Many API cleanups.

Fix a future-incompatible warning in some `io_util` macros due to trailing
semicolons.

Also fix those macros to work properly when expanded in contexts where
the types they're accessing aren't in scope, and allow them to work with
`impl AsRef<Path>` types rather than just `&String`.

Add trait implementations where appropriate, mostly `Default` but also a
few for `From` where it looked like it made sense given existing
conversion functions, in order to bring the APIs into conformance with
common naming conventions.

Fix a bunch of clippy lints around stuff that looks like
```rust
if f.is_err() {
    // thing
} else {
    let f = f.unwrap();
    // other thing
}
```
using `match` instead to avoid checking more than once whether the type
was an error.

Replace uses of `&Vec<T>` with `&[T]` where possible.  In some cases,
the extent to which this was possible is limited until some of the
upstream crates (from this repository) push a new version with their
APIs updated.

Replace uses of `&String` with either `&str` or `impl AsRef<Path>`
as appropriate for the use case.  Again, as with the `&Vec<T>` changes,
in some cases this means temporarily going through fallible steps
allocating String objects until the upstream crates publish versions
with updated APIs.

I could be wrong but I believe that `cargo clippy --fix` will be able
to fix the TODOs from the above two items automatically when the time
comes.

The above two changes _shouldn't_ affect existing users, at least in
most cases (`&Vec<T>` casts implicitly to `&[T]`, `&String` casts
implicitly to `&str`, and both `&String` and `&str` implement
`AsRef<Path>`).  At least, in theory.

Change the `segtype` field of `vdj_ann::refx::RefData` from a vector of
owned `String` to a `&'static str`, since everything added to that
vector is a constant value.  This should significantly reduce memory
consumption, as the vector is now a vector of pointers into constant
memory rather than a vector of pointers each into their own heap-
allocated copy of the source strings.  Because that field is public,
this is a breaking API change (though I don't know how many downstream
users actually access the field directly), so I'm bumping the version
of that crate.

In vdj_ann_ref, don't convert `&'static str` to `String` except in
cases where (similarly to above) we have to in order to satisfy the old
APIs.  Also, change the if-chaining logic to `match` statements to make
the different cases more clear.

Author: adam-azarchs

align_tools/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "align_tools"
-version = "0.1.9"
+version = "0.1.10"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

align_tools/src/lib.rs

@@ -198,7 +198,7 @@ pub fn affine_align(x: &DnaString, y: &DnaString) -> Alignment {
 // are used and would need to be tweaked if other operations are present.  You can set width to
 // the expected terminal width.
 
-pub fn vis_align(s1: &[u8], s2: &[u8], ops: &Vec<AlignmentOperation>, width: usize) -> String {
+pub fn vis_align(s1: &[u8], s2: &[u8], ops: &[AlignmentOperation], width: usize) -> String {
     let (mut pos1, mut pos2) = (0, 0);
     let (mut t1, mut t2) = (Vec::<u8>::new(), Vec::<u8>::new());
     let mut d = Vec::<u8>::new();

ansi_escape/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "ansi_escape"
-version = "0.1.2"
+version = "0.1.3"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

ansi_escape/src/ansi_to_html.rs

@@ -55,10 +55,11 @@ pub fn convert_text_with_ansi_escapes_to_svg(
 
     let lines0 = x.split('\n').collect::<Vec<&str>>();
     let height = vsep * (lines0.len() as f64 - 1.2);
-    let mut lines = Vec::<String>::new();
-    lines.push("<svg version=\"1.1\"".to_string());
-    lines.push("".to_string()); // PLACEHOLDER
-    lines.push("xmlns=\"http://www.w3.org/2000/svg\">".to_string());
+    let mut lines = vec![
+        "<svg version=\"1.1\"".to_string(),
+        String::default(), // PLACEHOLDER
+        "xmlns=\"http://www.w3.org/2000/svg\">".to_string(),
+    ];
     let mut max_width = 0;
     for m in 0..lines0.len() {
         let t = &lines0[m];
@@ -211,8 +212,8 @@ pub fn compress_ansi_escapes(x: &str) -> String {
                             end = Some(i);
                         }
                     }
-                    if end.is_some() {
-                        escapes = escapes[end.unwrap() + 1..escapes.len()].to_vec();
+                    if let Some(end) = end {
+                        escapes = escapes[end + 1..escapes.len()].to_vec();
                     }
                     if escapes.is_empty() {
                         // Emit end escape.
@@ -431,7 +432,7 @@ impl ColorState {
 
 // ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓
 
-fn merge(s: &Vec<ColorState>) -> ColorState {
+fn merge(s: &[ColorState]) -> ColorState {
     let mut x = ColorState::default();
     for i in 0..s.len() {
         if s[i].null() {

binary_vec_io/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "binary_vec_io"
-version = "0.1.11"
+version = "0.1.12"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

binary_vec_io/src/lib.rs

@@ -76,7 +76,7 @@ pub fn binary_read_to_ref<T>(f: &mut std::fs::File, p: &mut T, n: usize) -> Resu
 // in the first case, or to a vector, in the second case.
 
