alanorth/csv-metadata-quality
1
0
Fork 0
mirror of https://github.com/ilri/csv-metadata-quality.git synced 2024-11-17 19:47:03 +01:00
Code Releases Activity

Format with black

Browse Source
This commit is contained in:
Alan Orth 2019-08-29 01:10:39 +03:00
parent b375f0e895
commit d97dcd19db
Signed by: alanorth
GPG Key ID: 0FB860CC9C45B1B9
5 changed files with 117 additions and 79 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
csv_metadata_quality/__main__.py
View File

@ -6,5 +6,5 @@ def main():
app.run(argv)
if __name__ == '__main__':
if __name__ == "__main__":
main()

56
csv_metadata_quality/app.py
View File

@ -9,13 +9,37 @@ import sys
def parse_args(argv):
parser = argparse.ArgumentParser(description='Metadata quality checker and fixer.')
parser.add_argument('--agrovoc-fields', '-a', help='Comma-separated list of fields to validate against AGROVOC, for example: dc.subject,cg.coverage.country')
parser.add_argument('--input-file', '-i', help='Path to input file. Can be UTF-8 CSV or Excel XLSX.', required=True, type=argparse.FileType('r', encoding='UTF-8'))
parser.add_argument('--output-file', '-o', help='Path to output file (always CSV).', required=True, type=argparse.FileType('w', encoding='UTF-8'))
parser.add_argument('--unsafe-fixes', '-u', help='Perform unsafe fixes.', action='store_true')
parser.add_argument('--version', '-V', action='version', version=f'CSV Metadata Quality v{VERSION}')
parser.add_argument('--exclude-fields', '-x', help='Comma-separated list of fields to skip, for example: dc.contributor.author,dc.identifier.citation')
parser = argparse.ArgumentParser(description="Metadata quality checker and fixer.")
parser.add_argument(
"--agrovoc-fields",
"-a",
help="Comma-separated list of fields to validate against AGROVOC, for example: dc.subject,cg.coverage.country",
)
parser.add_argument(
"--input-file",
"-i",
help="Path to input file. Can be UTF-8 CSV or Excel XLSX.",
required=True,
type=argparse.FileType("r", encoding="UTF-8"),
)
parser.add_argument(
"--output-file",
"-o",
help="Path to output file (always CSV).",
required=True,
type=argparse.FileType("w", encoding="UTF-8"),
)
parser.add_argument(
"--unsafe-fixes", "-u", help="Perform unsafe fixes.", action="store_true"
)
parser.add_argument(
"--version", "-V", action="version", version=f"CSV Metadata Quality v{VERSION}"
)
parser.add_argument(
"--exclude-fields",
"-x",
help="Comma-separated list of fields to skip, for example: dc.contributor.author,dc.identifier.citation",
)
args = parser.parse_args()
return args
@ -40,11 +64,11 @@ def run(argv):
skip = False
# Split the list of excludes on ',' so we can test exact matches
# rather than fuzzy matches with regexes or "if word in string"
for exclude in args.exclude_fields.split(','):
for exclude in args.exclude_fields.split(","):
if column == exclude and skip is False:
skip = True
if skip:
print(f'Skipping {column}')
print(f"Skipping {column}")
continue
@ -58,7 +82,7 @@ def run(argv):
# Fix: missing space after comma. Only run on author and citation
# fields for now, as this problem is mostly an issue in names.
if args.unsafe_fixes:
match = re.match(r'^.*?(author|citation).*$', column)
match = re.match(r"^.*?(author|citation).*$", column)
if match is not None:
df[column] = df[column].apply(fix.comma_space, field_name=column)
@ -83,32 +107,32 @@ def run(argv):
# Check: invalid AGROVOC subject
if args.agrovoc_fields:
# Identify fields the user wants to validate against AGROVOC
for field in args.agrovoc_fields.split(','):
for field in args.agrovoc_fields.split(","):
if column == field:
df[column] = df[column].apply(check.agrovoc, field_name=column)
# Check: invalid language
match = re.match(r'^.*?language.*$', column)
match = re.match(r"^.*?language.*$", column)
if match is not None:
df[column] = df[column].apply(check.language)
# Check: invalid ISSN
match = re.match(r'^.*?issn.*$', column)
match = re.match(r"^.*?issn.*$", column)
if match is not None:
df[column] = df[column].apply(check.issn)
# Check: invalid ISBN
match = re.match(r'^.*?isbn.*$', column)
match = re.match(r"^.*?isbn.*$", column)
if match is not None:
df[column] = df[column].apply(check.isbn)
# Check: invalid date
match = re.match(r'^.*?date.*$', column)
match = re.match(r"^.*?date.*$", column)
if match is not None:
df[column] = df[column].apply(check.date, field_name=column)
# Check: filename extension
if column == 'filename':
if column == "filename":
df[column] = df[column].apply(check.filename_extension)
# Write

