From 04ea2824d317f8f358c027f62a74d1b40d5d3b41 Mon Sep 17 00:00:00 2001
From: Sebastian Schicho <64368773+schicho@users.noreply.github.com>
Date: Fri, 21 May 2021 12:18:08 +0200
Subject: [PATCH] os: implement File.read_bytes_into_newline method for reading
 into a buffer, line by line (#10129)

---
 vlib/os/file.c.v    | 40 +++++++++++++++++++++++++++++
 vlib/os/file_test.v | 61 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 101 insertions(+)

diff --git a/vlib/os/file.c.v b/vlib/os/file.c.v
index 6f2af20bb1..beedbfb28c 100644
--- a/vlib/os/file.c.v
+++ b/vlib/os/file.c.v
@@ -17,6 +17,8 @@ fn C.fseeko(voidptr, u64, int) int
 
 fn C._fseeki64(voidptr, u64, int) int
 
+fn C.getc(voidptr) int
+
 // open_file can be used to open or create a file with custom flags and permissions and returns a `File` object.
 pub fn open_file(path string, mode string, options ...int) ?File {
 	mut flags := 0
@@ -362,6 +364,44 @@ pub fn (f &File) read_bytes_at(size int, pos u64) []byte {
 	return arr[0..nreadbytes]
 }
 
+// read_bytes_into_newline reads from the beginning of the file into the provided buffer.
+// Each consecutive call on the same file continues reading where it previously ended.
+// A read call is either stopped, if the buffer is full, a newline was read or EOF.
+pub fn (f &File) read_bytes_into_newline(mut buf []byte) ?int {
+	if buf.len == 0 {
+		panic(@FN + ': `buf.len` == 0')
+	}
+	newline := 10
+	mut c := 0
+	mut buf_ptr := 0
+	mut nbytes := 0
+
+	for (buf_ptr < buf.len) {
+		c = C.getc(f.cfile)
+		match c {
+			C.EOF {
+				if C.feof(f.cfile) != 0 {
+					return nbytes
+				}
+				if C.ferror(f.cfile) != 0 {
+					return error('file read error')
+				}
+			}
+			newline {
+				buf[buf_ptr] = byte(c)
+				nbytes++
+				return nbytes
+			}
+			else {
+				buf[buf_ptr] = byte(c)
+				buf_ptr++
+				nbytes++
+			}
+		}
+	}
+	return nbytes
+}
+
 // read_bytes_into fills `buf` with bytes at the given position in the file.
 // `buf` *must* have length greater than zero.
 // Returns the number of read bytes, or an error.
diff --git a/vlib/os/file_test.v b/vlib/os/file_test.v
index c8ab607c7d..6144ad1aa1 100644
--- a/vlib/os/file_test.v
+++ b/vlib/os/file_test.v
@@ -59,6 +59,67 @@ fn testsuite_end() ? {
 	assert !os.is_dir(tfolder)
 }
 
+// test_read_bytes_into_newline_text tests reading text from a file with newlines.
+// This test simulates reading a larger text file step by step into a buffer and
+// returning on each newline, even before the buffer is full, and reaching EOF before
+// the buffer is completely filled.
+fn test_read_bytes_into_newline_text() ? {
+	mut f := os.open_file(tfile, 'w') ?
+	f.write_string('Hello World!\nGood\r morning.') ?
+	f.close()
+
+	f = os.open_file(tfile, 'r') ?
+	mut buf := []byte{len: 8}
+
+	n0 := f.read_bytes_into_newline(mut buf) ?
+	assert n0 == 8
+
+	n1 := f.read_bytes_into_newline(mut buf) ?
+	assert n1 == 5
+
+	n2 := f.read_bytes_into_newline(mut buf) ?
+	assert n2 == 8
+
+	n3 := f.read_bytes_into_newline(mut buf) ?
+	assert n3 == 6
+
+	f.close()
+}
+
+// test_read_bytes_into_newline_binary tests reading a binary file with NUL bytes.
+// This test simulates the scenario when a byte stream is read and a newline byte
+// appears in that stream and an EOF occurs before the buffer is full.
+fn test_read_bytes_into_newline_binary() ? {
+	os.rm(tfile) or {} // FIXME This is a workaround for macos, because the file isn't truncated when open with 'w'
+	mut bw := []byte{len: 15}
+	bw[9] = 0xff
+	bw[12] = 10 // newline
+
+	n0_bytes := bw[0..10]
+	n1_bytes := bw[10..13]
+	n2_bytes := bw[13..]
+
+	mut f := os.open_file(tfile, 'w') ?
+	f.write(bw) ?
+	f.close()
+
+	f = os.open_file(tfile, 'r') ?
+	mut buf := []byte{len: 10}
+
+	n0 := f.read_bytes_into_newline(mut buf) ?
+	assert n0 == 10
+	assert buf[..n0] == n0_bytes
+
+	n1 := f.read_bytes_into_newline(mut buf) ?
+	assert n1 == 3
+	assert buf[..n1] == n1_bytes
+
+	n2 := f.read_bytes_into_newline(mut buf) ?
+	assert n2 == 2
+	assert buf[..n2] == n2_bytes
+	f.close()
+}
+
 // test_read_eof_last_read_partial_buffer_fill tests that when reading a file
 // the end-of-file is detected and results in a none error being returned. This
 // test simulates file reading where the end-of-file is reached inside an fread