| 1 | // Take a look at the license at the top of the repository in the LICENSE file. | 
| 2 |  | 
|---|
| 3 | use std::{fmt, io}; | 
|---|
| 4 |  | 
|---|
| 5 | use crate::js::token::{self, Keyword, ReservedChar, Token, Tokens}; | 
|---|
| 6 | use crate::js::utils::{get_array, get_variable_name_and_value_positions, VariableNameGenerator}; | 
|---|
| 7 |  | 
|---|
| 8 | use std::collections::{HashMap, HashSet}; | 
|---|
| 9 |  | 
|---|
| 10 | /*#[derive(Debug, Clone, PartialEq, Eq)] | 
|---|
| 11 | enum Elem<'a> { | 
|---|
| 12 | Function(Function<'a>), | 
|---|
| 13 | Block(Block<'a>), | 
|---|
| 14 | Variable(Variable<'a>), | 
|---|
| 15 | Condition(token::Condition), | 
|---|
| 16 | Loop(Loop<'a>), | 
|---|
| 17 | Operation(Operation<'a>), | 
|---|
| 18 | } | 
|---|
| 19 |  | 
|---|
| 20 | impl<'a> Elem<'a> { | 
|---|
| 21 | fn is_condition(&self) -> bool { | 
|---|
| 22 | match *self { | 
|---|
| 23 | Elem::Condition(_) => true, | 
|---|
| 24 | _ => false, | 
|---|
| 25 | } | 
|---|
| 26 | } | 
|---|
| 27 | } | 
|---|
| 28 |  | 
|---|
| 29 | #[derive(Clone, Copy, PartialEq, Eq, Debug)] | 
|---|
| 30 | enum ConditionType { | 
|---|
| 31 | If, | 
|---|
| 32 | ElseIf, | 
|---|
| 33 | Else, | 
|---|
| 34 | Ternary, | 
|---|
| 35 | } | 
|---|
| 36 |  | 
|---|
| 37 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 38 | struct Block<'a> { | 
|---|
| 39 | elems: Vec<Elem<'a>>, | 
|---|
| 40 | } | 
|---|
| 41 |  | 
|---|
| 42 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 43 | struct Argument<'a> { | 
|---|
| 44 | name: &'a str, | 
|---|
| 45 | } | 
|---|
| 46 |  | 
|---|
| 47 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 48 | struct Function<'a> { | 
|---|
| 49 | name: Option<&'a str>, | 
|---|
| 50 | args: Vec<Argument<'a>>, | 
|---|
| 51 | block: Block<'a>, | 
|---|
| 52 | } | 
|---|
| 53 |  | 
|---|
| 54 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 55 | struct Variable<'a> { | 
|---|
| 56 | name: &'a str, | 
|---|
| 57 | value: Option<&'a str>, | 
|---|
| 58 | } | 
|---|
| 59 |  | 
|---|
| 60 | /*struct Condition<'a> { | 
|---|
| 61 | ty_: ConditionType, | 
|---|
| 62 | condition: &'a str, | 
|---|
| 63 | block: Block<'a>, | 
|---|
| 64 | }*/ | 
|---|
| 65 |  | 
|---|
| 66 | #[derive(Clone, Copy, PartialEq, Eq, Debug)] | 
|---|
| 67 | enum LoopType { | 
|---|
| 68 | Do, | 
|---|
| 69 | For, | 
|---|
| 70 | While, | 
|---|
| 71 | } | 
|---|
| 72 |  | 
|---|
| 73 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 74 | struct Loop<'a> { | 
|---|
| 75 | ty_: LoopType, | 
|---|
| 76 | condition: Vec<Elem<'a>>, | 
|---|
| 77 | block: Block<'a>, | 
|---|
| 78 | } | 
|---|
| 79 |  | 
|---|
| 80 | #[derive(Clone, PartialEq, Eq, Debug)] | 
|---|
| 81 | struct Operation<'a> { | 
|---|
| 82 | content: &'a str, | 
|---|
| 83 | } | 
|---|
| 84 |  | 
|---|
| 85 | fn get_while_condition<'a>(tokens: &[token::Token<'a>], pos: &mut usize) -> Result<Vec<Elem<'a>>, String> { | 
|---|
| 86 | let tmp = *pos; | 
|---|
| 87 | *pos += 1; | 
|---|
| 88 | if let Err(e) = match tokens.get(tmp) { | 
|---|
| 89 | Some(token::Token::Char(token::ReservedChar::OpenParenthese)) => Ok(()), | 
|---|
| 90 | Some(e) => Err(format!("Expected \"(\", found \"{:?}\"", e)), | 
|---|
| 91 | None => Err("Expected \"(\", found nothing...".to_owned()), | 
|---|
| 92 | } { | 
|---|
| 93 | return Err(e); | 
|---|
| 94 | } | 
|---|
| 95 | let mut elems: Vec<Elem<'a>> = Vec::with_capacity(1); | 
|---|
| 96 |  | 
|---|
| 97 | while let Some(e) = tokens.get(*pos) { | 
|---|
| 98 | *pos += 1; | 
|---|
| 99 | match e { | 
|---|
| 100 | token::Token::Char(token::ReservedChar::CloseParenthese) => return Ok(elems), | 
|---|
| 101 | token::Token::Condition(e) => { | 
|---|
| 102 | if let Some(cond) = elems.last() { | 
|---|
| 103 | if cond.is_condition() { | 
|---|
| 104 | return Err(format!("\"{:?}\" cannot follow \"{:?}\"", e, cond)); | 
|---|
| 105 | } | 
|---|
| 106 | } | 
|---|
| 107 | } | 
|---|
| 108 | _ => {} | 
|---|
| 109 | } | 
|---|
| 110 | } | 
|---|
| 111 | Err("Expected \")\", found nothing...".to_owned()) | 
|---|
| 112 | } | 
|---|
| 113 |  | 
|---|
| 114 | fn get_do<'a>(tokens: &[token::Token<'a>], pos: &mut usize) -> Result<Elem<'a>, String> { | 
|---|
| 115 | let tmp = *pos; | 
|---|
| 116 | *pos += 1; | 
|---|
| 117 | let block = match tokens.get(tmp) { | 
|---|
| 118 | Some(token::Token::Char(token::ReservedChar::OpenCurlyBrace)) => get_block(tokens, pos, true), | 
|---|
| 119 | Some(e) => Err(format!("Expected \"{{\", found \"{:?}\"", e)), | 
|---|
| 120 | None => Err("Expected \"{\", found nothing...".to_owned()), | 
|---|
| 121 | }?; | 
|---|
| 122 | let tmp = *pos; | 
|---|
| 123 | *pos += 1; | 
|---|
| 124 | let condition = match tokens.get(tmp) { | 
|---|
| 125 | Some(token::Token::Keyword(token::Keyword::While)) => get_while_condition(tokens, pos), | 
|---|
| 126 | Some(e) => Err(format!("Expected \"while\", found \"{:?}\"", e)), | 
|---|
| 127 | None => Err("Expected \"while\", found nothing...".to_owned()), | 
|---|
| 128 | }?; | 
|---|
| 129 | let mut loop_ = Loop { | 
|---|
| 130 | ty_: LoopType::Do, | 
|---|
| 131 | condition: condition, | 
|---|
| 132 | block, | 
|---|
| 133 | }; | 
|---|
| 134 | Ok(Elem::Loop(loop_)) | 
|---|
| 135 | } | 
|---|
| 136 |  | 
|---|
| 137 | fn get_block<'a>(tokens: &[token::Token<'a>], pos: &mut usize, | 
|---|
| 138 | start_with_paren: bool) -> Result<Block<'a>, String> { | 
|---|
| 139 | let mut block = Block { elems: Vec::with_capacity(2) }; | 
|---|
| 140 | while let Some(e) = tokens.get(*pos) { | 
|---|
| 141 | *pos += 1; | 
|---|
| 142 | block.elems.push(match e { | 
|---|
| 143 | token::Token::Keyword(token::Keyword::Do) => get_do(tokens, pos), | 
|---|
| 144 | token::Token::Char(token::ReservedChar::CloseCurlyBrace) => { | 
|---|
| 145 | if start_with_paren { | 
|---|
| 146 | return Ok(block); | 
|---|
| 147 | } | 
|---|
| 148 | return Err("Unexpected \"}\"".to_owned()); | 
|---|
| 149 | } | 
|---|
| 150 | }?); | 
|---|
| 151 | } | 
|---|
| 152 | if !start_with_paren { | 
|---|
| 153 | Ok(block) | 
|---|
| 154 | } else { | 
|---|
| 155 | Err("Expected \"}\" at the end of the block but didn't find one...".to_owned()) | 
|---|
| 156 | } | 
|---|
| 157 | } | 
|---|
| 158 |  | 
|---|
| 159 | fn build_ast<'a>(v: &[token::Token<'a>]) -> Result<Elem<'a>, String> { | 
|---|
| 160 | let mut pos = 0; | 
|---|
| 161 |  | 
|---|
| 162 | match get_block(v, &mut pos, false) { | 
|---|
| 163 | Ok(ast) => Ok(Elem::Block(ast)), | 
|---|
| 164 | Err(e) => Err(e), | 
|---|
| 165 | } | 
|---|
| 166 | }*/ | 
|---|
| 167 |  | 
|---|
| 168 | /// Minifies a given JS source code. | 
|---|
| 169 | /// | 
|---|
| 170 | /// # Example | 
|---|
| 171 | /// | 
|---|
