Website Summary Files Commits Branches @commit @diff @stats
path: root / homework / 81030 / 1,2-rust2html
diff options
context:
space:
mode:
Commit5729bc0ea96fc0b944c243230910c537d692bcad (patch)
Author
 Radoslav Georgiev <rgeorgiev583@gmail.com>
2018-11-17 19:11:03 +0200
Committer
 dtrendafilov <trendafilov.dn@gmail.com>
2018-11-17 19:11:02 +0200
Files43db75a7acf93c6c2c8e07c3a87590fe14fb4e4f /homework/81030/1,2-rust2html
IPL-5729bc0ea96fc0b944c243230910c537d692bcad.tar
IPL-5729bc0ea96fc0b944c243230910c537d692bcad.tar.gz
IPL-5729bc0ea96fc0b944c243230910c537d692bcad.zip
Parent9420791a9335fc174627bea5f2a25f434bbe7ae5 (diff)
implement the second homework assignment (#39)
* Rust syntax highlighter: fix some issues with the implementation of the first homework assignment * Rust syntax highlighter: pretty-printer: implement the second homework assignment
diff --git a/homework/81030/1,2-rust2html/.gitignore b/homework/81030/1,2-rust2html/.gitignore
new file mode 100644
index 0000000..691e40c
--- /dev/null
+++ b/homework/81030/1,2-rust2html/.gitignore
@@ -0,0 +1,5 @@
+a.out
+lex.yy.c
+
+/rust2html
+/example.html
diff --git a/homework/81030/1,2-rust2html/Makefile b/homework/81030/1,2-rust2html/Makefile
new file mode 100644
index 0000000..b6f2dde
--- /dev/null
+++ b/homework/81030/1,2-rust2html/Makefile
@@ -0,0 +1,21 @@
+PROGRAM_NAMES := rust2html rust2html-prettyprint
+PROGRAM_SOURCE_FILENAMES := $(patsubst %,%.lex.yy.c,PROGRAM_NAMES)
+
+LEX := flex
+
+ifdef DEBUG
+LEX_FLAGS := -d
+endif
+
+.PHONY: all clean
+
+all: $(PROGRAM_NAMES)
+
+%: %.lex.yy.c
+ $(CC) $(CFLAGS) -o $@ $<
+
+%.lex.yy.c: %.l
+ $(LEX) $(LEX_FLAGS) -o $@ $<
+
+clean:
+ $(RM) $(PROGRAM_NAMES) $(PROGRAM_SOURCE_FILENAMES) lex.yy.c example.html
diff --git a/homework/81030/1,2-rust2html/README.md b/homework/81030/1,2-rust2html/README.md
new file mode 100644
index 0000000..577a3d0
--- /dev/null
+++ b/homework/81030/1,2-rust2html/README.md
@@ -0,0 +1,50 @@
+# Rust syntax highlighter
+
+This is an implementation of a syntax highlighter for the Rust programming language. It takes a Rust program as input (or a Rust source file as an argument) and outputs a HTML page with a colorized version of the program. The C source for the highlighter is generated from a Flex-based scanner template (`rust2html.l`). (This is a solution to the first homework assignment.)
+
+There is also an extended and improved version of the highlighter which implements pretty-printing (`rust2html-prettyprint.l`). (This is a solution to the second homework assignment.)
+
+## Generating and building
+
+On Unix-like OS:
+
+* the regular highlighter:
+
+ flex rust2html.l
+ cc -o rust2html lex.yy.c
+
+* the extended (pretty-printing) highlighter:
+
+ flex rust2html-prettyprint.l
+ cc -o rust2html-prettyprint lex.yy.c
+
+Using the provided `Makefile`:
+
+* the regular highlighter:
+
+ make rust2html
+
+* the extended (pretty-printing) highlighter:
+
+ make rust2html-prettyprint
+
+## Usage
+
+This will generate a HTML document named `example.html` with a colorized version of the `example.rs` Rust program:
+
+ ./rust2html example.rs > example.html
+
+This will generate a HTML document named `example.html` with a colorized and pretty-printed version of the `example.rs` Rust program:
+
+ ./rust2html-prettyprint example.rs > example.html
+
+## Debugging
+
+If you want to be able to debug the `RUST2HTML` scanner (i.e. `rust2html` or `rust2html-prettyprint`), generate its code using the `-d` flag with Flex:
+
+ flex -d RUST2HTML.l
+ cc -o RUST2HTML lex.yy.c
+
+The following invocation of `RUST2HTML` will also write its standard error output to the `example.log` file (which is the main thing to look at when trying to debug the program):
+
+ ./RUST2HTML example.rs > example.html 2> example.log \ No newline at end of file
diff --git a/homework/81030/1,2-rust2html/example-chains.rs b/homework/81030/1,2-rust2html/example-chains.rs
new file mode 100644
index 0000000..be38300
--- /dev/null
+++ b/homework/81030/1,2-rust2html/example-chains.rs
@@ -0,0 +1,871 @@
+// Copyright 2015 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+//! Formatting of chained expressions, i.e. expressions which are chained by
+//! dots: struct and enum field access, method calls, and try shorthand (?).
+//!
+//! Instead of walking these subexpressions one-by-one, as is our usual strategy
+//! for expression formatting, we collect maximal sequences of these expressions
+//! and handle them simultaneously.
+//!
+//! Whenever possible, the entire chain is put on a single line. If that fails,
+//! we put each subexpression on a separate, much like the (default) function
+//! argument function argument strategy.
+//!
+//! Depends on config options: `chain_indent` is the indent to use for
+//! blocks in the parent/root/base of the chain (and the rest of the chain's
+//! alignment).
+//! E.g., `let foo = { aaaa; bbb; ccc }.bar.baz();`, we would layout for the
+//! following values of `chain_indent`:
+//! Block:
+//!
+//! ```ignore
+//! let foo = {
+//! aaaa;
+//! bbb;
+//! ccc
+//! }.bar
+//! .baz();
+//! ```
+//!
+//! Visual:
+//!
+//! ```ignore
+//! let foo = {
+//! aaaa;
+//! bbb;
+//! ccc
+//! }
+//! .bar
+//! .baz();
+//! ```
+//!
+//! If the first item in the chain is a block expression, we align the dots with
+//! the braces.
+//! Block:
+//!
+//! ```ignore
+//! let a = foo.bar
+//! .baz()
+//! .qux
+//! ```
+//!
+//! Visual:
+//!
