Cleanup and formatting (#17)

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name    = "flaco"
-version = "0.4.2"
+version = "0.5.0"
 edition = "2018"
 license = "Unlicense/MIT"
 

flaco/__init__.py

@@ -1,2 +1,2 @@
-__version__ = "0.4.2"
+__version__ = "0.5.0"
 from .io import read_sql, Database

flaco/io.pyx

@@ -66,41 +66,37 @@ cpdef dict read_sql(str stmt, Database db, int n_rows=-1, int size_hint=-1):
         lib.RowDataArrayPtr row_data_ptr
         lib.Data *data
 
-    while True:
-        if row_iterator == NULL:
-            break
-        else:
-
-            if row_idx == 0:
-                # Initialize arrays for output
-                # will resize at `n_increment` if `n_rows` is not set.
+    while row_iterator != NULL:
+        if row_idx == 0:
+            # Initialize arrays for output
+            # will resize at `n_increment` if `n_rows` is not set.
+            for i in range(0, n_columns):
+                data = lib.index_row(row_data_array_ptr, n_columns, i)
+                output.append(
+                    array_init(deref(data), n_increment if n_rows == -1 else n_rows)
+                )
+
+        # grow arrays if next insert is passed current len
+        if _n_rows == -1 and current_array_len <= row_idx:
                 for i in range(0, n_columns):
-                    data = lib.index_row(row_data_array_ptr, n_columns, i)
-                    output.append(
-                        array_init(deref(data), n_increment if n_rows == -1 else n_rows)
-                    )
+                    resize(output[i], current_array_len + n_increment)
+                current_array_len += n_increment
 
-            # grow arrays if next insert is passed current len
-            if _n_rows == -1 and current_array_len <= row_idx:
-                    for i in range(0, n_columns):
-                        resize(output[i], current_array_len + n_increment)
-                    current_array_len += n_increment
+        for i in range(0, n_columns):
+            data = lib.index_row(row_data_array_ptr, n_columns, i)
+            output[i] = insert_data_into_array(deref(data), output[i], row_idx)
 
-            for i in range(0, n_columns):
-                data = lib.index_row(row_data_array_ptr, n_columns, i)
-                output[i] = insert_data_into_array(deref(data), output[i], row_idx)
-
-            row_idx += one
-
-            lib.next_row(
-                &row_iterator,
-                &row_data_array_ptr,
-                &n_columns,
-                &column_names,
-                &exc
-            )
-            if exc != NULL:
-                raise FlacoException(exc.decode())
+        row_idx += one
+
+        lib.next_row(
+            &row_iterator,
+            &row_data_array_ptr,
+            &n_columns,
+            &column_names,
+            &exc
+        )
+        if exc != NULL:
+            raise FlacoException(exc.decode())
 
     # Ensure arrays are correct size; only if n_rows not set
     if _n_rows == -1 and current_array_len != row_idx:

setup.py

@@ -13,18 +13,14 @@
     libraries.extend(["ntdll", "ws2_32", "bcrypt", "advapi32"])
     flaco_lib_file = flaco_lib_dir.joinpath("flaco.lib")
     extra_link_args = []
-    extra_compile_args = [
-        "/link",
-        "/SUBSYSTEM:WINDOWS",
-    ]
+    extra_compile_args = ["/link", "/SUBSYSTEM:WINDOWS"]
 else:
     flaco_lib_file = flaco_lib_dir.joinpath("libflaco.a")
     extra_link_args = ["-l:libflaco.a"]
     extra_compile_args = ["-fopenmp", "-O3"]
 
 assert flaco_lib_file.is_file(), "flaco lib not built; run 'cargo build --release'"
 
-
 extension = Extension(
     name="*",
     sources=[str(pathlib.Path("flaco/*.pyx"))],
@@ -51,7 +47,7 @@
 
 setup(
     name="flaco",
-    version="0.4.2",
+    version="0.5.0",
     author="Miles Granger",
     author_email="miles59923@gmail.com",
     description="Fast and Efficient PostgreSQL data into numpy/pandas",

