odilf/cse2000-software-project
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

feat: upgrade pre-commit to use single stage

Browse source
This commit is contained in:
Roman Knyazhitskiy 2024-05-18 12:27:18 +00:00
parent f879765ed0
commit 594cdd86d1
23 changed files with 167 additions and 107 deletions
Download patch file Download diff file Expand all files Collapse all files

2
.gitignore vendored
View file

@ -1 +1 @@
.apikeys.env
.apikeys.env

7
.gitlab-ci.yml
View file

@ -12,7 +12,7 @@ stages:
- build
- lint
- test
build:app:
stage: build
image: 'node:22.1.0-slim'
@ -30,7 +30,7 @@ lint:app:
stage: lint
image: 'node:22.1.0-slim'
script:
- *pre_pnpm
- *pre_pnpm
- pnpm install --prefer-offline
- pnpm lint
cache:
@ -45,7 +45,7 @@ lint:app:
test:app:
stage: test
image: 'node:22.1.0-slim'
script:
script:
- *pre_pnpm
- pnpm install --prefer-offline
- pnpm test:unit --coverage
@ -55,4 +55,3 @@ test:app:
- frontend/pnpm-lock.yaml
paths:
- frontend/.pnpm-store

20
.pre-commit-config.yaml
View file

@ -1,4 +1,10 @@
repos:
- repo: https://github.com/pre-commit/pre-commit-hooks
rev: v2.3.0
hooks:
- id: check-yaml
- id: end-of-file-fixer
- id: trailing-whitespace
- repo: https://github.com/compilerla/conventional-pre-commit
rev: v3.2.0
hooks:
@ -10,15 +16,27 @@ repos:
- id: ruff
types_or: [python, pyi, jupyter]
args: [--fix]
stages: [pre-commit]
- id: ruff-format
types_or: [python, pyi, jupyter]
stages: [pre-commit]
- repo: https://github.com/RobertCraigie/pyright-python
rev: v1.1.351
hooks:
- id: pyright
types_or: [python, pyi, jupyter]
additional_dependencies: [numpy, pytest, fastapi, praat-parselmouth, orjson, pydantic, scipy, psycopg, deepgram-sdk, pydub, ffmpeg-python]
stages: [pre-commit]
- repo: https://github.com/crate-ci/typos
rev: v1.21.0
hooks:
- id: typos
stages: [pre-commit]
- repo: local
hooks:
- id: pnpm-lint
name: pnpm lint
language: system
entry: bash -c 'cd frontend; pnpm prettier --write . && pnpm eslint --fix .'
stages: [pre-commit]
types_or: [ts, javascript, svelte]
entry: bash -c 'cd frontend; pnpm prettier --write . && pnpm eslint --fix .'

2
.variables.env
View file

@ -2,4 +2,4 @@ POSTGRES_DB= spectral_db
POSTGRES_USER= user
POSTGRES_PASSWORD= password
POSTGRES_HOST= postgres
POSTGRES_PORT= 5432
POSTGRES_PORT= 5432

2
CONTRIBUTING.md
View file

@ -31,7 +31,7 @@ To use the conventional-pre-commit, which enforces [Conventional Commits](https:
pre-commit install --hook-type commit-msg
```
## Making code changes
## Making code changes
Now make your changes. Make sure to include additional tests if necessary.

1
backend/.gitignore vendored
View file

@ -1,2 +1 @@
**/__pycache__/**

2
backend/README.md
View file

@ -16,4 +16,4 @@ To be able to use the deepgram model for the endpoint api/transcription/{model}/
```
DG_KEY = <your_api_key>
```
```

2
backend/pyproject.toml
View file

@ -3,7 +3,7 @@ name = "spectral"
version = "0.0.1"
description = "Backend for spectral: atypical speech analysis and recognition platform"
authors = [
"Roman Knyazhitskiy <mail@knyaz.tech>",
"Roman Knyazhitskiy <mail@knyaz.tech>",
"Ody Machairas",
"Quinton Den Haan",
"Thijs Houben",