| 172 | /// ```rust | 
|---|
| 173 | /// use minifier::js::minify; | 
|---|
| 174 | /// | 
|---|
| 175 | /// let js = r#" | 
|---|
| 176 | ///     function forEach(data, func) { | 
|---|
| 177 | ///        for (var i = 0; i < data.length; ++i) { | 
|---|
| 178 | ///            func(data[i]); | 
|---|
| 179 | ///        } | 
|---|
| 180 | ///     }"#.into(); | 
|---|
| 181 | /// let js_minified = minify(js); | 
|---|
| 182 | /// assert_eq!( | 
|---|
| 183 | ///     &js_minified.to_string(), | 
|---|
| 184 | /// "function forEach(data,func){for(var i=0;i<data.length;++i){func(data[i])}}", | 
|---|
| 185 | /// ); | 
|---|
| 186 | /// ``` | 
|---|
| 187 | #[ inline] | 
|---|
| 188 | pub fn minify(source: &str) -> Minified<'_> { | 
|---|
| 189 | Minified(token::tokenize(source).apply(func:crate::js::clean_tokens)) | 
|---|
| 190 | } | 
|---|
| 191 |  | 
|---|
| 192 | pub struct Minified<'a>(token::Tokens<'a>); | 
|---|
| 193 |  | 
|---|
| 194 | impl Minified<'_> { | 
|---|
| 195 | pub fn write<W: io::Write>(self, w: W) -> io::Result<()> { | 
|---|
| 196 | self.0.write(w) | 
|---|
| 197 | } | 
|---|
| 198 | } | 
|---|
| 199 |  | 
|---|
| 200 | impl fmt::Display for Minified<'_> { | 
|---|
| 201 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { | 
|---|
| 202 | self.0.fmt(f) | 
|---|
| 203 | } | 
|---|
| 204 | } | 
|---|
| 205 |  | 
|---|
| 206 | // TODO: No scope handling or anything. Might be nice as a second step to add it... | 
|---|
| 207 | fn get_variables_name<'a>( | 
|---|
| 208 | tokens: &'a Tokens<'a>, | 
|---|
| 209 | ) -> (HashSet<&'a str>, HashMap<&'a str, (usize, usize)>) { | 
|---|
| 210 | let mut ret = HashSet::new(); | 
|---|
| 211 | let mut variables = HashMap::new(); | 
|---|
| 212 | let mut pos = 0; | 
|---|
| 213 |  | 
|---|
| 214 | while pos < tokens.len() { | 
|---|
| 215 | if tokens[pos].is_keyword() || tokens[pos].is_other() { | 
|---|
| 216 | if let Some((var_pos, Some(value_pos))) = | 
|---|
| 217 | get_variable_name_and_value_positions(tokens, pos) | 
|---|
| 218 | { | 
|---|
| 219 | pos = value_pos; | 
|---|
| 220 | if let Some(var_name) = tokens[var_pos].get_other() { | 
|---|
| 221 | if !var_name.starts_with( "r_") { | 
|---|
| 222 | pos += 1; | 
|---|
| 223 | continue; | 
|---|
| 224 | } | 
|---|
| 225 | ret.insert(var_name); | 
|---|
| 226 | } | 
|---|
| 227 | if let Some(s) = tokens[value_pos].get_string() { | 
|---|
| 228 | variables.insert(s, (var_pos, value_pos)); | 
|---|
| 229 | } | 
|---|
| 230 | } | 
|---|
| 231 | } | 
|---|
| 232 | pos += 1; | 
|---|
| 233 | } | 
|---|
| 234 | (ret, variables) | 
|---|
| 235 | } | 
|---|
| 236 |  | 
|---|
| 237 | fn aggregate_strings_inner<'a, 'b: 'a>( | 
|---|
| 238 | mut tokens: Tokens<'a>, | 
|---|
| 239 | separation_token: Option<Token<'b>>, | 
|---|
| 240 | ) -> Tokens<'a> { | 
|---|
| 241 | let mut new_vars = Vec::with_capacity(50); | 
|---|
| 242 | let mut to_replace: Vec<(usize, usize)> = Vec::new(); | 
|---|
| 243 |  | 
|---|
| 244 | for (var_name, positions) in { | 
|---|
| 245 | let mut strs: HashMap<&Token<'_>, Vec<usize>> = HashMap::with_capacity(1000); | 
|---|
| 246 | let mut validated: HashMap<&Token<'_>, String> = HashMap::with_capacity(100); | 
|---|
| 247 |  | 
|---|
| 248 | let mut var_gen = VariableNameGenerator::new(Some( "r_"), 2); | 
|---|
| 249 | let mut next_name = var_gen.to_string(); | 
|---|
| 250 |  | 
|---|
| 251 | let (all_variables, values) = get_variables_name(&tokens); | 
|---|
| 252 | while all_variables.contains(&next_name.as_str()) { | 
|---|
| 253 | var_gen.next(); | 
|---|
| 254 | next_name = var_gen.to_string(); | 
|---|
| 255 | } | 
|---|
| 256 |  | 
|---|
| 257 | for pos in 0..tokens.len() { | 
|---|
| 258 | let token = &tokens[pos]; | 
|---|
| 259 | if let Some(str_token) = token.get_string() { | 
|---|
| 260 | if let Some((var_pos, string_pos)) = values.get(&str_token) { | 
|---|
| 261 | if pos != *string_pos { | 
|---|
| 262 | to_replace.push((pos, *var_pos)); | 
|---|
| 263 | } | 
|---|
| 264 | continue; | 
|---|
| 265 | } | 
|---|
| 266 | let x = strs.entry(token).or_insert_with(|| Vec::with_capacity(1)); | 
|---|
| 267 | x.push(pos); | 
|---|
| 268 | if x.len() > 1 && !validated.contains_key(token) { | 
|---|
| 269 | let len = str_token.len(); | 
|---|
| 270 | // Computation here is simple, we declare new variables when creating this so | 
|---|
| 271 | // the total of characters must be shorter than: | 
|---|
| 272 | // `var r_aa=...;` -> 10 + `r_aa` -> 14 | 
|---|
| 273 | if (x.len() + 2/* quotes */) * len | 
|---|
| 274 | > next_name.len() + str_token.len() + 6 /* var _=_;*/ + x.len() * next_name.len() | 
|---|
| 275 | { | 
|---|
| 276 | validated.insert(token, next_name.clone()); | 
|---|
| 277 | var_gen.next(); | 
|---|
| 278 | next_name = var_gen.to_string(); | 
|---|
| 279 | while all_variables.contains(&next_name.as_str()) { | 
|---|
| 280 | var_gen.next(); | 
|---|
| 281 | next_name = var_gen.to_string(); | 
|---|
| 282 | } | 
|---|
| 283 | } | 
|---|
| 284 | } | 
|---|
| 285 | } | 
|---|
| 286 | } | 
|---|
| 287 | let mut ret = Vec::with_capacity(validated.len()); | 
|---|
| 288 |  | 
|---|
| 289 | // We need this macro to avoid having to sort the set when not testing the crate. | 
|---|
| 290 | //#[cfg(test)] | 
|---|
| 291 | macro_rules! inner_loop { | 
|---|
| 292 | ($x:ident) => {{ | 
|---|
| 293 | let mut $x = $x.into_iter().collect::<Vec<_>>(); | 
|---|
| 294 | $x.sort_unstable_by(|a, b| a.1.cmp(&b.1)); | 
|---|
| 295 | $x | 
|---|
| 296 | }}; | 
|---|
| 297 | } | 
|---|
| 298 | /*#[cfg(not(test))] | 
|---|
| 299 | macro_rules! inner_loop { | 
|---|
| 300 | ($x:ident) => { | 
|---|
| 301 | $x.into_iter() | 
|---|
| 302 | } | 
|---|
| 303 | }*/ | 
|---|
| 304 |  | 
|---|
| 305 | for (token, var_name) in inner_loop!(validated) { | 
|---|
| 306 | ret.push((var_name, strs.remove(&token).unwrap())); | 
|---|
| 307 | var_gen.next(); | 
|---|
| 308 | } | 
|---|
| 309 | ret | 
|---|
| 310 | } { | 
|---|
| 311 | if new_vars.is_empty() { | 
|---|
| 312 | new_vars.push(Token::Keyword(Keyword::Var)); | 
|---|
| 313 | } else { | 
|---|
| 314 | new_vars.push(Token::Char(ReservedChar::Comma)); | 
|---|
| 315 | } | 
|---|
| 316 | new_vars.push(Token::CreatedVarDecl(format!( | 
|---|
| 317 | "{} ={} ", | 
|---|
| 318 | var_name, tokens[positions[0]] | 
|---|
| 319 | ))); | 
|---|
| 320 | for pos in positions { | 
|---|
| 321 | tokens.0[pos] = Token::CreatedVar(var_name.clone()); | 
|---|
| 322 | } | 
|---|
| 323 | } | 
|---|
| 324 | if !new_vars.is_empty() { | 
|---|
| 325 | new_vars.push(Token::Char(ReservedChar::SemiColon)); | 
|---|
| 326 | } | 
|---|
| 327 | for (to_replace_pos, variable_pos) in to_replace { | 
|---|
| 328 | tokens.0[to_replace_pos] = tokens.0[variable_pos].clone(); | 
