Adds support for tabs and semicolons as delimiters

mailmerge/__main__.py

@@ -255,25 +255,47 @@ def create_sample_input_files(template_path, database_path, config_path):
     """))
 
 
+def detect_database_format(database_file):
+    """Automatically detect the database format.
+
+    Automatically detect the format ("dialect") using the CSV library's sniffer
+    class.  For example, comma-delimited, tab-delimited, etc.  Default to
+    StrictExcel if automatic detection fails.
+
+    """
+    class StrictExcel(csv.excel):
+        # Our helper class is really simple
+        # pylint: disable=too-few-public-methods, missing-class-docstring
+        strict = True
+
+    # Read a sample from database
+    sample = database_file.read(1024)
+    database_file.seek(0)
+
+    # Attempt automatic format detection, fall back on StrictExcel default
+    try:
+        csvdialect = csv.Sniffer().sniff(sample, delimiters=",;\t")
+    except csv.Error:
+        csvdialect = StrictExcel
+
+    return csvdialect
+
+
 def read_csv_database(database_path):
     """Read database CSV file, providing one line at a time.
 
-    We'll use a class to modify the csv library's default dialect ('excel') to
-    enable strict syntax checking.  This will trigger errors for things like
+    Use strict syntax checking, which will trigger errors for things like
     unclosed quotes.
 
     We open the file with the utf-8-sig encoding, which skips a byte order mark
     (BOM), if any.  Sometimes Excel will save CSV files with a BOM.  See Issue
     #93 https://github.com/awdeorio/mailmerge/issues/93
 
     """
-    class StrictExcel(csv.excel):
-        # Our helper class is really simple
-        # pylint: disable=too-few-public-methods, missing-class-docstring
-        strict = True
-
     with database_path.open(encoding="utf-8-sig") as database_file:
-        reader = csv.DictReader(database_file, dialect=StrictExcel)
+        csvdialect = detect_database_format(database_file)
+        csvdialect.strict = True
+        reader = csv.DictReader(database_file, dialect=csvdialect)
         try:
             for row in reader:
                 yield row

tests/test_main.py

@@ -859,3 +859,109 @@ def test_database_bom(tmpdir):
         >>> Limit was 1 message.  To remove the limit, use the --no-limit option.
         >>> This was a dry run.  To send messages, use the --no-dry-run option.
     """)  # noqa: E501
+
+
+def test_database_tsv(tmpdir):
+    """Automatically detect TSV database format."""
+    # Simple template
+    template_path = Path(tmpdir/"mailmerge_template.txt")
+    template_path.write_text(textwrap.dedent("""\
+        TO: {{email}}
+        FROM: My Self <myself@mydomain.com>
+
+        Hello {{name}}
+    """), encoding="utf8")
+
+    # Tab-separated format database
+    database_path = Path(tmpdir/"mailmerge_database.csv")
+    database_path.write_text(textwrap.dedent("""\
+        email\tname
+        to@test.com\tMy Name
+    """), encoding="utf8")
+
+    # Simple unsecure server config
+    config_path = Path(tmpdir/"mailmerge_server.conf")
+    config_path.write_text(textwrap.dedent("""\
+        [smtp_server]
+        host = open-smtp.example.com
+        port = 25
+    """), encoding="utf8")
+
+    # Run mailmerge
+    runner = click.testing.CliRunner()
+    with tmpdir.as_cwd():
+        result = runner.invoke(main, ["--output-format", "text"])
+    assert not result.exception
+    assert result.exit_code == 0
+
+    # Verify output
+    stdout = copy.deepcopy(result.output)
+    stdout = re.sub(r"Date:.+", "Date: REDACTED", stdout, re.MULTILINE)
+    assert stdout == textwrap.dedent("""\
+        >>> message 1
+        TO: to@test.com
+        FROM: My Self <myself@mydomain.com>
+        MIME-Version: 1.0
+        Content-Type: text/plain; charset="us-ascii"
+        Content-Transfer-Encoding: 7bit
+        Date: REDACTED
+
+        Hello My Name
+
+        >>> message 1 sent
+        >>> Limit was 1 message.  To remove the limit, use the --no-limit option.
+        >>> This was a dry run.  To send messages, use the --no-dry-run option.
+    """)  # noqa: E501
+
+
+def test_database_semicolon(tmpdir):
+    """Automatically detect semicolon-delimited database format."""
+    # Simple template
+    template_path = Path(tmpdir/"mailmerge_template.txt")
+    template_path.write_text(textwrap.dedent("""\
+        TO: {{email}}
+        FROM: My Self <myself@mydomain.com>
+
+        Hello {{name}}
+    """), encoding="utf8")
+
+    # Semicolon-separated format database
+    database_path = Path(tmpdir/"mailmerge_database.csv")
+    database_path.write_text(textwrap.dedent("""\
+        email;name
+        to@test.com;My Name
+    """), encoding="utf8")
+
+    # Simple unsecure server config
+    config_path = Path(tmpdir/"mailmerge_server.conf")
+    config_path.write_text(textwrap.dedent("""\
+        [smtp_server]
+        host = open-smtp.example.com
+        port = 25
+    """), encoding="utf8")
+
+    # Run mailmerge
+    runner = click.testing.CliRunner()
+    with tmpdir.as_cwd():
+        result = runner.invoke(main, ["--output-format", "text"])
+    assert not result.exception
+    assert result.exit_code == 0
+
+    # Verify output
+    stdout = copy.deepcopy(result.output)
+    stdout = re.sub(r"Date:.+", "Date: REDACTED", stdout, re.MULTILINE)
+    assert stdout == textwrap.dedent("""\
+        >>> message 1
+        TO: to@test.com
+        FROM: My Self <myself@mydomain.com>
+        MIME-Version: 1.0
+        Content-Type: text/plain; charset="us-ascii"
+        Content-Transfer-Encoding: 7bit
+        Date: REDACTED
+
+        Hello My Name
+
+        >>> message 1 sent
+        >>> Limit was 1 message.  To remove the limit, use the --no-limit option.
+        >>> This was a dry run.  To send messages, use the --no-dry-run option.
+    """)  # noqa: E501