12
backend/spectral/data_objects.py
View file

@ -1,21 +1,24 @@
from pydantic import BaseModel
from typing import Optional
class Frame(BaseModel):
"""
Frame model representing a frame of data with its sampling frequency.
Attributes:
data (list): The data contained in the frame.
fs (float): The sampling frequency of the data.
"""
data: list
fs: float
class Signal(BaseModel):
"""
Signal model representing a signal which contains both various attributes related to its
sampling frequency and values, and paramaters for calculating the pitches, spectogram and formants
Signal model representing a signal which contains both various attributes related to its
sampling frequency and values, and parameters for calculating the pitches, spectogram and formants
Attributes:
data (list): The data contained in the signal.
@ -27,6 +30,7 @@ class Signal(BaseModel):
formants_time_step (Optional[float]): The time step for formants analysis. Defaults to None.
formants_window_length (float): The window length for formants analysis. Defaults to 0.025 seconds.
"""
data: list
fs: float
pitch_time_step: Optional[float] = None
@ -34,4 +38,4 @@ class Signal(BaseModel):
spectogram_window_length: float = 0.005
spectogram_frequency_step: float = 20.0
formants_time_step: Optional[float] = None
formants_window_length: float = 0.025
formants_window_length: float = 0.025

83
backend/spectral/database.py
View file

@ -1,6 +1,7 @@
import psycopg
import psycopg
import uuid
class Database:
"""
Database class for interacting with a PostgreSQL database.
@ -10,13 +11,14 @@ class Database:
Attributes:
conn (psycopg.Connection): The connection object to the database.
cursor (psycopg.Cursor): The cursor object to execute database queries.
Methods:
fetch_file(id: int) -> dict:
Fetches a file record from the database by its ID.
close():
Closes the database connection and cursor.
"""
def __init__(self, user, password, host, port, dbname):
"""
Initializes the Database object and opens a connection to the specified PostgreSQL database.
@ -29,16 +31,12 @@ class Database:
dbname (str): The name of the database.
"""
self.conn = psycopg.connect(
dbname=dbname,
user=user,
password=password,
host=host,
port=port
dbname=dbname, user=user, password=password, host=host, port=port
)
print("database connection opened")
self.cursor = self.conn.cursor()
def fetch_file(self,id):
def fetch_file(self, id):
"""
Fetches a file record from the database by its ID.
@ -48,10 +46,19 @@ class Database:
Returns:
dict: A dictionary containing the file record's details.
"""
self.cursor.execute("SELECT * FROM files WHERE id = %s",[id])
res = self.cursor.fetchone()
return {"id":res[0],"name":res[1],"data":res[2],"creationTime":res[3],"modifiedTime":res[4],"uploader":res[5],"session":res[6],"emphemeral":res[7]}
self.cursor.execute("SELECT * FROM files WHERE id = %s", [id])
res = self.cursor.fetchone() # type: ignore
return {
"id": res[0], # type: ignore
"name": res[1], # type: ignore
"data": res[2], # type: ignore
"creationTime": res[3], # type: ignore
"modifiedTime": res[4], # type: ignore
"uploader": res[5], # type: ignore
"session": res[6], # type: ignore
"emphemeral": res[7], # type: ignore
}
def store_transcription(self, file_id, file_transcription):
"""
Stores a transcription record in the database.
@ -61,16 +68,28 @@ class Database:
file_transcription (list): A list of transcription entries to store, each containing "start", "end", and "value" keys.
"""
file_transcription_id = str(uuid.uuid4())
self.cursor.execute("""
self.cursor.execute(
"""
INSERT INTO file_transcription (id, file)
VALUES (%s, %s);
""",[file_transcription_id,file_id])
""",
[file_transcription_id, file_id],
)
for transcription in file_transcription:
self.cursor.execute("""
self.cursor.execute(
"""
INSERT INTO transcription (id, file_transcription, start, "end", value)
VALUES (%s, %s, %s, %s, %s);
""",[str(uuid.uuid4()),file_transcription_id,transcription["start"],transcription["end"],transcription["value"]])
""",
[
str(uuid.uuid4()),
file_transcription_id,
transcription["start"],
transcription["end"],
transcription["value"],
],
)
def get_transcriptions(self, file_id):
"""
Fetches transcriptions associated with a file from the database.
@ -81,24 +100,36 @@ class Database:
Returns:
list: A list of lists containing transcription entries, where each inner list represents a file transcription and contains dictionaries with "start", "end", and "value" keys.
"""
self.cursor.execute("""
self.cursor.execute(
"""
SELECT id FROM file_transcription
WHERE file = %s
""",[file_id])
WHERE file = %s
""",
[file_id],
)
file_transcriptions = self.cursor.fetchall()
res = []
for file_transcription in file_transcriptions:
self.cursor.execute("""
self.cursor.execute(
"""
SELECT start, "end", value FROM transcription
WHERE file_transcription = %s
""",[file_transcription[0]])
WHERE file_transcription = %s
""",
[file_transcription[0]],
)
transcriptions = self.cursor.fetchall()
parsed_file_transcriptions = []
for transcription in transcriptions:
parsed_file_transcriptions.append({"start":transcription[0],"end":transcription[1],"value":transcription[2]})
parsed_file_transcriptions.append(
{
"start": transcription[0],
"end": transcription[1],
"value": transcription[2],
}
)
res.append(parsed_file_transcriptions)
return res
def close(self):
"""
Closes the database connection and cursor.

