pattern.c

libxml,在UNIX/LINUX下非常重要的一个库,为XML相关应用提供方便.目前上载的是最新版本,若要取得最新版本,请参考里面的readme.

			    break;
			lst = lst->next;
		    }
		    if (lst != NULL)
			continue;
		}
		goto rollback;
	    }
            case XML_OP_ATTR:
		if (node->type != XML_ATTRIBUTE_NODE)
		    goto rollback;
		if (step->value != NULL) {
		    if (step->value[0] != node->name[0])
			goto rollback;
		    if (!xmlStrEqual(step->value, node->name))
			goto rollback;
		}
		/* Namespace test */
		if (node->ns == NULL) {
		    if (step->value2 != NULL)
			goto rollback;
		} else if (step->value2 != NULL) {
		    if (!xmlStrEqual(step->value2, node->ns->href))
			goto rollback;
		}
		continue;
            case XML_OP_PARENT:
		if ((node->type == XML_DOCUMENT_NODE) ||
		    (node->type == XML_HTML_DOCUMENT_NODE) ||
#ifdef LIBXML_DOCB_ENABLED
		    (node->type == XML_DOCB_DOCUMENT_NODE) ||
#endif
		    (node->type == XML_NAMESPACE_DECL))
		    goto rollback;
		node = node->parent;
		if (node == NULL)
		    goto rollback;
		if (step->value == NULL)
		    continue;
		if (step->value[0] != node->name[0])
		    goto rollback;
		if (!xmlStrEqual(step->value, node->name))
		    goto rollback;
		/* Namespace test */
		if (node->ns == NULL) {
		    if (step->value2 != NULL)
			goto rollback;
		} else if (node->ns->href != NULL) {
		    if (step->value2 == NULL)
			goto rollback;
		    if (!xmlStrEqual(step->value2, node->ns->href))
			goto rollback;
		}
		continue;
            case XML_OP_ANCESTOR:
		/* TODO: implement coalescing of ANCESTOR/NODE ops */
		if (step->value == NULL) {
		    i++;
		    step = &comp->steps[i];
		    if (step->op == XML_OP_ROOT)
			goto found;
		    if (step->op != XML_OP_ELEM)
			goto rollback;
		    if (step->value == NULL)
			return(-1);
		}
		if (node == NULL)
		    goto rollback;
		if ((node->type == XML_DOCUMENT_NODE) ||
		    (node->type == XML_HTML_DOCUMENT_NODE) ||
#ifdef LIBXML_DOCB_ENABLED
		    (node->type == XML_DOCB_DOCUMENT_NODE) ||
#endif
		    (node->type == XML_NAMESPACE_DECL))
		    goto rollback;
		node = node->parent;
		while (node != NULL) {
		    if (node == NULL)
			goto rollback;
		    if ((node->type == XML_ELEMENT_NODE) &&
			(step->value[0] == node->name[0]) &&
			(xmlStrEqual(step->value, node->name))) {
			/* Namespace test */
			if (node->ns == NULL) {
			    if (step->value2 == NULL)
				break;
			} else if (node->ns->href != NULL) {
			    if ((step->value2 != NULL) &&
			        (xmlStrEqual(step->value2, node->ns->href)))
				break;
			}
		    }
		    node = node->parent;
		}
		if (node == NULL)
		    goto rollback;
		/*
		 * prepare a potential rollback from here
		 * for ancestors of that node.
		 */
		if (step->op == XML_OP_ANCESTOR)
		    xmlPatPushState(&states, i, node);
		else
		    xmlPatPushState(&states, i - 1, node);
		continue;
            case XML_OP_NS:
		if (node->type != XML_ELEMENT_NODE)
		    goto rollback;
		if (node->ns == NULL) {
		    if (step->value != NULL)
			goto rollback;
		} else if (node->ns->href != NULL) {
		    if (step->value == NULL)
			goto rollback;
		    if (!xmlStrEqual(step->value, node->ns->href))
			goto rollback;
		}
		break;
            case XML_OP_ALL:
		if (node->type != XML_ELEMENT_NODE)
		    goto rollback;
		break;
	}
    }
found:
    if (states.states != NULL) {
        /* Free the rollback states */
	xmlFree(states.states);
    }
    return(1);
rollback:
    /* got an error try to rollback */
    if (states.states == NULL)
	return(0);
    if (states.nbstates <= 0) {
	xmlFree(states.states);
	return(0);
    }
    states.nbstates--;
    i = states.states[states.nbstates].step;
    node = states.states[states.nbstates].node;
#if 0
    fprintf(stderr, "Pop: %d, %s\n", i, node->name);
#endif
    goto restart;
}

