Merge branch 'main' of github.com:ztjhz/DesktopEnv
This commit is contained in:
David Chang
@@ -1,7 +1,7 @@
|
||||
from .chrome import is_expected_tabs, is_expected_bookmarks, compare_pdfs, is_cookie_deleted, is_shortcut_on_desktop
|
||||
from .docs import compare_font_names, compare_subscript_contains, has_page_numbers_in_footers
|
||||
from .docs import compare_font_names, compare_subscript_contains, has_page_numbers_in_footers, compare_docx_lines
|
||||
from .docs import find_default_font, contains_page_break, compare_docx_files, compare_docx_tables, compare_line_spacing, \
|
||||
compare_insert_equation
|
||||
compare_insert_equation, compare_highlighted_text
|
||||
from .docs import is_first_line_centered, check_file_exists, compare_contains_image
|
||||
from .docs import evaluate_colored_words_in_tables, check_highlighted_words, evaluate_strike_through_last_paragraph, \
|
||||
evaluate_conversion, evaluate_spacing, check_italic_font_size_14, evaluate_alignment, get_unique_train_ids, \
|
||||
|
||||
@@ -357,3 +357,31 @@ def check_no_duplicates(initial_file, processed_file):
|
||||
|
||||
# No duplicates found and at least one valid line was processed
|
||||
return 1
|
||||
|
||||
def compare_docx_lines(file1, file2):
|
||||
# Read the text of the document, line by line
|
||||
doc1 = Document(file1)
|
||||
doc1_lines = [p.text.strip() for p in doc1.paragraphs if p.text.strip()]
|
||||
|
||||
doc2 = Document(file2)
|
||||
doc2_lines = [p.text.strip() for p in doc2.paragraphs if p.text.strip()]
|
||||
|
||||
# Convert the list of lines to sets and compare
|
||||
return set(doc1_lines) == set(doc2_lines)
|
||||
|
||||
def compare_highlighted_text(file1, file2):
|
||||
def extract_highlighted_text(doc):
|
||||
highlighted_text = []
|
||||
# Iterate through each run in each paragraph to check for highlight
|
||||
for paragraph in doc.paragraphs:
|
||||
for run in paragraph.runs:
|
||||
if run.font.highlight_color: # Checks if the run is highlighted
|
||||
highlighted_text.append(run.text.strip())
|
||||
return highlighted_text
|
||||
|
||||
# Read the highlighted text from both documents
|
||||
doc1_highlighted = extract_highlighted_text(Document(file1))
|
||||
doc2_highlighted = extract_highlighted_text(Document(file2))
|
||||
|
||||
# Compare the sets of highlighted text to check if they are the same
|
||||
return set(doc1_highlighted) == set(doc2_highlighted)
|
||||
|
||||
@@ -1,6 +1,36 @@
|
||||
import logging
|
||||
import xml.etree.ElementTree as ET
|
||||
from math import sqrt
|
||||
|
||||
from pptx import Presentation
|
||||
from pptx.util import Inches
|
||||
|
||||
logger = logging.getLogger("desktopenv.metric.slides")
|
||||
|
||||
|
||||
def check_presenter_console_disable(config_file_path):
|
||||
try:
|
||||
tree = ET.parse(config_file_path)
|
||||
root = tree.getroot()
|
||||
|
||||
namespaces = {
|
||||
'oor': 'http://openoffice.org/2001/registry'
|
||||
}
|
||||
|
||||
for item in root.findall(
|
||||
".//item[@oor:path='/org.openoffice.Office.Impress/Misc/Start']/prop[@oor:name='EnablePresenterScreen']",
|
||||
namespaces):
|
||||
# Check if the value of the configuration item indicates that the presenter console has been disabled
|
||||
presenter_screen_enabled = item.find('value').text
|
||||
if presenter_screen_enabled.lower() == 'false':
|
||||
return 1.
|
||||
else:
|
||||
return 0.
|
||||
return 0.
|
||||
except Exception as e:
|
||||
logger.error(f"Error: {e}")
|
||||
return 0.
