Parsing refers to breaking an artifact into its constituent elements and capturing the relationship between those elements. This tag isn't for questions about the self hosted Parse Platform (use the [parse-platform] tag) or parse errors in a particular programming language (use the appropriate language tag instead).
Parsing refers to the action by software of breaking an artifact into its constituent elements and capturing the relationship between those elements.
When the artifact is a stream of arbitrary text, parsing is often used to mean breaking the stream into constituent atoms (called words, tokens or lexemes).
When the artifact is a stream of natural language text, parsing is used to mean breaking the stream into natural language elements (words and punctuation) and discovering the structure of the text as phrases or sentences.
When the artifact is a stream of text corresponding to a computer language (or other formal language), parsing consists of applying any of a variety of parsing algorithms (ad hoc, recursive descent, LL, LR, Packrat, Earley or other) to the source text (often broken into lexemes by another lower level parser called a "lexer") to verify the validity of the source language, and often to construct a parse tree representing the grammar productions used to tile the text.
The term can be applied more generally to analyzing any complex structure such as a binary data file or a graph.
