JarCastingBase data structures, algorithms and utility function
This project contains a set of base data structures, algorithms and utility functions which are used in several other projects. These components have been developed to be generic and of wide applicability. When an implementation of a component in this project has subsequently become available in the JDK, it should be deprecated in favour of the JDK implementation.
Install
To use through Maven, include the following dependency in your pom.xml file:
<dependency>
<groupId>com.vikmad</groupId>
<artifactId>base</artifactId>
<version>0.2.7</version>
</dependency>
or in Gradle, add the following line to your dependencies list in your build.gradle file:
compile "com.vikmad:base:0.2.7"
Components and example usage
Following are overview of the some of the components in this project. More information is available in the javadocs.
Fast and simple streaming-read of XML files
The ma.vi.base.xml.XmlReader class provides a simplified iterator interface to a streaming XML parser with support for backtracking an arbitrary number of steps, specified in the constructor or, better, through ma.vi.base.xml.XmlReaderBuilder. XmlReader is an iterator of Fragment with each Fragment carrying information on its type (e.g. element start, element end, comment, text, etc.) and content. Example use:
XmlReader xml = XmlReaderBuilder.newBuilder(new StringReader(xmlText))
.discardInterElementSpaces(false) // do not discard spaces between elements
.discardComments(false) // do not discard comments
.rewindCapacity(10) // allow rewinding up to 10 positions back
.build();
for (Fragment fragment: xml) {
// do something with the fragment
}
Tuples
ma.vi.base.tuple.T1 to ma.vi.base.tuple.T5 are one to five elements tuples inheriting from a common interface and intended as a convenience when a function need to return several values. Tuples with more components can be created by subclassing the AbstractTuple class or implementing the Tuple interface.
Tries
ma.vi.base.trie.Trie is a generic Trie structure which can be used to implement tries with any sequence types. The StringTrie subclass implements the classic tries on strings and the PathTrie subclass extends this concept to paths which are sequence of path elements (files and folders).
Union-Find
ma.vi.base.unionfind.UnionFind is an extended implementation of the union-find structure backed by as hashmap, implementing path-compression, and allowing for arbitrary values to be associated to each disjoint set. Operations to find and merge sets are accomplished in amortized constant time in this structure.
String escaping and utilities
ma.vi.base.string.Escape can be used to find, escape and unescape certain characters in strings which is useful when parsing certain arbitrary character sequences. It works by temporarily mapping a set of special characters to the unicode private area and remapping them back to unescape.
ma.vi.base.string.Strings contains static functions to generate random strings and to manipulate strings.
ma.vi.base.crypt.Obfuscator can obfuscate phrases which need to be stored or transmitted in some form other than plain text. This is not an encryption function as it does not use a key; it merely manipulates the characters in the text and inject certain random characters to obscure the original phrase in a manner that it can be revealed with a reverse function.
Reflection
ma.vi.base.reflect.Dissector uses Java reflection to provide access to the different elements of a class. ma.vi.base.reflect.Property is an abstraction of fields and properties (get and set accessors) allowing their values to be read and written irrespective of how they are implemented (e.g., as private field with or without accessors).
Collections
ma.vi.base.collections.DiskBasedCollection is an ordered collection of objects that is transparently stored on disk allowing it to grow to huge size with minimal impact on RAM. The objects are stored in two files on disk, with an index file containing pointers to a data file. The index file is structured as a self-balancing AA tree and both files are accessed through ma.vi.base.cache.CachedRandomAccessFile which optimizes read and write access to random-access files.
The ma.vi.base.collections package contains other classes to facilitate working with maps, sets, arrays, etc.