https://github.com/illarion/lr1

Simple LR(1) parser
https://github.com/illarion/lr1

Last synced: about 2 months ago
JSON representation

Simple LR(1) parser

• Host: GitHub
• URL: https://github.com/illarion/lr1
• Owner: illarion
• License: mit
• Created: 2020-08-01T05:31:58.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-08-01T11:02:36.000Z (over 4 years ago)
• Last Synced: 2024-05-01T14:16:58.798Z (8 months ago)
• Language: Go
• Size: 2.93 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# lr1 parser 

This project is an example of LR(1) parser, and is useful for quick and dirty parsing of files that you don't have parser library for.

It consists of 2 elements -- `lr1.Parse(items []interface{}, rules ...Rule) ([]interface{}, error)` function and a number of rules `Rule(items []interface{}) (interface{}, error)`. You have to specify rules yourself. 

Algorithm is extremely simple - given the number of items, initially it may be just a list of tokenized strings from some text, Parse() takes last 1 element, then last 2 elements, 3, 4 and so on, and applies all specified rules (one by one) to this sublist. If some rule is applicable and returns the new value (`[]interface{}`), parser replaces the current end of the list with the replacement provided by rule. 

For example, let's say we parse following text, consisting of some imaginary language function calls:

```

foo(1,2);

bar(a,b,c);

```

## Step 1 - tokenize like this (see https://golang.org/pkg/text/scanner/ for more details) : 

`["foo", "(", "1", ",", "2", ")", ";", "bar", "(", "a", ",", "b", ",", "c", ")", ";"]`

## Step 2 - implement following rule functions

```

[token, ARGS, SEMICOLON] --> struct Function{ name: ..., }

["(", END_OF_LIST] -> ARGS

[token, ",", END_OF_LIST] -> END_OF_LIST // with appended token

[")"] -> END_OF_LIST`

[";"] -> SEMICOLON

```

## Step 3 - run the Parser

Result - list of Function{} structs, parsed from the initial list of string tokens, or error if it's not possible to parse

## Links

https://en.wikipedia.org/wiki/Canonical_LR_parser