From 3406af37efb7ec3d027d3e4b3afff6b007079e6f Mon Sep 17 00:00:00 2001
From: Douglas Mennella <douglas.mennella@gmail.com>
Date: Fri, 29 Apr 2022 19:33:13 +0900
Subject: [PATCH 1/2] Fixes #68.  Protect against sniffing failing.  Fallback
 to overrides

---
 csv_reconcile/initdb.py          |  9 +++++++--
 tests/conftest.py                | 18 ++++++++++++++++++
 tests/main/test_csv_reconcile.py | 19 +++++++++++++++++++
 3 files changed, 44 insertions(+), 2 deletions(-)

diff --git a/csv_reconcile/initdb.py b/csv_reconcile/initdb.py
index 807e435..15d5668 100644
--- a/csv_reconcile/initdb.py
+++ b/csv_reconcile/initdb.py
@@ -73,9 +73,14 @@ def init_db(db,
     with db:
         # Create a table with ids (as PRIMARY ID), words and bigrams
         with open(csvfilenm, newline='', **enckwarg) as csvfile:
-            dialect = csv.Sniffer().sniff(csvfile.read(1024))
+            dialect = None
+            try:
+                dialect = csv.Sniffer().sniff(csvfile.read(1024))
+            except:
+                pass
+
             csvfile.seek(0)
-            reader = csv.reader(csvfile, dialect, **csvkwargs)
+            reader = csv.reader(csvfile, dialect=dialect, **csvkwargs)
             header = next(reader)
 
             # Throws if col doesn't exist
diff --git a/tests/conftest.py b/tests/conftest.py
index 6359a4e..8e6cabb 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -34,6 +34,13 @@ def ambiguous_csvcontents():
 However, above all, columns matter most
 '''.strip()
 
+@pytest.fixture
+def sniffer_throwing_csvcontents():
+    '''Try to throw off csv.Sniffer() to test overrides'''
+    return '''
+a,b,c\n1,2
+'''.strip()
+
 
 @pytest.fixture
 def formContentHeader():
@@ -78,6 +85,17 @@ def getSetup(idnm):
 
     return getSetup
 
+@pytest.fixture
+def sniffer_throwing_setup(tmp_path, sniffer_throwing_csvcontents):
+    '''mock csv file with id and name columns indicated'''
+
+    def getSetup(idnm):
+        p = tmp_path / "snfthrw_csvfile"
+        p.write_text(sniffer_throwing_csvcontents)
+        return (p, *idnm)
+
+    return getSetup
+
 @pytest.fixture
 def cfgContents():
     return '''
diff --git a/tests/main/test_csv_reconcile.py b/tests/main/test_csv_reconcile.py
index 8d80d8a..0c7b606 100644
--- a/tests/main/test_csv_reconcile.py
+++ b/tests/main/test_csv_reconcile.py
@@ -305,3 +305,22 @@ def test_csv_sniffer_overrides(app, ambiguous_setup, ambiguous_csvcontents, conf
     with chk.app_context():
         headernms = [name for _,name in getCSVCols()]
         assert headernms == items(SEP)
+
+def test_csv_sniffer_throwing(app, sniffer_throwing_setup, sniffer_throwing_csvcontents, config, mkConfig):
+
+    topline = sniffer_throwing_csvcontents.splitlines()[0]
+    items = lambda sep: [ h.strip() for h in topline.split(sep)]
+
+    # First guess is that the , is a separator
+    SEP = ','
+    chk = app(sniffer_throwing_setup(items(SEP)[:2]), config)
+    with chk.app_context():
+        headernms = [name for _,name in getCSVCols()]
+        assert headernms == items(SEP)
+
+    # Now parse with override
+    cfg = mkConfig('CSVKWARGS = {"delimiter": ","}')
+    chk = app(sniffer_throwing_setup(items(SEP)[:2]), cfg)
+    with chk.app_context():
+        headernms = [name for _,name in getCSVCols()]
+        assert headernms == items(SEP)

From 5112a0cd111c4793ff70eca65fdb335abd5fd0ae Mon Sep 17 00:00:00 2001
From: Douglas Mennella <douglas.mennella@gmail.com>
Date: Sat, 30 Apr 2022 14:45:54 +0900
Subject: [PATCH 2/2] Automated csv file encoding detection (defensive)

---
 csv_reconcile/__init__.py        | 2 +-
 pyproject.toml                   | 2 +-
 tests/main/test_csv_reconcile.py | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/csv_reconcile/__init__.py b/csv_reconcile/__init__.py
index 928a955..3575fc2 100644
--- a/csv_reconcile/__init__.py
+++ b/csv_reconcile/__init__.py
@@ -22,7 +22,7 @@
 except:
     from importlib import metadata
 
-__version__ = '0.3.1'
+__version__ = '0.3.2'
 #------------------------------------------------------------------
 # Implement reconciliation API
 # [[https://reconciliation-api.github.io/specs/latest/]]
diff --git a/pyproject.toml b/pyproject.toml
index 8db7102..5538297 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "csv-reconcile"
-version = "0.3.1"
+version = "0.3.2"
 description = "OpenRefine reconciliation service backed by csv resource"
 authors = ["Douglas Mennella <trx2358-pypi@yahoo.com>"]
 license = "MIT"
diff --git a/tests/main/test_csv_reconcile.py b/tests/main/test_csv_reconcile.py
index 0c7b606..36cd001 100644
--- a/tests/main/test_csv_reconcile.py
+++ b/tests/main/test_csv_reconcile.py
@@ -8,7 +8,7 @@
 
 
 def test_version():
-    assert __version__ == '0.3.1'
+    assert __version__ == '0.3.2'
 
 
 def test_manifest(basicClient):