+//! ```ignore
+//! let a = foo.bar
+//! .baz()
+//! .qux
+//! ```
+
+use comment::{rewrite_comment, CharClasses, FullCodeCharKind, RichChar};
+use config::IndentStyle;
+use expr::rewrite_call;
+use lists::extract_pre_comment;
+use macros::convert_try_mac;
+use rewrite::{Rewrite, RewriteContext};
+use shape::Shape;
+use source_map::SpanUtils;
+use utils::{
+ self, first_line_width, last_line_extendable, last_line_width, mk_sp, rewrite_ident,
+ trimmed_last_line_width, wrap_str,
+};
+
+use std::borrow::Cow;
+use std::cmp::min;
+
+use syntax::source_map::{BytePos, Span};
+use syntax::{ast, ptr};
+
+pub fn rewrite_chain(expr: &ast::Expr, context: &RewriteContext, shape: Shape) -> Option<String> {
+ let chain = Chain::from_ast(expr, context);
+ debug!("rewrite_chain {:?} {:?}", chain, shape);
+
+ // If this is just an expression with some `?`s, then format it trivially and
+ // return early.
+ if chain.children.is_empty() {
+ return chain.parent.rewrite(context, shape);
+ }
+
+ chain.rewrite(context, shape)
+}
+
+#[derive(Debug)]
+enum CommentPosition {
+ Back,
+ Top,
+}
+
+// An expression plus trailing `?`s to be formatted together.
+#[derive(Debug)]
+struct ChainItem {
+ kind: ChainItemKind,
+ tries: usize,
+ span: Span,
+}
+
+// FIXME: we can't use a reference here because to convert `try!` to `?` we
+// synthesise the AST node. However, I think we could use `Cow` and that
+// would remove a lot of cloning.
+#[derive(Debug)]
+enum ChainItemKind {
+ Parent(ast::Expr),
+ MethodCall(
+ ast::PathSegment,
+ Vec<ast::GenericArg>,
+ Vec<ptr::P<ast::Expr>>,
+ ),
+ StructField(ast::Ident),
+ TupleField(ast::Ident, bool),
+ Comment(String, CommentPosition),
+}
+
+impl ChainItemKind {
+ fn is_block_like(&self, context: &RewriteContext, reps: &str) -> bool {
+ match self {
+ ChainItemKind::Parent(ref expr) => utils::is_block_expr(context, expr, reps),
+ ChainItemKind::MethodCall(..)
+ | ChainItemKind::StructField(..)
+ | ChainItemKind::TupleField(..)
+ | ChainItemKind::Comment(..) => false,
+ }
+ }
+
+ fn is_tup_field_access(expr: &ast::Expr) -> bool {
+ match expr.node {
+ ast::ExprKind::Field(_, ref field) => {
+ field.name.to_string().chars().all(|c| c.is_digit(10))
+ }
+ _ => false,
+ }
+ }
+
+ fn from_ast(context: &RewriteContext, expr: &ast::Expr) -> (ChainItemKind, Span) {
+ let (kind, span) = match expr.node {
+ ast::ExprKind::MethodCall(ref segment, ref expressions) => {
+ let types = if let Some(ref generic_args) = segment.args {
+ if let ast::GenericArgs::AngleBracketed(ref data) = **generic_args {
+ data.args.clone()
+ } else {
+ vec![]
+ }
+ } else {
+ vec![]
+ };
+ let span = mk_sp(expressions[0].span.hi(), expr.span.hi());
+ let kind = ChainItemKind::MethodCall(segment.clone(), types, expressions.clone());
+ (kind, span)
+ }
+ ast::ExprKind::Field(ref nested, field) => {
+ let kind = if Self::is_tup_field_access(expr) {
+ ChainItemKind::TupleField(field, Self::is_tup_field_access(nested))
+ } else {
+ ChainItemKind::StructField(field)
+ };
+ let span = mk_sp(nested.span.hi(), field.span.hi());
+ (kind, span)
+ }
+ _ => return (ChainItemKind::Parent(expr.clone()), expr.span),
+ };
+
+ // Remove comments from the span.
+ let lo = context.snippet_provider.span_before(span, ".");
+ (kind, mk_sp(lo, span.hi()))
+ }
+}
+
+impl Rewrite for ChainItem {
+ fn rewrite(&self, context: &RewriteContext, shape: Shape) -> Option<String> {
+ let shape = shape.sub_width(self.tries)?;
+ let rewrite = match self.kind {
+ ChainItemKind::Parent(ref expr) => expr.rewrite(context, shape)?,
+ ChainItemKind::MethodCall(ref segment, ref types, ref exprs) => {
+ Self::rewrite_method_call(segment.ident, types, exprs, self.span, context, shape)?
+ }
+ ChainItemKind::StructField(ident) => format!(".{}", rewrite_ident(context, ident)),
+ ChainItemKind::TupleField(ident, nested) => format!(
+ "{}.{}",
+ if nested { " " } else { "" },
+ rewrite_ident(context, ident)
+ ),
+ ChainItemKind::Comment(ref comment, _) => {
+ rewrite_comment(comment, false, shape, context.config)?
+ }
+ };
+ Some(format!("{}{}", rewrite, "?".repeat(self.tries)))
+ }
+}
+
+impl ChainItem {
+ fn new(context: &RewriteContext, expr: &ast::Expr, tries: usize) -> ChainItem {
+ let (kind, span) = ChainItemKind::from_ast(context, expr);
+ ChainItem { kind, tries, span }
+ }
+
+ fn comment(span: Span, comment: String, pos: CommentPosition) -> ChainItem {
+ ChainItem {
+ kind: ChainItemKind::Comment(comment, pos),
+ tries: 0,
+ span,
+ }
+ }
+
+ fn is_comment(&self) -> bool {
+ match self.kind {
+ ChainItemKind::Comment(..) => true,
+ _ => false,
+ }
+ }
+
+ fn rewrite_method_call(
+ method_name: ast::Ident,
+ types: &[ast::GenericArg],
+ args: &[ptr::P<ast::Expr>],
+ span: Span,
+ context: &RewriteContext,
+ shape: Shape,
+ ) -> Option<String> {
+ let type_str = if types.is_empty() {
+ String::new()
+ } else {
+ let type_list = types
+ .iter()
+ .map(|ty| ty.rewrite(context, shape))
+ .collect::<Option<Vec<_>>>()?;
+
+ format!("::<{}>", type_list.join(", "))
+ };
+ let callee_str = format!(".{}{}", rewrite_ident(context, method_name), type_str);
+ rewrite_call(context, &callee_str, &args[1..], span, shape)
+ }
+}
+
+#[derive(Debug)]
+struct Chain {
+ parent: ChainItem,
+ children: Vec<ChainItem>,
+}
+
+impl Chain {
+ fn from_ast(expr: &ast::Expr, context: &RewriteContext) -> Chain {
+ let subexpr_list = Self::make_subexpr_list(expr, context);
+
+ // Un-parse the expression tree into ChainItems
+ let mut rev_children = vec![];
+ let mut sub_tries = 0;
+ for subexpr in &subexpr_list {
+ match subexpr.node {
+ ast::ExprKind::Try(_) => sub_tries += 1,
+ _ => {
+ rev_children.push(ChainItem::new(context, subexpr, sub_tries));
+ sub_tries = 0;
+ }
+ }
+ }
+
+ fn is_tries(s: &str) -> bool {
+ s.chars().all(|c| c == '?')
