Set the 1.1.0 release date

Reading the whole file is a bit too much to detect if the file
looks like an xlsx file.

.github/workflows/test.yml

@@ -9,7 +9,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: [3.5, 3.6, 3.7]
+        python-version: [3.5, 3.6, 3.7, 3.8]
         os: [ubuntu-latest, macOS-latest, windows-latest]
 
     steps:
@@ -28,6 +28,5 @@ jobs:
 
     - name: Tox tests
       shell: bash
-      # Drop the dot: py3.7-tests -> py37-tests
       run: |
-        tox -e py`echo ${{ matrix.python-version }} | tr -d .`-tests
+        tox -e py

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
   - repo: https://github.com/asottile/pyupgrade
-    rev: v1.25.2
+    rev: v1.26.2
     hooks:
       - id: pyupgrade
         args: ["--py3-plus"]
@@ -12,7 +12,7 @@ repos:
         additional_dependencies: [toml]
 
   - repo: https://github.com/pre-commit/pygrep-hooks
-    rev: v1.4.2
+    rev: v1.4.4
     hooks:
       - id: python-check-blanket-noqa
       - id: rst-backticks

HISTORY.md

@@ -1,5 +1,19 @@
 # History
 
+## 1.1.0 (2020-02-13)
+
+### Deprecations
+
+- Upcoming breaking change in Tablib 2.0.0: the `Row.lpush/rpush` logic is reversed.
+  `lpush` is appending while `rpush` and `append` are prepending. The broken behavior
+  will remain in Tablib 1.x and will be fixed (reversed) in Tablib 2.0.0 (#453). If you
+  count on the broken behavior, please update your code when you upgrade to Tablib 2.x.
+
+### Improvements
+
+- Tablib is now able to import CSV content where not all rows have the same
+  length. Missing columns on any line receive the empty string (#226).
+
 ## 1.0.0 (2020-01-13)
 
 ### Breaking changes

src/tablib/formats/__init__.py

@@ -17,7 +17,7 @@ uninstalled_format_messages = {
     "df": {"package_name": "pandas package", "extras_name": "pandas"},
     "html": {"package_name": "MarkupPy package", "extras_name": "html"},
     "ods": {"package_name": "odfpy package", "extras_name": "ods"},
-    "xls": {"package_name": "odfpy and xlwt packages", "extras_name": "ods"},
+    "xls": {"package_name": "xlrd and xlwt packages", "extras_name": "xls"},
     "xlsx": {"package_name": "openpyxl package", "extras_name": "xlsx"},
     "yaml": {"package_name": "pyyaml package", "extras_name": "yaml"},
 }

src/tablib/formats/_csv.py

@@ -46,6 +46,8 @@ class CSVFormat:
             if (i == 0) and (headers):
                 dset.headers = row
             elif row:
+                if i > 0 and len(row) < dset.width:
+                    row += [''] * (dset.width - len(row))
                 dset.append(row)
 
     @classmethod

src/tablib/formats/_xlsx.py

@@ -3,12 +3,12 @@
 
 from io import BytesIO
 
-import openpyxl
 import tablib
-
-Workbook = openpyxl.workbook.Workbook
-ExcelWriter = openpyxl.writer.excel.ExcelWriter
-get_column_letter = openpyxl.utils.get_column_letter
+from openpyxl.reader.excel import ExcelReader, load_workbook
+from openpyxl.styles import Alignment, Font
+from openpyxl.utils import get_column_letter
+from openpyxl.workbook import Workbook
+from openpyxl.writer.excel import ExcelWriter
 
 
 class XLSXFormat:
@@ -19,7 +19,10 @@ class XLSXFormat:
     def detect(cls, stream):
         """Returns True if given stream is a readable excel file."""
         try:
-            openpyxl.reader.excel.load_workbook(stream, read_only=True)
+            # No need to fully load the file, it should be enough to be able to
+            # read the manifest.
+            reader = ExcelReader(stream, read_only=False)
+            reader.read_manifest()
             return True
         except Exception:
             return False
@@ -60,7 +63,7 @@ class XLSXFormat:
 
         dset.wipe()
 
-        xls_book = openpyxl.reader.excel.load_workbook(in_stream, read_only=True)
+        xls_book = load_workbook(in_stream, read_only=True)
         sheet = xls_book.active
 
         dset.title = sheet.title
@@ -78,7 +81,7 @@ class XLSXFormat:
 
         dbook.wipe()
 
-        xls_book = openpyxl.reader.excel.load_workbook(in_stream, read_only=True)
+        xls_book = load_workbook(in_stream, read_only=True)
 
         for sheet in xls_book.worksheets:
             data = tablib.Dataset()
@@ -102,8 +105,8 @@ class XLSXFormat:
             _offset = i
             _package.insert((sep[0] + _offset), (sep[1],))
 
-        bold = openpyxl.styles.Font(bold=True)
-        wrap_text = openpyxl.styles.Alignment(wrap_text=True)
+        bold = Font(bold=True)
+        wrap_text = Alignment(wrap_text=True)
 
         for i, row in enumerate(_package):
             row_number = i + 1

tests/test_tablib.py

@@ -804,6 +804,25 @@ class CSVTests(BaseTestCase):
         data.csv = csv_text
         self.assertEqual(data.width, 7)
 
+    def test_csv_import_set_ragged(self):
+        """Import CSV set when not all rows have the same length."""
+        csv_text = (
+            "H1,H2,H3\n"
+            "A,B\n"
+            "C,D,E\n"
+            "\n"
+            "F\n"
+        )
+        dataset = tablib.import_set(csv_text, format="csv")
+        self.assertEqual(
+            str(dataset),
+            'H1|H2|H3\n'
+            '--|--|--\n'
+            'A |B |  \n'
+            'C |D |E \n'
+            'F |  |  '
+        )
+
     def test_csv_export(self):
         """Verify exporting dataset object as CSV."""
 

tox.ini

@@ -8,12 +8,16 @@ envlist =
 
 [testenv]
 deps =
-    tests: -rtests/requirements.txt
-    docs: sphinx
+    -rtests/requirements.txt
 extras = pandas
 commands =
-    tests: pytest {posargs:tests}
-    docs: sphinx-build -b html -d {envtmpdir}/doctrees docs {envtmpdir}/html
+    pytest {posargs:tests}
+
+[testenv:docs]
+deps =
+    sphinx
+commands =
+    sphinx-build -b html -d {envtmpdir}/doctrees docs {envtmpdir}/html
 
 [testenv:lint]
 deps =