1.4.0

Author: bcpeinhardt

CHANGELOG.md

@@ -2,6 +2,9 @@
 
 ## Unreleased
 
+## v1.4.0 - 29 March 2024
+- Fix bug where trailing comma was causing error
+
 ## v1.3.1 - 1 February 2024
 - Update to gleam_stdlib = "~> 0.34 or ~> 1.0" in preparation for 1.0
 

gleam.toml

@@ -1,5 +1,5 @@
 name = "gsv"
-version = "1.3.1"
+version = "1.4.0"
 gleam = ">= 0.32.0"
 description = "A simple csv parser and generator written in gleam "
 

src/gsv.gleam

@@ -1,9 +1,9 @@
-import gsv/internal/ast.{ParseError}
-import gsv/internal/token.{Location}
+import gleam/int
 import gleam/list
-import gleam/string
 import gleam/result
-import gleam/int
+import gleam/string
+import gsv/internal/ast.{ParseError}
+import gsv/internal/token.{Location}
 
 /// Parses a csv string to a list of lists of strings.
 /// Automatically handles Windows and Unix line endings.

src/gsv/internal/ast.gleam

@@ -92,33 +92,46 @@ fn parse_p(
 
     // 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
-    ] ->
+    [#(Textdata(str), _), ..remaining_tokens],
+      JustParsedComma,
+      [curr_line, ..previously_parsed_lines] ->
       parse_p(remaining_tokens, JustParsedField, [
         [str, ..curr_line],
         ..previously_parsed_lines
       ])
 
-    [#(Doublequote, _), ..remaining_tokens], JustParsedComma, [
-      curr_line,
-      ..previously_parsed_lines
-    ] ->
+    [#(Doublequote, _), ..remaining_tokens],
+      JustParsedComma,
+      [curr_line, ..previously_parsed_lines] ->
       parse_p(remaining_tokens, InsideEscapedString, [
         ["", ..curr_line],
         ..previously_parsed_lines
       ])
 
-    [#(Comma, _), ..remaining_tokens], JustParsedComma, [
-      curr_line,
-      ..previously_parsed_lines
-    ] ->
+    [#(Comma, _), ..remaining_tokens],
+      JustParsedComma,
+      [curr_line, ..previously_parsed_lines] ->
       parse_p(remaining_tokens, JustParsedComma, [
         ["", ..curr_line],
         ..previously_parsed_lines
       ])
 
+    [#(CR, _), ..remaining_tokens],
+      JustParsedComma,
+      [curr_line, ..previously_parsed_lines] ->
+      parse_p(remaining_tokens, JustParsedCR, [
+        ["", ..curr_line],
+        ..previously_parsed_lines
+      ])
+
+    [#(LF, _), ..remaining_tokens],
+      JustParsedComma,
+      [curr_line, ..previously_parsed_lines] ->
+      parse_p(remaining_tokens, JustParsedNewline, [
+        ["", ..curr_line],
+        ..previously_parsed_lines
+      ])
+
     [#(tok, loc), ..], JustParsedComma, _ ->
       Error(ParseError(
         loc,
@@ -130,10 +143,9 @@ fn parse_p(
     [#(Textdata(str), _), ..remaining_tokens], JustParsedNewline, llf ->
       parse_p(remaining_tokens, JustParsedField, [[str], ..llf])
 
-    [#(Doublequote, _), ..remaining_tokens], JustParsedNewline, [
-      curr_line,
-      ..previously_parsed_lines
-    ] ->
+    [#(Doublequote, _), ..remaining_tokens],
+      JustParsedNewline,
+      [curr_line, ..previously_parsed_lines] ->
       parse_p(remaining_tokens, InsideEscapedString, [
         ["", ..curr_line],
         ..previously_parsed_lines
@@ -148,10 +160,9 @@ fn parse_p(
 
     // If we're inside an escaped string, we can take anything until we get a double quote,
     // but a double double quote "" escapes the double quote and we keep parsing
-    [#(Doublequote, _), #(Doublequote, _), ..remaining_tokens], InsideEscapedString, [
-      [str, ..rest_curr_line],
-      ..previously_parsed_lines
-    ] ->
+    [#(Doublequote, _), #(Doublequote, _), ..remaining_tokens],
+      InsideEscapedString,
+      [[str, ..rest_curr_line], ..previously_parsed_lines] ->
       parse_p(remaining_tokens, InsideEscapedString, [
         [str <> "\"", ..rest_curr_line],
         ..previously_parsed_lines
@@ -160,10 +171,9 @@ fn parse_p(
     [#(Doublequote, _), ..remaining_tokens], InsideEscapedString, llf ->
       parse_p(remaining_tokens, JustParsedField, llf)
 
-    [#(other_token, _), ..remaining_tokens], InsideEscapedString, [
-      [str, ..rest_curr_line],
-      ..previously_parsed_lines
-    ] ->
+    [#(other_token, _), ..remaining_tokens],
+      InsideEscapedString,
+      [[str, ..rest_curr_line], ..previously_parsed_lines] ->
       parse_p(remaining_tokens, InsideEscapedString, [
         [str <> token.to_lexeme(other_token), ..rest_curr_line],
         ..previously_parsed_lines

src/gsv/internal/token.gleam

@@ -8,8 +8,8 @@
 ////  escaped = DQUOTE *(TEXTDATA / COMMA / CR / LF / 2DQUOTE) DQUOTE
 ////  non-escaped = *TEXTDATA
 
-import gleam/string
 import gleam/list
+import gleam/string
 
 pub type CsvToken {
   Comma

test/gsv_test.gleam

@@ -1,14 +1,14 @@
+import gleam/int
+import gleam/list
+import gleam/result
+import gleam/string
 import gleeunit
 import gleeunit/should
+import gsv.{Unix, Windows}
+import gsv/internal/ast.{ParseError, parse}
 import gsv/internal/token.{
   CR, Comma, Doublequote, LF, Location, Textdata, scan, with_location,
 }
-import gsv/internal/ast.{ParseError, parse}
-import gsv.{Unix, Windows}
-import gleam/list
-import gleam/result
-import gleam/int
-import gleam/string
 
 pub fn main() {
   gleeunit.main()
@@ -161,13 +161,13 @@ pub fn error_cases_test() {
 
   produce_error("Ben, 25,\n, TRUE")
   |> should.equal(#(
-    Location(1, 9),
-    "Expected escaped or non-escaped string after comma, found: \n",
+    Location(2, 1),
+    "Expected escaped or non-escaped string after newline, found: ,",
   ))
   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: \n",
+    Location(3, 1),
+    "Expected escaped or non-escaped string after newline, found: ,",
   ))
 }
 
@@ -180,3 +180,11 @@ pub fn totally_errors_test() {
   |> gsv.to_lists_or_error
   |> should.equal(Ok([["Ben", " 25", "", " TRUE"]]))
 }
+
+pub fn trailing_commas_fine_test() {
+  "Ben, 25, TRUE, Hello\nAustin, 25,\n"
+  |> gsv.to_lists
+  |> should.equal(
+    Ok([["Ben", " 25", " TRUE", " Hello"], ["Austin", " 25", ""]]),
+  )
+}