+ }
+
+ fn is_post_comment(s: &str) -> bool {
+ let comment_start_index = s.chars().position(|c| c == '/');
+ if comment_start_index.is_none() {
+ return false;
+ }
+
+ let newline_index = s.chars().position(|c| c == '\n');
+ if newline_index.is_none() {
+ return true;
+ }
+
+ comment_start_index.unwrap() < newline_index.unwrap()
+ }
+
+ fn handle_post_comment(
+ post_comment_span: Span,
+ post_comment_snippet: &str,
+ prev_span_end: &mut BytePos,
+ children: &mut Vec<ChainItem>,
+ ) {
+ let white_spaces: &[_] = &[' ', '\t'];
+ if post_comment_snippet
+ .trim_matches(white_spaces)
+ .starts_with('\n')
+ {
+ // No post comment.
+ return;
+ }
+ let trimmed_snippet = trim_tries(post_comment_snippet);
+ if is_post_comment(&trimmed_snippet) {
+ children.push(ChainItem::comment(
+ post_comment_span,
+ trimmed_snippet.trim().to_owned(),
+ CommentPosition::Back,
+ ));
+ *prev_span_end = post_comment_span.hi();
+ }
+ }
+
+ let parent = rev_children.pop().unwrap();
+ let mut children = vec![];
+ let mut prev_span_end = parent.span.hi();
+ let mut iter = rev_children.into_iter().rev().peekable();
+ if let Some(first_chain_item) = iter.peek() {
+ let comment_span = mk_sp(prev_span_end, first_chain_item.span.lo());
+ let comment_snippet = context.snippet(comment_span);
+ if !is_tries(comment_snippet.trim()) {
+ handle_post_comment(
+ comment_span,
+ comment_snippet,
+ &mut prev_span_end,
+ &mut children,
+ );
+ }
+ }
+ while let Some(chain_item) = iter.next() {
+ let comment_snippet = context.snippet(chain_item.span);
+ // FIXME: Figure out the way to get a correct span when converting `try!` to `?`.
+ let handle_comment =
+ !(context.config.use_try_shorthand() || is_tries(comment_snippet.trim()));
+
+ // Pre-comment
+ if handle_comment {
+ let pre_comment_span = mk_sp(prev_span_end, chain_item.span.lo());
+ let pre_comment_snippet = trim_tries(context.snippet(pre_comment_span));
+ let (pre_comment, _) = extract_pre_comment(&pre_comment_snippet);
+ match pre_comment {
+ Some(ref comment) if !comment.is_empty() => {
+ children.push(ChainItem::comment(
+ pre_comment_span,
+ comment.to_owned(),
+ CommentPosition::Top,
+ ));
+ }
+ _ => (),
+ }
+ }
+
+ prev_span_end = chain_item.span.hi();
+ children.push(chain_item);
+
+ // Post-comment
+ if !handle_comment || iter.peek().is_none() {
+ continue;
+ }
+
+ let next_lo = iter.peek().unwrap().span.lo();
+ let post_comment_span = mk_sp(prev_span_end, next_lo);
+ let post_comment_snippet = context.snippet(post_comment_span);
+ handle_post_comment(
+ post_comment_span,
+ post_comment_snippet,
+ &mut prev_span_end,
+ &mut children,
+ );
+ }
+
+ Chain { parent, children }
+ }
+
+ // Returns a Vec of the prefixes of the chain.
+ // E.g., for input `a.b.c` we return [`a.b.c`, `a.b`, 'a']
+ fn make_subexpr_list(expr: &ast::Expr, context: &RewriteContext) -> Vec<ast::Expr> {
+ let mut subexpr_list = vec![expr.clone()];
+
+ while let Some(subexpr) = Self::pop_expr_chain(subexpr_list.last().unwrap(), context) {
+ subexpr_list.push(subexpr.clone());
+ }
+
+ subexpr_list
+ }
+
+ // Returns the expression's subexpression, if it exists. When the subexpr
+ // is a try! macro, we'll convert it to shorthand when the option is set.
+ fn pop_expr_chain(expr: &ast::Expr, context: &RewriteContext) -> Option<ast::Expr> {
+ match expr.node {
+ ast::ExprKind::MethodCall(_, ref expressions) => {
+ Some(Self::convert_try(&expressions[0], context))
+ }
+ ast::ExprKind::Field(ref subexpr, _) | ast::ExprKind::Try(ref subexpr) => {
+ Some(Self::convert_try(subexpr, context))
+ }
+ _ => None,
+ }
+ }
+
+ fn convert_try(expr: &ast::Expr, context: &RewriteContext) -> ast::Expr {
+ match expr.node {
+ ast::ExprKind::Mac(ref mac) if context.config.use_try_shorthand() => {
+ if let Some(subexpr) = convert_try_mac(mac, context) {
+ subexpr
+ } else {
+ expr.clone()
+ }
+ }
+ _ => expr.clone(),
+ }
+ }
+}
+
+impl Rewrite for Chain {
+ fn rewrite(&self, context: &RewriteContext, shape: Shape) -> Option<String> {
+ debug!("rewrite chain {:?} {:?}", self, shape);
+
+ let mut formatter = match context.config.indent_style() {
+ IndentStyle::Block => Box::new(ChainFormatterBlock::new(self)) as Box<ChainFormatter>,
+ IndentStyle::Visual => Box::new(ChainFormatterVisual::new(self)) as Box<ChainFormatter>,
+ };
+
+ formatter.format_root(&self.parent, context, shape)?;
+ if let Some(result) = formatter.pure_root() {
+ return wrap_str(result, context.config.max_width(), shape);
+ }
+
+ // Decide how to layout the rest of the chain.
+ let child_shape = formatter.child_shape(context, shape)?;
+
+ formatter.format_children(context, child_shape)?;
+ formatter.format_last_child(context, shape, child_shape)?;
+
+ let result = formatter.join_rewrites(context, child_shape)?;
+ wrap_str(result, context.config.max_width(), shape)
+ }
+}
+
+// There are a few types for formatting chains. This is because there is a lot
+// in common between formatting with block vs visual indent, but they are
+// different enough that branching on the indent all over the place gets ugly.
+// Anything that can format a chain is a ChainFormatter.
+trait ChainFormatter {
+ // Parent is the first item in the chain, e.g., `foo` in `foo.bar.baz()`.
