zero-parser/tests/file_lexical_parser_tests.rs

mod tokenizer;

use crate::tokenizer::{nom_lexical_parser, zero_lexical_parser};
use anyhow::anyhow;
use std::fs::File;
use std::io::{BufReader, Read};
use std::path::PathBuf;
use std::{env, fs};
use zero_parser::parser::ParserContext;

pub fn ensure_sysy_sources() -> anyhow::Result<Vec<PathBuf>> {
    let current_dir = env::current_dir()?;

    for entry in fs::read_dir(&current_dir)? {
        let entry = entry?;

        if entry.file_name() == "sysy_sets" {
            let set_path = entry.path();

            return Ok(fs::read_dir(&set_path)?
                .filter_map(|f| {
                    f.ok().and_then(|f| {
                        f.file_name()
                            .into_string()
                            .ok()
                            .and_then(|file_name| file_name.ends_with(".sy").then_some(f.path()))
                    })
                })
                .collect());
        }
    }

    Err(anyhow!(
        "Failed to find `sys_sets` directory in current dir {:?}.",
        current_dir.file_name().unwrap()
    ))
}

#[derive(Debug)]
pub struct SourceFile {
    pub filename: String,
    pub content: String,
}

impl SourceFile {
    pub fn new(path: &PathBuf) -> anyhow::Result<Self> {
        let file = File::open(path)?;
        let mut reader = BufReader::new(file);

        let mut content = String::new();
        reader.read_to_string(&mut content)?;

        Ok(Self {
            filename: path
                .file_name()
                .and_then(|x| x.to_str())
                .ok_or(anyhow!("Failed to get filename from path."))?
                .to_owned(),
            content,
        })
    }
}

#[test]
fn file_lexical_parser_test() -> anyhow::Result<()> {
    let sysy_set = ensure_sysy_sources()?;
    let source_files = sysy_set
        .into_iter()
        .map(|p| SourceFile::new(&p))
        .collect::<Result<Vec<SourceFile>, anyhow::Error>>()?;

    for source_file in source_files {
        println!("Start to test file '{}'.", source_file.filename);

        let (_, nom_tokens) =
            nom_lexical_parser(source_file.content.as_str()).or_else(|e| Err(e.to_owned()))?;

        let context = ParserContext::new_with_str(source_file.content.as_str(), ());
        let borrowed_context = context.borrow();
        let (_, zero_tokens) = zero_lexical_parser(context.clone(), borrowed_context.input_slice())
            .or_else(|e| Err(anyhow!("{}", e.1)))?;

        assert_eq!(nom_tokens.len(), zero_tokens.len());

        for (except, actual) in nom_tokens.iter().zip(zero_tokens.iter()) {
            assert_eq!(
                except,
                actual,
                "The literal value of actual token is {}.",
                actual
                    .literal_value
                    .iter()
                    .map(|x| x.clone())
                    .collect::<String>()
            );
        }
        println!(
            "Succeed to test file '{}', testing {} tokens.",
            source_file.filename,
            zero_tokens.len()
        );
    }

    Ok(())
}