def test_write_to_sqlite(input_file):
"""
Test production of sqlite output
GIVEN an input file
WHEN writing to sqlite
THEN check output exists and contains content
"""
# GIVEN an input file
# WHEN writing to sqlite
output_filename = Path(f"{uuid4().hex}.db")
tscribe.write(input_file, save_as=output_filename, format="sqlite")
# THEN check output exists and contains content
assert output_filename.is_file(), "Output file should exist"
conn = sqlite3.connect(str(output_filename))
c = conn.cursor()
c.execute("SELECT * FROM transcript")
query = c.fetchall()
data = tscribe.load_json(input_file)
df = tscribe.decode_transcript(data)
assert len(query) == len(df), "Database table should be length of dataframe"
# Teardown
os.remove(output_filename)
def test_write_to_csv(input_file):
"""
Test production of csv output
GIVEN an input file
WHEN writing to csv
THEN check output exists and contains content
"""
# GIVEN an input file
# WHEN writing to csv
output_filename = Path(f"{uuid4().hex}.csv")
tscribe.write(input_file, save_as=output_filename, format="csv")
# THEN check output exists and contains content
assert output_filename.is_file(), "Output file should exist"
with open(output_filename, "r") as file:
lines = file.readlines()
data = tscribe.load_json(input_file)
df = tscribe.decode_transcript(data)
assert len(lines) == len(df) + 1, "CSV should be length of dataframe + headers"
# Teardown
os.remove(output_filename)
def test_decode_transcript(input_file):
"""
Test transcript decoding function
GIVEN a data dict
WHEN calling decode_transcript(...)
THEN
"""
# GIVEN a data dict
data = tscribe.load_json(input_file)
# WHEN calling decode_transcript(...)
df = tscribe.decode_transcript(data)
# THEN
assert isinstance(
df, pandas.DataFrame
), "decode_transcript should return a Pandas Data Frame"
rows, cols = df.shape
assert cols == 4, "Dataframe should have four columns"
if input_file == "sample_single.json":
# TODO
pass
if input_file == "sample_multiple.json":
assert rows == len(
data["results"]["speaker_labels"]["segments"]
), "Rows should match number of segments"
def test_write_to_docx(input_file):
"""
Test production of docx output
GIVEN an input file
WHEN writing to docx
THEN check output exists and contains content
"""
logging.info("test_write_to_docx")
# GIVEN an input file
# WHEN writing to docx
output_filename = Path(f"{uuid4().hex}.docx")
tscribe.write(input_file, save_as=output_filename, format="docx")
# THEN check output exists and contains content
assert output_filename.is_file(), "Output file should exist"
document = Document(output_filename)
assert (len(document.tables) == 2
), "Document should contain two tables, stats and transcript"
t_conf = document.tables[0].cell(0, 0).text
t_count = document.tables[0].cell(0, 1).text
t_perc = document.tables[0].cell(0, 2).text
assert (t_conf, t_count, t_perc) == (
"Confidence",
"Count",
"Percentage",
), "First table should be stats headers"
assert len(
document.tables[0].rows) == 12, "Stats table should hold 12 rows"
t_time = document.tables[1].cell(0, 0).text
t_speaker = document.tables[1].cell(0, 1).text
t_content = document.tables[1].cell(0, 2).text
assert (t_time, t_speaker, t_content) == (
"Time",
"Speaker",
"Content",
), "Second table should be transcript headers"
data = tscribe.load_json(input_file)
df = tscribe.decode_transcript(data)
assert (len(document.tables[1].rows) == len(df) +
1), "Second table should be length of dataframe + headers"
assert (
"chart.png"
in document.paragraphs[6]._p.xml), "Chart should be in paragraph six"
# Teardown
os.remove(output_filename)
def test_write_to_vtt(input_file):
"""
Test production of vtt format
GIVEN an input file
WHEN writing to vtt
THEN check output exists and contains content
"""
logging.info("test_write_to_vtt")
# GIVEN an input file
# WHEN writing to vtt
output_filename = Path(f"{uuid4().hex}.vtt")
tscribe.write(input_file, save_as=output_filename, format="vtt")
# THEN check output exists and contains content
vtt = webvtt.read(output_filename)
data = tscribe.load_json(input_file)
df = tscribe.decode_transcript(data)
assert len(vtt.captions) == len(
df), "vtt file should have equal captions to df rows"
for caption in vtt.captions:
assert hasattr(caption,
"start"), "each caption should have a start_time"
assert hasattr(caption, "end"), "each caption should have a end_time"
assert hasattr(caption, "text"), "each caption should have text"
assert len(caption.lines) >= len(
caption.text) / 80, "text should be split into max 80 long lines"
if input_file != "sample_single.json":
assert hasattr(
caption,
"identifier"), "each caption should have an identifier"
# Teardown
os.remove(output_filename)