+ // Root is the parent plus any other chain items placed on the first line to
+ // avoid an orphan. E.g.,
+ // ```ignore
+ // foo.bar
+ // .baz()
+ // ```
+ // If `bar` were not part of the root, then foo would be orphaned and 'float'.
+ fn format_root(
+ &mut self,
+ parent: &ChainItem,
+ context: &RewriteContext,
+ shape: Shape,
+ ) -> Option<()>;
+ fn child_shape(&self, context: &RewriteContext, shape: Shape) -> Option<Shape>;
+ fn format_children(&mut self, context: &RewriteContext, child_shape: Shape) -> Option<()>;
+ fn format_last_child(
+ &mut self,
+ context: &RewriteContext,
+ shape: Shape,
+ child_shape: Shape,
+ ) -> Option<()>;
+ fn join_rewrites(&self, context: &RewriteContext, child_shape: Shape) -> Option<String>;
+ // Returns `Some` if the chain is only a root, None otherwise.
+ fn pure_root(&mut self) -> Option<String>;
+}
+
+// Data and behaviour that is shared by both chain formatters. The concrete
+// formatters can delegate much behaviour to `ChainFormatterShared`.
+struct ChainFormatterShared<'a> {
+ // The current working set of child items.
+ children: &'a [ChainItem],
+ // The current rewrites of items (includes trailing `?`s, but not any way to
+ // connect the rewrites together).
+ rewrites: Vec<String>,
+ // Whether the chain can fit on one line.
+ fits_single_line: bool,
+ // The number of children in the chain. This is not equal to `self.children.len()`
+ // because `self.children` will change size as we process the chain.
+ child_count: usize,
+}
+
+impl<'a> ChainFormatterShared<'a> {
+ fn new(chain: &'a Chain) -> ChainFormatterShared<'a> {
+ ChainFormatterShared {
+ children: &chain.children,
+ rewrites: Vec::with_capacity(chain.children.len() + 1),
+ fits_single_line: false,
+ child_count: chain.children.len(),
+ }
+ }
+
+ fn pure_root(&mut self) -> Option<String> {
+ if self.children.is_empty() {
+ assert_eq!(self.rewrites.len(), 1);
+ Some(self.rewrites.pop().unwrap())
+ } else {
+ None
+ }
+ }
+
+ // Rewrite the last child. The last child of a chain requires special treatment. We need to
+ // know whether 'overflowing' the last child make a better formatting:
+ //
+ // A chain with overflowing the last child:
+ // ```ignore
+ // parent.child1.child2.last_child(
+ // a,
+ // b,
+ // c,
+ // )
+ // ```
+ //
+ // A chain without overflowing the last child (in vertical layout):
+ // ```ignore
+ // parent
+ // .child1
+ // .child2
+ // .last_child(a, b, c)
+ // ```
+ //
+ // In particular, overflowing is effective when the last child is a method with a multi-lined
+ // block-like argument (e.g. closure):
+ // ```ignore
+ // parent.child1.child2.last_child(|a, b, c| {
+ // let x = foo(a, b, c);
+ // let y = bar(a, b, c);
+ //
+ // // ...
+ //
+ // result
+ // })
+ // ```
+ fn format_last_child(
+ &mut self,
+ may_extend: bool,
+ context: &RewriteContext,
+ shape: Shape,
+ child_shape: Shape,
+ ) -> Option<()> {
+ let last = self.children.last()?;
+ let extendable = may_extend && last_line_extendable(&self.rewrites[0]);
+ let prev_last_line_width = last_line_width(&self.rewrites[0]);
+
+ // Total of all items excluding the last.
+ let almost_total = if extendable {
+ prev_last_line_width
+ } else {
+ self.rewrites.iter().fold(0, |a, b| a + b.len())
+ } + last.tries;
+ let one_line_budget = if self.child_count == 1 {
+ shape.width
+ } else {
+ min(shape.width, context.config.width_heuristics().chain_width)
+ }
+ .saturating_sub(almost_total);
+
+ let all_in_one_line = !self.children.iter().any(ChainItem::is_comment)
+ && self.rewrites.iter().all(|s| !s.contains('\n'))
+ && one_line_budget > 0;
+ let last_shape = if all_in_one_line {
+ shape.sub_width(last.tries)?
+ } else if extendable {
+ child_shape.sub_width(last.tries)?
+ } else {
+ child_shape.sub_width(shape.rhs_overhead(context.config) + last.tries)?
+ };
+
+ let mut last_subexpr_str = None;
+ if all_in_one_line || extendable {
+ // First we try to 'overflow' the last child and see if it looks better than using
+ // vertical layout.
+ let one_line_shape = if context.use_block_indent() {
+ last_shape.offset_left(almost_total)
+ } else {
+ last_shape
+ .visual_indent(almost_total)
+ .sub_width(almost_total)
+ };
+
+ if let Some(one_line_shape) = one_line_shape {
+ if let Some(rw) = last.rewrite(context, one_line_shape) {
+ // We allow overflowing here only if both of the following conditions match:
+ // 1. The entire chain fits in a single line except the last child.
+ // 2. `last_child_str.lines().count() >= 5`.
+ let line_count = rw.lines().count();
+ let could_fit_single_line = first_line_width(&rw) <= one_line_budget;
+ if could_fit_single_line && line_count >= 5 {
+ last_subexpr_str = Some(rw);
+ self.fits_single_line = all_in_one_line;
+ } else {
+ // We could not know whether overflowing is better than using vertical
+ // layout, just by looking at the overflowed rewrite. Now we rewrite the
+ // last child on its own line, and compare two rewrites to choose which is
+ // better.
+ let last_shape = child_shape
+ .sub_width(shape.rhs_overhead(context.config) + last.tries)?;
+ match last.rewrite(context, last_shape) {
+ Some(ref new_rw) if !could_fit_single_line => {
+ last_subexpr_str = Some(new_rw.clone());
+ }
+ Some(ref new_rw) if new_rw.lines().count() >= line_count => {
+ last_subexpr_str = Some(rw);
+ self.fits_single_line = could_fit_single_line && all_in_one_line;
+ }
+ new_rw @ Some(..) => {
+ last_subexpr_str = new_rw;
+ }
+ _ => {
+ last_subexpr_str = Some(rw);
+ self.fits_single_line = could_fit_single_line && all_in_one_line;
+ }
+ }
+ }
+ }
+ }
+ }
+
+ let last_shape = if context.use_block_indent() {
+ last_shape
+ } else {
+ child_shape.sub_width(shape.rhs_overhead(context.config) + last.tries)?
