Skip to content

Commit

Permalink
chore: cleaning up InputObservation (added type hints)
Browse files Browse the repository at this point in the history 
see issue #27.
- added typehints (6) for all arguments, except (date, time) which
  always seem to be none, and need investigation to resolve (see 2)
- updated the attribute uploaded_filename to uploaded_file, since it
  is *not* a filename, but a BytesIO-like object, `UploadedFile`.
  • Loading branch information
@rmm-ch
rmm-ch committed Jan 30, 2025
1 parent 18e57c7 commit d437f49
Show file tree
Hide file tree
Showing 2 changed files with 29 additions and 24 deletions.
4 changes: 2 additions & 2 deletions src/input/input_handling.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -155,7 +155,7 @@ def process_one_file(file:UploadedFile, ix:int=0) -> Tuple[np.ndarray, str, str,
observation = InputObservation(image=file, latitude=latitude, longitude=longitude,
author_email=author_email, date=image_datetime, time=None,
date_option=date_option, time_option=time_option,
uploaded_filename=file,
uploaded_file=file,
)

#the_data = [] \
Expand Down Expand Up @@ -306,7 +306,7 @@ def metadata_inputs_one_file(file:UploadedFile, image_hash:str, dbg_ix:int=0) ->
observation = InputObservation(image=image, latitude=latitude, longitude=longitude,
author_email=author_email, date=image_datetime, time=None,
date_option=date_option, time_option=time_option,
uploaded_filename=file, image_md5=image_hash
uploaded_file=file, image_md5=image_hash
)

# TODO: pass in the hash to InputObservation, so it is done once only. (need to refactor the class a bit)
Expand Down
49 changes: 27 additions & 22 deletions src/input/input_observation.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,24 +7,24 @@ class InputObservation:
A class to hold an input observation and associated metadata
Attributes:
image (Any):
image (ndarray):
The image associated with the observation.
latitude (float):
The latitude where the observation was made.
longitude (float):
The longitude where the observation was made.
author_email (str):
The email of the author of the observation.
date (str):
date (Any): # FIXME: specify the type
The date when the observation was made.
time (str):
time (Any): # FIXME: specify the type
The time when the observation was made.
date_option (str):
date_option (datetime.date):
Additional date option for the observation.
time_option (str):
time_option (datetime.time):
Additional time option for the observation.
uploaded_filename (Any):
The uploaded filename associated with the observation.
uploaded_file (UploadedFile):
The uploaded file associated with the observation.
image_md5 (str):
The MD5 hash of the image associated with the observation.
Expand All @@ -37,6 +37,8 @@ class InputObservation:
Checks if two observations are equal.
__ne__(other):
Checks if two observations are not equal.
show_diff(other):
Shows the differences between two observations.
to_dict():
Converts the observation to a dictionary.
from_dict(data):
Expand All @@ -47,9 +49,12 @@ class InputObservation:

_inst_count = 0

def __init__(self, image=None, latitude=None, longitude=None,
author_email=None, date=None, time=None, date_option=None, time_option=None,
uploaded_filename=None, image_md5=None):
def __init__(
self, image:ndarray=None, latitude:float=None, longitude:float=None,
author_email:str=None, date=None, time=None, date_option:datetime.date=None,
time_option:datetime.time=None,
uploaded_file:UploadedFile=None, image_md5:str=None):

self.image = image
self.latitude = latitude
self.longitude = longitude
Expand All @@ -58,7 +63,7 @@ def __init__(self, image=None, latitude=None, longitude=None,
self.time = time
self.date_option = date_option
self.time_option = time_option
self.uploaded_filename = uploaded_filename
self.uploaded_file = uploaded_file
self.image_md5 = image_md5
self._top_predictions = []

