papis-extract/papis_extract/annotation.py

146 lines
4.7 KiB
Python
Raw Normal View History

2023-08-28 08:28:06 +00:00
import math
2024-01-24 12:39:01 +00:00
from dataclasses import dataclass
from typing import Any, Optional, cast
2023-08-28 08:28:06 +00:00
import chevron
2023-08-28 08:28:06 +00:00
import papis.config
from papis.document import Document
TEXT_SIMILARITY_MINIMUM = 0.75
COLOR_SIMILARITY_MINIMUM = 0.833
2024-01-24 07:55:43 +00:00
COLORS: dict[str, tuple[float, float, float]] = {
2023-08-28 08:28:06 +00:00
"blue": (0, 0, 1),
2024-01-24 07:55:43 +00:00
"green": (0, 1, 0),
"red": (1, 0, 0),
"cyan": (0, 1, 1),
2023-08-28 08:28:06 +00:00
"yellow": (1, 1, 0),
2024-01-24 07:55:43 +00:00
"magenta": (1, 0, 1),
2023-08-28 08:28:06 +00:00
"purple": (0.5, 0, 0.5),
2024-01-24 07:55:43 +00:00
"pink": (1, 0.75, 0.8),
2023-08-28 08:28:06 +00:00
"orange": (1, 0.65, 0),
}
2024-01-24 07:55:43 +00:00
2023-08-28 08:28:06 +00:00
@dataclass
class Annotation:
2023-09-19 15:52:45 +00:00
"""A PDF annotation object.
Contains all information necessary for the annotation itself, content and metadata.
"""
2023-08-28 08:28:06 +00:00
def __init__(
self,
file: str,
color: tuple[float, float, float] = (0.0, 0.0, 0.0),
content: str = "",
note: str = "",
page: int = 0,
tag: str = "",
type: str = "Highlight",
minimum_similarity_color: float = 1.0,
) -> None:
self.minimum_similarity_color = minimum_similarity_color
self.file = file
self._color = color
self.content = content
self.note = note
self.page = page
self.tag = tag or self._tag_from_colorname(self.colorname or "")
self.type = type
def format(self, formatting: str, doc: Document = Document()):
2023-08-28 08:28:06 +00:00
"""Return a formatted string of the annotation.
Given a provided formatting pattern, this method returns the annotation
formatted with the correct marker replacements and removals, ready
for display or writing.
"""
data = {
"file": self.file,
"quote": self.content,
"note": self.note,
"page": self.page,
"tag": self.tag,
"type": self.type,
"doc": doc,
2023-08-28 08:28:06 +00:00
}
return chevron.render(formatting, data)
2023-08-28 08:28:06 +00:00
@property
def color(self):
return self._color
@color.setter
def color(self, value: tuple[float, float, float]):
self._color = value
self.tag = self._tag_from_colorname(self.colorname or "")
2023-08-28 08:28:06 +00:00
@property
def colorname(self):
"""Return the stringified version of the annotation color.
Finds the closest named color to the annotation and returns it,
using euclidian distance between the two color vectors.
"""
annot_colors = self.color or (0.0, 0.0, 0.0)
2023-08-28 08:28:06 +00:00
nearest = None
minimum_similarity = (
2023-08-28 10:55:01 +00:00
papis.config.getfloat("minimum_similarity_color", "plugins.extract") or 1.0
2023-08-28 08:28:06 +00:00
)
minimum_similarity = self.minimum_similarity_color
2023-08-28 08:28:06 +00:00
for name, values in COLORS.items():
similarity_ratio = self._color_similarity_ratio(values, annot_colors)
if similarity_ratio >= minimum_similarity:
2023-08-28 08:28:06 +00:00
minimum_similarity = similarity_ratio
nearest = name
return nearest
def _color_similarity_ratio(
self,
color_one: tuple[float, float, float],
color_two: tuple[float, float, float],
) -> float:
2023-08-28 08:28:06 +00:00
"""Return the similarity of two colors between 0 and 1.
Takes two rgb color tuples made of floats between 0 and 1,
e.g. (1, 0.65, 0) for orange, and returns the similarity
between them, with 1 being the same color and 0 being the
difference between full black and full white, as a float.
"""
return 1 - (abs(math.dist([*color_one], [*color_two])) / 3)
def _tag_from_colorname(self, colorname: str) -> str:
color_mapping: dict[str, str] = self._getdict("tags", "plugins.extract")
if not color_mapping:
return ""
return color_mapping.get(colorname, "")
# mimics the functions in papis.config.{getlist,getint,getfloat} etc.
def _getdict(self, key: str, section: Optional[str] = None) -> dict[str, str]:
"""Dict getter
:returns: A python dict
:raises SyntaxError: Whenever the parsed syntax is either not a valid
python object or a valid python dict.
"""
rawvalue: Any = papis.config.general_get(key, section=section)
if isinstance(rawvalue, dict):
return cast(dict[str, str], rawvalue)
try:
rawvalue = eval(rawvalue)
except Exception:
raise SyntaxError(
"The key '{}' must be a valid Python object: {}".format(key, rawvalue)
)
else:
if not isinstance(rawvalue, dict):
raise SyntaxError(
"The key '{}' must be a valid Python dict. Got: {} (type {!r})".format(
key, rawvalue, type(rawvalue).__name__
)
)
return cast(dict[str, str], rawvalue)