1 files changed, 136 insertions, 116 deletions
diff --git a/src/xmlreader.h b/src/xmlreader.h
index 708a386..c8f7202 100644
--- a/src/xmlreader.h
+++ b/src/xmlreader.h
@@ -1,121 +1,141 @@
-#ifndef XML_READER_H
+#ifndef XMLREADER
-#define XML_READER_H
+#define XMLREADER
+#include <stdio.h>
+#include "xmldocument.h"
+#include "flexbuf.h"
+#include "hashtable.h"
+#include "staticstring.h"
+/**
+ * Takes care of reading in xml formatted data from a file.  This could/should
+ * be made more arbitrary in the future so that we can read the data from any
+ * source.  This is actually made quite simple already since all data read in
+ * is handled by one single helper function and then palced into a FlexBuf for
+ * easy access by the other functions.  The FlexBuf also allows for block
+ * reading from disk, which improves speed by a noticable amount.
+ * <br>
+ * There are also some extra features implemented that allow you to break the
+ * standard XML reader specs and eliminate leading and trailing whitespace in
+ * all read content.  This is useful in situations where you allow additional
+ * whitespace in the files to make them easily human readable.  The resturned
+ * content will be NULL in sitautions where all content between nodes was
+ * stripped.
+ *@author Mike Buland
+ */
+class XmlReader : public XmlDocument
+{
+public:
+        /**
+         * Create a standard XmlReader.  The optional parameter bStrip allows you to
+         * create a reader that will strip out all leading and trailing whitespace
+         * in content, a-la html.
+         *@param bStrip Strip out leading and trailing whitespace?
+         */
+        XmlReader( bool bStrip=false );
-#include <stdint.h>
+        /**
-#include "bu/stream.h"
+         * Destroy this XmlReader.
-#include "bu/fstring.h"
+         */
-#include "bu/xmlnode.h"
+        virtual ~XmlReader();
+        /**
+         * Build a document based on some kind of input.  This is called
+         * automatically by the constructor.
+         */
+        bool buildDoc();
+private:
+        /**
+         * This is called by the low level automoton in order to get the next
+         * character.  This function should return a character at the current
+         * position plus nIndex, but does not increment the current character.
+         *@param nIndex The index of the character from the current stream position.
+         *@returns A single character at the requested position, or 0 for end of
+         * stream.
+         */
+        virtual char getChar( int nIndex = 0 ) = 0;
+        /**
+         * Called to increment the current stream position by a single character.
+         */
+        virtual void usedChar( int nAmnt = 1) = 0;
+        /**
+         * Automoton function: is whitespace.
+         *@param chr A character
+         *@returns True if chr is whitespace, false otherwise.
+         */
+        bool isws( char chr );
+        /**
+         * Automoton function: ws.  Skips sections of whitespace.
+         *@returns True if everything was ok, False for end of stream.
+         */
+        bool ws();
+        /**
+         * Automoton function: node.  Processes an XmlNode
+         *@returns True if everything was ok, False for end of stream.
+         */
+        bool node();
-namespace Bu
-{
        /**
-         * An Xml 1.1 reader.  I've decided to write this, this time, based on the
+         * Automoton function: startNode.  Processes the begining of a node.
-         * official W3C reccomendation, now included with the source code.  I've
+         *@returns True if everything was ok, False for end of stream.
-         * named the productions in the parser states the same as in that document,
+         */
-         * which may make them easier to find, etc, although possibly slightly less
+        bool startNode();
-         * optimized than writing my own reduced grammer.
-         *
+        /**
-         * Below I will list differences between my parser and the official standard
+         * Automoton function: name.  Processes the name of a node.
-         * as I come up with them.
+         *@returns True if everything was ok, False for end of stream.
-         *  - Encoding and Standalone headings are ignored for the moment. (4.3.3,
+         */
-         *    2.9)
+        bool name();
-         *  - The standalone heading attribute can have any standard whitespace
-         *    before it (the specs say only spaces, no newlines). (2.9)
+        /**
-         *  - Since standalone is ignored, it is currently allowed to have any
+         * Automoton function: textDecl.  Processes the xml text decleration, if
-         *    value (should be restricted to "yes" or "no"). (2.9)
+         * there is one.
-         *  - Currently only UTF-8 / ascii are parsed.
+         */
-         *  - [optional] The content of comments is thrown away. (2.5)
+        void textDecl();
-         *  - The content of processing instruction blocks is parsed properly, but
-         *    thrown away. (2.6)
+        /**
-         */
+         * Automoton function: entity.  Processes an entity from the header.
-        class XmlReader
+         */
-        {
+        void entity();
-        public:
-                XmlReader( Bu::Stream &sIn );
+        /**
-                virtual ~XmlReader();
+         * Adds an entity to the list, if it doesn't already exist.
+         *@param name The name of the entity
-                XmlNode *read();
+         *@param value The value of the entity
+         */
-        private:
+        void addEntity( const char *name, const char *value );
-                Bu::Stream &sIn;
-                Bu::FString sBuf;
+        StaticString *getEscape();
-        private: // Helpers
+        /**
-                const char *lookahead( int nAmnt );
+         * Automoton function: paramlist.  Processes a list of node params.
