From 7e612ebf7a66614f0dea8114f907cc5a85a235f0 Mon Sep 17 00:00:00 2001
From: BENJAMIN PEINHARDT <benjaminpeinhardt@gmail.com>
Date: Fri, 5 Jan 2024 19:41:56 -0600
Subject: [PATCH] fix comma bug

---
 CHANGELOG.md               |  3 +++
 gleam.toml                 |  2 +-
 src/gsv.gleam              | 25 +++++++++++++++++--------
 src/gsv/internal/ast.gleam | 22 +++++++++++++++-------
 test/gsv_test.gleam        | 12 +++++-------
 5 files changed, 41 insertions(+), 23 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index a3fbdd1..97dfe27 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,9 @@
 
 ## Unreleased
 
+## v1.2.1 - 5 January 2024
+- Fix bug with double commas producing error, now produce empty string
+
 ## v1.2.0 - 29 December 2023
 - Add `to_lists_or_error` function
 
diff --git a/gleam.toml b/gleam.toml
index b12ae8f..267793a 100644
--- a/gleam.toml
+++ b/gleam.toml
@@ -1,5 +1,5 @@
 name = "gsv"
-version = "1.2.0"
+version = "1.2.1"
 gleam = ">= 0.32.0"
 description = "A simple csv parser and generator written in gleam "
 
diff --git a/src/gsv.gleam b/src/gsv.gleam
index 1e1ce51..297a819 100644
--- a/src/gsv.gleam
+++ b/src/gsv.gleam
@@ -29,9 +29,13 @@ pub fn to_lists_or_panic(input: String) -> List(List(String)) {
     Ok(lol) -> lol
     Error(ParseError(Location(line, column), msg)) -> {
       panic as {
-        "[" <> "line " <> int.to_string(line) <> " column " <> int.to_string(
-          column,
-        ) <> "] of csv: " <> msg
+        "["
+        <> "line "
+        <> int.to_string(line)
+        <> " column "
+        <> int.to_string(column)
+        <> "] of csv: "
+        <> msg
       }
       [[]]
     }
@@ -48,7 +52,13 @@ pub fn to_lists_or_error(input: String) -> Result(List(List(String)), String) {
   |> ast.parse
   |> result.map_error(fn(e) {
     let ParseError(Location(line, column), msg) = e
-    "[" <> "line " <> int.to_string(line) <> " column " <> int.to_string(column) <> "] of csv: " <> msg
+    "["
+    <> "line "
+    <> int.to_string(line)
+    <> " column "
+    <> int.to_string(column)
+    <> "] of csv: "
+    <> msg
   })
 }
 
@@ -85,10 +95,9 @@ pub fn from_lists(
 
       // If the string contains a , \n \r\n or " it needs to be escaped by wrapping in double quotes
       case
-        string.contains(entry, separator) || string.contains(entry, "\n") || string.contains(
-          entry,
-          "\"",
-        )
+        string.contains(entry, separator)
+        || string.contains(entry, "\n")
+        || string.contains(entry, "\"")
       {
         True -> "\"" <> entry <> "\""
         False -> entry
diff --git a/src/gsv/internal/ast.gleam b/src/gsv/internal/ast.gleam
index 2c99ff3..7a526a6 100644
--- a/src/gsv/internal/ast.gleam
+++ b/src/gsv/internal/ast.gleam
@@ -90,7 +90,8 @@ fn parse_p(
         "Expected \"\\n\" after \"\\r\", found: " <> token.to_lexeme(tok),
       ))
 
-    // If we just parsed a comma, we're expecting an Escaped or Non-Escaped string
+    // If we just parsed a comma, we're expecting an Escaped or Non-Escaped string, or another comma
+    // (indicating an empty string)
     [#(Textdata(str), _), ..remaining_tokens], JustParsedComma, [
       curr_line,
       ..previously_parsed_lines
@@ -109,12 +110,20 @@ fn parse_p(
         ..previously_parsed_lines
       ])
 
+    [#(Comma, _), ..remaining_tokens], JustParsedComma, [
+      curr_line,
+      ..previously_parsed_lines
+    ] ->
+      parse_p(remaining_tokens, JustParsedComma, [
+        ["", ..curr_line],
+        ..previously_parsed_lines
+      ])
+
     [#(tok, loc), ..], JustParsedComma, _ ->
       Error(ParseError(
         loc,
-        "Expected escaped or non-escaped string after comma, found: " <> token.to_lexeme(
-          tok,
-        ),
+        "Expected escaped or non-escaped string after comma, found: "
+          <> token.to_lexeme(tok),
       ))
 
     // If we just parsed a new line, we're expecting an escaped or non-escaped string
@@ -133,9 +142,8 @@ fn parse_p(
     [#(tok, loc), ..], JustParsedNewline, _ ->
       Error(ParseError(
         loc,
-        "Expected escaped or non-escaped string after newline, found: " <> token.to_lexeme(
-          tok,
-        ),
+        "Expected escaped or non-escaped string after newline, found: "
+          <> token.to_lexeme(tok),
       ))
 
     // If we're inside an escaped string, we can take anything until we get a double quote,
diff --git a/test/gsv_test.gleam b/test/gsv_test.gleam
index 0b04f7f..ea1042c 100644
--- a/test/gsv_test.gleam
+++ b/test/gsv_test.gleam
@@ -159,15 +159,15 @@ pub fn error_cases_test() {
     }
   }
 
-  produce_error("Ben, 25,, TRUE")
+  produce_error("Ben, 25,\n, TRUE")
   |> should.equal(#(
     Location(1, 9),
-    "Expected escaped or non-escaped string after comma, found: ,",
+    "Expected escaped or non-escaped string after comma, found: \n",
   ))
-  produce_error("Austin, 25, FALSE\n\"Ben Peinhardt\", 25,, TRUE")
+  produce_error("Austin, 25, FALSE\n\"Ben Peinhardt\", 25,\n, TRUE")
   |> should.equal(#(
     Location(2, 21),
-    "Expected escaped or non-escaped string after comma, found: ,",
+    "Expected escaped or non-escaped string after comma, found: \n",
   ))
 }
 
@@ -178,7 +178,5 @@ pub fn error_cases_test() {
 pub fn totally_errors_test() {
   "Ben, 25,, TRUE"
   |> gsv.to_lists_or_error
-  |> should.equal(Error(
-    "[line 1 column 9] of csv: Expected escaped or non-escaped string after comma, found: ,",
-  ))
+  |> should.equal(Ok([["Ben", " 25", "", " TRUE"]]))
 }