admin/scribe
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

Auto fixes from PEP8, fixes from flake8.

Browse Source
This commit is contained in:
Marko Henning
2024-05-15 15:18:17 +02:00
parent 9f526a8f3b
commit 4bcd28d0ea
15 changed files with 391 additions and 417 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scraibe/transcript_exporter.py
+66 -64
View File
@@ -1,5 +1,6 @@
import json
import time
from json.decoder import JSONDecodeError
from typing import Union
@@ -8,13 +9,12 @@ from .hallucinations import KNOWN_HALLUCINATIONS
ALPHABET = [*"abcdefghijklmnopqrstuvwxyz"]
class Transcript:
"""
Class for storing transcript data, including speaker information and text segments,
and exporting it to various file formats such as JSON, HTML, and LaTeX.
"""
def __init__(self, transcript: dict) -> None:
"""
Initializes the Transcript object with the given transcript data.
@@ -30,7 +30,7 @@ class Transcript:
self.speakers = self._extract_speakers()
self.segments = self._extract_segments()
self.annotation = {}
def annotate(self, *args, **kwargs) -> dict:
"""
Annotates the transcript to associate specific names with speakers.
@@ -46,36 +46,41 @@ class Transcript:
ValueError: If the number of speaker names does not match the number
of speakers, or if an unknown speaker is found.
"""
annotations = {}
if args and len(args) != len(self.speakers):
raise ValueError("Number of speaker names does not match number of speakers")
raise ValueError(
"Number of speaker names does not match number of speakers")
if args:
for arg, speaker in zip(args, sorted(self.speakers)):
annotations[speaker] = arg
invalid_speakers = set(kwargs.keys()) - set(self.speakers)
if invalid_speakers:
raise ValueError(f"These keys are not speakers: {', '.join(invalid_speakers)}")
raise ValueError(
f"These keys are not speakers: {', '.join(invalid_speakers)}")
annotations.update({key: kwargs[key] for key in self.speakers if key in kwargs})
annotations.update({key: kwargs[key]
for key in self.speakers if key in kwargs})
self.annotation = annotations
return self
def _remove_hallucinations(self) -> None:
"""
Removes all occurances of known hallucinations from all segments of the transcript.
Segments that are identical to empty strings afterwards are removed from the transcript.
"""
segments_to_drop=[]
segments_to_drop = []
for id in self.transcript:
for snippet in KNOWN_HALLUCINATIONS:
self.transcript[id]['text']=self.transcript[id]['text'].replace(snippet,'')
if self.transcript[id]['text'] == '': segments_to_drop.append(id)
self.transcript[id]['text'] = self.transcript[id]['text'].replace(
snippet, '')
if self.transcript[id]['text'] == '':
segments_to_drop.append(id)
for id in segments_to_drop:
del self.transcript[id]
@@ -87,9 +92,9 @@ class Transcript:
Returns:
list: List of unique speaker names in the transcript.
"""
return list(set([self.transcript[id]["speakers"] for id in self.transcript]))
def _extract_segments(self) -> list:
"""
Extracts all the text segments from the transcript.
@@ -109,23 +114,23 @@ class Transcript:
time stamps for each segment.
"""
fstring = ""
for _id in self.transcript:
seq = self.transcript[_id]
if self.annotation:
speaker = self.annotation[seq["speakers"]]
else:
speaker = seq["speakers"]
segm = seq["segments"]
sseg = time.strftime("%H:%M:%S",time.gmtime(segm[0]))
eseg = time.strftime("%H:%M:%S",time.gmtime(segm[1]))
sseg = time.strftime("%H:%M:%S", time.gmtime(segm[0]))
eseg = time.strftime("%H:%M:%S", time.gmtime(segm[1]))
fstring += f"{speaker} ({sseg} ; {eseg}):\t{seq['text']}\n"
return fstring
def __repr__(self) -> str:
"""Return a string representation of the Transcript object.
@@ -133,8 +138,8 @@ class Transcript:
str: A string that provides an informative description of the object.