|---|
| 329 | } | 
|---|
| 330 | if let Some(token) = separation_token { | 
|---|
| 331 | new_vars.push(token); | 
|---|
| 332 | } | 
|---|
| 333 | new_vars.append(&mut tokens.0); | 
|---|
| 334 | Tokens(new_vars) | 
|---|
| 335 | } | 
|---|
| 336 |  | 
|---|
| 337 | /// Aggregate litteral strings. For instance, if the string litteral "Oh look over there!" | 
|---|
| 338 | /// appears more than once, a variable will be created with this value and used everywhere the | 
|---|
| 339 | /// string appears. Of course, this replacement is only performed when it allows to take | 
|---|
| 340 | /// less space. | 
|---|
| 341 | /// | 
|---|
| 342 | /// # Example | 
|---|
| 343 | /// | 
|---|
| 344 | /// ```rust,no_run | 
|---|
| 345 | /// extern crate minifier; | 
|---|
| 346 | /// use minifier::js::{aggregate_strings, clean_tokens, simple_minify}; | 
|---|
| 347 | /// use std::fs; | 
|---|
| 348 | /// | 
|---|
| 349 | /// fn main() { | 
|---|
| 350 | ///     let content = fs::read( "some_file.js").expect( "file not found"); | 
|---|
| 351 | ///     let source = String::from_utf8_lossy(&content); | 
|---|
| 352 | ///     let s = simple_minify(&source);    // First we get the tokens list. | 
|---|
| 353 | ///     let s = s.apply(aggregate_strings) // This `apply` aggregates string litterals. | 
|---|
| 354 | ///              .apply(clean_tokens)      // This one is used to remove useless chars. | 
|---|
| 355 | ///              .to_string();             // And we finally convert to string. | 
|---|
| 356 | ///     println!( "result: {}", s); | 
|---|
| 357 | /// } | 
|---|
| 358 | /// ``` | 
|---|
| 359 | #[ inline] | 
|---|
| 360 | pub fn aggregate_strings(tokens: Tokens<'_>) -> Tokens<'_> { | 
|---|
| 361 | aggregate_strings_inner(tokens, separation_token:None) | 
|---|
| 362 | } | 
|---|
| 363 |  | 
|---|
| 364 | /// Exactly like `aggregate_strings` except this one expects a separation token | 
|---|
| 365 | /// to be passed. This token will be placed between the created variables for the | 
|---|
| 366 | /// strings aggregation and the rest. | 
|---|
| 367 | /// | 
|---|
| 368 | /// # Example | 
|---|
| 369 | /// | 
|---|
| 370 | /// Let's add a backline between the created variables and the rest of the code: | 
|---|
| 371 | /// | 
|---|
| 372 | /// ```rust,no_run | 
|---|
| 373 | /// extern crate minifier; | 
|---|
| 374 | /// use minifier::js::{ | 
|---|
| 375 | ///     aggregate_strings_with_separation, | 
|---|
| 376 | ///     clean_tokens, | 
|---|
| 377 | ///     simple_minify, | 
|---|
| 378 | ///     Token, | 
|---|
| 379 | ///     ReservedChar, | 
|---|
| 380 | /// }; | 
|---|
| 381 | /// use std::fs; | 
|---|
| 382 | /// | 
|---|
| 383 | /// fn main() { | 
|---|
| 384 | ///     let content = fs::read( "some_file.js").expect( "file not found"); | 
|---|
| 385 | ///     let source = String::from_utf8_lossy(&content); | 
|---|
| 386 | ///     let s = simple_minify(&source);    // First we get the tokens list. | 
|---|
| 387 | ///     let s = s.apply(|f| { | 
|---|
| 388 | ///                  aggregate_strings_with_separation(f, Token::Char(ReservedChar::Backline)) | 
|---|
| 389 | ///              })                   // We add a backline between the variable and the rest. | 
|---|
| 390 | ///              .apply(clean_tokens) // We clean the tokens. | 
|---|
| 391 | ///              .to_string();        // And we finally convert to string. | 
|---|
| 392 | ///     println!( "result: {}", s); | 
|---|
| 393 | /// } | 
|---|
| 394 | /// ``` | 
|---|
| 395 | #[ inline] | 
|---|
| 396 | pub fn aggregate_strings_with_separation<'a, 'b: 'a>( | 
|---|
| 397 | tokens: Tokens<'a>, | 
|---|
| 398 | separation_token: Token<'b>, | 
|---|
| 399 | ) -> Tokens<'a> { | 
|---|
| 400 | aggregate_strings_inner(tokens, separation_token:Some(separation_token)) | 
|---|
| 401 | } | 
|---|
| 402 |  | 
|---|
| 403 | fn aggregate_strings_into_array_inner<'a, 'b: 'a, T: Fn(&Tokens<'a>, usize) -> bool>( | 
|---|
| 404 | mut tokens: Tokens<'a>, | 
|---|
| 405 | array_name: &str, | 
|---|
| 406 | separation_token: Option<Token<'b>>, | 
|---|
| 407 | filter: T, | 
|---|
| 408 | ) -> Tokens<'a> { | 
|---|
| 409 | let mut to_insert = Vec::with_capacity(100); | 
|---|
| 410 | let mut to_replace = Vec::with_capacity(100); | 
|---|
| 411 |  | 
|---|
| 412 | { | 
|---|
| 413 | let mut to_ignore = HashSet::new(); | 
|---|
| 414 | // key: the token string | 
|---|
| 415 | // value: (position in the array, positions in the tokens list, need creation) | 
|---|
| 416 | let mut strs: HashMap<&str, (usize, Vec<usize>, bool)> = HashMap::with_capacity(1000); | 
|---|
| 417 | let (current_array_values, need_recreate, mut end_bracket) = | 
|---|
| 418 | match get_array(&tokens, array_name) { | 
|---|
| 419 | Some((s, p)) => (s, false, p), | 
|---|
| 420 | None => (Vec::new(), true, 0), | 
|---|
| 421 | }; | 
|---|
| 422 | let mut validated: HashSet<&str> = HashSet::new(); | 
|---|
| 423 |  | 
|---|
| 424 | let mut array_pos = 0; | 
|---|
| 425 | for s in current_array_values.iter() { | 
|---|
| 426 | if let Some(st) = tokens.0[*s].get_string() { | 
|---|
| 427 | strs.insert(&st[1..st.len() - 1], (array_pos, vec![], false)); | 
|---|
| 428 | array_pos += 1; | 
|---|
| 429 | validated.insert(&st[1..st.len() - 1]); | 
|---|
| 430 | to_ignore.insert(*s); | 
|---|
| 431 | } | 
|---|
| 432 | } | 
|---|
| 433 |  | 
|---|
| 434 | let mut array_pos_str = array_pos.to_string(); | 
|---|
| 435 | for pos in 0..tokens.len() { | 
|---|
| 436 | if to_ignore.contains(&pos) { | 
|---|
| 437 | continue; | 
|---|
| 438 | } | 
|---|
| 439 | let token = &tokens[pos]; | 
|---|
| 440 | if let Some(str_token) = token.get_string() { | 
|---|
| 441 | if !filter(&tokens, pos) { | 
|---|
| 442 | continue; | 
|---|
| 443 | } | 
|---|
| 444 | let s = &str_token[1..str_token.len() - 1]; | 
|---|
| 445 | let x = strs | 
|---|
| 446 | .entry(s) | 
|---|
| 447 | .or_insert_with(|| (0, Vec::with_capacity(1), true)); | 
|---|
| 448 | x.1.push(pos); | 
|---|
| 449 | if x.1.len() > 1 && !validated.contains(s) { | 
|---|
| 450 | let len = s.len(); | 
|---|
| 451 | if len * x.1.len() | 
|---|
| 452 | > (array_name.len() + array_pos_str.len() + 2) * x.1.len() | 
|---|
| 453 | + array_pos_str.len() | 
|---|
| 454 | + 2 | 
|---|
| 455 | { | 
|---|
| 456 | validated.insert(&str_token[1..str_token.len() - 1]); | 
|---|
| 457 | x.0 = array_pos; | 
|---|
| 458 | array_pos += 1; | 
|---|
| 459 | array_pos_str = array_pos.to_string(); | 
|---|
| 460 | } | 
|---|
| 461 | } | 
|---|
| 462 | } | 
|---|
| 463 | } | 
|---|
| 464 |  | 
|---|
| 465 | // TODO: | 
|---|
| 466 | // 1. Sort strings by length (the smallest should take the smallest numbers | 
|---|
