chore: fixed to_dataframe header issue (#20)

galz10 · web-flow · commit 75c82ec0bd0b · 2022-11-23T20:57:17.000-08:00
* chore: fix to_dataframe header issue
diff --git a/packages/google-cloud-documentai-toolbox/.github/CODEOWNERS b/packages/google-cloud-documentai-toolbox/.github/CODEOWNERS
@@ -9,4 +9,4 @@
 *     @googleapis/yoshi-python @googleapis/cdpe-cloudai
 
 # @googleapis/python-samples-reviewers @googleapis/cdpe-cloudai are the default owners for samples changes
-# /samples/   @googleapis/python-samples-reviewers @googleapis/cdpe-cloudai
+/samples/   @googleapis/python-samples-reviewers @googleapis/cdpe-cloudai
diff --git a/packages/google-cloud-documentai-toolbox/.github/workflows/docs.yml b/packages/google-cloud-documentai-toolbox/.github/workflows/docs.yml
@@ -12,7 +12,7 @@ jobs:
     - name: Setup Python
       uses: actions/setup-python@v4
       with:
-        python-version: "3.10"
+        python-version: "3.9"
     - name: Install nox
       run: |
         python -m pip install --upgrade setuptools pip wheel
@@ -28,7 +28,7 @@ jobs:
     - name: Setup Python
       uses: actions/setup-python@v4
       with:
-        python-version: "3.10"
+        python-version: "3.9"
     - name: Install nox
       run: |
         python -m pip install --upgrade setuptools pip wheel
diff --git a/packages/google-cloud-documentai-toolbox/.github/workflows/lint.yml b/packages/google-cloud-documentai-toolbox/.github/workflows/lint.yml
@@ -12,7 +12,7 @@ jobs:
     - name: Setup Python
       uses: actions/setup-python@v4
       with:
-        python-version: "3.10"
+        python-version: "3.8"
     - name: Install nox
       run: |
         python -m pip install --upgrade setuptools pip wheel
diff --git a/packages/google-cloud-documentai-toolbox/.github/workflows/unittest.yml b/packages/google-cloud-documentai-toolbox/.github/workflows/unittest.yml
@@ -41,7 +41,7 @@ jobs:
     - name: Setup Python
       uses: actions/setup-python@v4
       with:
-        python-version: "3.10"
+        python-version: "3.8"
     - name: Install coverage
       run: |
         python -m pip install --upgrade setuptools pip wheel
diff --git a/packages/google-cloud-documentai-toolbox/.kokoro/docker/docs/Dockerfile b/packages/google-cloud-documentai-toolbox/.kokoro/docker/docs/Dockerfile
@@ -60,16 +60,16 @@ RUN apt-get update \
   && rm -rf /var/lib/apt/lists/* \
   && rm -f /var/cache/apt/archives/*.deb
 
-###################### Install python 3.8.11
+###################### Install python 3.9.13
 
-# Download python 3.8.11
-RUN wget https://www.python.org/ftp/python/3.8.11/Python-3.8.11.tgz
+# Download python 3.9.13
+RUN wget https://www.python.org/ftp/python/3.9.13/Python-3.9.13.tgz
 
 # Extract files
-RUN tar -xvf Python-3.8.11.tgz
+RUN tar -xvf Python-3.9.13.tgz
 
-# Install python 3.8.11
-RUN ./Python-3.8.11/configure --enable-optimizations
+# Install python 3.9.13
+RUN ./Python-3.9.13/configure --enable-optimizations
 RUN make altinstall
 
 ###################### Install pip
diff --git a/packages/google-cloud-documentai-toolbox/.kokoro/release.sh b/packages/google-cloud-documentai-toolbox/.kokoro/release.sh
@@ -16,7 +16,7 @@
 set -eo pipefail
 
 # Start the releasetool reporter
-python3 -m pip install --require-hashes -r .kokoro/requirements.txt
+python3 -m pip install --require-hashes -r github/python-documentai-toolbox/.kokoro/requirements.txt
 python3 -m releasetool publish-reporter-script > /tmp/publisher-script; source /tmp/publisher-script
 
 # Disable buffering, so that the logs stream through.
diff --git a/packages/google-cloud-documentai-toolbox/.kokoro/requirements.in b/packages/google-cloud-documentai-toolbox/.kokoro/requirements.in
@@ -5,4 +5,6 @@ typing-extensions
 twine
 wheel
 setuptools
-nox
+nox
+charset-normalizer<3
+click<8.1.0
diff --git a/packages/google-cloud-documentai-toolbox/.kokoro/requirements.txt b/packages/google-cloud-documentai-toolbox/.kokoro/requirements.txt
diff --git a/packages/google-cloud-documentai-toolbox/docs/_templates/layout.html b/packages/google-cloud-documentai-toolbox/docs/_templates/layout.html
@@ -5,7 +5,7 @@
   <div class="document">
     {{ sidebar() }}
     {%- block document %}
-      <div class="Document">
+      <div class="documentwrapper">
       {%- if render_sidebar %}
         <div class="bodywrapper">
       {%- endif %}
diff --git a/packages/google-cloud-documentai-toolbox/google/cloud/documentai_toolbox/wrappers/page.py b/packages/google-cloud-documentai-toolbox/google/cloud/documentai_toolbox/wrappers/page.py
@@ -54,14 +54,18 @@ def to_dataframe(self) -> pd.DataFrame:
                 The DataFrame of the table.
 
