Spaces:
Running
Running
File size: 10,748 Bytes
cb5b71d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 |
import enum
from typing import Any
import streamlit as st
from core.state import Field
from core.state import RecordSet
from events.fields import ExtractType
from events.fields import FieldEvent
from events.fields import handle_field_change
from events.fields import TransformType
import mlcroissant as mlc
from utils import needed_field
class SourceType:
"""The type of the source (distribution or field)."""
DISTRIBUTION = "distribution"
FIELD = "field"
EXTRACT_TYPES = [
ExtractType.COLUMN,
ExtractType.JSON_PATH,
ExtractType.FILE_CONTENT,
ExtractType.FILE_NAME,
ExtractType.FILE_PATH,
ExtractType.FILE_FULLPATH,
ExtractType.FILE_LINES,
ExtractType.FILE_LINE_NUMBERS,
]
TRANSFORM_TYPES = [
TransformType.FORMAT,
TransformType.JSON_PATH,
TransformType.REGEX,
TransformType.REPLACE,
TransformType.SEPARATOR,
]
def _get_extract(source: mlc.Source) -> str | None:
if source.extract.column:
return ExtractType.COLUMN
elif source.extract.file_property:
file_property = source.extract.file_property
if file_property == mlc.FileProperty.content:
return ExtractType.FILE_CONTENT
elif file_property == mlc.FileProperty.filename:
return ExtractType.FILE_NAME
elif file_property == mlc.FileProperty.filepath:
return ExtractType.FILE_PATH
elif file_property == mlc.FileProperty.fullpath:
return ExtractType.FILE_FULLPATH
elif file_property == mlc.FileProperty.lines:
return ExtractType.FILE_LINES
elif file_property == mlc.FileProperty.lineNumbers:
return ExtractType.FILE_LINE_NUMBERS
else:
return None
elif source.extract.json_path:
return ExtractType.JSON_PATH
return None
def _get_extract_index(source: mlc.Source) -> int | None:
extract = _get_extract(source)
if extract in EXTRACT_TYPES:
return EXTRACT_TYPES.index(extract)
return None
def _get_transforms(source: mlc.Source) -> list[str]:
transforms = source.transforms
return [_get_transform(transform) for transform in transforms]
def _get_transform(transform: mlc.Transform) -> str | None:
if transform.format:
return TransformType.FORMAT
elif transform.json_path:
return TransformType.JSON_PATH
elif transform.regex:
return TransformType.REGEX
elif transform.replace:
return TransformType.REPLACE
elif transform.separator:
return TransformType.SEPARATOR
return None
def _get_transforms_indices(source: mlc.Source) -> list[int]:
transforms = _get_transforms(source)
return [
TRANSFORM_TYPES.index(transform) if transform in TRANSFORM_TYPES else None
for transform in transforms
]
def _handle_remove_reference(field):
"""Removes the reference from a field."""
field.references = mlc.Source()
def render_source(
record_set_key: int,
record_set: RecordSet,
field: Field,
field_key: int,
possible_sources: list[str],
):
"""Renders the form for the source."""
source = field.source
prefix = f"source-{record_set.name}-{field.name}"
col1, col2, col3 = st.columns([1, 1, 1])
index = (
possible_sources.index(source.uid) if source.uid in possible_sources else None
)
key = f"{prefix}-source"
col1.selectbox(
needed_field("Source"),
index=index,
options=[s for s in possible_sources if not s.startswith(record_set.name)],
key=key,
on_change=handle_field_change,
args=(FieldEvent.SOURCE, field, key),
)
if source.node_type == "distribution":
extract = col2.selectbox(
needed_field("Extract"),
index=_get_extract_index(source),
key=f"{prefix}-extract",
options=EXTRACT_TYPES,
on_change=handle_field_change,
args=(FieldEvent.SOURCE_EXTRACT, field, key),
)
if extract == ExtractType.COLUMN:
key = f"{prefix}-columnname"
col3.text_input(
needed_field("Column name"),
value=source.extract.column,
key=key,
on_change=handle_field_change,
args=(FieldEvent.SOURCE_EXTRACT_COLUMN, field, key),
)
if extract == ExtractType.JSON_PATH:
key = f"{prefix}-jsonpath"
col3.text_input(
needed_field("JSON path"),
value=source.extract.json_path,
key=key,
on_change=handle_field_change,
args=(FieldEvent.SOURCE_EXTRACT_JSON_PATH, field, key),
)