+ };
+
+ last_subexpr_str = last_subexpr_str.or_else(|| last.rewrite(context, last_shape));
+ self.rewrites.push(last_subexpr_str?);
+ Some(())
+ }
+
+ fn join_rewrites(&self, context: &RewriteContext, child_shape: Shape) -> Option<String> {
+ let connector = if self.fits_single_line {
+ // Yay, we can put everything on one line.
+ Cow::from("")
+ } else {
+ // Use new lines.
+ if *context.force_one_line_chain.borrow() {
+ return None;
+ }
+ child_shape.to_string_with_newline(context.config)
+ };
+
+ let mut rewrite_iter = self.rewrites.iter();
+ let mut result = rewrite_iter.next().unwrap().clone();
+ let children_iter = self.children.iter();
+ let iter = rewrite_iter.zip(children_iter);
+
+ for (rewrite, chain_item) in iter {
+ match chain_item.kind {
+ ChainItemKind::Comment(_, CommentPosition::Back) => result.push(' '),
+ ChainItemKind::Comment(_, CommentPosition::Top) => result.push_str(&connector),
+ _ => result.push_str(&connector),
+ }
+ result.push_str(&rewrite);
+ }
+
+ Some(result)
+ }
+}
+
+// Formats a chain using block indent.
+struct ChainFormatterBlock<'a> {
+ shared: ChainFormatterShared<'a>,
+ root_ends_with_block: bool,
+}
+
+impl<'a> ChainFormatterBlock<'a> {
+ fn new(chain: &'a Chain) -> ChainFormatterBlock<'a> {
+ ChainFormatterBlock {
+ shared: ChainFormatterShared::new(chain),
+ root_ends_with_block: false,
+ }
+ }
+}
+
+impl<'a> ChainFormatter for ChainFormatterBlock<'a> {
+ fn format_root(
+ &mut self,
+ parent: &ChainItem,
+ context: &RewriteContext,
+ shape: Shape,
+ ) -> Option<()> {
+ let mut root_rewrite: String = parent.rewrite(context, shape)?;
+
+ let mut root_ends_with_block = parent.kind.is_block_like(context, &root_rewrite);
+ let tab_width = context.config.tab_spaces().saturating_sub(shape.offset);
+
+ while root_rewrite.len() <= tab_width && !root_rewrite.contains('\n') {
+ let item = &self.shared.children[0];
+ if let ChainItemKind::Comment(..) = item.kind {
+ break;
+ }
+ let shape = shape.offset_left(root_rewrite.len())?;
+ match &item.rewrite(context, shape) {
+ Some(rewrite) => root_rewrite.push_str(rewrite),
+ None => break,
+ }
+
+ root_ends_with_block = last_line_extendable(&root_rewrite);
+
+ self.shared.children = &self.shared.children[1..];
+ if self.shared.children.is_empty() {
+ break;
+ }
+ }
+ self.shared.rewrites.push(root_rewrite);
+ self.root_ends_with_block = root_ends_with_block;
+ Some(())
+ }
+
+ fn child_shape(&self, context: &RewriteContext, shape: Shape) -> Option<Shape> {
+ Some(
+ if self.root_ends_with_block {
+ shape.block_indent(0)
+ } else {
+ shape.block_indent(context.config.tab_spaces())
+ }
+ .with_max_width(context.config),
+ )
+ }
+
+ fn format_children(&mut self, context: &RewriteContext, child_shape: Shape) -> Option<()> {
+ for item in &self.shared.children[..self.shared.children.len() - 1] {
+ let rewrite = item.rewrite(context, child_shape)?;
+ self.shared.rewrites.push(rewrite);
+ }
+ Some(())
+ }
+
+ fn format_last_child(
+ &mut self,
+ context: &RewriteContext,
+ shape: Shape,
+ child_shape: Shape,
+ ) -> Option<()> {
+ self.shared
+ .format_last_child(true, context, shape, child_shape)
+ }
+
+ fn join_rewrites(&self, context: &RewriteContext, child_shape: Shape) -> Option<String> {
+ self.shared.join_rewrites(context, child_shape)
+ }
+
+ fn pure_root(&mut self) -> Option<String> {
+ self.shared.pure_root()
+ }
+}
+
+// Format a chain using visual indent.
+struct ChainFormatterVisual<'a> {
+ shared: ChainFormatterShared<'a>,
+ // The extra offset from the chain's shape to the position of the `.`
+ offset: usize,
+}
+
+impl<'a> ChainFormatterVisual<'a> {
+ fn new(chain: &'a Chain) -> ChainFormatterVisual<'a> {
+ ChainFormatterVisual {
+ shared: ChainFormatterShared::new(chain),
+ offset: 0,
+ }
+ }
+}
+
+impl<'a> ChainFormatter for ChainFormatterVisual<'a> {
+ fn format_root(
+ &mut self,
+ parent: &ChainItem,
+ context: &RewriteContext,
+ shape: Shape,
+ ) -> Option<()> {
+ let parent_shape = shape.visual_indent(0);
+ let mut root_rewrite = parent.rewrite(context, parent_shape)?;
+ let multiline = root_rewrite.contains('\n');
+ self.offset = if multiline {
+ last_line_width(&root_rewrite).saturating_sub(shape.used_width())
+ } else {
+ trimmed_last_line_width(&root_rewrite)
+ };
+
+ if !multiline || parent.kind.is_block_like(context, &root_rewrite) {
+ let item = &self.shared.children[0];
+ if let ChainItemKind::Comment(..) = item.kind {
+ self.shared.rewrites.push(root_rewrite);
+ return Some(());
+ }
+ let child_shape = parent_shape
+ .visual_indent(self.offset)
+ .sub_width(self.offset)?;
+ let rewrite = item.rewrite(context, child_shape)?;
+ match wrap_str(rewrite, context.config.max_width(), shape) {
+ Some(rewrite) => root_rewrite.push_str(&rewrite),
+ None => {
+ // We couldn't fit in at the visual indent, try the last
+ // indent.
+ let rewrite = item.rewrite(context, parent_shape)?;
+ root_rewrite.push_str(&rewrite);
+ self.offset = 0;
+ }
+ }
+
+ self.shared.children = &self.shared.children[1..];
+ }
+
+ self.shared.rewrites.push(root_rewrite);
+ Some(())
+ }
+
+ fn child_shape(&self, context: &RewriteContext, shape: Shape) -> Option<Shape> {
+ shape
+ .with_max_width(context.config)
+ .offset_left(self.offset)
+ .map(|s| s.visual_indent(0))
+ }
+
+ fn format_children(&mut self, context: &RewriteContext, child_shape: Shape) -> Option<()> {
+ for item in &self.shared.children[..self.shared.children.len() - 1] {
+ let rewrite = item.rewrite(context, child_shape)?;
+ self.shared.rewrites.push(rewrite);
+ }
+ Some(())
+ }
+
+ fn format_last_child(
+ &mut self,
+ context: &RewriteContext,
+ shape: Shape,
+ child_shape: Shape,
+ ) -> Option<()> {
+ self.shared
+ .format_last_child(false, context, shape, child_shape)
+ }
+
+ fn join_rewrites(&self, context: &RewriteContext, child_shape: Shape) -> Option<String> {
+ self.shared.join_rewrites(context, child_shape)
+ }
+
+ fn pure_root(&mut self) -> Option<String> {
+ self.shared.pure_root()
+ }
+}
+
+/// Remove try operators (`?`s) that appear in the given string. If removing
+/// them leaves an empty line, remove that line as well unless it is the first
+/// line (we need the first newline for detecting pre/post comment).
