polish: more polishing work

Touching up docs, restyling code in places, etc.

.github/workflows/ci.yml

@@ -31,7 +31,7 @@ jobs:
         include:
         - build: pinned
           os: ubuntu-18.04
-          rust: 1.39.0
+          rust: 1.40.0
         - build: stable
           os: ubuntu-18.04
           rust: stable

Cargo.toml

@@ -26,6 +26,7 @@ levenshtein = ["utf8-ranges"]
 utf8-ranges = { version = "1.0.4", optional = true }
 
 [dev-dependencies]
+doc-comment = "0.3.1"
 fnv = "1.0.6"
 memmap = "0.7"
 quickcheck = { version = "0.9.2", default-features = false }

README.md

@@ -23,11 +23,10 @@ Dual-licensed under MIT or the [UNLICENSE](http://unlicense.org).
 https://docs.rs/fst
 
 The
-[`fst-regex`](https://docs.rs/fst-regex)
-and
-[`fst-levenshtein`](https://docs.rs/fst-levenshtein)
-crates provide regular expression matching and fuzzy searching on FSTs,
-respectively.
+[`regex-automata`](https://docs.rs/regex-automata)
+crate provides implementations of the `fst::Automata` trait when its
+`transducer` feature is enabled. This permits using DFAs compiled by
+`regex-automata` to search finite state transducers produced by this crate.
 
 
 ### Installation
@@ -36,27 +35,21 @@ Simply add a corresponding entry to your `Cargo.toml` dependency list:
 
 ```toml,ignore
 [dependencies]
-fst = "0.3"
+fst = "0.4"
 ```
 
 
 ### Example
 
 This example demonstrates building a set in memory and executing a fuzzy query
-against it. You'll need `fst = "0.3"` and `fst-levenshtein = "0.2"` in your
-`Cargo.toml`.
+against it. You'll need `fst = "0.4"` with the `levenshtein` feature enabled in
+your `Cargo.toml`.
 
 ```rust
-extern crate fst;
-extern crate fst_levenshtein;
-
-use std::error::Error;
-use std::process;
-
 use fst::{IntoStreamer, Set};
-use fst_levenshtein::Levenshtein;
+use fst::automaton::Levenshtein;
 
-fn try_main() -> Result<(), Box<Error>> {
+fn main() -> Result<(), Box<dyn std::error::Error>> {
   // A convenient way to create sets in memory.
   let keys = vec!["fa", "fo", "fob", "focus", "foo", "food", "foul"];
   let set = Set::from_iter(keys)?;
@@ -71,13 +64,13 @@ fn try_main() -> Result<(), Box<Error>> {
   assert_eq!(keys, vec!["fo", "fob", "foo", "food"]);
   Ok(())
 }
-
-fn main() {
-  if let Err(err) = try_main() {
-    eprintln!("{}", err);
-    process::exit(1);
-  }
-}
 ```
 
 Check out the documentation for a lot more examples!
+
+
+### Cargo features
+
+* `levenshtein` - **Disabled** by default. This adds the `Levenshtein`
+  automaton to the `automaton` sub-module. This includes an additional
+  dependency on `utf8-ranges`.

fst-bin/src/merge.rs

@@ -32,7 +32,7 @@ impl<I> Merger<I>
 where
     I: Iterator<Item = Result<(BString, u64), Error>> + Send + 'static,
 {
-    pub fn new<T, P>(it: T, output: P) -> Self
+    pub fn new<T, P>(it: T, output: P) -> Merger<I>
     where
         P: AsRef<Path>,
         T: IntoIterator<IntoIter = I, Item = I::Item>,
@@ -49,35 +49,35 @@ where
         }
     }
 
-    pub fn value_merger<F>(mut self, f: F) -> Self
+    pub fn value_merger<F>(mut self, f: F) -> Merger<I>
     where
         F: Fn(u64, u64) -> u64 + Send + Sync + 'static,
     {
         self.value_merger = Some(Arc::new(f));
         self
     }
 
-    pub fn fd_limit(mut self, fd_limit: u32) -> Self {
+    pub fn fd_limit(mut self, fd_limit: u32) -> Merger<I> {
         self.fd_limit = fd_limit;
         self
     }
 
-    pub fn batch_size(mut self, batch_size: u32) -> Self {
+    pub fn batch_size(mut self, batch_size: u32) -> Merger<I> {
         self.batch_size = batch_size;
         self
     }
 
-    pub fn threads(mut self, threads: u32) -> Self {
+    pub fn threads(mut self, threads: u32) -> Merger<I> {
         self.threads = threads;
         self
     }
 
-    pub fn tmp_dir<P: AsRef<Path>>(mut self, path: P) -> Self {
+    pub fn tmp_dir<P: AsRef<Path>>(mut self, path: P) -> Merger<I> {
         self.tmp_dir = path.as_ref().to_path_buf();
         self
     }
 