-                void burn( int nAmnt );
+         *@returns True if everything was ok, False for end of stream.
-                void checkString( const char *str, int nLen );
+         */
+        bool paramlist();
-        private: // States
-                /**
+        /**
-                 * The headers, etc.
+         * Automoton function: param.  Processes a single parameter.
-                 */
+         *@returns True if everything was ok, False for end of stream.
-                void prolog();
+         */
+        bool param();
-                /**
-                 * The xml decleration (version, encoding, etc).
+        /**
-                 */
+         * Automoton function: content.  Processes node content.
-                void XMLDecl();
+         *@returns True if everything was ok, False for end of stream.
+         */
-                /**
+        bool content();
-                 * Misc things, Includes Comments and PIData (Processing Instructions).
+        
-                 */
+        FlexBuf fbContent;              /**< buffer for the current node's content. */
-                void Misc();
+        FlexBuf fbParamName;    /**< buffer for the current param's name. */
+        FlexBuf fbParamValue;   /**< buffer for the current param's value. */
-                /**
+        bool bStrip;    /**< Are we stripping whitespace? */
-                 * Comments
-                 */
+        HashTable htEntity;             /**< Entity type definitions. */
-                void Comment();
+};
-                /**
-                 * Processing Instructions
-                 */
-                void PI();
-                /**
-                 * Whitespace eater.
-                 */
-                void S();
-                /**
-                 * Optional whitespace eater.
-                 */
-                void Sq();
-                /**
-                 * XML Version spec
-                 */
-                void VersionInfo();
-                /**
-                 * Your basic equals sign with surrounding whitespace.
-                 */
-                void Eq();
-                /**
-                 * Read in an attribute value.
-                 */
-                FString AttValue();
-                /**
-                 * Read in the name of something.
-                 */
-                FString Name();
-                /**
-                 * Encoding decleration in the header
-                 */
-                void EncodingDecl();
-                /**
-                 * Standalone decleration in the header
-                 */
-                void SDDecl();
-                bool isS( unsigned char c )
-                {
-                        return ( c == 0x20 || c == 0x9 || c == 0xD || c == 0xA );
-                }
-        };
-}
 #endif

diff --git a/src/xmlreader.h b/src/xmlreader.h index 708a386..c8f7202 100644 --- a/src/xmlreader.h +++ b/src/xmlreader.h
@@ -1,121 +1,141 @@
1	#ifndef XML_READER_H	1	#ifndef XMLREADER
2	#define XML_READER_H	2	#define XMLREADER
		3
		4	#include <stdio.h>
		5	#include "xmldocument.h"
		6	#include "flexbuf.h"
		7	#include "hashtable.h"
		8	#include "staticstring.h"
		9
		10	/**
		11	* Takes care of reading in xml formatted data from a file. This could/should
		12	* be made more arbitrary in the future so that we can read the data from any
		13	* source. This is actually made quite simple already since all data read in
		14	* is handled by one single helper function and then palced into a FlexBuf for
		15	* easy access by the other functions. The FlexBuf also allows for block
		16	* reading from disk, which improves speed by a noticable amount.
		17	* <br>
		18	* There are also some extra features implemented that allow you to break the
		19	* standard XML reader specs and eliminate leading and trailing whitespace in
		20	* all read content. This is useful in situations where you allow additional
		21	* whitespace in the files to make them easily human readable. The resturned
		22	* content will be NULL in sitautions where all content between nodes was
		23	* stripped.
		24	*@author Mike Buland
		25	*/
		26	class XmlReader : public XmlDocument
		27	{
		28	public:
		29	/**
		30	* Create a standard XmlReader. The optional parameter bStrip allows you to
		31	* create a reader that will strip out all leading and trailing whitespace
		32	* in content, a-la html.
		33	*@param bStrip Strip out leading and trailing whitespace?
		34	*/
		35	XmlReader( bool bStrip=false );
3		36
4	#include <stdint.h>	37	/**
5	#include "bu/stream.h"	38	* Destroy this XmlReader.
6	#include "bu/fstring.h"	39	*/
7	#include "bu/xmlnode.h"	40	virtual ~XmlReader();
		41
		42	/**
		43	* Build a document based on some kind of input. This is called
		44	* automatically by the constructor.
		45	*/
		46	bool buildDoc();
		47
		48	private:
		49	/**
		50	* This is called by the low level automoton in order to get the next
		51	* character. This function should return a character at the current
		52	* position plus nIndex, but does not increment the current character.
		53	*@param nIndex The index of the character from the current stream position.
		54	*@returns A single character at the requested position, or 0 for end of
		55	* stream.
		56	*/
		57	virtual char getChar( int nIndex = 0 ) = 0;
		58
		59	/**
		60	* Called to increment the current stream position by a single character.
		61	*/
		62	virtual void usedChar( int nAmnt = 1) = 0;
		63
		64	/**
		65	* Automoton function: is whitespace.
