lzy/sci-gui-agent-benchmark
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update multi-apps examples

Browse Source
This commit is contained in:
Timothyxxx
2024-03-07 22:15:08 +08:00
parent fd9f6cbc59
commit 1aa2a43908
7 changed files with 445 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files

103
desktop_env/evaluators/metrics/docs.py
View File

@@ -11,9 +11,9 @@ from docx.shared import RGBColor
from odf.opendocument import load
from odf.text import P
from odf.text import Span
from rapidfuzz import fuzz
from skimage.color import deltaE_ciede2000
from skimage.color import rgb2lab
from rapidfuzz import fuzz
logger = logging.getLogger("desktopenv.metric.docs")
@@ -173,9 +173,11 @@ def compare_docx_tables(docx_file1, docx_file2):
return 1
from io import BytesIO
from PIL import Image
def compare_docx_images(docx_file1, docx_file2):
doc1 = Document(docx_file1)
doc2 = Document(docx_file2)
@@ -187,7 +189,7 @@ def compare_docx_images(docx_file1, docx_file2):
img_data = rel.target_part.blob
images.append(BytesIO(img_data))
return images
images1 = extract_images(doc1)
images2 = extract_images(doc2)
if len(images1) != len(images2):
@@ -197,8 +199,10 @@ def compare_docx_images(docx_file1, docx_file2):
return 0
return 1
import pytesseract
def compare_image_text(image_path, rule):
img = Image.open(image_path)
img_text = pytesseract.image_to_string(img)
@@ -207,6 +211,7 @@ def compare_image_text(image_path, rule):
else:
raise ValueError("Unsupported rule type")
def compare_line_spacing(docx_file1, docx_file2):
if not compare_docx_files(docx_file1, docx_file2):
return 0
@@ -313,7 +318,7 @@ def check_tabstops(docx_file1, docx_file2, **kwargs) -> float:
section = doc2.sections[0]
paragraph_width = section.page_width - section.left_margin - section.right_margin
ignore_tabs = lambda x: x.alignment == WD_TAB_ALIGNMENT.CLEAR or (
x.alignment == WD_TAB_ALIGNMENT.LEFT and x.position == 0)
x.alignment == WD_TAB_ALIGNMENT.LEFT and x.position == 0)
minus = .0
for p1, p2 in zip(para1, para2):
# filter CLEAR tabstop and default left-0 tabstop
@@ -583,3 +588,95 @@ def compare_highlighted_text(file1, file2):
return 1
else:
return 0
def compare_references(file1, file2, **options):
reference_indicator = options.get('reference_indicator', 'References')
reference_base_result = options.get('reference_base_result', 0.5)
# Determine file types and load documents
if file1.endswith('.docx') and file2.endswith('.docx'):
doc1 = Document(file1)
doc2 = Document(file2)
doc1_paragraphs = [p.text for p in doc1.paragraphs]
doc2_paragraphs = [p.text for p in doc2.paragraphs]
else:
# Unsupported file types or mismatch
print("Unsupported file types or mismatch between file types.")
return 0
# Find the references section in the paragraphs, find the idx of the last reference_indicator in the paragraph list
ref1_idx = doc1_paragraphs.index(reference_indicator) if reference_indicator in doc1_paragraphs else -1
ref2_idx = doc2_paragraphs.index(reference_indicator) if reference_indicator in doc2_paragraphs else -1
if ref1_idx == -1 and ref2_idx == -1:
return 1
if ref1_idx == -1 or ref2_idx == -1:
return 0
# split the reference section into reference items, and remove the empty string items
ref1 = [p for p in doc1_paragraphs[ref1_idx + 1:] if p.strip()]
ref2 = [p for p in doc2_paragraphs[ref2_idx + 1:] if p.strip()]
# Compare the references
if len(ref1) != len(ref2):
return 0
total_similarity = 0
for r1, r2 in zip(ref1, ref2):
# fuzzy match the references
similarity = fuzz.ratio(r1, r2) / 100.0
total_similarity += similarity
result = total_similarity / len(ref1)
if result >= reference_base_result:
return (result - reference_base_result) / (1 - reference_base_result)
else:
return 0
def compare_answer(file1, file2, **options):
"""This is a specific function to compare the """
# Determine file types and load documents
if file1.endswith('.docx') and file2.endswith('.docx'):
doc1 = Document(file1)
doc2 = Document(file2)
doc1_paragraphs = [p.text for p in doc1.paragraphs]
doc2_paragraphs = [p.text for p in doc2.paragraphs]
else:
# Unsupported file types or mismatch
print("Unsupported file types or mismatch between file types.")
return 0
# Find the references section in the paragraphs, find the idx of the last reference_indicator in the paragraph list
ref1_idx = doc1_paragraphs.index(reference_indicator) if reference_indicator in doc1_paragraphs else -1
ref2_idx = doc2_paragraphs.index(reference_indicator) if reference_indicator in doc2_paragraphs else -1
if ref1_idx == -1 and ref2_idx == -1:
return 1
if ref1_idx == -1 or ref2_idx == -1:
return 0
# split the reference section into reference items, and remove the empty string items
ref1 = [p for p in doc1_paragraphs[ref1_idx + 1:] if p.strip()]
ref2 = [p for p in doc2_paragraphs[ref2_idx + 1:] if p.strip()]
# Compare the references
if len(ref1) != len(ref2):
return 0
total_similarity = 0
for r1, r2 in zip(ref1, ref2):
# fuzzy match the references
similarity = fuzz.ratio(r1, r2) / 100.0
total_similarity += similarity
result = total_similarity / len(ref1)
if result >= reference_base_result:
return (result - reference_base_result) / (1 - reference_base_result)
else:
return 0