11
backend/spectral/frame_analysis.py
View file

@ -1,6 +1,7 @@
import parselmouth
import numpy as np
def simple_frame_info(frame, fs, frame_info):
"""
Extracts and returns basic information from a given audio frame.
@ -30,15 +31,15 @@ def simple_frame_info(frame, fs, frame_info):
"""
if frame_info is None:
return None
data = frame[frame_info["startIndex"]:frame_info["endIndex"]]
data = frame[frame_info["startIndex"] : frame_info["endIndex"]]
res = {}
res["duration"] = calculate_frame_duration(data,fs)
res["pitch"] = calculate_frame_pitch(data,fs)
formants = calculate_frame_f1_f2(data,fs)
res["duration"] = calculate_frame_duration(data, fs)
res["pitch"] = calculate_frame_pitch(data, fs)
formants = calculate_frame_f1_f2(data, fs)
res["f1"] = formants[0]
res["f2"] = formants[1]
return res
def calculate_frame_duration(frame, fs):
"""

32
backend/spectral/main.py
View file

@ -14,18 +14,18 @@ from .frame_analysis import (
calculate_frame_f1_f2,
)
from .mode_handler import (
simple_info_mode,
spectogram_mode,
simple_info_mode,
spectogram_mode,
vowel_space_mode,
transcription_mode
)
from .transcription import (
get_transcription
transcription_mode,
)
from .transcription import get_transcription
from .data_objects import Frame, Signal
from .database import Database
import orjson
import io
import os
from pydub import AudioSegment
@ -154,10 +154,8 @@ async def analyze_signal_mode(
try:
file = database.fetch_file(id)
except Exception as _:
raise HTTPException(
status_code=404, detail="File not found"
)
raise HTTPException(status_code=404, detail="File not found")
audio = AudioSegment.from_file(io.BytesIO(file["data"]))
fs = audio.frame_rate
data = audio.get_array_of_samples()
@ -176,6 +174,7 @@ async def analyze_signal_mode(
case _:
raise HTTPException(status_code=400, detail="Mode not found")
@app.get("/transcription/{model}/{id}")
async def transcribe_file(
model: Annotated[str, Path(title="The transcription model")],
@ -193,7 +192,7 @@ async def transcribe_file(
- id (str): The ID of the file to transcribe.
Returns:
- list: A list of dictionaires with keys 'start', 'end' and 'value' containing the transcription of the audio file.
- list: A list of dictionaries with keys 'start', 'end' and 'value' containing the transcription of the audio file.
Raises:
- HTTPException: If the file is not found or an error occurs during transcription or storing the transcription.
@ -201,15 +200,14 @@ async def transcribe_file(
try:
file = database.fetch_file(id)
except Exception as _:
raise HTTPException(
status_code=404, detail="File not found"
)
transcription = get_transcription(model,file)
raise HTTPException(status_code=404, detail="File not found")
transcription = get_transcription(model, file)
try:
database.store_transcription(id,transcription)
database.store_transcription(id, transcription)
except Exception as _:
raise HTTPException(
status_code=500, detail="Something went wrong while storing the transcription"
status_code=500,
detail="Something went wrong while storing the transcription",
)
return transcription