74
csv_metadata_quality/check.py
View File

@ -18,10 +18,10 @@ def issn(field):
return
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
if not issn.is_valid(value):
print(f'Invalid ISSN: {value}')
print(f"Invalid ISSN: {value}")
return field
@ -43,10 +43,10 @@ def isbn(field):
return
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
if not isbn.is_valid(value):
print(f'Invalid ISBN: {value}')
print(f"Invalid ISBN: {value}")
return field
@ -64,13 +64,13 @@ def separators(field):
return
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
# After splitting, see if there are any remaining "|" characters
match = re.findall(r'^.*?\|.*$', value)
match = re.findall(r"^.*?\|.*$", value)
if match:
print(f'Invalid multi-value separator: {field}')
print(f"Invalid multi-value separator: {field}")
return field
@ -88,22 +88,22 @@ def date(field, field_name):
from datetime import datetime
if pd.isna(field):
print(f'Missing date ({field_name}).')
print(f"Missing date ({field_name}).")
return
# Try to split multi-value field on "||" separator
multiple_dates = field.split('||')
multiple_dates = field.split("||")
# We don't allow multi-value date fields
if len(multiple_dates) > 1:
print(f'Multiple dates not allowed ({field_name}): {field}')
print(f"Multiple dates not allowed ({field_name}): {field}")
return field
try:
# Check if date is valid YYYY format
datetime.strptime(field, '%Y')
datetime.strptime(field, "%Y")
return field
except ValueError:
@ -111,7 +111,7 @@ def date(field, field_name):
try:
# Check if date is valid YYYY-MM format
datetime.strptime(field, '%Y-%m')
datetime.strptime(field, "%Y-%m")
return field
except ValueError:
@ -119,11 +119,11 @@ def date(field, field_name):
try:
# Check if date is valid YYYY-MM-DD format
datetime.strptime(field, '%Y-%m-%d')
datetime.strptime(field, "%Y-%m-%d")
return field
except ValueError:
print(f'Invalid date ({field_name}): {field}')
print(f"Invalid date ({field_name}): {field}")
return field
@ -140,7 +140,7 @@ def suspicious_characters(field, field_name):
return
# List of suspicious characters, for example: ́ˆ~`
suspicious_characters = ['\u00B4', '\u02C6', '\u007E', '\u0060']
suspicious_characters = ["\u00B4", "\u02C6", "\u007E", "\u0060"]
for character in suspicious_characters:
# Find the position of the suspicious character in the string
@ -156,8 +156,10 @@ def suspicious_characters(field, field_name):
# character and spanning enough of the rest to give a preview,
# but not too much to cause the line to break in terminals with
# a default of 80 characters width.
suspicious_character_msg = f'Suspicious character ({field_name}): {field_subset}'
print(f'{suspicious_character_msg:1.80}')
suspicious_character_msg = (
f"Suspicious character ({field_name}): {field_subset}"
)
print(f"{suspicious_character_msg:1.80}")
return field
@ -177,22 +179,22 @@ def language(field):
# need to handle "Other" values here...
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
# After splitting, check if language value is 2 or 3 characters so we
# can check it against ISO 639-2 or ISO 639-3 accordingly.
if len(value) == 2:
if not languages.get(alpha_2=value):
print(f'Invalid ISO 639-2 language: {value}')
print(f"Invalid ISO 639-2 language: {value}")
pass
elif len(value) == 3:
if not languages.get(alpha_3=value):
print(f'Invalid ISO 639-3 language: {value}')
print(f"Invalid ISO 639-3 language: {value}")
pass
else:
print(f'Invalid language: {value}')
print(f"Invalid language: {value}")
return field
@ -220,12 +222,16 @@ def agrovoc(field, field_name):
return
# Try to split multi-value field on "||" separator
for value in field.split('||'):
request_url = f'http://agrovoc.uniroma2.it/agrovoc/rest/v1/agrovoc/search?query={value}'
for value in field.split("||"):
request_url = (
f"http://agrovoc.uniroma2.it/agrovoc/rest/v1/agrovoc/search?query={value}"
)
# enable transparent request cache with thirty days expiry
expire_after = timedelta(days=30)
requests_cache.install_cache('agrovoc-response-cache', expire_after=expire_after)
requests_cache.install_cache(
"agrovoc-response-cache", expire_after=expire_after
)
request = requests.get(request_url)
@ -236,8 +242,8 @@ def agrovoc(field, field_name):
data = request.json()
# check if there are any results
if len(data['results']) == 0:
print(f'Invalid AGROVOC ({field_name}): {value}')
if len(data["results"]) == 0:
print(f"Invalid AGROVOC ({field_name}): {value}")
return field
@ -260,10 +266,18 @@ def filename_extension(field):
return
# Try to split multi-value field on "||" separator
values = field.split('||')
values = field.split("||")
# List of common filename extentions
common_filename_extensions = ['.pdf', '.doc', '.docx', '.ppt', '.pptx', '.xls', '.xlsx']
common_filename_extensions = [
".pdf",
".doc",
".docx",
".ppt",
".pptx",
".xls",
".xlsx",
]
# Iterate over all values
for value in values:
@ -272,7 +286,7 @@ def filename_extension(field):
for filename_extension in common_filename_extensions:
# Check for extension at the end of the filename
pattern = re.escape(filename_extension) + r'$'
pattern = re.escape(filename_extension) + r"$"
match = re.search(pattern, value, re.IGNORECASE)
if match is not None:
@ -282,6 +296,6 @@ def filename_extension(field):
break
if filename_extension_match is False:
print(f'Filename with uncommon extension: {value}')
print(f"Filename with uncommon extension: {value}")
return field