| 467 | //    for bigger gains). | 
|---|
| 468 | // 2. Compute "score" for all strings of the same length and sort the strings | 
|---|
| 469 | //    of the same length with this score. | 
|---|
| 470 | // 3. Loop again over strings and remove those who shouldn't be there anymore. | 
|---|
| 471 | // 4. Repeat. | 
|---|
| 472 | // | 
|---|
| 473 | // ALTERNATIVE: | 
|---|
| 474 | // | 
|---|
| 475 | // Compute the score based on: | 
|---|
| 476 | // current number of digits * str length * str occurence | 
|---|
| 477 | // | 
|---|
| 478 | // ^ This second solution should bring even better results. | 
|---|
| 479 | // | 
|---|
| 480 | // ALSO: if an array with such strings already exists, it'd be worth it to recompute | 
|---|
| 481 | // everything again. | 
|---|
| 482 | let mut validated = validated.iter().map(|v| (strs[v].0, v)).collect::<Vec<_>>(); | 
|---|
| 483 | validated.sort_unstable_by(|(p1, _), (p2, _)| p2.cmp(p1)); | 
|---|
| 484 |  | 
|---|
| 485 | if need_recreate && !validated.is_empty() { | 
|---|
| 486 | if let Some(token) = separation_token { | 
|---|
| 487 | to_insert.push((0, token)); | 
|---|
| 488 | } | 
|---|
| 489 | to_insert.push((0, Token::Char(ReservedChar::SemiColon))); | 
|---|
| 490 | to_insert.push((0, Token::Char(ReservedChar::CloseBracket))); | 
|---|
| 491 | to_insert.push((0, Token::Char(ReservedChar::OpenBracket))); | 
|---|
| 492 | to_insert.push((0, Token::CreatedVarDecl(format!( "var {} =", array_name)))); | 
|---|
| 493 |  | 
|---|
| 494 | end_bracket = 2; | 
|---|
| 495 | } | 
|---|
| 496 |  | 
|---|
| 497 | let mut iter = validated.iter().peekable(); | 
|---|
| 498 | while let Some((array_pos, s)) = iter.next() { | 
|---|
| 499 | let (_, ref tokens_pos, create_array_entry) = strs[*s]; | 
|---|
| 500 | let array_index = Token::CreatedVar(format!( "{} [{} ]", array_name, array_pos)); | 
|---|
| 501 | for token in tokens_pos.iter() { | 
|---|
| 502 | to_replace.push((*token, array_index.clone())); | 
|---|
| 503 | } | 
|---|
| 504 | if !create_array_entry { | 
|---|
| 505 | continue; | 
|---|
| 506 | } | 
|---|
| 507 | to_insert.push((end_bracket, Token::CreatedVar(format!( "\"{}\" ", *s)))); | 
|---|
| 508 | if iter.peek().is_none() && current_array_values.is_empty() { | 
|---|
| 509 | continue; | 
|---|
| 510 | } | 
|---|
| 511 | to_insert.push((end_bracket, Token::Char(ReservedChar::Comma))); | 
|---|
| 512 | } | 
|---|
| 513 | } | 
|---|
| 514 | for (pos, rep) in to_replace.into_iter() { | 
|---|
| 515 | tokens.0[pos] = rep; | 
|---|
| 516 | } | 
|---|
| 517 | for (pos, rep) in to_insert.into_iter() { | 
|---|
| 518 | tokens.0.insert(pos, rep); | 
|---|
| 519 | } | 
|---|
| 520 | tokens | 
|---|
| 521 | } | 
|---|
| 522 |  | 
|---|
| 523 | /// Exactly like `aggregate_strings_into_array` except this one expects a separation token | 
|---|
| 524 | /// to be passed. This token will be placed between the created array for the | 
|---|
| 525 | /// strings aggregation and the rest. | 
|---|
| 526 | /// | 
|---|
| 527 | /// # Example | 
|---|
| 528 | /// | 
|---|
| 529 | /// Let's add a backline between the created variables and the rest of the code: | 
|---|
| 530 | /// | 
|---|
| 531 | /// ```rust,no_run | 
|---|
| 532 | /// extern crate minifier; | 
|---|
| 533 | /// use minifier::js::{ | 
|---|
| 534 | ///     aggregate_strings_into_array_with_separation, | 
|---|
| 535 | ///     clean_tokens, | 
|---|
| 536 | ///     simple_minify, | 
|---|
| 537 | ///     Token, | 
|---|
| 538 | ///     ReservedChar, | 
|---|
| 539 | /// }; | 
|---|
| 540 | /// use std::fs; | 
|---|
| 541 | /// | 
|---|
| 542 | /// fn main() { | 
|---|
| 543 | ///     let content = fs::read( "some_file.js").expect( "file not found"); | 
|---|
| 544 | ///     let source = String::from_utf8_lossy(&content); | 
|---|
| 545 | ///     let s = simple_minify(&source);    // First we get the tokens list. | 
|---|
| 546 | ///     let s = s.apply(|f| { | 
|---|
| 547 | ///                  aggregate_strings_into_array_with_separation(f, "R", Token::Char(ReservedChar::Backline)) | 
|---|
| 548 | ///              })                   // We add a backline between the variable and the rest. | 
|---|
| 549 | ///              .apply(clean_tokens) // We clean the tokens. | 
|---|
| 550 | ///              .to_string();        // And we finally convert to string. | 
|---|
| 551 | ///     println!( "result: {}", s); | 
|---|
| 552 | /// } | 
|---|
| 553 | /// ``` | 
|---|
| 554 | #[ inline] | 
|---|
| 555 | pub fn aggregate_strings_into_array_with_separation<'a, 'b: 'a>( | 
|---|
| 556 | tokens: Tokens<'a>, | 
|---|
| 557 | array_name: &str, | 
|---|
| 558 | separation_token: Token<'b>, | 
|---|
| 559 | ) -> Tokens<'a> { | 
|---|
| 560 | aggregate_strings_into_array_inner(tokens, array_name, separation_token:Some(separation_token), |_, _| true) | 
|---|
| 561 | } | 
|---|
| 562 |  | 
|---|
| 563 | /// Same as [`aggregate_strings_into_array_with_separation`] except it allows certain strings to | 
|---|
| 564 | /// not be aggregated thanks to the `filter` parameter. If it returns `false`, then the string will | 
|---|
| 565 | /// be ignored. | 
|---|
| 566 | #[ inline] | 
|---|
| 567 | pub fn aggregate_strings_into_array_with_separation_filter<'a, 'b: 'a, T>( | 
|---|
| 568 | tokens: Tokens<'a>, | 
|---|
| 569 | array_name: &str, | 
|---|
| 570 | separation_token: Token<'b>, | 
|---|
| 571 | filter: T, | 
|---|
| 572 | ) -> Tokens<'a> | 
|---|
| 573 | where | 
|---|
| 574 | T: Fn(&Tokens<'a>, usize) -> bool, | 
|---|
| 575 | { | 
|---|
| 576 | aggregate_strings_into_array_inner(tokens, array_name, separation_token:Some(separation_token), filter) | 
|---|
| 577 | } | 
|---|
| 578 |  | 
|---|
| 579 | /// Aggregate litteral strings. For instance, if the string litteral "Oh look over there!" | 
|---|
| 580 | /// appears more than once, it will be added to the generated array and used everywhere the | 
|---|
| 581 | /// string appears. Of course, this replacement is only performed when it allows to take | 
|---|
| 582 | /// less space. | 
|---|
| 583 | /// | 
|---|
| 584 | /// # Example | 
|---|
| 585 | /// | 
|---|
| 586 | /// ```rust,no_run | 
|---|
| 587 | /// extern crate minifier; | 
|---|
| 588 | /// use minifier::js::{aggregate_strings_into_array, clean_tokens, simple_minify}; | 
|---|
| 589 | /// use std::fs; | 
|---|
| 590 | /// | 
|---|
| 591 | /// fn main() { | 
|---|
| 592 | ///     let content = fs::read( "some_file.js").expect( "file not found"); | 
|---|
| 593 | ///     let source = String::from_utf8_lossy(&content); | 
|---|
| 594 | ///     let s = simple_minify(&source);    // First we get the tokens list. | 
|---|
| 595 | ///     let s = s.apply(|f| aggregate_strings_into_array(f, "R")) // This `apply` aggregates string litterals. | 