40
backend/spectral/mode_handler.py
View file

@ -1,12 +1,10 @@
from fastapi import HTTPException
from .signal_analysis import simple_signal_info
from .frame_analysis import (
simple_frame_info,
calculate_frame_f1_f2
)
from .frame_analysis import simple_frame_info, calculate_frame_f1_f2
def simple_info_mode(data,fs,file,frame_index):
def simple_info_mode(data, fs, file, frame_index):
"""
Extracts and returns basic information about a signal and its corresponding frame.
@ -28,22 +26,22 @@ def simple_info_mode(data,fs,file,frame_index):
result = simple_info_mode(data, fs, file, frame_index)
```
"""
result = simple_signal_info(data,fs)
result = simple_signal_info(data, fs)
result["fileSize"] = len(file["data"])
result["fileCreationDate"] = file["creationTime"]
result["frame"] = simple_frame_info(data,fs,frame_index)
result["frame"] = simple_frame_info(data, fs, frame_index)
return result
def spectogram_mode(data,fs,frame_index):
""" TBD
def spectogram_mode(data, fs, frame_index):
"""TBD
Raises:
HTTPException: 501 not implemented
"""
raise HTTPException(
status_code=501, detail="spectogram_mode is not implemented"
)
raise HTTPException(status_code=501, detail="spectogram_mode is not implemented")
def vowel_space_mode(data,fs,frame_index):
def vowel_space_mode(data, fs, frame_index):
"""
Extracts and returns the first and second formants of a specified frame.
@ -71,11 +69,12 @@ def vowel_space_mode(data,fs,frame_index):
raise HTTPException(
status_code=400, detail="Vowel-space mode was not given frame"
)
frame_data = data[frame_index["startIndex"]:frame_index["endIndex"]]
formants = calculate_frame_f1_f2(frame_data,fs)
return {"f1":formants[0],"f2":formants[1]}
frame_data = data[frame_index["startIndex"] : frame_index["endIndex"]]
formants = calculate_frame_f1_f2(frame_data, fs)
return {"f1": formants[0], "f2": formants[1]}
def transcription_mode(id,database):
def transcription_mode(id, database):
"""
Retrieve transcriptions of a file from the database.
@ -95,7 +94,6 @@ def transcription_mode(id,database):
return database.get_transcriptions(id)
except Exception as _:
raise HTTPException(
status_code=500, detail="Something went wrong when retrieving the transcriptions of this file"
status_code=500,
detail="Something went wrong when retrieving the transcriptions of this file",
)

12
backend/spectral/signal_analysis.py
View file

@ -1,6 +1,7 @@
import parselmouth
import numpy as np
def simple_signal_info(signal, fs):
"""
Extracts and returns basic information from a given audio signal.
@ -19,9 +20,12 @@ def simple_signal_info(signal, fs):
result = simple_signal_info(signal, fs)
```
"""
duration = calculate_signal_duration(signal=signal,fs=fs)
avg_pitch = np.mean(calculate_sound_pitch(signal_to_sound(signal=signal,fs=fs))["data"]).item()
return {"duration":duration,"averagePitch":avg_pitch}
duration = calculate_signal_duration(signal=signal, fs=fs)
avg_pitch = np.mean(
calculate_sound_pitch(signal_to_sound(signal=signal, fs=fs))["data"] # type: ignore
).item()
return {"duration": duration, "averagePitch": avg_pitch}
def signal_features(signal, fs):
"""
@ -52,7 +56,7 @@ def signal_features(signal, fs):
"spectogram": spectrogram,
"formants": formants,
}
def signal_to_sound(signal, fs):
"""