 #[allow(dead_code)]
-pub fn binary_write_vec<T>(f: &mut std::fs::File, x: &Vec<T>) -> Result<(), Error>
+pub fn binary_write_vec<T>(f: &mut std::fs::File, x: &[T]) -> Result<(), Error>
 where
     T: BinaryInputOutputSafe,
 {
@@ -112,7 +112,7 @@ where
     binary_read_to_ref::<T>(f, &mut x[len], n)
 }
 
-pub fn binary_write_vec_vec<T>(f: &mut std::fs::File, x: &Vec<Vec<T>>) -> Result<(), Error>
+pub fn binary_write_vec_vec<T>(f: &mut std::fs::File, x: &[Vec<T>]) -> Result<(), Error>
 where
     T: BinaryInputOutputSafe,
 {

dna/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "dna"
-version = "0.1.2"
+version = "0.1.3"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

dna/src/lib.rs

@@ -44,19 +44,16 @@ pub fn tm_nearest_neighbor(s: &str) -> f64 {
     tm_nearest_neighbor_full(s, 0.00000025, 0.05, &locked)
 }
 
-pub fn tm_nearest_neighbor_full(s: &str, s_mol: f64, na_mol: f64, locked: &Vec<bool>) -> f64 {
+pub fn tm_nearest_neighbor_full(s: &str, s_mol: f64, na_mol: f64, locked: &[bool]) -> f64 {
     // Allow for + symbols.
 
     if s.contains('+') {
         assert_eq!(locked.len(), 0);
         assert!(!s.contains("++"));
         assert!(!s.ends_with('+'));
         let mut sx = String::new();
-        let mut lockedx = Vec::<bool>::new();
-        let mut schars = Vec::<char>::new();
-        for c in s.chars() {
-            schars.push(c);
-        }
+        let schars: Vec<char> = s.chars().collect();
+        let mut lockedx = Vec::<bool>::with_capacity(schars.len());
         let mut i = 0;
         while i < schars.len() {
             if schars[i] != '+' {
@@ -365,7 +362,7 @@ pub fn thermodynamic_sums_dna(
     dg_sum: &mut f64,
     include_symmetry_correction: bool,
     include_initiation_terms: bool,
-    locked: &Vec<bool>,
+    locked: &[bool],
 ) {
     // defaults for last: true, true, empty
 

exons/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "exons"
-version = "0.1.4"
+version = "0.1.5"
 authors = ["David Jaffe <[email protected]"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

exons/src/lib.rs

@@ -14,7 +14,7 @@ use std::{
 use string_utils::TextUtils;
 use vector_utils::unique_sort;
 
-pub fn fetch_exons(species: &String, exons: &mut Vec<(String, i32, i32, bool, String, i32)>) {
+pub fn fetch_exons(species: &str, exons: &mut Vec<(String, i32, i32, bool, String, i32)>) {
     assert!(species == "human" || species == "mouse");
 
     // Define gtf file location.  See notes in bin/build_vdj_ref.fs.

expr_tools/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "expr_tools"
-version = "0.1.1"
+version = "0.1.2"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."

expr_tools/src/lib.rs

@@ -8,7 +8,7 @@
 use evalexpr::{ContextWithMutableFunctions, ContextWithMutableVariables, HashMapContext};
 use evalexpr::{Function, Value};
 use statrs::distribution::ContinuousCDF;
-use string_utils::*;
+use string_utils::TextUtils;
 use vector_utils::{bin_member, unique_sort};
 
 // ================================================================================================
@@ -124,7 +124,7 @@ macro_rules! evalexpr_fn3 {
 // to be called on arbitrary strings, but if the strings are not all f64, then the return value is
 // null.
 
-pub fn define_evalexpr_context(vars: &Vec<String>, vals: &Vec<String>) -> evalexpr::HashMapContext {
+pub fn define_evalexpr_context(vars: &[String], vals: &[String]) -> evalexpr::HashMapContext {
     assert_eq!(vars.len(), vals.len());
     let mut c = HashMapContext::new();
 

fasta_tools/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "fasta_tools"
-version = "0.1.5"
+version = "0.1.6"
 authors = ["David Jaffe <[email protected]>"]
 license = "MIT"
 description = "Some tools that are 'internal' for now because they are insufficiently refined and unstable, but which are used by other 'public' crates."