+fn trim_tries(s: &str) -> String {
+ let mut result = String::with_capacity(s.len());
+ let mut line_buffer = String::with_capacity(s.len());
+ for (kind, rich_char) in CharClasses::new(s.chars()) {
+ match rich_char.get_char() {
+ '\n' => {
+ if result.is_empty() || !line_buffer.trim().is_empty() {
+ result.push_str(&line_buffer);
+ result.push('\n')
+ }
+ line_buffer.clear();
+ }
+ '?' if kind == FullCodeCharKind::Normal => continue,
+ c => line_buffer.push(c),
+ }
+ }
+ if !line_buffer.trim().is_empty() {
+ result.push_str(&line_buffer);
+ }
+ result
+}
diff --git a/homework/81030/1,2-rust2html/example-ugly.rs b/homework/81030/1,2-rust2html/example-ugly.rs
new file mode 100644
index 0000000..2b17e2f
--- /dev/null
+++ b/homework/81030/1,2-rust2html/example-ugly.rs
@@ -0,0 +1,29 @@
+/*
+ * The following function returns the first answer.
+ * It takes no arguments and returns an unsigned integer.
+ */
+fn answer1 () -> usize {
+ return 42; // answer to the Ultimate Question of Life, the Universe, and Everything
+}
+/*
+ * The following function returns the second answer.
+ * It takes no arguments and returns a floating-point number.
+ */
+fn answer2( ) -> f64
+{
+ return -4.2;
+}
+
+/*
+ * The following function returns the third answer.
+ * It takes no arguments and returns a single character.
+ */
+fn answer3() -> char {
+ return 'E' ; }
+
+
+fn main( ) {
+ println! ("The \"First Answer\" is: '{}'", answer1( ) );
+ println!("The \"Second Answer\" is: '{}'" , answer2()) ;
+ println!( "The \"Third Answer\" is: '{}'" , answer3 () );
+}
diff --git a/homework/81030/1,2-rust2html/example.rs b/homework/81030/1,2-rust2html/example.rs
new file mode 100644
index 0000000..396aa0c
--- /dev/null
+++ b/homework/81030/1,2-rust2html/example.rs
@@ -0,0 +1,29 @@
+/*
+ * The following function returns the first answer.
+ * It takes no arguments and returns an unsigned integer.
+ */
+fn answer1() -> usize {
+ return 42; // answer to the Ultimate Question of Life, the Universe, and Everything
+}
+
+/*
+ * The following function returns the second answer.
+ * It takes no arguments and returns a floating-point number.
+ */
+fn answer2() -> f64 {
+ return -4.2;
+}
+
+/*
+ * The following function returns the third answer.
+ * It takes no arguments and returns a single character.
+ */
+fn answer3() -> char {
+ return 'E';
+}
+
+fn main() {
+ println!("The \"First Answer\" is: '{}'", answer1());
+ println!("The \"Second Answer\" is: '{}'", answer2());
+ println!("The \"Third Answer\" is: '{}'", answer3());
+}
diff --git a/homework/81030/1,2-rust2html/rust2html-prettyprint.l b/homework/81030/1,2-rust2html/rust2html-prettyprint.l
new file mode 100644
index 0000000..2d90931
--- /dev/null
+++ b/homework/81030/1,2-rust2html/rust2html-prettyprint.l
@@ -0,0 +1,344 @@
+%option nounistd
+%option noyywrap
+%option stack
+
+%{
+#include <stdbool.h>
+
+#define INDENTATION " "
+#define SPACE " "
+#define HIGHLIGHT_TERMINATOR "</span>"
+
+#define PRINT_HIGHLIGHTED_TOKEN_INITIATOR(type) \
+ printf("<span class=\"" #type "\">%s", yytext)
+
+#define PRINT_HIGHLIGHTED_TOKEN_TERMINATOR \
+ printf("%s" HIGHLIGHT_TERMINATOR, yytext)
+
+#define PRINT_HIGHLIGHTED_TOKEN(type) \
+ printf("<span class=\"" #type "\">%s</span>", yytext)
+
+#define PRINT_TOKEN \
+ printf("%s", yytext)
+
+#define PRINT_INDENTED_NEWLINE \
+ printf("\n"); \
+ for (int i = 0; i < indentation_level; i++) \
+ printf(INDENTATION)
+
+#define CLEAR_MODES \
+ require_space_mode = false; \
+ adjacent_parenthesis_mode = false
+
+#define HANDLE_COMMENT_INITIATOR(type) \
+ PRINT_HIGHLIGHTED_TOKEN_INITIATOR(comment); \
+ yy_push_state(type); \
+ CLEAR_MODES
+
+#define HANDLE_LITERAL(type) \
+ if (require_space_mode) \
+ printf(SPACE); \
+ PRINT_HIGHLIGHTED_TOKEN(type); \
+ require_space_mode = true; \
+ adjacent_parenthesis_mode = false
+
+#define HANDLE_IDENTIFIER \
+ if (require_space_mode) \
+ printf(SPACE); \
+ PRINT_HIGHLIGHTED_TOKEN(identifier); \
+ require_space_mode = true; \
+ adjacent_parenthesis_mode = true
+
+#define HANDLE_EOL \
+ yy_pop_state(); \
+ adjacent_parenthesis_mode = true; \
+ yyless(0)
+
+static int indentation_level = 0;
+static bool require_space_mode = false;
+static bool adjacent_parenthesis_mode = false;
+%}
+
+%s multiline_list
+%s path
+%s directive
+
+%x line_comment
+%x block_comment
+%x eol
+
+NONEOL_CHAR .