		66	*@param chr A character
		67	*@returns True if chr is whitespace, false otherwise.
		68	*/
		69	bool isws( char chr );
		70
		71	/**
		72	* Automoton function: ws. Skips sections of whitespace.
		73	*@returns True if everything was ok, False for end of stream.
		74	*/
		75	bool ws();
		76
		77	/**
		78	* Automoton function: node. Processes an XmlNode
		79	*@returns True if everything was ok, False for end of stream.
		80	*/
		81	bool node();
8		82
9	namespace Bu
10	{
11	/**	83	/**
12	* An Xml 1.1 reader. I've decided to write this, this time, based on the	84	* Automoton function: startNode. Processes the begining of a node.
13	* official W3C reccomendation, now included with the source code. I've	85	*@returns True if everything was ok, False for end of stream.
14	* named the productions in the parser states the same as in that document,	86	*/
15	* which may make them easier to find, etc, although possibly slightly less	87	bool startNode();
16	* optimized than writing my own reduced grammer.	88
17	*	89	/**
18	* Below I will list differences between my parser and the official standard	90	* Automoton function: name. Processes the name of a node.
19	* as I come up with them.	91	*@returns True if everything was ok, False for end of stream.
20	* - Encoding and Standalone headings are ignored for the moment. (4.3.3,	92	*/
21	* 2.9)	93	bool name();
22	* - The standalone heading attribute can have any standard whitespace	94
23	* before it (the specs say only spaces, no newlines). (2.9)	95	/**
24	* - Since standalone is ignored, it is currently allowed to have any	96	* Automoton function: textDecl. Processes the xml text decleration, if
25	* value (should be restricted to "yes" or "no"). (2.9)	97	* there is one.
26	* - Currently only UTF-8 / ascii are parsed.	98	*/
27	* - [optional] The content of comments is thrown away. (2.5)	99	void textDecl();
28	* - The content of processing instruction blocks is parsed properly, but	100
29	* thrown away. (2.6)	101	/**
30	*/	102	* Automoton function: entity. Processes an entity from the header.
31	class XmlReader	103	*/
32	{	104	void entity();
33	public:	105
34	XmlReader( Bu::Stream &sIn );	106	/**
35	virtual ~XmlReader();	107	* Adds an entity to the list, if it doesn't already exist.
36		108	*@param name The name of the entity
37	XmlNode *read();	109	*@param value The value of the entity
38		110	*/
39	private:	111	void addEntity( const char name, const char value );
40	Bu::Stream &sIn;	112
41	Bu::FString sBuf;	113	StaticString *getEscape();
42		114
43	private: // Helpers	115	/**
44	const char *lookahead( int nAmnt );	116	* Automoton function: paramlist. Processes a list of node params.
45	void burn( int nAmnt );	117	*@returns True if everything was ok, False for end of stream.
46	void checkString( const char *str, int nLen );	118	*/
47		119	bool paramlist();
48	private: // States	120
49	/**	121	/**
50	* The headers, etc.	122	* Automoton function: param. Processes a single parameter.
51	*/	123	*@returns True if everything was ok, False for end of stream.
52	void prolog();	124	*/
53		125	bool param();
54	/**	126
55	* The xml decleration (version, encoding, etc).	127	/**
56	*/	128	* Automoton function: content. Processes node content.
57	void XMLDecl();	129	*@returns True if everything was ok, False for end of stream.
58		130	*/
59	/**	131	bool content();
60	* Misc things, Includes Comments and PIData (Processing Instructions).	132
61	*/	133	FlexBuf fbContent; /*< buffer for the current node's content. /
62	void Misc();	134	FlexBuf fbParamName; /*< buffer for the current param's name. /
63		135	FlexBuf fbParamValue; /*< buffer for the current param's value. /
64	/**	136	bool bStrip; /*< Are we stripping whitespace? /
65	* Comments	137
66	*/	138	HashTable htEntity; /*< Entity type definitions. /
67	void Comment();	139	};
68
69	/**
70	* Processing Instructions
71	*/
72	void PI();
73
74	/**
75	* Whitespace eater.
76	*/
77	void S();
78
79	/**
80	* Optional whitespace eater.
81	*/
82	void Sq();
83
84	/**
85	* XML Version spec
86	*/
87	void VersionInfo();
88
89	/**
90	* Your basic equals sign with surrounding whitespace.
91	*/
92	void Eq();
93
94	/**
95	* Read in an attribute value.
96	*/
97	FString AttValue();
98
99	/**
100	* Read in the name of something.
101	*/
102	FString Name();
103
104	/**
105	* Encoding decleration in the header
106	*/
107	void EncodingDecl();
108
109	/**
110	* Standalone decleration in the header
111	*/
112	void SDDecl();
113
114	bool isS( unsigned char c )
115	{
116	return ( c == 0x20 \|\| c == 0x9 \|\| c == 0xD \|\| c == 0xA );
117	}
118	};
119	}
120		140
121	#endif	141	#endif