micro/src/preprocessor.rs

use crate::tokenizer::Token;
use regex::{Captures, Match, Regex};
use toml::{Table, Value};

// MetaRules
// Struct containing all meta rules.
pub struct MetaRules {
    replacement_rules: Vec<(String, (String, String))>,
    interpolation_rules: Vec<(String, (String, String))>,
    token_rules: Vec<(String, String)>,
    pub special_tokens: Vec<(String, Token)>,
}

// Implementation of MetaRules
// Trait implementation
impl MetaRules {
    // @name new
    // @return MetaRules
    // @brief Create a new rule struct by reading from a configuration file.
    // @param configuration_filename: &str
    pub fn new(configuration_filename: &str) -> MetaRules {
        let configuration_content: String = std::fs::read_to_string(configuration_filename)
            .expect("[ERROR] Could not open configuration file!");
        let mut replacements: Vec<(String, (String, String))> = vec![];
        let mut interpolation: Vec<(String, (String, String))> = vec![];
        let mut meta_token_rules: Vec<(String, String)> = vec![];
        let meta_tokens: Vec<(String, Token)> = vec![];
        let configuration = gtoml::parse(configuration_content.as_str())
            .expect("[ERROR] TOML invalid in preprocessor!");
        let configuration_unpacked: Table = Table::try_from(configuration).unwrap();
        let meta_configuration: Table = match configuration_unpacked.get("meta") {
            Some(config) => config.as_table().unwrap().clone(),
            None => Table::new(),
        };

        if !meta_configuration.is_empty() {
            if meta_configuration.contains_key("replacements") {
                println!("[INFO] Found replacement rules.");
                let replacement_rules: Table = meta_configuration
                    .get("replacements")
                    .unwrap()
                    .as_table()
                    .unwrap()
                    .clone();
                for key in replacement_rules.keys() {
                    let value: Vec<Value> = replacement_rules
                        .get(key)
                        .unwrap()
                        .as_array()
                        .unwrap()
                        .clone();
                    let name: String = key.clone();
                    let pattern: String = value[0].as_str().unwrap().to_owned();
                    let replacement: String = value[1].as_str().unwrap().to_owned();
                    replacements.push((name, (pattern, replacement)));
                }
            }

            if meta_configuration.contains_key("interpolation") {
                println!("[INFO] Found interpolation rules.");
                let interpolation_rules: Table = meta_configuration
                    .get("interpolation")
                    .unwrap()
                    .as_table()
                    .unwrap()
                    .clone();
                for key in interpolation_rules.keys() {
                    let value: Vec<Value> = interpolation_rules
                        .get(key)
                        .unwrap()
                        .as_array()
                        .unwrap()
                        .clone();
                    let name: String = key.clone();
                    let pattern: String = value[0].as_str().unwrap().to_owned();
                    let cmd: &str = value[1].as_str().unwrap();
                    interpolation.push((name, (pattern, String::from(cmd))));
                }
            }

            if meta_configuration.contains_key("token") {
                println!("[INFO] Found token rules.");
                let token_rules: Table = meta_configuration
                    .get("token")
                    .unwrap()
                    .as_table()
                    .unwrap()
                    .clone();
                for rule in token_rules.keys() {
                    let pattern: String =
                        token_rules.get(rule).unwrap().as_str().unwrap().to_owned();
                    meta_token_rules.push((rule.clone(), pattern));
                }
            }
        } else {
            println!("[WARNING] No meta configuration, skipping preprocessor.");
        }

        MetaRules {
            replacement_rules: replacements,
            interpolation_rules: interpolation,
            token_rules: meta_token_rules,
            special_tokens: meta_tokens,
        }
    }

    // @name process
    // @return String
    // @brief Run preprocessor on raw code.
    // @param rule_set: MetaRules, raw_code: String
    pub fn process(&mut self, raw_code: String) -> String {
        let mut processed_code: String = raw_code.clone();

        // replacement rules
        for rule in self.replacement_rules.iter() {
            println!("[INFO] Applying rule {}", rule.0);
            let base_pattern: Regex = Regex::new((rule.1 .0).as_str()).unwrap();
            processed_code = base_pattern
                .replace_all(processed_code.as_str(), rule.1 .1.as_str())
                .to_string();
        }

        // interpolation rules
        for rule in self.interpolation_rules.iter() {
            println!("[INFO] Applying rule {}", rule.0);
            let base_pattern: Regex = Regex::new((rule.1 .0).as_str()).unwrap();
            let processed_code_replacement = processed_code.clone();

            let captures: Option<Captures> =
                base_pattern.captures(processed_code_replacement.as_str());
            let directive: String;
            match captures {
                Some(n) => directive = n.get(0).map_or("", |m| m.as_str()).to_string(),
                None => continue,
            };
            let command: &str = &base_pattern.replace(directive.as_str(), rule.1 .1.as_str());
            let subprocess = std::process::Command::new("/bin/bash")
                .arg("-c")
                .arg(String::from("echo \"$(") + command + ")\"")
                .output()
                .expect((String::from("") + "Failed to run command " + command + "!").as_str());
            processed_code = base_pattern
                .replace(
                    processed_code.as_str(),
                    String::from_utf8(subprocess.stdout).unwrap(),
                )
                .to_string();
        }

        for token_style in self.token_rules.iter() {
            println!("[INFO] Searching meta tokens of style {}", token_style.0);

            // Search all occurrences
            let token_pattern: Regex =
                Regex::new(token_style.1.as_str()).expect("Could not assign pattern.");
            let match_list: Match;
            match_list = match token_pattern.find(processed_code.as_str()) {
                Some(n) => n,
                None => continue,
            };

            // Create id for each occurrence
            let meta_id: String = String::from("meta_token_")
                + match_list.start().to_string().as_str()
                + "__"
                + match_list.end().to_string().as_str();

            // Replace token by id
            let meta_value: String = match_list.as_str().to_string();
            let value_regex: Regex =
                Regex::new(meta_value.as_str()).expect("Could not create pattern.");
            processed_code = value_regex
                .replace(processed_code.as_str(), meta_id.as_str())
                .to_string();
            println!("Replace {} with {}.", meta_value, meta_id);

            // Safe id and token
            self.special_tokens.push((
                meta_id,
                Token {
                    token: meta_value,
                    token_type: crate::TokenType::IDENTIFIER,
                },
            ));
        }

        return processed_code;
    }
}