genode/repos/base/include/util/xml_node.h

1167 lines
27 KiB
C
Raw Normal View History

2011-12-22 16:19:25 +01:00
/*
* \brief XML parser
* \author Norman Feske
* \date 2007-08-21
*/
/*
* Copyright (C) 2007-2017 Genode Labs GmbH
2011-12-22 16:19:25 +01:00
*
* This file is part of the Genode OS framework, which is distributed
* under the terms of the GNU Affero General Public License version 3.
2011-12-22 16:19:25 +01:00
*/
#ifndef _INCLUDE__UTIL__XML_NODE_H_
#define _INCLUDE__UTIL__XML_NODE_H_
#include <base/log.h>
2011-12-22 16:19:25 +01:00
#include <util/token.h>
#include <util/noncopyable.h>
2011-12-22 16:19:25 +01:00
#include <base/exception.h>
namespace Genode {
class Xml_attribute;
class Xml_node;
class Xml_unquoted;
}
2011-12-22 16:19:25 +01:00
/**
* Representation of an XML-node attribute
*
* An attribute has the form 'name="value"'.
*/
class Genode::Xml_attribute
{
private:
/**
* Scanner policy that accepts hyphens in identifiers
*/
struct Scanner_policy_xml_identifier {
static bool identifier_char(char c, unsigned i) {
return is_letter(c) || c == '_' || c == ':'
|| (i && (c == '-' || c == '.' || is_digit(c))); } };
/**
* Define tokenizer that matches XML tags (with hyphens) as identifiers
*/
typedef ::Genode::Token<Scanner_policy_xml_identifier> Token;
struct Tokens
{
Token name;
Token value;
Tokens(Token t)
: name(t.eat_whitespace()), value(name.next().next()) { };
bool valid() const
{
bool const tag_present = (name.type() == Token::IDENT);
bool const value_present = (name.next()[0] == '=' &&
value.type() == Token::STRING);
return tag_present && value_present;
}
} _tokens;
friend class Xml_node;
/*
* Even though 'Tag' is part of 'Xml_node', the friendship
* to 'Xml_node' does not apply for 'Tag' when compiling
* the code with 'gcc-3.4'. Hence, we need to add an
* explicit friendship to 'Tag'.
*/
friend class Tag;
/**
* Return true if token refers to a valid attribute
*/
static bool _valid(Token t) { return Tokens(t).valid(); }
/**
* Constructor
*
* This constructor is meant to be used as implicitly to
* construct an 'Xml_attribute' from a token sequence via an
* assignment from the leading 'Token'.
*/
explicit Xml_attribute(Token t) : _tokens(t)
{
if (_tokens.name.type() != Token::IDENT)
throw Nonexistent_attribute();
if (!_tokens.valid())
throw Invalid_syntax();
}
/**
* Return token following the attribute declaration
*/
Token _next_token() const { return _tokens.name.next().next().next(); }
2011-12-22 16:19:25 +01:00
public:
2011-12-22 16:19:25 +01:00
/*********************
** Exception types **
*********************/
2011-12-22 16:19:25 +01:00
class Invalid_syntax : public Exception { };
class Nonexistent_attribute : public Exception { };
2011-12-22 16:19:25 +01:00
/**
* Return attribute type as null-terminated string
*
* \deprecated
* \noapi
*/
void type(char *dst, size_t max_len) const __attribute__((deprecated))
{
/*
* Limit number of characters by token length, take
* null-termination into account.
