Text parsing is a variation of parsing which refers to the action of breaking a stream of text into different components, and capturing the relationship between those components.
When the stream of text is arbitrary, parsing is often used to mean breaking the stream into constituent atoms (words or lexemes).
When the stream of text corresponds to natural language, 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 string of text corresponds to a computer source 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.
