RomanNum3ral
/
Cybersecurity-Projects
mirror of https://github.com/CarterPerez-dev/Cybersecurity-Projects.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

metadata-scrubber-tool: 

added the png_metadata processing functionalities,
added an exception class to handle various error
This commit is contained in:
HERITAGE-XION 2026-01-04 21:13:37 +01:00
parent ceb5ef6d55
commit a517d774c2
5 changed files with 137 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
src/__pycache__/main.cpython-314.pyc
View File

Binary file not shown.

7
src/services/image_handler.py
View File

@ -7,12 +7,17 @@ from PIL import Image
from src.services.metadata_handler import MetadataHandler
from src.utils.jpeg_metadata import JpegProcessaor
from src.utils.png_metadata import PngProcessor
class ImageHandler(MetadataHandler):
def __init__(self, filepath: str):
super().__init__(filepath)
self.processors = {".jpeg": JpegProcessaor(), ".jpg": JpegProcessaor()}
self.processors = {
".jpeg": JpegProcessaor(),
".jpg": JpegProcessaor(),
".png": PngProcessor(),
}
self.tags_to_delete = []
def read(self):

22
src/utils/exceptions.py Normal file
View File

@ -0,0 +1,22 @@
class MetadataException(Exception):
"""Base class for other exceptions in this module."""
pass
class UnsupportedFormatError(MetadataException):
"""Exception raised when an unsupported file format is encountered."""
pass
class MetadataNotFoundError(MetadataException):
"""Exception raised when metadata is not found."""
pass
class MetadataProcessingError(MetadataException):
"""Exception raised when an error occurs during metadata processing."""
pass

89
src/utils/jpeg_metadata.py
View File

@ -1,4 +1,6 @@
import piexif
import piexif # pyright: ignore[reportMissingTypeStubs]
from src.utils.exceptions import MetadataNotFoundError, MetadataProcessingError
class JpegProcessaor:
@ -7,48 +9,55 @@ class JpegProcessaor:
self.data = {}
def get_metadata(self, img):
if "exif" in img.info:
try:
if "exif" in img.info:
exif_dict = piexif.load(img.info["exif"])
for ifd, value in exif_dict.items():
# this exclude thumbnail IFD (its the thumbnails blob data. it can be removed but the image will take a couple more seconds to load)
if not isinstance(exif_dict[ifd], dict):
continue
# iterate through the IFD
for tag, tag_value in exif_dict[ifd].items():
tag_info = piexif.TAGS[ifd].get(tag, {})
# Get the human-readable name for the tag
tag_name = (
tag_info.get("name", "Unknown Tag")
if tag_info
else "Unknown Tag"
)
# exculudes tags that are necessary for image display integrity
if (
tag_name == "Orientation"
or tag_name == "ColorSpace"
or tag_name == "ExifTag"
):
continue
# save to list and dict
self.tags_to_delete.append(tag)
self.data[tag_name] = tag_value
return {"data": self.data, "tags_to_delete": self.tags_to_delete}
else:
raise MetadataNotFoundError("No EXIF data found in the image.")
except MetadataNotFoundError as e:
return f"Error: {str(e)}"
def delete_metadata(self, img, tags_to_delete):
try:
exif_dict = piexif.load(img.info["exif"])
for ifd, value in exif_dict.items():
# this exclude thumbnail IFD (its the thumbnails blob data. it can be removed but the image will take a couple more seconds to load)
# exclude thumbnail IFD (its the thumbnails blob data so i dont wanna deal with that)
if not isinstance(exif_dict[ifd], dict):
continue
# iterate through the IFD
for tag, tag_value in exif_dict[ifd].items():
tag_info = piexif.TAGS[ifd].get(tag, {})
# iterate through and delete tags
for tag in list(exif_dict[ifd]):
if tag in tags_to_delete:
del exif_dict[ifd][tag]
# Get the human-readable name for the tag
tag_name = (
tag_info.get("name", "Unknown Tag")
if tag_info
else "Unknown Tag"
)
# exculudes tags that are necessary for image display integrity
if (
tag_name == "Orientation"
or tag_name == "ColorSpace"
or tag_name == "ExifTag"
):
continue
# save to list and dict
self.tags_to_delete.append(tag)
self.data[tag_name] = tag_value
return {"data": self.data, "tags_to_delete": self.tags_to_delete}
def delete_metadata(self, img, tags_to_delete):
exif_dict = piexif.load(img.info["exif"])
for ifd, value in exif_dict.items():
# exclude thumbnail IFD (its the thumbnails blob data so i dont wanna deal with that)
if not isinstance(exif_dict[ifd], dict):
continue
# iterate through and delete tags
for tag in list(exif_dict[ifd]):
if tag in tags_to_delete:
del exif_dict[ifd][tag]
return exif_dict
return exif_dict
except Exception as e:
raise MetadataProcessingError(f"Error Processing: {str(e)}")

60
src/utils/png_metadata.py Normal file
View File

@ -0,0 +1,60 @@
from PIL import ExifTags
from src.utils.exceptions import MetadataNotFoundError, MetadataProcessingError
class PngProcessor:
def __init__(self):
self.tags_to_delete = []
self.data = {}
def get_metadata(self, img):
img.load()
exif = img.getexif()
try:
if exif:
# iterate through the (0th) IFD
for tag, value in exif.items():
# Get the human-readable name for the tag
tag_name = ExifTags.TAGS.get(tag, tag)
# save to list and dict
self.tags_to_delete.append(tag)
self.data[tag_name] = value
print(f"{tag_name}: {value}")
# terate through the (GPS) IFD
gps_ifd = exif.get_ifd(ExifTags.IFD.GPSInfo)
for tag, value in gps_ifd.items():
# Get the human-readable name for the tag
tag_name = ExifTags.GPSTAGS.get(tag, tag)
# save to list and dict
self.tags_to_delete.append(tag)
self.data[tag_name] = value
print(f"{tag_name}: {value}")
return {"data": self.data, "tags_to_delete": self.tags_to_delete}
else:
raise MetadataNotFoundError("No EXIF data found in the image.")
except MetadataNotFoundError as e:
return f"Error: {str(e)}"
def delete_metadata(self, img, tags_to_delete):
img.load()
exif = img.getexif()
try:
# iterate through the (0th) IFD
for tag_id, value in exif.items():
if tag_id in tags_to_delete:
del exif[tag_id]
# terate through the (GPS) IFD
gps_ifd = exif.get_ifd(ExifTags.IFD.GPSInfo)
for tag_id, value in gps_ifd.items():
if tag_id in tags_to_delete:
del gps_ifd[tag_id]
return exif + gps_ifd
except Exception as e:
raise MetadataProcessingError(f"Error Processing image: {str(e)}")