*/
max_len = min(max_len, _tokens.name.len() + 1);
copy_cstring(dst, _tokens.name.start(), max_len);
}
2011-12-22 16:19:25 +01:00
typedef String<64> Name;
Name name() const {
return Name(Cstring(_tokens.name.start(), _tokens.name.len())); }
/**
* Return true if attribute has specified type
*/
bool has_type(char const *type) {
return strlen(type) == _tokens.name.len() &&
strcmp(type, _tokens.name.start(), _tokens.name.len()) == 0; }
2011-12-22 16:19:25 +01:00
/**
* Return size of value
*
* \deprecated use 'with_raw_node' instead
* \noapi
*/
char const *value_base() const __attribute__((deprecated))
{
return _tokens.value.start() + 1;
}
2011-12-22 16:19:25 +01:00
/**
* Return size of the value in bytes
*/
size_t value_size() const
{
/*
* The size of the actual value content excludes both the starting
* and the trailing quote character.
*
* The invariant 'len >= 2' is enforced by the 'Xml_attribute'
* constructor by checking the '_value' type for being a 'STRING'.
*/
return _tokens.value.len() - 2;
}
/**
* Return true if attribute has the specified value
*/
bool has_value(char const *value) const {
return strlen(value) == (_tokens.value.len() - 2)
&& !strcmp(value, _tokens.value.start() + 1, _tokens.value.len() - 2); }
/**
* Call functor 'fn' with the data of the attribute value as argument
*
* The functor is called with the start pointer ('char const *') and
* size (size_t) of the attribute value as arguments.
*
* Note that the content of the buffer is not null-terminated but
* delimited by the size argument.
*/
template <typename FN>
void with_raw_value(FN const &fn) const
{
/*
* Skip leading quote of the '_value' to access the actual value.
*/
fn(_tokens.value.start() + 1, value_size());
}
/**
* Return attribute value as null-terminated string
*
* \deprecated
* \noapi
*/
void value(char *dst, size_t max_len) const __attribute__((deprecated))
{
with_raw_value([&] (char const *start, size_t length) {
copy_cstring(dst, start, min(max_len, length + 1)); });
}
/**
* Return attribute value as typed value
*
* \param T type of value to read
* \return true on success, or
* false if attribute is invalid or value
* conversion failed
*/
template <typename T>
bool value(T &out) const
{
bool result = false;
with_raw_value([&] (char const *start, size_t length) {
result = (ascii_to(start, out) == length); });
return result;
}
2011-12-22 16:19:25 +01:00
/**
* Return attribute value as 'Genode::String'
*/
template <size_t N>
void value(String<N> &out) const
{
with_raw_value([&] (char const *start, size_t length) {
out = String<N>(Cstring(start, length)); });
}
/**
* Return attribute value as 'Genode::String'
*
* \deprecated use 'value(String<N> &out' instead
* \noapi
*/
template <size_t N>
__attribute__((deprecated))
void value(String<N> *out) const
{
with_raw_value([&] (char const *start, size_t length) {
*out = String<N>(Cstring(start, length)); });
}
/**
* Return attribute value as typed value
*
* \deprecated use 'value(T &out)' instead
* \noapi
*/
template <typename T>
__attribute__((deprecated))
bool value(T *out) const
{
bool result = false;
with_raw_value([&] (char const *start, size_t length) {
result = (ascii_to(start, *out) == length); });
return result;
}
/**
* Return next attribute in attribute list
*/
Xml_attribute next() const { return Xml_attribute(_next_token()); }
};
2011-12-22 16:19:25 +01:00
/**
* Representation of an XML node