6
desktop_env/evaluators/metrics/table.py
View File

@@ -16,9 +16,9 @@ from openpyxl.utils import get_column_letter
from openpyxl.worksheet.datavalidation import DataValidation
from openpyxl.worksheet.worksheet import Worksheet
from .utils import _match_value_to_rule, _read_cell_style, read_cell_value
from .utils import load_charts, load_sparklines, load_rows_or_cols, load_xlsx_styles\
, load_filters, load_pivot_tables
from desktop_env.evaluators.metrics.utils import _match_value_to_rule, _read_cell_style, read_cell_value
from desktop_env.evaluators.metrics.utils import load_charts, load_sparklines, load_rows_or_cols, load_xlsx_styles \
, load_filters, load_pivot_tables
from rapidfuzz import fuzz
# from openpyxl.utils import coordinate_to_tuple

116
evaluation_examples/examples/multi_apps/1f18aa87-af6f-41ef-9853-cdb8f32ebdea.json Normal file
View File

@@ -0,0 +1,116 @@
{
"id": "1f18aa87-af6f-41ef-9853-cdb8f32ebdea",
"snapshot": "libreoffice_calc",
"instruction": "I've prepared some grammar tests and placed them in the 'Grammar test' folder. I've already provided the multiple-choice answers for Test 1 in the 'answer doc' file. Could you please follow the same format to write out the answers for the remaining two tests in the doc file? This way, I can distribute them to the students as a reference. Thank you.",
"source": "authors",
"config": [
{
"type": "command",
"parameters": {
"command": [
"mkdir",
"-p",
"/home/user/Desktop/students work/",
"/home/user/Desktop/Lec powerpoint/",
"/home/user/Desktop/Grammar test/",
"/home/user/Desktop/Grammar rules PDF/",
"/home/user/Desktop/FDI/"
]
}
},
{
"type": "download",
"parameters": {
"files": [
{
"path": "/home/user/Desktop/Grammer test 1.docx",
"url": "https://drive.google.com/uc?id=1VaXQ9XdzMv079xKFs0Y2XrwdmwFHIvBK&export=download"
},
{
"path": "/home/user/Desktop/Grammer test 2.docx",
"url": "https://drive.google.com/uc?id=1k2T88WreTwi-Yyp9mEJnreEQC3DdkJ2x&export=download"
},
{
"path": "/home/user/Desktop/Grammer test 3.docx",
"url": "https://drive.google.com/uc?id=1QgyQWVOcAJuPaSlrywb9nuFiQDySsTb2&export=download"
},
{
"path": "/home/user/Desktop/Answer.docx",
"url": "https://drive.google.com/uc?id=1BC2DuWJuZggmf6fXl6Ys9xQMZzU6a1br&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/irregularrules02.pdf",
"url": "https://drive.google.com/uc?id=1Eln9ehX6y6Df2-S_Hp7Ao1teKRu6I1Tg&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/irregularrules01.pdf",
"url": "https://drive.google.com/uc?id=1krdEEdNWvTwMKZU14QtI_xc2lCFVeVcl&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/fragrules.pdf",
"url": "https://drive.google.com/uc?id=1IXyI2KeiXsuh6XV2LelcmhZ2PDh_dBQf&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/csfsrules.pdf",
"url": "https://drive.google.com/uc?id=1ernwGGrjhYNoHVNAevdb2qNKQ0I5n3RP&export=download"
},
{
"path": "/home/user/Desktop/Public Lecture Teaching Plan.docx",
"url": "https://drive.google.com/uc?id=1ywfVFTEbiSkypZpzLjLmq_ppSbQIC8s8&export=download"