Expand All @@ -80,10 +85,10 @@ def top_predictions(self):
def assign_image_md5(self):
raise DeprecationWarning("This method is deprecated. hash is a required constructor argument.")
if not self.image_md5:
self.image_md5 = hashlib.md5(self.uploaded_filename.read()).hexdigest() if self.uploaded_filename else generate_random_md5()
self.image_md5 = hashlib.md5(self.uploaded_file.read()).hexdigest() if self.uploaded_file else generate_random_md5()

# new comment / hybj hunk
self._cprint(f"[D] Assigned image md5: {self.image_md5} for {self.uploaded_filename}")
self._cprint(f"[D] Assigned image md5: {self.image_md5} for {self.uploaded_file}")

def _cprint(self, msg:str, color:str=OKGREEN):
"""Print colored message"""
Expand All @@ -94,7 +99,7 @@ def __str__(self):
return (
f"Observation: {_im_str}, {self.latitude}, {self.longitude}, "
f"{self.author_email}, {self.date}, {self.time}, {self.date_option}, "
f"{self.time_option}, {self.uploaded_filename}, {self.image_md5}"
f"{self.time_option}, {self.uploaded_file}, {self.image_md5}"
)

def __repr__(self):
Expand All @@ -109,7 +114,7 @@ def __repr__(self):
f"Time: {self.time}, "
f"Date Option: {self.date_option}, "
f"Time Option: {self.time_option}, "
f"Uploaded Filename: {self.uploaded_filename}"
f"Uploaded Filename: {self.uploaded_file}"
f"Image MD5 hash: {self.image_md5}"
)

Expand All @@ -132,7 +137,7 @@ def __eq__(self, other):
self.date_option == other.date_option and
# temporarily skip time_option, it is followed by the clock and that is always differnt
#self.time_option == other.time_option and
self.uploaded_filename == other.uploaded_filename and
self.uploaded_file == other.uploaded_file and
self.image_md5 == other.image_md5
)
return equality
Expand Down Expand Up @@ -165,7 +170,7 @@ def show_diff(self, other):
differences.append(f" Date option is different. (self: {self.date_option}, other: {other.date_option})")
if self.time_option != other.time_option:
differences.append(f" Time option is different. (self: {self.time_option}, other: {other.time_option})")
if self.uploaded_filename != other.uploaded_filename:
if self.uploaded_file != other.uploaded_file:
differences.append(" Uploaded filename is different.")
if self.image_md5 != other.image_md5:
differences.append(" Image MD5 hash is different.")
Expand All @@ -183,17 +188,17 @@ def __ne__(self, other):
def to_dict(self):
return {
#"image": self.image,
"image_filename": self.uploaded_filename.name if self.uploaded_filename else None,
"image_filename": self.uploaded_file.name if self.uploaded_file else None,
"image_md5": self.image_md5,
#"image_md5": hashlib.md5(self.uploaded_filename.read()).hexdigest() if self.uploaded_filename else generate_random_md5(),
#"image_md5": hashlib.md5(self.uploaded_file.read()).hexdigest() if self.uploaded_file else generate_random_md5(),
"latitude": self.latitude,
"longitude": self.longitude,
"author_email": self.author_email,
"date": self.date,
"time": self.time,
"date_option": str(self.date_option),
"time_option": str(self.time_option),
"uploaded_filename": self.uploaded_filename
"uploaded_file": self.uploaded_file
}

@classmethod
Expand All @@ -207,7 +212,7 @@ def from_dict(cls, data):
time=data.get("time"),
date_option=data.get("date_option"),
time_option=data.get("time_option"),
uploaded_filename=data.get("uploaded_filename"),
uploaded_file=data.get("uploaded_file"),
image_hash=data.get("image_md5")
)

Expand All @@ -222,7 +227,7 @@ def from_input(cls, input):
time=input.time,
date_option=input.date_option,
time_option=input.time_option,
uploaded_filename=input.uploaded_filename,
uploaded_file=input.uploaded_file,
image_hash=input.image_hash
)

Expand Down

0 comments on commit d437f49

Please sign in to comment.