*/
class Genode::Xml_node
{
private:
2011-12-22 16:19:25 +01:00
typedef Xml_attribute::Token Token;
2011-12-22 16:19:25 +01:00
/**
* Forward declaration needed for befriending Tag with Xml_attribute
*/
class Tag;
friend class Xml_unquoted;
public:
2011-12-22 16:19:25 +01:00
/*********************
** Exception types **
*********************/
typedef Genode::Exception Exception;
typedef Xml_attribute::Nonexistent_attribute Nonexistent_attribute;
typedef Xml_attribute::Invalid_syntax Invalid_syntax;
2011-12-22 16:19:25 +01:00
class Nonexistent_sub_node : public Exception { };
/**
* Type definition for maintaining backward compatibility
*/
typedef Xml_attribute Attribute;
private:
class Tag
{
public:
enum Type { START, END, EMPTY, INVALID };
private:
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Token _token { };
Token _name { };
Type _type { INVALID };
public:
/**
* Constructor
*
* \param start first token of the tag
*
* At construction time, the validity of the tag is checked and
* the tag type is determined. A valid tag consists of:
* # Leading '<' tag delimiter
* # '/' for marking an end tag
* # Tag name
* # Optional attribute sequence (if tag is no end tag)
* # '/' for marking an empty-element tag (if tag is no end tag)
* # Closing '>' tag delimiter
*/
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Tag(Token start) : _token(start)
{
Type supposed_type = START;
2011-12-22 16:19:25 +01:00
if (_token[0] != '<')
return;
2011-12-22 16:19:25 +01:00
if (_token.next()[0] == '/')
supposed_type = END;
2011-12-22 16:19:25 +01:00
if (_token.next().type() != Token::IDENT && _token.next()[0] != '/')
return;
2011-12-22 16:19:25 +01:00
_name = _token.next()[0] == '/' ? _token.next().next() : _token.next();
if (_name.type() != Token::IDENT)
return;
2011-12-22 16:19:25 +01:00
/* skip attributes to find tag delimiter */
Token delimiter = _name.next();
if (supposed_type != END)
while (Xml_attribute::_valid(delimiter))
delimiter = Xml_attribute(delimiter)._next_token();
2011-12-22 16:19:25 +01:00
delimiter = delimiter.eat_whitespace();
2011-12-22 16:19:25 +01:00
/*
* Now we expect the '>' delimiter. For empty-element tags,
* the delimiter is prefixed with a '/'.
*/
if (delimiter[0] == '/') {
2011-12-22 16:19:25 +01:00
/* if a '/' was already at the start, the tag is invalid */
if (supposed_type == END)
return;
2011-12-22 16:19:25 +01:00
supposed_type = EMPTY;
2011-12-22 16:19:25 +01:00
/* skip '/' */
delimiter = delimiter.next();
}
2011-12-22 16:19:25 +01:00
if (delimiter[0] != '>') return;
2011-12-22 16:19:25 +01:00
_type = supposed_type;
}
2011-12-22 16:19:25 +01:00
/**
* Default constructor produces invalid Tag
*/
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Tag() { }
2011-12-22 16:19:25 +01:00
/**
* Return type of tag
*/
Type type() const { return _type; }
2011-12-22 16:19:25 +01:00
/**
* Return true if tag is the start of a valid XML node
*/
bool node() const { return _type == START || _type == EMPTY; }
2011-12-22 16:19:25 +01:00
/**
* Return first token of tag
*/
Token token() const { return _token; }
2011-12-22 16:19:25 +01:00
/**
* Return name of tag
*/
Token name() const { return _name; }
2011-12-22 16:19:25 +01:00
/**
* Return token after the closing tag delimiter
*/
Token next_token() const
{
/*
* Search for next closing delimiter, skip potential
* attributes and '/' delimiter prefix of empty-element
* tags.