"""
return f"Transcript(speakers = {self.speakers},"\
f"segments = {self.segments}, annotation = {self.annotation})"
f"segments = {self.segments}, annotation = {self.annotation})"
def get_dict(self) -> dict:
"""
Get transcript as dict
@@ -142,10 +147,10 @@ class Transcript:
:return: transcript as dict
:rtype: dict
"""
return self.transcript
def get_json(self, *args, use_annotation : bool = True, **kwargs) -> str:
def get_json(self, *args, use_annotation: bool = True, **kwargs) -> str:
"""
Get transcript as json string
:return: transcript as json string
@@ -153,14 +158,14 @@ class Transcript:
"""
if "indent" not in kwargs:
kwargs["indent"] = 3
if use_annotation and self.annotation:
for _id in self.transcript:
seq = self.transcript[_id]
seq["speakers"] = self.annotation[seq["speakers"]]
return json.dumps(self.transcript, *args, **kwargs)
def get_html(self) -> str:
"""
Get transcript as html string
@@ -171,9 +176,9 @@ class Transcript:
html = "<p>" + self.__str__().replace("\n", "<br>") + "</p>"
html = "<html><body>" + html + "</body></html>"
html = html.replace("\t", "&nbsp;&nbsp;&nbsp;&nbsp;")
return html
return html
def get_md(self) -> str:
"""Get transcript as Markdown string, using HTML formatting.
@@ -181,7 +186,7 @@ class Transcript:
str: Transcript as a Markdown string.
"""
return self.get_html()
def get_tex(self) -> str:
"""Get transcript as LaTeX string. If no annotations are present, the speakers will
be annotated with the first letters of the alphabet.
@@ -192,43 +197,42 @@ class Transcript:
if not self.annotation:
self.annotate(*ALPHABET[:len(self.speakers)])
fstring ="\\begin{drama}"
fstring = "\\begin{drama}"
for speaker in self.speakers:
fstring += "\n\t\\Character{"+ str(self.annotation[speaker]) + "}" \
"{"+ str(self.annotation[speaker]) + "}"
fstring += "\n\t\\Character{" + str(self.annotation[speaker]) + "}" \
"{" + str(self.annotation[speaker]) + "}"
for id in self.transcript:
seq = self.transcript[id]
speaker = self.annotation[seq["speakers"]]
fstring += f"\n\\{speaker}speaks:\n{seq['text']}"
fstring += "\n\\end{drama}"
return fstring
def to_json(self,path, *args, **kwargs) -> None:
def to_json(self, path, *args, **kwargs) -> None:
"""Save transcript as json file
Args:
path (str): path to save file
"""
with open(path, "w") as f:
json.dump(self.transcript, f, *args, **kwargs)
def to_txt(self, path: str) -> None:
"""Save transcript as a LaTeX file (placeholder function, implementation needed).
Args:
path (str): Path to save the LaTeX file.
"""
with open(path, "w") as f:
f.write(self.__str__())
def to_md(self, path: str) -> None:
"""Get transcript as Markdown string, using HTML formatting.
@@ -236,7 +240,7 @@ class Transcript:
str: Transcript as a Markdown string.
"""
return self.to_html(path)
def to_html(self, path: str) -> None:
"""
Save transcript as html file
@@ -244,10 +248,10 @@ class Transcript:
:param path: path to save file
:type path: str
"""
with open(path, "w") as file:
file.write(self.get_html())
def to_tex(self, path: str) -> None:
"""Save transcript as a LaTeX file (placeholder function, implementation needed).
@@ -255,7 +259,7 @@ class Transcript:
path (str): Path to save the LaTeX file.
"""
pass
def to_pdf(self, path: str) -> None:
"""Save transcript as a PDF file (placeholder function, implementation needed).
@@ -263,7 +267,7 @@ class Transcript:
path (str): Path to save the PDF file.
"""
pass
def save(self, path: str, *args, **kwargs) -> None:
"""Save transcript to file with the given path and file format.
@@ -279,7 +283,7 @@ class Transcript:
Raises:
ValueError: If the file format specified in the path is unknown.
"""
if path.endswith(".json"):
self.to_json(path, *args, **kwargs)
elif path.endswith(".txt"):
@@ -294,7 +298,7 @@ class Transcript:
self.to_pdf(path, *args, **kwargs)
else:
raise ValueError("Unknown file format")
@classmethod
def from_json(cls, json: Union[dict, str]) -> "Transcript":
"""Load transcript from json file
@@ -310,10 +314,8 @@ class Transcript:
else:
try:
transcript = json.loads(json)
except:
except (TypeError, JSONDecodeError):
with open(json, "r") as f:
transcript = json.load(f)
return cls(transcript)
return cls(transcript)