         """
-        rows = self.body_rows if self.body_rows != [] else self.header_rows
-
-        rows.append("\n")
-        rows.append("\n")
-
-        dataframe = pd.DataFrame(rows)
-        if self.body_rows != [] and self.header_rows != []:
-            dataframe.columns = self.header_rows
+        dataframe = None
+
+        if not self.body_rows:
+            dataframe = pd.DataFrame(columns=self.header_rows)
+        else:
+            if self.header_rows != []:
+                dataframe = pd.DataFrame(self.body_rows)
+                dataframe.columns = self.header_rows
+            else:
+
+                dataframe = pd.DataFrame(self.body_rows)
+                dataframe.columns = [None] * len(self.body_rows[0])
 
         return dataframe
 
diff --git a/packages/google-cloud-documentai-toolbox/noxfile.py b/packages/google-cloud-documentai-toolbox/noxfile.py
@@ -272,12 +272,16 @@ def cover(session):
     session.run("coverage", "erase")
 
 
-@nox.session(python=DEFAULT_PYTHON_VERSION)
+@nox.session(python="3.9")
 def docs(session):
     """Build the docs for this library."""
 
     session.install("-e", ".")
-    session.install("sphinx==4.0.1", "alabaster", "recommonmark")
+    session.install(
+        "sphinx==4.0.1",
+        "alabaster",
+        "recommonmark",
+    )
 
     shutil.rmtree(os.path.join("docs", "_build"), ignore_errors=True)
     session.run(
@@ -294,13 +298,16 @@ def docs(session):
     )
 
 
-@nox.session(python=DEFAULT_PYTHON_VERSION)
+@nox.session(python="3.9")
 def docfx(session):
     """Build the docfx yaml files for this library."""
 
     session.install("-e", ".")
     session.install(
-        "sphinx==4.0.1", "alabaster", "recommonmark", "gcp-sphinx-docfx-yaml"
+        "sphinx==4.0.1",
+        "alabaster",
+        "recommonmark",
+        "gcp-sphinx-docfx-yaml",
     )
 
     shutil.rmtree(os.path.join("docs", "_build"), ignore_errors=True)
diff --git a/packages/google-cloud-documentai-toolbox/renovate.json b/packages/google-cloud-documentai-toolbox/renovate.json
@@ -5,7 +5,7 @@
     ":preserveSemverRanges",
     ":disableDependencyDashboard"
   ],
-  "ignorePaths": [".pre-commit-config.yaml", ".kokoro/requirements.txt"],
+  "ignorePaths": [".pre-commit-config.yaml", ".kokoro/requirements.txt", "setup.py"],
   "pip_requirements": {
     "fileMatch": ["requirements-test.txt", "samples/[\\S/]*constraints.txt", "samples/[\\S/]*constraints-test.txt"]
   }
diff --git a/packages/google-cloud-documentai-toolbox/tests/unit/test_page.py b/packages/google-cloud-documentai-toolbox/tests/unit/test_page.py
@@ -19,8 +19,11 @@
 
 
 def test_table_to_csv():
-    header_rows = [["This", "Is", "A", "Header", "Test"]]
-    body_rows = [["This", "Is", "A", "Body", "Test"]]
+    header_rows = [
+        ["This", "Is", "A", "Header", "Test"],
+        ["", "", "A", "Sub", "Header"],
+    ]
+    body_rows = [["This", "Is", "A", "Body", "Test"], ["1", "2", "3", "4", "5"]]
     table = page.Table(
         documentai_table=None, header_rows=header_rows, body_rows=body_rows
     )
@@ -29,11 +32,9 @@ def test_table_to_csv():
     assert (
         contents
         == """This,Is,A,Header,Test
+,,A,Sub,Header
 This,Is,A,Body,Test
-"
-",,,,
-"
-",,,,
+1,2,3,4,5
 """
     )
 
@@ -46,11 +47,38 @@ def test_table_to_csv_with_empty_body_rows():
 
     assert (
         contents
-        == """0,1,2,3,4
-This,Is,A,Header,Test
-"
-",,,,
-"
-",,,,
+        == """This,Is,A,Header,Test
+"""
+    )
+
+
+def test_table_to_csv_with_empty_header_rows():
+    body_rows = [["This"], ["Is"], ["A"], ["Body"], ["Test"]]
+    table = page.Table(documentai_table=None, header_rows=[], body_rows=body_rows)
+
+    contents = table.to_csv()
+
+    assert (
+        contents
+        == """""
+This
+Is
+A
+Body
+Test
+"""
+    )
+
+
+def test_table_to_csv_with_empty_header_rows_and_single_body():
+    body_rows = [["Body"]]
+    table = page.Table(documentai_table=None, header_rows=[], body_rows=body_rows)
+
+    contents = table.to_csv()
+
+    assert (
+        contents
+        == """""
+Body
 """
     )