62
csv_metadata_quality/fix.py
View File

@ -16,23 +16,23 @@ def whitespace(field):
values = list()
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
# Strip leading and trailing whitespace
value = value.strip()
# Replace excessive whitespace (>2) with one space
pattern = re.compile(r'\s{2,}')
pattern = re.compile(r"\s{2,}")
match = re.findall(pattern, value)
if match:
print(f'Excessive whitespace: {value}')
value = re.sub(pattern, ' ', value)
print(f"Excessive whitespace: {value}")
value = re.sub(pattern, " ", value)
# Save cleaned value
values.append(value)
# Create a new field consisting of all values joined with "||"
new_field = '||'.join(values)
new_field = "||".join(values)
return new_field
@ -48,21 +48,21 @@ def separators(field):
values = list()
# Try to split multi-value field on "||" separator
for value in field.split('||'):
for value in field.split("||"):
# After splitting, see if there are any remaining "|" characters
pattern = re.compile(r'\|')
pattern = re.compile(r"\|")
match = re.findall(pattern, value)
if match:
print(f'Fixing invalid multi-value separator: {value}')
print(f"Fixing invalid multi-value separator: {value}")
value = re.sub(pattern, '||', value)
value = re.sub(pattern, "||", value)
# Save cleaned value
values.append(value)
# Create a new field consisting of all values joined with "||"
new_field = '||'.join(values)
new_field = "||".join(values)
return new_field
@ -86,36 +86,36 @@ def unnecessary_unicode(field):
return
# Check for zero-width space characters (U+200B)
pattern = re.compile(r'\u200B')
pattern = re.compile(r"\u200B")
match = re.findall(pattern, field)
if match:
print(f'Removing unnecessary Unicode (U+200B): {field}')
field = re.sub(pattern, '', field)
print(f"Removing unnecessary Unicode (U+200B): {field}")
field = re.sub(pattern, "", field)
# Check for replacement characters (U+FFFD)
pattern = re.compile(r'\uFFFD')
pattern = re.compile(r"\uFFFD")
match = re.findall(pattern, field)
if match:
print(f'Removing unnecessary Unicode (U+FFFD): {field}')
field = re.sub(pattern, '', field)
print(f"Removing unnecessary Unicode (U+FFFD): {field}")
field = re.sub(pattern, "", field)
# Check for no-break spaces (U+00A0)
pattern = re.compile(r'\u00A0')
pattern = re.compile(r"\u00A0")
match = re.findall(pattern, field)
if match:
print(f'Removing unnecessary Unicode (U+00A0): {field}')
field = re.sub(pattern, '', field)
print(f"Removing unnecessary Unicode (U+00A0): {field}")
field = re.sub(pattern, "", field)
# Check for soft hyphens (U+00AD), sometimes preceeded with a normal hyphen
pattern = re.compile(r'\u002D*?\u00AD')
pattern = re.compile(r"\u002D*?\u00AD")
match = re.findall(pattern, field)
if match:
print(f'Replacing unnecessary Unicode (U+00AD): {field}')
field = re.sub(pattern, '-', field)
print(f"Replacing unnecessary Unicode (U+00AD): {field}")
field = re.sub(pattern, "-", field)
return field
@ -128,7 +128,7 @@ def duplicates(field):
return
# Try to split multi-value field on "||" separator
values = field.split('||')
values = field.split("||")
# Initialize an empty list to hold the de-duplicated values
new_values = list()
@ -139,10 +139,10 @@ def duplicates(field):
if value not in new_values:
new_values.append(value)
else:
print(f'Dropping duplicate value: {value}')
print(f"Dropping duplicate value: {value}")
# Create a new field consisting of all values joined with "||"
new_field = '||'.join(new_values)
new_field = "||".join(new_values)
return new_field
@ -169,11 +169,11 @@ def newlines(field):
return
# Check for Unix line feed (LF)
match = re.findall(r'\n', field)
match = re.findall(r"\n", field)
if match:
print(f'Removing newline: {field}')
field = field.replace('\n', '')
print(f"Removing newline: {field}")
field = field.replace("\n", "")
return field
@ -193,10 +193,10 @@ def comma_space(field, field_name):
return
# Check for comma followed by a word character
match = re.findall(r',\w', field)
match = re.findall(r",\w", field)
if match:
print(f'Adding space after comma ({field_name}): {field}')
field = re.sub(r',(\w)', r', \1', field)
print(f"Adding space after comma ({field_name}): {field}")
field = re.sub(r",(\w)", r", \1", field)
return field

2
csv_metadata_quality/version.py
View File

@ -1 +1 @@
VERSION = '0.2.2'
VERSION = "0.2.2"