|
||||
|
||||
|
||||
def check_image_stretch_and_center(modified_ppt, original_ppt):
|
||||
# fixme: this func is overfit to this example libreoffice_impress
|
||||
@@ -72,32 +102,60 @@ def compare_pptx_files(file1_path, file2_path, **options):
|
||||
prs1 = Presentation(file1_path)
|
||||
prs2 = Presentation(file2_path)
|
||||
|
||||
examine_number_of_slides = options.get("examine_number_of_slides", True)
|
||||
examine_shape = options.get("examine_shape", True)
|
||||
examine_text = options.get("examine_text", True)
|
||||
examine_font_name = options.get("examine_font_name", True)
|
||||
examine_font_size = options.get("examine_font_size", True)
|
||||
examine_font_bold = options.get("examine_font_bold", True)
|
||||
examine_font_italic = options.get("examine_font_italic", True)
|
||||
examine_color_rgb = options.get("examine_color_rgb", True)
|
||||
examine_font_underline = options.get("examine_font_underline", True)
|
||||
|
||||
# compare the number of slides
|
||||
if len(prs1.slides) != len(prs2.slides):
|
||||
if len(prs1.slides) != len(prs2.slides) and examine_number_of_slides:
|
||||
return False
|
||||
|
||||
# compare the content of each slide
|
||||
for slide1, slide2 in zip(prs1.slides, prs2.slides):
|
||||
# check if the shapes are the same
|
||||
for shape1, shape2 in zip(slide1.shapes, slide2.shapes):
|
||||
if shape1.left != shape2.left or shape1.top != shape2.top or shape1.width != shape2.width or shape1.height != shape2.height:
|
||||
if (
|
||||
shape1.left != shape2.left or shape1.top != shape2.top or shape1.width != shape2.width or shape1.height != shape2.height) and examine_shape:
|
||||
return False
|
||||
|
||||
if hasattr(shape1, "text") and hasattr(shape2, "text"):
|
||||
if shape1.text != shape2.text:
|
||||
if shape1.text != shape2.text and examine_text:
|
||||
return False
|
||||
|
||||
# check if the paragraphs are the same
|
||||
for para1, para2 in zip(shape1.text_frame.paragraphs, shape2.text_frame.paragraphs):
|
||||
# check if the runs are the same
|
||||
for run1, run2 in zip(para1.runs, para2.runs):
|
||||
if run1.text != run2.text:
|
||||
if run1.text != run2.text and examine_text:
|
||||
return False
|
||||
|
||||
# check if the font properties are the same
|
||||
if run1.font.name != run2.font.name or run1.font.size != run2.font.size or run1.font.bold != run2.font.bold or run1.font.italic != run2.font.italic or run1.font.color.rgb != run2.font.color.rgb:
|
||||
if run1.font.name != run2.font.name and examine_font_name:
|
||||
return False
|
||||
|
||||
if run1.font.size != run2.font.size and examine_font_size:
|
||||
return False
|
||||
|
||||
if run1.font.bold != run2.font.bold and examine_font_bold:
|
||||
return False
|
||||
|
||||
if run1.font.italic != run2.font.italic and examine_font_italic:
|
||||
return False
|
||||
|
||||
if run1.font.color.rgb != run2.font.color.rgb and examine_color_rgb:
|
||||
return False
|
||||
|
||||
if run1.font.underline != run2.font.underline and examine_font_underline:
|
||||
return False
|
||||
|
||||
# fixme: Actually there are more properties to be compared, but we cannot get them through pptx
|
||||
|
||||
return True
|
||||
|
||||
|
||||
@@ -126,32 +184,6 @@ def check_strikethrough(pptx_path, rules):
|
||||
return True
|
||||
|
||||
|
||||
def has_audio_on_page(slide):
|
||||
for shape in slide.shapes:
|
||||
if shape.shape_type == 13:
|
||||
return True
|
||||
return False
|
||||
|
||||
|
||||
def check_for_audio(prs):
|
||||
prs = Presentation(prs)
|
||||
for i, slide in enumerate(prs.slides):
|
||||
if has_audio_on_page(slide):
|
||||
return 1
|
||||
return 0
|
||||
|
||||
|
||||
def check_formula_shape(prs):
|
||||
prs = Presentation(prs)
|
||||
slide = prs.slides[13]
|
||||
|
||||
for shape in slide.shapes:
|
||||
|
||||
if shape.has_text_frame and shape.shape_type == 1:
|
||||
return 1
|
||||
return 0
|
||||
|
||||
|
||||
def check_slide_orientation_Portrait(pptx_path):
|
||||
presentation = Presentation(pptx_path)
|
||||
|
||||
@@ -163,20 +195,40 @@ def check_slide_orientation_Portrait(pptx_path):
|
||||
return 0
|
||||
|
||||
|
||||
def contains_mp4_video(pptx_path):
|
||||
prs = Presentation(pptx_path)
|
||||
for slide in prs.slides:
|
||||
for shape in slide.shapes:
|
||||
if shape.shape_type == 16:
|
||||
if shape.media_type == 3:
|
