row_base

A xlang::meta::reader::row_base<Row> structure represents a row in a database table, as well as a random-access iterator to that row.

The row_base<Row> is unusual in that it is its own iterator. (Dereferencing the row_base<Row> returns itself!)

A row_base<Row> can be empty. All operations on an empty row_base<Row> are invalid, aside from the operator bool() that reports whether the object is valid or empty.

Do not construct a row_base<Row> object directly. Use only row_base<Row> objects obtained from the database.

A row_base<Row> object is valid only as long as its corresponding database is loaded.

Members

operator bool

operator bool() const noexcept;

Returns true if the row_base<Row> is valid, or false if it is empty. If the row_base<Row> is empty, then no other methods may be called.

Performance: O(1). Very fast.

get_database

xlang::meta::reader::database const& get_database() const noexcept;

Returns a reference to the database that contains this row.

Performance: O(1). Very fast.

get_cache

xlang::meta::reader::cache const& get_cache() const noexcept;

Returns a reference to the cache that contains this row's database.

Performance: O(1). Very fast.

Iterator members

The row_base<Row> defines the operations and namespace types as required for it to be considered a random-access iterator by standard algorithms.

Namespace types

• iterator_category is std::random_access_iterator_tag.
• value_type is Row, the type this row represents.
• difference_type is int32_t.
• pointer is Row*.
• reference is Row&.
• const_reference is Row const&.

Operations

• row + n and row - n return a new iterator that is ahead of or behind the current iterator.
• ++row, row++, --row, and row-- move the iterator forward or backward one step.
• row +=n and row -= n move the iterator forward or backward by multiple steps.
• *row accesses the object pointed to by the iterator.
• row[n] accesses the object n steps ahead of (or behind, if negative) the current iterator.
• row1 - row2 returns the signed distance between the two rows.
• All relational operators are supported in the natural way.

Row access members

You will typically not call these members directly, but will instead use wrapper functions provided by derived types.

get_value

template<typename T> T
get_value(uint32_t column) const;

Returns the value in the specified column of the row.

T must be an enumeration or an unsigned integer type.

If T is not large enough to represent the value of the column, the value is truncated.

Performance: O(1). This looks up data in the table and parses it.

Exceptions: Throws std::invalid_argument if the row is not valid. This typically is the result of a corrupted database.

get_list

template<typename T> std::pair<T, T>
get_list(uint32_t column) const;

Returns a range of T objects (in the form of a pair of iterators) referenced by the specified column of the row.

Performance: O(1). This looks up multiple pieces of data in the table.

Exceptions: Throws std::invalid_argument if the row is not valid. This typically is the result of a corrupted database.

• get_target_row

template<typename T> xmr::row_base<T>
get_target_row(uint32_t column) const;

Returns a row from the T table as indexed by the value in the indicated column. The returned row is not validated.

Performance: O(1). This looks up data in the table and parses it.

Exceptions: Throws std::invalid_argument if the row is not valid. This typically is the result of a corrupted database.

• get_parent_row

template<typename T, uint32_t ParentColumn>
xmr::row_base<T> get_parent_row() const;

Returns a row from the T table as indexed by the value in the indicated column of the T table. The returned row is not validated.

This is the inverse of get_target_row. The T table must be sorted on the ParentColumn, and the ParentColumn must contain the index of a row in the current table.

Performance: O(log n) where n is the size of the T table.

Exceptions: Throws std::invalid_argument if the row is not valid. This typically is the result of a corrupted database.

Helper members

These members are primarily for internal use by the database management code.

index

uint32_t index() const noexcept;

Returns the row number of the row within its table.

Performance: O(1). Very fast.

coded_index<T>

template<typename T> xmr::coded_index<T> coded_index() const noexcept;

Returns a coded_index that captures information on how to locate this row.

This method is useful if you have a row, and you need to produce a particular kind of coded_index that refers to it. For example, you have a TypeDef object, but you need to pass it to a function that expects a coded_index<TypeDefOrRef>. Call coded_index<TypeDefOrRef>() to produce the value that resolves back to the original TypeDef object.

Arguably, since this is a conversion, it should be named as_coded_index.

Performance: O(1). Very fast.