE                    	|
+CONVERT                     	|
+CREATE{s}VIEW               	|
+CREATE{s}INDEX              	|
+CREATE{s}FUNCTION           	|
+CREATE{s}DATABASE           	|
+CREATE{s}PROCEDURE          	|
+CREATE{s}SCHEMA             	|
+CREATE{s}TRIGGER            	|
+CREATE{s}TABLE              	|
+CREATE{s}USER               	|
+CROSS                       	|
+CURRENT_DATE                	|
+CURRENT_TIME                	|
+CURRENT_TIMESTAMP           	|
+CURRENT_USER                	|
+CURSOR                      	|
+DATABASE                    	|
+DATABASES                   	|
+DAY_HOUR                    	|
+DAY_MICROSECOND             	|
+DAY_MINUTE                  	|
+DAY_SECOND                  	|
+DEC                         	|
+DECIMAL                     	|
+DECLARE                     	|
+DEFAULT                     	|
+DELAYED                     	|
+DELETE                      	|
+DESC                        	|
+DESCRIBE                    	|
+DETERMINISTIC               	|
+DISTINCT                    	|
+DISTINCTROW                 	|
+DIV                         	|
+DOUBLE                      	|
+DROP{s}TABLE                	|
+DROP{s}TRIGGER              	|
+DROP{s}VIEW                 	|
+DROP{s}SCHEMA               	|
+DROP{s}USER                 	|
+DROP{s}PROCEDURE            	|
+DROP{s}FUNCTION             	|
+DROP{s}FOREIGN{s}KEY        	|
+DROP{s}INDEX                	|
+DROP{s}PREPARE              	|
+DROP{s}PRIMARY{s}KEY        	|
+DROP{s}DATABASE             	|
+DUAL                        	|
+EACH                        	|
+ELSE                        	|
+ELSEIF                      	|
+ENCLOSED                    	|
+ESCAPED                     	|
+EXISTS                      	|
+EXIT                        	|
+EXPLAIN                     	|
+FALSE                       	|
+FETCH                       	|
+FIELDS                      	|
+FLOAT                       	|
+FOR                         	|
+FORCE                       	|
+FOREIGN{s}KEY               	|
+FOUND                       	|
+FROM                        	|
+FULLTEXT                    	|
+GOTO                        	|
+GRANT                       	|
+GROUP                       	|
+HAVING                      	|
+HIGH_PRIORITY               	|
+HOUR_MICROSECOND            	|
+HOUR_MINUTE                 	|
+HOUR_SECOND                 	|
+IGNORE                      	|
+INDEX                       	|
+INFILE                      	|
+INNER                       	|
+INOUT                       	|
+INSENSITIVE                 	|
+INSERT                      	|
+INT                         	|
+INTEGER                     	|
+INTERVAL                    	|
+INTO                        	|
+ITERATE                     	|
+JOIN                        	|
+KEY                         	|
+KEYS                        	|
+KILL                        	|
+LEADING                     	|
+LEAVE                       	|
+LEFT                        	|
+LIKE                        	|
+LIMIT                       	|
+LINES                       	|
+LOAD                        	|
+LOCALTIME                   	|
+LOCALTIMESTAMP              	|
+LOCK                        	|
+LONG                        	|
+LONGBLOB                    	|
+LONGTEXT                    	|
+LOOP                        	|
+LOW_PRIORITY                	|
+MATCH                       	|
+MEDIUMBLOB                  	|
+MEDIUMINT                   	|
+MEDIUMTEXT                  	|
+MIDDLEINT                   	|
+MINUTE_MICROSECOND          	|
+MINUTE_SECOND               	|
+MOD                         	|
+NATURAL                     	|
+NOT                         	|
+NO_WRITE_TO_BINLOG          	|
+NULL                        	|
+NUMERIC                     	|
+ON                          	|
+OPTIMIZE                    	|
+OPTION                      	|
+OPTIONALLY                  	|
+ORDER                       	|
+OUT                         	|
+OUTER                       	|
+OUTFILE                     	|
+PRECISION                   	|
+PRIMARY                     	|
+PRIVILEGES                  	|
+PROCEDURE                   	|
