From 7b48f7ac9ecad9e6d174592814533b79b925b495 Mon Sep 17 00:00:00 2001
From: joe-conigliaro <joe-conigliaro@users.noreply.github.com>
Date: Wed, 14 Aug 2019 16:45:56 +1000
Subject: [PATCH] encoding.csv module

---
 vlib/encoding/csv/csv_test.v |  33 ++++++++
 vlib/encoding/csv/reader.v   | 144 +++++++++++++++++++++++++++++++++++
 2 files changed, 177 insertions(+)
 create mode 100644 vlib/encoding/csv/csv_test.v
 create mode 100644 vlib/encoding/csv/reader.v

diff --git a/vlib/encoding/csv/csv_test.v b/vlib/encoding/csv/csv_test.v
new file mode 100644
index 0000000000..7ecf29424a
--- /dev/null
+++ b/vlib/encoding/csv/csv_test.v
@@ -0,0 +1,33 @@
+import encoding.csv
+
+fn test_encoding_csv() {
+	// test reading
+	data := 'name,email,phone,other\njoe,joe@blow.com,0400000000,test\nsam,sam@likesham.com,0433000000,"test quoted field"\n#chris,chris@nomail.com,94444444,"commented row"\nmike,mike@mikesbikes.com,98888888,"bike store"\n'
+	mut csv_reader := csv.new_reader(data)
+
+	mut row_count := 0
+	for {
+		row := csv_reader.read() or {
+			break
+		}
+		row_count++
+		if row_count== 1 {
+			assert row[0] == 'name'
+		}
+		if row_count == 2 {
+			assert row[0] == 'joe'
+		}
+		if row_count == 3 {
+			assert row[0] == 'sam'
+			// quoted field
+			assert row[3] == 'test quoted field'
+		}
+		if row_count == 4 {
+			assert row[0] == 'mike'
+		}
+	}
+
+	assert row_count == 4
+
+	// test writing to come
+}
diff --git a/vlib/encoding/csv/reader.v b/vlib/encoding/csv/reader.v
new file mode 100644
index 0000000000..6ee69124ec
--- /dev/null
+++ b/vlib/encoding/csv/reader.v
@@ -0,0 +1,144 @@
+module csv
+
+// Copyright (c) 2019 Alexander Medvednikov. All rights reserved.
+// Use of this source code is governed by an MIT license
+// that can be found in the LICENSE file.
+
+// Once interfaces are further along the idea would be to have something similar to
+// go's io.reader & bufio.reader rather than reading the whole file into string, this
+// would then satisfy that interface. I designed it this way to be easily adapted.
+
+const (
+	err_comment_is_delim = error('encoding.csv: comment cannot be the same as delimiter')
+	err_invalid_delim    = error('encoding.csv: invalid delimiter')
+	err_eof              = error('encoding.csv: end of file')
+	err_invalid_le       = error('encoding.csv: could not find any valid line endings.')
+)
+
+
+struct Reader  {
+	// not used yet
+	// has_header        bool
+	// headings          []string
+	data              string
+mut:
+	delimiter         byte
+	comment           byte
+	is_mac_pre_osx_le bool
+	row_pos           int
+}
+
+pub fn new_reader(data string) *Reader {
+	return &Reader{
+		delimiter: `,`,
+		comment: `#`,
+		data: data
+	}
+}
+
+// read() reads one row from the csv file
+pub fn (r mut Reader) read() ?[]string {
+	l := r.read_record() or {
+		return error(err)
+	}
+	return l
+}
+
+// Once we have multi dimensional array
+// pub fn (r mut Reader) read_all() ?[][]string {
+// 	mut records := []string
+// 	for {
+// 		record := r.read_record() or {
+// 			if error(err).error == err_eof.error {
+// 				return records
+// 			} else {
+// 				return error(err)
+// 			}
+// 		}
+// 		records << record
+// 	}
+// 	return records
+// }
+
+fn (r mut Reader) read_line() ?string {
+	// last record
+	if r.row_pos == r.data.len {
+		return err_eof
+	}
+	le := if r.is_mac_pre_osx_le { '\r' } else { '\n' }
+	mut i := r.data.index_after(le, r.row_pos)
+	if i == -1 {
+		if r.row_pos == 0 {
+			// check for pre osx mac line endings
+			i = r.data.index_after('\r', r.row_pos)
+			if i != -1 {
+				r.is_mac_pre_osx_le = true
+			} else {
+				// no valid line endings found
+				return err_invalid_le
+			}
+		}
+	}
+	mut line := r.data.substr(r.row_pos, i)
+	r.row_pos = i+1
+	// normalize win line endings (remove extra \r)
+	if !r.is_mac_pre_osx_le && (line.len >= 1 && line[line.len-1] == `\r`) {
+		line = line.left(line.len-1)
+	}
+	return line
+}
+
+fn (r mut Reader) read_record() ?[]string {
+	if r.delimiter == r.comment {
+		return err_comment_is_delim
+	}
+	mut line := ''
+	for {
+		l := r.read_line() or {
+			return error(err)
+		}
+		line = l
+		// skip commented lines
+		if line[0] == r.comment {
+			continue
+		}
+		break
+	}
+	mut fields := []string
+	mut i := -1
+	for {
+		// not quoted
+		if line[0] != `"` {
+			i = line.index(r.delimiter.str())
+			if i == -1 {
+				// last
+				break
+			}
+			fields << line.left(i)
+			line = line.right(i+1)
+			continue
+		}
+		// quoted
+		else {
+			line = line.right(1)
+			i = line.index('"')
+			if i+1 == line.len {
+				// last record
+				fields << line.left(i)
+				break
+			}
+			next := line[i+1]
+			if next == r.delimiter {
+				fields << line.left(i)
+				line = line.right(i)
+				continue
+			}
+			line = line.right(1)
+		}
+		if i <= -1 && fields.len == 0 {
+			return err_invalid_delim
+		}
+	}
+	
+	return fields
+}