src/lib.rs

@@ -1,14 +1,14 @@
 //#![warn(missing_docs)]
 use postgres as pg;
 use postgres::fallible_iterator::FallibleIterator;
+use postgres::types::{FromSql, Type};
 use postgres::RowIter;
 use rust_decimal::prelude::{Decimal, ToPrimitive};
+use std::error::Error;
 use std::ffi::CString;
 use std::net::IpAddr;
 use std::os::raw::c_char;
 use std::{ffi, mem};
-use std::error::Error;
-use postgres::types::{FromSql, Type};
 use time;
 
 type DatabasePtr = *mut u32;
@@ -148,7 +148,7 @@ impl From<CString> for StringPtr {
 #[repr(C)]
 pub struct DateInfo {
     /// The value represents the number of days since January 1st, 2000.
-    offset: i32
+    offset: i32,
 }
 
 impl FromSql<'_> for DateInfo {
@@ -189,7 +189,7 @@ impl FromSql<'_> for TimeInfo {
 #[repr(C)]
 pub struct DateTimeInfo {
     /// The value represents the number of microseconds since midnight, January 1st, 2000.
-    offset: i64
+    offset: i64,
 }
 impl FromSql<'_> for DateTimeInfo {
     fn from_sql(_: &Type, raw: &[u8]) -> std::result::Result<Self, Box<dyn Error + Sync + Send>> {
@@ -248,7 +248,6 @@ simple_from!(TimeInfo, Time);
 simple_from!(BytesPtr, Bytes);
 simple_from!(StringPtr, String);
 
-
 #[no_mangle]
 pub extern "C" fn free_db(ptr: DatabasePtr) {
     unsafe { Box::from_raw(ptr as DatabasePtr) };
@@ -325,28 +324,13 @@ fn row_data(row: pg::Row, array_ptr: &mut RowDataArrayPtr) -> Result<()> {
         // TODO: postgres-types: expose Inner enum which these variations
         // and have postgres Row.type/(or something) expose the variant
         let value: Data = match type_.name() {
-            "bytea" => row
-                .get::<_, Option<Vec<u8>>>(i)
-                .map(BytesPtr::from)
-                .into(),
-            "char" => row
-                .get::<_, Option<i8>>(i)
-                .into(),
-            "smallint" | "smallserial" | "int2" => row
-                .get::<_, Option<i16>>(i)
-                .into(),
-            "oid" => row
-                .get::<_, Option<u32>>(i)
-                .into(),
-            "int4" | "int" | "serial" => row
-                .get::<_, Option<i32>>(i)
-                .into(),
-            "bigint" | "int8" | "bigserial" => row
-                .get::<_, Option<i64>>(i)
-                .into(),
-            "bool" => row
-                .get::<_, Option<bool>>(i)
-                .into(),
+            "bytea" => row.get::<_, Option<Vec<u8>>>(i).map(BytesPtr::from).into(),
+            "char" => row.get::<_, Option<i8>>(i).into(),
+            "smallint" | "smallserial" | "int2" => row.get::<_, Option<i16>>(i).into(),
+            "oid" => row.get::<_, Option<u32>>(i).into(),
+            "int4" | "int" | "serial" => row.get::<_, Option<i32>>(i).into(),
+            "bigint" | "int8" | "bigserial" => row.get::<_, Option<i64>>(i).into(),
+            "bool" => row.get::<_, Option<bool>>(i).into(),
             "double precision" | "float8" => row
                 .get::<_, Option<f64>>(i)
                 .or_else(|| Some(f64::NAN))
@@ -360,15 +344,11 @@ fn row_data(row: pg::Row, array_ptr: &mut RowDataArrayPtr) -> Result<()> {
                 .map(|v| CString::new(v).unwrap())
                 .map(StringPtr::from)
                 .into(),
-            "timestamp" | "timestamp with time zone" | "timestamptz" => row
-                .get::<_, Option<DateTimeInfo>>(i)
-                .into(),
-            "date" => row
-                .get::<_, Option<DateInfo>>(i)
-                .into(),
-            "time" => row
-                .get::<_, Option<TimeInfo>>(i)
-                .into(),
+            "timestamp" | "timestamp with time zone" | "timestamptz" => {
+                row.get::<_, Option<DateTimeInfo>>(i).into()
+            }
+            "date" => row.get::<_, Option<DateInfo>>(i).into(),
+            "time" => row.get::<_, Option<TimeInfo>>(i).into(),
             "json" | "jsonb" => row
                 .get::<_, Option<serde_json::Value>>(i)
                 .map(|v| CString::new(v.to_string()).unwrap())