2011-12-22 16:19:25 +01:00
*/
Token t = _name;
for (; t && t[0] != '>'; t = t.next());
2011-12-22 16:19:25 +01:00
/* if 't' is invalid, 't.next()' is invalid too */
return t.next();
}
2011-12-22 16:19:25 +01:00
/**
* Return true if tag as at least one attribute
*/
bool has_attribute() const { return Xml_attribute::_valid(_name.next()); }
/**
* Return first attribute of tag
*/
Xml_attribute attribute() const { return Xml_attribute(_name.next()); }
};
2011-12-22 16:19:25 +01:00
class Comment
{
private:
2011-12-22 16:19:25 +01:00
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Token _next { }; /* token following the comment */
bool _valid { false }; /* true if comment is well formed */
2011-12-22 16:19:25 +01:00
public:
2011-12-22 16:19:25 +01:00
/**
* Constructor
*
* \param start first token of the comment tag
*/
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Comment(Token t)
{
/* check for comment start */
if (!t.matches("<!--"))
return;
2011-12-22 16:19:25 +01:00
/* skip four single characters for "<!--" */
t = t.next().next().next().next();
2011-12-22 16:19:25 +01:00
/* find token after comment delimiter */
_next = t.next_after("-->");
_valid = _next.valid();
}
2011-12-22 16:19:25 +01:00
/**
* Default constructor produces invalid Comment
*/
Follow practices suggested by "Effective C++" The patch adjust the code of the base, base-<kernel>, and os repository. To adapt existing components to fix violations of the best practices suggested by "Effective C++" as reported by the -Weffc++ compiler argument. The changes follow the patterns outlined below: * A class with virtual functions can no longer publicly inherit base classed without a vtable. The inherited object may either be moved to a member variable, or inherited privately. The latter would be used for classes that inherit 'List::Element' or 'Avl_node'. In order to enable the 'List' and 'Avl_tree' to access the meta data, the 'List' must become a friend. * Instead of adding a virtual destructor to abstract base classes, we inherit the new 'Interface' class, which contains a virtual destructor. This way, single-line abstract base classes can stay as compact as they are now. The 'Interface' utility resides in base/include/util/interface.h. * With the new warnings enabled, all member variables must be explicitly initialized. Basic types may be initialized with '='. All other types are initialized with braces '{ ... }' or as class initializers. If basic types and non-basic types appear in a row, it is nice to only use the brace syntax (also for basic types) and align the braces. * If a class contains pointers as members, it must now also provide a copy constructor and assignment operator. In the most cases, one would make them private, effectively disallowing the objects to be copied. Unfortunately, this warning cannot be fixed be inheriting our existing 'Noncopyable' class (the compiler fails to detect that the inheriting class cannot be copied and still gives the error). For now, we have to manually add declarations for both the copy constructor and assignment operator as private class members. Those declarations should be prepended with a comment like this: /* * Noncopyable */ Thread(Thread const &); Thread &operator = (Thread const &); In the future, we should revisit these places and try to replace the pointers with references. In the presence of at least one reference member, the compiler would no longer implicitly generate a copy constructor. So we could remove the manual declaration. Issue #465
2017-12-21 15:42:15 +01:00
Comment() { }
2011-12-22 16:19:25 +01:00
/**
* Return true if comment is valid
2011-12-22 16:19:25 +01:00
*/
bool valid() const { return _valid; }
2011-12-22 16:19:25 +01:00
/**
* Return token after the closing comment delimiter
*/
Token next_token() const { return _next; }
};
2011-12-22 16:19:25 +01:00
/**
* Helper class to decode XML character entities
*/
struct Decoded_character
{
char character = 0;
size_t encoded_len = 1;
struct Translation
{
char character;
char const *seq;
size_t seq_len;
};
static Translation translate(char const *src, size_t src_len)
{
enum { NUM = 6 };
static Translation translations[NUM] = {
{ '>', "&gt;", 4 },
{ '<', "&lt;", 4 },
{ '&', "&amp;", 5 },
{ '"', "&quot;", 6 },
{ '\'', "&apos;", 6 },
{ 0, "&#x00;", 6 }
};
if (src_len == 0)
return { 0, nullptr, 0 };
for (unsigned i = 0; i < NUM; i++) {
Translation const &translation = translations[i];
if (src_len < translation.seq_len
|| memcmp(src, translation.seq, translation.seq_len))
continue;
/* translation matches */
return translation;
}
/* sequence is not known, pass single character as is */
return { *src, nullptr, 1 };
}
Decoded_character(char const *src, size_t src_len)
{
if (*src != '&' || src_len == 0) {
character = *src;
return;
}
Translation const translation = translate(src, src_len);
character = translation.character;
encoded_len = translation.seq_len;
}
};
char const * _addr; /* first character of XML data */
size_t _max_len; /* length of XML data in characters */
2011-12-22 16:19:25 +01:00
/**
* Search matching end tag for given start tag and detemine number of
* immediate sub nodes along the way.
