/*
*******************************************************************************
*
* Copyright (C) 2004-2010, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
* file name: xmlparser.cpp
* encoding: US-ASCII
* tab size: 8 (not used)
* indentation:4
*
* created on: 2004jul21
* created by: Andy Heninger
*/
#include <stdio.h>
#include "unicode/uchar.h"
#include "unicode/ucnv.h"
#include "unicode/regex.h"
#include "filestrm.h"
#include "xmlparser.h"
#if !UCONFIG_NO_REGULAR_EXPRESSIONS && !UCONFIG_NO_CONVERSION
// character constants
enum {
x_QUOT=0x22,
x_AMP=0x26,
x_APOS=0x27,
x_LT=0x3c,
x_GT=0x3e,
x_l=0x6c
};
#define XML_SPACES "[ \\u0009\\u000d\\u000a]"
// XML #4
#define XML_NAMESTARTCHAR "[[A-Z]:_[a-z][\\u00c0-\\u00d6][\\u00d8-\\u00f6]" \
"[\\u00f8-\\u02ff][\\u0370-\\u037d][\\u037F-\\u1FFF][\\u200C-\\u200D]" \
"[\\u2070-\\u218F][\\u2C00-\\u2FEF][\\u3001-\\uD7FF][\\uF900-\\uFDCF]" \
"[\\uFDF0-\\uFFFD][\\U00010000-\\U000EFFFF]]"
// XML #5
#define XML_NAMECHAR "[" XML_NAMESTARTCHAR "\\-.[0-9]\\u00b7[\\u0300-\\u036f][\\u203f-\\u2040]]"
// XML #6
#define XML_NAME XML_NAMESTARTCHAR "(?:" XML_NAMECHAR ")*"
U_NAMESPACE_BEGIN
UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UXMLParser)
UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UXMLElement)
//
// UXMLParser constructor. Mostly just initializes the ICU regexes that are
// used for parsing.
//
UXMLParser::UXMLParser(UErrorCode &status) :
// XML Declaration. XML Production #23.
// example: "<?xml version=1.0 encoding="utf-16" ?>
// This is a sloppy implementation - just look for the leading <?xml and the closing ?>
// allow for a possible leading BOM.
mXMLDecl(UnicodeString("(?s)\\uFEFF?<\\?xml.+?\\?>", -1, US_INV), 0, status),
// XML Comment production #15
// example: "<!-- whatever -->
// note, does not detect an illegal "--" within comments
mXMLComment(UnicodeString("(?s)<!--.+?-->", -1, US_INV), 0, status),
// XML Spaces
// production [3]
mXMLSP(UnicodeString(XML_SPACES "+", -1, US_INV), 0, status),
// XML Doctype decl production #28
// example "<!DOCTYPE foo SYSTEM "somewhere" >
// or "<!DOCTYPE foo [internal dtd]>
// TODO: we don't actually parse the DOCTYPE or internal subsets.
// Some internal dtd subsets could confuse this simple-minded
// attempt at skipping over them, specifically, occcurences
// of closeing square brackets. These could appear in comments,
// or in parameter entity declarations, for example.
mXMLDoctype(UnicodeString(
"(?s)<!DOCTYPE.*?(>|\\[.*?\\].*?>)", -1, US_INV
), 0, status),
// XML PI production #16
// example "<?target stuff?>
mXMLPI(UnicodeString("(?s)<\\?.+?\\?>", -1, US_INV), 0, status),
// XML Element Start Productions #40, #41
// example <foo att1='abc' att2="d e f" >
// capture #1: the tag name
//
mXMLElemStart (UnicodeString("(?s)<(" XML_NAME ")" // match "<tag_name"
"(?:"
XML_SPACES "+" XML_NAME XML_SPACES "*=" XML_SPACES "*" // match "ATTR_NAME = "
"(?:(?:\\\'[^<\\\']*?\\\')|(?:\\\"[^<\\\"]*?\\\"))" // match '"attribute value"'
")*" // * for zero or more attributes.