+DEC_DIGIT 0|{NONZERO_DEC_DIGIT}
+NONZERO_DEC_DIGIT [1-9]
+OCT_DIGIT [0-7]
+HEX_DIGIT {DEC_DIGIT}|[a-fA-F]
+
+IDENTIFIER {IDENTIFIER_FIRST_CHAR}({DEC_DIGIT}|{IDENTIFIER_FIRST_CHAR})*
+IDENTIFIER_FIRST_CHAR [A-Za-z_]
+
+LINE_COMMENT_BEGIN "//"
+BLOCK_COMMENT_INITIATOR "/*"
+BLOCK_COMMENT_TERMINATOR "*/"
+
+KEYWORD _|abstract|alignof|as|become|box|break|const|continue|crate|do|else|enum|extern|false|final|fn|for|if|impl|in|let|loop|macro|match|mod|move|mut|offsetof|override|priv|proc|pub|pure|ref|return|Self|self|sizeof|static|struct|super|trait|true|type|typeof|unsafe|unsized|use|virtual|where|while|yield
+
+ESCAPE {COMMON_ESCAPE}|{UNICODE_ESCAPE}
+COMMON_ESCAPE \\|n|r|t|0|x{HEX_DIGIT}{2}
+UNICODE_ESCAPE "u{"{HEX_DIGIT}{6}"}"
+
+CHAR_LITERAL '{CHAR_BODY}'
+CHAR_BODY [^']|(\\('|{ESCAPE}))
+STRING_LITERAL \"{STRING_BODY}*\"
+STRING_BODY [^\"]|(\\(\"|{ESCAPE}))
+BYTE_CHAR_LITERAL b'{BYTE_CHAR_BODY}'
+BYTE_CHAR_BODY [^']|(\\('|{COMMON_ESCAPE}))
+BYTE_STRING_LITERAL b\"{BYTE_STRING_BODY}*\"
+BYTE_STRING_BODY [^\"]|(\\(\"|{ESCAPE}))
+
+NUMBER_LITERAL {NONZERO_DEC_DIGIT}{DEC_DIGIT}*{FLOAT_SUFFIX}?|(0{DEC_DIGIT_LITERAL}*{FLOAT_SUFFIX}?|b[10_]+|o{OCT_DIGIT_LITERAL}+|x{HEX_DIGIT_LITERAL}+)
+FLOAT_SUFFIX {EXPONENT}|"."{DEC_DIGIT_LITERAL}+{EXPONENT}?
+EXPONENT [Ee][-+]?{DEC_DIGIT_LITERAL}+
+DEC_DIGIT_LITERAL {DEC_DIGIT}|_
+HEX_DIGIT_LITERAL {HEX_DIGIT}|_
+OCT_DIGIT_LITERAL {OCT_DIGIT}|_
+
+%%
+
+<eol>[[:space:]]+
+
+<eol>";"|"}" {
+ HANDLE_EOL;
+}
+
+<eol>"else" {
+ HANDLE_EOL;
+ printf(SPACE);
+}
+
+<eol>. {
+ HANDLE_EOL;
+ PRINT_INDENTED_NEWLINE;
+}
+
+<line_comment>[[:blank:]]*\n {
+ yy_pop_state();
+ printf(HIGHLIGHT_TERMINATOR);
+ yy_push_state(eol);
+}
+
+<block_comment>{BLOCK_COMMENT_TERMINATOR} {
+ yy_pop_state();
+ PRINT_HIGHLIGHTED_TOKEN_TERMINATOR;
+ yy_push_state(eol);
+}
+
+<line_comment>. {
+ PRINT_TOKEN;
+}
+
+<block_comment>(.|\n) {
+ PRINT_TOKEN;
+}
+
+[[:space:]]+
+
+"use" {
+ HANDLE_LITERAL(keyword);
+ yy_push_state(path);
+}
+
+{KEYWORD} {
+ HANDLE_LITERAL(keyword);
+}
+
+{IDENTIFIER} {
+ HANDLE_IDENTIFIER;
+}
+
+{NUMBER_LITERAL} {
+ HANDLE_LITERAL(number);
+}
+
+{CHAR_LITERAL}|{STRING_LITERAL}|{BYTE_CHAR_LITERAL}|{BYTE_STRING_LITERAL} {
+ HANDLE_LITERAL(string);
+}
+
+{LINE_COMMENT_BEGIN} {
+ HANDLE_COMMENT_INITIATOR(line_comment);
+}
+
+{BLOCK_COMMENT_INITIATOR} {
+ HANDLE_COMMENT_INITIATOR(block_comment);
+}
+
+<multiline_list>"," {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(eol);
+}
+
+","|":" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ require_space_mode = true;
+ adjacent_parenthesis_mode = false;
+}
+
+<path>";" {
+ yy_pop_state();
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(eol);
+}
+
+";" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(eol);
+}
+
+"#" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(directive);
+}
+
+"?" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ require_space_mode = true;
+ adjacent_parenthesis_mode = false;
+}
+
+"!" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ require_space_mode = false;
+ adjacent_parenthesis_mode = true;
+}
+
+"."|"::"|"'"|"$"|"@"|"~" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+}
+
+"=="|"=>"|"="|"!="|"<="|"<<"|"<<="|"<"|">="|">>"|">>="|">"|"<-"|"->"|"-"|"-="|"&&"|"&"|"&="|"||"|"|"|"|="|"+"|"+="|"*"|"*="|"/"|"/="|"^"|"^="|"%"|"%="|".."|"..." {
+ printf(SPACE);
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ require_space_mode = true;
+ adjacent_parenthesis_mode = false;
+}
+
+"(" {
+ if (!adjacent_parenthesis_mode)
+ printf(SPACE);
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(INITIAL);
+}
+
+<directive>"[" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+}
+
+"[" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(INITIAL);
+}
+
+<directive>"]" {
+ yy_pop_state();
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(eol);
+}
+
+")"|"]" {
+ yy_pop_state();
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ require_space_mode = true;
+ adjacent_parenthesis_mode = false;
+}
+
+<path>"{" {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+}
+
+"{" {
+ if (require_space_mode)
+ printf(SPACE);
+ indentation_level++;
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+ yy_push_state(multiline_list);
+ yy_push_state(eol);
+}
+
+<path>"}" {
+ yy_pop_state();
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ CLEAR_MODES;
+}
+
+"}" {
+ yy_pop_state();
+ indentation_level--;
+ PRINT_INDENTED_NEWLINE;
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+ if (0 == indentation_level)
+ printf("\n");
+ CLEAR_MODES;
+ yy_push_state(eol);
+}
+
+. {
+ PRINT_TOKEN; /* echo the rest */
+}
+
+%%
+
+int main(int argc, const char* const argv[]) {
+ yyin = argc > 1 ? fopen(argv[1], "r") : stdin;
+
+ printf(
+ "<!DOCTYPE html>\n"
+ "<html>\n"
+ "\n"
+ "<head>\n"
+ " <style>\n"
+ " .comment {\n"
+ " color: gray;\n"
+ " }\n"
+ "\n"
+ " .keyword {\n"
+ " color: red;\n"
+ " }\n"
+ "\n"
+ " .string {\n"
+ " color: green;\n"
+ " }\n"
+ "\n"
+ " .number {\n"
+ " color: blue;\n"
+ " }\n"
+ "\n"
+ " .symbol {\n"
+ " font-weight: bold;\n"
+ " }\n"
+ " </style>\n"
+ "</head>\n"
+ "\n"
+ "<body>\n"
+ " <pre class=\"code\">\n"
+ );
+ yylex();
+ printf(
+ "\n"
+ " </pre>\n"
+ "</body>\n"
+ "\n"
+ "</html>"
+ );
+
+ return 0;
+}
diff --git a/homework/81030/1,2-rust2html/rust2html.l b/homework/81030/1,2-rust2html/rust2html.l
new file mode 100644
index 0000000..de49285
--- /dev/null
+++ b/homework/81030/1,2-rust2html/rust2html.l
@@ -0,0 +1,130 @@
+%option nounistd
+%option noyywrap
+
+%{
+#define PRINT_HIGHLIGHTED_TOKEN(type) \
+ printf("<span class=\"" #type "\">%s</span>", yytext)
+%}
+
+NONEOL_CHAR .
