dedup_csv 0.1.0-x86_64-linux-musl → 0.1.1-x86_64-linux-musl

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3fd679aa79b8440e5b65df12e5691e3dc7bc28cbc5ec7854b5000f9d42ba9b22
-  data.tar.gz: 5d53342cf426de5305da821025ed27e5b9017dcf4127671877377f14fde9bd79
+  metadata.gz: c8d96c1554006e43d6327794ec8ffa67730660d50afba12addcca2e72551f1fc
+  data.tar.gz: 609cf2a07320a051bfbfb46868d4f3d268ee7ee3765ce2834a7c7a64e00887ad
 SHA512:
-  metadata.gz: cae9dbfcc4814989551c2208bbf1d305bedbf1d8ab7d7ba985ea24cbe8697c385edbea613168227d1b99bf4c79542089b70665c4ca15d76c1cf3067c21034d10
-  data.tar.gz: 923ed9468fee9785b6b6bb98a7b314bd9776dbd8a1045db4daab81ecd3bb3d8226e2a9fb2044779ecbea80424eedb0f9ef26d14748e6262b7795a72c40c06166
+  metadata.gz: 80fece252c1ddd8701fdb0fa8e070dff2545a82ae49e4d65e7d7c0754cc6fc07b973a5f8c0d8d654c99dedd9207d6be6ad4e68f80e35b44f4bfbda5804b61acd
+  data.tar.gz: e78e7dbffaae44de88742c093ca33a374cad402aed48369941a1cb2bfe380af48f9ca5c30d1b36fde7141af23c3d8a420dc2c22544e4c37cbf63a277d0bcf650

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    dedup_csv (0.1.0)
+    dedup_csv (0.1.1)
 
 GEM
   remote: https://rubygems.org/
@@ -67,4 +67,4 @@ DEPENDENCIES
   rubocop (~> 1.21)
 
 BUNDLED WITH
-   2.4.10
+   2.4.4

data/README.md

@@ -2,6 +2,19 @@
 
 Given 2 CSV files, this gem will create a third CSV file that contains rows from the first CSV file that are not present in the second CSV file.
 
+## Installation
+
+```bash
+gem install dedup_csv
+```
+
+## Usage
+
+```irb
+require 'dedup_csv/3.2/dedup_csv'
+DedupCsv.dedup('file1.csv', 'file2.csv', 'output.csv')
+```
+
 ## Contributing
 
 Bug reports and pull requests are welcome on GitHub at https://github.com/kingsleyh/dedup_csv.

data/Rakefile

@@ -14,11 +14,11 @@ require 'rb_sys/extensiontask'
 task build: :compile
 
 spec = Bundler.load_gemspec('dedup_csv.gemspec')
-spec.requirements.clear
-spec.required_ruby_version = nil
-spec.required_rubygems_version = nil
-spec.extensions.clear
-spec.files -= Dir['ext/**/*']
+# spec.requirements.clear
+# spec.required_ruby_version = nil
+# spec.required_rubygems_version = nil
+# spec.extensions.clear
+# spec.files -= Dir['ext/**/*']
 
 Rake::ExtensionTask.new('dedup_csv', spec) do |c|
   c.lib_dir = 'lib/dedup_csv'

data/ext/dedup_csv/Cargo.toml

@@ -0,0 +1,14 @@
+[package]
+name = "dedup_csv"
+version = "0.1.0"
+edition = "2021"
+authors = ["kingsley.hendrickse <kingsley.hendrickse@patchwork.health>"]
+publish = false
+
+[lib]
+crate-type = ["cdylib"]
+
+[dependencies]
+magnus = { version = "0.7.1" }
+csv = "1.3.0"
+eyre = "0.6.12"

data/ext/dedup_csv/extconf.rb

@@ -0,0 +1,6 @@
+# frozen_string_literal: true
+
+require 'mkmf'
+require 'rb_sys/mkmf'
+
+create_rust_makefile('dedup_csv/dedup_csv')