*
* \return end tag or invalid tag
*
* The method searches for a end tag that matches the same depth level
* and the same name as the start tag of the XML node. If the XML
* structure is invalid, the search results is an invalid Tag.
*/
static Tag _search_end_tag(Tag start_tag, int &sub_nodes_count)
{
/*
* If the start tag is invalid or an empty-element tag,
* we use the same tag as end tag.
*/
if (start_tag.type() != Tag::START)
return start_tag;
2011-12-22 16:19:25 +01:00
int depth = 1;
Token curr_token = start_tag.next_token();
2011-12-22 16:19:25 +01:00
while (curr_token.type() != Token::END) {
2011-12-22 16:19:25 +01:00
/* eat XML comment */
Comment curr_comment(curr_token);
if (curr_comment.valid()) {
curr_token = curr_comment.next_token();
continue;
}
2011-12-22 16:19:25 +01:00
/* skip all tokens that are no tags */
Tag curr_tag(curr_token);
if (curr_tag.type() == Tag::INVALID) {
curr_token = curr_token.next();
continue;
}
2011-12-22 16:19:25 +01:00
/* count sub nodes at depth 1 */
if (depth == 1 && curr_tag.node())
sub_nodes_count++;
2011-12-22 16:19:25 +01:00
/* keep track of the current depth */
depth += (curr_tag.type() == Tag::START);
depth -= (curr_tag.type() == Tag::END);
2011-12-22 16:19:25 +01:00
/* within sub nodes, continue after current token */
if (depth > 0) {
2011-12-22 16:19:25 +01:00
/* continue search with token after current tag */
curr_token = curr_tag.next_token();
continue;
2011-12-22 16:19:25 +01:00
}
/* reaching the same depth as the start tag */
char const *start_name = start_tag.name().start();
size_t start_len = start_tag.name().len();
char const *curr_name = curr_tag.name().start();
size_t curr_len = curr_tag.name().len();
2011-12-22 16:19:25 +01:00
/* on mismatch of start tag and end tag, return invalid tag */
if (start_len != curr_len
|| strcmp(start_name, curr_name, curr_len))
return Tag();
2011-12-22 16:19:25 +01:00
/* end tag corresponds to start tag */
return curr_tag;
}
return Tag();
}
2011-12-22 16:19:25 +01:00
/**
* Find next non-whitespace and non-comment token
*/
static Token skip_non_tag_characters(Token t)
{
while (true) {
t = t.eat_whitespace();
/* eat comment */
Comment comment(t);
if (comment.valid()) {
t = comment.next_token();
continue;
2011-12-22 16:19:25 +01:00
}
/* skip token if it is valid but does not start a tag */
Tag curr_tag(t);
if (curr_tag.type() == Tag::INVALID && curr_tag.token()) {
t = t.next();
continue;
}
break;
2011-12-22 16:19:25 +01:00
}
return t;
}
2011-12-22 16:19:25 +01:00
struct Tags
{
int num_sub_nodes = 0;
Tag start;
Tag end;
Tags(char const *addr, size_t max_len)
:
start(skip_non_tag_characters(Token(addr, max_len))),
end(_search_end_tag(start, num_sub_nodes))
{ }
} _tags;
/**
* Return true if specified buffer contains a valid XML node
*/
static bool _valid(Tags const &tags)
{
if (tags.start.type() == Tag::EMPTY)
return true;
if (tags.start.type() == Tag::START && tags.end.type() == Tag::END)
return true;
return false;
}
bool _valid_node_at(char const *at) const
{
bool const in_range = (at >= _addr && (size_t)(at - _addr) < _max_len);
return in_range && _valid(Tags(at, _max_len - (at - _addr)));
}
/**
* Create sub node from XML node
*
* \throw Nonexistent_sub_node
* \throw Invalid_syntax
*/
Xml_node _node_at(char const *at) const
{
if (!_valid_node_at(at))
throw Nonexistent_sub_node();
2011-12-22 16:19:25 +01:00
return Xml_node(at, _max_len - (at - _addr));
}
2011-12-22 16:19:25 +01:00
/**
* Return pointer to start of content
*/
char const *_content_base() const { return _tags.start.next_token().start(); }
public:
2011-12-22 16:19:25 +01:00
/**
* Constructor
*
* The constructor validates if the start tag has a matching end tag of
* the same depth and counts the number of immediate sub nodes.