+PURGE                       	|
+READ                        	|
+REAL                        	|
+REFERENCES                  	|
+REGEXP                      	|
+RENAME                      	|
+REPEAT                      	|
+REPLACE                     	|
+REQUIRE                     	|
+RESTRICT                    	|
+RETURN                      	|
+REVOKE                      	|
+RIGHT                       	|
+RLIKE                       	|
+SECOND_MICROSECOND          	|
+SELECT                      	|
+SENSITIVE                   	|
+SEPARATOR                   	|
+SET                         	|
+SHOW{s}PROCEDURE{s}STATUS   	|
+SHOW{s}PROCESSLIST          	|
+SHOW{s}SCHEMAS              	|
+SHOW{s}SLAVE{s}HOSTS        	|
+SHOW{s}PRIVILEGES           	|
+SHOW{s}OPEN{s}TABLES        	|
+SHOW{s}MASTER{s}STATUS      	|
+SHOW{s}SLAVE{s}STATUS       	|
+SHOW{s}PLUGIN               	|
+SHOW{s}STORAGE{s}ENGINES    	|
+SHOW{s}VARIABLES            	|
+SHOW{s}WARNINGS             	|
+SHOW{s}TRIGGERS             	|
+SHOW{s}TABLES               	|
+SHOW{s}MASTER{s}LOGS        	|
+SHOW{s}TABLE{s}STATUS       	|
+SHOW{s}TABLE{s}TYPES        	|
+SHOW{s}STATUS               	|
+SHOW{s}INNODB{s}STATUS      	|
+SHOW{s}CREATE{s}DATABASE    	|
+SHOW{s}CREATE{s}FUNCTION    	|
+SHOW{s}CREATE{s}PROCEDURE   	|
+SHOW{s}CREATE{s}SCHEMA      	|
+SHOW{s}COLUMNS              	|
+SHOW{s}COLLATION            	|
+SHOW{s}BINARY{s}LOGS        	|
+SHOW{s}BINLOG{s}EVENTS      	|
+SHOW{s}CHARACTER{s}SET      	|
+SHOW{s}CREATE{s}TABLE       	|
+SHOW{s}CREATE{s}VIEW        	|
+SHOW{s}FUNCTION{s}STATUS    	|
+SHOW{s}GRANTS               	|
+SHOW{s}INDEX                	|
+SHOW{s}FIELDS               	|
+SHOW{s}ERRORS               	|
+SHOW{s}DATABASES            	|
+SHOW{s}ENGINE               	|
+SHOW{s}ENGINES              	|
+SHOW{s}KEYS                 	|
+SMALLINT                    	|
+SONAME                      	|
+SPATIAL                     	|
+SPECIFIC                    	|
+SQL                         	|
+SQLEXCEPTION                	|
+SQLSTATE                    	|
+SQLWARNING                  	|
+SQL_BIG_RESULT              	|
+SQL_CALC_FOUND_ROWS         	|
+SQL_SMALL_RESULT            	|
+SSL                         	|
+STARTING                    	|
+STRAIGHT_JOIN               	|
+TABLE                       	|
+TABLES                      	|
+TERMINATED                  	|
+THEN                        	|
+TINYBLOB                    	|
+TINYINT                     	|
+TINYTEXT                    	|
+TRAILING                    	|
+TRIGGER                     	|
+TRUE                        	|
+UNDO                        	|
+UNION                       	|
+UNIQUE                      	|
+UNLOCK                      	|
+UNSIGNED                    	|
+UPDATE                      	|
+USAGE                       	|
+USE                         	|
+USING                       	|
+UTC_DATE                    	|
+UTC_TIME                    	|
+UTC_TIMESTAMP               	|
+VALUES                      	|
+VARBINARY                   	|
+VARCHAR                     	|
+VARCHARACTER                	|
+VARYING                     	|
+WHEN                        	|
+WHERE                       	|
+WHILE                       	|
+WITH                        	|
+WRITE                       	|
+XOR                         	|
+YEAR_MONTH                  	|
+ZEROFILL            			{ return SPT_RESERVED_WORD;        } /* all the mysql reserved words  */
+{word}+							{ return SPT_OTHER;                } /* return any word  */
+{nonword}						{ return SPT_OTHER;                } /* return anything else  */
+                    			
+<<EOF>>   						{
+                                    BEGIN(INITIAL);
+            						yy_delete_buffer(YY_CURRENT_BUFFER);
+            						return 0;
+          						}
+%%
+
+int utf8strlen(const char *s)
+/*
+This simple function calculates the length of an UTF8 string in characters (not bytes)
+It's not especially fast, but it's easy to comprehend
+*/
+{
+    int j=0;
+	while (*s)
+    {
+		if ((*s & 0xC0) != 0x80) j++;
+		s++;
+	}
+	return (j);
+}
\ No newline at end of file