XML_SPACES "*?>", -1, US_INV), 0, status), // match " >"
// XML Element End production #42
// example </foo>
mXMLElemEnd (UnicodeString("</(" XML_NAME ")" XML_SPACES "*>", -1, US_INV), 0, status),
// XML Element Empty production #44
// example <foo att1="abc" att2="d e f" />
mXMLElemEmpty (UnicodeString("(?s)<(" XML_NAME ")" // match "<tag_name"
"(?:"
XML_SPACES "+" XML_NAME XML_SPACES "*=" XML_SPACES "*" // match "ATTR_NAME = "
"(?:(?:\\\'[^<\\\']*?\\\')|(?:\\\"[^<\\\"]*?\\\"))" // match '"attribute value"'
")*" // * for zero or more attributes.
XML_SPACES "*?/>", -1, US_INV), 0, status), // match " />"
// XMLCharData. Everything but '<'. Note that & will be dealt with later.
mXMLCharData(UnicodeString("(?s)[^<]*", -1, US_INV), 0, status),
// Attribute name = "value". XML Productions 10, 40/41
// Capture group 1 is name,
// 2 is the attribute value, including the quotes.
//
// Note that attributes are scanned twice. The first time is with
// the regex for an entire element start. There, the attributes
// are checked syntactically, but not separted out one by one.
// Here, we match a single attribute, and make its name and
// attribute value available to the parser code.
mAttrValue(UnicodeString(XML_SPACES "+(" XML_NAME ")" XML_SPACES "*=" XML_SPACES "*"
"((?:\\\'[^<\\\']*?\\\')|(?:\\\"[^<\\\"]*?\\\"))", -1, US_INV), 0, status),
mAttrNormalizer(UnicodeString(XML_SPACES, -1, US_INV), 0, status),
// Match any of the new-line sequences in content.
// All are changed to \u000a.
mNewLineNormalizer(UnicodeString("\\u000d\\u000a|\\u000d\\u0085|\\u000a|\\u000d|\\u0085|\\u2028", -1, US_INV), 0, status),
// & char references
// We will figure out what we've got based on which capture group has content.
// The last one is a catchall for unrecognized entity references..
// 1 2 3 4 5 6 7 8
mAmps(UnicodeString("&(?:(amp;)|(lt;)|(gt;)|(apos;)|(quot;)|#x([0-9A-Fa-f]{1,8});|#([0-9]{1,8});|(.))"),
0, status),
fNames(status),
fElementStack(status),
fOneLF((UChar)0x0a) // Plain new-line string, used in new line normalization.
{
}
UXMLParser *
UXMLParser::createParser(UErrorCode &errorCode) {
if (U_FAILURE(errorCode)) {
return NULL;
} else {
return new UXMLParser(errorCode);
}
}
UXMLParser::~UXMLParser() {}
UXMLElement *
UXMLParser::parseFile(const char *filename, UErrorCode &errorCode) {
char bytes[4096], charsetBuffer[100];
FileStream *f;
const char *charset, *pb;
UnicodeString src;
UConverter *cnv;
UChar *buffer, *pu;
int32_t fileLength, bytesLength, length, capacity;
UBool flush;
if(U_FAILURE(errorCode)) {
return NULL;
}
f=T_FileStream_open(filename, "rb");
if(f==NULL) {
errorCode=U_FILE_ACCESS_ERROR;
return NULL;
}
bytesLength=T_FileStream_read(f, bytes, (int32_t)sizeof(bytes));
if(bytesLength<(int32_t)sizeof(bytes)) {
// we have already read the entire file
fileLength=bytesLength;
} else {
// get the file length
fileLength=T_FileStream_size(f);
}
/*
* get the charset:
* 1. Unicode signature
* 2. treat as ISO-8859-1 and read XML encoding="charser"
* 3. default to UTF-8
*/
charset=ucnv_detectUnicodeSignature(bytes, bytesLength, NULL, &errorCode);
if(U_SUCCESS(errorCode) && charset!=NULL) {
// open converter according to Unicode signature
cnv=ucnv_open(charset, &errorCode);
} else {
// read as Latin-1 and parse the XML declaration and encoding
cnv=ucnv_open("ISO-8859-1", &errorCode);
if(U_FAILUR
