From 4be9b785135b287ce7605db9772fd082cddbe5c6 Mon Sep 17 00:00:00 2001
From: Hesam Pakdaman <hesam.pakdaman@gmail.com>
Date: Sat, 25 May 2024 22:16:29 +0200
Subject: [PATCH] Fix 337: Look ahead for b' ' if skip_initial_space is enabled

---
 csv-core/src/reader.rs | 44 ++++++++++++++++++++++++++++++++++++++++++
 src/reader.rs          |  9 +++++++++
 2 files changed, 53 insertions(+)

diff --git a/csv-core/src/reader.rs b/csv-core/src/reader.rs
index dbd6dc3..563abd6 100644
--- a/csv-core/src/reader.rs
+++ b/csv-core/src/reader.rs
@@ -126,6 +126,8 @@ pub struct Reader {
     has_read: bool,
     /// The current position in the output buffer when reading a record.
     output_pos: usize,
+    /// Skip any space following the delimiter
+    skip_initial_space: bool,
 }
 
 impl Default for Reader {
@@ -145,6 +147,7 @@ impl Default for Reader {
             line: 1,
             has_read: false,
             output_pos: 0,
+            skip_initial_space: false,
         }
     }
 }
@@ -256,6 +259,15 @@ impl ReaderBuilder {
         self.rdr.use_nfa = yes;
         self
     }
+
+    /// Enable or disable initial space.
+    ///
+    /// When enabled skip any space following the field delimiter when
+    /// parsing CSV.
+    pub fn skip_initial_space(&mut self, yes: bool) -> &mut ReaderBuilder {
+        self.rdr.skip_initial_space = yes;
+        self
+    }
 }
 
 /// The result of parsing at most one field from CSV data.
@@ -671,6 +683,12 @@ impl Reader {
             if has_out {
                 output[nout] = input[nin];
                 nout += 1;
+            } else if self.skip_initial_space
+                && nin < input.len() - 1
+                && input[nin] == self.delimiter
+                && input[nin + 1] == b' '
+            {
+                nin += 1;
             }
             nin += 1;
             if state >= self.dfa.final_field {
@@ -731,6 +749,12 @@ impl Reader {
             if has_out {
                 output[nout] = b;
                 nout += 1;
+            } else if self.skip_initial_space
+                && nin < input.len() - 1
+                && input[nin] == self.delimiter
+                && input[nin + 1] == b' '
+            {
+                nin += 1;
             }
             nin += 1;
             if state >= self.dfa.final_field {
@@ -893,6 +917,12 @@ impl Reader {
                     nin += 1;
                 }
                 NfaInputAction::Discard => {
+                    if self.skip_initial_space
+                        && nin < input.len() - 1
+                        && input[nin + 1] == b' '
+                    {
+                        nin += 1;
+                    }
                     nin += 1;
                 }
                 NfaInputAction::Epsilon => {}
@@ -944,6 +974,12 @@ impl Reader {
                     nin += 1;
                 }
                 NfaInputAction::Discard => {
+                    if self.skip_initial_space
+                        && nin < input.len() - 1
+                        && input[nin + 1] == b' '
+                    {
+                        nin += 1;
+                    }
                     nin += 1;
                 }
                 NfaInputAction::Epsilon => (),
@@ -1725,6 +1761,14 @@ mod tests {
             b.comment(Some(b'#'));
         }
     );
+    parses_to!(
+        issue_337,
+        "a, \"b\", c",
+        csv![["a", "b", "c"]],
+        |b: &mut ReaderBuilder| {
+            b.skip_initial_space(true);
+        }
+    );
 
     macro_rules! assert_read {
         (
diff --git a/src/reader.rs b/src/reader.rs
index 82f5d2c..c6509dd 100644
--- a/src/reader.rs
+++ b/src/reader.rs
@@ -630,6 +630,15 @@ impl ReaderBuilder {
         self.builder.nfa(yes);
         self
     }
+
+    /// Enable or disable initial space.
+    ///
+    /// When enabled skip any space following the field delimiter when
+    /// parsing CSV.
+    pub fn skip_initial_space(&mut self, yes: bool) -> &mut ReaderBuilder {
+        self.builder.skip_initial_space(yes);
+        self
+    }
 }
 
 /// A already configured CSV reader.