|---|
| 596 | ///              .apply(clean_tokens)      // This one is used to remove useless chars. | 
|---|
| 597 | ///              .to_string();             // And we finally convert to string. | 
|---|
| 598 | ///     println!( "result: {}", s); | 
|---|
| 599 | /// } | 
|---|
| 600 | /// ``` | 
|---|
| 601 | #[ inline] | 
|---|
| 602 | pub fn aggregate_strings_into_array<'a>(tokens: Tokens<'a>, array_name: &str) -> Tokens<'a> { | 
|---|
| 603 | aggregate_strings_into_array_inner(tokens, array_name, separation_token:None, |_, _| true) | 
|---|
| 604 | } | 
|---|
| 605 |  | 
|---|
| 606 | /// Same as [`aggregate_strings_into_array`] except it allows certain strings to not be aggregated | 
|---|
| 607 | /// thanks to the `filter` parameter. If it returns `false`, then the string will be ignored. | 
|---|
| 608 | #[ inline] | 
|---|
| 609 | pub fn aggregate_strings_into_array_filter<'a, T>( | 
|---|
| 610 | tokens: Tokens<'a>, | 
|---|
| 611 | array_name: &str, | 
|---|
| 612 | filter: T, | 
|---|
| 613 | ) -> Tokens<'a> | 
|---|
| 614 | where | 
|---|
| 615 | T: Fn(&Tokens<'a>, usize) -> bool, | 
|---|
| 616 | { | 
|---|
| 617 | aggregate_strings_into_array_inner(tokens, array_name, separation_token:None, filter) | 
|---|
| 618 | } | 
|---|
| 619 |  | 
|---|
| 620 | /// Simple function to get the untouched token list. Useful in case you want to perform some | 
|---|
| 621 | /// actions directly on it. | 
|---|
| 622 | /// | 
|---|
| 623 | /// # Example | 
|---|
| 624 | /// | 
|---|
| 625 | /// ```rust,no_run | 
|---|
| 626 | /// extern crate minifier; | 
|---|
| 627 | /// use minifier::js::simple_minify; | 
|---|
| 628 | /// use std::fs; | 
|---|
| 629 | /// | 
|---|
| 630 | /// fn main() { | 
|---|
| 631 | ///     let content = fs::read( "some_file.js").expect( "file not found"); | 
|---|
| 632 | ///     let source = String::from_utf8_lossy(&content); | 
|---|
| 633 | ///     let s = simple_minify(&source); | 
|---|
| 634 | ///     println!( "result: {:?}", s); // We now have the tokens list. | 
|---|
| 635 | /// } | 
|---|
| 636 | /// ``` | 
|---|
| 637 | #[ inline] | 
|---|
| 638 | pub fn simple_minify(source: &str) -> Tokens<'_> { | 
|---|
| 639 | token::tokenize(source) | 
|---|
| 640 | } | 
|---|
| 641 |  | 
|---|
| 642 | #[ test] | 
|---|
| 643 | fn aggregate_strings_in_array() { | 
|---|
| 644 | let source = r#"var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 645 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 646 | let expected_result = "var R=[\" a nice string\" ,\" cake!\" ];var x=[R[0],R[0],\ | 
|---|
| 647 |                            \" another nice string\" ,R[1],R[1],R[0],R[1],R[1],R[1]]"; | 
|---|
| 648 |  | 
|---|
| 649 | let result = simple_minify(source) | 
|---|
| 650 | .apply(crate::js::clean_tokens) | 
|---|
| 651 | .apply(|c| aggregate_strings_into_array(c, "R")) | 
|---|
| 652 | .to_string(); | 
|---|
| 653 | assert_eq!(result, expected_result); | 
|---|
| 654 |  | 
|---|
| 655 | let source = r#"var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 656 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 657 | let expected_result = "var R=[\" a nice string\" ,\" cake!\" ];\n var x=[R[0],R[0],\ | 
|---|
| 658 |                            \" another nice string\" ,R[1],R[1],R[0],R[1],R[1],R[1]]"; | 
|---|
| 659 |  | 
|---|
| 660 | let result = simple_minify(source) | 
|---|
| 661 | .apply(crate::js::clean_tokens) | 
|---|
| 662 | .apply(|c| { | 
|---|
| 663 | aggregate_strings_into_array_with_separation( | 
|---|
| 664 | c, | 
|---|
| 665 | "R", | 
|---|
| 666 | Token::Char(ReservedChar::Backline), | 
|---|
| 667 | ) | 
|---|
| 668 | }) | 
|---|
| 669 | .to_string(); | 
|---|
| 670 | assert_eq!(result, expected_result); | 
|---|
| 671 |  | 
|---|
| 672 | let source = r#"var x = ["a nice string", "a nice string", "another nice string", "another nice string", "another nice string", "another nice string","cake!","cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 673 | let expected_result = "var R=[\" a nice string\" ,\" another nice string\" ,\" cake!\" ];\n \ | 
|---|
| 674 |                            var x=[R[0],R[0],R[1],R[1],R[1],R[1],R[2],R[2],R[0],R[2],\ | 
|---|
| 675 |                            R[2],R[2]]"; | 
|---|
| 676 |  | 
|---|
| 677 | let result = simple_minify(source) | 
|---|
| 678 | .apply(crate::js::clean_tokens) | 
|---|
| 679 | .apply(|c| { | 
|---|
| 680 | aggregate_strings_into_array_with_separation( | 
|---|
| 681 | c, | 
|---|
| 682 | "R", | 
|---|
| 683 | Token::Char(ReservedChar::Backline), | 
|---|
| 684 | ) | 
|---|
| 685 | }) | 
|---|
| 686 | .to_string(); | 
|---|
| 687 | assert_eq!(result, expected_result); | 
|---|
| 688 | } | 
|---|
| 689 |  | 
|---|
| 690 | #[ test] | 
|---|
| 691 | fn aggregate_strings_in_array_filter() { | 
|---|
| 692 | let source = r#"var searchIndex = {};searchIndex['duplicate_paths'] = {'aaaaaaaa': 'bbbbbbbb', 'bbbbbbbb': 'aaaaaaaa', 'duplicate_paths': 'aaaaaaaa'};"#; | 
|---|
| 693 | let expected_result = "var R=[\" bbbbbbbb\" ,\" aaaaaaaa\" ];\n var searchIndex={};searchIndex['duplicate_paths']={R[1]:R[0],R[0]:R[1],'duplicate_paths':R[1]}"; | 
|---|
| 694 |  | 
|---|
| 695 | let result = simple_minify(source) | 
|---|
| 696 | .apply(crate::js::clean_tokens) | 
|---|
| 697 | .apply(|c| { | 
|---|
| 698 | aggregate_strings_into_array_with_separation_filter( | 
|---|
| 699 | c, | 
|---|
| 700 | "R", | 
|---|
| 701 | Token::Char(ReservedChar::Backline), | 
|---|
| 702 | |tokens, pos| { | 
|---|
| 703 | pos < 2 | 
|---|
| 704 | || !tokens[pos - 1].eq_char(ReservedChar::OpenBracket) | 
|---|
| 705 | || tokens[pos - 2].get_other() != Some( "searchIndex") | 
|---|
| 706 | }, | 
|---|
| 707 | ) | 
|---|
| 708 | }) | 
|---|
| 709 | .to_string(); | 
|---|
| 710 | assert_eq!(result, expected_result); | 
|---|
| 711 |  | 
|---|
| 712 | let source = r#"var searchIndex = {};searchIndex['duplicate_paths'] = {'aaaaaaaa': 'bbbbbbbb', 'bbbbbbbb': 'aaaaaaaa', 'duplicate_paths': 'aaaaaaaa', 'x': 'duplicate_paths'};"#; | 
|---|
| 713 | let expected_result = "var R=[\" bbbbbbbb\" ,\" aaaaaaaa\" ,\" duplicate_paths\" ];\n var searchIndex={};searchIndex['duplicate_paths']={R[1]:R[0],R[0]:R[1],R[2]:R[1],'x':R[2]}"; | 
|---|
| 714 |  | 
|---|
| 715 | let result = simple_minify(source) | 
|---|
| 716 | .apply(crate::js::clean_tokens) | 
|---|
| 717 | .apply(|c| { | 
|---|
| 718 | aggregate_strings_into_array_with_separation_filter( | 
|---|
| 719 | c, | 
|---|
| 720 | "R", | 
|---|
| 721 | Token::Char(ReservedChar::Backline), | 
|---|
| 722 | |tokens, pos| { | 
|---|
| 723 | pos < 2 | 
|---|
| 724 | || !tokens[pos - 1].eq_char(ReservedChar::OpenBracket) | 