||||
return 1
|
||||
return 0
|
||||
def evaluate_presentation_fill_to_rgb_distance(pptx_file, rules):
|
||||
rgb = rules["rgb"]
|
||||
|
||||
def slide_fill_distance_to_rgb(_slide, _rgb):
|
||||
fill = _slide.background.fill
|
||||
if fill.type == 1:
|
||||
r1, g1, b1 = fill.fore_color.rgb
|
||||
r2, g2, b2 = _rgb
|
||||
return sqrt((r1 - r2) ** 2 + (g1 - g2) ** 2 + (b1 - b2) ** 2) / sqrt(255 ** 2 + 255 ** 2 + 255 ** 2)
|
||||
|
||||
return 1
|
||||
|
||||
prs = Presentation(pptx_file)
|
||||
similarity = 1 - sum(slide_fill_distance_to_rgb(slide, rgb) for slide in prs.slides) / len(prs.slides)
|
||||
return similarity
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
path1 = "../../任务数据/LibreOffice Impress/Change_Color_Slide_Number_gold_textbox.pptx"
|
||||
presentation = Presentation(path1)
|
||||
def check_left_panel(accessibility_tree):
|
||||
namespaces = {
|
||||
'st': 'uri:deskat:state.at-spi.gnome.org',
|
||||
'cp': 'uri:deskat:component.at-spi.gnome.org'
|
||||
}
|
||||
|
||||
for i, sl in enumerate(presentation.slides):
|
||||
for j, sh in enumerate(sl.shapes):
|
||||
print(i, j, sh, sh.name, sh.shape_type, sh.text)
|
||||
root = ET.fromstring(accessibility_tree)
|
||||
|
||||
for root_pane in root.iter('root-pane'):
|
||||
for panel in root_pane.iter('panel'):
|
||||
for split_pane in panel.iter('split-pane'):
|
||||
# Get the left panel
|
||||
if split_pane.attrib.get("{{{}}}parentcoord".format(namespaces['cp'])) == "(0, 0)":
|
||||
# Get the visible attribute
|
||||
visible = split_pane.attrib.get("{{{}}}visible".format(namespaces['st']))
|
||||
if visible:
|
||||
# decide if it is left panel
|
||||
return 1.
|
||||
|
||||
return 0.
|
||||
@@ -10,6 +10,8 @@ import imagehash
|
||||
import librosa
|
||||
import numpy as np
|
||||
from PIL import Image
|
||||
from fastdtw import fastdtw
|
||||
from scipy.spatial.distance import cosine
|
||||
from skimage.metrics import structural_similarity as ssim
|
||||
|
||||
logger = logging.getLogger("desktopenv.metrics.vlc")
|
||||
@@ -111,11 +113,10 @@ def compare_images(image1_path, image2_path):
|
||||
return similarity_index
|
||||
|
||||
|
||||
def compare_audios(audio_path_1, audio_path_2, max_distance=1000):
|
||||
def compare_audios(audio_path_1, audio_path_2):
|
||||
"""
|
||||
Compare two audio files and return a similarity score in the range [0, 1].
|
||||
audio_path_1, audio_path_2: paths to the audio files to compare
|
||||
max_distance: an empirically determined maximum expected DTW distance
|
||||
"""
|
||||
# Example Usage:
|
||||
# similarity = compare_audios_simple('path_to_audio1.mp3', 'path_to_audio2.mp3')
|
||||
@@ -125,21 +126,31 @@ def compare_audios(audio_path_1, audio_path_2, max_distance=1000):
|
||||
if not audio_path_1 or not audio_path_2:
|
||||
return 0
|
||||
|
||||
# Load the audio files and extract MFCC features
|
||||
y1, sr1 = librosa.load(audio_path_1)
|
||||
y2, sr2 = librosa.load(audio_path_2)
|
||||
|
||||
# Extract MFCC features
|
||||
mfcc1 = librosa.feature.mfcc(y=y1, sr=sr1)
|
||||
|
||||
y2, sr2 = librosa.load(audio_path_2)
|
||||
mfcc2 = librosa.feature.mfcc(y=y2, sr=sr2)
|
||||
|
||||
# Compute Dynamic Time Warping distance
|
||||
distance, path = librosa.sequence.dtw(mfcc1.T, mfcc2.T)
|
||||
# Normalize the MFCC features
|
||||
mfcc1 = librosa.util.normalize(mfcc1, axis=1)
|
||||
mfcc2 = librosa.util.normalize(mfcc2, axis=1)
|
||||
|
||||
# Normalize distance to get a similarity score
|
||||
normalized_distance = np.mean(distance) / max_distance
|
||||
similarity_score = 1 - min(normalized_distance, 1) # Ensure the score is within [0, 1]
|
||||
# Define a lambda function to compute cosine distance
|
||||
dist_func = lambda x, y: cosine(x, y)
|
||||
|
||||
return similarity_score
|
||||
# Use the DTW algorithm to find the best alignment path
|
||||
distance, path = fastdtw(mfcc1.T, mfcc2.T, dist=dist_func)
|
||||
|
||||
# Calculate the similarity score, here we use 1/(1+distance) to convert distance to a similarity score
|
||||
similarity = 1 / (1 + distance)
|
||||
|
||||
return similarity
|
||||
|
||||
|
||||
def compare_audios_by_dl_model(audio_path_1, audio_path_2):
|
||||
pass
|
||||
|
||||
|
||||
def compare_videos(video_path1, video_path2, max_frames_to_check=100, threshold=5):
|
||||
|
||||
Reference in New Issue
Block a user