22
backend/spectral/transcription.py
View file

@ -2,6 +2,7 @@ from deepgram import DeepgramClient, PrerecordedOptions, FileSource
from fastapi import HTTPException
import os
def get_transcription(model, file):
"""
Get transcription of an audio file using the specified model.
@ -22,10 +23,8 @@ def get_transcription(model, file):
case "deepgram":
return deepgram_transcription(file["data"])
case _:
raise HTTPException(
status_code=404, detail="Model was not found"
)
raise HTTPException(status_code=404, detail="Model was not found")
def deepgram_transcription(data):
"""
@ -44,7 +43,12 @@ def deepgram_transcription(data):
"""
try:
# STEP 1: Create a Deepgram client using the API key
deepgram = DeepgramClient(os.getenv("DG_KEY"))
key = os.getenv("DG_KEY")
deepgram = None
if key is None:
raise Exception("No API key for Deepgram is found")
else:
deepgram = DeepgramClient(key)
payload: FileSource = {
"buffer": data,
@ -59,11 +63,13 @@ def deepgram_transcription(data):
# STEP 3: Call the transcribe_file method with the text payload and options
response = deepgram.listen.prerecorded.v("1").transcribe_file(payload, options)
res = []
for word in response["results"]["channels"][0]["alternatives"][0]["words"]:
res.append({"value":word["word"],"start":word["start"],"end":word["end"]})
res.append(
{"value": word["word"], "start": word["start"], "end": word["end"]}
)
return res
except Exception as e:
print(f"Exception: {e}")
print(f"Exception: {e}")

2
backend/tests/data/frames.json
View file

File diff suppressed because one or more lines are too long

7
backend/tests/test_fast_api.py
View file

@ -8,11 +8,14 @@ import os
client = TestClient(app)
# Load the JSON file
with open(os.path.join(os.path.realpath(__file__),"../data/frames.json"), "r") as file:
with open(os.path.join(os.path.realpath(__file__), "../data/frames.json"), "r") as file:
frame_data = json.load(file)
typical_1_fs, typical_1_data = wv.read(
os.path.join(os.path.realpath(__file__),"../data/torgo-dataset/MC02_control_head_sentence1.wav")
os.path.join(
os.path.realpath(__file__),
"../data/torgo-dataset/MC02_control_head_sentence1.wav",
)
)
typical_1_data = typical_1_data.tolist()

2
backend/tests/test_frame_analysis.py
View file

@ -9,7 +9,7 @@ import json
import os
# Load the JSON file
with open(os.path.join(os.path.realpath(__file__),"../data/frames.json"), "r") as file:
with open(os.path.join(os.path.realpath(__file__), "../data/frames.json"), "r") as file:
frame_data = json.load(file)

3
docker-compose.yaml
View file

@ -47,7 +47,7 @@ services:
env_file:
- '.variables.env'
- '.apikeys.env'
volumes:
postgres_data:
@ -55,4 +55,3 @@ volumes:
networks:
default:
external: false

2
docs/CODE_OF_CONDUCT.md
View file

@ -146,4 +146,4 @@ And, of course, we care about the project and building a good product. We are mo
4. Hard-working
5. Willingness to help others when in need
6. Good attitude toward project/teammates
7. Willingness to propose novel, disrupting solutions
7. Willingness to propose novel, disrupting solutions

4
docs/issues-and-mrs.md
View file

@ -3,7 +3,7 @@
- **Type**: Choose Incident if it’s a bug, otherwise just a normal issue
- **Description**: Follow a template
- **Assignees**: assign yourself. if there is a person who is an expert in the topic, and will help you, you should also assign them
- **Reviewers**: Assign two people. If there is a particular person that should review, assign them. Otherwise assign people pseudo-arbitrarily, while trying to keep it evenly distributed.
- **Reviewers**: Assign two people. If there is a particular person that should review, assign them. Otherwise assign people pseudo-arbitrarily, while trying to keep it evenly distributed.
- **Milestone**: one of MVP (0.1), 1.0
- **Iteration**: Corresponds to each weekly sprint (from 1 to 10, deadlines are on Mondays)
- **Labels**: any number of
@ -43,4 +43,4 @@
- Try to merge MRs in 3 days or less. The limit is 5 days to merge an MR, otherwise we will just close it, as we don't want to clutter the repo with stale MRs (barring exceptional cases, that should be justified). To facilitate this...
- Make MRs as small as possible and generally self-contained.
- Squash commits if history is messy/too detailed.
- Squash commits if history is messy/too detailed.

2
frontend/src/lib/analysis/AnalysisPane.svelte
View file

@ -3,7 +3,7 @@
A pane of analysis. TODO: Expand docs
The analysis pane is responsible for fetching the data that the mode components need.
The analysis pane is responsible for fetching the data that the mode components need.
-->
<script lang="ts">

2
frontend/src/routes/session/[sessionId]/FileExplorer.svelte
View file

@ -1,7 +1,7 @@
<!--
@component
File explorer for the audio files.
File explorer for the audio files.
-->
<script lang="ts">