|---|
| 725 | || tokens[pos - 2].get_other() != Some( "searchIndex") | 
|---|
| 726 | }, | 
|---|
| 727 | ) | 
|---|
| 728 | }) | 
|---|
| 729 | .to_string(); | 
|---|
| 730 | assert_eq!(result, expected_result); | 
|---|
| 731 | } | 
|---|
| 732 |  | 
|---|
| 733 | #[ test] | 
|---|
| 734 | fn aggregate_strings_in_array_existing() { | 
|---|
| 735 | let source = r#"var R=[];var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 736 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 737 | let expected_result = "var R=[\" a nice string\" ,\" cake!\" ];var x=[R[0],R[0],\ | 
|---|
| 738 |                            \" another nice string\" ,R[1],R[1],R[0],R[1],R[1],R[1]]"; | 
|---|
| 739 |  | 
|---|
| 740 | let result = simple_minify(source) | 
|---|
| 741 | .apply(crate::js::clean_tokens) | 
|---|
| 742 | .apply(|c| aggregate_strings_into_array(c, "R")) | 
|---|
| 743 | .to_string(); | 
|---|
| 744 | assert_eq!(result, expected_result); | 
|---|
| 745 |  | 
|---|
| 746 | let source = r#"var R=["a nice string"];var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 747 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 748 | let expected_result = "var R=[\" a nice string\" ,\" cake!\" ];var x=[R[0],R[0],\ | 
|---|
| 749 |                            \" another nice string\" ,R[1],R[1],R[0],R[1],R[1],R[1]]"; | 
|---|
| 750 |  | 
|---|
| 751 | let result = simple_minify(source) | 
|---|
| 752 | .apply(crate::js::clean_tokens) | 
|---|
| 753 | .apply(|c| aggregate_strings_into_array(c, "R")) | 
|---|
| 754 | .to_string(); | 
|---|
| 755 | assert_eq!(result, expected_result); | 
|---|
| 756 |  | 
|---|
| 757 | let source = r#"var y = 12;var R=["a nice string"];var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 758 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 759 | let expected_result = "var y=12;var R=[\" a nice string\" ,\" cake!\" ];var x=[R[0],R[0],\ | 
|---|
| 760 |                            \" another nice string\" ,R[1],R[1],R[0],R[1],R[1],R[1]]"; | 
|---|
| 761 |  | 
|---|
| 762 | let result = simple_minify(source) | 
|---|
| 763 | .apply(crate::js::clean_tokens) | 
|---|
| 764 | .apply(|c| aggregate_strings_into_array(c, "R")) | 
|---|
| 765 | .to_string(); | 
|---|
| 766 | assert_eq!(result, expected_result); | 
|---|
| 767 |  | 
|---|
| 768 | let source = r#"var R=["osef1", "o2", "damn"]; | 
|---|
| 769 |                     var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 770 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 771 | let expected_result = "var R=[\" osef1\" ,\" o2\" ,\" damn\" ,\" a nice string\" ,\" cake!\" ];\ | 
|---|
| 772 |                            var x=[R[3],R[3],\" another nice string\" ,R[4],R[4],R[3],R[4],R[4],R[4]]"; | 
|---|
| 773 |  | 
|---|
| 774 | let result = simple_minify(source) | 
|---|
| 775 | .apply(crate::js::clean_tokens) | 
|---|
| 776 | .apply(|c| aggregate_strings_into_array(c, "R")) | 
|---|
| 777 | .to_string(); | 
|---|
| 778 | assert_eq!(result, expected_result); | 
|---|
| 779 | } | 
|---|
| 780 |  | 
|---|
| 781 | #[ test] | 
|---|
| 782 | fn string_duplicates() { | 
|---|
| 783 | let source = r#"var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 784 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 785 | let expected_result = "var r_aa=\" a nice string\" ,r_ba=\" cake!\" ;var x=[r_aa,r_aa,\ | 
|---|
| 786 |                            \" another nice string\" ,r_ba,r_ba,r_aa,r_ba,r_ba,r_ba]"; | 
|---|
| 787 |  | 
|---|
| 788 | let result = simple_minify(source) | 
|---|
| 789 | .apply(aggregate_strings) | 
|---|
| 790 | .apply(crate::js::clean_tokens) | 
|---|
| 791 | .to_string(); | 
|---|
| 792 | assert_eq!(result, expected_result); | 
|---|
| 793 | } | 
|---|
| 794 |  | 
|---|
| 795 | #[ test] | 
|---|
| 796 | fn already_existing_var() { | 
|---|
| 797 | let source = r#"var r_aa = "a nice string"; var x = ["a nice string", "a nice string", | 
|---|
| 798 |                     "another nice string", "cake!", | 
|---|
| 799 |                     "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 800 | let expected_result = "var r_ba=\" cake!\" ;var r_aa=\" a nice string\" ;var x=[r_aa,r_aa,\ | 
|---|
| 801 |                            \" another nice string\" ,r_ba,r_ba,r_aa,r_ba,r_ba,r_ba]"; | 
|---|
| 802 |  | 
|---|
| 803 | let result = simple_minify(source) | 
|---|
| 804 | .apply(aggregate_strings) | 
|---|
| 805 | .apply(crate::js::clean_tokens) | 
|---|
| 806 | .to_string(); | 
|---|
| 807 | assert_eq!(result, expected_result); | 
|---|
| 808 | } | 
|---|
| 809 |  | 
|---|
| 810 | #[ test] | 
|---|
| 811 | fn string_duplicates_variables_already_exist() { | 
|---|
| 812 | let source = r#"var r_aa=1;var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 813 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 814 | let expected_result = "var r_ba=\" a nice string\" ,r_ca=\" cake!\" ;\ | 
|---|
| 815 |                            var r_aa=1;var x=[r_ba,r_ba,\ | 
|---|
| 816 |                            \" another nice string\" ,r_ca,r_ca,r_ba,r_ca,r_ca,r_ca]"; | 
|---|
| 817 |  | 
|---|
| 818 | let result = simple_minify(source) | 
|---|
| 819 | .apply(aggregate_strings) | 
|---|
| 820 | .apply(crate::js::clean_tokens) | 
|---|
| 821 | .to_string(); | 
|---|
| 822 | assert_eq!(result, expected_result); | 
|---|
| 823 | } | 
|---|
| 824 |  | 
|---|
| 825 | #[ test] | 
|---|
| 826 | fn string_duplicates_with_separator() { | 
|---|
| 827 | use self::token::ReservedChar; | 
|---|
| 828 |  | 
|---|
| 829 | let source = r#"var x = ["a nice string", "a nice string", "another nice string", "cake!", | 
|---|
| 830 |                              "cake!", "a nice string", "cake!", "cake!", "cake!"];"#; | 
|---|
| 831 | let expected_result = "var r_aa=\" a nice string\" ,r_ba=\" cake!\" ;\n var x=[r_aa,r_aa,\ | 
|---|
| 832 |                            \" another nice string\" ,r_ba,r_ba,r_aa,r_ba,r_ba,r_ba]"; | 
|---|
| 833 | let result = simple_minify(source) | 
|---|
| 834 | .apply(crate::js::clean_tokens) | 
|---|
| 835 | .apply(|f| aggregate_strings_with_separation(f, Token::Char(ReservedChar::Backline))) | 
|---|
| 836 | .to_string(); | 
|---|
| 837 | assert_eq!(result, expected_result); | 
|---|
| 838 | } | 
|---|
| 839 |  | 
|---|
| 840 | #[ test] | 
|---|
| 841 | fn clean_except() { | 
|---|
| 842 | use self::token::ReservedChar; | 
|---|
| 843 |  | 
|---|
| 844 | let source = r#"var x = [1, 2, 3]; | 
|---|
| 845 | var y = "salut"; | 
|---|
| 846 | var z = "ok!";"#; | 
|---|
| 847 | let expected = r#"var x=[1,2,3]; | 
|---|
| 848 | var y="salut"; | 
|---|
| 849 | var z="ok!""#; | 
|---|
| 850 |  | 
|---|
| 851 | let result = simple_minify(source) | 
|---|
| 852 | .apply(|f| { | 