/************************************************************************
 *									*
 *			Dedicated parser for templates			*
 *									*
 ************************************************************************/

#define TODO 								\
    xmlGenericError(xmlGenericErrorContext,				\
	    "Unimplemented block at %s:%d\n",				\
            __FILE__, __LINE__);
#define CUR (*ctxt->cur)
#define SKIP(val) ctxt->cur += (val)
#define NXT(val) ctxt->cur[(val)]
#define CUR_PTR ctxt->cur

#define SKIP_BLANKS 							\
    while (IS_BLANK_CH(CUR)) NEXT

#define CURRENT (*ctxt->cur)
#define NEXT ((*ctxt->cur) ?  ctxt->cur++: ctxt->cur)


#define PUSH(op, val, val2) 						\
    if (xmlPatternAdd(ctxt, ctxt->comp, (op), (val), (val2))) goto error;

#define XSLT_ERROR(X)							\
    { xsltError(ctxt, __FILE__, __LINE__, X);			\
      ctxt->error = (X); return; }

#define XSLT_ERROR0(X)							\
    { xsltError(ctxt, __FILE__, __LINE__, X);			\
      ctxt->error = (X); return(0); }

#if 0
/**
 * xmlPatScanLiteral:
 * @ctxt:  the XPath Parser context
 *
 * Parse an XPath Litteral:
 *
 * [29] Literal ::= '"' [^"]* '"'
 *                | "'" [^']* "'"
 *
 * Returns the Literal parsed or NULL
 */

static xmlChar *
xmlPatScanLiteral(xmlPatParserContextPtr ctxt) {
    const xmlChar *q, *cur;
    xmlChar *ret = NULL;
    int val, len;

    SKIP_BLANKS;
    if (CUR == '"') {
        NEXT;
	cur = q = CUR_PTR;
	val = xmlStringCurrentChar(NULL, cur, &len);
	while ((IS_CHAR(val)) && (val != '"')) {
	    cur += len;
	    val = xmlStringCurrentChar(NULL, cur, &len);
	}
	if (!IS_CHAR(val)) {
	    ctxt->error = 1;
	    return(NULL);
	} else {
	    ret = xmlStrndup(q, cur - q);
        }
	cur += len;
	CUR_PTR = cur;
    } else if (CUR == '\'') {
        NEXT;
	cur = q = CUR_PTR;
	val = xmlStringCurrentChar(NULL, cur, &len);
	while ((IS_CHAR(val)) && (val != '\'')) {
	    cur += len;
	    val = xmlStringCurrentChar(NULL, cur, &len);
	}
	if (!IS_CHAR(val)) {
	    ctxt->error = 1;
	    return(NULL);
	} else {
	    ret = xmlStrndup(q, cur - q);
        }
	cur += len;
	CUR_PTR = cur;
    } else {
	/* XP_ERROR(XPATH_START_LITERAL_ERROR); */
	ctxt->error = 1;
	return(NULL);
    }
    return(ret);
}
#endif

/**
 * xmlPatScanName:
 * @ctxt:  the XPath Parser context
 *
 * [4] NameChar ::= Letter | Digit | '.' | '-' | '_' | 
 *                  CombiningChar | Extender
 *
 * [5] Name ::= (Letter | '_' | ':') (NameChar)*
 *
 * [6] Names ::= Name (S Name)*
 *
 * Returns the Name parsed or NULL
 */

static xmlChar *
xmlPatScanName(xmlPatParserContextPtr ctxt) {
    const xmlChar *q, *cur;
    xmlChar *ret = NULL;
    int val, len;