# Transforms
indices = _get_transforms_indices(field.source)
if source.transforms:
for number, (index, transform) in enumerate(zip(indices, source.transforms)):
_, col2, col3, col4 = st.columns([4.5, 4, 4, 1])
key = f"{prefix}-{number}-transform"
selected = col2.selectbox(
"Transform",
index=index,
key=key,
options=TRANSFORM_TYPES,
on_change=handle_field_change,
args=(FieldEvent.TRANSFORM, field, key),
kwargs={"number": number},
)
if selected == TransformType.FORMAT:
key = f"{prefix}-{number}-transform-format"
col3.text_input(
needed_field("Format"),
value=transform.format,
key=key,
on_change=handle_field_change,
args=(selected, field, key),
kwargs={"number": number, "type": "format"},
)
elif selected == TransformType.JSON_PATH:
key = f"{prefix}-{number}-jsonpath"
col3.text_input(
needed_field("JSON path"),
value=transform.json_path,
key=key,
on_change=handle_field_change,
args=(selected, field, key),
kwargs={"number": number, "type": "format"},
)
elif selected == TransformType.REGEX:
key = f"{prefix}-{number}-regex"
col3.text_input(
needed_field("Regular expression"),
value=transform.regex,
key=key,
on_change=handle_field_change,
args=(selected, field, key),
kwargs={"number": number, "type": "format"},
)
elif selected == TransformType.REPLACE:
key = f"{prefix}-{number}-replace"
col3.text_input(
needed_field("Replace pattern"),
value=transform.replace,
key=key,
on_change=handle_field_change,
args=(selected, field, key),
kwargs={"number": number, "type": "format"},
)
elif selected == TransformType.SEPARATOR:
key = f"{prefix}-{number}-separator"
col3.text_input(
needed_field("Separator"),
value=transform.separator,
key=key,
on_change=handle_field_change,
args=(selected, field, key),
kwargs={"number": number, "type": "format"},
)
def _handle_remove_transform(field, number):
del field.source.transforms[number]
col4.button(
"✖️",
key=f"{prefix}-{number}-remove-transform",
on_click=_handle_remove_transform,
args=(field, number),
)
def _handle_add_transform(field):
if not field.source:
field.source = mlc.Source(transforms=[])
field.source.transforms.append(mlc.Transform())
col1, _, _ = st.columns([1, 1, 1])
col1.button(
"Add transform on data",
key=f"{prefix}-close-fields",
on_click=_handle_add_transform,
args=(field,),
)
def render_references(
record_set_key: int,
record_set: RecordSet,
field: Field,
field_key: int,
possible_sources: list[str],
):
"""Renders the form for references."""
key = f"references-{record_set.name}-{field.name}"
button_key = f"{key}-add-reference"
has_clicked_button = st.session_state.get(button_key)
references = field.references
if references or has_clicked_button:
col1, col2, col3, col4 = st.columns([4.5, 4, 4, 1])
index = (
possible_sources.index(references.uid)
if references.uid in possible_sources
else None
)
key = f"{key}-reference"
col1.selectbox(
"Reference",
index=index,
options=[s for s in possible_sources if not s.startswith(record_set.name)],
key=key,
on_change=handle_field_change,
args=(FieldEvent.REFERENCE, field, key),
)
if references.node_type == "distribution":
key = f"{key}-extract-references"
extract = col2.selectbox(
needed_field("Extract the reference"),
index=_get_extract_index(references),
key=key,
options=EXTRACT_TYPES,
on_change=handle_field_change,
args=(FieldEvent.REFERENCE_EXTRACT, field, key),
)
if extract == ExtractType.COLUMN:
key = f"{key}-columnname"
col3.text_input(
needed_field("Column name"),
value=references.extract.column,
key=key,
on_change=handle_field_change,
args=(FieldEvent.REFERENCE_EXTRACT_COLUMN, field, key),
)
if extract == ExtractType.JSON_PATH:
key = f"{key}-jsonpath"
col3.text_input(
needed_field("JSON path"),
value=references.extract.json_path,
key=key,
on_change=handle_field_change,
args=(FieldEvent.REFERENCE_EXTRACT_JSON_PATH, field, key),
)
col4.button(
"✖️",
key=f"{key}-remove-reference",
on_click=_handle_remove_reference,
args=(field,),
)
elif not has_clicked_button:
st.button(
"Add a join with another column/field",
key=button_key,
)
|