|---|
| 853 | crate::js::clean_tokens_except(f, |c| c.get_char() != Some(ReservedChar::Backline)) | 
|---|
| 854 | }) | 
|---|
| 855 | .to_string(); | 
|---|
| 856 | assert_eq!(result, expected); | 
|---|
| 857 | } | 
|---|
| 858 |  | 
|---|
| 859 | #[ test] | 
|---|
| 860 | fn clean_except2() { | 
|---|
| 861 | use self::token::ReservedChar; | 
|---|
| 862 |  | 
|---|
| 863 | let source = "let x = [ 1, 2, \t 3];"; | 
|---|
| 864 | let expected = "let x = [ 1, 2, 3];"; | 
|---|
| 865 |  | 
|---|
| 866 | let result = simple_minify(source) | 
|---|
| 867 | .apply(|f| { | 
|---|
| 868 | crate::js::clean_tokens_except(f, |c| { | 
|---|
| 869 | c.get_char() != Some(ReservedChar::Space) | 
|---|
| 870 | && c.get_char() != Some(ReservedChar::SemiColon) | 
|---|
| 871 | }) | 
|---|
| 872 | }) | 
|---|
| 873 | .to_string(); | 
|---|
| 874 | assert_eq!(result, expected); | 
|---|
| 875 | } | 
|---|
| 876 |  | 
|---|
| 877 | #[ test] | 
|---|
| 878 | fn clean_except3() { | 
|---|
| 879 | use self::token::ReservedChar; | 
|---|
| 880 |  | 
|---|
| 881 | let source = "let x = [ 1, 2, \t 3];"; | 
|---|
| 882 | let expected = "let x=[1,2,\t 3];"; | 
|---|
| 883 |  | 
|---|
| 884 | let result = simple_minify(source) | 
|---|
| 885 | .apply(|f| { | 
|---|
| 886 | crate::js::clean_tokens_except(f, |c| { | 
|---|
| 887 | c.get_char() != Some(ReservedChar::Tab) | 
|---|
| 888 | && c.get_char() != Some(ReservedChar::SemiColon) | 
|---|
| 889 | }) | 
|---|
| 890 | }) | 
|---|
| 891 | .to_string(); | 
|---|
| 892 | assert_eq!(result, expected); | 
|---|
| 893 | } | 
|---|
| 894 |  | 
|---|
| 895 | #[ test] | 
|---|
| 896 | fn name_generator() { | 
|---|
| 897 | let s = "a".repeat(36); | 
|---|
| 898 | // We need to generate enough long strings to reach the point that the name generator | 
|---|
| 899 | // generates names with 3 characters. | 
|---|
| 900 | let s = std::iter::repeat(s) | 
|---|
| 901 | .take(20000) | 
|---|
| 902 | .enumerate() | 
|---|
| 903 | .map(|(pos, s)| format!( "{}{}", s, pos)) | 
|---|
| 904 | .collect::<Vec<_>>(); | 
|---|
| 905 | let source = format!( | 
|---|
| 906 | "var x = [{}];", | 
|---|
| 907 | s.iter() | 
|---|
| 908 | .map(|s| format!( "\" {0}\" ,\" {0}\" ", s)) | 
|---|
| 909 | .collect::<Vec<_>>() | 
|---|
| 910 | .join( ",") | 
|---|
| 911 | ); | 
|---|
| 912 | let result = simple_minify(&source) | 
|---|
| 913 | .apply(crate::js::clean_tokens) | 
|---|
| 914 | .apply(aggregate_strings) | 
|---|
| 915 | .to_string(); | 
|---|
| 916 | assert!(result.contains( ",r_aaa=")); | 
|---|
| 917 | assert!(result.find( ",r_ab=").unwrap() < result.find( ",r_ba=").unwrap()); | 
|---|
| 918 | } | 
|---|
| 919 |  | 
|---|
| 920 | #[ test] | 
|---|
| 921 | fn simple_quote() { | 
|---|
| 922 | let source = r#"var x = "\\";"#; | 
|---|
| 923 | let expected_result = r#"var x="\\""#; | 
|---|
| 924 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 925 | } | 
|---|
| 926 |  | 
|---|
| 927 | #[ test] | 
|---|
| 928 | fn js_minify_test() { | 
|---|
| 929 | let source = r##" | 
|---|
| 930 | var foo = "something"; | 
|---|
| 931 |  | 
|---|
| 932 | var another_var = 2348323; | 
|---|
| 933 |  | 
|---|
| 934 | // who doesn't like comments? | 
|---|
| 935 | /* and even longer comments? | 
|---|
| 936 |  | 
|---|
| 937 | like | 
|---|
| 938 | on | 
|---|
| 939 | a | 
|---|
| 940 | lot | 
|---|
| 941 | of | 
|---|
| 942 | lines! | 
|---|
| 943 |  | 
|---|
| 944 | Fun! | 
|---|
| 945 | */ | 
|---|
| 946 | function far_away(x, y) { | 
|---|
| 947 |     var x2 = x + 4; | 
|---|
| 948 |     return x * x2 + y; | 
|---|
| 949 | } | 
|---|
| 950 |  | 
|---|
| 951 | // this call is useless | 
|---|
| 952 | far_away(another_var, 12); | 
|---|
| 953 | // this call is useless too | 
|---|
| 954 | far_away(another_var, 12); | 
|---|
| 955 | "##; | 
|---|
| 956 |  | 
|---|
| 957 | let expected_result = "var foo=\" something\" ;var another_var=2348323;function far_away(x,y){\ | 
|---|
| 958 |                            var x2=x+4;return x*x2+y}far_away(another_var,12);far_away(another_var,\ | 
|---|
| 959 |                            12)"; | 
|---|
| 960 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 961 | } | 
|---|
| 962 |  | 
|---|
| 963 | #[ test] | 
|---|
| 964 | fn another_js_test() { | 
|---|
| 965 | let source = r#" | 
|---|
| 966 | /*! let's keep this license | 
|---|
| 967 |  * | 
|---|
| 968 |  * because everyone likes licenses! | 
|---|
| 969 |  * | 
|---|
| 970 |  * right? | 
|---|
| 971 |  */ | 
|---|
| 972 |  | 
|---|
| 973 | function forEach(data, func) { | 
|---|
| 974 |     for (var i = 0; i < data.length; ++i) { | 
|---|
| 975 |         func(data[i]); | 
|---|
| 976 |     } | 
|---|
| 977 | } | 
|---|
| 978 |  | 
|---|
| 979 | forEach([0, 1, 2, 3, 4, | 
|---|
| 980 |          5, 6, 7, 8, 9], function (x) { | 
|---|
| 981 |             console.log(x); | 
|---|
| 982 |          }); | 
|---|
| 983 | // I think we're done? | 
|---|
| 984 | console.log('done!'); | 
|---|
| 985 | "#; | 
|---|
| 986 |  | 
|---|
| 987 | let expected_result = r#"/*! let's keep this license | 
|---|
| 988 |  * | 
|---|
| 989 |  * because everyone likes licenses! | 
|---|
| 990 |  * | 
|---|
| 991 |  * right? | 
|---|
| 992 |  */function forEach(data,func){for(var i=0;i<data.length;++i){func(data[i])}}forEach([0,1,2,3,4,5,6,7,8,9],function(x){console.log(x)});console.log('done!')"#; | 
|---|
| 993 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 994 | } | 
|---|
| 995 |  | 
|---|
| 996 | #[ test] | 
|---|
| 997 | fn comment_issue() { | 
|---|
| 998 | let source = r#" | 
|---|
| 999 | search_input.onchange = function(e) { | 
|---|
| 1000 |     // Do NOT e.preventDefault() here. It will prevent pasting. | 
|---|
| 1001 |     clearTimeout(searchTimeout); | 
|---|
| 1002 |     // zero-timeout necessary here because at the time of event handler execution the | 
|---|
| 1003 |     // pasted content is not in the input field yet. Shouldn’t make any difference for | 
|---|
| 1004 |     // change, though. | 
|---|
| 1005 |     setTimeout(search, 0); | 
|---|
| 1006 | }; | 
|---|
| 1007 | "#; | 
|---|
| 1008 | let expected_result = "search_input.onchange=function(e){clearTimeout(searchTimeout);\ | 
|---|
| 1009 |                            setTimeout(search,0)}"; | 
|---|
| 1010 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1011 | } | 
|---|
| 1012 |  | 
|---|
| 1013 | #[ test] | 
|---|
| 1014 | fn missing_whitespace() { | 
|---|
| 1015 | let source = r#" | 
|---|
| 1016 | for (var entry in results) { | 
|---|
| 1017 |     if (results.hasOwnProperty(entry)) { | 
|---|
| 1018 |         ar.push(results[entry]); | 
|---|
| 1019 |     } | 
|---|
| 1020 | }"#; | 
|---|