*
* \throw Invalid_syntax
*/
Xml_node(char const *addr, size_t max_len = ~0UL)
:
_addr(addr), _max_len(max_len), _tags(addr, max_len)
{
if (!_valid(_tags))
throw Invalid_syntax();
}
2011-12-22 16:19:25 +01:00
/**
* Request type name of XML node as null-terminated string
*
* \noapi
*/
void type_name(char *dst, size_t max_len) const __attribute__((deprecated))
{
_tags.start.name().string(dst, max_len);
}
2011-12-22 16:19:25 +01:00
/**
* Return size of node including start and end tags in bytes
*/
size_t size() const { return _tags.end.next_token().start() - _addr; }
/**
* Return pointer to start of node
*
* \deprecated use 'with_raw_node' instead
* \noapi
*/
char const *addr() const __attribute__((deprecated)) { return _addr; }
/**
* Return size of node content
*/
size_t content_size() const
{
if (_tags.start.type() == Tag::EMPTY)
return 0;
return _tags.end.token().start() - _content_base();
}
/**
* Request type name of XML node as null-terminated string
*/
typedef String<64> Type;
Type type() const
{
Token name = _tags.start.name();
return Type(Cstring(name.start(), name.len()));
}
/**
* Return true if tag is of specified type
*/
bool has_type(char const *type) const {
return (!strcmp(type, _tags.start.name().start(),
_tags.start.name().len())
&& strlen(type) == _tags.start.name().len()); }
2011-12-22 16:19:25 +01:00
/**
* Call functor 'fn' with the node data '(char const *, size_t)'
*/
template <typename FN>
void with_raw_node(FN const &fn) const
{
char const *start_ptr = _tags.start.token().start();
fn(start_ptr, _tags.end.next_token().start() - start_ptr);
}
2011-12-22 16:19:25 +01:00
/**
* Call functor 'fn' with content '(char const *, size_t) as argument'
*
* Note that the content is not null-terminated. It points directly
* into a sub range of the unmodified 'Xml_node' data.
*
* If the node has no content, the functor 'fn' is not called.
*/
template <typename FN>
void with_raw_content(FN const &fn) const
{
if (_tags.start.type() == Tag::EMPTY)
return;
2011-12-22 16:19:25 +01:00
fn(_content_base(), content_size());
}
/**
* Return pointer to start of content
*
* \deprecated use 'with_raw_content()' instead
* \noapi
*/
char *content_addr() const __attribute__((deprecated))
{
return _tags.start.next_token().start();
}
2011-12-22 16:19:25 +01:00
/**
* Return pointer to start of content
*
* \deprecated use 'with_raw_content()' instead
*
* \noapi
*/
char const *content_base() const __attribute__((deprecated))
{
return _tags.start.next_token().start();
}
2011-12-22 16:19:25 +01:00
/**
* Export decoded node content from XML node
*
* \param dst destination buffer
* \param dst_len size of destination buffer in bytes
* \return number of bytes written to the destination buffer
*
* This function transforms XML character entities into their
* respective characters.