    SKIP_BLANKS;

    cur = q = CUR_PTR;
    val = xmlStringCurrentChar(NULL, cur, &len);
    if (!IS_LETTER(val) && (val != '_') && (val != ':'))
	return(NULL);

    while ((IS_LETTER(val)) || (IS_DIGIT(val)) ||
           (val == '.') || (val == '-') ||
	   (val == '_') || 
	   (IS_COMBINING(val)) ||
	   (IS_EXTENDER(val))) {
	cur += len;
	val = xmlStringCurrentChar(NULL, cur, &len);
    }
    ret = xmlStrndup(q, cur - q);
    CUR_PTR = cur;
    return(ret);
}

/**
 * xmlPatScanNCName:
 * @ctxt:  the XPath Parser context
 *
 * Parses a non qualified name
 *
 * Returns the Name parsed or NULL
 */

static xmlChar *
xmlPatScanNCName(xmlPatParserContextPtr ctxt) {
    const xmlChar *q, *cur;
    xmlChar *ret = NULL;
    int val, len;

    SKIP_BLANKS;

    cur = q = CUR_PTR;
    val = xmlStringCurrentChar(NULL, cur, &len);
    if (!IS_LETTER(val) && (val != '_'))
	return(NULL);

    while ((IS_LETTER(val)) || (IS_DIGIT(val)) ||
           (val == '.') || (val == '-') ||
	   (val == '_') ||
	   (IS_COMBINING(val)) ||
	   (IS_EXTENDER(val))) {
	cur += len;
	val = xmlStringCurrentChar(NULL, cur, &len);
    }
    ret = xmlStrndup(q, cur - q);
    CUR_PTR = cur;
    return(ret);
}

#if 0
/**
 * xmlPatScanQName:
 * @ctxt:  the XPath Parser context
 * @prefix:  the place to store the prefix
 *
 * Parse a qualified name
 *
 * Returns the Name parsed or NULL
 */

static xmlChar *
xmlPatScanQName(xmlPatParserContextPtr ctxt, xmlChar **prefix) {
    xmlChar *ret = NULL;

    *prefix = NULL;
    ret = xmlPatScanNCName(ctxt);
    if (CUR == ':') {
        *prefix = ret;
	NEXT;
	ret = xmlPatScanNCName(ctxt);
    }
    return(ret);
}
#endif

/**
 * xmlCompileAttributeTest:
 * @ctxt:  the compilation context
 *
 * Compile an attribute test.
 */
static void
xmlCompileAttributeTest(xmlPatParserContextPtr ctxt) {
    xmlChar *token = NULL;
    xmlChar *name = NULL;
    xmlChar *URL = NULL;
    
    name = xmlPatScanNCName(ctxt);
    if (name == NULL) {
	if (CUR == '*') {
	    PUSH(XML_OP_ATTR, NULL, NULL);
	} else {
	    ERROR(NULL, NULL, NULL,
		"xmlCompileAttributeTest : Name expected\n");
	    ctxt->error = 1;
	}
	return;
    }
    if (CUR == ':') {
	int i;
	xmlChar *prefix = name;
	
