https://github.com/azrdev/coco-idea-plugin
Jetbrains IDE (Intellij IDEA, PyStorm, CLion, ...) support for Coco/R compiler generator grammar files
https://github.com/azrdev/coco-idea-plugin
intellij intellij-plugin jetbrains-ides
Last synced: 2 months ago
JSON representation
Jetbrains IDE (Intellij IDEA, PyStorm, CLion, ...) support for Coco/R compiler generator grammar files
- Host: GitHub
- URL: https://github.com/azrdev/coco-idea-plugin
- Owner: azrdev
- Created: 2016-12-21T21:03:46.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2018-03-17T21:41:23.000Z (about 7 years ago)
- Last Synced: 2025-02-07T19:35:01.022Z (4 months ago)
- Topics: intellij, intellij-plugin, jetbrains-ides
- Language: Java
- Homepage: https://plugins.jetbrains.com/idea/plugin/9457-coco-r-grammar-support
- Size: 231 KB
- Stars: 2
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Coco/R grammar support for Jetbrains IDEs
Plugin for [Jetbrains](https://www.jetbrains.com/) IDEs
to support grammar files (\*.atg) of the [Coco/R](http://ssw.jku.at/Coco/) compiler generator.
Check it out from [Jetbrains' plugin repository](https://plugins.jetbrains.com/idea/plugin/9457-coco-r-grammar-support).
## Implementation
Following the [Custom Language Support Tutorial](http://www.jetbrains.org/intellij/sdk/docs/tutorials/custom_language_support_tutorial.html), the lexer is generated from Coco.flex using JFlex, and the parser from Coco.bnf using [Grammar-Kit](https://github.com/JetBrains/Grammar-Kit).
# TODOs
* language injection
- configurable language and/or good guess
- prefix & suffix as needed
- file header
- folding: comments, injected blocks (esp. at file start)
- references: not resolved in IGNORE, COMMENTS, PRAGMAS
- customizable syntax highlighting: [Color Settings Page](http://www.jetbrains.org/intellij/sdk/docs/tutorials/custom_language_support/syntax_highlighter_and_color_settings_page.html)
- syntax highlighting: different colors for keywords, charsets, tokens, productions
- PRAGMAs in structure view
- `$CNF` file header
- code completion, templates, "new file" (template), build system integration, quick fix, code style, ...
# Tutorial deficits
- "register" occurs in plugin.xml
- for references: methods from .bnf are delegated to psiImplUtilClass, *if signature fits*
- bnf: methods(nonterminals)=[] are not resolved correctly in "attributes" header
- bug in `SimpleStructureViewModel.isAlwaysLeaf`: `element instanceof SimpleFile` is always false, it should be `treeElement.psiElement instanceof ...`
- `PsiReference.resolve()` is only called when you return a valid value from `.rangeInsideHost()` (which `PsiReferenceBase` does *not* do)
What may help understanding the tutorial:
- Rust [JFlex Lexer](https://github.com/intellij-rust/intellij-rust/blob/master/src/main/grammars/RustLexer.flex) and [Grammar-Kit Parser](https://github.com/intellij-rust/intellij-rust/blob/master/src/main/grammars/RustParser.bnf)
- Monkey C [Grammar-Kit Parser](https://github.com/liias/monkey/blob/master/src/main/java/io/github/liias/monkey/lang/grammar/Monkey.bnf)
- meta-plugin for using ANTLR-grammar in jetbrains IDEs: [Code](https://github.com/antlr/jetbrains) and [forum announcement](https://intellij-support.jetbrains.com/hc/en-us/community/posts/206103369-Using-ANTLR-v4-to-lex-parse-custom-file-formats)
## howto language embedding, other plugins & forum posts
-
gitter.im Alexander Zolotov @zolotov 2017-08-15 09:17
> You can properly lex/parse A language inside B language to get rid of injection A into B. We do so for CSS and JS in HTML, or for Python in SQL, so you can write JS in HTML and inject something into that JS.
Take a look at com.intellij.psi.impl.source.tree.LazyParseablePsiElement and com.intellij.lexer.EmbeddedTokenTypesProvider (this one is for embedding into HTML)
[This thread about the *Play!* plugin](https://intellij-support.jetbrains.com/hc/en-us/community/posts/206780275-Example-of-a-custom-language-plugin-for-a-templating-language)[^1] explains the differences between the three possible embedding "types":
- LanguageSubstitutor: replace Language of an already handled file (e.g. HTML) with your own lang
* (Perl?)
- Language injection: embed your language (fragments) in already handled file, e.g. properties in Java string[^2]
* Documentation for [Language Injections](https://www.jetbrains.com/help/idea/2016.3/using-language-injections.html). References from embedded fragments are explained [in the tutorial](http://www.jetbrains.org/intellij/sdk/docs/tutorials/custom_language_support/reference_contributor.html#define-a-reference-contributor)
* (CUP)
* (HTL, but there are other plugins for HTL, too)
- MultiplePsiFilesPerDocumentFileViewProvider: create several different Psi trees for one file. Have to use a layered Lexer (like [Scala](https://github.com/JetBrains/intellij-scala/blob/32dd1c4/src/org/jetbrains/plugins/scala/lang/lexer/LayeredLexer.java) or *Latte*)?
* (Latte)
* (Perl/Pod), [Perl plugin dev thread](https://intellij-support.jetbrains.com/hc/en-us/community/posts/206102159-Perl5-plugin-for-Intellij-IDEA)
* (Elixir/ERB)
* (Handlebars)
* thoughts on multiple languages' parser interaction
* how to implement references between different language Psis (another HTL plugin)
The last paragraph in [Custom Language Tutorial/Lexer](http://www.jetbrains.org/intellij/sdk/docs/reference_guide/custom_language_support/implementing_lexer.html) tells about mixing languages using chameleon tokens implementing `ILazyParseableElementType`, probably that's the method with multiple Psi files.
- ,
[^1]: For *Play!* see also