*/
size_t decoded_content(char *dst, size_t dst_len) const
{
size_t result_len = 0;
char const *src = _content_base();
size_t src_len = content_size();
for (; dst_len && src_len; dst_len--, result_len++) {
Decoded_character const decoded_character(src, src_len);
*dst++ = decoded_character.character;
src += decoded_character.encoded_len;
src_len -= decoded_character.encoded_len;
}
return result_len;
}
/**
* Read decoded node content as Genode::String
*/
template <typename STRING>
STRING decoded_content() const
{
char buf[STRING::capacity() + 1];
size_t const len = decoded_content(buf, sizeof(buf) - 1);
buf[min(len, sizeof(buf) - 1)] = 0;
return STRING(Cstring(buf));
}
2011-12-22 16:19:25 +01:00
/**
* Return the number of the XML node's immediate sub nodes
*/
size_t num_sub_nodes() const { return _tags.num_sub_nodes; }
2011-12-22 16:19:25 +01:00
/**
* Return XML node following the current one
*
* \throw Nonexistent_sub_node subsequent node does not exist
*/
Xml_node next() const
{
Token after_node = _tags.end.next_token();
after_node = skip_non_tag_characters(after_node);
try {
return _node_at(after_node.start());
}
catch (Invalid_syntax) { throw Nonexistent_sub_node(); }
}
2011-12-22 16:19:25 +01:00
/**
* Return next XML node of specified type
*
* \param type type of XML node, or nullptr for matching any type
*
* \throw Nonexistent_sub_node subsequent node does not exist
*/
Xml_node next(char const *type) const
{
Xml_node node = next();
for (; type && !node.has_type(type); node = node.next());
return node;
}
2011-12-22 16:19:25 +01:00
/**
* Return true if node is the last of a node sequence
*/
bool last(char const *type = nullptr) const
{
Token after = _tags.end.next_token();
after = skip_non_tag_characters(after);
for (;;) {
if (!_valid_node_at(after.start()))
return true;
Xml_node node = _node_at(after.start());
if (!type || node.has_type(type))
return false;
after = node._tags.end.next_token();
after = skip_non_tag_characters(after);
}
}
/**
* Return sub node with specified index
*
* \param idx index of sub node,
* default is the first node
* \throw Nonexistent_sub_node no such sub node exists
*/
Xml_node sub_node(unsigned idx = 0U) const
{
if (_tags.num_sub_nodes > 0) {
try {
Xml_node curr_node = _node_at(_content_base());
for (; idx > 0; idx--)
curr_node = curr_node.next();
return curr_node;
} catch (Invalid_syntax) { }
2011-12-22 16:19:25 +01:00
}
throw Nonexistent_sub_node();
}
2011-12-22 16:19:25 +01:00
/**
* Return first sub node that matches the specified type
*
* \throw Nonexistent_sub_node no such sub node exists
*/
Xml_node sub_node(char const *type) const
{
if (_tags.num_sub_nodes > 0) {
/* search for sub node of specified type */
try {
Xml_node curr_node = _node_at(_content_base());
for ( ; true; curr_node = curr_node.next())
if (!type || curr_node.has_type(type))
return curr_node;
} catch (...) { }
2011-12-22 16:19:25 +01:00
}
throw Nonexistent_sub_node();
}
/**
* Apply functor 'fn' to first sub node of specified type
*
* The functor is called with the sub node as argument.
* If no matching sub node exists, the functor is not called.