+DEC_DIGIT 0|{NONZERO_DEC_DIGIT}
+NONZERO_DEC_DIGIT [1-9]
+OCT_DIGIT [0-7]
+HEX_DIGIT {DEC_DIGIT}|[a-fA-F]
+
+IDENTIFIER {IDENTIFIER_FIRST_CHAR}({DEC_DIGIT}|{IDENTIFIER_FIRST_CHAR})*
+IDENTIFIER_FIRST_CHAR [A-Za-z_]
+
+LINE_COMMENT "//"{NONEOL_CHAR}*
+BLOCK_COMMENT_INITIATOR "/*"
+BLOCK_COMMENT_TERMINATOR "*/"
+
+KEYWORD _|abstract|alignof|as|become|box|break|const|continue|crate|do|else|enum|extern|false|final|fn|for|if|impl|in|let|loop|macro|match|mod|move|mut|offsetof|override|priv|proc|pub|pure|ref|return|Self|self|sizeof|static|struct|super|trait|true|type|typeof|unsafe|unsized|use|virtual|where|while|yield
+
+ESCAPE {COMMON_ESCAPE}|{UNICODE_ESCAPE}
+COMMON_ESCAPE \\|n|r|t|0|x{HEX_DIGIT}{2}
+UNICODE_ESCAPE "u{"{HEX_DIGIT}{6}"}"
+
+CHAR_LITERAL '{CHAR_BODY}'
+CHAR_BODY [^']|(\\('|{ESCAPE}))
+STRING_LITERAL \"{STRING_BODY}*\"
+STRING_BODY [^\"]|(\\(\"|{ESCAPE}))
+BYTE_CHAR_LITERAL b'{BYTE_CHAR_BODY}'
+BYTE_CHAR_BODY [^']|(\\('|{COMMON_ESCAPE}))
+BYTE_STRING_LITERAL b\"{BYTE_STRING_BODY}*\"
+BYTE_STRING_BODY [^\"]|(\\(\"|{ESCAPE}))
+
+NUMBER_LITERAL {NONZERO_DEC_DIGIT}{DEC_DIGIT}*{FLOAT_SUFFIX}?|(0{DEC_DIGIT_LITERAL}*{FLOAT_SUFFIX}?|b[10_]+|o{OCT_DIGIT_LITERAL}+|x{HEX_DIGIT_LITERAL}+)
+FLOAT_SUFFIX {EXPONENT}|"."{DEC_DIGIT_LITERAL}+{EXPONENT}?
+EXPONENT [Ee][-+]?{DEC_DIGIT_LITERAL}+
+DEC_DIGIT_LITERAL {DEC_DIGIT}|_
+HEX_DIGIT_LITERAL {HEX_DIGIT}|_
+OCT_DIGIT_LITERAL {OCT_DIGIT}|_
+
+SYMBOL "::"|"->"|"#"|"["|"]"|"("|")"|"{"|"}"|","|";"|":"|"?"|"!"|"."|"'"|"$"|"@"|"~"|"=="|"=>"|"="|"!="|"<="|"<<"|"<<="|"<"|">="|">>"|">>="|">"|"<-"|"-"|"-="|"&&"|"&"|"&="|"||"|"|"|"|="|"+"|"+="|"*"|"*="|"/"|"/="|"^"|"^="|"%"|"%="|".."|"..."
+
+%%
+
+{SYMBOL} {
+ PRINT_HIGHLIGHTED_TOKEN(symbol);
+}
+
+{KEYWORD} {
+ PRINT_HIGHLIGHTED_TOKEN(keyword);
+}
+
+{IDENTIFIER} {
+ PRINT_HIGHLIGHTED_TOKEN(identifier);
+}
+
+{NUMBER_LITERAL} {
+ PRINT_HIGHLIGHTED_TOKEN(number);
+}
+
+{CHAR_LITERAL}|{STRING_LITERAL}|{BYTE_CHAR_LITERAL}|{BYTE_STRING_LITERAL} {
+ PRINT_HIGHLIGHTED_TOKEN(string);
+}
+
+{LINE_COMMENT} {
+ PRINT_HIGHLIGHTED_TOKEN(comment);
+}
+
+{BLOCK_COMMENT_INITIATOR} {
+ printf("<span class=\"comment\">%s", yytext);
+}
+
+{BLOCK_COMMENT_TERMINATOR} {
+ printf("%s</span>", yytext);
+}
+
+. {
+ printf("%s", yytext); /* echo the rest */
+}
+
+%%
+
+int main(int argc, const char* const argv[]) {
+ yyin = argc > 1 ? fopen(argv[1], "r") : stdin;
+
+ printf(
+ "<!DOCTYPE html>\n"
+ "<html>\n"
+ "\n"
+ "<head>\n"
+ " <style>\n"
+ " .comment {\n"
+ " color: gray;\n"
+ " }\n"
+ "\n"
+ " .keyword {\n"
+ " color: red;\n"
+ " }\n"
+ "\n"
+ " .string {\n"
+ " color: green;\n"
+ " }\n"
+ "\n"
+ " .number {\n"
+ " color: blue;\n"
+ " }\n"
+ "\n"
+ " .symbol {\n"
+ " font-weight: bold;\n"
+ " }\n"
+ " </style>\n"
+ "</head>\n"
+ "\n"
+ "<body>\n"
+ " <pre class=\"code\">\n"
+ );
+ yylex();
+ printf(
+ "\n"
+ " </pre>\n"
+ "</body>\n"
+ "\n"
+ "</html>"
+ );
+
+ return 0;
+}