},
{
"path": "/home/user/Desktop/Course Timetable.xlsx",
"url": "https://drive.google.com/uc?id=1NGtahknRq_kXsXlw0tRQ1_CZp9SljoVg&export=download"
}
]
}
}
],
"trajectory": "trajectories/1f18aa87-af6f-41ef-9853-cdb8f32ebdea",
"related_apps": [
"os",
"libreoffice_writer"
],
"evaluator": {
"postconfig": [
{
"type": "activate_window",
"parameters": {
"window_name": "Answer.docx - LibreOffice Writer",
"strict": true
}
},
{
"type": "sleep",
"parameters": {
"seconds": 0.5
}
},
{
"type": "execute",
"parameters": {
"command": [
"python",
"-c",
"import pyautogui; import time; pyautogui.hotkey('ctrl', 's'); time.sleep(0.5); "
]
}
}
],
"func": "compare_docx_files",
"expected": {
"type": "cloud_file",
"path": "https://drive.google.com/uc?id=1TOMGWC3OFuP6yEGQuRJMEFWdg2NcBPSs&export=download",
"dest": "Answer gold.docx"
},
"result": {
"type": "vm_file",
"path": "/home/user/Desktop/Answer.docx",
"dest": "Answer.docx"
},
"options": {
"ignore_case": true,
"ignore_blanks": true
}
}
}

42
evaluation_examples/examples/multi_apps/2c1ebcd7-9c6d-4c9a-afad-900e381ecd5e.json
View File

@@ -1,7 +1,7 @@
{
"id": "2c1ebcd7-9c6d-4c9a-afad-900e381ecd5e",
"snapshot": "libreoffice_calc",
"instruction": "Could you please take a moment to review the 'case study' file located within the 'student work' folder? I'm particularly interested in ensuring that the references section at the end of the document adheres to the APA 7th edition formatting guidelines. If it turns out that the current formatting does not align with APA 7 standards, I would greatly appreciate your assistance in making the necessary adjustments to comply with those guidelines. ",
"instruction": "Could you please take a moment to review the 'case study' file located within the 'student work' folder? I'm particularly interested in ensuring that the references section at the end of the document adheres to the APA 7th edition formatting guidelines. Making the necessary adjustments if it turns out that the current formatting does not align with APA 7 standards or exists some errors.",
"source": "authors",
"config": [
{
@@ -90,13 +90,45 @@
"related_apps": [
],
"evaluator": {
"postconfig": [],
"func": "",
"result": {
},
"postconfig": [
{
"type": "activate_window",
"parameters": {
"window_name": "case study.docx - LibreOffice Writer",
"strict": true
}
},
{
"type": "sleep",
"parameters": {
"seconds": 0.5
}
},
{
"type": "execute",
"parameters": {
"command": [
"python",
"-c",
"import pyautogui; import time; pyautogui.hotkey('ctrl', 's'); time.sleep(0.5); "
]
}
}
],
"func": "compare_references",
"expected": {
"type": "cloud_file",
"path": "https://drive.google.com/uc?id=1325Qfch0JaJ_wJ20ICxMoHeW8KLpK8v0&export=download",
"dest": "case study gold.docx"
},
"result": {
"type": "vm_file",
"path": "/home/user/Desktop/students work/case study.docx",
"dest": "case study.docx"
},
"options": {
"content_only": true,
"reference_base_result": 0.92
}
}
}