	NEXT;
	/*
	* This is a namespace match
	*/
	token = xmlPatScanName(ctxt);
	for (i = 0;i < ctxt->nb_namespaces;i++) {
	    if (xmlStrEqual(ctxt->namespaces[2 * i + 1], prefix)) {
		URL = xmlStrdup(ctxt->namespaces[2 * i]);
		break;
	    }
	}
	if (i >= ctxt->nb_namespaces) {
	    ERROR5(NULL, NULL, NULL,
		"xmlCompileAttributeTest : no namespace bound to prefix %s\n",
		prefix);
	    ctxt->error = 1;	    
	    goto error;
	}
	
	xmlFree(prefix);
	if (token == NULL) {
	    if (CUR == '*') {
		NEXT;
		PUSH(XML_OP_ATTR, NULL, URL);
	    } else {
		ERROR(NULL, NULL, NULL,
		    "xmlCompileAttributeTest : Name expected\n");
		ctxt->error = 1;
		goto error;
	    }	    
	} else {
	    PUSH(XML_OP_ATTR, token, URL);
	}
    } else {
	PUSH(XML_OP_ATTR, name, NULL);
    }
    return;
error:
    if (URL != NULL)
	xmlFree(URL);
    if (token != NULL)
	xmlFree(token);
}


/**
 * xmlCompileStepPattern:
 * @ctxt:  the compilation context
 *
 * Compile the Step Pattern and generates a precompiled
 * form suitable for fast matching.
 *
 * [3]    Step    ::=    '.' | NameTest
 * [4]    NameTest    ::=    QName | '*' | NCName ':' '*' 
 */

static void
xmlCompileStepPattern(xmlPatParserContextPtr ctxt) {
    xmlChar *token = NULL;
    xmlChar *name = NULL;
    xmlChar *URL = NULL;

    SKIP_BLANKS;
    if (CUR == '.') {
	NEXT;
	PUSH(XML_OP_ELEM, NULL, NULL);
	return;
    }
    name = xmlPatScanNCName(ctxt);
    if (name == NULL) {
	if (CUR == '*') {
	    NEXT;
	    PUSH(XML_OP_ALL, NULL, NULL);
	    return;
	} else if (CUR == '@') {
	    NEXT;
	    xmlCompileAttributeTest(ctxt);
	    if (ctxt->error != 0) 
		goto error;
	    return;
	} else {
	    ERROR(NULL, NULL, NULL,
		    "xmlCompileStepPattern : Name expected\n");
	    ctxt->error = 1;
	    return;
	}
    }
    SKIP_BLANKS;
    if (CUR == ':') {
	NEXT;
	if (CUR != ':') {
	    xmlChar *prefix = name;
	    int i;

	    /*
	     * This is a namespace match
	     */
	    token = xmlPatScanName(ctxt);
	    for (i = 0;i < ctxt->nb_namespaces;i++) {
	        if (xmlStrEqual(ctxt->namespaces[2 * i + 1], prefix)) {
		    URL = xmlStrdup(ctxt->namespaces[2 * i]);
		    break;
		}
	    }
	    if (i >= ctxt->nb_namespaces) {
		ERROR5(NULL, NULL, NULL,
		    "xmlCompileStepPattern : no namespace bound to prefix %s\n",
		    prefix);
		ctxt->error = 1;
		goto error;
	    }
	    xmlFree(prefix);
	    if (token == NULL) {
		if (CUR == '*') {
		    NEXT;
		    PUSH(XML_OP_NS, URL, NULL);
		} else {
		    ERROR(NULL, NULL, NULL,
			    "xmlCompileStepPattern : Name expected\n");
		    ctxt->error = 1;
		    goto error;
		}
	    } else {
		PUSH(XML_OP_ELEM, token, URL);
	    }
	} else {
	    NEXT;
	    if (xmlStrEqual(name, (const xmlChar *) "child")) {
		xmlFree(name);
		name = xmlPatScanName(ctxt);
		if (name == NULL) {
		    if (CUR == '*') {
			NEXT;
			PUSH(XML_OP_ALL, NULL, NULL);
			return;
		    } else {
			ERROR(NULL, NULL, NULL,
			    "xmlCompileStepPattern : QName expected\n");
			ctxt->error = 1;
			goto error;
		    }
		}
		if (CUR == ':') {
		    xmlChar *prefix = name;
		    int i;
		    
		    NEXT;
		    /*
		    * This is a namespace match
		    */
		    token = xmlPatScanName(ctxt);
		    for (i = 0;i < ctxt->nb_namespaces;i++) {