From 141c0ae35d70363d0b36d9a6eb9df41846bde0e8 Mon Sep 17 00:00:00 2001
From: Joel Natividad <1980690+jqnatividad@users.noreply.github.com>
Date: Sun, 28 Jan 2024 09:50:29 -0500
Subject: [PATCH] `joinp`: automatically set tab delimiter when reading TSV/TAB
 input files

---
 src/cmd/joinp.rs | 22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/src/cmd/joinp.rs b/src/cmd/joinp.rs
index 46efa0b20..b2891f0ed 100644
--- a/src/cmd/joinp.rs
+++ b/src/cmd/joinp.rs
@@ -162,7 +162,7 @@ joinp options:
 Common options:
     -h, --help             Display this message
     -o, --output <file>    Write output to <file> instead of stdout.
-    -d, --delimiter <arg>  The field delimiter for reading CSV data.
+    -d, --delimiter <arg>  The field delimiter for reading/writing CSV data.
                            Must be a single character. (default: ,)
     -Q, --quiet            Do not return join shape to stderr.
 "#;
@@ -520,7 +520,7 @@ impl Args {
                 .has_header(true)
                 .with_missing_is_null(self.flag_nulls)
                 .with_comment_prefix(comment_char.as_deref())
-                .with_separator(delim)
+                .with_separator(tsvtab_delim(&self.arg_input1, delim))
                 .with_infer_schema_length(num_rows)
                 .with_try_parse_dates(try_parsedates)
                 .low_memory(low_memory)
@@ -545,7 +545,7 @@ impl Args {
                 .has_header(true)
                 .with_missing_is_null(self.flag_nulls)
                 .with_comment_prefix(comment_char.as_deref())
-                .with_separator(delim)
+                .with_separator(tsvtab_delim(&self.arg_input2, delim))
                 .with_infer_schema_length(num_rows)
                 .with_try_parse_dates(try_parsedates)
                 .low_memory(low_memory)
@@ -580,3 +580,19 @@ impl Args {
         })
     }
 }
+
+fn tsvtab_delim(file: &str, orig_delim: u8) -> u8 {
+    // if the file has a TSV or TAB extension, we automatically use tab as the delimiter
+    let inputfile_extension = Path::new(file)
+        .extension()
+        .and_then(std::ffi::OsStr::to_str)
+        .unwrap_or_default();
+
+    if inputfile_extension.eq_ignore_ascii_case("tsv")
+        || inputfile_extension.eq_ignore_ascii_case("tab")
+    {
+        b'\t'
+    } else {
+        orig_delim
+    }
+}