All Downloads are FREE. Search and download functionalities are using the official Maven repository.

Java.target.apidocs.org.antlr.v4.runtime.CharStreams.html Maven / Gradle / Ivy

There is a newer version: 4.9.0
Show newest version






CharStreams (ANTLR 4 Runtime (Optimized) 4.7.3 API)











org.antlr.v4.runtime

Enum CharStreams

  • All Implemented Interfaces:
    Serializable, Comparable<CharStreams>


    public enum CharStreams
    extends Enum<CharStreams>
    This class represents the primary interface for creating CharStreams from a variety of sources as of 4.7. The motivation was to support Unicode code points > U+FFFF. ANTLRInputStream and ANTLRFileStream are now deprecated in favor of the streams created by this interface. DEPRECATED: new ANTLRFileStream("myinputfile") NEW: CharStreams.fromFileName("myinputfile") WARNING: If you use both the deprecated and the new streams, you will see a nontrivial performance degradation. This speed hit is because the Lexer's internal code goes from a monomorphic to megamorphic dynamic dispatch to get characters from the input stream. Java's on-the-fly compiler (JIT) is unable to perform the same optimizations so stick with either the old or the new streams, if performance is a primary concern. See the extreme debugging and spelunking needed to identify this issue in our timing rig: https://github.com/antlr/antlr4/pull/1781 The ANTLR character streams still buffer all the input when you create the stream, as they have done for ~20 years. If you need unbuffered access, please note that it becomes challenging to create parse trees. The parse tree has to point to tokens which will either point into a stale location in an unbuffered stream or you have to copy the characters out of the buffer into the token. That defeats the purpose of unbuffered input. Per the ANTLR book, unbuffered streams are primarily useful for processing infinite streams *during the parse.* The new streams also use 8-bit buffers when possible so this new interface supports character streams that use half as much memory as the old ANTLRFileStream, which assumed 16-bit characters. A big shout out to Ben Hamilton (github bhamiltoncx) for his superhuman efforts across all targets to get true Unicode 3.1 support for U+10FFFF.
    Since:
    4.7

Copyright © 1992–2019 Tunnel Vision Laboratories, LLC. All rights reserved.





© 2015 - 2024 Weber Informatics LLC | Privacy Policy