189
evaluation_examples/examples/multi_apps/3a93cae4-ad3e-403e-8c12-65303b271818.json
View File

@@ -1,26 +1,169 @@
{
"id": "3a93cae4-ad3e-403e-8c12-65303b271818",
"snapshot": "libreoffice_calc",
"instruction": "Could you please add a two-hour lecture slot to my weekly course timetable, scheduled for every Wednesday at 12 PM? It seems I accidentally omitted that when setting up my schedule. I'd appreciate you taking care of that for me. Thanks!",
"source": "authors",
"config": [
],
"trajectory": "trajectories/3a93cae4-ad3e-403e-8c12-65303b271818",
"related_apps": [
],
"evaluator": {
"postconfig": [],
"func": "",
"result": {
},
"expected": {
},
"options": {
"id": "3a93cae4-ad3e-403e-8c12-65303b271818",
"snapshot": "libreoffice_calc",
"instruction": "Could you please add a two-hour lecture slot to my weekly course timetable, scheduled for every Wednesday at 12 PM? It seems I accidentally omitted that when setting up my schedule. I'd appreciate you taking care of that for me. Thanks!",
"source": "authors",
"config": [
{
"type": "command",
"parameters": {
"command": [
"mkdir",
"-p",
"/home/user/Desktop/students work/",
"/home/user/Desktop/Lec powerpoint/",
"/home/user/Desktop/Grammar test/",
"/home/user/Desktop/Grammar rules PDF/",
"/home/user/Desktop/FDI/"
]
}
},
{
"type": "download",
"parameters": {
"files": [
{
"path": "/home/user/Desktop/students work/Zheng He .docx",
"url": "https://drive.google.com/uc?id=1wI4141LAthnY5m6qcCUaGgDooe4wiTgz&export=download"
},
{
"path": "/home/user/Desktop/students work/cassie.docx",
"url": "https://drive.google.com/uc?id=1cW9TGJy56vossXxDsdnutPyCbR70af7M&export=download"
},
{
"path": "/home/user/Desktop/students work/case study.docx",
"url": "https://drive.google.com/uc?id=11GzpoZvp4qnL2ukXdpbhH-a3zOIHhtDx&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/irregularrules02.pdf",
"url": "https://drive.google.com/uc?id=1Eln9ehX6y6Df2-S_Hp7Ao1teKRu6I1Tg&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/irregularrules01.pdf",
"url": "https://drive.google.com/uc?id=1krdEEdNWvTwMKZU14QtI_xc2lCFVeVcl&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/fragrules.pdf",
"url": "https://drive.google.com/uc?id=1IXyI2KeiXsuh6XV2LelcmhZ2PDh_dBQf&export=download"
},
{
"path": "/home/user/Desktop/Grammar rules PDF/csfsrules.pdf",
"url": "https://drive.google.com/uc?id=1ernwGGrjhYNoHVNAevdb2qNKQ0I5n3RP&export=download"
},
{
"path": "/home/user/Desktop/Public Lecture Teaching Plan.docx",
"url": "https://drive.google.com/uc?id=1ywfVFTEbiSkypZpzLjLmq_ppSbQIC8s8&export=download"
},
{
"path": "/home/user/Desktop/Course Timetable.xlsx",
"url": "https://drive.google.com/uc?id=1DSjRYgofPK2jldKwIsAygz2x8XWlXCK6&export=download"
}
]
}
}
],
"trajectory": "trajectories/3a93cae4-ad3e-403e-8c12-65303b271818",
"related_apps": [
"os",
"libreoffice_calc"
],
"evaluator": {
"postconfig": [
{
"type": "activate_window",
"parameters": {
"window_name": "Course Timetable.xlsx - LibreOffice Calc",
"strict": true
}
}
},
{
"type": "sleep",
"parameters": {
"seconds": 0.5
}
},
{
"type": "execute",
"parameters": {
"command": [
"python",
"-c",
"import pyautogui; import time; pyautogui.hotkey('ctrl', 's'); time.sleep(0.5); "
]
}
}
],
"func": [
"compare_table",
"compare_table",
"compare_table"
],
"result": [
{
"type": "vm_file",
"path": "/home/user/Desktop/Course Timetable.xlsx",
"dest": "Course Timetable.xlsx"
},
{
"type": "vm_file",
"path": "/home/user/Desktop/Course Timetable.xlsx",
"dest": "Course Timetable.xlsx"
},
{
"type": "vm_file",
"path": "/home/user/Desktop/Course Timetable.xlsx",
"dest": "Course Timetable.xlsx"
}
],
"expected": [
{
"type": "cloud_file",
"path": "https://drive.google.com/uc?id=1VMOon8byWuoCW2Uk5etGMJLMzAfwFVyB&export=download",
"dest": "Course Timetable gold.xlsx"
},
{
"type": "cloud_file",
"path": "https://drive.google.com/uc?id=1jAThiIqILZ5t-RFPHVniSvAL8ZJO1H3P&export=download",
"dest": "Course Timetable gold 2.xlsx"
},
{
"type": "cloud_file",
"path": "https://drive.google.com/uc?id=1U0THDtPCgsw-Rb0N9fjF8DeOepPeUajP&export=download",
"dest": "Course Timetable gold 3.xlsx"
}
],
"options": [
{
"rules": [
{
"type": "sheet_data",
"sheet_idx0": "RNSheet1",
"sheet_idx1": "ENSheet1",
"ignore_case": true
}
]
},
{
"rules": [
{
"type": "sheet_data",
"sheet_idx0": "RNSheet1",
"sheet_idx1": "ENSheet1",
"ignore_case": true
}
]
},
{
"rules": [
{
"type": "sheet_data",
"sheet_idx0": "RNSheet1",
"sheet_idx1": "ENSheet1",
"ignore_case": true
}
]
}
],
"conj": "or"
}
}