*/
template <typename FN>
void with_sub_node(char const *type, FN const &fn) const
{
if (has_sub_node(type))
fn(sub_node(type));
}
/**
* Execute functor 'fn' for each sub node of specified type
*/
template <typename FN>
void for_each_sub_node(char const *type, FN const &fn) const
{
if (!has_sub_node(type))
return;
for (Xml_node node = sub_node(type); ; node = node.next()) {
if (!type || node.has_type(type))
fn(node);
if (node.last())
break;
}
}
/**
* Execute functor 'fn' for each sub node
*/
template <typename FN>
void for_each_sub_node(FN const &fn) const
{
for_each_sub_node(nullptr, fn);
}
2011-12-22 16:19:25 +01:00
/**
* Return Nth attribute of XML node
*
* \param idx attribute index,
* first attribute has index 0
* \throw Nonexistent_attribute no such attribute exists
* \return XML attribute
*/
Xml_attribute attribute(unsigned idx) const
{
Xml_attribute attr = _tags.start.attribute();
for (unsigned i = 0; i < idx; i++)
attr = Xml_attribute(attr._next_token());
2011-12-22 16:19:25 +01:00
return attr;
}
/**
* Return attribute of specified type
*
* \param type name of attribute type
* \throw Nonexistent_attribute no such attribute exists
* \return XML attribute
*/
Xml_attribute attribute(char const *type) const
{
for (Xml_attribute attr = _tags.start.attribute(); ;) {
if (attr.has_type(type))
return attr;
attr = Xml_attribute(attr._next_token());
}
throw Nonexistent_attribute();
}
/**
* Read attribute value from XML node
*
* \param type attribute name
* \param default_value value returned if no attribute with the
* name 'type' is present.
* \return attribute value or specified default value
*
* The type of the return value corresponds to the type of the
* default value.
*/
template <typename T>
T attribute_value(char const *type, T const default_value) const
{
T result = default_value;
if (!_tags.start.has_attribute())
return result;
for (Xml_attribute attr = _tags.start.attribute(); ; ) {
/* match */
if (attr.has_type(type)) {
attr.value(result);
return result;
}
/* end of attribute */
Token const next = attr._next_token();
if (!Xml_attribute::_valid(next))
break;
/* keep searching */
attr = Xml_attribute(next);
}
return result;
}
2013-11-19 14:18:08 +01:00
/**
* Return true if attribute of specified type exists
*/
bool has_attribute(char const *type) const
{
if (!_tags.start.has_attribute())
return false;
if (type == nullptr)
return true;
for (Xml_attribute attr = _tags.start.attribute(); ; ) {
if (attr.has_type(type))
return true;
Token const next = attr._next_token();
if (!Xml_attribute::_valid(next))
return false;
attr = Xml_attribute(next);
}
}
/**
* Return true if sub node of specified type exists
*/
bool has_sub_node(char const *type) const
{
if (_tags.num_sub_nodes == 0)
return false;
if (!_valid_node_at(_content_base()))
return false;
if (type == nullptr)
return true;
/* search for node of given type */
for (Xml_node node = _node_at(_content_base()); ; node = node.next()) {
if (node.has_type(type))
return true;
if (node.last())
break;
}
return false;
}
void print(Output &output) const {
output.out_string(_addr, size()); }
/**
* Return true if this node differs from 'another'
*/
bool differs_from(Xml_node const &another) const
{
return size() != another.size() ||
memcmp(_addr, another._addr, size()) != 0;
}
};
2011-12-22 16:19:25 +01:00
/**
* Utility for unquoting XML attributes
*
* The 'Xml_unquoted' utility can be used to revert quoted XML attribute
* values. Such quoting is needed whenever an attribute value can contain '"'
* characters.
*/
class Genode::Xml_unquoted : Noncopyable
{
private:
struct
{
char const *base;
size_t len;
} const _content_ptr;
public:
template <size_t N>
Xml_unquoted(String<N> const &string)
: _content_ptr({ string.string(), string.length() - 1})
{ }
void print(Output &out) const
{
char const *src = _content_ptr.base;
size_t len = _content_ptr.len;
while (len > 0) {
Xml_node::Decoded_character const decoded_character(src, len);
Genode::print(out, Char(decoded_character.character));
src += decoded_character.encoded_len;
len -= decoded_character.encoded_len;
}
}
};
2011-12-22 16:19:25 +01:00
#endif /* _INCLUDE__UTIL__XML_NODE_H_ */