Moved CSV and JSON testing away from using files. (#290)

.github/workflows/test.yml

@@ -159,11 +159,11 @@ jobs:
         env:
           RUST_BACKTRACE: full
           RUST_LOG: 'trace'
-        # --skip io: miri does not handle IO very well, unfortunately.
+        # --skip io: miri can't handle opening of files, so we skip those
         run: |
           cargo miri setup
           cargo clean
-          cargo miri test -- --skip io
+          cargo miri test -- --skip io::parquet --skip io::ipc
 
   coverage:
     name: Coverage

src/io/json/read/infer_schema.rs

@@ -170,20 +170,18 @@ fn generate_schema(spec: HashMap<String, HashSet<DataType>>) -> Result<Schema> {
 ///
 /// # Examples
 /// ```
-/// use std::fs::File;
-/// use std::io::{BufReader, SeekFrom, Seek};
-/// use flate2::read::GzDecoder;
+/// use std::io::{BufReader, Cursor, SeekFrom, Seek};
 /// use arrow2::io::json::infer_json_schema;
 ///
-/// let mut file = File::open("test/data/mixed_arrays.json.gz").unwrap();
+/// let data = r#"{"a":1, "b":[2.0, 1.3, -6.1], "c":[false, true], "d":4.1}
+/// {"a":-10, "b":[2.0, 1.3, -6.1], "c":null, "d":null}
+/// {"a":2, "b":[2.0, null, -6.1], "c":[false, null], "d":"text"}
+/// {"a":3, "b":4, "c": true, "d":[1, false, "array", 2.4]}
+/// "#;
 ///
 /// // file's cursor's offset at 0
-/// let mut reader = BufReader::new(GzDecoder::new(&file));
+/// let mut reader = BufReader::new(Cursor::new(data));
 /// let inferred_schema = infer_json_schema(&mut reader, None).unwrap();
-/// // cursor's offset at end of file
-///
-/// // seek back to start so that the original file is usable again
-/// file.seek(SeekFrom::Start(0)).unwrap();
 /// ```
 pub fn infer_json_schema<R: Read>(
     reader: &mut BufReader<R>,
@@ -345,14 +343,17 @@ where
 /// # Examples
 /// ```
 /// use std::fs::File;
-/// use std::io::BufReader;
+/// use std::io::{BufReader, Cursor};
 /// use arrow2::io::json::infer_json_schema_from_seekable;
 ///
-/// let file = File::open("test/data/mixed_arrays.json").unwrap();
-/// // file's cursor's offset at 0
-/// let mut reader = BufReader::new(file);
+/// let data = r#"{"a":1, "b":[2.0, 1.3, -6.1], "c":[false, true], "d":4.1}
+/// {"a":-10, "b":[2.0, 1.3, -6.1], "c":null, "d":null}
+/// {"a":2, "b":[2.0, null, -6.1], "c":[false, null], "d":"text"}
+/// {"a":3, "b":4, "c": true, "d":[1, false, "array", 2.4]}
+/// "#;
+/// let mut reader = BufReader::new(Cursor::new(data));
 /// let inferred_schema = infer_json_schema_from_seekable(&mut reader, None).unwrap();
-/// // file's cursor's offset automatically set at 0
+/// // cursor's position automatically set at 0
 /// ```
 pub fn infer_json_schema_from_seekable<R: Read + Seek>(
     reader: &mut BufReader<R>,

src/io/json/read/reader.rs

@@ -133,19 +133,21 @@ impl Decoder {
 /// use std::sync::Arc;
 /// use arrow2::datatypes::{DataType, Field, Schema};
 /// use arrow2::io::json;
-/// use std::fs::File;
-/// use std::io::BufReader;
+/// use std::io::{Cursor, BufReader};
 ///
 /// let schema = Arc::new(Schema::new(vec![
-///     Field::new("a", DataType::Float64, false),
-///     Field::new("b", DataType::Float64, false),
-///     Field::new("c", DataType::Float64, false),
+///     Field::new("a", DataType::Int64, true),
+///     Field::new("b", DataType::Float32, true),
+///     Field::new("c", DataType::Boolean, true),
+///     Field::new("d", DataType::Utf8, true),
 /// ]));
 ///
-/// let file = File::open("test/data/basic.json").unwrap();
-///
-/// let mut json = json::Reader::new(BufReader::new(file), schema, 1024, None);
-/// let batch = json.next().unwrap().unwrap();
+/// let data = r#"{"a":1, "b":2.0, "c":false, "d":"4"}
+/// {"a":-10, "b":-3.5, "c":true, "d":null}
+/// {"a":100000000, "b":0.6, "d":"text"}"#;
+/// let mut reader = BufReader::new(Cursor::new(data));
+/// let mut reader = json::Reader::new(&mut reader, schema, 1024, None);
+/// let batch = reader.next().unwrap().unwrap();
 /// ```
 #[derive(Debug)]
 pub struct Reader<R: Read> {