25
evaluation_examples/examples/multi_apps/767a3271-56db-4745-ac5d-846ef05e6fe5.json
View File

@@ -1,25 +0,0 @@
{
"id": "767a3271-56db-4745-ac5d-846ef05e6fe5",
"snapshot": "libreoffice_calc",
"instruction": "Hey there! I've been swamped with emails lately, and I'm trying to get organized. I'm part of a local community group, and we've been receiving a lot of emails about different events and volunteer opportunities. I need to sort through these emails and pull out specific information to keep track of everything. Could you help me extract details from emails that mention 'volunteer opportunities' and organize them into a spreadsheet? I'm looking for the event name, date, location, and contact person's email. It would be a huge help if we could have this info neatly laid out so we can easily see what's coming up and who to contact. Thanks a bunch!",
"source": "authors",
"config": [
{
}
],
"trajectory": "trajectories/767a3271-56db-4745-ac5d-846ef05e6fe5",
"related_apps": [
"thunderbird",
"libreoffice_calc"
],
"evaluator": {
"postconfig": [],
"func": "",
"result": {
},
"expected": {
},
"options": {
}
}
}

23
evaluation_examples/examples/multi_apps/82e3c869-49f6-4305-a7ce-f3e64a0618e7.json
View File

@@ -52,6 +52,29 @@
"image"
],
"evaluator": {
"postconfig": [
{
"type": "execute",
"parameters": {
"command": [
"rm",
"-rf",
"/home/user/Desktop/presenter"
]
}
},
{
"type": "execute",
"parameters": {
"command": [
"unzip",
"/home/user/Desktop/presenter.zip",
"-d",
"/home/user/Desktop"
]
}
}
],
"func": "compare_image_list",
"result": {
"type": "vm_file",