-    pub fn keep_tmp_dir(mut self, yes: bool) -> Self {
+    pub fn keep_tmp_dir(mut self, yes: bool) -> Merger<I> {
         self.keep_tmp_dir = yes;
         self
     }
@@ -117,7 +117,7 @@ where
             for (i, union_batch) in batches.into_iter().enumerate() {
                 sorters.create_fst(UnionBatch {
                     tmp_dir: tmp_dir_path.clone(),
-                    gen: gen,
+                    gen,
                     index: i,
                     fsts: union_batch?,
                     value_merger: self.value_merger.clone(),
@@ -178,7 +178,7 @@ struct Sorters<B> {
 }
 
 impl<B: Batchable + Send + 'static> Sorters<B> {
-    fn new(threads: u32) -> Self {
+    fn new(threads: u32) -> Sorters<B> {
         let (bsend, brecv) = chan::bounded::<B>(0);
         let (rsend, rrecv) = chan::bounded(0);
         for _ in 0..threads {

fst-bin/src/util.rs

@@ -103,14 +103,14 @@ type Lines = bstr::io::ByteLines<
 impl ConcatLines {
     pub fn new(mut inputs: Vec<PathBuf>) -> ConcatLines {
         inputs.reverse(); // treat it as a stack
-        ConcatLines { inputs: inputs, cur: None }
+        ConcatLines { inputs, cur: None }
     }
 }
 
 impl Iterator for ConcatLines {
     type Item = io::Result<BString>;
 
-    fn next(&mut self) -> Option<Self::Item> {
+    fn next(&mut self) -> Option<io::Result<BString>> {
         loop {
             if self.cur.is_none() {
                 match self.inputs.pop() {
@@ -143,7 +143,7 @@ type Rows = csv::DeserializeRecordsIntoIter<Reader, (BString, u64)>;
 impl ConcatCsv {
     pub fn new(mut inputs: Vec<PathBuf>) -> ConcatCsv {
         inputs.reverse(); // treat it as a stack
-        ConcatCsv { inputs: inputs, cur: None }
+        ConcatCsv { inputs, cur: None }
     }
 
     fn read_row(&mut self) -> Option<Result<(BString, u64), Error>> {
@@ -162,7 +162,7 @@ impl ConcatCsv {
 impl Iterator for ConcatCsv {
     type Item = Result<(BString, u64), Error>;
 
-    fn next(&mut self) -> Option<Self::Item> {
+    fn next(&mut self) -> Option<Result<(BString, u64), Error>> {
         loop {
             if self.cur.is_none() {
                 match self.inputs.pop() {

fst-levenshtein/src/lib.rs

@@ -1,6 +1,3 @@
-
-
-
 use std::cmp;
 use std::collections::hash_map::Entry;
 use std::collections::{HashMap, HashSet};
@@ -90,13 +87,13 @@ impl Levenshtein {
     /// A `Levenshtein` value satisfies the `Automaton` trait, which means it
     /// can be used with the `search` method of any finite state transducer.
     #[inline]
-    pub fn new(query: &str, distance: u32) -> Result<Self, Error> {
+    pub fn new(query: &str, distance: u32) -> Result<Levenshtein, Error> {
         let lev = DynamicLevenshtein {
             query: query.to_owned(),
             dist: distance as usize,
         };
         let dfa = DfaBuilder::new(lev.clone()).build()?;
-        Ok(Levenshtein { prog: lev, dfa: dfa })
+        Ok(Levenshtein { prog: lev, dfa })
     }
 }
 
@@ -197,10 +194,10 @@ struct DfaBuilder {
 }
 
 impl DfaBuilder {
-    fn new(lev: DynamicLevenshtein) -> Self {
+    fn new(lev: DynamicLevenshtein) -> DfaBuilder {
         DfaBuilder {
             dfa: Dfa { states: Vec::with_capacity(16) },
-            lev: lev,
+            lev,
             cache: HashMap::with_capacity(1024),
         }
     }
@@ -251,9 +248,7 @@ impl DfaBuilder {
             Entry::Occupied(v) => (*v.get(), true),
             Entry::Vacant(v) => {
                 let is_match = self.lev.is_match(lev_state);
-                self.dfa
-                    .states
-                    .push(State { next: [None; 256], is_match: is_match });
+                self.dfa.states.push(State { next: [None; 256], is_match });
                 (*v.insert(self.dfa.states.len() - 1), false)
             }
         })
@@ -314,7 +309,7 @@ impl DfaBuilder {
     }
 
     fn new_state(&mut self, is_match: bool) -> usize {
-        self.dfa.states.push(State { next: [None; 256], is_match: is_match });
+        self.dfa.states.push(State { next: [None; 256], is_match });
         self.dfa.states.len() - 1
     }
 }

src/automaton/levenshtein.rs

@@ -109,7 +109,10 @@ impl Levenshtein {
     /// A `Levenshtein` value satisfies the `Automaton` trait, which means it
     /// can be used with the `search` method of any finite state transducer.
     #[inline]
-    pub fn new(query: &str, distance: u32) -> Result<Self, LevenshteinError> {
+    pub fn new(
+        query: &str,
+        distance: u32,
+    ) -> Result<Levenshtein, LevenshteinError> {
         let lev = DynamicLevenshtein {
             query: query.to_owned(),
             dist: distance as usize,
@@ -216,7 +219,7 @@ struct DfaBuilder {
 }
 
 impl DfaBuilder {
-    fn new(lev: DynamicLevenshtein) -> Self {
+    fn new(lev: DynamicLevenshtein) -> DfaBuilder {
         DfaBuilder {
             dfa: Dfa { states: Vec::with_capacity(16) },
             lev,