data/ext/dedup_csv/src/lib.rs

@@ -0,0 +1,77 @@
+use std::error::Error;
+use std::ffi::OsStr;
+use std::fs::File;
+use std::path::Path;
+use csv::{StringRecord, Writer};
+use magnus::{define_module, function, prelude::*, Ruby};
+
+fn dedup(ruby: &Ruby, previous_csv_path: String, new_csv_path: String, target_path: String) -> magnus::error::Result<()> {
+    if !previous_csv_path.has_extension(&["csv"]) {
+        return Err(magnus::Error::new(ruby.exception_exception(), "previous_csv_path must be a csv file".to_string()));
+    }
+    if !new_csv_path.has_extension(&["csv"]) {
+        return Err(magnus::Error::new(ruby.exception_exception(), "new_csv_path must be a csv file".to_string()));
+    }
+
+    let csv1 = File::open(previous_csv_path).map_err(|e| magnus_err(ruby, e, "previous_csv_path"))?;
+    let csv2 = File::open(new_csv_path).map_err(|e| magnus_err(ruby, e, "new_csv_path"))?;
+
+    let mut previous_csv: csv::Reader<File> = csv::Reader::from_reader(csv1);
+    let mut new_csv: csv::Reader<File> = csv::Reader::from_reader(csv2);
+
+    let mut wtr = Writer::from_path(target_path).map_err(|e| magnus_err(ruby, e, "target_path"))?;
+
+    let previous_headers = previous_csv.headers().map_err(|e| magnus_err(ruby, e, "previous_csv_path headers"))?;
+    let new_headers = new_csv.headers().map_err(|e| magnus_err(ruby, e, "new_csv_path headers"))?;
+
+    if previous_headers != new_headers {
+        return Err(magnus::Error::new(ruby.exception_exception(), "headers of both csv files must be the same".to_string()));
+    }
+
+    wtr.write_byte_record(previous_headers.as_byte_record()).unwrap();
+
+    let mut previous_records = vec![];
+    for previous_record in previous_csv.records() {
+        let previous_record = previous_record.map_err(|e| magnus_err(ruby, e, "previous_record"))?;
+        previous_records.push(previous_record)
+    }
+
+    for new_record in new_csv.records() {
+        let new_record = new_record.map_err(|e| magnus_err(ruby, e, "new_record"))?;
+        if !previous_records.contains(&new_record) {
+            let new_record = new_record.into_iter().map(|r| r.trim_end()).collect::<StringRecord>();
+            wtr.write_byte_record(new_record.as_byte_record()).unwrap();
+        }
+    }
+
+    wtr.flush().unwrap();
+
+    Ok(())
+}
+
+fn magnus_err<E: Error>(ruby: &Ruby, e: E, msg: &str) -> magnus::Error {
+    magnus::Error::new(ruby.exception_exception(), format!("{}: {}", msg, e.to_string()))
+}
+
+#[magnus::init]
+fn init() -> Result<(), magnus::Error> {
+    let module = define_module("DedupCsv")?;
+    module.define_singleton_method("dedup", function!(dedup, 3))?;
+    Ok(())
+}
+
+pub trait FileExtension {
+    fn has_extension<S: AsRef<str>>(&self, extensions: &[S]) -> bool;
+}
+
+impl<P: AsRef<Path>> FileExtension for P {
+    fn has_extension<S: AsRef<str>>(&self, extensions: &[S]) -> bool {
+        if let Some(ref extension) = self.as_ref().extension().and_then(OsStr::to_str) {
+            return extensions
+                .iter()
+                .any(|x| x.as_ref().eq_ignore_ascii_case(extension));
+        }
+
+        false
+    }
+}

data/lib/dedup_csv/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module DedupCsv
-  VERSION = '0.1.0'
+  VERSION = '0.1.1'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: dedup_csv
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: x86_64-linux-musl
 authors:
 - kingsley.hendrickse
@@ -25,6 +25,9 @@ files:
 - Gemfile.lock
 - README.md
 - Rakefile
+- ext/dedup_csv/Cargo.toml
+- ext/dedup_csv/extconf.rb
+- ext/dedup_csv/src/lib.rs
 - lib/dedup_csv.rb
 - lib/dedup_csv/2.7/dedup_csv.so
 - lib/dedup_csv/3.0/dedup_csv.so
@@ -51,6 +54,9 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 3.3.dev
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: 3.0.0
   - - ">="
     - !ruby/object:Gem::Version
       version: 3.3.22