| 1021 | let expected_result = "for(var entry in results){if(results.hasOwnProperty(entry)){\ | 
|---|
| 1022 |                            ar.push(results[entry])}}"; | 
|---|
| 1023 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1024 | } | 
|---|
| 1025 |  | 
|---|
| 1026 | #[ test] | 
|---|
| 1027 | fn weird_regex_issue() { | 
|---|
| 1028 | let source = r#" | 
|---|
| 1029 | val = val.replace(/\_/g, ""); | 
|---|
| 1030 |  | 
|---|
| 1031 | var valGenerics = extractGenerics(val);"#; | 
|---|
| 1032 | let expected_result = "val=val.replace(/\\ _/g,\"\" );var valGenerics=extractGenerics(val)"; | 
|---|
| 1033 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1034 | } | 
|---|
| 1035 |  | 
|---|
| 1036 | #[ test] | 
|---|
| 1037 | fn keep_space() { | 
|---|
| 1038 | fn inner_double_checks(source: &str, expected: &str) { | 
|---|
| 1039 | assert_eq!(minify(source).to_string(), expected); | 
|---|
| 1040 | let s = minify(source); | 
|---|
| 1041 | let mut out: Vec<u8> = Vec::new(); | 
|---|
| 1042 | s.write(&mut out).unwrap(); | 
|---|
| 1043 | assert_eq!(String::from_utf8(out).unwrap(), expected); | 
|---|
| 1044 | } | 
|---|
| 1045 |  | 
|---|
| 1046 | inner_double_checks( "return 12;return x;", "return 12;return x"); | 
|---|
| 1047 | inner_double_checks( "t in e", "t in e"); | 
|---|
| 1048 | inner_double_checks( "t + 1 in e", "t+1 in e"); | 
|---|
| 1049 | inner_double_checks( "t - 1 in e", "t-1 in e"); | 
|---|
| 1050 | inner_double_checks( "'a' in e", "'a'in e"); | 
|---|
| 1051 | inner_double_checks( "/a/g in e", "/a/g in e"); | 
|---|
| 1052 | inner_double_checks( "/a/i in e", "/a/i in e"); | 
|---|
| 1053 |  | 
|---|
| 1054 | inner_double_checks( "t instanceof e", "t instanceof e"); | 
|---|
| 1055 | inner_double_checks( "t + 1 instanceof e", "t+1 instanceof e"); | 
|---|
| 1056 | inner_double_checks( "t - 1 instanceof e", "t-1 instanceof e"); | 
|---|
| 1057 | inner_double_checks( "'a' instanceof e", "'a'instanceof e"); | 
|---|
| 1058 | inner_double_checks( "/a/g instanceof e", "/a/g instanceof e"); | 
|---|
| 1059 | inner_double_checks( "/a/i instanceof e", "/a/i instanceof e"); | 
|---|
| 1060 |  | 
|---|
| 1061 | inner_double_checks( "function foo() { let x = 12; }", "function foo(){let x=12}"); | 
|---|
| 1062 | inner_double_checks( | 
|---|
| 1063 | r#""use strict"; | 
|---|
| 1064 |  | 
|---|
| 1065 | (function() { | 
|---|
| 1066 |     const itemTypes = [ | 
|---|
| 1067 |         "mod", | 
|---|
| 1068 |         "externcrate", | 
|---|
| 1069 |         "import", | 
|---|
| 1070 |         "struct", | 
|---|
| 1071 |     ]; | 
|---|
| 1072 |     const TY_PRIMITIVE = itemTypes; | 
|---|
| 1073 |     function hasOwnPropertyRustdoc() {} | 
|---|
| 1074 | })();"#, | 
|---|
| 1075 | "\" use strict\" ;(function(){const itemTypes=[\" mod\" ,\" externcrate\" ,\" import\" ,\" struct\" \ | 
|---|
| 1076 |          ,];const TY_PRIMITIVE=itemTypes;function hasOwnPropertyRustdoc(){}})()", | 
|---|
| 1077 | ); | 
|---|
| 1078 | } | 
|---|
| 1079 |  | 
|---|
| 1080 | #[ test] | 
|---|
| 1081 | fn test_remove_extra_whitespace_before_typeof() { | 
|---|
| 1082 | let source = "var x = typeof 'foo';var y = typeof x;case typeof 'foo': 'bla'"; | 
|---|
| 1083 |  | 
|---|
| 1084 | let expected_result = "var x=typeof'foo';var y=typeof x;case typeof'foo':'bla'"; | 
|---|
| 1085 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1086 | } | 
|---|
| 1087 |  | 
|---|
| 1088 | #[ test] | 
|---|
| 1089 | fn test_remove_extra_whitespace_before_in() { | 
|---|
| 1090 | let source = r#"if ("key" in ev && typeof ev) { return true; } | 
|---|
| 1091 | if (x in ev && typeof ev) { return true; } | 
|---|
| 1092 | if (true in ev) { return true; }"#; | 
|---|
| 1093 |  | 
|---|
| 1094 | let expected_result = r#"if("key"in ev&&typeof ev){return true}if(x in ev&&typeof ev){return true}if(true in ev){return true}"#; | 
|---|
| 1095 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1096 | } | 
|---|
| 1097 |  | 
|---|
| 1098 | #[ test] | 
|---|
| 1099 | fn test_remove_extra_whitespace_before_operator() { | 
|---|
| 1100 | let source = "( x ) / 2; x / y;x /= y"; | 
|---|
| 1101 |  | 
|---|
| 1102 | let expected_result = "(x)/2;x/y;x/=y"; | 
|---|
| 1103 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1104 | } | 
|---|
| 1105 |  | 
|---|
| 1106 | #[ test] | 
|---|
| 1107 | fn check_regex_syntax() { | 
|---|
| 1108 | let source = "console.log(/MSIE|Trident|Edge/.test(window.navigator.userAgent));"; | 
|---|
| 1109 | let expected = "console.log(/MSIE|Trident|Edge/.test(window.navigator.userAgent))"; | 
|---|
| 1110 | assert_eq!(minify(source).to_string(), expected); | 
|---|
| 1111 | } | 
|---|
| 1112 |  | 
|---|
| 1113 | #[ test] | 
|---|
| 1114 | fn minify_minified() { | 
|---|
| 1115 | let source = "function (i, n, a) { i[n].type.replace(/ *;(.|\\ s)*/,\"\" )===t&&a.push(i[n].MathJax.elementJax);return a}"; | 
|---|
| 1116 | let expected = "function(i,n,a){i[n].type.replace(/ *;(.|\\ s)*/,\"\" )===t&&a.push(i[n].MathJax.elementJax);return a}"; | 
|---|
| 1117 | assert_eq!(minify(source).to_string(), expected); | 
|---|
| 1118 | } | 
|---|
| 1119 |  | 
|---|
| 1120 | #[ test] | 
|---|
| 1121 | fn check_string() { | 
|---|
| 1122 | let source = r###" | 
|---|
| 1123 |         const a = 123; | 
|---|
| 1124 |         const b = "123"; | 
|---|
| 1125 |         const c = `the number is ${a}  <-- note the spaces here`; | 
|---|
| 1126 |         const d = `      ${a}         ${b}      `; | 
|---|
| 1127 |     "###; | 
|---|
| 1128 | let expected = "const a=123;const b=\" 123\" ;const c=`the number is ${a}  <-- note the spaces \ | 
|---|
| 1129 |     here`;const d=`      ${a}         ${b}      `"; | 
|---|
| 1130 | assert_eq!(minify(source).to_string(), expected); | 
|---|
| 1131 | } | 
|---|
| 1132 |  | 
|---|
| 1133 | // TODO: requires AST to fix this issue! | 
|---|
| 1134 | /*#[test] | 
|---|
| 1135 | fn no_semi_colon() { | 
|---|
| 1136 | let source = r#" | 
|---|
| 1137 | console.log(1) | 
|---|
| 1138 | console.log(2) | 
|---|
| 1139 | var x = 12; | 
|---|
| 1140 | "#; | 
|---|
| 1141 | let expected_result = r#"console.log(1);console.log(2);var x=12;"#; | 
|---|
| 1142 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1143 | }*/ | 
|---|
| 1144 |  | 
|---|
| 1145 | // TODO: requires AST to fix this issue! | 
|---|
| 1146 | /*#[test] | 
|---|
| 1147 | fn correct_replace_for_backline() { | 
|---|
| 1148 | let source = r#" | 
|---|
| 1149 | function foo() { | 
|---|
| 1150 | return | 
|---|
| 1151 | 12; | 
|---|
| 1152 | } | 
|---|
| 1153 | "#; | 
|---|
| 1154 | let expected_result = r#"function foo(){return 12;}"#; | 
|---|
| 1155 | assert_eq!(minify(source).to_string(), expected_result); | 
|---|
| 1156 | }*/ | 
|---|
| 1157 |  | 
|---|